Valorization of aquaculture waste in removal of cadmium from aqueous solution: optimization by kinetics and ANN analysis

Valorization of aquaculture waste in removal of cadmium from aqueous solution: optimization by... Cadmium is one of the most hazardous heavy metal concerning human health and aquatic pollution. The removal of cadmium through biosorption is a feasible option for restoration of the ecosystem health of the contaminated freshwater ecosystems. In compliance with this proposition and considering the efficiency of calcium carbonate as biosorbent, the shell dust of the economically important snail Bellamya bengalensis was tested for the removal of cadmium from aqueous medium. Follow- ing use of the flesh as a cheap source of protein, the shells of B. bengalensis made up of CaC O are discarded as aquaculture waste. The biosorption was assessed through batch sorption studies along with studies to characterize the morphology and surface structures of waste shell dust. The data on the biosorption were subjected to the artificial neural network (ANN) model for optimization of the process. The biosorption process changed as functions of pH of the solution, concentration of heavy metal, biomass of the adsorbent and time of exposure. The kinetic process was well represented by pseudo second 2 2 −1 order (R = 0.998), and Langmuir equilibrium (R = 0.995) had better fits in the equilibrium process with 30.33 mg g of maximum sorption capacity. The regression equation (R = 0.948) in the ANN model supports predicted values of Cd removal 2+ satisfactorily. The normalized importance analysis in ANN predicts Cd concentration, and pH has the most influence in removal than biomass dose and time. The SEM and EDX studies show clear peaks for Cd confirming the biosorption process while the FTIR study depicts the main functional groups (–OH, C–H, C=O, C=C) responsible for the biosorption process. The study indicated that the waste shell dust can be used as an ec ffi ient, low cost, environment friendly, sustainable adsorbent for the removal of cadmium from aqueous solution. Keywords Cadmium · Shell · Valorization · Kinetics · ANN Abbreviations Introduction AAS Atomic absorption spectroscopy ANN Artificial neural network The heavy metals in aquatic ecosystem can be detrimental SEM Scanning electron microscope to different organisms depending on the tolerance levels and BSD Bellamya shell dust the complexity of the food web. In course of the different FT-IR Fourier transformed infrared industrial and agricultural processes, the heavy metals are EDX Energy dispersive X-ray released and eventually contaminate the aquatic ecosystems. BCF Bio-concentr ation factor For instance, cadmium, considered as one of the important heavy metal pollutant, is added to the aquatic ecosystems through activities in mining, electroplating, battery, paint and ceramic industries, in addition to the natural deposits. The disposal of heavy metals in aquatic systems concen- * Asif Hossain trates as a cascading effect in the trophic levels, frequently asifhossain.bu@gmail.com recognized in the form of bioaccumulation and biomagni- Gautam Aditya fications (Brooks et al. 2004; Tao et al. 2012). Subsequent Gautamaditya2001@gmail.com entry of heavy metals in the living system can create a chain Department of Zoology, University of Calcutta, 35 of physiological, biochemical and genetic changes that are Ballygunge Circular Road, Kolkata 700 019, India concern from public health viewpoint. In case of cadmium, Department of Zoology, Sidho Kanho Birsha University, entry in the human body can damage the kidney, substitute Ranchi Road, Purulia 723104, India Vol.:(0123456789) 1 3 68 Page 2 of 14 Applied Water Science (2018) 8:68 calcium in the bones, can cause liver damage, cancer and (Mollusca: Gastropoda: Viviparidae). A common inhabitant hypertension (Godt et al. 2006; Bernard 2008). Accumula- of ponds, ditches, rivers and varied types of freshwater wet- tion of cadmium in the food web causes damage to the wild lands, the operculate snail B. bengalensis a prolific breeder life and their diversity (Johri et al. 2010; Bernard 2008). As and can easily be cultured (Khan and Choudhury 1984). a consequence, the removal of cadmium and similar heavy The flesh of the snail is used as a cheap source of protein metals are given priority for the sustenance of the ecological and the shells are discarded. The discarded shells that are functions of the aquatic ecosystem. aquacultural waste may be used as cheap source of calcium The heavy metal pollutants are non-biodegradable and carbonate and the Bellamya shell dust (BSD) can be used in thus adsorption is considered as one of the convenient way the removal of cadmium from aquatic system. of removal from the aquatic ecosystems. In recent past, To promote the biosorption as an effective process, the several methods based on the principle of ion exchange, solution for the optimization is required since the process is chemical precipitation, coagulation, activated charcoal, influenced by several factors (Witek-Krowiak et al. 2014). electrochemical processes and membrane technology alone Improving the performance of the cadmium biosorption in or in combination have been promoted for the purpose of terms of the process efficiency can be achieved through the the heavy metal removal (Du et al. 2011; Liu et al. 2009). optimization, which provides an idea about the best opera- However, many of these processes increase the risk of gen- tional condition to yield the best possible response (adsorp- erating secondary pollutants, and therefore pose a concern tion) (Witek-Krowiak et al. 2011). Apart from the adsorp- from ecosystem health and public health viewpoints. As a tion kinetics which forms the basis for the selection of the consequence, the use of biological materials for the removal conditions, the modeling of the adsorption process, includ- of these heavy metals is being promoted to minimize the ing those of heavy metals such as cadmium, is carried out secondary pollutant level in the system as well as cost-effect to portray the changes in the efficiency depending on the benefit (Gifford et al. 2006). Application of the biological various physico-chemical parameters (Çelekli and Geyik materials that are component of the ecosystem reduces the 2011; Çelekli et al. 2013). Since the biosorption process is possibility of yielding unwanted secondary pollutants. These linked with different variables in a non-linear manner and are substantiated in the observations on the metal adsorption the mechanism of the process is multifaceted, it is difficult ability of different microorganisms (Hetzer et al. 2006; Sari to model using conventional mathematical simulation. In and Tuzen 2009) and hydrophytes (Sinha et al. 2007) many recent past, artificial neural network (ANN) model has been of which are hyperaccumulator of heavy metals. Application used in describing problems in different fields of chemical of the aquatic animals in the removal of the heavy metals and environmental engineering (Çelekli et al. 2012; Hossain from the aquatic system has been tested as evident from the et al. 2015; Maghsoudi et al. 2015). In a generalized ANN, studies on living freshwater bivalve (Jana and Das 1997), the outputs provide an overview of the relative importance bivalve shell (Du et al. 2011; Liu et al. 2009; Pena-Rodri- of the input parameters (factors influencing the biosorp- guez et al. 2010) crab and arca shell biomass (Dahiya et al. tion process) that drive the sorption process (Çelekli and 2008). However, the unregulated use of the living forms of Geyik 2011; Çelekli et al. 2013; Ahmad et al. 2014; Witek- the biological resources may facilitate species invasion, such Krowiak et al. 2014). In the present instance, the applica- as many of the plant hyperaccumulators (Sinha et al. 2007). tion of the ANN was meant for deducing the efficacy of the In other instances, such as the case of the freshwater bivalve, B. bengalensis shell dust (BSD) as biosorbent of cadmium, despite the metal removal being high (Jana and Das 1997), under varied physico-chemical conditions that influence the their potential as aquaculture resources limits their use in adsorption process. Thus, the cadmium absorption efficiency bioremediation. Recent studies have demonstrated that cal- of BSD was judged through multilayer perceptron in ANN cium carbonate and its derivatives may act as biosorbents model, apart from description of the equilibrium and kinetic of different heavy metals (Du et al. 2011; Pena-Rodriguez models. The results will be useful in supplementing the et al. 2010). The shells of different molluscs are composed of required information in utilizing the BSD, a waste generated 95–99% calcite and/or aragonite (principally made up of cal- from aquaculture as a biosorbent and thus in bioremediation cium carbonate) oriented in a matrix of proteins (0.1–5%), of heavy metal contaminated aquatic ecosystem. and act as natural ceramic with excellent strength and tough- ness (Kaplan 1998; Boro et al. 2012). The shells of aquatic gastropods can be considered as cheap and available source Materials and methods of calcium carbonate (Hossain and Aditya 2013) and thus can be considered as low-cost biosorbent of heavy met- Preparation of the metal solution als. In compliance with this proposition, the present study was aimed at evaluation of cadmium removal capacity of The experiments were initiated through preparation of a cad- −1 shell dust of freshwater gastropod Bellamya bengalensis mium chloride stock solution of 1000 mg L using double 1 3 Applied Water Science (2018) 8:68 Page 3 of 14 68 distilled water and further working solutions were prepared where q is the amount of metal adsorbed, v is the volume by appropriate dilution. The pH of the solution was adjusted of solution, m is the mass of adsorbent, c is the initial con- by adding HNO (0.1 N) and NaOH (0.1 N) as required in centration of the Cd solution and c is the equilibrium con- 3 e course of the experiments. All the inorganic chemicals that centration of the Cd solution. have been used in these experiments were purchased from Merck India Ltd., India. Estimation of the metal content in tissues and shells by AAS Preparation of the Bellamya shell dust For estimation of metal, required amount of Bellamya ben- The waste shells of the snail B. bengalensis were collected galensis tissue/shell of a particular dose was taken into Tef- from local fish markets in Burdwan, India. Following pro- lon container for microwave digestion. 