Subscriber access provided by - Access paid by the | UCSB Libraries
Separations
Analysis of breakthrough behaviours of hydrophilic and hydrophobic pharmaceuticals in novel clay composite adsorbent column in the presence and absence of biofilm Arya Vijayanandan, Ligy Philip, and S. Murty Bhallamudi Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.8b00987 • Publication Date (Web): 15 Jun 2018 Downloaded from http://pubs.acs.org on June 22, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
Analysis of breakthrough behaviours of hydrophilic and hydrophobic pharmaceuticals in novel clay composite adsorbent column in the presence and absence of biofilm Arya Vijayanandan, Ligy Philip*, S. Murty Bhallamudi * Corresponding Author Department of Civil Engineering, Indian Institute of Technology Madras, India - 600036, E-mail:
[email protected]; Phone No: +91-44-22574274; Fax No: +91-44-22574252 Abstract Present study investigated the use of novel clay composite adsorbent in simultaneous removal of hydrophilic and hydrophobic pharmaceuticals in fixed bed column. Potential of biologically active clay composite adsorbent in removing the pharmaceuticals was examined in detail. Mechanism of adsorption was elucidated based on equilibrium sorption and mass transfer approach. Effects of dispersion, mass transfer zone, empty bed contact time and interfering substance such as humic acid on column operation were investigated in detail. It was observed that adsorption was the dominating mechanism of removal in biologically active adsorbent column and the amount of biodegradation gradually increased with increase in contact time. Breakthrough behaviors of pharmaceuticals were numerically simulated using equilibrium sorption approach as well as mass transfer approach. Although both equilibrium sorption model (EQM) and linear driving force (LDF) model predicted breakthrough behaviors satisfactorily, tailing of the breakthrough curve was better predicted by LDF model. Based on LDF model, surface diffusion coefficients for atenolol, ciprofloxacin and gemfibrozil were estimated to be 6.5×10-4, 9.4×10-4 and 1.2×10-3 cm/h, respectively. Keywords: LDF model, EQM model, Pharmaceuticals, Adsorption, Clay composite adsorbent, Biologically active adsorbent
1
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1. Introduction The presence of pharmaceuticals in water sources poses a threat to the aquatic life.1 Even though there are no clear evidences of human toxicity, cumulative exposure to these compounds even at low concentrations is of significant concern. Pharmaceuticals are getting released into the environment through human and livestock excreta and, effluents from hospitals and drug manufacturing units. There are evidences of groundwater contamination by pharmaceutically active compounds (PhACs) released from effluents of pharmaceutical industries.2,3 Municipal wastewater and pharmaceutical industry effluents should be treated in such a manner that the entry of pharmaceuticals into the water bodies through treated effluents is minimized. Adsorption is a widely used treatment technology for pollutant removal as it is less energy intensive and easy to implement. Even though activated carbon is commonly used adsorbent, recent studies have focused on developing low cost clay based adsorbents as the selectivity of these adsorbents could be enhanced by suitable modifications in the surface chemistry. Modified forms of clay include pillared clay, organo clay, magnetic clay composite and polymer clay composites. Several studies have reported the removal of pharmaceuticals using clay and its modified forms.4–8 Most of these studies were carried out in batch scale. In order to scale up and design an adsorption column, column experiments need to be conducted to obtain the necessary fixed bed operating parameters. Detailed removal mechanisms and mass transfer limitations can be identified only by closely analyzing the breakthrough curves. Studies focusing on the analysis of breakthrough behavior of PhACs with the help of models are very scanty. Sotelo et al.9 studied the fixed bed sorption of caffeine on sepiolite clay and modelled the breakthrough behaviour using BDST model. Cabrera-Lafaurie et al.10 conducted column studies on the adsorption of salicylic acid, clofibric acid, carbamazepine and caffeine on transition metal modified inorganic– organic pillared clays and fixed bed operation was modelled using Yoon–Nelson, Clark, and the modified dose-response models. Álvarez-Torrellas et al.11 compared the performance of three adsorbents including sepiolite clay on the breakthrough behaviour of caffeine and experimental data were modelled using Thomas model and Yoon Nelson model. Thiebault et al.12 evaluated the removal of 14 pharmaceutically active compounds in clay-based filter and modelled the 2
ACS Paragon Plus Environment
Page 2 of 30
Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
experimental results using Lagergren first order model, second order model and Bangham model. It has been observed that models based on mass transfer approaches could give better predictions compared to other empirical models as the shape of the breakthrough curve is predominantly controlled by intra particle mass transfer and dispersion.13 Adsorption column performance can be improved by biofilm formation on the adsorbent. These biologically active adsorbents offer increased removal of pollutants due to coupled effect of adsorption and biodegradation. Not many studies are available on the use of biologically active adsorbent for pharmaceutical removal. A few studies have reported the removal of pharmaceuticals in biological activated carbon (BAC) filter. Paredes et al.14 compared the removal of 18 micropollutants on biofilm coated sand and GAC. Reungoat et al.15 studied the removal of micropollutants in waste water treatment plant (WWTP) effluent using BAC filter. Rivera-Utrilla et al.16 conducted the effect of bio-adsorption on AC in removing three tetracycline compounds. It has been shown that BAC can be useful in removing even the recalcitrant organic compounds, by employing properly acclimatized organisms. Extensive studies on the performance of a biologically active adsorbent column are still lacking. The efficiency of biologically active adsorbent in removing the pharmaceuticals can be understood in detail by conducting fixed bed column studies to interpret the mass transfer mechanism and influence of parameters controlling the performance of the column operation. Even though earlier studies have modeled the adsorption of pollutants in fixed bed columns, not many studies have looked into the prediction of column dynamics using reactive transport model for biologically active adsorbent column. In the present study, breakthrough behaviors of three pharmaceuticals, namely, atenolol, ciprofloxacin and gemfibrozil through saturated columns packed with novel clay composite adsorbent in the presence and absence of biofilm were investigated in detail. The adsorbent used in the present study has been proven to remove both hydrophilic and hydrophobic pharmaceuticals simultaneously.17 Control experiments were conducted in column packed with sand. Effects of operating parameters such as flow rate, bed depth and presence of humic acid were studied. The study also attempted to investigate the mechanism of adsorption in detail by comparing two models- equilibrium sorption (EQM) model and linear driving force (LDF) 3
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
model. Reactive transport model coupling adsorption and biodegradation was developed and was validated for experimental conditions. The validated model can be used to simulate the performance of existing fixed bed absorbers and to design adsorbent column for the removal of pharmaceuticals. 2. Materials and Methods 2.1. Chemicals Atenolol, ciprofloxacin and gemfibrozil were used in the present study as target pharmaceuticals. Structure and properties of the target pharmaceuticals are presented in Table S1 in Supporting information. The pharmaceuticals compounds were procured from Sigma Aldrich (Germany). Stock solutions of the target pharmaceuticals were prepared in Millipore water. HPLC grade acetonitrile and potassium dihydrogen phosphate were purchased from Fisher scientific, India. Dextrose was procured from Merck, India. In order to synthesize the adsorbent, naturally available bentonite was used. Chitosan was purchased from TCI, Japan. All the remaining chemicals used were purchased from Rankem, India. 2.2. Synthesis of composite adsorbent Clay composite adsorbent was synthesised using clay, chitosan, powdered activated carbon (PAC) and magnetic nano particles (MNP) in the ratio 1:0.5:0.3:0.3. Detailed synthesis procedure is explained in Arya et al.
