Competitive removal of textile dyes from solution by pine bark-compost in batch and fixed bed column experiments

Environmental Technology & Innovation 27 (2022) 102421 Contents lists available at ScienceDirect Environmental Technology & Innovation journal homepage: www.elsevier.com/locate/eti Competitive removal of textile dyes from solution by pine bark-compost in batch and fixed bed column experiments Khaled Al-Zawahreha, María Teresa Barralb, Yahya Al-Degsc, Remigio Paradelob,∗ aDepartment of Earth Sciences and Environment, Prince El-Hassan bin Talal Faculty for Natural Resources and Environment, The Hashemite University, Zarqa 13133, Jordan bCRETUS, Department of Soil Science and Agricultural Chemistry, University of Santiago de Compostela, 15782 Santiago de Compostela, Spain cDepartment of Chemistry, Faculty of Science, The Hashemite University, P.O. Box 330127, Zarqa 13133, Jordan article info Article history: Received 14 December 2021 Received in revised form 22 January 2022 Accepted 8 February 2022 Available online 16 February 2022 Keywords: Pine bark compost Column adsorber Competitive adsorption Thomas model Bed depth service time model Textile wastewater abstract Compost from pine bark has been previously suggested as an effective low-cost biosorbent for different classes of textile dyes, although the existing studies have been performed in non-competitive batch conditions, so the effect of competition or adsorption in continuous-flow conditions has not been assessed. In this work, the removal of Basic Violet 10 (BV10) and Direct Blue 151 (DB151) by pine bark compost from single and bi-solute mixtures has been studied in batch and fixed-bed column experiments. Adsorption capacity of pine bark compost was three times higher for BV10 than for DB151 in batch conditions, where competition reduced the uptake of both dyes, with competition factors of 0.63 for DB151 and 0.82 for BV10. Dye adsorption capacity was lower in column than in batch tests, with 112.6 and 34.7 mg g−1for BV10 and DB151, respectively, versus 127.1 and 42.1 mg g−1in batch conditions. The presence of both dyes in solution also reduced their affinities with respect to non-competitive conditions in column tests, with saturation capacities of 71.6 mg g−1for BV10 and 16.8 mg g−1 for DB151. The effect of competition between dyes was higher in columns than in batch conditions, with competition factors of 0.76 for BV10 and 0.59 for DB151. The column biosorbent was effectively regenerated using ethanol, thus enabling reuse in the practical application of compost for textile dye removal. The concentration of dyes in the eluted ethanol was higher than the influent concentration, what would give compost value for pre-concentration of textile dyes. ©2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction A large number of synthetic dyes are commercially available and currently employed in numerous applications. These dyes are classified by their source (natural and manufactured), chemical structure (indigoid, anthraquinone, azo, polymethine, nitro and nitroso, and polycyclic aromatic carbonyl), and substrate use (reactive, acidic, basic, and disperse) (Raval et al.,2017). Cationic and direct dyes are widely employed in the textile industry, with advantages such as high color brightness, a broad color palette, reasonable substantivity and cost-effectiveness (Raval et al.,2017). Industrial coloring, however, is frequently the source of water contaminants that can be harmful (predominantly water-soluble poisonous, ∗Correspondence to: Facultade de Farmacia, Praza Seminario de Estudos Galegos s/n, 15782 Santiago de Compostela, Spain. E-mail address: [email protected] (R. Paradelo). https://doi.org/10.1016/j.eti.2022.102421 2352-1864/©2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/). K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 carcinogenic, and/or mutagenic) to people and other life forms, as well as having unappealing aesthetic consequences. In addition, textile dye effluents and tannery wastewaters contribute to the build-up of chemical oxygen demand and biological oxygen demand in water bodies (Durai and Rajasimman,2011;Saharimoghaddam et al.,2019;Sy et al.,2019), making them unsuitable for other secondary purposes. It is also worth noting that the majority of synthetic dyes are resistant to water, light and chemical substances (Raval et al.,2017;Bilal et al.,2019). As a result, it is critical to design systems for their removal from wastewater. It is already well