Retention of inorganic arsenic by coryneform mutant strains
ARTICLE IN PRESS WAT E R R E S E A R C H
41 (2007) 531 – 542
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Retention of inorganic arsenic by coryneform mutant strains J.C. Feoa, E. Ordon˜ezb, M. Letekb, M.A. Castroa, M.I. Mun˜oza, J.A. Gilb, L.M. Mateosb, A.J. Allera, a
Department of Biochemistry, Area of Analytical Chemistry, Leo´n, Spain Department of Ecology, Genetic and Microbiology, Area of Microbiology; University of Leo´n, E 24071 Leo´n, Spain
b
art i cle info
ab st rac t
Article history:
The natural resistance mechanisms of corynebacteria to respond to the environments
Received 12 June 2006
containing high levels of arsenic were successfully adopted to develop inexpensive and
Received in revised form
selective extractants for submicrogram amounts of arsenic. Kinetic and equilibrium
19 October 2006
characteristics were evaluated, and a preliminary exploration of the capability of these
Accepted 7 November 2006
strains to be used for arsenic speciation was also made in this work. Three kinetics models were used to fit the experimental data. It was found that the pseudo-first-order kinetics
Keywords:
model was not quite adequate to describe the retention process, while the intraparticle
Arsenate
diffusion and the pseudo-second-order kinetics models provide the best fits. The
Arsenite
equilibrium isotherm showed that the retention of arsenic was consistent with the
Corynebacteria
Langmuir equation and that the Freundlich and Dubinin–Radushkevich models provided
Mutant strains
poorer fits to the experimental data. The maximum effective retention capacity for arsenic
Isotherm
was about 15.4 ng As/mg biomass. The amount of arsenic retained was directly measured
Kinetics
in the biomass by forward planning a slurry electrothermal atomic absorption spectro-
Electrothermal atomic absorption
metric procedure.
spectrometry
& 2006 Elsevier Ltd. All rights reserved.
Slurry sampling
1.
Introduction
Heavy metal contamination of natural water is a major problem in industrialised areas, as they can cause a detrimental effect on environment and finally on human health. Arsenic is categorised as a heavy metal and comes into the environment as a result of the contribution of several natural sources, but mainly from human intervention, usually through mining, industrial, agricultural, medicinal, and like other activities (Langdon et al., 2003). The adverse health effects arsenic can produce in humans are well known (Duker et al., 2003). However, toxicity of arsenic differs with the chemical forms involved, resulting in a great challenge for arsenic speciation. For a better management of contaminaCorresponding author. Tel.: +34 987 291536; fax: +34 987 291883.
E-mail address: [email protected] (A.J. Aller). 0043-1354/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2006.11.015
tion of this metal, an adequate monitoring and control of the arsenic species levels are required. The most advantageous large-scale continuous operations for the separation of metals are liquid–liquid extraction and solid retention (Maity et al., 2005; Daus et al., 2004; Lin and Wu, 2001). More recently, however, many efforts have been addressed on the development of new sorbents, where bacteria-based biosorption has been considered a suitable wastewater technology to remove efficiently heavy metals. In other fields, such as speciation analyses, diverse analytical instrumentation has generally been used (Oliveira et al., 2005; Fattorini et al., 2004; Rosen and Hieftje, 2004; Jain and Ali, 2000), but microorganisms, particularly bacteria and mainly those evolving resistance mechanisms to many toxic metals
ARTICLE IN PRESS 532
WA T E R R E S E A R C H
Nomenclature
C0 Ce E KF KL KS kd k1 k2
concentration of arsenic in the initial sample solution (ng ml1) concentration of arsenic in the sample solution at equilibrium (ng ml1) mean free energy of retention (kJ mol1) retention capacity (Freundlich isotherm constant) (mg g1) instability equilibrium constant (Langmuir isotherm constant) (ml ng1) stability equilibrium constant Intraparticle diffusion rate (ng ml1 h1/2) the equilibrium rate constant of pseudo-firstorder sorption (h1) the equilibrium rate constant of pseudo-secondorder sorption (mg ng1 h1)
(Silver and Phung, 1996), have also been considered for preconcentration and/or separation of diverse metal species. The strategies developed by bacteria to circumvent the toxicity of arsenic usually include: (i) decreasing in the uptake of arsenate through the phosphate uptake system (Cervantes et al., 1994), (ii) evolving peroxidation reactions through membrane lipids (Abdrashitova et al., 1986), and (iii) using the best characterized microbial arsenic detoxification pathway, which involves genetic determinants typically organized in the ars operon (Silver and Phung, 1996). The ars operon is generally constituted of either three (arsRBC) or five (arsRDABC) genes, organized into a single transcriptional unit. The three-gene system encodes the arsenic transcriptional repressor (arsR), an arsenite permease, (arsB), a membranelocated arsenite efflux pump, and an arsenate reductase (arsC), which converts arsenate to arsenite before arsenic is pumped out of the cell through the ArsB anion pump; the five-genes operon (arsRDABC) encodes, besides the above described genes, a negative regulatory protein which provides additional fine tuning (arsD) and an arsenic-specific ATPase (arsA) (Silver and Phung, 1996). As a general rule, in bacteria, the interaction of arsenite with the repressor protein (ArsR) undergoes a conformational change of the protein and dissociates from the regulatory sequences of the ars operon, leading to the expression of downstream genes. Metal resistance mechanisms are clearly regulated by specific metalloregulatory proteins and the high affinity and specificity of the regulatory protein of the ars operon, ArsR, have been used for developing whole-cell bacterial biosensors for arsenic (Scott et al., 1997; Cai and DuBow, 1997; Ramanathan et al., 1997; Tauriainen et al., 1997), as well as for arsenic remediation (Kostal et al., 2004). In these studies, ArsR and other proteins, such as ArsB, have been fused to the bacterial luciferase genes (lux operon) (Cai and DuBow, 1997; Ramanathan et al., 1997; Roberto et al., 2002; Tauriainen et al., 1999) or to firefly luciferase (luc operon) (Tauriainen et al., 1997). Bioavailability studies of arsenic have also been tested using several bacteria-based biosensors (Tauriainen et al., 1997, 1999; Ji and Silver, 1992; Corbisier
41 (2007) 531– 542
k n qe qt qm RL R T t t1/2 e n0
constant related to retention energy (mol2 kJ2) intensity of the sorbent amount of As(III) retained per mass unit of biomass at equilibrium (ng mg1) amount of As(III) retained per mass unit of biomass at any time t (ng ml1) amount of As(III) retained at saturation (ng ml1) dimensionless equilibrium parameter the ideal gas constant (kJ mol1 K1) the absolute temperature (K) retention time (h) half-retention time (h) Polanyi potential initial retention rate (ng mg1 h1)
et al., 1993; Peta¨nen and Romantschuk, 2002). In general, the best-suited host strains are native bacteria from the environment for which the analysis is though (Peta¨nen and Romantschuk, 2003), although different recombinant plasmids have been used with similar successful results (Peta¨nen et al., 2001). The saprophytic soil bacterium Corynebacterium glutamicum is a Gram positive microorganism with a high resistance to arsenic; this resistance is associated to the presence of two arsenic detoxification operons (ars1 and ars2), which are basically structured following the gene arrangement arsR1arsB1-arsC1-arsC1’ for ars1 operon and arsR2-arsB2-arsC2 for ars2 (Ordo´n˜ez et al., 2005). The aim of this work was to evaluate the capability of four C. glutamicum strains, the wildtype strain (ATCC 13032) and three mutants generated by disruption of the arsenite permease genes (arsB1, arsB2 and both arsB1-arsB2) for the retention of As(III) and As(V) species. The biosorption process was followed by forward planning a slurry sampling living electrothermal atomic absorption spectrometry (ETAAS) procedure, which allow us to measure in a direct way the arsenic retained by the living bacterial biomass.