17
Initially, solution containing required amounts of FeCl3
and FeCl2 (Fe3+/Fe2+ ratio as 2:1) was added to clay solution. Iron coated clay was synthesized by increasing the solution pH to 11 by adding 0.1M NaOH under constant stirring and nitrogen gas purging. In the second step, solution containing required amounts of iron coated clay and PAC was slowly added to chitosan solution (2% w/v) under constant stirring. The composite solution was pelletized by adding it drop wise to sodium tripolyphosphate solution (1% w/v) at pH 3. After ageing for 24 h, the pellets were washed and dried at 60 ºC. These pellets were used for the experiments.
4
ACS Paragon Plus Environment
Page 4 of 30
Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
2.3. Experimental set up Three identical columns made of acrylic (30 mm dia; 20 cm long) were used for the present study. Columns were packed with sand, adsorbent and biologically active adsorbent. Control experiments were conducted in columns packed with sand. Sand of mean size 0.4 mm was used in the study. In order to avoid clay, silt and other impurities attached to the sand, sand was washed before using it for experiments. Sand was taken in 0.2 mm sieve and was kept under running tap with continuous stirring. Washing was continued till the turbidity of water leaving the sieve was similar to that of the tap water used for washing. Adsorbent used in the present study was magnetic clay polymer composite which was synthesized in order to remove both hydrophilic and hydrophobic pharmaceuticals, and the details of characterization and performance are presented elsewhere.17 Pore volume and surface area of the clay composite adsorbent were 0.12 cm3/g and 94.81 m2/g, respectively. Clay composite comprised of mesoporous structure with an average pore size of 39 Å. Sizes of atenolol, ciprofloxacin and gemfibrozil molecules were 9.12, 7.48 and 7.01 Å, respectively. Mean diameter of the adsorbent particle was 2.5 mm. Initially, water was filled in the column, and sand or adsorbent was loaded stepwise, to prevent any air entrapment in the bed. In order to prepare biologically active adsorbent, required amount of adsorbent was mixed thoroughly with microbial solution which was already acclimatized to selected pharmaceuticals. The microbial consortium which was used for degrading the volatile organic compounds on pharmaceutical industry wastewater18 was used in the present study. In order to acclimatize the microbial consortium to the selected pharmaceuticals, 5 mg/L of selected pharmaceuticals were added to the microbial consortia and bacterial suspension was grown in minimal salt medium (MSM) in presence of dextrose (500 mg/L). The composition of MSM is as follows: Na2HPO4.2H2O- 3.5 g/L, KH2PO4-1 g/L, (NH4)2SO4- 0.5 g/L, MgCl2.6H2O- 0.1 g/L, Ca(NO3)2.4H2O- 0.05 g/L and trace elements- 1 mL. Trace elements’ solution composition was: ZnSO4.7H2O- 0.1 g/L, FeSO4.7H2O- 0.2 g/L, and Na2MoO4.2H2O- 0.03 g/L. It was assumed that 5 mg/L was sufficient to induce necessary enzymes for pharmaceutical degradation. The degradation was continuously monitored and whenever TOC in the system went below 20 mg/L, the medium was refreshed and pharmaceuticals were added. This was continued until the degradation rates of pharmaceuticals were increased. After that, pharmaceutical concentration was reduced to experimental 5
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
concentration of 1 mg/L. This acclimatized bacterial consortium was used for biofilm development on clay composite adsorbent pellets. This biologically active adsorbent was packed in the third column. Amount of biomass attached to the adsorbent, at the start of the experiment, was 1.67 mg/g. Bed depth was maintained at 15 cm. Sampling ports were provided at 5 cm and 10 cm from the inlet. Schematic diagram of experimental setup is presented in Fig. 1. Inlet solution was prepared by spiking pharmaceuticals so as to maintain inlet pharmaceutical concentration to be 1 mg/L and each pharmaceutical was passed through the column separately. In case of column packed with biologically active adsorbent, dextrose (COD= 400 mg/L) was also added to the inlet solution to act as primary substrate to the microbes and all the pharmaceuticals were passed through the column simultaneously. Inlet solution consisted of mineral media to support the growth of biofilm, composition of which is given below: Na2HPO4.2H2O- 3.5 g/L, KH2PO4-1 g/L, (NH4)2SO4- 0.5 g/L, MgCl2.6H2O- 0.1 g/L, Ca(NO3)2.4H2O- 0.05 g/L and trace elements- 1 mL. Trace elements’ solution composition was: ZnSO4.7H2O- 0.1 g/L, FeSO4.7H2O- 0.2 g/L, and Na2MoO4.2H2O- 0.03 g/L. Columns were operated in down flow mode using peristaltic pumps (Miclins, India). Experiments were conducted at different flow rates, namely, 1, 2 and 5 mg/L and at different bed depths, namely, 5, 10 and 15 cm. Chloride was used as tracer in order to determine the dispersion coefficient to be used in mathematical model. 2.4. Analytical procedures Target pharmaceuticals were analyzed in HPLC-UV (Dionex, USA) using reverse phase Acclaim C-18 column (4.6 × 250 mm; 5 µm). Detailed analytical procedure has been presented elsewhere.19 Concentration of chloride ions was found using APHA method 4500 B- Clmethod.20 Protein content in samples was determined using modified Lowry’s method.21 In order to determine the protein content, collected samples were sonicated and centrifuged, and supernatant was used for optical density measurement using UV- Spectrophotometer (Shimadzu, Japan). Morphologies of the adsorbent surfaces during different stages of operations were analyzed in scanning electron microscope (FEI, US).