known that adsorption is a highly recommended methodology for removing dye molecules from effluent streams, due to advantages such as high efficiency, ease of application, adaptability and economic viability (Tahira et al.,2016). In general, solid adsorbents are classified into three categories (Patel,2019): (a) Synthetic adsorbent: Different porous materials with high adsorption capabilities are produced in the laboratory using various procedures. The expensive production process is the main disadvantage of these adsorbents. (b) Natural adsorbent: Natural materials such as plant roots, leaves, and agricultural waste are dried, crushed, and washed with distilled water before being utilized as an adsorbent for target pollutants. Although natural adsorbents are inexpensive, its adsorption capacity is rather limited. (c) Semi-synthetic adsorbent: Natural materials are activated biologically, chemically or physically to produce a highly porous surface. The main benefits of these adsorbents are low cost, high efficiency, no chemical or biological sludge as by-products, and possible absorbent. In this regard, composted organic residuals have recently been investigated as a potential natural material for dyes removal from solution (Paradelo et al.,2019,2020;Al-Zawahreh et al.,2021,2022). Composting organic waste is the most sustainable way to avoid leachate and greenhouse gas emissions after disposal or landfilling. Composting reduces the initial waste volume by up to 65% and yields a nutrient-rich end product that can be used as organic amendment (Ayilara et al.,2020). Besides agricultural uses of compost, increasing attention is being paid to its remediation applications, including removal of textile dyes (Anastopoulos and Kyzas,2015). The published research indicates the good performance of composts for removing different classes of dyes, with high affinity for cationic dyes compared to anionic ones (Kausar et al., 2018;Benkhaya et al.,2020;Al-Zawahreh et al.,2021). In this sense, the adsorption capacity of compost for basic dyes has been observed to be 10 to 20 times higher than for reactive and acid dyes (Paradelo et al.,2019;Al-Zawahreh et al.,2021). However, this has been tested in non-competitive batch conditions, with no systematic studies for competitive removal available and very few in column conditions (Kamarudzaman et al.,2015). In general, very few studies have reported competitive adsorption of textile dyes by any biosorbent (Al-Ghouti et al.,2016;Issa et al.,2017;Yu et al.,2019). Taking this into account, the current work has the following objectives: (i) to assess the effect of competition on textile dye removal from water by pine bark compost, (ii) to assess the performance of pine bark compost as biosorbent in continuous-flow conditions, and (iii) to understand the factors that influence the performance of pine bark compost for dye removal. 2. Materials and methods 2.1. Compost The compost used as biosorbent was produced in an industrial composting facility in the region of Galicia (Spain) and derived from the abundant and unused pine bark residues. The compost was generated via aerobic composting of pine bark in windrows and supplied by the company Costiña Orgánica (A Coruña, Spain). It presents a high content of organic matter (95.3%) and acidic nature, with pH in water of 5.3 and a point of zero charge (pHPZC) of 4.4. It presents low electrical conductivity (0.98 dS m−1), low nutrient contents (0.01, 0.32, and 0.93% for N, P and K, respectively), and very low levels of Pb and Zn, 7.4 and 33.8 mg kg−1, respectively. Cation exchange and anion exchange capacity values are 26.2 and 4.6 cmolckg−1, respectively. N2adsorption–desorption analysis indicated a specific surface area of 22.4 m2 g−1and a mean pore size diameter of 135 Å. Further information about characteristics of the adsorbent, including FTIR spectra, can be found in previous publications (Paradelo et al.,2020;Al-Zawahreh et al.,2021,2022). For the adsorption tests, the sample was crushed and sieved at a particle size 250–500 µm. 2.2. Textile dyes Two textile dyes were selected for the study, belonging to basic and direct dye classes. Basic Violet 10 (BV10, also known as Rhodamine B) and Direct Blue 151 (DB151) were obtained from Panreac (Barcelona, Spain) and Sigma–Aldrich® (USA) chemicals, respectively. The pKa values of the dyes, determined from the half-point of pH-VNaOH titration plot of 0.010 M dye solution and chemical structures are shown in Fig. 1. BV10 has no azo bond and can form a zwitterion at pH 3, while DB151 has a large hydrophobic skeleton with two azo bonds, polar groups (N–H and O–H) and two ionizable sulfonate groups. 