Ind. Eng. Chem. Res. 1998, 37, 1107-1111
1107
Kinetic Modeling of Copper Biosorption by Immobilized Biomass F. Veglio’,* F. Beolchini, and L. Toro Dipartimento di Chimica, Ingegneria Chimica e Materiali, Facolta´ di Ingegneria, Universita´ degli Studi dell’Aquila, 67040 Monteluco di Roio, L’Aquila, Italy
The kinetic modeling of copper biosorption by Arthrobacter sp. immobilized in a hydroxyethyl methacrylate-based matrix is reported in this work. The resin-biomass complex (RBC) has been used for copper biosorption in different conditions according to a factorial experiment. Factors investigated were cross-linker (trimethylolpropane trimethacrylate) concentration, biomass concentration in the solid, and particles’ granulometry. A maximum copper specific uptake of about 7 mg of Cu/g of biomass (dry weight) has been observed, in the case of a RBC with the following characteristics: 2% (w/w) cross-linker concentration, 8% (w/w) biomass concentration, and 425-750 µm granulometry. The shrinking core model has been used for the fitting of experimental data. A good fit has been found in the case of controlling intraparticle diffusion in all experimental trials. The copper diffusion coefficient in RBC has been estimated from the slope of the regression lines. Values obtained for the diffusion coefficients do not differ from one another with respect to the estimated standard error. An average apparent copper diffusion coefficient of about 3 × 10-6 cm2 /s has been found. Introduction Biosorption of heavy metals is one of the most promising technologies involved in the removal of toxic metals from industrial waste streams and natural waters. It is a potential alternative to conventional processes for the removal of metals, such as ionexchange processes. Regarding the case of the utilization of waste biomass, biosorption actually represents a cheap alternative to conventional processes because of the application of a low-cost sorbent material. For the industrial application of biosorption, the use of an immobilized biomass in a polymeric matrix or in other supports improves the biomass performance and allows its use in many subsequent cycles in the usual unit processes characteristic of chemical engineering (Tsezos, 1984). Established technologically relevant hydrobiometallurgical processes are based on the support of bacteria (BIOCLAIM), algae (AlgaSORB), and peatmoss (BIOFIX) on poly(ethylenimine)glutaraldehyde, silica gel, and polysulfone, respectively (Brierley, 1990; Veglio’ and Beolchini, 1997). Many microorganisms are able to accumulate heavy metals from solutions. Accumulation mechanisms are various. A passive physical-chemical phenomenon seems to be the most common. It is based on adsorption, ion exchange, complexation, and/or microprecipitation (Volesky, 1986). Biomass cell walls, mainly consisting of polysaccharides, proteins, and lipids, offer many functional groups which can bind metal ions such as carboxylate, hydroxyl, sulfate, phosphate, and amino groups. The study of sorption performances of Arthrobacter sp. dispersed in aqueous solution (Veglio’ et al., 1997b) and inside the polymer network of a macro- and microporous cross-linked resin based on poly(hydroxyethyl methacrylate) (polyHEMA)-trimethylolpropane tri* Author to whom correspondence should be addressed. Telephone: +39 862 434223. Fax: +39 862 434203. E-mail:
[email protected].