2.
Materials and methods
2.1.
Instruments and operating conditions
A Thermo Jarrel Ash atomic-absorption spectrophotometer (SH 11), equipped with a CTF-188 graphite atomizer and Smith–Hieftje background correction, was used for the measurements. The operating parameters used were as follows: wavelength, 193.7 nm; the arsenic pulsed hollow cathode lamp, Visimax II, was used under the recommended conditions; bandwidth, 1 nm. Standard uncoated rectangular graphite tubes, and standard pyrolytic graphite-coated graphite platforms were used for atomization. A syringe was used to inject manually 10 ml of the solutions and slurries into the graphite atomizer at room temperature. The temperature
ARTICLE IN PRESS WAT E R R E S E A R C H
533
41 (20 07) 53 1 – 542
Table 1 – Temperature programme Parameter
Dry
Pyrolysis 1
Pyrolysis 2
Atomization
Cleaning
Temperature, 1C Ramp time, s Hold time, s Argon flow Read on, s
150 2 0 Low
375 15 0 Medium
1100 15 0 Low
2100 0 4 Off 4
2500 0 0 High
programme of the atomiser is shown in Table 1. Argon, 99.995% purity, was used as the purge gas. At least three replicate determinations based on the integrated absorbances of the atomic absorption signals of arsenic were used for each measurement. A pH-meter (Crison model Digit 505) was used to measure the acidity of the aqueous phase when necessary. A Mettler AE 240 semi-micro analytical balance (sensitivity 70.01 mg) was used to weigh the chemicals and biomass. The slurry was maintained during sample introduction in the atomizer using a Brasonic sonicator (Model B-5).
2.2.
Chemicals
Arsenic stock solutions, 1000 mg l1, were prepared by dissolving suitable amounts of sodium arsenite and di-sodium hydrogen arsenate in an appropriate volume of de-mineralized water. Stock solutions of those elements used as chemical modifiers were prepared from Ni(NO3)2 and Pd(NO3)2, dissolved in de-mineralised water. Working solutions were prepared by the appropriate dilution of the stock solutions immediately prior to their use. The Tryptone Soy Agar (TSA; Oxoid) medium (40 g l1) was composed of tryptone (1.5% m/m), soy peptone (0.5% m/m), sodium chloride (0.5% m/m) and agar (1.5% m/m) and adjusted to pH 7.3 70.2; TSB medium (Tryptone Soy Broth) was the same than TSA but without agar. The composition of minimal medium for corynebacteria (MMC) was as follow: glucose (2 g l1), ammonium sulphate (10 g l1), urea (3 g l1), monopotassium phosphate (1 g l1), magnesium sulphate heptahydrate (0.41 g l1), sodium chloride (50 mg l1), glucose (2 g l1), iron(II) sulphate monohydrate (0.2 mg l1), manganese sulphate monohydrate (0.2 mg l1), biotin (5 mg l1) and tyamine (20 mg l1) for liquid MMC and containing 15% of agar for solid MMC (Kaneko and Sakaguchi, 1979). All media were sterilized by autoclaving for 20 min at 120 1C. All chemicals used in this study were of analytical reagent grade and were obtained from Merck. Distilled, deionized water (resistivity 18 M Ocm) was used for the preparation of the samples and standards. The pH was adjusted by using HNO3 and NaOH as necessary. The TAB buffer used for washing at the As retention analysis contains: HEPES (75 mM), potassium chloride (150 mM) and magnesium chloride (1 mM); the buffer was adjusted to pH 7.370.2.
2.3.
Strains and mutants
C. glutamicum ATCC 13032 was used as the type strain. Escherichia coli S17-1 was used for subcloning and as donor
strain in conjugation (Mateos et al., 1996). Bacterial cultures were grown in complex medium (TSA/TSB) for 24–48 h at 37 1C (E. coli) or 30 1C (corynebacteria). For preservation, cultures were stored at 4 1C until required for use. When necessary, antibiotics kanamycin and apramycin were added in cultures at final concentrations of 12.5 mg l1 for corynebacteria and 50 mg l1 for E. coli. Internal fragments from the C. glutamicum arsB genes were subcloned in E. coli vectors, and then transferred to corynebacteria by conjugation using standard protocols (Mateos et al., 1996). C. glutamicum transconjugants were selected on TSA medium containing 50 mg l1 of nalidixic acid and an additional antibiotic (kanamycin, or apramycin), depending on the mobilizable plasmid used for the mating. The single and double C. glutamicum arsenite permease mutants (ArsB1, ArsB2 and ArsB1-B2) were constructed as elsewhere (Ordo´n˜ez et al., 2005).
2.4.
Procedures
For the As retention procedure, the living bacterial cells (C. glutamicum strains) were cultivated in 50 ml of liquid MMC mixed with arsenic in the concentration range 0–200 ng ml1, using a 250-ml Erlenmeyer flasks for 5 h, as otherwise was noted, with continuous stirring (200 rev min1). The inoculated liquid cultures started with the same optical density (0.05) measured at l ¼ 598 nm (OD598 ¼ 0.05). After 5 h, the metal-biomass pellet was collected by centrifugation at 12 000 g, washed twice with TAB buffer, and treated with a 5.8 M nitric acid solution containing the optimum amount of palladium nitrate (2000 mg l1) as a chemical modifier. Then, maintaining this slurry by sonication, the amount of arsenic was determined by slurry sampling ETAAS. The concentration of arsenic in the sample can be evaluated from a calibration graph prepared by treating arsenic standards in the same way as samples via bacterial preconcentration.
3.
Results and discussion
3.1.
