Removal of organic compounds by a biofilm supported on GAC: modelling of batch and column data
The performance of a biofilm of Arthrobacter viscosus supported on granular activated carbon on the retention of organic compounds was evaluated. The presence of functional groups on the cell wall surface of the biomass that may interact with the organic compounds was confirmed by Fourier transform infrared spectroscopy, to assess the applicability of this system to the removal of those compounds. The batch assays showed that the removal percentage decreases with the increasing initial concentration. The removal of phenol ranged from 99.5 to 93.4%, the chlorophenol removal ranged from 99.3 to 61.6% and the o-cresol removal ranged from 98.7 to 73.5%, for initial concentrations between 100 and 1,700 mg/L. The batch data were described by Freundlich, Langmuir, Redlich–Peterson, Dubinin-Radushkevich, Sips and Toth model isotherms and the best fit for the retention of phenol and for the retention of o-cresol was obtained with the Sips model, while for chlorophenol, the best fit was obtained with the Freundlich model. The column tests showed that the retention performance followed the order: phenol > chlorophenol > o-cresol, and increased with the increasing initial organic compound concentration. Data from column runs were described by Adams–Bohart, Wolborska and Yoon and Nelson models with good fitting for all the models.
KeywordsArthrobacter viscosus Biodegradation Biosorption Hazardous compounds
List of symbols
- Qe (mg/g)
Ratio between mass of compound sorbed by the biofilm and the mass of GAC, at the equilibrium
- Qmax (mg/g)
Maximum mass of compound sorbed per mass of GAC
- Ce (mg/L)
Concentration of compound in solution at equilibrium
- b (L/mg)
Langmuir adsorption equilibrium constant
Capacity of adsorption
Intensity of adsorption
- KR (L/g), aR (L/mg) and β
Redlich–Peterson constants. β varies between 0 and 1
- KS (Lbsmg1−bs/g), aS (L/mg)bs and bS
Sips isotherm parameters
- Kt (mg/g), at and t
Toth isotherm constants
Related to the mean free energy of sorption per gram of the sorbate as it is transferred to the surface of the solid from infinite distance in the solution
Universal gas constant
Kinetic constant (L/(mg·min) for the Adams–Bohart model
Saturation concentration (mg/L) for the Adams–Bohart model
Inlet compound concentration (mg/L)
Effluent compound concentrations (mg/L)
Compound concentration at the solid/liquid interface (mg/L)
Axial diffusion coefficient (cm2/min)
Is the migration rate (cm/min)
Kinetic coefficient of the external mass transfer (min−1)
External mass transfer coefficient with a negligible axial dispersion coefficient D
Rate constant (min−1)
Time required for 50% adsorbate breakthrough (min)
Breakthrough time (min)
The chemical pollution caused by the presence of heavy metals and organic compounds in the environment is, at present, one of the most serious problems faced by nature. This pollution can cause changes in physical (turbidity, colour, temperature, viscosity, surface tension), chemical (chemical oxygen demand, pH, acidity, alkalinity, dissolved oxygen, toxicity level, nutrients) and biological characteristics of an aquatic medium compromising the water quality for human consumption. A variety of industries including olive mills, oil refineries, plastics, leather, paint, pharmaceutical and steel industries are responsible for the presence of phenolic compounds in wastewaters. Cresol is used in disinfectants and fumigants, in the manufacture of synthetic resins, in photographic developers and in explosives. This compound is highly toxic, corrosive and causes nervous system depression (Tallur et al. 2006). Chlorophenols are used in petrochemical refinery, plastic, pesticide, herbicide, fiberglass manufacturing and coal conversion industries, hence, they are present in the respective effluents. General weakness, fatigue, ataxia, headache, anorexia, sweating, hyperpyrexia, nausea, vomiting, tachycardia, abdominal pain, terminal spasms and death characterize acute poisoning by dichlorophenol (Sathishkumar et al. 2007).
The conventional treatments, physical or chemical, applied to the removal of hazardous compounds are effective but present several limitations that include excessive usage of chemicals, expensive plant requirements, high operational costs and sensitivity to variations in the wastewater input (Prigione et al. 2008). The development of a robust, highly competitive process, with high performance and efficiency, based on the ability of certain biological materials to accumulate and eventually to transform molecules from effluents by physico-chemical or metabolic reactions, is strongly recommended. This process would include a biosorption step followed by a biodegradation step, depending on the microorganism ability to degrade the pollutants. The main advantages of such system are the reusability of biomaterial, low operating cost, improved selectivity for specific pollutants, short operation time and no production of secondary compounds which might be toxic (Mungasavalli et al. 2007).