4 mL of aqua regia curement, the remnant tissue portion and dusts in the shells was added and container was placed in microwave oven at were cleared in warm water and dried in sunlight for 2 days. 450  W and the sample was digested for 7  min. 4  mL of Next, the shells were placed in hot air oven for 2 days at 60 °C hydrogen peroxide was then added to the mixture and was and kept in plastic zipper bags. For preparation of the shell again digested for another 7 min. It was diluted with dis- dust, the shells were pulverized in mortar and pestle to fine tilled water, and filtered in a 25-mL volumetric flask. Final granules and dried in oven for 2 days at 60 °C. The granules volume was made up with distilled water. Reagent blank were then sieved through 500 µm and consequently through was prepared in the same way. The final samples were then 200 µm net, to yield two different sized granules, respectively, estimated for cadmium by atomic absorption spectroscopy 500–200 µm and ≥ 200 µm. Initially, the larger-sized granules (GBC Avanta 1.3, India) at 228.8 nm wavelength. adsorbed less amount of cadmium and thus were not studied further. The ash content of the dust was 72.34% and the water Analyses of FT‑IR absorbance spectra of BSD −1 content capacity was 1859.0 mg g . The specific surface 2+ area, pore volume and pore diameter were determined by the IR spectra of protonated or Cd loaded BSDs were recorded Brunauer–Emmett–Teller (BET) method using Quantachrome in a (Perkin-Elmer FTIR, Model RX1) Fourier transform Autosorb automate with nitrogen gas (version 4.0). The sur- infrared (FT-IR) spectroscopy. Samples of 100 mg of KBr 2 −1 face area of the biosorbent was 10.143 m  g , while the pore disks contain 1% of the finally ground powder of each sam- −1 volume was 0.079 cc g and the pore diameter was 3.336 nm. ple were prepared less than 24 h before recording. Batch sorption procedure The scanning electron micrograph (SEM) study of BSD The batch sorption experiments were performed in a 250-mL Erlenmeyer’s flask that contained 100 mL solution of the Raw and metal-adsorbed BSDs were dried and prepared for particular cadmium ion concentration at required pH and scanning electron microscopic studies. The samples were relevant amount of snail shell dust (BSD). The flasks were attached with the stubs by both side cello tapes and gold sealed with wax paper and shaken in a shaking incubator plated in a sputter coater before use in the SEM. Electron (Lab Companion, SI-300R, India) at a required temperature acceleration potential of 20 kV was used for the microscopic at 150 rpm for required time. After shaking for particular observations. Photographs were taken in a HITACHI S530 time, the solution of the flasks was centrifuged at 2000 rpm scanning electron microscope at 800× magnification. for 15 min and the supernatant was taken for estimation of metal concentration by atomic absorption spectroscopy Energy dispersive X‑ray (EDX) analysis (GBC Avanta 1.3, India). The influence of different levels of pH on biosorption equilibrium was studied through chang- For energy dispersive X-ray analysis, the samples were stud- ing the pH of the solution in range of 2–7. The effect of ied by field emission gun-based scanning electron micro- contact times between solution and the BSD were monitored scope with energy dispersive X-ray analysis (SEM/EDX) by varying it from 10 to 80 min. For equilibrium studies, the (Quanta model by FEI Co., the Netherlands) for the morpho- metal ion concentrations were used in a varied concentration logical as well as for the presence of elemental information −1 of 25–1000 mg L , and for optimum biosorption study, the on the samples. The SEM studies were performed at 30 kV BSD biomass was varied between 200 and 1000 mg. The in the low vacuum mode. EDX spectra were taken from the amount of Cadmium ion adsorbed on the BSD was estimated area corresponding to the SEM image shown in the in-set. following the equation (Sharma et al. 2011): The EDX spectra were taken in the ‘region mode’ with the bombardment of energetic electrons for duration of electron q = c − c v∕m, e o e of 100 s. 1 3 68 Page 4 of 14 Applied Water Science (2018) 8:68 Equilibrium modeling ln q = ln x −  , e m where β and x (maximum Cd adsorption capacity) are the The adsorption equilibriums were studied for the estima- m constants obtained from the plot of ln q vs. ɛ (Polanyi tion of maximum cadmium biosorption by the BSD. For the e potential). ‘ɛ’ is calculated from the following equation: equilibrium study, the experiments were performed at dif- −1 ferent initial cadmium ion concentration (25–1000 mg g ). = RT ln 1 + 1∕c , Langmuir and Freundlich isotherms were used in describ- −1 where R is the universal gas constant (J mol ) and T is the ing the equilibrium between adsorbed cadmium ions on the absolute temperature in Kelvin and ‘β’ is associated to the BSD (q ) and in solution (c ) at a particular temperature. The e e adsorption free energy (E). Adsorption energy is calculated parameters of the Langmuir equation (Langmuir 1918) were with the following equation (Tanzifi et al. 2017): determined from a linear form of the following equation: E = 1∕− 2. c ∕q = c ∕a + 1∕ab, e e e where a is the maximum amount of metal ions/unit mass of Kinetic modeling adsorbent to form a monolayer and b is the equilibrium con- stant. The correlation coefficient of the linear plot of c /q e e The pseudo second order model for prediction of biosorption against c indicates the applicability of the Langmuir mod- is not suitable for a long period of adsorption process (Ho eling in the experiment (Tan and Hameed 2017). and McKay 1999) and the pseudo second order equation: Freundlich equation (Freundlich 1906) is described as below: dq ∕dt = k q −q , t 2 e t 1∕n q = k c , or e f q = 1∕ k q + t∕q , where k and 1/n are the Freundlich constants indicating the t 2 e adsorption capacity and intensity, respectively. The linear where q is the amount of adsorbed metal ion on biosorbent model of the isotherm can be expressed logarithmically as at equilibrium, q is the amount of adsorbed metal ion on in the following: biosorbent at time ‘t’, and k is the second order rate constant −1 −1 (g mg  min ). A linear plot of t/q vs. t indicates whether log q = 1∕n log c + log K , e e F this model of biosorption is applicable for this case. where the value of K and n can be determined from the y-intercept and slope of the plot of log c against log q . The e e Artificial neural network (ANN) model linear plot of log q against log c indicates the applicability e e of the Freundlich modeling in the experiment. ANN is a non-linear mathematical model that is inspired by Temkin’s isotherm model is an empirical equation com- the structural and functional aspects of neuron in explora- prises a factor that clearly indicates the adsorbent–adsorb- tion of a group of input data (through training and testing) ate interactions. Through overlooking the lower and higher as function of output data (including relative error) and to value of concentrations, it assumes that heat of adsorption envisage the performance of the given system (Cavas et al. of the interacting molecules in the layer would decrease lin- 2011). A three-layered ANN model is used in the analyses early rather than logarithmically. The model is given by the of the complex interactions in finding the pattern of experi- following equation (Foo and Hameed 2010; Tanzifi et al. mental data. Multilayer perceptron uses supervised learn- 2017; 2018; Ghaedi and Vafaei 2017): ing technique in relating output responses to the expected q = RT∕b ln A c , e T T e responses by regulating the loads of the input signals and remind it. The memory is used in the next time of data feed or, when the output response will be much closer to the wanted q = RT∕b ln A + RT∕b ln c , e T T T e one after the same input signals are delivered (Shanmu- where A is the Temkin isotherm equilibrium binding con- gaprakash and Sivakumar 2013). The weight of any input −1 stant (L g ), b is the Temkin isotherm constant, R is the signal is calculated by the following equation: −1 −1 universal gas constant (8.314 J   mol  K), and T is the temperature at 298 K. A and B values are obtained from T T W = w x , i ij j q versus ln k plot for the adsorbent. e e j=1 Dubinin–Radushkevich (D–R) isotherm was used in investigation of Cd adsorption nature based on the equilib- rium data. The isotherm is expressed in linear form as: 1 3 Applied Water Science (2018) 8:68 Page 5 of 14 68 where w is the corresponding load of connection between Results and discussion ij each neuron ‘j’ in input layer and each neuron ‘i’ at hidden layer and x is the value of input signal ‘j’ at input layer Bioaccumulation of cadmium on shell dust (Çelekli et al. 2013). For a particular function, the output is predicted from the following equation: Eec ff t of pH of the solution on batch sorption y =  (w + b ), k i j The biosorption procedure was maintained over the experi- where w is the sum weight of each neuron and b is the bias, i j mental range of pH 2–7. The sorption procedure was affected i.e., constant weight of a neuron representing the relative by the pH of the medium in two ways—metal solubility and error (Çelekli and Geyik 2011). In predicting the biosorp- total charge of the functional groups of the biosorbent. The tion of cadmium (output), the combined effect of the vari- optimum pH, at which the procedure shows the highest 2+ ables—(1) pH, (2) Cd concentration, (3) biosorbent dose adsorption for the biosorbent was estimated to be 6 which and (4) contact time were used in the analysis. Sixty-nine is nearer to the environmental pH of most of the freshwater experimental sets were used in training, validation and test- aquatic system. At high pH, that is, at alkaline condition, ing the ANN model by SPSS 20, trial version. precipitation of the metal takes place, and at low pH due to high protonation, metal sorption capacity decreases (Guo et al. 2008). The experiment was carried out in 100 mL of −1 2+ solution having 100 mg L Cd and 100 mg of the BSD at 30 °C in reference to varying pH of the solution. Figure 1a indicates the pH-dependent adsorption of the metal ion by BSD as the metal sorption was negligible at pH 2 and it Fig. 1 a Effect of pH of the A B solution on biosorption of cad- 10 30 mium on BSD at 30 °C, initial cadmium ion concentration of −1 100 mg L , contact time of 8 60 min and biosorbent dose of 1 g per 100 mL solution. b Effect of initial cadmium ion concentration of solution on biosorption on BSD at 30 °C, pH 6, contact time of 60 min and biosorbent dose of 1 g per 100 mL solution. c Effect of time on biosorption of cadmium on BSD at 30 °C, pH 6, initial cadmium ion concentration 01234567 8 0200 400600 80010001200 −1 of 100 mg L , and biosorb- 2+ -1 pH Cd conc. in solution (mg L ) ent dose of 1 g per 100 mL C D solution. d Effect of biosorbent dose on biosorption of cadmium 10 10 on BSD at 30 °C, pH 6, initial cadmium ion concentration of −1 100 mg L , contact time of 60 min 0200 400600 80010001200 Adsorption time (min) Amount of BSD (mg/100ml) 1 3 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 68 Page 6 of 14 Applied Water Science (2018) 8:68 increases dramatically as increase in pH and at pH 6 it shows spectrum of the BSD was performed. IR spectra of proto- the highest adsorption. The metal sorption declined as the nated and cadmium loaded BSD are shown in Fig. 2a, b, −1 pH increase further. In a particular pH range, most metal respectively. A peak at 3424 cm is indicating the presence −1 sorption is enhanced with pH, increasing to a certain value of hydroxyl (–OH) groups. Strong band at 2921 cm is due followed by a reduction on further pH increase (Guo et al. to C–H stretching frequency (Semerciöz et al. 2017) and −1 2008; Pavasant et al. 2006; Semerciöz et al. 2017). peak at 1634 cm is due to C=O stretching mode of the primary and secondary amides (Reddy et al. 2010). Weak −1 Eec ff t of initial metal ion concentration on batch sorption band at 1462 cm is attributed to aromatic C=C and the −1 strong band at 1041 cm is due to C–O stretching of alco- Increase in metal ion concentration added on metal adsorp- holic groups (Blazquez et al. 2011). FTIR study reveals that −1 tion as concentration of the ion increases. 25–1000 mg L OH, CH, CO, and C=C may be the responsible groups in the ion concentrations were used for the study taking seven biosorption process. The FTIR of metal loaded BSD shows −1 different doses in series. The 25 mg L initial metal ion that distinct shift of the above mentioned bands as well as concentration showed the lowest adsorption while the change in intensity informs some ion exchange behavior of −1 400 mg L concentration showed the highest adsorption the BSD. and the adsorption remained same in further increase in 2+ Cd ion increase in the solution (Fig. 1b). The initial metal The scanning electron micrograph (SEM) analysis ion concentration can modify the metal removal efficiency of the shell dust over a combination of factors, the availability of specific surface functional groups and ability of surface functional The surface structure of the free and cadmium-loaded BSD groups to bind metal ions (Taty-Costodes et al. 2003; Pino was analyzed under scanning electron microscope. The et al. 2006) scanning electron micrographs of the dried BSD before and 2+ after the Cd treatment at 800× magnification are shown Eec ff t of contact time on batch sorption in Fig.  3a, b, respectively. It indicates the irregular mor- phological structure of the particles and lamellar stratified The sorption potentials of the BSD over time were moni- surface of the BSD. The SEM image of the biosorbent after 2+ tored from 10 min to 20, 40, 60, 80 min using 100 mL of exposure to the Cd shows a spongy layer indicating surface −1 2+ 100 mg L Cd at pH 6 (Fig. 1c). At the beginning, metal precipitation occurred during the sorption (Du et al. 2011). adsorption was less due to more binding sites remained In case of cadmium sorption by CaCO compound gener- free when treated for the short period of time and increased ally follows surface precipitation due to similar ionic radii rapidly as the treatment time