Optimization of the analytical procedure
As slurry ETAAS was though to be used for evaluating the retention process, those factors affecting the atomic absorption signal of arsenic injected as slurry were firstly optimized: the chemical modifier concentration, pryrolysis and atomization temperatures, and the slurry characteristics (the washing acid concentration and slurry concentration). All these experiments were carried out by using an As-biomass slurry,
ARTICLE IN PRESS 534
WA T E R R E S E A R C H
prepared after the retention of 50 ng ml1 of arsenic as As(III) by the C. glutamicum wild-type strain under standard culture conditions. Owing to the difficulties associated with the determination of arsenic by ETAAS in biological matrices, the use of a chemical modifier is mandatory. In order to optimize the suitable amount of the modifier, several studies working at the tentatively selected preliminary pyrolysis and atomization temperatures of 900 and 1800 1C, respectively, were carried out. The effect of the amount of two metals (Ni and Pd, as nitrates) on the stabilization of arsenic in the graphite tube was evaluated for platform atomization using the temperature programme shown in Table 1 and the results were shown in Fig. 1. From Fig. 1, it is possible to conclude that nickel and palladium concentrations of 2000 mg l1 in the final slurry were sufficient to stabilize the atomic absorption signal of arsenic. Nonetheless, the modifier selected for the subsequent comparative studies was palladium at the concentration of 2000 mg l1 in the final slurry. In order to find out the maximum allowable pyrolysis temperature for arsenic, introduced as a As-biomass slurry in the presence of palladium, a curve was deduced of the analyte signal (integrated absorbance) obtained at different temperatures for a pyrolysis step constituted of two pyrolysis temperatures. Pyrolysis 1 temperature was maintained fixed at 375 1C, while pyrolysis 2 temperature was modified from 400 to 1300 1C. In each case, the same atomization temperature (1800 1C) was used throughout. Pyrolysis curves were established for slurries prepared using biomass materials with and without modifier. Fig. 2 shows the effect of pyrolysis temperature on the atomic absorption signals of arsenic. When no modifier was used, arsenic was readily loss owing to its high volatility and the low absorption signal of arsenic decreased rapidly with pyrolysis temperature. By contrary, using chemical modifiers, higher and most stabilized arsenic atomic absorption signals were obtained. Nonetheless, as it is shown in Fig. 2, even in the presence of modifier, the atomic absorption signal of arsenic decreased rapidly at pyrolysis temperatures above 1100 1C. In conclusion, the pyrolysis
41 (2007) 531– 542
temperature value of 1100 1C was selected for further experiments. A second set of experiments exploring the relationship between the atomization temperature and the integrated absorbance of arsenic showed that the maximum and most stabilised atomic absorption signals (integrated absorbances) of arsenic were obtained at an atomization temperature of about 2100 1C when a 1100 1C pyrolysis 2 step was used. Using living cells, the removal of arsenic from the sample solution depends not only on the uptake rate but also on the growing rate of the bacterial cells, because the number of cells increases during cultivation and, consequently, the retention capacity for arsenic. However, this behaviour not necessarily should be observed in the atomic absorption signal of arsenic if analyte is introduced into the graphite tube as slurry. Biomass may affects on the atomization path of arsenic, but also it can alter the sample introduction stage by clogging the aspiration capillary. The amount of biomass injected into the tube affected on the atomic absorption signal of arsenic. Thus, the presence of organic material decreased the atomization efficiency of arsenic when the chemical modifier was absent, as the signals derived from arsenic aqueous solutions were slightly higher than those obtained for As-biomass slurry. However, in the presence of palladium as modifier, sensitivity was not only restored, but improved for the As-biomass slurries compared to Asaqueous solutions. Therefore, the bacterial mass favours atomization of arsenic, but only in the presence of palladium. The most likely explanation for this behaviour is that the organic matter may act as reducing agent for both palladium and arsenic, thus facilitating their interaction in reduced form. In consequence, the effect of slurry density was studied in the presence of the chemical modifier. For these situations, the atomic absorbance of arsenic increases gradually with the bacterial mass used in the retention process and, consequently, injected into the tube, but only till about 40 mg l1. Then, the atomic absorption signal of arsenic decreases suddenly. In conclusion, we have found that the best slurry density for further experiments was 40 mg l1.
Integrated absorbance. s
0.065 0.060 0.055 0.050 0.045 Palladium
0.040
Nickel 0.035 0
500
1000 1500 2000 Modifier concentration, mg l−1
2500
Fig. 1 – Effect of the modifier concentration on the integrated absorbance of arsenic.
ARTICLE IN PRESS WAT E R R E S E A R C H
41 (20 07) 53 1 – 542
535
Integrated absorbance,s
0.07 0.06 0.05 0.04 0.03 As-biomass As-biomass As-biomass-Pd As-biomass-Pd
0.02 0.01 400
600
800
1000 1200 1400 1600 1800 2000 2200 2400 Temperature,°C
Fig. 2 – Pyrolysis (solid symbols) and atomization (open symbols) temperatures for As-biomass slurry alone (squares) and As-biomass slurry together with the optimum amount of palladium as modifier (circles).
Nitric acid usually has an effect on the atomic absorption signals of arsenic, mainly if injected as an organic matrix. Thus, arsenic integrated absorbance growths as proton concentration increases, but only up to 5.8 M nitric acid solutions. Above this proton concentration value, atomic absorption of arsenic decreases sharply, maybe as a result of deterioration of the graphite tube by nitrate anion. The 5.8 M nitric acid acts as a digestion agent breaking or destroying organic compounds of the biomass and releasing in part arsenic ions under ultrasonic agitation. The sensitivity of the determination is an average of 2.5 times higher when the slurry is prepared with 5.8 M nitric acid in comparison with the water-prepared slurry, suggesting that the nitric acid solution favours mainly the atomization rate of arsenic. For that reason, the aforementioned nitric acid concentration was used in all experiments. Arsenic was finally determined by the slurry sampling because integrated absorbance for this element was higher than that derived from the supernatant (the nitric acid washing solution) after separation by centrifugation. These results can be explained on the basis that arsenic species remain scavenged (at least partly) by biomass even after being treated with an acidic solution, or alternatively that biomass favours atomization of arsenic much more at relatively high nitric acid concentrations. The washing time over intervals up to 24 h (using 5.8 M nitric acid every time) did not show a strong effect on the atomic absorption signal of arsenic (50 ng ml1) by ETAAS. Hence, a washing time interval of 20 minutes was selected in all experiments. An additional advantage derived from the use of the slurry sampling procedure was to follow the retention of arsenic by bacteria in a direct way. Different elements could compete with arsenic on the retention process at the wall sites showing interferent effects on the final stage of the determination of arsenic by ETAAS, but the most plausible competitors for arsenic in the bacterial uptake processes, Bi(III) and Sb(III), did not interfere (o5%) on the As ETAAS when present in a 100-fold concentration.
Calibration lines were derived from As-biomass slurries with enough sensitivity to be used for analytical purposes. The slope derived for this calibration line of such situation was equal to 6.05 104 s (ng ml1)1 (r ¼ 0.9989). This means that the characteristic mass was 72.7 pg for 10 ml of the sample volume injected.
3.2.
Optimization of the retention system
A comparison of the efficiency of four bacterial strains, C. glutamicum 13032 (wild type), C. glutamicum ArsB1, C. glutamicum ArsB2 and C. glutamicum ArsB1-B2, for the retention of As(III) and As(V) was performed (Fig. 3). In general, the genetically modified C. glutamicum bacteria showed higher retention capacity for both arsenic species, As(III) and As(V), than that derived with the wild-type strain. However, as a general rule and considering accumulation of arsenic as the analytical parameter, the best results, obtained as determined by ETAAS, were followed by the double-permease mutant (Fig. 3). The efficiency in the retention of arsenic by the mutant C. glutamicum ArsB1-B2 was particularly suited for As(III), which under the usual growth conditions showed a 35 times improved retention capacity compared to the wild-type strain. In conclusion, the following experiments were particularly stressed on the double-permease mutant. The effect of the growth/contact time on the retention of As(III) and As(V) at several concentrations mixed together with the liquid MMC is shown in Fig. 4 The maximum retention level of As(III) by the mutant C. glutamicum ArsB1B2, followed by discrete slurry sampling ETAAS determinations of arsenic, occurred after a growth/contact time of 5 h (Fig. 4A). The inserted derivative plot in Fig. 4 confirms this conclusion because it is in the initial stage of the growth process where the uptake rate shows the highest values. It may be concluded that arsenic-binding sites became saturated during the initial 5 h. Results for As(V) accumulation showed a poor retention capacity using the standard growth conditions (Fig. 4B). As a result of this, only the retention
ARTICLE IN PRESS 536
WA T E R R E S E A R C H
7
ng As / mg biomass
6
41 (2007) 531– 542
C. glutamicum ArsB1-B2
As(III) As(V)
5 4 C. glutamicum ArsB1 C. glutamicum ArsB2
3 2 1
C. glutamicum 13032 0 1
2
3
4
Bacterial strains Fig. 3 – Relative amounts of arsenic retained by four Corynebacterium strains from a liquid MMC standard culture containing 100 ng ml1 of As(III) and As(V).
process for As(III) was deeply evaluated. Nonetheless, we must to bear in mind that owing to the large number and assortment of different functional groups on the membrane surface, different types of adsorbent–adsorbate interactions may occur and any kinetics or mass transfer model is likely to be global.
3.3.