2 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 Fig. 1. Chemical structures of Basic Violet 10 (BV10) (pKa 3.70) and Direct Blue 151 (DB151) (pKa 2.68). 2.3. Adsorption isotherms of single and bi-solute systems Adsorption isotherms (at 25 ±1◦C) for both dyes were determined individually in batch experiments using the concentration–variation method. Samples of 0.50 g (±0.01 g) of compost (particle range 250–500 µm) were added to 50.0 mL of solutions containing different concentrations of dyes (5.0–2000 mg L−1for BV10 and 5.0–700 mg L−1for DB151). To have equal chance for dyes adsorption from solution, competitive tests were carried using equal molar concentration of both dyes. The tests were performed at pH 3.0 to maximize the affinity of both dyes toward compost, based on the results of previous works (Al-Zawahreh et al.,2021). The suspensions were sealed and placed in a thermostated shaker (GFL, Germany) for 24 h, the equilibrium time as determined from earlier kinetic studies (Al-Zawahreh et al.,2021). After equilibrium, the supernatants were separated and dye concentrations were determined by measuring absorbance values of the solutions using a double-beam spectrophotometer (Thermo evolution 100 electro Corporation, USA) at 547 nm and 554 nm for BV10 and DB151, respectively. For single-solute solution, dye concentration was estimated from Beer‘s law, while for bi-solute solution modified Beer‘s law was solved at the maximum wavelengths of both dyes. Blank solutions, containing no compost, were also included in the study. The amount of dye absorbed, qe(mg g−1) is estimated as follows: qe=(C0−Ce)·V/m(1) where C0is the initial solute concentration (mg L−1), Ceis the equilibrium concentration in the extract (mg L−1), V is the volume of solution (L), and m is the mass of adsorbent (g). Adsorption isotherms were performed in triplicate and the average values were reported. 2.4. Fixed bed column adsorber tests for single and bi-solute systems In addition to adsorption isotherms, column or fixed bed adsorber experiments are often required for testing practical application. For this, a glass column with internal diameter of 1.0 cm was filled with a known amount of pine bark compost. In order to provide a uniform inlet flow, a layer of 3.0 mm-glass beads was placed at the top of the packed compost. The column tests were carried out by pumping the dye solution through the column in downward flow mode, using a peristaltic pump (Smith and Nephew Watson-Marlow, England). All experiments were conducted at 25.0 ◦C and pH 3.0. The effect of various operational factors including flow rate (10–30 mL min−1), bed-depth height (5.1–17.8 cm) and influent dye levels (50–150 mg L−1) were investigated. The column was considered to be essentially exhausted when the effluent concentration reached 90% (i.e., Ceff/Cinf =0.9) of initial concentration and the breakthrough point at Ceff/Cinf =0.01. The effluent solutions from the column were collected at different intervals and dye concentrations analyzed as described above. In addition to single dye uptake, bi-solute competitive adsorption was examined at different bed-depths of adsorber. For the binary mixtures, dyes were quantified as outlined earlier. 2.5. Regeneration of compost fixed-bed adsorber Desorption tests can clarify the nature of adsorption and the potential recycling of the exhausted compost. If the adsorbed dyes can be eluted by distilled water, then the dye is attached to the adsorbent by weak bonds. Hence, desorption tests were performed for batch and column systems using different solvents including water. In the batch study, 0.1 M sulfuric acid, 0.1 M acetic acid, or ethanol were added to the dye-loaded compost and shaken for 5 h. The suspensions were centrifuged at 5000 rpm for 5.0 min and the supernatants were analyzed for dye content. The chemical reagent 3 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 with the highest dye recovery was selected for desorption in the column test. In this case, the 17.8-cm columns saturated with dyes were regenerated by passing the solvent through them at a flow rate of 5.0 mL min−1, a lower flow rate than that used in the adsorption tests for a better elution from compost. Effluents were recovered and analyzed as explained above, until the eluted dye content was below 1.0 mg L−1. 