increases. It showed lowest (Prieto et al. 2003; Perez-Garrido et al. 2007) of divalent adsorption when treated for 10 min and increased over time calcium and cadmium. to saturate at 60  min and after that the uptake remained almost same. The variation in uptake of the cadmium ions Energy dispersive X‑ray (EDX) analysis with time was used in fitting the kinetic models. The surface structure of the shell dust makes accessible a Influence of the biosorbent dose on batch sorption large unadsorbed surface area for the cadmium ion of the solution. The elemental profile of the BSD before and after The cadmium biosorption potential of the BSD augmented the treatment of the cadmium solution was estimated using over its amount increased in treating the metal solution. the energy dispersive X-ray analysis. The projecting peaks The more the amount of biosorbent the more the free bind- in the EDX spectra correspond to CKα, OKα, AlKα, SiKα, ing sites or exchanging group to adsorb the metal ion from PKα, CaKα, CaKβ, FeKα, etc., in the untreated BSD and −1 the solution. For a 100 mg L metal ion concentration, the furthermore to it a CdLα peak in treated BSD (Fig. 4a, b). increase in biosorbent resulted increase in metal ion adsorp- Both the treated and untreated BSD display strong peaks tion and above a certain dose it remained same or slightly corresponding to calcium. In addition, the calcium of BSD higher due to comparatively higher number of free sites and may facilitate adsorption of cadmium because of the simi- lesser number of metal ions (Fig. 1d) (Al-Anber and Matouq larities in ionic radii that enhance ion exchange (Purkayastha 2008; Ghodbane et al. 2008). et al. 2014). FT‑IR study of the shell dust Equilibrium modeling To study the mechanism of cadmium removal and the main The biosorption isotherm is important in waste water treat- 2+ functional groups responsible for Cd binding, the FTIR ment as it implies estimation of biosorption capacity of 1 3 Applied Water Science (2018) 8:68 Page 7 of 14 68 Fig. 2 FTIR absorbance spectra of snail shell dust (BSD) before 13.3 (a) and after (b) biosorption of cadmium 864.28 618.01 1082.87 %T 1467.99 1618.13 1563.27 3420.87 2.0 4400.0 4000 3000 2000 1500 1000 400.0 cm-1 11.2 864.51 626.36 1078.90 2921.85 %T 1424.31 1563.27 1635.92 3428.08 1.0 4400.0 4000 3000 2000 1500 1000 400.0 cm-1 Fig. 3 Scanning electron micrograph (SEM) of snail shell dust (BSD) before (a) and after (b) the biosorption of cadmium (magnification at ×800) 1 3 68 Page 8 of 14 Applied Water Science (2018) 8:68 Fig. 4 EDX analyses for elemental composition in the snail shell dust for BSD before (a) and after (b) biosorption of cadmium clearly showing the peaks for cadmium, in the treated shell dust, indicating the adsorption of the metal on the shell dust the adsorbent. The linear representations of Langmuir and cadmium biosorption support that under optimum condi- 2+ Freundlich isotherm of cadmium adsorption at 30 °C are tions (pH 6, biosorbent dose of 1 g, Cd concentration of −1 −1 given in Fig. 5a, b, respectively. The correlation coefficient 100 mg L and 60 min time period) 30.33 mg g is the and constants obtained from the equations are presented in maximum biosorption capacity of BSD. The Temkin iso- Table 1. The correlation of determination is high in Lang- therm shows B = 533, T = 1.15 and the R = 0.896 showing T A 2 2 muir equation (R = 0.995) contrast to Freundlich equation moderate t fi . The R–D equation ( R = 0.766) for the biosorp- (R = 0.760) (Fig. 5). It indicates that Langmuir model is tion shows the value of the adsorption energy (E) is 13.89 J/ more suitable for describing the biosorption equilibrium of mol that corresponds to chemisorption type of uptake based 2+ −1 Cd on the snail shell dust. High q value (30.33 mg g ) on ion exchange (Wu et al. 2012; Markou et al. 2016). Maxi- max from Langmuir equation indicates the biosorption potential mum cadmium biosorption capacities of similar low cost of the material. The value of b (0.04) (Table 1) indicates b biosorbents are shown in Table 2. BSD is mainly composed the affinity of the binding sites and the energy of adsorption of calcium carbonate, degradation of which, if any, will not (L/mg) (Blazquez et al. 2011). The equilibrium models of yield unwanted compound to the ecosystem. 1 3 Applied Water Science (2018) 8:68 Page 9 of 14 68 Fig. 5 a Langmuir isotherm, b AB Freundlich isotherm, c Temkin 40 2 isotherm and d Dubinin–Radu- shkevich isotherm plot for y = 0.0338x + 0.6614 1.5 biosorption of cadmium on R² = 0.9954 BSD (pH 6, temperature 30 °C, biosorbent = 1 g, cadmium ion −1 y = 0.3747x + 0.5136 concentration = 100 mg g ) 0.5 R² = 0.7608 0200 400600 800 e log c CD 40 4 y = 4.6404x + 0.6756 R² = 0.8946 y = -17.489x + 2.9892 R² = 0.7663 00.050.1 0.15 0246 8 [ln(1+1/ Ce)] ln c Table 1 Coefficients of the Langmuir, Freundlich, Temkin isotherm and Dubinin–Radushkevich models for cadmium biosorption by BSD Langmuir coefficients Freundlich coefficients Temkin coefficients D–R coefficients −1 −1 2 2 −1 2 −1 2 q (mg g ) b (L mg ) R n k R T (L mg ) b R q (mg g ) E R max F A T max 30.33 0.049 0.995 2.67 1.67 0.760 1.15 533 0.894 9.211 13.89 0.766 −1 adsorption (9.26 mg g ) is much nearer to the expected Kinetic modeling −1 value (9.302 mg g ). These suggest that the biosorption process is based on the pseudo second order model. The kinetic model is necessary for determination of opti- mal condition of the biosorption process. For the evalua- Optimization results through ANN model tion of differences in sorption process, the kinetics of metal uptake were described by pseudo second order model (Ho The ANN model established in this study comprised three and McKay 1999). The linear plots obtained from pseudo −1 2+ layers as shown in Fig. 7a consisting of a hyperbolic tan- second order model at 100 mg L initial Cd concentra- gent transfer function at hidden layer and a linear transfer tion, pH 6 and at studied temperature are shown in Fig. 6a, function at output layer. The input layers have four predic- b, respectively. The rate constants, expected metal uptake tors, viz. biomass, pH, Cd concentration and time, the hid- and correlation coefficients have been described in Table  3. den layer and the output layer have one response variable, The pseudo second order reaction in biosorption is based the amount of Cd absorbed(Table 4) (Çelekli et al. 2012, on the sorption capacity on the solid phase. From Fig. 6b 2016). The weight of the neurons was used in the study and Table 3, the correlation coefficient in the pseudo second of the relative influence of each of the input variables on order reaction (R = 0.998) is high and the calculated metal 1 3 e c /q e e log q ln q e 68 Page 10 of 14 Applied Water Science (2018) 8:68 Table 2 Comparative data of biosorption capacities (q —maximum 10 max metal uptake capacity) for cadmium by different biosorbents −1 Adsorbent q (mg g ) References max Raw corn stalk 3.39 Zheng et al. (2010) Olive waste 6.55 Azouaou et al. (2008) y = 0.108x + 0.1836 Brewer’s yeast 10.17 Cui et al. (2010) 2 R² = 0.999 Corncob 4.73 Ramos et al. (2005) Rice straw 13.9 Ding et al. (2012) Wheat bran 15.71 Nouri et al. (2007) Time (min) Castor seed hull 6.98 Sen et al. (2010) Bamboo charcoal 12.08 Wang et al. (2010) Fig. 6 Pseudo second order plot for biosorption of cadmium on BSD Walnut tree sawdust 5.76 Yasemin and Zek (2007) (pH 6, temperature 30 °C, biosorbent = 1 g, cadmium ion concentra- −1 Chitosan/bentonite 12.05 Arvand and Pakseresht tion = 100 mg g ) (2012) Coconut copra meal 4.99 Ho and Ofomaja (2006) Table 3 Rate constant and equilibrium uptake for cadmium binding S. platensis 73.64 Çelekli and Bozkurt −1 by BSD at pH 6 and initial metal ion concentration of 100 mg L (2011) Metal Second order model Sweet potato 18.00 Asuquo and Martin (2016) −1 −1 −1 2 P. pubescens biochars 14.70 Zhang et al. (2017) k (g mg  min ) q (mg g ) R 2 e cal. Bellamya shell dust 30.33 The present study Cadmium 0.064 9.26 0.998 (BSD) the biosorption process, i.e., on output layer. In the present Conclusion instance, the sum of squares error (SSE error, a measure of the predicted and the observed values; Çelekli and Geyik The biosorption process works as function of pH of the solu- 2011) is 0.467; and the relative error being 0.056, justify- tion, biomass, Cd ion concentration and time of contact. ing the suitability of the multilayer perceptron ANN model. The biosorption study indicates that at pH 6, the maximum −1 The regression equation of the output (predicted adsorp- cadmium adsorption capacity of BSD is 30.33 mg g . The tion through ANN model, y) and the corresponding target isotherm model follows Langmuir model (R = 0.995) better 2 2 (observed through the experiments, x) complied with the lin- than Temkin isotherm (R = 0.896), R–D (R = 0.766) and ear form as y = 0.37 + 0.95x (Fig. 7b). The high (R = 0.943) Freundlich model (R = 0.760). The main functional group coefficient of determination fits well with the experimental responsible for chelation is OH, C=O, C=C and C–C, as dataset with that of the ANN model of the system (Yang supported by FTIR analysis. EDX study shows clear peaks et al. 2011; Khataee et al. 2011). Upon comparison, the data for cadmium in the treated biomaterial, conveying further complied well with both the pseudo second order (Fig. 6) as support of clear adsorption. The biosorption process fol- well as the ANN model (Fig. 7b), with high values of coef- lowed pseudo second order (R = 0.998) kinetics. A multi- ficient of determination. Among the predictors, the pH of the layer perceptron model in artificial neural network (ANN) solution (0.347) appeared to be comparable to the cadmium model successfully portrayed biosorption of cadmium ion concentration (0.339) in terms of the relative importance in on the LSD with high correlation coefficient (R = 0.943) shaping the adsorption process, similar to those observed between predicted and observed removal. BSD can be for the dye removal by Chara contraria (Çelekli and Geyik considered as efficient, low cost and environment friendly 2011). The contact time and the biomass bear comparative biosorbent for cadmium bioremediation, opening a new low influence in the adsorption, contrast to those observed aspect of the economic value of shells of the freshwater snail for the dye removal by the walnut husk (Çelekli et al. 2016). B. bengalensis, apart from its value as cheap protein source. 1 3 t/q Applied Water Science (2018) 8:68 Page 11 of 14 68 Fig. 7 Basic design (a) of the multilayer perceptron artificial neural network (ANN) model used in the study along with the regression line (b) endorsing the linearity of predicted and observed cadmium removal by BSD 1 3 68 Page 12 of 14 Applied Water Science (2018) 8:68 Table 4 Basic information of Network information the artificial neural network used in describing the Input layer Covariates biosorption of cadmium by  1 pH BSD  2 Cd concentration  3 Biomass  4 Time Number of units 4 Rescaling method for covariates Standardized Hidden layer(s) Number of hidden layers 1 Number of units in hidden layer 1 2 Activation function Hyperbolic tangent Output layer Dependent variables  1 Removal Number of units 1 Rescaling method for scale dependents Standardized Activation function Identity Error function Sum of squares Excluding the bias unit Azouaou N, Sadaoui Z, Mokaddem H (2008) Removal of cadmium Acknowledgements The authors are thankful to the respective Heads, from aqueous solution by adsorption on vegetable wastes. J Department of Zoology, Department of Chemistry, and Department of Appl Sci 8:4638–4643 Environmental Science, the University of Burdwan, Burdwan, West Bernard A (2008) Cadmium and its adverse effects on human health. Bengal, India for the facilities provided. We thankfully acknowledge Indian J Med Res 128:557–564 the anonymous reviewers for their thoughtful comments and kind Blazquez G, Martin-Lara MA, Tenorio G, Calero M (2011) Batch suggestions to enrich this manuscript. We express our gratitude and biosorption of lead(II) from aqueous solutions by olive tree regards to Prof. Enrico Drioli for his kind cooperation in revising the earlier version of the manuscript. AH thankfully acknowledges pruning waste: equilibrium, kinetics and thermodynamic study. the financial assistance provided Council of Scientific and Industrial Chem Eng J 168:170–177 Research (CSIR), New Delhi, India. Boro J, Deka D, Thakur AJ (2012) A review on solid oxide derived from waste shells as catalyst for biodiesel production. Renew Sustain Energy Rev 16:904–910 Open Access This article is distributed under the terms of the Crea- Brooks BW, Stanley JK, White JC, Turner PK, Wu KB, La Point TW tive Commons Attribution 4.0 International License (http://creat iveco (2004) Laboratory and field responses to cadmium: an experi- mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- mental study in effluent-dominated stream mesocosms. Environ tion, and reproduction in any medium, provided you give appropriate Toxicol Chem 23:1057–1064 credit to the original author(s) and the source, provide a link to the Cavas L, Karabay Z, Alyuruk H, Dogan H, Demir GK (2011) Thomas Creative Commons license, and indicate if changes were made. and artificial neural network models for the fixed bed adsorption of methylene blue by a beach waste, Posidonia oceanea (L.) dead leaves. Chem Eng J 171(2):557–562 Çelekli A, Bozkurt H (2011) Bio-sorption of cadmium and nickel ions using Spirulina platensis: kinetic and equilibrium studies. References Desalination 275:141–147 Çelekli A, Geyik F (2011) Artificial neural networks (ANN) Ahmad MF, Haydar S, Bhatti AA, Bari AJ (2014) Application of arti- approach for modeling of removal of Lanaset Red G on Char- ficial neural network for the prediction of biosorption capacity of acontraria. Bioresour Technol 102:5634–5638 immobilized Bacillus subtilis for the removal of cadmium ions Çelekli A, Birecikligil SS, Geyik F, Bozkurt H (2012) Prediction from aqueous solution. Biochem Eng J 84:83–90 of removal efficiency of Lanaset Red G on walnut husk using Al-Anber ZA, Matouq MAD (2008) Batch adsorption of cadmium ions artificial neural network model. Bioresour Technol 103:64–70 from aqueous solution by means of olive cake. J Hazard Mater Çelekli A, Bozkurt H, Geyik F (2013) Use of artificial neural network 151:194–201 and genetic algorithms for prediction of sorption of an azo–metal Arvand M, Pakseresht MA (2012) Cadmium adsorption on modified complex dye onto lentil straw. Bioresour Technol 129:396–401 chitosan coated bentonite: batch experimental studies. J Chem Çelekli A, Bozkurt H, Geyik F (2016) Artificial neural network and Technol Biotechnol 88(4):572–578 genetic algorithms for modeling of removal of an azo dye on wal- Asuquo ED, Martin AD (2016) Sorption of cadmium (II) ion from nut husk. Desalin Water Treat 57:15580–15591 aqueous solution onto sweet potato (Ipomoea batatas L.) peel Cui L, Wu G, Jeong T (2010) Adsorption performance of nickel and adsorbent: characterisation, kinetic and isotherm studies. J Envi- cadmium ions onto brewer’s yeast. Can J Chem Eng 88:109–115 ron Chem Eng 4:4207–4228 1 3 Applied Water Science (2018) 8:68 Page 13 of 14 68 Dahiya S, Tripathi RM, Hegde AG (2008) Biosorption of lead and Ca(OH) -pretreated biomass of Phragmites sp. J Environ Sci copper from aqueous solutions by pre-treated crab and arca shell 45(49–5):9 biomass. Bioresour Technol 99:179–187 Nouri L, Ghodbane I, Hamdaoui O, Chiha M (2007) Batch sorption Ding Y, Jing D, Gong H, Zhou L, Yang X (2012) Biosorption of dynamics and equilibrium for the removal of cadmium ions from aquatic cadmium (II) by unmodified rice straw. Bioresour Tech- aqueous phase using wheat bran. J Hazard Mater 149:115–125 nol 114:20–25 Pavasant P, Apiratikul R, Sungkhum V, Suthiparinyanont P, Wattan- 2+ 2+ 2+ Du Y, Lian F, Zhu L (2011) Biosorption of divalent Pb, Cd, Zn on arag- achira S, Marhaba TF (2006) Biosorption of Cu, Cd, Pb , 2+ onite and calcite mollusk shells. Environ Pollut 159:1763–1768and Zn using dried marine green macroalga Caulerpa lentillif- Foo KY, Hameed BH (2010) Insights into the modeling of adsorption era. Bioresour Technol 97(18):2321–2329 isotherm systems. Chem Eng J 156:2–10 Pena-Rodriguez S, Fernandez-Calvino D, Novoa-Munoz JC, Nunez- Freundlich HMF (1906) Uber die adsorption in losungen. Z Phys Chem Delgado A, Fernandez-Sanjurjo MJ, Alvarez-Rodriguez A (2010) 57A:385–470 Kinetics of Hg(II) adsorption and desorption in calcined mussel Ghaedi AM, Vafaei A (2017) Application of artificial neural networks shells. J Hazard Mater 180:622–627 for adsorption removal of dyes from aqueous solution: a review. Perez-Garrido C, Fernandez-Diaz L, Pina CM, Prieto M (2007) In situ Adv Colloid Interface Sci 245:20–39 AFM observations of the interaction between calcite surfaces and Ghodbane I, Nouri L, Hamdaoui O, Chiha M (2008) Kinetic and equi- Cd-bearing aqueous solutions. Surf Sci 601:5499–5509 librium study for the sorption of cadmium (II) ions from aqueous Pino GH, deMesquita LMS, Torem ML, Pinto GAS (2006) Biosorp- phase by eucalyptus bark. J Hazard Mater 152:148–158 tion of cadmium by green coconut shell powder. Miner Eng Gifford S, Dunstan RH, O’Connor W, Koller CE, MacFarlane GR 19:380–387 (2006) Aquatic zooremediation: deploying animals to remediate Prieto M, Cubillas P, Fernandez-Gonzalez A (2003) Uptake of dis- contaminated aquatic environment. Trends Biotechnol 25:60–65 solved Cd by biogenic and abiogenic aragonite: a comparison with Godt J, Scheidig F, Grosse-Siestrup C, Esche V, Brandenburg P, Reich sorption onto calcite. Geochim Cosmochim Acta 67:3859–3869 A, Groneberg DA (2006) The toxicity of cadmium and resulting Purkayastha D, Mishra U, Biswas S (2014) A comprehensive review hazards for human health. J Occup Med Toxicol 1(1):22 on Cd(II) removal from aqueous solution. J Water Process Eng Guo X, Zhang S, Shan XQ (2008) Adsorption of metal ions on lignin. 2:105–128 J Hazard Mater 151(1):134–142 Ramos RL, Jacome LAB, Rodriguez IA (2005) Adsorption of cadmium Hetzer A, Daughney CJ, Morgan HW (2006) Cadmium ion biosorption (II) from aqueous solution on natural and oxidized corncob. Sep by the thermophilic bacteria Geobacillus stearothermophilus and Purif Technol 45:41–49 G. thermocatenulatus. Appl Environ Microbiol 72:4020–4027 Reddy HKD, Seshaiah K, Reddy AVR, Rao MM, Wang MC (2010) 2+ Ho YS, McKay G (1999) Pseudo-second order model for sorption pro- Biosorption of Pb from aqueous solutions by Moringa oleif- cess. Process Biochem 34:451–465 era bark: equilibrium and kinetic studies. J Hazard Mater Ho YS, Ofomaja AE (2006) Biosorption thermodynamics of cadmium 174:831–838 on coconut copra meal as biosorbent. Biochem Eng J 30:117–123 Sari A, Tuzen M (2009) Kinetic and equilibrium studies of biosorp- Hossain A, Aditya G (2013) Cadmium biosorption potential of shell tion of Pb(II) and Cd(II) from aqueous solution by macrofungus dust of the fresh water invasive snail Physa acuta. J Environ Chem (Amanita rubescens) biomass. J Hazard Mater 164:1004–1011 Eng 1:574–580 Semerciöz AS, Göğüş F, Çelekli A, Bozkurt H (2017) Development Hossain A, Bhattacharyya SR, Aditya G (2015) Biosorption of cad- of carbonaceous material from grapefruit peel with microwave mium from aqueous solution by shell dust of the freshwater snail implemented-low temperature hydrothermal carbonization tech- Lymnaealuteola. Environ Technol Innov 4:82–91 nique for the adsorption of Cu (II). J Clean Prod 65:599–610 Jana BB, Das S (1997) Potential of freshwater mussel (Lamellidens Sen TK, Mohammod M, Maitra S, Dutta BK (2010) Removal of Cad- marginalis) for cadmium clearance in a model system. Ecol Eng mium from aqueous solution using castor seed hull: a kinetic and 8:179–193 equilibrium study. Clean Soil Air Water 38:850–858 Johri N, Jacquillet G, Unwin R (2010) Heavy metal poisoning: the Shanmugaprakash M, Sivakumar V (2013) Development of experi- effects of cadmium on the kidney. Biometals 23:783–792 mental design approach and ANN-based models for determination Kaplan DL (1998) Mollusc shell structure: novel design strategies for of Cr(VI) ions uptake rate from aqueous solution onto the solid synthetic materials. Curr Opin Solid State Mater Sci 3:232–236 biodiesel waste residue. Bioresour Technol 148:550–559 Khan RA, Choudhury S (1984) The population and production ecol- Sharma M, Kaushik A, Kaushik CP (2011) Biosorption of reactive ogy of a freshwater snail, Bellamya bengalensis (Lamarck) (Gas- dye by waste biomass of Nostoc lincia. Ecol Eng 37:1589–1594 tropoda: Viviparidae) in an artificial lake of Calcutta, India. Bull Sinha RK, Herat S, Tandon PK (2007) Phytoremediation: role of plants Zool Surv India 5:59–76 in contaminated site management. In: Singh N, Tripathi RD (eds) Khataee AR, Dehghan G, Zarei M, Ebadia E, Pourhassan M (2011) Environ. Biorem. Technol. Springer, New York, pp 315–330 Neural network modelling of biotreatment of triphenylmeth- Tan KL, Hameed BH (2017) Insight into the adsorption kinetics mod- ane dye solution by a green macroalgae. Chem Eng Res Des els for the removal of contaminants from aqueous solutions. J 89:172–178 Taiwan Inst Chem Eng 74:25–48 Langmuir I (1918) The adsorption of gases on plane surface of glass, Tanzifi M, Hossein SH, Kiadehi AD, Olazar M, Karimipour K, mica and platinum. J Am Chem Soc 40:1361–1403 Rezaiemehr R, Ali I (2017) Artificial neural network optimiza- Liu Y, Sun C, Xu J, Li Y (2009) The use of raw and acid pretreated tion for methyl orange adsorption onto polyaniline nano-adsor- bivalve mollusk to remove metals from aqueous solutions. J Haz- bent: kinetic, isotherm and thermodynamic studies. J Mol Liq ard Mater 168:156–162 244:189–200 Maghsoudi M, Ghaedi M, Zinali A, Ghaedi AM, Habibi MH (2015) Tanzifi M, Yaraki MT, Kiadehi AD, Hosseini SH, Olazar M, Bhati Artificial neural network (ANN) method for modeling of sunset AK, Agarwal S, Gupta VK, Kazemi A (2018) Adsorption of yellow dye adsorption using zinc oxide nanorods loaded on acti- Amido Black 10B from aqueous solution using polyaniline/SiO vated carbon: kinetic and isotherm study. Spectrochim Acta A nanocomposite: experimental investigation and artificial neural Mol Biomol Spectrosc 134:1–9 network modeling. J Colloid Interface Sci 510:246–261 Markou G, Mitrogiannis D, Muylaert K, Çelekli A, Bozkurt H Tao T, Yuan Z, XiaonaH Wei M (2012) Distribution and bioaccumu- (2016) Biosorption and retention of orthophosphate onto lation of heavy metals in aquatic organisms of different trophic 1 3 68 Page 14 of 14 Applied Water Science (2018) 8:68 levels and potential health risk assessment from Taihu lake, China. Yang Y, Wang G, Wang B, Li Z, Jia X, Zhou Q, Zhao Y (2011) Ecotoxicol Environ Saf 81:55–64 Biosorption of Acid Black 172 and Congo Red from aqueous Taty-Costodes VC, Fauduet H, Porte C, Delacroix A (2003) Removal solution by non-viable Penicillium YW 01: kinetic study, equilib- of Cd(II) and Pb(II) ions, from aqueous solutions, by adsorption rium isotherm and artificial neural network modelling. Bioresour onto sawdust of Pinus sylvestris. J Hazard Mater 105:121–142 Technol 102:828–834 Wang FY, Wang H, Ma JW (2010) Adsorption of cadmium (II) ions Yasemin B, Zek T (2007) Removal of heavy metals from aqueous solu- from aqueous solution by a new low-cost adsorbent—bamboo tion by sawdust adsorption. J Environ Sci 19:160–166 charcoal adsorption of cadmium (II) ions from aqueous solution Zhang C, Shan B, Tang W, Zhu Y (2017) Comparison of cadmium by a new low-cost adsorbent—bamboo charcoal. J Hazard Mater and lead sorption by Phyllostachys pubescens biochar produced 177:300–306 under a low-oxygen pyrolysis atmosphere. Bioresour Technol Witek-Krowiak A, Szafran RG, Modelski S (2011) Biosorption of 238:352–360 heavy metals from aqueous solutions onto peanut shell as a low- Zheng L, Dang L, Yi X, Zhang H (2010) Equilibrium and kinetic stud- cost biosorbent. Desalination 265:126–134 ies of adsorption of Cd(II) from aqueous solution using modified Witek-Krowiak A, Chojnacka K, Podstawczyk D, Dawiec A, Pokom- corn stalk. J Hazard Mater 176:650–656 eda K (2014) Application of response surface methodology and artificial neural network methods in modelling and optimization Publisher’s Note Springer Nature remains neutral with regard to of biosorption process. Bioresour Technol 160:150–160 jurisdictional claims in published maps and institutional affiliations. Wu Y, Wen Y, Zhou J, Dai Q, Wu Y (2012) The characteristics of waste Saccharomyces cerevisiae biosorption of arsenic(III). Environ Sci Pollut Res 19:3371–3379 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Water Science Springer Journals

Valorization of aquaculture waste in removal of cadmium from aqueous solution: optimization by kinetics and ANN analysis

Free
14 pages
Loading next page...