6
ACS Paragon Plus Environment
Page 6 of 30
Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
2.5. Mathematical model used for column packed with sand In the case of column packed with sand, one dimensional advection-dispersion equation for the transport of contaminant was used for modeling column experiment. (1)
where, C = concentration of pharmaceutical compound, u = pore velocity in x direction, Dx = longitudinal hydrodynamic dispersion coefficient, x = distance along longitudinal direction and t = time. Dx = u.α, where α = dispersivity of the packing medium. Adsorption and biodegradation terms were not included in the model as the effect of these processes in transport would be negligible in the case of clean sand. Eq. 1 was used for simulating tracer transport through columns packed with adsorbent and biologically active adsorbent also. 2.6. EQM model used for columns packed with adsorbent and bioactive adsorbent In order to incorporate effect of adsorption, retardation factor was included in the one dimensional advection-dispersion-reaction equation for the transport of pharmaceuticals. Equation used for simulating transport through adsorbent in EQM model is given below. (2)
where, Rd = retardation factor resulting from the adsorption of the contaminant. Retardation factor, Rd in case of column packed with adsorbent, was determined using Eq. 3. (3)
where, ρb is bulk density of adsorbent, ε is the porosity of adsorbent packing, and Kf and n are Freundlich coefficients obtained from batch isotherm studies.17
7
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 30
In case of column packed with adsorbent, fixed bed process parameters such as depth of mass transfer zone (dMTZ) and empty bed contact time (EBCT) were determined according to the equations given below.13 (4)
(5)
where, h is bed depth, ts is the time to reach saturation (C/C0= 0.95), tb is the time to reach breakthrough (C/C0= 0.05), Fs is the symmetry factor of the breakthrough curve, V is the volume of the adsorbent and Qf is the flow rate. In case of column packed with biologically active adsorbent, all the pharmaceuticals and substrate were passed through the column simultaneously. In order to model transport of pharmaceuticals through column packed with bioactive adsorbent, 1D advection-dispersion equation with reaction term represented by Monod co-metabolic model was used. Equations used for modeling the simultaneous transport of pharmaceuticals are given below. (6)
(7)
(8)
(9)
8
ACS Paragon Plus Environment
Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
(10)
(11)
(12)
where µmax is the maximum specific growth rate, YT is the yield coefficient, Ks is the half saturation constant, KiT is the total inhibition concentration which is taken as the average of inhibition concentrations of each pharmaceutical, η is efficiency, λ is proportionality constant which takes care of the differences in biomass activity in suspended and attached systems, M is the biomass concentration in the column, Kfs and ns are Freundlich parameters corresponding to substrate, S is the substrate concentration and C is the pharmaceutical concentration. Subscript i denotes equation corresponding to each pharmaceutical in multi pollutant transport. Biokinetic parameters were estimated from batch biodegradation studies carried out earlier.19 Assumptions of the above model are: 1. Packing medium is homogeneous. 2. Porosity remains constant throughout the column operation. 3. Adsorption occurs under equilibrium conditions. 4. Adsorption occurs under isotherm conditions. 5. Microbes in the packing medium are immobile and are equally exposed to the substrate and pharmaceuticals. 2.7. Linear driving force (LDF) model used for columns packed with adsorbent and bioactive adsorbent EQM model assumes that local equilibrium is established in any cross section of the column and the influence of adsorption kinetics on column operation is not considered in the model. In LDF model, the effect of film and surface diffusion are considered along with the effect of dispersion. Hence in this model, spreading of the breakthrough curve can be effectively explained by mass 9
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 30
transfer and dispersion which offers better prediction compared to EQM model.13 Mass balance equation for pharmaceuticals during column transport considering non-equilibrium adsorption can be represented as in Eq. (13). In the present model, it is assumed that transport of contaminant through the column is controlled by dispersion, surface diffusion and film diffusion.22
(13)
Film diffusion which is also termed as external transport is represented using Fick’s law.
(14) (15)
where, kF is the film mass transfer coefficient, am is the total surface area corresponding to the adsorbent mass, rp is the radius of adsorbent particle, C is the concentration in the bulk phase and Cs is the adsorbate concentration on the surface of the particle. Intra particle diffusion is represented by LDF approach as given below.
(16) (17)
(18) where, ks is the surface (intra particle) diffusion coefficient, av is the surface area related to the volume of adsorbent, q is the mean adsorbent loading and qs is the adsorbed amount on the surface of the adsorbent which is described by the equilibrium adsorption. ks value obtained from batch kinetic data was used in the model. kF was varied for each flow rate. kF was used as a 10
ACS Paragon Plus Environment
Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
fitting parameter in the model. kF for a particular flow rate was fitted for atenolol and the same value was applied to remaining pharmaceuticals. In case of column packed with biologically active adsorbent, following equations were used for simultaneous transport of pharmaceuticals and substrate. (19)
(20)
In Eq. (19), subscript i denotes the transport equation corresponding to the particular pharmaceutical in multi pollutant transport. Eqs. (8) to (10) were used for estimating biodegradation
corresponding
to
pharmaceuticals
and
substrate.