Retention kinetics
The retention of As(III) on the biomass may depend on two transport processes acting in parallel, i.e. film diffusion and intraparticle diffusion. In this regard, the Weber–Morris equation (Weber and Morris, 1963) was tested for kinetic data pffiffi (1) qt ¼ kd t, where qt is the amount of As(III) retained per mass unit of biomass at time t, and kd is the retention rate (ng mg1 h1/2). The values of the intraparticle diffusion rate (kd) for As(III) (Table 2) were derived from the plot of qt versus t1/2 for two concentration ranges. The plot for the lower concentration range was linear, passing very near through the origin, which means that intraparticle diffusion was one process taking relatively long contact time, while film diffusion contributes in a much smaller extension to the overall rate of transport. In this regard, film diffusion would represent an instantaneous retention stage for low concentrations, but not for increasing amount of analyte. Kinetic data were also treated with the pseudo-first-order kinetic model (Lagergren, 1898) using the integrated law for a pseudo-first-order reaction, which was rearranged to obtain the linear form logðqe qt Þ ¼ log qe
k1 t, 2:303
(2)
where qe and qt refer to the amount of arsenic retained at equilibrium and at any time t, respectively, and k1 is the rate
constant of pseudo-first-order sorption. Values of the rate constant, k1, and equilibrium retention capacity, qe, (Table 2) were calculated from the plots of log(qe–qt) versus t, for arsenic aqueous solutions. The heterogeneous nature of the active sites onto the biosorbent suggests that a metal retention mechanism with an order greater than one may occur. Then, the kinetic data of retention of arsenic were modelled (Fig. 5) using the pseudosecond-order rate equation developed by Ho and McKay (1998). The kinetic rate equation is dq1 ¼ k2 ðqe qt Þ2 , dt
(3)
where k2 (ng mg1 h1) is the rate constant of retention and qe and qt have the same meaning as above. So the rate of arsenic retention is proportional to the square of the number of remaining sites. Integrating this equation for the boundary conditions t ¼ 0 (qt ¼ 0) to te (qt ¼ qe) and then linearizing, it leads to t 1 1 ¼ þ t. qt k2 q2e qe
(4)
By plotting t/qt versus t, a straight line is obtained allowing the determination of qe and k2 (Table 2). The initial retention rate n0 (n0 ¼ k2 q2e ) was equal to 46.88 ng mg1 h1, while the half-retention time, t1=2 ¼ 1=k2 qe , at which the half of the maximum retention is reached was 0.23 h. This fast sorption kinetics is a key point when effluent treatment systems are designed. Success with the pseudo-second-order kinetics model suggests chemisorption as one of the rate-limiting steps (Ho and McKay, 2000). The correlation coefficient for the pseudo-first-order kinetic model was the lowest from the three kinetic models under study, suggesting that the retention system is better described by the pseudo-secondorder kinetic model, which provides a good correlation for the
ARTICLE IN PRESS
A 11
5
10 25 50 100 200
9 ng As/mg biomass
8
200 ng/ml
4 d(ng As/mg biomass)/dt
10
7
537
41 (20 07) 53 1 – 542
WAT E R R E S E A R C H
As(III)
3 2 100 ng/ml 1 0
50 ng/ml 25 ng/ml
10 ng/ml
−1
6
−2
5
0
2
4
6 Time, h
8
10
12
4 3 2 1 0 0
B 11
6 Growth time, h
8
0.5
8
10
12
50 ng/ml
0.4 0.3 d(ng As/mg biomass)/dt
9 ng As/mg biomass
4
10 25 50 100 200
10
7
2
As(V)
200 ng/ml
0.2 0.1
25 ng/ml
0.0 10 ng/ml
−0.1 −0.2
100 ng/ml
−0.3 −0.4
6
−0.5
5
0
2
4
6 Time, h
8
10
12
4 3 2 1 0 0
2
4
6 Growth time, h
8
10
12
Fig. 4 – Effect of the growth time on the retention of arsenic by the C. glutamicum ArsB1-B2 mutant strain for several concentrations of (A): As(III) and (B): As(V), 10, 25, 50, 100 and 200 ng ml1. The inserted figures represent the corresponding derivative plots.
Table 2 – Kinetics models and parameters Model Intraparticle diffusion Pseudo-first-order Pseudo-second-order
Rate constant
qe, ng mg1
r
kd, ng mg1 h1/2 (3.11–8.90) k1, h1 (1.748) k2, mg ng1 h1 (0.397)
10.70 10.87
0.9950 0.9789 0.9907
whole retention process. The intraparticle diffusion model fits better for an initial period of the retention process. Hence, it seems that the intraparticle diffusion would be rate limiting, followed by the pseudo-second-order kinetic model.
3.4.
Retention isotherms
The amount (q, in ng As mg1 bacteria) of the arsenic retained per mass unit (dry weight) of the bacterial cells increases with
ARTICLE IN PRESS 538
WA T E R R E S E A R C H
41 (2007) 531– 542
A 10
10
8 qt,ngmg−1
qe, ng mg−1
8
6
6 4 2
4
Weber-Morris equation
0 0.0
0.4
0.8
1.2
1.6
Time1/2, h1/2
2
Pseudo-second-order model Experimental results Pseudo-first-order model
0 0.0
B
0.5
1.0
2.5 Time, h
2.0
2.5
3.0
14 12
qe, ng mg−1
10 8 6 4
Langmuir isotherm Experimental results
2
Freundlich isotherm Dubinin-Radushkevich isotherm
0 0
20
40
60
80
100
120
Ce, ng ml−1 Fig. 5 – Plot of (A) the retained amount of As(III) versus time and (B) the three retention isotherms obtained using the linear method for the retention of As(III) onto living bacterial cells. The insert on (A) represents the Weber–Morris plot.