 
/lp/springer_journal/valorization-of-aquaculture-waste-in-removal-of-cadmium-from-aqueous-tks0Bu0E38
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Earth Sciences; Hydrogeology; Water Industry/Water Technologies; Industrial and Production Engineering; Waste Water Technology / Water Pollution Control / Water Management / Aquatic Pollution; Nanotechnology; Private International Law, International & Foreign Law, Comparative Law
ISSN
2190-5487
eISSN
2190-5495
D.O.I.
10.1007/s13201-018-0712-z
Publisher site
See Article on Publisher Site

Abstract

Cadmium is one of the most hazardous heavy metal concerning human health and aquatic pollution. The removal of cadmium through biosorption is a feasible option for restoration of the ecosystem health of the contaminated freshwater ecosystems. In compliance with this proposition and considering the efficiency of calcium carbonate as biosorbent, the shell dust of the economically important snail Bellamya bengalensis was tested for the removal of cadmium from aqueous medium. Follow- ing use of the flesh as a cheap source of protein, the shells of B. bengalensis made up of CaC O are discarded as aquaculture waste. The biosorption was assessed through batch sorption studies along with studies to characterize the morphology and surface structures of waste shell dust. The data on the biosorption were subjected to the artificial neural network (ANN) model for optimization of the process. The biosorption process changed as functions of pH of the solution, concentration of heavy metal, biomass of the adsorbent and time of exposure. The kinetic process was well represented by pseudo second 2 2 −1 order (R = 0.998), and Langmuir equilibrium (R = 0.995) had better fits in the equilibrium process with 30.33 mg g of maximum sorption capacity. The regression equation (R = 0.948) in the ANN model supports predicted values of Cd removal 2+ satisfactorily. The normalized importance analysis in ANN predicts Cd concentration, and pH has the most influence in removal than biomass dose and time. The SEM and EDX studies show clear peaks for Cd confirming the biosorption process while the FTIR study depicts the main functional groups (–OH, C–H, C=O, C=C) responsible for the biosorption process. The study indicated that the waste shell dust can be used as an ec ffi ient, low cost, environment friendly, sustainable adsorbent for the removal of cadmium from aqueous solution. Keywords Cadmium · Shell · Valorization · Kinetics · ANN Abbreviations Introduction AAS Atomic absorption spectroscopy ANN Artificial neural network The heavy metals in aquatic ecosystem can be detrimental SEM Scanning electron microscope to different organisms depending on the tolerance levels and BSD Bellamya shell dust the complexity of the food web. In course of the different FT-IR Fourier transformed infrared industrial and agricultural processes, the heavy metals are EDX Energy dispersive X-ray released and eventually contaminate the aquatic ecosystems. BCF Bio-concentr ation factor For instance, cadmium, considered as one of the important heavy metal pollutant, is added to the aquatic ecosystems through activities in mining, electroplating, battery, paint and ceramic industries, in addition to the natural deposits. The disposal of heavy metals in aquatic systems concen- * Asif Hossain trates as a cascading effect in the trophic levels, frequently asifhossain.bu@gmail.com recognized in the form of bioaccumulation and biomagni- Gautam Aditya fications (Brooks et al. 2004; Tao et al. 2012). Subsequent Gautamaditya2001@gmail.com entry of heavy metals in the living system can create a chain Department of Zoology, University of Calcutta, 35 of physiological, biochemical and genetic changes that are Ballygunge Circular Road, Kolkata 700 019, India concern from public health viewpoint. In case of cadmium, Department of Zoology, Sidho Kanho Birsha University, entry in the human body can damage the kidney, substitute Ranchi Road, Purulia 723104, India Vol.:(0123456789) 1 3 68 Page 2 of 14 Applied Water Science (2018) 8:68 calcium in the bones, can cause liver damage, cancer and (Mollusca: Gastropoda: Viviparidae). A common inhabitant hypertension (Godt et al. 2006; Bernard 2008). Accumula- of ponds, ditches, rivers and varied types of freshwater wet- tion of cadmium in the food web causes damage to the wild lands, the operculate snail B. bengalensis a prolific breeder life and their diversity (Johri et al. 2010; Bernard 2008). As and can easily be cultured (Khan and Choudhury 1984). a consequence, the removal of cadmium and similar heavy The flesh of the snail is used as a cheap source of protein metals are given priority for the sustenance of the ecological and the shells are discarded. The discarded shells that are functions of the aquatic ecosystem. aquacultural waste may be used as cheap source of calcium The heavy metal pollutants are non-biodegradable and carbonate and the Bellamya shell dust (BSD) can be used in thus adsorption is considered as one of the convenient way the removal of cadmium from aquatic system. of removal from the aquatic ecosystems. In recent past, To promote the biosorption as an effective process, the several methods based on the principle of ion exchange, solution for the optimization is required since the process is chemical precipitation, coagulation, activated charcoal, influenced by several factors (Witek-Krowiak et al. 2014). electrochemical processes and membrane technology alone Improving the performance of the cadmium biosorption in or in combination have been promoted for the purpose of terms of the process efficiency can be achieved through the the heavy metal removal (Du et al. 2011; Liu et al. 2009). optimization, which provides an idea about the best opera- However, many of these processes increase the risk of gen- tional condition to yield the best possible response (adsorp- erating secondary pollutants, and therefore pose a concern tion) (Witek-Krowiak et al. 2011). Apart from the adsorp- from ecosystem health and public health viewpoints. As a tion kinetics which forms the basis for the selection of the consequence, the use of biological materials for the removal conditions, the modeling of the adsorption process, includ- of these heavy metals is being promoted to minimize the ing those of heavy metals such as cadmium, is carried out secondary pollutant level in the system as well as cost-effect to portray the changes in the efficiency depending on the benefit (Gifford et al. 2006). Application of the biological various physico-chemical parameters (Çelekli and Geyik materials that are component of the ecosystem reduces the 2011; Çelekli et al. 2013). Since the biosorption process is possibility of yielding unwanted secondary pollutants. These linked with different variables in a non-linear manner and are substantiated in the observations on the metal adsorption the mechanism of the process is multifaceted, it is difficult ability of different microorganisms (Hetzer et al. 2006; Sari to model using conventional mathematical simulation. In and Tuzen 2009) and hydrophytes (Sinha et al. 2007) many recent past, artificial neural network (ANN) model has been of which are hyperaccumulator of heavy metals. Application used in describing problems in different fields of chemical of the aquatic animals in the removal of the heavy metals and environmental engineering (Çelekli et al. 2012; Hossain from the aquatic system has been tested as evident from the et al. 2015; Maghsoudi et al. 2015). In a generalized ANN, studies on living freshwater bivalve (Jana and Das 1997), the outputs provide an overview of the relative importance bivalve shell (Du et al. 2011; Liu et al. 2009; Pena-Rodri- of the input parameters (factors influencing the biosorp- guez et al. 2010) crab and arca shell biomass (Dahiya et al. tion process) that drive the sorption process (Çelekli and 2008). However, the unregulated use of the living forms of Geyik 2011; Çelekli et al. 2013; Ahmad et al. 2014; Witek- the biological resources may facilitate species invasion, such Krowiak et al. 2014). In the present instance, the applica- as many of the plant hyperaccumulators (Sinha et al. 2007). tion of the ANN was meant for deducing the efficacy of the In other instances, such as the case of the freshwater bivalve, B. bengalensis shell dust (BSD) as biosorbent of cadmium, despite the metal removal being high (Jana and Das 1997), under varied physico-chemical conditions that influence the their potential as aquaculture resources limits their use in adsorption process. Thus, the cadmium absorption efficiency bioremediation. Recent studies have demonstrated that cal- of BSD was judged through multilayer perceptron in ANN cium carbonate and its derivatives may act as biosorbents model, apart from description of the equilibrium and kinetic of different heavy metals (Du et al. 2011; Pena-Rodriguez models. The results will be useful in supplementing the et al. 2010). The shells of different molluscs are composed of required information in utilizing the BSD, a waste generated 95–99% calcite and/or aragonite (principally made up of cal- from aquaculture as a biosorbent and thus in bioremediation cium carbonate) oriented in a matrix of proteins (0.1–5%), of heavy metal contaminated aquatic ecosystem. and act as natural ceramic with excellent strength and tough- ness (Kaplan 1998; Boro et al. 2012). The shells of aquatic gastropods can be considered as cheap and available source Materials and methods of calcium carbonate (Hossain and Aditya 2013) and thus can be considered as low-cost biosorbent of heavy met- Preparation of the metal solution als. In compliance with this proposition, the present study was aimed at evaluation of cadmium removal capacity of The experiments were initiated through preparation of a cad- −1 shell dust of freshwater gastropod Bellamya bengalensis mium chloride stock solution of 1000 mg L using double 1 3 Applied Water Science (2018) 8:68 Page 3 of 14 68 distilled water and further working solutions were prepared where q is the amount of metal adsorbed, v is the volume by appropriate dilution. The pH of the solution was adjusted of solution, m is the mass of adsorbent, c is the initial con- by adding HNO (0.1 N) and NaOH (0.1 N) as required in centration of the Cd solution and c is the equilibrium con- 3 e course of the experiments. All the inorganic chemicals that centration of the Cd solution. have been used in these experiments were purchased from Merck India Ltd., India. Estimation of the metal content in tissues and shells by AAS Preparation of the Bellamya shell dust For estimation of metal, required amount of Bellamya ben- The waste shells of the snail B. bengalensis were collected galensis tissue/shell of a particular dose was taken into Tef- from local fish markets in Burdwan, India. Following pro- lon container for microwave digestion. 