Transport
equations
corresponding to substrate and pharmaceuticals were solved simultaneously to obtain the simulated results. Values of input parameters used for model are presented in Table 1. System of partial differential equations was solved numerically in MATLAB. Contaminant transport equations were solved using semi implicit-explicit scheme. Advection term was discretized using Essentially NonOscillating (ENO) scheme employing MINMOD limiter to suppress numerical oscillations. Dispersion term was discretized using central difference scheme.23 Model performance was evaluated by calculating modified coefficient of efficiency, E which can be represented as follows.23
(21)
where, E(ti) is the predicted value at time ti, O(ti) is the experimental observation at same time, and
is the mean of experimentally observed values of the particular pharmaceutical. E value
can vary from -∞ to 1. E > 0.5 represents good simulation and the experimental and predicted results for each experiment were used for calculating E. 11
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
3. Results and Discussions 3.1. Transport through column packed with sand Experiments were carried out in column packed with clean sand in order to eliminate the effects of adsorption and biodegradation. These results were useful in understanding only the transport under controlled conditions, without adsorption and biodegradation. The experimental data was used for validating the numerical model. Experimental and numerically simulated results corresponding to the transport of tracer and pharmaceuticals through column packed with sand at different flow rates are presented in Fig. S1 to S4 in Supporting information. It can be observed from the results that the pharmaceuticals behaved like a tracer in column packed with clean sand, with very less retardation. Breakthrough concentrations of tracer at a flow rate of 1 mL/min and at a bed depth of 15 cm were used to back fit the dispersivity value and this value was used to simulate the results for all other experimental conditions. Dispersivity in this experimental study was estimated to be 0.72 cm. E values for all the experimental conditions are presented in Table 1. E values for tracer were found to be varying from 0.93 to 0.97. In case of pharmaceuticals, E values varied from 0.83 to 0.94 indicating good estimation of dispersivity. This indicates that the mathematical model is able to simulate the experimental results well. Complete breakthrough was observed at the outlet which indicated that there was no reaction taking place in the column and the tracer like behavior of pharmaceuticals clearly showed that sorption was insignificant. 3.2. Transport through column packed with clay composite adsorbent Breakthrough curve provides information on the capability of adsorbent to contain the pharmaceuticals. Tracer results are shown in Fig. 2a at different flow rates at a bed depth of 15 cm. Dispersivity was estimated to be 2.7 cm which is higher compared to that of column filled with sand. This may be due to the larger size of adsorbent pellet compared to that of sand. Size of sand particles was 0.4 mm and size of adsorbent pellet was 2.5 mm. Freundlich isotherm parameters estimated from batch experiments17 were used for fitting experimental data and are presented in Table S2 in Supporting information. Experimental data along with numerically simulated results at different flow rates at a bed depth of 15 cm for individual pharmaceutical transport are shown in Figs. 2b to 2d. Transport of pharmaceuticals at bed depths of 5 cm and 10 12
ACS Paragon Plus Environment
Page 12 of 30
Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
cm are given in Figs. 3a to 3c. Comparing the results of sand column and adsorbent column, it is very clear that retardation played a major role in the transport of the pharmaceuticals. It was also observed that ciprofloxacin was retarded the most compared to other two compounds. This can be explained by the higher Kf value of ciprofloxacin. It can be seen from Eq. 3 that Kf has direct correlation to retardation. Among the three pharmaceuticals, Kf was higher for ciprofloxacin and thus effect of retardation was more in case of ciprofloxacin. From Fig. 2c to 2e, it is evident that both the EQM and LDF models are able to predict the breakthrough of all the pharmaceuticals. In transport model, it can be assumed that local equilibrium exists in any cross section in which case, spreading of the breakthrough curve can be explained by diffusion and dispersion.22 Sorption process can be explained by equilibrium adsorption isotherm. However, if mass transfer between the liquid and sorbent phases occurs slowly, non-equilibrium sorption models should be used to predict the transport. Tailing or spreading of the breakthrough curve could be explained by dispersion together with mass transfer coefficients with the help of mass transfer model. Mass transfer coefficients used in the present study are presented in Table 2. Sensitivity analyses for ks and kF were performed and the results are presented in Fig. S5 (sensitivity analysis for ks) and Fig. S6 (sensitivity analysis for kF) in Supporting information. ks and kF values were varied to 200%, 50% and 10% of the actual values for sensitivity analysis. All the other parameters were held constant. From Fig. S5, it is clear that intra particle mass transfer coefficient is an important factor determining the shape of the breakthrough curve. Breakthrough curve is significantly affected at very low values of ks. From Fig. S6, it can be seen that very low values of kF results in earlier breakthrough. However, the effect kF values which were 200 and 50% of the actual values were less significant on the model output. When comparing the E values, LDF model was performing better compared to EQM model in most of the cases. This shows that the intra particle mass transfer coefficients calculated from batch kinetic studies could satisfactorily predict the column dynamics of pharmaceuticals. ks values for atenolol, ciprofloxacin and gemfibrozil were estimated to be 6.5×10-4, 9.4×10-4 and 1.2×10-3 cm/h, respectively. Moreover, the tailing of breakthrough curve is effectively simulated by LDF model. From these observations, it can be concluded that transport of pharmaceuticals through adsorbent column is limited by mass transfer processes. Film diffusion coefficients used in the model were 1.21, 1.53 and 1.78 cm/h for flow rates 1, 2 and 5 mL/min. kF value was found 13
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