the concentration (C0, in ng As ml1) of arsenic in the original sample solution, showing the following relationship, q ¼ 0.003+0.109 C0, in the arsenic concentration range 0–50 ng ml1 (r ¼ 0.9989). This relationship means that the living bacteria could be successfully used as extractant from the analytical point of view. In order to obtain additional data to provide explanations about the retention mechanism, the following experiments were carried out. We used three isotherms, Langmuir, Freundlich and Dubin–Rudhesevich, which tend to mimic, respectively, three different retention processes: (i) monolayer retention, (ii) sorption on a heterogeneous surface, not limited by monolayer capacity and usually showing maximum retention for dilute solutions, and (iii) adsorption on microporous solids for which maximum adsorption capacity is usually analogous to monolayer capacity. Consequently, the Langmuir theoretical
model was used for the estimation of the maximum adsorption capacity corresponding to complete monolayer coverage on the biomass surface. The basic assumption of the Langmuir equation is a homogeneous surface of the outer membrane, averaging rather than specific description of the retention characterization. The Freundlich empirical model was chosen to estimate the adsorption intensity of the sorbate on the heterogeneous sorbent surface. On the other hand, the Dubinin–Radushkevich (DR) model was used to estimate the characteristic porosity, the maximum adsorption capacity, and the apparent free energy of adsorption. The retention isotherm of As(III) for C. glutamicum ArsB1-B2 (Fig. 5) suggest that the uptake of As(III) shows a Langmuirian shape, being fairly well fitted to a two-site Langmuir model, at least for a wider concentration range of As(III). That is, binding of As(III) decreases as As(III) loading of the bacterial
ARTICLE IN PRESS WAT E R R E S E A R C H
mass increases, which means that As(III) bind first to the ligands with the highest affinity and subsequently to those of lesser activity. It is also possible that for increasing analyte bacterial mass loads, the cell modifies the As(III) binding in a less extension or even a few of the earlier bound As(III) may be leached again. However, the results obtained, assuming a one-site Langmuirian sorption, can also explain the retention of As(III) with reasonable precision (Fig. 5). Thus, from the retention results of the bacterial cells with increments of As(III), we may derive equilibrium constant, KL, assuming a Langmuir linear plot (Langmuir, 1918) using the following equation: 1 1 1 ¼ þ , qe qm KL Ce qm
(5)
where the term qe represents the amount of analyte bound to bacteria (ng mg1), and Ce is the equilibrium concentration (ng ml1), while KL is the instability equilibrium constant (ml ng1) of the retention process, and qm is the amount of metal retained at saturation (ng mg1). The equilibrium constant, KL, obtained from the slope of the plot 1/qe versus 1/Ce in the Langmuir isotherm changes slightly with the concentration of arsenic. Such equilibrium constants for As(III) in the concentration range 10–200 ng ml1 are shown in Table 3, suggesting that the affinity of the bacterial cells with As(III) is relatively large, but lower than that for other metals and bacteria (Aller et al., 1996; Aller and Robles, 1998). Nonetheless, in order to predict the retention efficiency of the process, the dimensionless equilibrium parameter (RL) (Hall et al., 1966) was determined from Langmuir’s isotherm using the following equation: 1 , (6) RL ¼ 1 þ KL C0 where C0 is the As(III) concentration (ng ml1) in the initial sample solution and KL is the Langmuir isotherm constant. All values of RL were found to be lower than unity and in the range 0.05–0.5, depending on the initial concentration of arsenic, which means favourable retention. Results were also derived by Freundlich’s isotherm (Freundlich, 1906), using the following Freundlich’s equation: 1 log Ce , (7) log qe ¼ log KF þ n where KF and 1/n are the constants representing the retention capacity (mg g1) and intensity of the retention, respectively, with the other terms having the same meaning as above. The reported values for KF and 1/n (Table 3) were obtained from the slope and intercept of the plot of log qe versus log Ce.
539
41 (20 07) 53 1 – 542
Langmuir and Freundlich isotherms do not provide information about the retention type, and equilibrium data were also applied to the DR isotherm (Dubinin and Radushkevich, 1947), for which the amount retained is usually assumed to be a Gaussian function of the Polanyi potential (Hasany and Saeed, 1992), Lnqe ¼ Lnqm k2
(8)
with 1 ðPolanyi potentialÞ ¼ RT Ln 1 þ Ce
(9)
where, k is a constant related to retention energy, R the gas constant (kJ mol1 K1), T the absolute temperature (K), and assuming the prevalent meaning for the other terms. The DR isotherm constants, k and qm, were calculated from the slope and intercept of the plot of Ln qe versus e2 (Table 3). One additional advantage of the DR isotherm is that the k parameter can be used to estimate an average free energy value unique to the analyte–sorbent system. The mean free energy of retention (E), defined as the free energy change when 1 mol of ion is transferred from infinity in solution to the surface of the cellular membrane, was calculated from the k-value using the following equation (Hobson, 1969): 1 E ¼ pffiffiffiffiffiffi . 2k
(10)
It is known that magnitude of E is useful for estimating the retention type. Thus, if values for E are between 8 and 16 kJ mol1, the retention type can be explained by an ion exchange (chemisorption) process (Saeed et al., 2003). However, the calculated value for E ( ¼ 1.45 kJ mol1) in this study is within the energy range of physical retention (Eo8 kJ mol1) mainly attributed to the weak van der Waals forces. The results from Table 3 show that retention data fit reasonably well in the three models, but for a different analyte concentration range. In any case, a better fit in terms of r was obtained with Langmuir isotherm, giving a better representation of the experimental isotherm data in the whole concentration range studied. This may be due to the relatively low analyte concentrations used, which fit better to the monolayer condition generally recognized as for the Langmuir model. The decrease in the curvature of the isotherm, tending to a monolayer considerably increasing the Ce values for a small increase in qe, may be explained as due to the less active sites being available at the end of the adsorption process, arsenic species partially covering the surface sites. According to the DR isotherm, the relatively low qm value (Table 3) indicates that the biomass had a small
Table 3 – Retention isotherms and parameters Isotherm
qm, ng mg1
Isotherm constant
r
Langmuir Freundlich
15.4
0.9926 0.9865
Dubinin–Radushkevich
5.75
KL ¼ 0.10 KF ¼ 1.53 ng mg1 n ¼ 2.39 k ¼ 0.237 mol kJ2
0.9798
ARTICLE IN PRESS 540
WA T E R R E S E A R C H
affinity for arsenic. Hence, sorption of arsenic by biomass may not be significant also assuming that the porosity factor, k, was found to be less than unity. The correlation coefficient for Freundlich equation takes an intermediate value between than the other two isotherms. The KF value hardly above unity suggests a favourable but not great retention tendency towards the biomass. The value of 1/n was found to be smaller than unity indicating that no significant adsorption may take place. The isotherms rise sharply in the initial stage for low Ce and qe values, which indicates that there are plenty of readily accessible sites, but eventually a plateau is reached, particularly for the DR isotherm, indicating that the adsorbent (porosity in the DR case) is saturated at this level. The above results suggest that in the arsenic concentration range studied, the living cells of the mutant C. glutamicum ArsB1-B2 are long way from the saturation relating to their capacity of assimilation by metabolic processes. So, we deduced the maximum effective retention capacity, defined
41 (2007) 531– 542
as the amount of As(III) retained by mass unit of the bacterial cells corresponding to the maximum atomic absorption signal of As(III) when it is introduced as slurry. The maximum retention of arsenic was about 12 ng of arsenic per mg of bacteria mass (dry weight) for the best experimental conditions. This value is a little smaller than that derived from the Langmuir isotherm for the amount adsorbed at saturation, and higher than those deduced from the Freundlich and DR isotherms. In order to investigate the stability constants of binding sites of the outer membrane with As(III) species, the Scatchard equation (Scatchard, 1949) was also tested, written in the following form: qe ¼ KS ðqm qe Þ, Ce
(11)
where KS is the stability constant of As(III) species on the biomass, and the other terms taking the current meaning.
A 14 As(V) 1/20 As(V) 1/10 As(V) 1/1
ng As / mg biomass
12 10 8 6 4 2 0 0
50
100 As concentration, ng ml−1
150
200
100 As concentration, ng ml-1
150
200
B 14 As(III) 1/1 As(III) 1/10 As(III) 1/20
ng As / mg biomass
12 10 8 6 4 2 0 0
50
Fig. 6 – Effect of the amount of phosphate present in the liquid MMC on the retention of several concentrations of (A) As(V) and (B) As(III) by the mutant C. glutamicum ArsB1-B2. Concentration of phosphate was as in the standard culture medium (1/1), and diluted 1:10 (1/10) and 1:20 (1/20) times.
ARTICLE IN PRESS WAT E R R E S E A R C H
The plot of qe/Ce versus qe showed a curvature, which means that more than one class of complexes has been formed and each complex has its own unique formation constant, KS. The resultant curve was suitably resolved into two linear portions that are obviously due to the presence of two different types of binding site on biomass, which agrees with that found from Langmuir’s isotherm. The stability constants, KS1 ¼ 0.676 and KS2 ¼ 0.085, for the retained As(III) species were deduced from the respective slopes, whereas the values of the parameters qm1 ( ¼ 3.9670.05 ng mg1) and qm2 ( ¼ 10.7870.05 ng mg1) were derived from the corresponding intercepts. The values of the stability constants indicate that both bonding sites of biomass are not highly active towards As(III) species. Since this model involves only macroscopic parameters, therefore, it does not give any information on the molecular structure of the retained complex and only provides apparent conditional stability constants, which are not in accord with either the stoichiometric or molecular configuration.