Several authors have been studying biosorption applied to the removal of organic compounds from wastewaters. Wu and Yu (2006, 2007) used a fungus, Phanerochaete chrysosporium, for the removal of phenol and chlorophenols. A wider approach was used by Tallur et al. (2006) who studied the biodegradation of p-cresol using Bacillus sp. Moreover, the fungus Trametes versicolor was used for the treatment of phenol and o-cresol contaminated wastewaters (Ryan et al. 2007). Bacteria tend to retain strongly many organic compounds including dyes, phenolics and pesticides. According to Xiao et al. (2007), as the interior plasma membrane is impermeable to organic pollutants, the bacterial cell wall is expected to be the primary element responsible for organic biosorption.
Bacteria spend most of their natural existence growing as a biofilm. It is possible that the presence of a suitable substrate for attachment is all that is required to trigger biofilm formation (Jefferson 2004). Biofilms have also been used for the treatment of wastewaters contaminated with organic contaminants (Quintelas et al. 2006; Wicke et al. 2007). The use of activated carbon for the removal of organic compounds was extensively studied by several authors (Cañizares et al. 2006; Mourão et al. 2006; Li et al. 2002). Aktas and Çeçen (2007) studied the adsorption and desorption of chlorophenol on activated carbon and activated sludge. The use of bacteria supported on activated carbon allows combining the good capabilities of both to retain and/or degrade organic compounds and, eventually, a synergetic effect is expected between support and biofilm.
The design and analysis of the adsorption process require equilibrium characterization. Sorption equilibria provide fundamental physicochemical data to evaluate the applicability of the sorption process as a unit operation and to describe the fixation capacity of the biosorbents. The prediction of the concentration–time profile or breakthrough curve for the effluent to be treated is one of the requirements for a successful design of a column adsorption process.
The main focus of this study was the evaluation of the retention ability of a biofilm of Arthrobacter viscosus supported on granular activated carbon (GAC), for the removal of organic compounds from liquid solutions.
The presence of functional groups in the suspended biomass that may have a role in the retention process was confirmed by Fourier transform infrared spectroscopy aiming to evaluate the applicability of this microorganism to the removal of organic compounds. The removal ability was investigated using batch and column studies. All the equilibrium isotherms for the biosorption of phenol, chlorophenol and o-cresol by the biofilms were described by Freundlich, Langmuir, Redlich–Peterson, Dubinin–Radushkevich, Sips and Toth isotherms. The dynamic behaviour of the columns with respect to the inlet organic compound concentration was analysed by the Adams–Bohart, Wolborska and Yoon and Nelson models.
Materials and methods
The bacterium A. viscosus (CECT 908) was obtained from the Spanish Type Culture Collection of the University of Valência. The organic compounds solutions were prepared by diluting extra pure o-cresol (Riedel–de Haen), phenol and chlorophenol (Merck) in distilled water. Glassware used for experimental purposes were washed in 60% nitric acid and subsequently rinsed with deionised water to remove any possible interference by other compounds. The support was GAC from MERCK, characterised by N2 adsorption (77 K) with an ASAP Micromeritics 2001. This support has an average particle size of 2.5 mm, a Langmuir area of 1,270 m2/g and an average pore diameter of 2 nm.
Fourier transform infrared spectroscopy
Infrared spectra of suspended biomass, with and without previous contact with the organic compound, were obtained using a Fourier transform infrared spectrometer (FTIR BOMEM MB 104). For the FTIR study, biomass were centrifuged and dried (24 h at 60°C), followed by weighting. Then, 10 mg of finely ground biomass was encapsulated in 100 mg of KBr (Riedel) in order to prepare translucent sample disks. Background correction for atmospheric air was used for each spectrum. The resolution was 4 cm−1 and a minimum of five scans were conducted for each spectrum with the range between 500 and 4,000 wavenumbers.
The biofilm formation was prepared according to previous studies (Quintelas and Tavares 2001, 2002). A medium with 10 g/L of glucose, 5 g/L of peptone, 3 g/L of malt extract and 3 g/L of yeast extract was used for growth and maintenance of the microorganism. Bacteria are always harvested during the exponential phase of the growth curve, as described in previous publications. The biosorption isotherms for the organic compounds by the biofilm supported on GAC were obtained from batch experiments at 28°C. The experiments were performed with 250 mL Erlenmeyer flasks containing 150 mL of the organic compound solution and 1.5 g of GAC covered with biofilm. The concentrations tested ranged between 100 and 1,000 mg/L, for phenol, between 100 and 1,500 mg/L, for chlorophenol and between 150 and 1,700 mg/L, for o-cresol. The flasks were rotated at a constant rate of 150 rpm until equilibrium was reached. Previous assays were made to determine the time needed for equilibrium to be reached (5 days). Samples of 5 mL were taken after reaching equilibrium, centrifuged at 2,500g during 5 min and the supernatant liquid was analysed for the organic compounds using spectrometry with the 4-aminoantipyrine method (Clesceri et al. 1989). This method is based on the fact that phenolic compounds react with 4-aminoantipyrine in alkaline solution, in the presence of ferricyanide to produce a red reaction product.