to be high compared to ks values for all the pharmaceuticals. It implies that mass transfer is dominated by intra particle diffusion. Film diffusion played a less significant role in the transport. Moreover, tailing is observed in the upper part of the breakthrough curve when intra particle diffusion dominated the kinetics of adsorption. Similar results were reported by other researchers also. Lin et al.24 conducted experimental studies on adsorption of levulinic acid using resin and modeled it using non-equilibrium generic rate model and found that intraparticle diffusion is predominant compared to film diffusion. García-Mateos et al.25 also reported similar results while studying paracetamol adsorption onto activated carbon. 3.2.1. Effect of operating parameters Effect of flow rate on adsorption breakthrough was found out by conducting experiments at different flow rates, namely, 1, 2 and 5 mL/min at a bed depth of 15 cm, results of which are shown in Figs. 2c to 2e. With increase in flow rate, the time required to reach the breakthrough and saturation concentrations decreased due to increased mixing in the column.26 It can be observed from the results that at higher flow rates, sharp breakthrough curves were obtained at short times. This may be due to the reduction in contact time needed for adsorption to take place. The empty bed contact time (EBCT) decreased from 1.76 h to 0.35 h with increase in flow rate from 1 mL/min to 5 mL/min. Moreover, as flow rate increased, breakthrough curve became steeper indicating the decrease in mass transfer resistance.27 It is evident from Table 2 that kF value increased from 1.21 to 1.78 cm/h when flow rate was increased from 1 to 5 mL/min. Similar results were reported in other studies also.24,28 Lin et al.24 also reported that kF increased from 0.09 to 0.158 cm/min when the flow rate increased from 1 mL/min to 5 mL/min. Breakthrough in fixed bed column is affected by bed depth as adsorbent mass in the column varies with change in bed depth. In the present study, studies were conducted at bed depths 5, 10 and 15 cm at a flow rate of 1 mL/min, results of which are shown in Figs. 3a to 3c. As bed depth of the column decreased, time required to obtain complete breakthrough also decreased. However, in this case, steepness of the breakthrough curves were similar owing to the similar experimental conditions. As the bed depth increased or adsorbent mass increased, time for attaining breakthrough increased by manyfold for all the pharmaceuticals. This could be 14
ACS Paragon Plus Environment
Page 14 of 30
Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
explained by the availability of more adsorption sites. Besides, as bed depth increases, more contact time is available for adsorbent-pharmaceuticals interaction. Delay in the occurrence of breakthrough indicates the strong affinity of pharmaceuticals to the adsorbent. In fixed bed adsorbent column, the entire bed is not used for adsorption. Adsorption proceeds layer by layer equilibriating each layer and moving down to the next layer. This layer or zone in which stepwise adsorption take place in adsorbent bed is referred to as mass transfer zone (MTZ). Depth of MTZ gives insight into the mechanism of adsorption taking place in the column at various experimental conditions. Depth of MTZ is shown in Table S3 in Supporting information for various experimental conditions. Depth of MTZ increased with increase in flow rate. This is due to the increase in Reynolds number which leads to the expansion of mass transfer zone.24 In case of ciprofloxacin, depth of MTZ varied from 13.4 to 14.4 cm when flow rate was increased from 1 to 5 mL/min. Depth of MTZ increased from 12.5 to 13.9 cm in case of gemfibrozil, due to increase in flow rate from 1 to 5 mL/min. In case of atenolol, depth of MTZ was found to be varying from 12 to 14.6 cm with an increase in flow rate from 1 to 5 mL/min. Broadened mass transfer zones were observed in the present study. Mass transfer resistances and dispersion affect the broadening of MTZ.13 Dispersion coefficient was estimated to be higher in the adsorption column which may have led to the expansion of the mass transfer zone. As adsorption is a non-selective process, other interfering substances can significantly affect the performance of fixed bed adsorption column. Humic acid is one such component in raw water which interferes with the sorption of compounds. Studies were conducted by adding humic acid to the inlet solution to maintain humic acid concentration as 10 mg/L and pharmaceutical concentration was maintained at 1 mg/L. In this study, humic acid and all the pharmaceuticals were simultaneously passed through the column. Experimental results are presented in Fig. 3d. Results indicated earlier breakthrough of pharmaceuticals compared to the results in the absence of humic acid. Presence of humic acid can lead to competitive adsorption in which case pharmaceuticals and humic acid compete for adsorption sites. Larger humic acid molecules can block even the mesopores of the adsorbent which is referred to as pore blockage.13 Morphologies of the adsorbent were analyzed before and after humic acid adsorption and the results are shown in Fig. 4. It can be seen from the images that the rough and porous surface of the clay composite 15
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
adsorbent was smoothened and the pores on the surface were filled after humic acid adsorption. This study also confirmed that humic acid is an interfering substance competing for available adsorption sites, thus decreasing the efficiency of the adsorption column. 3.3. Transport through column packed with biologically active adsorbent In order to understand the contribution of combined adsorption and biodegradation, transport studies were carried out in biologically active adsorbent column. Experiments were performed for multiple pollutant transport. Tracer results are shown in Fig. 5a at different flow rates at 15 cm bed depth. Dispersivity was estimated to be 2.8 cm by back fitting tracer data obtained at 15 cm at a flow rate of 1 mL/min. Slight change in dispersivity of column packed with bioactive adsorbent compared to that of adsorbent might be due to the presence of biomass in the column. Biokinetic parameters obtained from batch biodegradation experiments and Freundlich parameters obtained from adsorption isotherm experiments were used as the input parameters for the model. It is assumed in the model that the biofilm formed around the adsorbent offers negligible resistance to the mass transfer across it. Only film diffusion and intra particle diffusion control the mass transfer process. The parameter, λ was introduced in the Monod model for reaction term in order to account for the differences in microbial activity in suspended and attached systems. This value was estimated by back fitting the experimental results for atenolol at a flow rate of 1 mL/min at a depth of 15 cm. Proportionality constant, λ was estimated to be 0.85. Experimental data along with numerically simulated results at different flow rates are shown in Fig. 5. E values ranged from 0.76 to 0.91 in case of tracer indicating fairly good estimation of dispersivity. It can be seen from the results that LDF model gave better prediction compared to EQM model. In comparison with transport through adsorbent column, it can be observed from Table 2 that film diffusion coefficient has decreased in bioactive adsorbent column. This shows the increased resistance across the liquid boundary layer. Possible explanation of this could be due to the presence of biofilm around the adsorbent which indirectly increases the mass transfer resistance. In order to model adsorption of pharmaceuticals in bioactive adsorbent column, Freundlich coefficient estimated for each pharmaceutical in presence of other two pharmaceuticals and 16