3.5.
Having a look at speciation
Fig. 3 shows that the amount of As(III) retained by the mutant C. glutamicum ArsB1-B2 was higher than that of As(V). This means that this mutant could be used for speciation of As(III) in the presence of As(V), as for similar concentrations of both ions in the original sample solution As(III) was largely retained (15 fold) than As(V). However, as it was widely described that the uptake of As(V) follows the phosphate route (Cervantes et al., 1994), we evaluated the effect of the phosphate concentration of the culture medium on the retention of inorganic arsenic species by the mutant C. glutamicum ArsB1-B2 (Fig. 6). Fig. 6 shows that increased retention of As(V) by C. glutamicum ArsB1-B2 resulted from decreased phosphate concentrations in the culture media; however, retention of As(III) by this strain seems also to be slightly and inversely dependent on the phosphate concentration. Therefore, the mutant C. glutamicum ArsB1-B2 seems to be more adequate to preconcentrate As(III) than As(V) when cultured under standard conditions (liquid MMC). However, this suitability would be taken turns by changing the growth medium composition, and both inorganic arsenic species could be simultaneously retained by selecting properly the culture medium.
4.
Conclusion
The retention of arsenic ions from aqueous solutions using double mutants of coryneforms was studied here. The above results made evident that retention, and even speciation, of arsenite and arsenate can be possible by a proper selection of both the bacterial strain and the composition of the culture medium. The equilibrium retention data were well described by Langmuir isotherm. The intraparticle diffusion kinetics model was evidenced, although the retention rate can also be described satisfactorily by a pseudo-second-order model. Our results suggest that this bacteria type is a potential biomaterial removing efficiently arsenic from dilute wastewater streams. However, there still remain analytical challenges to
41 (20 07) 53 1 – 542
541
be solved to harness this class of bacteria in complex matrices. Nonetheless, the relative simplicity and low cost of this kind of material will surely contribute to further refinement of these new bio-extraction systems for analytical purposes.
Acknowledgements This work was supported by funding provided by grants from the Junta de Castilla y Leo´n, Consejerı´a de Educacio´n, Ref.: LE04/04, and the Spanish Ministerio de Educacio´n y Ciencia, Ref.: BIO2005-02723. One of the authors (MIM) is thankful to the Spanish Ministerio de Educacio´n y Ciencia for support. R E F E R E N C E S
Abdrashitova, S.A., Abdullina, G.G., Ilialetdinov, A.N., 1986. Role of arsenites in lipid peroxidation in Pseudomonas putida cells oxidizing arsenite. Mikrobiologiya 55, 212–216. Aller, A.J., Robles, L.C., 1998. Speciation of selenomethionine and selenourea using living bacterial cells. Analyst 123, 919–927. Aller, A.J., Lumbreras, J.M., Robles, L.C., Ferna´ndez, G.M., 1996. Stability of bacterium–mercury complexes and speciation of soluble inorganic mercury species. Anal. Chim. Acta 330, 89–105. Cai, J., DuBow, M.S., 1997. Use of a luminescent bacterial biosensor for biomonitoring and characterization of arsenic toxicity of chromated copper arsenate (CCA). Biodegradation 8, 105–111. Cervantes, C., Ji, G., Ramirez, J.L., Silver, S., 1994. Resistance to arsenic compounds in microorganisms. FEMS Microbiol. Rev. 15, 355–367. Corbisier, P., Ji, G., Nuyts, G., Mergeay, M., Silver, S., 1993. luxAB gene fusions with the arsenic and cadmium resistance operons of Staphylococcus aureus plasmid pI258. FEMS Microbiol. Lett. 110, 231–238. Daus, B., Wennrich, R., Weiss, H., 2004. Sorption materials for arsenic removal from water: a comparative study. Water Res. 38, 2948–2954. Dubinin, M.M., Radushkevich, L.V., 1947. The equation of the characteristic curve of activated charcoal. Proc. Acad. Sci. USSR Phys. Chem. Sect. 55, 331–333. Duker, A.A., Carranza, E.J.M., Hale, M., 2003. Arsenic geochemistry and health. Environ. Int. 31, 631–641. Fattorini, D., Alonso-Hernandez, C.M., Diaz-Asencio, M., MunozCaravaca, A., Pannacciulli, F.G., Tangherlini, M., Regoli, F., 2004. Chemical speciation of arsenic in different marine organisms: importance in monitoring studies. Mar. Environ. Res. 58, 845–850. ¨ ber die adsorption in lo¨sungen. Z. Phys. Freundlich, H., 1906. U Chem. (Leipzig) 57A, 385–470. Hall, K.R., Eagleton, L.C., Acrivos, A., Vermeulen, T., 1966. Poreand solid-diffusion kinetics in fixed-bed adsorption under constant-pattern conditions. Ind. Eng. Chem. Fund. 5 (2), 212–223. Hasany, S.M., Saeed, M.M., 1992. A kinetic and thermodynamic study of silver sorption onto manganese dioxide from acid solutions. Sep. Sci. Technol. 27, 1789–1800. Ho, Y.S., McKay, G., 1998. Sorption of dye from aqueous solution by peat. Chem. Eng. J. 70, 115–124. Ho, Y.S., McKay, G., 2000. Batch sorber design using equilibrium and control time data for the removal of lead. Air Soil Pollut. 124, 141–153.
ARTICLE IN PRESS 542
WA T E R R E S E A R C H
Hobson, J.P., 1969. Physical adsorption isotherms extending from ultrahigh vacuum to vapor pressure. J. Phys. Chem. 73, 2720–2727. Jain, C.K., Ali, I., 2000. Arsenic: occurrence, toxicity and speciation techniques. Water Res. 34, 4304–4312. Ji, G., Silver, S., 1992. Regulation and expression of the arsenic resistance operon from Staphylococcus aureus plasmid pI258. J. Bacteriol. 174, 3684–3694. Kaneko, H., Sakaguchi, K., 1979. Fusion of protoplasts and genetic recombination of Brevibacterium flavum. Agric. Biol. Chem. 43, 1007–1013. Kostal, J., Yang, R., Wu, C.H., Mulchandani, A., Chen, W., 2004. Enhanced arsenic accumulation in engineered bacterial cells expressing ArsR. Appl. Environ. Microbiol. 70, 4582–4587. Lagergren, S., 1898. About the theory of so-called adsorption of soluble substances. K. Seven. Vetenskapsakad. Handl. Band 24, 1–39. Langdon, C.J., Piearce, T.G., Meharg, A.A., Semple, K.T., 2003. Interactions between earthworms and arsenic in the soil environment: a review. Environ. Pollut. 124, 361–373. Langmuir, I., 1918. The adsorption of gases on plane surfaces of glass, mica and platinum. J. Am. Chem. Soc. 40, 1361–1403. Lin, T.F., Wu, J.K., 2001. Adsorption of arsenite and arsenate within activated alumina grains: equilibrium and kinetics. Water Res. 35, 2049–2057. Maity, S., Chakravarty, S., Bhattacharjee, S., Roy, B.C., 2005. A study on arsenic adsorption on polymetallic sea nodule in aqueous medium. Water Res. 39, 2579–2590. Mateos, L.M., Schafer, A., Kalinowski, J., Martin, J.F., Puhler, A., 1996. Integration of narrow-host-range vectors from Escherichia coli into the genomes of amino acid-producing corynebacteria after intergeneric conjugation. J. Bacteriol. 178, 5768–5775. Oliveira, V., Go´mez-Ariza, J.L., Sa´nchez-Rodas, D., 2005. Extraction procedures for chemical speciation of arsenic in atmospheric total suspended particles. Anal. Bioanal. Chem. 382, 335–340. Ordo´n˜ez, E., Letek, M., Valbuena, N., Gil, J.A., Mateos, L.M., 2005. Analysis of genes involved in arsenic resistance in corynebacterium glutamicum ATCC 13032. Appl. Environ. Microbiol. 71, 6206–6215.