methacrylate (TMPTM) (Veglio’ et al., 1997a) has already been reported. Arthrobacter sp. lives in natural waters and it represents a cheap and available biosorbing material. There is quite a lack of information in the literature about its sorption performances. Arthrobacter sp. has been shown to be able to absorb manganese, copper, nickel, and lead from aqueous solution (Veglio’ et al., 1997b), and the equilibrium of the process was in all cases well-described by the Langmuir isotherm. After demonstrating that it is an effective metalsorbing material, Arthrobacter sp has been immobilized in a polymeric matrix (HEMA-based) in order to have a biosorbent material with proper characteristics for the use in operation typical of chemical engineering, such as fixed beds. Preliminary results have already been reported. The aim of this paper is to report an expansion of the previous work, aimed at the kinetic modeling of copper biosorption by Arthrobacter sp. entrapped inside polyHEMA-based resins. Materials and Methods Microorganism. Arthrobacter sp., harvested from natural waters near L’Aquila (Italy), was kindly supplied by the Dipartimento di Biologia di Base ed Applicata (University of L’Aquila, Italy). Further details about cell cultivation, harvesting, and use can be found elsewhere (Veglio’ et al., 1997b). Preparation of the Immobilized Biomass. The resin-biomass complex (RBC) has been prepared with different characteristics according to a factorial experiment (see Table 1, further information about factors and levels is given in the Results section). In a typical experiment, the proper amount of lyophilized biomass was suspended in distilled water for one night inside a plastic 10-mL test tube. To the just opalescent suspension, HEMA and TMPTM were added (in this order) and the obtained mixture was then exposed to γ-rays (from a 60Co source) for 5 h at a total dose equal to 1 Mrad. This established procedure (Arshady et al., 1989; Corain et al., 1989a,b; Kaetsu, 1981) produces a glassy phase made by HEMA and TMPTM suitable to homogeneously
S0888-5885(97)00419-3 CCC: $15.00 © 1998 American Chemical Society Published on Web 02/03/1998
1108 Ind. Eng. Chem. Res., Vol. 37, No. 3, 1998
entrapped biomass and ice microcrystals. At the end of the irradiation process, the material was allowed to reach room temperature; it was then ground, sieved, and washed with water (3 × 50 cm3) under vigorous stirring. The RBC material was then dried and stored at 5 °C. Analytical Determinations. The total amounts of biomass entrapped was estimated as already reported (Veglio’ et al., 1997a). Copper concentration was determined by using a Varian Spectr AA 200 atomic absorption spectrophotometer. Biosorption Trials. The biosorption experiments were carried out in batches as follows, 1 g of biosorbing material was added to copper-containing solutions. The final volume was 80 cm3 and the copper initial concentration 25 mg/dm. Flasks were kept at room temperature. Aliquot amounts (1.5 cm3) were collected, periodically, for residual metal concentration determination. Before analysis, solids were removed by centrifugation (6000 g for 10 min). Kinetic Model. In the case of immobilized biomass mass-transfer resistances inside the particle may take place. Hence, a kinetic model has to be taken into consideration, in order to estimate mass-transfer characteristic parameters. The shrinking core model (Rao and Gupta, 1982) has been used in this work. In the case of a process controlled by the diffusion of metal ions through the liquid film (film diffusion control) the extent of the biosorption process as a function of time will be given by the following expression:
χ)
∫0tC dt
3D δRC0
(1)
Consequently, if the film diffusion is controlling, a plot of χ vs ∫t0C dt yields a straightline relationship. If the process is controlled by the diffusion through the reacted shell (particle diffusion control), the model is represented by the following expression:
F(χ) ) 1 - 3(1 - χ)2/3 + 2(1 - χ) )
6Da
t C dt ∫ RC 2
0 0
(2)
Consequently, in the case of particle diffusion control a plot of function F(χ) vs ∫t0C dt will give a straightline relationship and the apparent diffusivity in the RBC could be obtained from the slope of such a plot as follows:
R2 Da ) [slope]C0 6
(3)