2.6. Mathematical modelling 2.6.1. Adsorption isotherms Different adsorption models, including two and three parameter models, were adjusted to experimental data in order to predict the maximum uptake capacity, formation of multilayer adsorption, variability of active sites, porosity, and heterogeneity of sorbent. Among known models, only Langmuir and Freundlich were found to be applicable. Langmuir equation has the following form: qe=QLKLCe 1+KLCe (2) where QL(mg g−1) and KL(L mg−1) are respectively the maximum adsorption capacity and the affinity constant of the dye toward the compost. The model assumes equal-energy active sites, no interaction with adsorbed solutes, and one mono-layer coverage. The Freundlich model expression defines the heterogeneity of the surface as well as the exponential distribution of the active sites and the active sites energies: qe=KFCe1/n(3) where KF(Lnmg1−ng−1) is the Freundlich constant and can indicate uptake capacity, while 1/n (dimensionless) measures the favorability of the process, and both are system specific constants. Competitive adsorption was assessed by estimating the competition factor (CF) as follows (Issa et al.,2017): CF =Qmax(bi-solute-system)/Qmax(single-solute-system) (4) where Qmax (bi-solute-system) and Qmax (single-solute-system) are the maximum retention capacities estimated from single and bi-solute systems, respectively. In case of positive competition, CF is higher than unity and adsorption is enhanced by the other solutes. For CF =1, adsorption takes place without competition. The general case is CF<1, meaning that solute affinity is reduced due to negative competition with other solutes. The relative error of prediction (REP%) was used as a criterion to select the optimum model as equilibrium data was fitted by more than one model. REP% is estimated as follows: REP%=100 ·   √∑n i=1(qi,pred −qi,act )2 ∑n i=1(qi,act )2(5) where qi,pred, qi,act, and n are predicted adsorption value, actual adsorption value, and number of experimental points, respectively. Lower REP% indicates better model fitting to the data. 2.6.2. Mass transfer zone and breakthrough curve The relation between the nature of breakthrough curve and fixed-bed adsorption is adequately explained by the mass transfer zone model (Walker and Weatherley,2000). The pollutant is removed most rapidly and effectively in a mass transfer zone (MTZ) at the upper fraction of adsorbent during the initial step of the interaction. This is due to the higher volume of adsorbent and lower amount of adsorbate present in these upper layers, allowing adsorbate to easily escape in the lower strata of the bed and no adsorbate runoff from the adsorbent at the first stage. As a result, a MTZ forms at the front of the column where adsorption occurs, and goes until it reaches the adsorber end, when the effluent solute concentration begins to rise in the aqueous phase (Faust and Aly,1987;Walker and Weatherley,2000). The faster adsorption kinetics, the shallower is the MTZ. The time needed for mass transfer zone (TMTZ) to develop and move down to the end of column adsorber is estimated as Vex/u, where Vex and u are total volume of solution needed for saturation and volumetric flow rate, respectively. The height of mass transfer zone (HMTZ) often depends on adsorption rate and flow rate of solution and is estimated as follows: HMTZ =H·(tex−tb)/tex (6) where H is bed-depth (cm), tex is the time needed for adsorber exhaustion (min), and tbis the time for breakthrough (min). The breakthrough curve plots the ratio of effluent to influent dye concentration (Ceff/Cinf) as a function of process time (t) or treated volume (V). At a certain saturation point, the total capacity of the column is calculated as follows: qbed =u·C0 m·1000 ∫t=tsaturation t=0(1−Ceff Cinf )dt (7) 4 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 where uis the feed flow rate (mL min−1), Ceff is the effluent concentration at time t(mg L−1), C0is the inlet concentration (mg L−1) and tsaturation is the time required for the bed to become saturated (min) at Ceff/Cinf =0.9. Two common models were used to model the curves: the Thomas model was used to estimate dye saturation values and influence of competition, while bed depth service time (BDST) model was used to estimate service time of the adsorber at different bed-depths. The Thomas model allows to calculate the rate constant and maximum solid-phase concentration of adsorbate on adsorbent and it is used to establish breakthrough curves over long