ACS Paragon Plus Environment
Page 16 of 30
Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
dextrose were used. Freundlich coefficients used for simultaneous transport of all the pharmaceuticals can be found from Table S2 in Supporting information. It is clear from Table 1 that E values corresponding to LDF model are higher compared to that of EQM model except in few cases. Influence of biodegradation is less visible at flow rates of 2 and 5 mL/min for all the pharmaceuticals. At higher flow rates, EBCT would be reduced which leads to the reduction in removal efficiency. Similar results were reported by other researchers14,15 in their study on the removal of micropollutants using biological GAC columns. It can be observed from the results that adsorption was the dominant process as retardation played a major role compared to biodegradation. Subsequently, there was a reduction in the pharmaceutical concentration which could be explained by the microbial biodegradation which was more evident at a flow rate of 1 mL/min as visible from Fig. 5. In case of coupled adsorption and biodegradation, dispersion and adsorption controlled the steepness of breakthrough curve whereas biodegradation component led to the reduction in final concentration. Biodegradation of pharmaceuticals was responsible for maintaining the reduced concentration in the outlet unlike almost complete saturation observed in the column packed with adsorbent. Concentrations obtained at the outlet of the column were 62, 79 and 86% of the inlet concentrations for atenolol, ciprofloxacin and gemfibrozil, respectively, at a flow rate of 1 mL/min. Results of the present study were compared with earlier studies. Previous studies on biofilters and bioactive GAC columns also showed good removal of pharmaceuticals.14,15,29 Reungoat et al.15 reported poor removal of atenolol in biofilter and >90% of atenolol removal in BAC whereas gemfibrozil was degraded upto 50% and 90% in biofilter and BAC, respectively. Salem Attia et al.30 reported biological acclimated sand was able to remove 83% of gemfibrozil. Gemfibrozil was degraded upto 82% after the biofiltration process which employed biological acclimated sand for filtration in the study conducted by Bertelkamp et al.29 Katsigiannis et al.31 observed the removal of selected emerging contaminants in BAC. 3.3.1. Biological activity in the column In the bioactive adsorbent column, adsorption and biotransformation would be taking place, step by step or simultaneously. During the initial time, microbial population would be less and 17
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
pollutants would be getting adsorbed on the adsorbent. At later stages, microbes form a biofilm around the adsorbent. Then onwards, removal of pollutant would be contributed by both adsorption and biodegradation. In the present study, biological activity was monitored by analyzing the morphology and protein content in the biofilm. SEM images of the biofilm formed on the adsorbent are shown in Fig. 6. It can be seen that the biofilm is clustered on the surface of adsorbent. Clusters of bacterial communities are formed in the extracellular polymeric secretions (EPS) which gives the three dimensional structure to the biofilm.32 The fibrous nature of EPS is visible in the images. It can be seen that the biofilm formed was porous and non-continuous which would allow the passage of the target pharmaceuticals to active adsorption sites. Moreover, biological activity was increased in the column as organic loading rates were increased at higher flow rates. This is evident from the high protein content in the outlet of the column during the operation which would also be due to the shearing off of excess biomass due to high flow rate. Initial protein content in the column outlet was 72±6 mg/g during a flow rate of 1 mL/min which increased to 110±8.8 mg/g and 178±13 mg/g during flow rate of 2 mL/min and 5 mL/min, respectively. Growth of biofilm was also evident from the increase in dry weight of the biomass attached on adsorbent pellets. During start-up of column, dry weight of biomass was 1.67 mg/g of adsorbent which was increased to 3.83 mg/g by the end of operation. It should be noted that the effect of biodegradation was becoming evident at low flow rates which shows that if sufficient contact time is available, the pharmaceutical concentrations at the outlet of the column can be significantly reduced. 4. Conclusions Breakthrough behaviors of three pharmaceuticals, namely, atenolol, ciprofloxacin and gemfibrozil in clay composite adsorbent columns were investigated in detail in the present study. Breakthrough times were found to be higher for ciprofloxacin followed by atenolol and gemfibrozil. In case of atenolol, complete saturation was achieved in 25 h corresponding to adsorbent mass of 75 g and treated volume of 1.5 L. For ciprofloxacin, complete saturation was observed after 45 h corresponding to adsorbent mass of 75 g and treated volume of 2.7 L. In case of gemfibrozil, column was completely saturated after 20 h corresponding to adsorbent mass of 75 g and treated volume of 1.2 L. Experimental results were modelled using one dimensional 18
ACS Paragon Plus Environment
Page 18 of 30
Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
transport equation based on equilibrium and mass transfer approach. Intra particle mass transfer was found to be limiting the adsorption process and was responsible for the tailing of breakthrough curve. Column operation was positively affected by increase in bed depth and was negatively influenced by the presence of humic acid and increase in flow rate. Biologically active adsorbent was able to extend the operation period of the column and biodegradation contributed to 38, 21 and 14% removal for atenolol, ciprofloxacin, gemfibrozil, respectively. The study proposes the potential use of biologically active clay composite adsorbent as a fixed bed adsorbent column for simultaneous removal of mixture of pharmaceuticals. Supporting information Transport of tracer and pharmaceuticals through column packed with adsorbent; Sensitivity analysis for mass transfer coefficients; Properties of selected pharmaceuticals; Input parameters of the model; dMTZ and EBCT values of adsorbent column operation. Funding sources The authors declare no competing financial interest. References (1)
Fent, K.; Weston, A. A.; Caminada, D. Ecotoxicology of Human Pharmaceuticals. Aquat. Toxicol. 2006, 76 (2), 122–159.
(2)
Fick, J.; Söderström, H.; Lindberg, R. H.; Phan, C.; Tysklind, M.; Larsson, D. G. J. Contamination of Surface, Ground, and Drinking Water from Pharmaceutical Production. Environ. Toxicol. Chem. 2009, 28 (12), 2522–2527.
(3)
Larsson, D. G. J.; de Pedro, C.; Paxeus, N. Effluent from Drug Manufactures Contains Extremely High Levels of Pharmaceuticals. J. Hazard. Mater. 2007, 148 (3), 751–755.
(4)
Mabrouki, H.; Akretche, D. E. Diclofenac Potassium Removal from Water by Adsorption on Natural and Pillared Clay. Desalin. Water Treat. 2016, 57 (13), 6033–6043.