All experiments were conducted in triplicate. GAC was placed in Erlenmeyer flasks of 250 ml with 150 ml of distilled water. It was sterilised at 120°C for 20 min to release the air inside the pores. Then, it was placed in mini-columns (internal diameter = 2 cm, height = 30 cm) for open system studies. The microorganism culture and the nutrient broth, composed by glucose, peptone, yeast extract and malt extract, were pumped through at a flow rate of 25 mL/min, according with previous works of this group (Quintelas and Tavares 2001, 2002). The high flow rate used (25 mL/min) allows the formation of a compact biofilm, resistant to the erosion stress promoted by hydrodynamic forces. After biofilm formation, beds were washed out and 5 L of the organic compound solutions with concentrations between 10 and 100 mg/L, were passed through the columns with a flow rate of 5 mL/min for 17 h. At the end of each run, columns were washed out and samples of the effluent were seeded in Petri plates with nutrient agar to assess the metabolic activity of the microorganism. Organic compound concentration at the inlet and at the outlet of the columns was systematically measured by spectrometry with the 4-aminoantipyrine method (Clesceri et al. 1989).
Adsorption isotherm models
Isotherm models used to represent the sorption equilibria
Qe = (QmaxbCe)/(1 + bCe)
Establishes a relationship between the amount of gas sorbed on a surface and the pressure of gas. Assumes monolayer coverage of adsorbate over a homogenous adsorbent surface
Qe = KfCe1/n
This exponential equation assumes that as the adsorbate concentration in solution increases so does it on the adsorbent surface. Can be applied to nonideal sorption on heterogeneous surfaces as well as to multilayer sorption
Qe = (KRCe)/(1 + aRCeβ)
This isotherm model incorporates features of both the Langmuir and the Freundlich isotherms and may be used to represent adsorption equilibria over a wide concentration range
Reddlich and Peterson (1959)
Qe = (KSCe1/bs)/(1 + aSCe1/bs)
Is also called Langmuir–Freundlich isotherm, and the name derives from the limiting behaviour of the equation. At low sorbate concentrations it effectively reduces to a Freundlich isotherm and thus does not obey Henry’s law. At high sorbate concentrations, it predicts a monolayer sorption capacity, characteristic of the Langmuir isotherm
Qe = (KtCe)/[(at + Ce)1/t]
Derived from potencial theory, it is used in heterogeneous systems. It assumes a quasi-Gaussian energy distribution, i.e. most sites have an adsorption energy lower than the peak of maximum adsorption energy
Qe = qD exp (−BD[RT ln (1 + 1/Ce)]2)
The characteristic sorption curve is related to the porous structure of the sorbent
Dubinin and Radushkevich 1947
Modelling column biosorption data
The Adams–Bohart, Wolborska and Yoon and Nelson models
The prediction of the concentration–time profile or breakthrough curve for the effluent to be treated is one of the requirements for a successful design of a column adsorption process. The Adams–Bohart, Wolborska and Yoon and Nelson models can be used to predict the behavior of breakthrough curves. All parameters are referred to in the Nomenclature section, in order of appearance.
The Adams–Bohart model
The Wolborska model
The Yoon and Nelson model
Results and discussion
FTIR spectral analysis
Batch biosorption studies
The adsorption isotherms express the specific relation between the concentration of adsorbate and its degree of accumulation onto the adsorbent surface at constant temperature (Sathishkumar et al. 2007) or, in other words, the amount adsorbed per unit mass of adsorbent as a function of the equilibrium concentration in solution (Mourão et al. 2006). The retention capacity of organic compounds by a biofilm of A. viscosus supported on GAC was analysed using different isotherm models: Langmuir, Freundlich, Redlich–Peterson, Sips, Toth and Dubinin–Radushkevich. The sorption capacity was correlated with the variation of surface area and porosity of the support. Higher surface area and pore volume would result in higher retention capacity (Sathishkumar et al. 2007). Because of this, GAC with a biofilm presents improved characteristics of a good sorbent.