41 (2007) 531– 542
Peta¨nen, T., Romantschuk, M., 2002. Use of bioluminescent bacterial sensors as an alternative method for measuring heavy metals in soil extracts. Anal. Chim. Acta 456, 55–61. Peta¨nen, T., Romantschuk, M., 2003. Toxicity and bioavailability to bacteria of particle-associated arsenite and mercury. Chemosphere 50, 409–413. Peta¨nen, T., Virta, M., Karp, M., Romantschuk, M., 2001. Construction and use of broad host range mercury and arsenite biosensor plasmids in the soil bacterium Pseudomonas fluorescens OS8. Microb. Ecol. 41, 360–368. Ramanathan, S., Shi, W., Rosen, B.P., Daunert, S., 1997. Sensing antimonite and arsenite at the subattomole level with genetically engineered bioluminescent bacteria. Anal. Chem. 69, 3380–3384. Roberto, F.F., Barnes, J.M., Bruhn, D.F., 2002. Evaluation of a GFP reporter gene construct for environmental arsenic detection. Talanta 58, 181–188. Rosen, A.L., Hieftje, G.M., 2004. Inductively coupled plasma mass spectrometry and electrospray mass spectrometry for speciation analysis: applications and instrumentation. Spectrochim. Acta B 59, 135–146. Saeed, M.M., Ahmed, M., Chaudary, M.H., Gaffar, A., 2003. Kinetics, thermodynamics, and sorption profile of Eu(III) and Tm(III) on 4-(2-pyridylazo) resorcinol (PAR) imbedded polyurethane foam. Solvent Extr. Ion Exchange 21, 881–898. Scatchard, G., 1949. The attractions of proteins for small molecules and ions. Ann. NY Acad. Sci. 51, 660–672. Scott, D.L., Ramanathan, S., Shi, W., Rosen, B.P., Daunert, S., 1997. Genetically engineered bacteria: electrochemical sensing systems for antimonite and arsenite. Anal. Chem. 69, 16–20. Silver, S., Phung, L.T., 1996. Bacterial heavy metal resistance: new surprises. Annu. Rev. Microbiol. 50, 753–789. Tauriainen, S., Karp, M., Chang, W., Virta, M., 1997. Recombinant luminescent bacteria for measuring bioavailable arsenite and antimonite. Appl. Environ. Microbiol. 63, 4456–4461. Tauriainen, A., Virta, M., Chang, W., Karp, M., 1999. Measurement of firefly luciferase reporter gene activity from cells and lysates using E. coli arsenite and mercury sensors. Anal. Biochem. 272, 191–198. Weber, W.J., Morris, S.C., 1963. Kinetics of adsorption on carbon from solution. J. Sanit. Eng. Div. Am. Soc. Civ. Eng. 89, 31–60.
41 (2007) 531 – 542
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Retention of inorganic arsenic by coryneform mutant strains J.C. Feoa, E. Ordon˜ezb, M. Letekb, M.A. Castroa, M.I. Mun˜oza, J.A. Gilb, L.M. Mateosb, A.J. Allera, a
Department of Biochemistry, Area of Analytical Chemistry, Leo´n, Spain Department of Ecology, Genetic and Microbiology, Area of Microbiology; University of Leo´n, E 24071 Leo´n, Spain
b
art i cle info
ab st rac t
Article history:
The natural resistance mechanisms of corynebacteria to respond to the environments
Received 12 June 2006
containing high levels of arsenic were successfully adopted to develop inexpensive and
Received in revised form
selective extractants for submicrogram amounts of arsenic. Kinetic and equilibrium
19 October 2006
characteristics were evaluated, and a preliminary exploration of the capability of these
Accepted 7 November 2006
strains to be used for arsenic speciation was also made in this work. Three kinetics models were used to fit the experimental data. It was found that the pseudo-first-order kinetics
Keywords:
model was not quite adequate to describe the retention process, while the intraparticle
Arsenate
diffusion and the pseudo-second-order kinetics models provide the best fits. The
Arsenite
equilibrium isotherm showed that the retention of arsenic was consistent with the
Corynebacteria
Langmuir equation and that the Freundlich and Dubinin–Radushkevich models provided
Mutant strains
poorer fits to the experimental data. The maximum effective retention capacity for arsenic
Isotherm
was about 15.4 ng As/mg biomass. The amount of arsenic retained was directly measured
Kinetics
in the biomass by forward planning a slurry electrothermal atomic absorption spectro-
Electrothermal atomic absorption
metric procedure.
spectrometry
& 2006 Elsevier Ltd. All rights reserved.
Slurry sampling
1.
Introduction
Heavy metal contamination of natural water is a major problem in industrialised areas, as they can cause a detrimental effect on environment and finally on human health. Arsenic is categorised as a heavy metal and comes into the environment as a result of the contribution of several natural sources, but mainly from human intervention, usually through mining, industrial, agricultural, medicinal, and like other activities (Langdon et al., 2003). The adverse health effects arsenic can produce in humans are well known (Duker et al., 2003). However, toxicity of arsenic differs with the chemical forms involved, resulting in a great challenge for arsenic speciation. For a better management of contaminaCorresponding author. Tel.: +34 987 291536; fax: +34 987 291883.