(5)
Molu, Z. B.; Yurdakoç, K. Preparation and Characterization of Aluminum Pillared K10 and KSF for Adsorption of Trimethoprim. Microporous Mesoporous Mater. 2010, 127 (1– 19
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2), 50–60. (6)
Rakić, V.; Rajić, N.; Daković, A.; Auroux, A. The Adsorption of Salicylic Acid, Acetylsalicylic Acid and Atenolol from Aqueous Solutions onto Natural Zeolites and Clays: Clinoptilolite, Bentonite and Kaolin. Microporous Mesoporous Mater. 2013, 166, 185–194.
(7)
Wang, C.-J.; Li, Z.; Jiang, W.-T. Adsorption of Ciprofloxacin on 2:1 Dioctahedral Clay Minerals. Appl. Clay Sci. 2011, 53 (4), 723–728.
(8)
Wu, H.; Xie, H.; He, G.; Guan, Y.; Zhang, Y. Effects of the pH and Anions on the Adsorption of Tetracycline on Iron-Montmorillonite. Appl. Clay Sci. 2016, 119, 161–169.
(9)
Sotelo, J. L.; Ovejero, G.; Rodríguez, A.; Álvarez, S.; García, J. Study of Natural Clay Adsorbent Sepiolite for the Removal of Caffeine from Aqueous Solutions: Batch and Fixed-Bed Column Operation. Water. Air. Soil Pollut. 2013, 224 (3).
(10)
Cabrera-Lafaurie, W. A.; Román, F. R.; Hernández-Maldonado, A. J. Single and MultiComponent Adsorption of Salicylic Acid, Clofibric Acid, Carbamazepine and Caffeine from Water onto Transition Metal Modified and Partially Calcined Inorganic–organic Pillared Clay Fixed Beds. J. Hazard. Mater. 2015, 282, 174–182.
(11)
Álvarez-Torrellas, S.; Rodríguez, A.; Ovejero, G.; Gómez, J. M.; García, J. Removal of Caffeine from Pharmaceutical Wastewater by Adsorption: Influence of NOM, Textural and Chemical Properties of the Adsorbent. Environ. Technol. 2016, 37 (13), 1618–1630.
(12)
Thiebault, T.; Boussafir, M.; Guégan, R.; Le Milbeau, C.; Le Forestier, L. Clayey–sand Filter for the Removal of Pharmaceuticals from Wastewater Effluent: Percolation Experiments. Environ. Sci. Water Res. Technol. 2016, 2 (3), 529–538.
(13)
Worch, E. Adsorption Technology in Water Treatment-Fundamentals, Processes, and Modeling; DE GRUYTER: Berlin, Boston, 2012.
(14)
Paredes, L.; Fernandez-Fontaina, E.; Lema, J. M.; Omil, F.; Carballa, M. Understanding 20
ACS Paragon Plus Environment
Page 20 of 30
Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
the Fate of Organic Micropollutants in Sand and Granular Activated Carbon Biofiltration Systems. Sci. Total Environ. 2016, 551–552, 640–648. (15)
Reungoat, J.; Escher, B. I.; Macova, M.; Keller, J. Biofiltration of Wastewater Treatment Plant Effluent: Effective Removal of Pharmaceuticals and Personal Care Products and Reduction of Toxicity. Water Res. 2011, 45 (9), 2751–2762.
(16)
Rivera-Utrilla, J.; Gómez-Pacheco, C. V.; Sánchez-Polo, M.; López-Peñalver, J. J.; Ocampo-Pérez, R. Tetracycline Removal from Water by Adsorption/bioadsorption on Activated Carbons and Sludge-Derived Adsorbents. J. Environ. Manage. 2013, 131, 16– 24.
(17)
Arya, V.; Philip, L. Adsorption of Pharmaceuticals in Water Using Fe3O4 Coated Polymer Clay Composite. Microporous Mesoporous Mater. 2016, 232, 273–280.
(18)
Priya, V. S.; Philip, L. Biodegradation of Dichloromethane along with Other VOCs from Pharmaceutical Wastewater. Appl. Biochem. Biotechnol. 2013, 169 (4), 1197–1218.
(19)
Arya, V.; Philip, L.; Murty Bhallamudi, S. Performance of Suspended and Attached Growth Bioreactors for the Removal of Cationic and Anionic Pharmaceuticals. Chem. Eng. J. 2016, 284, 1295–1307.
(20)
APHA; AWWA; WEF. Standard Methods for the Examination of Water and Wastewater; 2012.
(21)
Lowry, O. H.; Roserbrough, N. J.; Farr, A. L.; Randall, R. J. Protein Measurement with the Folin Phenol Reagent. J. Biol. Chem. 1951, 193 (1), 265–275.
(22)
Worch, E. Modelling the Solute Transport under Nonequilibrium Conditions on the Basis of Mass Transfer Equations. J. Contam. Hydrol. 2004, 68 (1–2), 97–120.
(23)
Shashidhar, T.; Murty Bhallamudi, S.; Philip, L. Development and Validation of a Model of Bio-Barriers for Remediation of Cr(VI) Contaminated Aquifers Using Laboratory Column Experiments. J. Hazard. Mater. 2007, 145 (3), 437–452. 21
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(24)
Lin, X.; Qi, G.; Shi, S.; Xiong, L.; Huang, C.; Chen, X.; Li, H.; Chen, X. Estimation of Fixed-Bed Column Parameters and Mathematical Modeling of Breakthrough Behaviors for Adsorption of Levulinic Acid from Aqueous Solution Using SY-01 Resin. Sep. Purif. Technol. 2016, 174, 222–231.
(25)
García-Mateos, F. J.; Ruiz-Rosas, R.; Marqués, M. D.; Cotoruelo, L. M.; RodríguezMirasol, J.; Cordero, T. Removal of Paracetamol on Biomass-Derived Activated Carbon: Modeling the Fixed Bed Breakthrough Curves Using Batch Adsorption Experiments. Chem. Eng. J. 2015, 279, 18–30.
(26)
Monazam, E. R.; Spenik, J.; Shadle, L. J. Fluid Bed Adsorption of Carbon Dioxide on Immobilized Polyethylenimine (PEI): Kinetic Analysis and Breakthrough Behavior. Chem. Eng. J. 2013, 223, 795–805.