service time and often applied to model dye removal in column experiments (Negrea et al.,2020). It is based on the premise that rate driving forces follow second-order reversible reaction kinetics and that there is no axial dispersion in Langmuir adsorption–desorption kinetics (Thomas, 1944). Hence, it has been frequently adopted to study adsorbents of high capacity at dynamic conditions. The Thomas equation is linearly presented as follows: ln(Cinf Ceff −1) =KThqThm u−KThC0t(8) where Cinf is the concentration of influent solution (mg L−1); Ceff is the solution concentration at time t in the effluent solution (mg L−1); KTh is the Thomas rate constant, (L mg−1min−1); qTh is the equilibrium solute uptake per gram of compost (mg g−1); m is the mass of adsorbent (g); u is the volumetric flow rate (mL min−1). The BDST model, proposed by Bohart and Adams (1920), offers a simple approach and rapid prediction of adsorber design and performance (McKay and Bion,1990;Al-Degs et al.,2009). In terms of process concentrations and adsorption factors, this model is based on the relationship between bed depth and service time. This model assumes that the adsorption rate is proportional to both the residual capacity of sorbent and the concentration of the adsorbing solute. The linear form of the model is as follows: ln (Cinf Ceff −1)=ln (ekaNoH u−1)−kaC0t(9) where Cinf (mg L−1) is the inlet dye concentration and Ceff (mg L−1) is the maximum acceptable limit concentration which could be at 50% exhaustion, t is the fix-bed service time (s) and H is the bed-depth (m). N0is the column adsorption capacity (mg L−1), u is the volumetric flow rate (m3s−1) of solution, kais the BDST rate constant (L mg−1s−1). As the exponential term in Eq. (9) is usually much larger than unity, then the relation is reduced to: t=N0 Cinf uH−1 kaCinf ln(Cinf Ceff −1) (10) From the slope and intercept, both N0and kacan be calculated. At 50% breakthrough (Cinf/Ceff =0.5), the second term on the right-hand side of Eq. (10) is reduced to zero, generating the following simplified equation: t0.5=No/Cou·H(11) A plot of t0.5(service time at 50% breakthrough or Ceff/Cinf =0.5) versus H should yield a straight line with a slope equal to N0/C0·u, and this slope represents the time required to exhaust a unit length of the adsorber under the applied experimental variables. In fact, Eq. (10) can be used to estimate column service time at Ceff/Cinf range 0.1–0.5 or saturation% from 10–50. Finally, the adsorbent utilization factor (η) relates the total capacity achieved in adsorber (qbed) with the total capacity obtained in equilibrium batch isotherm (Qmax) and therefore represents the amount of active sites that are not utilized in the adsorber: η=qbed/Qmax (12) 3. Results and discussion 3.1. Dye adsorption from single and bi-solute mixtures in batch condition As a previous step to the dynamic column tests, equilibrium adsorption of BV10 and DB151 in single and bi-solute systems was investigated. The curves obtained are shown in Fig. 2 and the parameters of the models and quality of fit are given in Table 1. Both competitive and non-competitive adsorption of the two dyes presented L2-type isotherm shapes, according to Giles and Smith classification (Giles et al.,1960), in agreement with the literature (Al-Degs et al., 2007;Al-Ghouti et al.,2016;Issa et al.,2017). This reflects a high affinity between BV10 and DB151 and compost at low concentrations and a saturation of the sorbent at high concentrations, as well as a limited effect of competition between BV10 and DB151 toward active sites in the compost. In theory, competition between dyes should increase with concentration, thus disturbing the typical L2-isotherm, but this has not been observed here. As shown in Table 1, Langmuir model outperformed Freundlich model for presenting equilibrium data and also showed good prediction of saturation for both dye adsorption systems, with REP% values lower than 10 in all cases. The good applicability of Langmuir model to present bi-solute systems can be attributed to the basic assumption of this model of no interaction between sorbates. Although Freundlich model described adequately single-solute systems (REP% 9.2–2.4), it was not able to predict competitive adsorption in bi-solute systems (REP% 11.4–12.7). In the case of the Langmuir model, 5 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 Fig. 2. Single and bi-solute adsorption isotherms of dyes by pine bark derived compost. Competitive adsorption tests were