4 mL of aqua regia curement, the remnant tissue portion and dusts in the shells was added and container was placed in microwave oven at were cleared in warm water and dried in sunlight for 2 days. 450  W and the sample was digested for 7  min. 4  mL of Next, the shells were placed in hot air oven for 2 days at 60 °C hydrogen peroxide was then added to the mixture and was and kept in plastic zipper bags. For preparation of the shell again digested for another 7 min. It was diluted with dis- dust, the shells were pulverized in mortar and pestle to fine tilled water, and filtered in a 25-mL volumetric flask. Final granules and dried in oven for 2 days at 60 °C. The granules volume was made up with distilled water. Reagent blank were then sieved through 500 µm and consequently through was prepared in the same way. The final samples were then 200 µm net, to yield two different sized granules, respectively, estimated for cadmium by atomic absorption spectroscopy 500–200 µm and ≥ 200 µm. Initially, the larger-sized granules (GBC Avanta 1.3, India) at 228.8 nm wavelength. adsorbed less amount of cadmium and thus were not studied further. The ash content of the dust was 72.34% and the water Analyses of FT‑IR absorbance spectra of BSD −1 content capacity was 1859.0 mg g . The specific surface 2+ area, pore volume and pore diameter were determined by the IR spectra of protonated or Cd loaded BSDs were recorded Brunauer–Emmett–Teller (BET) method using Quantachrome in a (Perkin-Elmer FTIR, Model RX1) Fourier transform Autosorb automate with nitrogen gas (version 4.0). The sur- infrared (FT-IR) spectroscopy. Samples of 100 mg of KBr 2 −1 face area of the biosorbent was 10.143 m  g , while the pore disks contain 1% of the finally ground powder of each sam- −1 volume was 0.079 cc g and the pore diameter was 3.336 nm. ple were prepared less than 24 h before recording. Batch sorption procedure The scanning electron micrograph (SEM) study of BSD The batch sorption experiments were performed in a 250-mL Erlenmeyer’s flask that contained 100 mL solution of the Raw and metal-adsorbed BSDs were dried and prepared for particular cadmium ion concentration at required pH and scanning electron microscopic studies. The samples were relevant amount of snail shell dust (BSD). The flasks were attached with the stubs by both side cello tapes and gold sealed with wax paper and shaken in a shaking incubator plated in a sputter coater before use in the SEM. Electron (Lab Companion, SI-300R, India) at a required temperature acceleration potential of 20 kV was used for the microscopic at 150 rpm for required time. After shaking for particular observations. Photographs were taken in a HITACHI S530 time, the solution of the flasks was centrifuged at 2000 rpm scanning electron microscope at 800× magnification. for 15 min and the supernatant was taken for estimation of metal concentration by atomic absorption spectroscopy Energy dispersive X‑ray (EDX) analysis (GBC Avanta 1.3, India). The influence of different levels of pH on biosorption equilibrium was studied through chang- For energy dispersive X-ray analysis, the samples were stud- ing the pH of the solution in range of 2–7. The effect of ied by field emission gun-based scanning electron micro- contact times between solution and the BSD were monitored scope with energy dispersive X-ray analysis (SEM/EDX) by varying it from 10 to 80 min. For equilibrium studies, the (Quanta model by FEI Co., the Netherlands) for the morpho- metal ion concentrations were used in a varied concentration logical as well as for the presence of elemental information −1 of 25–1000 mg L , and for optimum biosorption study, the on the samples. The SEM studies were performed at 30 kV BSD biomass was varied between 200 and 1000 mg. The in the low vacuum mode. EDX spectra were taken from the amount of Cadmium ion adsorbed on the BSD was estimated area corresponding to the SEM image shown in the in-set. following the equation (Sharma et al. 2011): The EDX spectra were taken in the ‘region mode’ with the bombardment of energetic electrons for duration of electron q = c − c v∕m, e o e of 100 s. 1 3 68 Page 4 of 14 Applied Water Science (2018) 8:68 Equilibrium modeling ln q = ln x −  , e m where β and x (maximum Cd adsorption capacity) are the The adsorption equilibriums were studied for the estima- m constants obtained from the plot of ln q vs. ɛ (Polanyi tion of maximum cadmium biosorption by the BSD. For the e potential). ‘ɛ’ is calculated from the following equation: equilibrium study, the experiments were performed at dif- −1 ferent initial cadmium ion concentration (25–1000 mg g ). = RT ln 1 + 1∕c , Langmuir and Freundlich isotherms were used in describ- −1 where R is the universal gas constant (J mol ) and T is the ing the equilibrium between adsorbed cadmium ions on the absolute temperature in Kelvin and ‘β’ is associated to the BSD (q ) and in solution (c ) at a particular temperature. The e e adsorption free energy (E). Adsorption energy is calculated parameters of the Langmuir equation (Langmuir 1918) were with the following equation (Tanzifi et al. 2017): determined from a linear form of the following equation: E = 1∕− 2. c ∕q = c ∕a + 1∕ab, e e e where a is the maximum amount of metal ions/unit mass of Kinetic modeling adsorbent to form a monolayer and b is the equilibrium con- stant. The correlation coefficient of the linear plot of c /q e e The pseudo second order model for prediction of biosorption against c indicates the applicability of the Langmuir mod- is not suitable for a long period of adsorption process (Ho eling in the experiment (Tan and Hameed 2017). and McKay 1999) and the pseudo second order equation: Freundlich equation (Freundlich 1906) is described as below: dq ∕dt = k q −q , t 2 e t 1∕n q = k c , or e f q = 1∕ k q + t∕q , where k and 1/n are the Freundlich constants indicating the t 2 e adsorption capacity and intensity, respectively. The linear where q is the amount of adsorbed metal ion on biosorbent model of the isotherm can be expressed logarithmically as at equilibrium, q is the amount of adsorbed metal ion on in the following: biosorbent at time ‘t’, and k is the second order rate constant −1 −1 (g mg  min ). A linear plot of t/q vs. t indicates whether log q = 1∕n log c + log K , e e F this model of biosorption is applicable for this case. where the value of K and n can be determined from the y-intercept and slope of the plot of log c against log q . The e e Artificial neural network (ANN) model linear plot of log q against log c indicates the applicability e e of the Freundlich modeling in the experiment. ANN is a non-linear mathematical model that is inspired by Temkin’s isotherm model is an empirical equation com- the structural and functional aspects of neuron in explora- prises a factor that clearly indicates the adsorbent–adsorb- tion of a group of input data (through training and testing) ate interactions. Through overlooking the lower and higher as function of output data (including relative error) and to value of concentrations, it assumes that heat of adsorption envisage the performance of the given system (Cavas et al. of the interacting molecules in the layer would decrease lin- 2011). A three-layered ANN model is used in the analyses early rather than logarithmically. The model is given by the of the complex interactions in finding the pattern of experi- following equation (Foo and Hameed 2010; Tanzifi et al. mental data. Multilayer perceptron uses supervised learn- 2017; 2018; Ghaedi and Vafaei 2017): ing technique in relating output responses to the expected q = RT∕b ln A c , e T T e responses by regulating the loads of the input signals and remind it. The memory is used in the next time of data feed or, when the output response will be much closer to the wanted q = RT∕b ln A + RT∕b ln c , e T T T e one after the same input signals are delivered (Shanmu- where A is the Temkin isotherm equilibrium binding con- gaprakash and Sivakumar 2013). The weight of any input −1 stant (L g ), b is the Temkin isotherm constant, R is the signal is calculated by the following equation: −1 −1 universal gas constant (8.314 J   mol  K), and T is the temperature at 298 K. A and B values are obtained from T T W = w x , i ij j q versus ln k plot for the adsorbent. e e j=1 Dubinin–Radushkevich (D–R) isotherm was used in investigation of Cd adsorption nature based on the equilib- rium data. The isotherm is expressed in linear form as: 1 3 Applied Water Science (2018) 8:68 Page 5 of 14 68 where w is the corresponding load of connection between Results and discussion ij each neuron ‘j’ in input layer and each neuron ‘i’ at hidden layer and x is the value of input signal ‘j’ at input layer Bioaccumulation of cadmium on shell dust (Çelekli et al. 2013). For a particular function, the output is predicted from the following equation: Eec ff t of pH of the solution on batch sorption y =  (w + b ), k i j The biosorption procedure was maintained over the experi- where w is the sum weight of each neuron and b is the bias, i j mental range of pH 2–7. The sorption procedure was affected i.e., constant weight of a neuron representing the relative by the pH of the medium in two ways—metal solubility and error (Çelekli and Geyik 2011). In predicting the biosorp- total charge of the functional groups of the biosorbent. The tion of cadmium (output), the combined effect of the vari- optimum pH, at which the procedure shows the highest 2+ ables—(1) pH, (2) Cd concentration, (3) biosorbent dose adsorption for the biosorbent was estimated to be 6 which and (4) contact time were used in the analysis. Sixty-nine is nearer to the environmental pH of most of the freshwater experimental sets were used in training, validation and test- aquatic system. At high pH, that is, at alkaline condition, ing the ANN model by SPSS 20, trial version. precipitation of the metal takes place, and at low pH due to high protonation, metal sorption capacity decreases (Guo et al. 2008). The experiment was carried out in 100 mL of −1 2+ solution having 100 mg L Cd and 100 mg of the BSD at 30 °C in reference to varying pH of the solution. Figure 1a indicates the pH-dependent adsorption of the metal ion by BSD as the metal sorption was negligible at pH 2 and it Fig. 1 a Effect of pH of the A B solution on biosorption of cad- 10 30 mium on BSD at 30 °C, initial cadmium ion concentration of −1 100 mg L , contact time of 8 60 min and biosorbent dose of 1 g per 100 mL solution. b Effect of initial cadmium ion concentration of solution on biosorption on BSD at 30 °C, pH 6, contact time of 60 min and biosorbent dose of 1 g per 100 mL solution. c Effect of time on biosorption of cadmium on BSD at 30 °C, pH 6, initial cadmium ion concentration 01234567 8 0200 400600 80010001200 −1 of 100 mg L , and biosorb- 2+ -1 pH Cd conc. in solution (mg L ) ent dose of 1 g per 100 mL C D solution. d Effect of biosorbent dose on biosorption of cadmium 10 10 on BSD at 30 °C, pH 6, initial cadmium ion concentration of −1 100 mg L , contact time of 60 min 0200 400600 80010001200 Adsorption time (min) Amount of BSD (mg/100ml) 1 3 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 2+ -1 Cd uptake by BSD (mg g ) 68 Page 6 of 14 Applied Water Science (2018) 8:68 increases dramatically as increase in pH and at pH 6 it shows spectrum of the BSD was performed. IR spectra of proto- the highest adsorption. The metal sorption declined as the nated and cadmium loaded BSD are shown in Fig. 2a, b, −1 pH increase further. In a particular pH range, most metal respectively. A peak at 3424 cm is indicating the presence −1 sorption is enhanced with pH, increasing to a certain value of hydroxyl (–OH) groups. Strong band at 2921 cm is due followed by a reduction on further pH increase (Guo et al. to C–H stretching frequency (Semerciöz et al. 2017) and −1 2008; Pavasant et al. 2006; Semerciöz et al. 2017). peak at 1634 cm is due to C=O stretching mode of the primary and secondary amides (Reddy et al. 2010). Weak −1 Eec ff t of initial metal ion concentration on batch sorption band at 1462 cm is attributed to aromatic C=C and the −1 strong band at 1041 cm is due to C–O stretching of alco- Increase in metal ion concentration added on metal adsorp- holic groups (Blazquez et al. 2011). FTIR study reveals that −1 tion as concentration of the ion increases. 