ANOVA Analysis. The planning of experimental runs has been carried out using full factorial design. In this way, an orthogonal experimental plane is arranged in which it is possible to evaluate independently both the main effects and the interactions among the factors investigated for a given response. The effect of a factor is the change in response produced by a change in the level of the factor. When the effect of a factor depends on the level of another factor, the two factors are said to interact. The analysis of the variance (ANOVA) allows one to evaluate whether the effects and the interactions among the investigated factors are signifi-
Table 1. Factors and Levels Investigated factors
levels
cross-linker [% w/w] biomass [% w/w] granulometry [µm]
2
4 8 425-750
8 15 750-1000
Table 2. Copper Biosorption by RBC: Equilibrium Specific Uptake qeq trial
cross-linker [% w/w]
biomass [% w/w]
granulometry [µm]
qeq [mg/g]
1 2 3 4 5 6 7 8 9 10 11 12
2 2 2 2 4 4 4 4 8 8 8 8
8 8 15 15 8 8 15 15 8 8 15 15
425-750 750-1000 425-750 750-1000 425-750 750-1000 425-750 750-1000 425-750 750-1000 425-750 750-1000
6.6 4.7 3.5 4.6 6.5 5.0 4.1 4.1 6.2 5.5 4.4 3.4
cant with respect to the experimental error (Himmelblau, 1970). Results and Discussion Biosorption. The RBC sorption abilities have been investigated according to a factorial experiment. Table 1 shows factors and levels investigated. The levels are the values chosen for each factor. As shown by the table, factors investigated were cross-linker concentration, biomass concentration in the solid, and particles’ granulometry. Cross-linker concentration (and therefore polymer chain concentration) might influence the metal accessibility inside the biosorbing material, and hence the metal apparent diffusivity, Da. Biomass concentration in the solid has been chosen as a factor because its effect on the metal specific uptake (mg of metal accumulated/g of dry biomass) has been evident in the case of nonimmobilized biomass (Veglio’ et al, 1997b; Fourest and Roux, 1992; Gadd and White, 1989). One of the main goals of this work is to verify if biomass concentration is also relevant with immobilized cells. Granulometry might influence the kinetic controlling step: by increasing particle diameter the process might become controlled by the metal diffusion through the material. Table 2 reports the equilibrium specific uptake qeq values obtained for each biosorption trial. As shown by the table, a maximum copper uptake of about 6.6 mg of Cu/g of biomass (dry weight, d.w.) has been observed, in the case of a RBC with the following characteristics: 2% (w/w) cross-linker concentration, 8% (w/w) biomass concentration, and 425-750 µm granulometry. Results obtained have been examined through the ANOVA (see Materials and Methods section) considering qeq as response of the process. Figures 1-3 show the metal specific uptake vs time profiles in different experimental conditions. ANOVA results are also evidenced in those figures. The following aspects have come out from ANOVA (see Table 3): (i) It is evident there is a significant effect (significance 97%) of biomass concentration on the maximum specific uptake: by increasing biomass concentration
Ind. Eng. Chem. Res., Vol. 37, No. 3, 1998 1109
Figure 1. Granulometry effect on the specific uptake vs time profile (biomass 15%; cross-linker 4%). Continuous lines are calculated through SCM with particle diffusion control.
Figure 3. Biomass weight fraction effect on the specific uptake vs time profile (cross-linker 2%; granulometry 425-750 µm). Continuous lines are calculated through SCM with particle diffusion control. Table 3. ANOVA Results: Equilibrium Specific Uptake as a Response of the Process factora
mean square [(mg/g)2]
mean square/errorb
significance [%]
A B C AB AC BC ABC
0.007 8.998 1.357 0.039 0.053 1.327 0.632
0.02 21.85 3.29 0.09 0.13 3.22 1.53
0.7 97 70 3.8 5.1 70 43
a A ) cross-linker; B ) biomass; C ) granulometry. b Error variance ) 0.412 (mg/g)2.
Figure 2. Cross-linker weight fraction effect on the specific uptake vs time profile (biomass 8%; granulometry 425-750 µm). Continuous lines are calculated through SCM with particle diffusion control.
inside the RBC, the copper specific uptake decreases (see Figure 3). (ii) There are not significant effects of granulometry and cross-linker concentration on the maximum specific uptake (see Figures 1 and 2). Results in Figure 1 also give evidence of an obvious negative effect of granulometry on the process kinetics. The negative effect of biomass concentration on the metal specific uptake (Figure 3) has also been observed in the case of free biomass (i.e., suspended in solution) (Veglio’ et al, 1997b; Fourest and Roux, 1992; Gadd and White, 1989) especially for low values of the biomass concentration (