carried out using equal molar concentration of both dyes in the solution. Table 1 Isotherm parameters for competitive adsorption of BV10 and FD151 by compost. System Qexp Langmuir Freundlich Competition factor QLKLREP% KfnREP% BV10 127.0 127.1 0.039 1.5 13.40 2.78 2.4 0.82 BV10 (with BD151) 105.0 104.6 0.044 5.2 12.01 3.03 12.7 DB51 42.1 53.9 0.0134 2.9 3.02 2.10 9.2 0.63 DB151 (with BV10) 29.0 33.8 0.0166 6.1 3.21 2.61 11.4 Qexp: experimental saturation estimated from isotherms. QL:Langmuir maximum capacity (mg g−1); KL: Langmuir constant (L g−1);KF: Freundlich constant (Lnmg1−ng−1); n: Freundlich coefficient. KLcan be an indicator of the favorability of adsorption through the determination of the dimensionless separation factor: RL=1/(1+KLCo), with RLvalues between 0 and 1 indicative of favorable adsorption. The RLvalues obtained here were in the range of 0.99–0.96, also indicating a favorable retention of dyes in all systems. Indeed, compost showed a high affinity for both dyes, with a combined adsorption capacity of 138 mg g−1in the bi-solute system, which was larger than the capacities for each dye in single-solute systems. In binary systems, variable competitive effects exist among dyes for active sites, as shown by the competition factors estimated using the saturation values predicted by the Langmuir model. With a competition factor of 0.63, uptake of DB151 was more affected in bi-solute system than BV10, with a competition factor of 0.82. Compared to other materials, the adsorption capacity of CPB for BV10 (127 mg g−1,Table 1) is in the same range of the values obtained with more expensive synthetic adsorbents like graphene oxides (155 mg g−1) and activated carbon (264 mg g−1,Yu et al.,2013), and higher than the values reported for natural adsorbents such as zeolite (13.2 mg g−1), kaolinite (46.1 mg g−1,Yu et al.,2013), bentonite (98 mg g−1,Namasivayam et al.,2001), or banana peels (10 mg g−1,Oyekanmi et al.,2019). The values are also higher than those reported for other composts, for example 27.2 mg g−1obtained for a sewage sludge and plant residue compost (Jóźwiak et al.,2013) or 36.1 mg g−1for MSWC compost (Al-Zawahreh et al., 2021). Comparing the performance of the compost for DB151 is more difficult due to the scarcity of adsorption studies with this specific molecule, although our previous work has shown that CPB has higher adsorption capacity for this dye than municipal solid waste compost (Al-Zawahreh et al.,2021). In summary, composted pine bark is very efficient for dye removal compared to other sorbents, specially taking into account its low production cost, which is around 20–30 $ m−3. In previous works, the adsorption mechanism of different types of dyes (including basic and direct dyes) on compost has been investigated by assessing the effect of pH and ionic strength changes on adsorption and by studying FTIR spectra of dye-loaded composts (Paradelo et al.,2020;Al-Zawahreh et al.,2021). The results of these studies have shown that surface functional groups like carboxylic acids are involved in adsorption and that, in general, electrostatic interaction is an important mechanism of interaction in both cationic and anionic dyes. However, it cannot be the only mechanism behind dye removal. At the pH tested here, compost would present a net positive charge (pHsolution <pHPZC), but dyes would present charges of different sign: positive for BV10 and negative for DB151. Thus, electrostatic interaction is stronger for DB151 than for BV10, and if this was the main mechanism of interaction, adsorption capacity would also be higher for DB151, which was not the case. Considering the complex structure of the dye molecules and compost, the contribution of other mechanisms such as hydrophobic–hydrophobic and dipole–dipole forces should be envisaged to explain dye sorption under unfavorable charge conditions. Indeed, hydrophobic interaction is likely to play an important role, specially taking into account the high organic matter content of the compost (95%) and the large bands of polysaccharides and aromatic rings observed in the FTIR spectra of the compost (Al-Zawahreh et al.,2021). 