(27)
Sancho, J. L. S.; Rodríguez, A. R.; Torrellas, S. Á.; Rodríguez, J. G. Removal of an Emerging Pharmaceutical Compound by Adsorption in Fixed Bed Column. Desalin. Water Treat. 2012, 45, 305–314.
(28)
Long, Y.; Lei, D.; Ni, J.; Ren, Z.; Chen, C.; Xu, H. Packed Bed Column Studies on lead(II) Removal from Industrial Wastewater by Modified Agaricus Bisporus. Bioresour. Technol. 2014, 152, 457–463.
(29)
Bertelkamp, C.; Reungoat, J.; Cornelissen, E. R.; Singhal, N.; Reynisson, J.; Cabo, A. J.; van der Hoek, J. P.; Verliefde, A. R. D. Sorption and Biodegradation of Organic Micropollutants during River Bank Filtration: A Laboratory Column Study. Water Res. 2014, 52, 231–241.
(30)
Salem Attia, T. M.; Hu, X. L.; Yin, D. Q. Synthesized Magnetic Nanoparticles Coated Zeolite for the Adsorption of Pharmaceutical Compounds from Aqueous Solution Using Batch and Column Studies. Chemosphere 2013, 93 (9), 2076–2085.
(31)
Katsigiannis, A.; Noutsopoulos, C.; Mantziaras, J.; Gioldasi, M. Removal of Emerging Pollutants through Granular Activated Carbon. Chem. Eng. J. 2015, 280 (November), 49– 22
ACS Paragon Plus Environment
Page 22 of 30
Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
57. (32)
Ikuma, K.; Decho, A. W.; Lau, B. L. T. The Extracellular Bastions of Bacteria — A Biofilm Way of Life. Nat. Educ. Knowl. 2013, 4 (2), 1–11.
List of figures Figure 1. Schematic diagram of column setup Figure 2.Transport of (a) tracer, (b) atenolol, (c) ciprofloxacin and (d) gemfibrozil through column packed with adsorbent at different flow rates [Bed depth- 15 cm, at flow rates- 1, 2, and 5 mL/min.] Figure 3. Transport of (a) atenolol, (b) ciprofloxacin and (c) gemfibrozil through column packed with adsorbent at different bed depths [Flow rate- 1 mL/min, at bed depths- 5, 10 and 15 cm] Figure 4. SEM images of clay composite adsorbent (a) before and (b) after humic acid adsorption Figure 5. Transport of (a) tracer,(b) atenolol, (c) ciprofloxacin and (d) gemfibrozil through column packed with biologically active adsorbent at different flow rates [Bed depth- 15 cm, at flow rates- 1, 2, and 5 mL/min.] Figure 6. (a, b & c) SEM images of biofilm formed on the surface of adsorbent List of tables Table 1. E values for different experimental conditions Table 2. Mass transfer coefficients used in LDF model
23
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Fig. 1. Schematic diagram of column setup
24
ACS Paragon Plus Environment
Page 24 of 30
Page 25 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
Fig. 2.Transport of (a) tracer, (b) atenolol, (c) ciprofloxacin and (d) gemfibrozil through column packed with adsorbent at different flow rates [Bed depth- 15 cm, at flow rates- 1, 2, and 5mL/min.]
Fig. 3. Transport of (a) atenolol, (b) ciprofloxacin, (c) gemfibrozil through column packed with adsorbent at different bed depths [Flow rate- 1 mL/min, at bed depths- 5, 10 and 15 cm] and (d) transport of pharmaceuticals in presence of humic acid [Flow rate- 1 mL/min, Bed depth- 15 cm]
25
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Fig. 4. SEM images of clay composite adsorbent (a) before and (b) after humic acid adsorption
26
ACS Paragon Plus Environment
Page 26 of 30
Page 27 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
Fig. 5. Transport of (a) tracer,(b) atenolol, (c) ciprofloxacin and (d) gemfibrozil through column packed with biologically active adsorbent at different flow rates [Bed depth- 15 cm, at flow rates- 1, 2, and 5 mL/min.]
Fig. 6. (a, b& c) SEM images of biofilm formed on the surface of adsorbent
27
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 30
Table 1. E values for different experimental conditions Type of column
Flow rate
Gemfibrozil EQM LDF
Atenolol EQM LDF
Ciprofloxacin EQM LDF
Tracer
Sand
1 ml/min
0.93
0.92
0.94
0.97
2 ml/min
0.87
0.88
0.83
0.93
5 ml/min
0.89
0.94
0.85
0.96
Adsorbent
Bioactive adsorbent
1 ml/min
0.91
0.87
0.90
0.89
0.89
0.93
0.89
2 ml/min
0.88
0.92
0.93
0.93
0.92
0.85
0.88
5 ml/min
0.91
0.89
0.87
0.90
0.92
0.94
0.95
1 ml/min
0.81
0.8
0.66
0.77
0.72
0.8
0.76
2 ml/min
0.84
0.76
0.7
0.74
0.79
0.84
0.91
5 ml/min
0.7
0.74
0.71
0.7
0.91
0.65
0.86
Table 2. Mass transfer coefficients used in LDF model
ks
Flow rate, Atenolol ml/min -4 1 6.5×10 (cm/h)
for adsorbent and biologically active adsorbent
kF (cm/h) for adsorbent
kF (cm/h) for biologically active adsorbent
Gemfibrozil Ciprofloxacin 1.2×10
-3
9.4×10
-4
2
6.5×10
-4
1.2×10
-3
9.4×10
-4
5
6.5×10
-4
1.2×10
-3
9.4×10
-4
1
1.21
1.21
1.21
2
1.53
1.53
1.53
5
1.78
1.78
1.78
1
1.08
1.08
1.08
2
1.17
1.17
1.17
5
1.42
1.42
1.42
28
ACS Paragon Plus Environment
Page 29 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Industrial & Engineering Chemistry Research
29
ACS Paragon Plus Environment
Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
448x187mm (150 x 150 DPI)
ACS Paragon Plus Environment
Page 30 of 30