25–1000 mg L OH, CH, CO, and C=C may be the responsible groups in the ion concentrations were used for the study taking seven biosorption process. The FTIR of metal loaded BSD shows −1 different doses in series. The 25 mg L initial metal ion that distinct shift of the above mentioned bands as well as concentration showed the lowest adsorption while the change in intensity informs some ion exchange behavior of −1 400 mg L concentration showed the highest adsorption the BSD. and the adsorption remained same in further increase in 2+ Cd ion increase in the solution (Fig. 1b). The initial metal The scanning electron micrograph (SEM) analysis ion concentration can modify the metal removal efficiency of the shell dust over a combination of factors, the availability of specific surface functional groups and ability of surface functional The surface structure of the free and cadmium-loaded BSD groups to bind metal ions (Taty-Costodes et al. 2003; Pino was analyzed under scanning electron microscope. The et al. 2006) scanning electron micrographs of the dried BSD before and 2+ after the Cd treatment at 800× magnification are shown Eec ff t of contact time on batch sorption in Fig.  3a, b, respectively. It indicates the irregular mor- phological structure of the particles and lamellar stratified The sorption potentials of the BSD over time were moni- surface of the BSD. The SEM image of the biosorbent after 2+ tored from 10 min to 20, 40, 60, 80 min using 100 mL of exposure to the Cd shows a spongy layer indicating surface −1 2+ 100 mg L Cd at pH 6 (Fig. 1c). At the beginning, metal precipitation occurred during the sorption (Du et al. 2011). adsorption was less due to more binding sites remained In case of cadmium sorption by CaCO compound gener- free when treated for the short period of time and increased ally follows surface precipitation due to similar ionic radii rapidly as the treatment time increases. It showed lowest (Prieto et al. 2003; Perez-Garrido et al. 2007) of divalent adsorption when treated for 10 min and increased over time calcium and cadmium. to saturate at 60  min and after that the uptake remained almost same. The variation in uptake of the cadmium ions Energy dispersive X‑ray (EDX) analysis with time was used in fitting the kinetic models. The surface structure of the shell dust makes accessible a Influence of the biosorbent dose on batch sorption large unadsorbed surface area for the cadmium ion of the solution. The elemental profile of the BSD before and after The cadmium biosorption potential of the BSD augmented the treatment of the cadmium solution was estimated using over its amount increased in treating the metal solution. the energy dispersive X-ray analysis. The projecting peaks The more the amount of biosorbent the more the free bind- in the EDX spectra correspond to CKα, OKα, AlKα, SiKα, ing sites or exchanging group to adsorb the metal ion from PKα, CaKα, CaKβ, FeKα, etc., in the untreated BSD and −1 the solution. For a 100 mg L metal ion concentration, the furthermore to it a CdLα peak in treated BSD (Fig. 4a, b). increase in biosorbent resulted increase in metal ion adsorp- Both the treated and untreated BSD display strong peaks tion and above a certain dose it remained same or slightly corresponding to calcium. In addition, the calcium of BSD higher due to comparatively higher number of free sites and may facilitate adsorption of cadmium because of the simi- lesser number of metal ions (Fig. 1d) (Al-Anber and Matouq larities in ionic radii that enhance ion exchange (Purkayastha 2008; Ghodbane et al. 2008). et al. 2014). FT‑IR study of the shell dust Equilibrium modeling To study the mechanism of cadmium removal and the main The biosorption isotherm is important in waste water treat- 2+ functional groups responsible for Cd binding, the FTIR ment as it implies estimation of biosorption capacity of 1 3 Applied Water Science (2018) 8:68 Page 7 of 14 68 Fig. 2 FTIR absorbance spectra of snail shell dust (BSD) before 13.3 (a) and after (b) biosorption of cadmium 864.28 618.01 1082.87 %T 1467.99 1618.13 1563.27 3420.87 2.0 4400.0 4000 3000 2000 1500 1000 400.0 cm-1 11.2 864.51 626.36 1078.90 2921.85 %T 1424.31 1563.27 1635.92 3428.08 1.0 4400.0 4000 3000 2000 1500 1000 400.0 cm-1 Fig. 3 Scanning electron micrograph (SEM) of snail shell dust (BSD) before (a) and after (b) the biosorption of cadmium (magnification at ×800) 1 3 68 Page 8 of 14 Applied Water Science (2018) 8:68 Fig. 4 EDX analyses for elemental composition in the snail shell dust for BSD before (a) and after (b) biosorption of cadmium clearly showing the peaks for cadmium, in the treated shell dust, indicating the adsorption of the metal on the shell dust the adsorbent. The linear representations of Langmuir and cadmium biosorption support that under optimum condi- 2+ Freundlich isotherm of cadmium adsorption at 30 °C are tions (pH 6, biosorbent dose of 1 g, Cd concentration of −1 −1 given in Fig. 5a, b, respectively. The correlation coefficient 100 mg L and 60 min time period) 30.33 mg g is the and constants obtained from the equations are presented in maximum biosorption capacity of BSD. The Temkin iso- Table 1. The correlation of determination is high in Lang- therm shows B = 533, T = 1.15 and the R = 0.896 showing T A 2 2 muir equation (R = 0.995) contrast to Freundlich equation moderate t fi . The R–D equation ( R = 0.766) for the biosorp- (R = 0.760) (Fig. 5). It indicates that Langmuir model is tion shows the value of the adsorption energy (E) is 13.89 J/ more suitable for describing the biosorption equilibrium of mol that corresponds to chemisorption type of uptake based 2+ −1 Cd on the snail shell dust. High q value (30.33 mg g ) on ion exchange (Wu et al. 2012; Markou et al. 2016). Maxi- max from Langmuir equation indicates the biosorption potential mum cadmium biosorption capacities of similar low cost of the material. The value of b (0.04) (Table 1) indicates b biosorbents are shown in Table 2. BSD is mainly composed the affinity of the binding sites and the energy of adsorption of calcium carbonate, degradation of which, if any, will not (L/mg) (Blazquez et al. 2011). The equilibrium models of yield unwanted compound to the ecosystem. 1 3 Applied Water Science (2018) 8:68 Page 9 of 14 68 Fig. 5 a Langmuir isotherm, b AB Freundlich isotherm, c Temkin 40 2 isotherm and d Dubinin–Radu- shkevich isotherm plot for y = 0.0338x + 0.6614 1.5 biosorption of cadmium on R² = 0.9954 BSD (pH 6, temperature 30 °C, biosorbent = 1 g, cadmium ion −1 y = 0.3747x + 0.5136 concentration = 100 mg g ) 0.5 R² = 0.7608 0200 400600 800 e log c CD 40 4 y = 4.6404x + 0.6756 R² = 0.8946 y = -17.489x + 2.9892 R² = 0.7663 00.050.1 0.15 0246 8 [ln(1+1/ Ce)] ln c Table 1 Coefficients of the Langmuir, Freundlich, Temkin isotherm and Dubinin–Radushkevich models for cadmium biosorption by BSD Langmuir coefficients Freundlich coefficients Temkin coefficients D–R coefficients −1 −1 2 2 −1 2 −1 2 q (mg g ) b (L mg ) R n k R T (L mg ) b R q (mg g ) E R max F A T max 30.33 0.049 0.995 2.67 1.67 0.760 1.15 533 0.894 9.211 13.89 0.766 −1 adsorption (9.26 mg g ) is much nearer to the expected Kinetic modeling −1 value (9.302 mg g ). These suggest that the biosorption process is based on the pseudo second order model. The kinetic model is necessary for determination of opti- mal condition of the biosorption process. For the evalua- Optimization results through ANN model tion of differences in sorption process, the kinetics of metal uptake were described by pseudo second order model (Ho The ANN model established in this study comprised three and McKay 1999). The linear plots obtained from pseudo −1 2+ layers as shown in Fig. 7a consisting of a hyperbolic tan- second order model at 100 mg L initial Cd concentra- gent transfer function at hidden layer and a linear transfer tion, pH 6 and at studied temperature are shown in Fig. 6a, function at output layer. The input layers have four predic- b, respectively. The rate constants, expected metal uptake tors, viz. biomass, pH, Cd concentration and time, the hid- and correlation coefficients have been described in Table  3. den layer and the output layer have one response variable, The pseudo second order reaction in biosorption is based the amount of Cd absorbed(Table 4) (Çelekli et al. 2012, on the sorption capacity on the solid phase. From Fig. 6b 2016). The weight of the neurons was used in the study and Table 3, the correlation coefficient in the pseudo second of the relative influence of each of the input variables on order reaction (R = 0.998) is high and the calculated metal 1 3 e c /q e e log q ln q e 68 Page 10 of 14 Applied Water Science (2018) 8:68 Table 2 Comparative data of biosorption capacities (q —maximum 10 max metal uptake capacity) for cadmium by different biosorbents −1 Adsorbent q (mg g ) References max Raw corn stalk 3.39 Zheng et al. (2010) Olive waste 6.55 Azouaou et al. (2008) y = 0.108x + 0.1836 Brewer’s yeast 10.17 Cui et al. (2010) 2 R² = 0.999 Corncob 4.73 Ramos et al. (2005) Rice straw 13.9 Ding et al. (2012) Wheat bran 15.71 Nouri et al. (2007) Time (min) Castor seed hull 6.98 Sen et al. (2010) Bamboo charcoal 12.08 Wang et al. (2010) Fig. 6 Pseudo second order plot for biosorption of cadmium on BSD Walnut tree sawdust 5.76 Yasemin and Zek (2007) (pH 6, temperature 30 °C, biosorbent = 1 g, cadmium ion concentra- −1 Chitosan/bentonite 12.05 Arvand and Pakseresht tion = 100 mg g ) (2012) Coconut copra meal 4.99 Ho and Ofomaja (2006) Table 3 Rate constant and equilibrium uptake for cadmium binding S. platensis 73.64 Çelekli and Bozkurt −1 by BSD at pH 6 and initial metal ion concentration of 100 mg L (2011) Metal Second order model Sweet potato 18.00 Asuquo and Martin (2016) −1 −1 −1 2 P. pubescens biochars 14.70 Zhang et al. (2017) k (g mg  min ) q (mg g ) R 2 e cal. Bellamya shell dust 30.33 The present study Cadmium 0.064 9.26 0.998 (BSD) the biosorption process, i.e., on output layer. In the present Conclusion instance, the sum of squares error (SSE error, a measure of the predicted and the observed values; Çelekli and Geyik The biosorption process works as function of pH of the solu- 2011) is 0.467; and the relative error being 0.056, justify- tion, biomass, Cd ion concentration and time of contact. ing the suitability of the multilayer perceptron ANN model. The biosorption study indicates that at pH 6, the maximum −1 The regression equation of the output (predicted adsorp- cadmium adsorption capacity of BSD is 30.33 mg g . The tion through ANN model, y) and the corresponding target isotherm model follows Langmuir model (R = 0.995) better 2 2 (observed through the experiments, x) complied with the lin- than Temkin isotherm (R = 0.896), R–D (R = 0.766) and ear form as y = 0.37 + 0.95x (Fig. 7b). The high (R = 0.943) Freundlich model (R = 0.760). The main functional group coefficient of determination fits well with the experimental responsible for chelation is OH, C=O, C=C and C–C, as dataset with that of the ANN model of the system (Yang supported by FTIR analysis. EDX study shows clear peaks et al. 2011; Khataee et al. 2011). Upon comparison, the data for cadmium in the treated biomaterial, conveying further complied well with both the pseudo second order (Fig. 6) as support of clear adsorption. The biosorption process fol- well as the ANN model (Fig. 7b), with high values of coef- lowed pseudo second order (R = 0.998) kinetics. A multi- ficient of determination. Among the predictors, the pH of the layer perceptron model in artificial neural network (ANN) solution (0.347) appeared to be comparable to the cadmium model successfully portrayed biosorption of cadmium ion concentration (0.339) in terms of the relative importance in on the LSD with high correlation coefficient (R = 0.943) shaping the adsorption process, similar to those observed between predicted and observed removal. BSD can be for the dye removal by Chara contraria (Çelekli and Geyik considered as efficient, low cost and environment friendly 2011). The contact time and the biomass bear comparative biosorbent for cadmium bioremediation, opening a new low influence in the adsorption, contrast to those observed aspect of the economic value of shells of the freshwater snail for the dye removal by the walnut husk (Çelekli et al. 2016). B. bengalensis, apart from its value as cheap protein source. 