6 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 Table 2 Parameters of Thomas model, height of mass transfer zone (HMTZ), and time mass transfer zone (TMTZ) estimated at different operational conditions. Factor Value Model parameters BV10 KTh qTh (mg g−1) HMTZ (cm) TMTZ (h) REP% Bed-deptha(cm) 5.1 0.00055 24.9 4.6 2.3 9.4 12.7 0.00044 72.8 5.1 7.3 8.5 17.8 0.00037 102.8 6.5 12.8 7.4 Flow rateb(ml min−1) 10 0.00037 102.8 6.5 12.8 7.4 20 0.00082 75.4 5.1 4.6 7.8 30 0.00094 58.5 4.9 2.1 6.5 Concentrationc(mg L−1) 50 0.00044 57.4 3.1 13.9 9.8 100 0.00037 102.8 6.5 12.8 7.4 150 0.00029 112.6 7.2 9.1 7.5 DB151 Bed-deptha(cm) 5.1 0.00081 22.2 2.2 1.9 6.1 12.7 0.00072 26.6 3.8 3.0 8.0 17.8 0.00055 31.8 4.1 4.6 6.4 Flow rateb(ml min−1) 5 0.00031 36.5 8.7 8.1 8.3 10 0.00055 31.8 4.1 4.6 6.4 15 0.00056 25.4 2.4 2.6 9.2 Concentrationc( mg L−1) 50 0.00061 21.2 3.9 5.1 9.3 100 0.00055 31.8 4.1 4.6 6.4 150 0.00047 34.7 5.1 2.7 7.8 aFlow rate 10 ml min−1; Initial concentration 100 mg L−1. bBed-depth 17.8 cm; Initial concentration 100 mg L−1. cBed-depth 17.8 cm; flow rate 10 ml min−1. 3.2. Removal of dyes from single-solute solution by fixed-bed adsorber The breakthrough curves of dyes at different experimental conditions are provided in Figs. 3 and 4, whereas the results of their analysis using Thomas model and MTZ parameters are provided in Tables 2 and 3. The performance in the fixedbed column is shown by the evolution of the ratio between the effluent dye concentration and the influent concentration in the influent (Cinf/Ceff). Since mass transfer rates were finite, the removal of BV10 and DB151 by the compost adsorber at different operational conditions presented diffuse breakthrough curves with a typical S-shape of variable steepness, which have also been reported for large dye molecules adsorbed onto porous durian peel waste (Thuong et al.,2019). The curves presented a varying degree of steepness, in agreement with previous works by Al-Degs et al. (2009) and Han et al. (2009). Deformed S shapes have been reported for reactive dyes adsorption onto activated carbon, which is attributed to slow interaction in the system (Han et al.,2009). Moreover, the shape of breakthrough curve is also sensitive to the tested parameter, for instance, steeper S-curves are often reported at high flow rate and short bed-depth (Patel,2019). Bed depth.Figs. 3A and 4A show the breakthrough curves of BV10 and DB151 at different bed heights, at fixed flow rate and influent dye concentration. Dyes uptake increased with bed height from 5.1 to 17.8 cm, due to the increment in surface area of adsorbent, which provided more active sites for the adsorption process as well as longer contact times of interaction. In general, steeper breakthrough curves are obtained with smaller bed depth, indicating faster adsorption kinetics, shorter breakthrough time, and fast column saturation. The curves obtained for the lowest bed-depth tested (5.1 cm) were steeper, with earlier breakthrough and faster equilibrium times in contrast with the curves obtained at the highest bed-depth (17.8 cm), with longer breakthrough time and equilibrium times. Rate of adsorption process, as indicated from Thomas model, was reduced by increasing bed-depth. For instance, increasing bed-depth from 5.1 to 17.8 cm resulted in 32 and 37% rate reductions in KTh value for BV10 and DB151, respectively. The reduction in adsorption rate at higher bed-depth is correlated with the time needed to develop MTZ for each bed-depth. As shown in Table 2, longer TMTZ values were reported with longer bed depths and for both dyes. For BV10, time needed to create MTZ and movement along the adsorber was 12.8 h for 17.8 cm depth and only 2.3 h for 5.1 cm bed depth. Hence, adsorption rate of BV10 was slower (KTh 0.00037) at longer bed depth. On the other hand, higher saturation capacity is expected at longer MTZ. For BV10, saturation capacity increased 4-times by bed depth and this is attributed to the increment of MTZ with bed depth (Table 2). Compared to BV10, the faster kinetics of DB151 (KTh 0.00055–0.00088) was attributed to lower time needed to create MTZ (2.1–4.6 h) and faster breakthrough times (Fig. 4A). Higher column saturation was also reported for DB151, which is attributed to higher HMTZ. The early breakthrough time (estimated from the treated volume and flow rate) also increased with bed height: the estimated breakthrough times for BV10 were 11.2, 283.4, and 573.4 min for 5.1, 12.7 and 17.8 cm, respectively. The 7 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 Fig. 3. Breakthrough curves of BV10 removal by compost adsorber (25 ◦C) at different operational factors. 8 K. Al-Zawahreh, M.T. Barral, Y. Al-Degs et al. Environmental Technology & Innovation 27 (2022) 102421 Fig. 4. Breakthrough curves of DB151 removal by compost adsorber (25 ◦C) at different operational factors. 9