Equilibrium concentrations and removal percentages of phenol, chlorophenol and o-cresol, obtained for different initial concentrations of the organic compound (28°C, 150 rpm)
Comparing values of removal percentage for similar initial concentrations (values highlighted on Table 2), it may be registered a removal percentage of 99.1% of o-cresol, against a removal percentage of 93.4% for the phenol and 80.8% for the chlorophenol. The higher affinity between the biofilm and the o-cresol seems to be the reason for that behaviour. The different removal values can be explained by the differences in molecular size, solubility, dissociation equilibrium and benzene ring reactivity between the three organic compounds tested (Streat et al. 1995). Studies developed by Brasquet et al. (1997) showed that the adsorbability of organic compounds increased with molecular size and decreased with the number of heteroatoms. The molecular sizes of chlorophenol, o-cresol and phenol are 128.6, 108.1 and 94.1, respectively. The biosorption performance described in the present report, for the batch studies, followed the order: o-cresol > phenol > chlorophenol and it can be concluded that the results are in agreement with the study developed by Brasquet et al. (1997): the compound with higher number of heteroatoms (chlorophenol) showed the worst results and between phenol and o-cresol the best results were obtained for the compound with higher molecular size, o-cresol.
Uptake and removal percentages obtained for different biosorbents/adsorbents
Initial conc. (mg/L)
Thawornchaisit and Pakulanon (2007)
Rao and Viraraghavan (2002)
Pleurotus sajor caju
Denizli et al. (2005)
Aktas and Çeçen (2006)
Ziagova and Liakopoulou-Kyriakides (2007)
Aktas and Çeçen (2007)
Activated carbon (GAC)
Hamdaoui and Naffrechoux (2009)
Ziagova and Liakopoulou-Kyriakides (2007)
Activated carbon (GAC)
Lu and Sorial (2004)
Activated carbon (GAC)
Lu and Sorial (2004)
Adsorption isotherm constants and adjustment regression parameter for the models considered and for the three organic compounds
Column biosorption data
Application of the Adams–Bohart and the Wolborska models
Parameters predicted by the Adams–Bohart and Wolborska models and adjustment regression parameter at different inlet organic compound concentrations
Application of the Yoon and Nelson model
Parameters predicted from the Yoon and Nelson model and adjustment regression parameter at different inlet organic compound concentrations
A biofilm of A. viscosus supported on GAC was tested, in batch and column studies, for the removal of phenol, chlorophenol and o-cresol with promising results. The presence of functional groups on the cell wall surface of the biomass that may interact with the organic compounds were previous confirmed by FTIR, to assess the applicability of this system to the removal of organic compounds. The batch biosorption studies showed that the phenol removal ranged from 99.5 to 93.4%, the chlorophenol removal ranged from 99.3 to 61.6% and the o-cresol removal ranged from 98.7 to 73.5%, for the range of initial concentrations of each organic compound used. The best fit for the biosorption of phenol was obtained with the Sips model, followed by the Redlich–Peterson and Toth models, for the chlorophenol, the best fit was obtained with the Freundlich model and for the o-cresol the best fit was achieved with the Sips model. The modelling of the biosorption process allows the successful design of a column adsorption process aiming a future industrial usage. Data from column runs were described by Adams–Bohart, Wolborska and Yoon and Nelson models, with good agreement for both models.
This work was supported by Fundação para a Ciência e Tecnologia (FCT-Portugal), under programme POCTI/FEDER (POCTI/CTA/44449/2002). Cristina Quintelas gratefully acknowledges the Fundação para a Ciência e Tecnologia (FCT-Portugal) for a Post-Doc grant.
- Brasquet C, Subrenat E, Le Cloirec P (1997) Selective adsorption on fibrous activated carbon of organics from aqueous solutions: correlation between adsorption and molecular structure. Water Sci Technol 35:251–259Google Scholar
- Clesceri LS, Greenberg AE, Trussell RR (1989) Standard methods for the examination of water and wastewater, 17th edn. American Public Health Association, WashingtonGoogle Scholar
- Dubinin MM, Radushkevich LV (1947) The equation of the characteristic curve of activated charcoal. Dokl Akad Nauk SSSR 55:327–329Google Scholar
- Freundlich H (1906) Adsorption in solutions. Phys Chemie 57:384–410Google Scholar
- Gerente C, Lee VKC, Le Cloirec P, McKay G (2007) Application of chitosan for the removal of metals from wastewaters by adsorption—mechanisms and models review. Rev Environ Sci Bio/Technol 37:41–127Google Scholar
- Horsfall M Jr, Ogban F, Akporhonor EE (2006) Sorption of chromium (VI) from aqueous solution by cassava (Manihot sculenta CRANZ) waste biomass. Chem & Bio 3:161–173Google Scholar
- Quintelas C, Tavares T (2002) Lead (II) and iron (II) removal from aqueous solution: biosorption by a bacterial biofilm supported on granular activated carbon. J Res Environ Biotechnol 3:196–202Google Scholar
- Toth J (1971) State equations of the solid–gas interface layer. Acta Chim Acad Sci Hung 69:311–328Google Scholar