1 3 t/q Applied Water Science (2018) 8:68 Page 11 of 14 68 Fig. 7 Basic design (a) of the multilayer perceptron artificial neural network (ANN) model used in the study along with the regression line (b) endorsing the linearity of predicted and observed cadmium removal by BSD 1 3 68 Page 12 of 14 Applied Water Science (2018) 8:68 Table 4 Basic information of Network information the artificial neural network used in describing the Input layer Covariates biosorption of cadmium by  1 pH BSD  2 Cd concentration  3 Biomass  4 Time Number of units 4 Rescaling method for covariates Standardized Hidden layer(s) Number of hidden layers 1 Number of units in hidden layer 1 2 Activation function Hyperbolic tangent Output layer Dependent variables  1 Removal Number of units 1 Rescaling method for scale dependents Standardized Activation function Identity Error function Sum of squares Excluding the bias unit Azouaou N, Sadaoui Z, Mokaddem H (2008) Removal of cadmium Acknowledgements The authors are thankful to the respective Heads, from aqueous solution by adsorption on vegetable wastes. J Department of Zoology, Department of Chemistry, and Department of Appl Sci 8:4638–4643 Environmental Science, the University of Burdwan, Burdwan, West Bernard A (2008) Cadmium and its adverse effects on human health. Bengal, India for the facilities provided. We thankfully acknowledge Indian J Med Res 128:557–564 the anonymous reviewers for their thoughtful comments and kind Blazquez G, Martin-Lara MA, Tenorio G, Calero M (2011) Batch suggestions to enrich this manuscript. We express our gratitude and biosorption of lead(II) from aqueous solutions by olive tree regards to Prof. Enrico Drioli for his kind cooperation in revising the earlier version of the manuscript. AH thankfully acknowledges pruning waste: equilibrium, kinetics and thermodynamic study. the financial assistance provided Council of Scientific and Industrial Chem Eng J 168:170–177 Research (CSIR), New Delhi, India. Boro J, Deka D, Thakur AJ (2012) A review on solid oxide derived from waste shells as catalyst for biodiesel production. Renew Sustain Energy Rev 16:904–910 Open Access This article is distributed under the terms of the Crea- Brooks BW, Stanley JK, White JC, Turner PK, Wu KB, La Point TW tive Commons Attribution 4.0 International License (http://creat iveco (2004) Laboratory and field responses to cadmium: an experi- mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- mental study in effluent-dominated stream mesocosms. Environ tion, and reproduction in any medium, provided you give appropriate Toxicol Chem 23:1057–1064 credit to the original author(s) and the source, provide a link to the Cavas L, Karabay Z, Alyuruk H, Dogan H, Demir GK (2011) Thomas Creative Commons license, and indicate if changes were made. and artificial neural network models for the fixed bed adsorption of methylene blue by a beach waste, Posidonia oceanea (L.) dead leaves. Chem Eng J 171(2):557–562 Çelekli A, Bozkurt H (2011) Bio-sorption of cadmium and nickel ions using Spirulina platensis: kinetic and equilibrium studies. References Desalination 275:141–147 Çelekli A, Geyik F (2011) Artificial neural networks (ANN) Ahmad MF, Haydar S, Bhatti AA, Bari AJ (2014) Application of arti- approach for modeling of removal of Lanaset Red G on Char- ficial neural network for the prediction of biosorption capacity of acontraria. Bioresour Technol 102:5634–5638 immobilized Bacillus subtilis for the removal of cadmium ions Çelekli A, Birecikligil SS, Geyik F, Bozkurt H (2012) Prediction from aqueous solution. Biochem Eng J 84:83–90 of removal efficiency of Lanaset Red G on walnut husk using Al-Anber ZA, Matouq MAD (2008) Batch adsorption of cadmium ions artificial neural network model. Bioresour Technol 103:64–70 from aqueous solution by means of olive cake. J Hazard Mater Çelekli A, Bozkurt H, Geyik F (2013) Use of artificial neural network 151:194–201 and genetic algorithms for prediction of sorption of an azo–metal Arvand M, Pakseresht MA (2012) Cadmium adsorption on modified complex dye onto lentil straw. Bioresour Technol 129:396–401 chitosan coated bentonite: batch experimental studies. J Chem Çelekli A, Bozkurt H, Geyik F (2016) Artificial neural network and Technol Biotechnol 88(4):572–578 genetic algorithms for modeling of removal of an azo dye on wal- Asuquo ED, Martin AD (2016) Sorption of cadmium (II) ion from nut husk. Desalin Water Treat 57:15580–15591 aqueous solution onto sweet potato (Ipomoea batatas L.) peel Cui L, Wu G, Jeong T (2010) Adsorption performance of nickel and adsorbent: characterisation, kinetic and isotherm studies. J Envi- cadmium ions onto brewer’s yeast. Can J Chem Eng 88:109–115 ron Chem Eng 4:4207–4228 1 3 Applied Water Science (2018) 8:68 Page 13 of 14 68 Dahiya S, Tripathi RM, Hegde AG (2008) Biosorption of lead and Ca(OH) -pretreated biomass of Phragmites sp. J Environ Sci copper from aqueous solutions by pre-treated crab and arca shell 45(49–5):9 biomass. Bioresour Technol 99:179–187 Nouri L, Ghodbane I, Hamdaoui O, Chiha M (2007) Batch sorption Ding Y, Jing D, Gong H, Zhou L, Yang X (2012) Biosorption of dynamics and equilibrium for the removal of cadmium ions from aquatic cadmium (II) by unmodified rice straw. Bioresour Tech- aqueous phase using wheat bran. J Hazard Mater 149:115–125 nol 114:20–25 Pavasant P, Apiratikul R, Sungkhum V, Suthiparinyanont P, Wattan- 2+ 2+ 2+ Du Y, Lian F, Zhu L (2011) Biosorption of divalent Pb, Cd, Zn on arag- achira S, Marhaba TF (2006) Biosorption of Cu, Cd, Pb , 2+ onite and calcite mollusk shells. Environ Pollut 159:1763–1768and Zn using dried marine green macroalga Caulerpa lentillif- Foo KY, Hameed BH (2010) Insights into the modeling of adsorption era. Bioresour Technol 97(18):2321–2329 isotherm systems. Chem Eng J 156:2–10 Pena-Rodriguez S, Fernandez-Calvino D, Novoa-Munoz JC, Nunez- Freundlich HMF (1906) Uber die adsorption in losungen. Z Phys Chem Delgado A, Fernandez-Sanjurjo MJ, Alvarez-Rodriguez A (2010) 57A:385–470 Kinetics of Hg(II) adsorption and desorption in calcined mussel Ghaedi AM, Vafaei A (2017) Application of artificial neural networks shells. J Hazard Mater 180:622–627 for adsorption removal of dyes from aqueous solution: a review. Perez-Garrido C, Fernandez-Diaz L, Pina CM, Prieto M (2007) In situ Adv Colloid Interface Sci 245:20–39 AFM observations of the interaction between calcite surfaces and Ghodbane I, Nouri L, Hamdaoui O, Chiha M (2008) Kinetic and equi- Cd-bearing aqueous solutions. Surf Sci 601:5499–5509 librium study for the sorption of cadmium (II) ions from aqueous Pino GH, deMesquita LMS, Torem ML, Pinto GAS (2006) Biosorp- phase by eucalyptus bark. J Hazard Mater 152:148–158 tion of cadmium by green coconut shell powder. Miner Eng Gifford S, Dunstan RH, O’Connor W, Koller CE, MacFarlane GR 19:380–387 (2006) Aquatic zooremediation: deploying animals to remediate Prieto M, Cubillas P, Fernandez-Gonzalez A (2003) Uptake of dis- contaminated aquatic environment. Trends Biotechnol 25:60–65 solved Cd by biogenic and abiogenic aragonite: a comparison with Godt J, Scheidig F, Grosse-Siestrup C, Esche V, Brandenburg P, Reich sorption onto calcite. Geochim Cosmochim Acta 67:3859–3869 A, Groneberg DA (2006) The toxicity of cadmium and resulting Purkayastha D, Mishra U, Biswas S (2014) A comprehensive review hazards for human health. J Occup Med Toxicol 1(1):22 on Cd(II) removal from aqueous solution. J Water Process Eng Guo X, Zhang S, Shan XQ (2008) Adsorption of metal ions on lignin. 2:105–128 J Hazard Mater 151(1):134–142 Ramos RL, Jacome LAB, Rodriguez IA (2005) Adsorption of cadmium Hetzer A, Daughney CJ, Morgan HW (2006) Cadmium ion biosorption (II) from aqueous solution on natural and oxidized corncob. Sep by the thermophilic bacteria Geobacillus stearothermophilus and Purif Technol 45:41–49 G. thermocatenulatus. Appl Environ Microbiol 72:4020–4027 Reddy HKD, Seshaiah K, Reddy AVR, Rao MM, Wang MC (2010) 2+ Ho YS, McKay G (1999) Pseudo-second order model for sorption pro- Biosorption of Pb from aqueous solutions by Moringa oleif- cess. Process Biochem 34:451–465 era bark: equilibrium and kinetic studies. J Hazard Mater Ho YS, Ofomaja AE (2006) Biosorption thermodynamics of cadmium 174:831–838 on coconut copra meal as biosorbent. Biochem Eng J 30:117–123 Sari A, Tuzen M (2009) Kinetic and equilibrium studies of biosorp- Hossain A, Aditya G (2013) Cadmium biosorption potential of shell tion of Pb(II) and Cd(II) from aqueous solution by macrofungus dust of the fresh water invasive snail Physa acuta. J Environ Chem (Amanita rubescens) biomass. J Hazard Mater 164:1004–1011 Eng 1:574–580 Semerciöz AS, Göğüş F, Çelekli A, Bozkurt H (2017) Development Hossain A, Bhattacharyya SR, Aditya G (2015) Biosorption of cad- of carbonaceous material from grapefruit peel with microwave mium from aqueous solution by shell dust of the freshwater snail implemented-low temperature hydrothermal carbonization tech- Lymnaealuteola. Environ Technol Innov 4:82–91 nique for the adsorption of Cu (II). J Clean Prod 65:599–610 Jana BB, Das S (1997) Potential of freshwater mussel (Lamellidens Sen TK, Mohammod M, Maitra S, Dutta BK (2010) Removal of Cad- marginalis) for cadmium clearance in a model system. Ecol Eng mium from aqueous solution using castor seed hull: a kinetic and 8:179–193 equilibrium study. Clean Soil Air Water 38:850–858 Johri N, Jacquillet G, Unwin R (2010) Heavy metal poisoning: the Shanmugaprakash M, Sivakumar V (2013) Development of experi- effects of cadmium on the kidney. Biometals 23:783–792 mental design approach and ANN-based models for determination Kaplan DL (1998) Mollusc shell structure: novel design strategies for of Cr(VI) ions uptake rate from aqueous solution onto the solid synthetic materials. Curr Opin Solid State Mater Sci 3:232–236 biodiesel waste residue. Bioresour Technol 148:550–559 Khan RA, Choudhury S (1984) The population and production ecol- Sharma M, Kaushik A, Kaushik CP (2011) Biosorption of reactive ogy of a freshwater snail, Bellamya bengalensis (Lamarck) (Gas- dye by waste biomass of Nostoc lincia. Ecol Eng 37:1589–1594 tropoda: Viviparidae) in an artificial lake of Calcutta, India. Bull Sinha RK, Herat S, Tandon PK (2007) Phytoremediation: role of plants Zool Surv India 5:59–76 in contaminated site management. In: Singh N, Tripathi RD (eds) Khataee AR, Dehghan G, Zarei M, Ebadia E, Pourhassan M (2011) Environ. Biorem. Technol. Springer, New York, pp 315–330 Neural network modelling of biotreatment of triphenylmeth- Tan KL, Hameed BH (2017) Insight into the adsorption kinetics mod- ane dye solution by a green macroalgae. Chem Eng Res Des els for the removal of contaminants from aqueous solutions. J 89:172–178 Taiwan Inst Chem Eng 74:25–48 Langmuir I (1918) The adsorption of gases on plane surface of glass, Tanzifi M, Hossein SH, Kiadehi AD, Olazar M, Karimipour K, mica and platinum. J Am Chem Soc 40:1361–1403 Rezaiemehr R, Ali I (2017) Artificial neural network optimiza- Liu Y, Sun C, Xu J, Li Y (2009) The use of raw and acid pretreated tion for methyl orange adsorption onto polyaniline nano-adsor- bivalve mollusk to remove metals from aqueous solutions. J Haz- bent: kinetic, isotherm and thermodynamic studies. J Mol Liq ard Mater 168:156–162 244:189–200 Maghsoudi M, Ghaedi M, Zinali A, Ghaedi AM, Habibi MH (2015) Tanzifi M, Yaraki MT, Kiadehi AD, Hosseini SH, Olazar M, Bhati Artificial neural network (ANN) method for modeling of sunset AK, Agarwal S, Gupta VK, Kazemi A (2018) Adsorption of yellow dye adsorption using zinc oxide nanorods loaded on acti- Amido Black 10B from aqueous solution using polyaniline/SiO vated carbon: kinetic and isotherm study. Spectrochim Acta A nanocomposite: experimental investigation and artificial neural Mol Biomol Spectrosc 134:1–9 network modeling. J Colloid Interface Sci 510:246–261 Markou G, Mitrogiannis D, Muylaert K, Çelekli A, Bozkurt H Tao T, Yuan Z, XiaonaH Wei M (2012) Distribution and bioaccumu- (2016) Biosorption and retention of orthophosphate onto lation of heavy metals in aquatic organisms of different trophic 1 3 68 Page 14 of 14 Applied Water Science (2018) 8:68 levels and potential health risk assessment from Taihu lake, China. Yang Y, Wang G, Wang B, Li Z, Jia X, Zhou Q, Zhao Y (2011) Ecotoxicol Environ Saf 81:55–64 Biosorption of Acid Black 172 and Congo Red from aqueous Taty-Costodes VC, Fauduet H, Porte C, Delacroix A (2003) Removal solution by non-viable Penicillium YW 01: kinetic study, equilib- of Cd(II) and Pb(II) ions, from aqueous solutions, by adsorption rium isotherm and artificial neural network modelling. Bioresour onto sawdust of Pinus sylvestris. J Hazard Mater 105:121–142 Technol 102:828–834 Wang FY, Wang H, Ma JW (2010) Adsorption of cadmium (II) ions Yasemin B, Zek T (2007) Removal of heavy metals from aqueous solu- from aqueous solution by a new low-cost adsorbent—bamboo tion by sawdust adsorption. J Environ Sci 19:160–166 charcoal adsorption of cadmium (II) ions from aqueous solution Zhang C, Shan B, Tang W, Zhu Y (2017) Comparison of cadmium by a new low-cost adsorbent—bamboo charcoal. J Hazard Mater and lead sorption by Phyllostachys pubescens biochar produced 177:300–306 under a low-oxygen pyrolysis atmosphere. Bioresour Technol Witek-Krowiak A, Szafran RG, Modelski S (2011) Biosorption of 238:352–360 heavy metals from aqueous solutions onto peanut shell as a low- Zheng L, Dang L, Yi X, Zhang H (2010) Equilibrium and kinetic stud- cost biosorbent. Desalination 265:126–134 ies of adsorption of Cd(II) from aqueous solution using modified Witek-Krowiak A, Chojnacka K, Podstawczyk D, Dawiec A, Pokom- corn stalk. J Hazard Mater 176:650–656 eda K (2014) Application of response surface methodology and artificial neural network methods in modelling and optimization Publisher’s Note Springer Nature remains neutral with regard to of biosorption process. Bioresour Technol 160:150–160 jurisdictional claims in published maps and institutional affiliations. Wu Y, Wen Y, Zhou J, Dai Q, Wu Y (2012) The characteristics of waste Saccharomyces cerevisiae biosorption of arsenic(III). Environ Sci Pollut Res 19:3371–3379 1 3

Journal

Applied Water ScienceSpringer Journals

Published: Apr 23, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off