E-mail address: [email protected] (A.J. Aller). 0043-1354/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.watres.2006.11.015
tion of this metal, an adequate monitoring and control of the arsenic species levels are required. The most advantageous large-scale continuous operations for the separation of metals are liquid–liquid extraction and solid retention (Maity et al., 2005; Daus et al., 2004; Lin and Wu, 2001). More recently, however, many efforts have been addressed on the development of new sorbents, where bacteria-based biosorption has been considered a suitable wastewater technology to remove efficiently heavy metals. In other fields, such as speciation analyses, diverse analytical instrumentation has generally been used (Oliveira et al., 2005; Fattorini et al., 2004; Rosen and Hieftje, 2004; Jain and Ali, 2000), but microorganisms, particularly bacteria and mainly those evolving resistance mechanisms to many toxic metals
ARTICLE IN PRESS 532
WA T E R R E S E A R C H
Nomenclature
C0 Ce E KF KL KS kd k1 k2
concentration of arsenic in the initial sample solution (ng ml1) concentration of arsenic in the sample solution at equilibrium (ng ml1) mean free energy of retention (kJ mol1) retention capacity (Freundlich isotherm constant) (mg g1) instability equilibrium constant (Langmuir isotherm constant) (ml ng1) stability equilibrium constant Intraparticle diffusion rate (ng ml1 h1/2) the equilibrium rate constant of pseudo-firstorder sorption (h1) the equilibrium rate constant of pseudo-secondorder sorption (mg ng1 h1)
(Silver and Phung, 1996), have also been considered for preconcentration and/or separation of diverse metal species. The strategies developed by bacteria to circumvent the toxicity of arsenic usually include: (i) decreasing in the uptake of arsenate through the phosphate uptake system (Cervantes et al., 1994), (ii) evolving peroxidation reactions through membrane lipids (Abdrashitova et al., 1986), and (iii) using the best characterized microbial arsenic detoxification pathway, which involves genetic determinants typically organized in the ars operon (Silver and Phung, 1996). The ars operon is generally constituted of either three (arsRBC) or five (arsRDABC) genes, organized into a single transcriptional unit. The three-gene system encodes the arsenic transcriptional repressor (arsR), an arsenite permease, (arsB), a membranelocated arsenite efflux pump, and an arsenate reductase (arsC), which converts arsenate to arsenite before arsenic is pumped out of the cell through the ArsB anion pump; the five-genes operon (arsRDABC) encodes, besides the above described genes, a negative regulatory protein which provides additional fine tuning (arsD) and an arsenic-specific ATPase (arsA) (Silver and Phung, 1996). As a general rule, in bacteria, the interaction of arsenite with the repressor protein (ArsR) undergoes a conformational change of the protein and dissociates from the regulatory sequences of the ars operon, leading to the expression of downstream genes. Metal resistance mechanisms are clearly regulated by specific metalloregulatory proteins and the high affinity and specificity of the regulatory protein of the ars operon, ArsR, have been used for developing whole-cell bacterial biosensors for arsenic (Scott et al., 1997; Cai and DuBow, 1997; Ramanathan et al., 1997; Tauriainen et al., 1997), as well as for arsenic remediation (Kostal et al., 2004). In these studies, ArsR and other proteins, such as ArsB, have been fused to the bacterial luciferase genes (lux operon) (Cai and DuBow, 1997; Ramanathan et al., 1997; Roberto et al., 2002; Tauriainen et al., 1999) or to firefly luciferase (luc operon) (Tauriainen et al., 1997). Bioavailability studies of arsenic have also been tested using several bacteria-based biosensors (Tauriainen et al., 1997, 1999; Ji and Silver, 1992; Corbisier
41 (2007) 531– 542
k n qe qt qm RL R T t t1/2 e n0
constant related to retention energy (mol2 kJ2) intensity of the sorbent amount of As(III) retained per mass unit of biomass at equilibrium (ng mg1) amount of As(III) retained per mass unit of biomass at any time t (ng ml1) amount of As(III) retained at saturation (ng ml1) dimensionless equilibrium parameter the ideal gas constant (kJ mol1 K1) the absolute temperature (K) retention time (h) half-retention time (h) Polanyi potential initial retention rate (ng mg1 h1)
et al., 1993; Peta¨nen and Romantschuk, 2002). In general, the best-suited host strains are native bacteria from the environment for which the analysis is though (Peta¨nen and Romantschuk, 2003), although different recombinant plasmids have been used with similar successful results (Peta¨nen et al., 2001). The saprophytic soil bacterium Corynebacterium glutamicum is a Gram positive microorganism with a high resistance to arsenic; this resistance is associated to the presence of two arsenic detoxification operons (ars1 and ars2), which are basically structured following the gene arrangement arsR1arsB1-arsC1-arsC1’ for ars1 operon and arsR2-arsB2-arsC2 for ars2 (Ordo´n˜ez et al., 2005). The aim of this work was to evaluate the capability of four C. glutamicum strains, the wildtype strain (ATCC 13032) and three mutants generated by disruption of the arsenite permease genes (arsB1, arsB2 and both arsB1-arsB2) for the retention of As(III) and As(V) species. The biosorption process was followed by forward planning a slurry sampling living electrothermal atomic absorption spectrometry (ETAAS) procedure, which allow us to measure in a direct way the arsenic retained by the living bacterial biomass.
2.
Materials and methods
2.1.
Instruments and operating conditions
A Thermo Jarrel Ash atomic-absorption spectrophotometer (SH 11), equipped with a CTF-188 graphite atomizer and Smith–Hieftje background correction, was used for the measurements. The operating parameters used were as follows: wavelength, 193.7 nm; the arsenic pulsed hollow cathode lamp, Visimax II, was used under the recommended conditions; bandwidth, 1 nm. Standard uncoated rectangular graphite tubes, and standard pyrolytic graphite-coated graphite platforms were used for atomization. A syringe was used to inject manually 10 ml of the solutions and slurries into the graphite atomizer at room temperature. The temperature
ARTICLE IN PRESS WAT E R R E S E A R C H
533
41 (20 07) 53 1 – 542
Table 1 – Temperature programme Parameter
Dry
Pyrolysis 1
Pyrolysis 2
Atomization
Cleaning
Temperature, 1C Ramp time, s Hold time, s Argon flow Read on, s
150 2 0 Low
375 15 0 Medium
1100 15 0 Low
2100 0 4 Off 4
2500 0 0 High
programme of the atomiser is shown in Table 1. Argon, 99.995% purity, was used as the purge gas. At least three replicate determinations based on the integrated absorbances of the atomic absorption signals of arsenic were used for each measurement. A pH-meter (Crison model Digit 505) was used to measure the acidity of the aqueous phase when necessary. A Mettler AE 240 semi-micro analytical balance (sensitivity 70.01 mg) was used to weigh the chemicals and biomass. The slurry was maintained during sample introduction in the atomizer using a Brasonic sonicator (Model B-5).
2.2.
Chemicals
Arsenic stock solutions, 1000 mg l1, were prepared by dissolving suitable amounts of sodium arsenite and di-sodium hydrogen arsenate in an appropriate volume of de-mineralized water. Stock solutions of those elements used as chemical modifiers were prepared from Ni(NO3)2 and Pd(NO3)2, dissolved in de-mineralised water. Working solutions were prepared by the appropriate dilution of the stock solutions immediately prior to their use. The Tryptone Soy Agar (TSA; Oxoid) medium (40 g l1) was composed of tryptone (1.5% m/m), soy peptone (0.5% m/m), sodium chloride (0.5% m/m) and agar (1.5% m/m) and adjusted to pH 7.3 70.2; TSB medium (Tryptone Soy Broth) was the same than TSA but without agar. The composition of minimal medium for corynebacteria (MMC) was as follow: glucose (2 g l1), ammonium sulphate (10 g l1), urea (3 g l1), monopotassium phosphate (1 g l1), magnesium sulphate heptahydrate (0.41 g l1), sodium chloride (50 mg l1), glucose (2 g l1), iron(II) sulphate monohydrate (0.2 mg l1), manganese sulphate monohydrate (0.2 mg l1), biotin (5 mg l1) and tyamine (20 mg l1) for liquid MMC and containing 15% of agar for solid MMC (Kaneko and Sakaguchi, 1979). All media were sterilized by autoclaving for 20 min at 120 1C. All chemicals used in this study were of analytical reagent grade and were obtained from Merck. Distilled, deionized water (resistivity 18 M Ocm) was used for the preparation of the samples and standards. The pH was adjusted by using HNO3 and NaOH as necessary. The TAB buffer used for washing at the As retention analysis contains: HEPES (75 mM), potassium chloride (150 mM) and magnesium chloride (1 mM); the buffer was adjusted to pH 7.370.2.
2.3.
Strains and mutants
C. glutamicum ATCC 13032 was used as the type strain. Escherichia coli S17-1 was used for subcloning and as donor
strain in conjugation (Mateos et al., 1996). Bacterial cultures were grown in complex medium (TSA/TSB) for 24–48 h at 37 1C (E. coli) or 30 1C (corynebacteria). For preservation, cultures were stored at 4 1C until required for use. When necessary, antibiotics kanamycin and apramycin were added in cultures at final concentrations of 12.5 mg l1 for corynebacteria and 50 mg l1 for E. coli. Internal fragments from the C. glutamicum arsB genes were subcloned in E. coli vectors, and then transferred to corynebacteria by conjugation using standard protocols (Mateos et al., 1996). C. glutamicum transconjugants were selected on TSA medium containing 50 mg l1 of nalidixic acid and an additional antibiotic (kanamycin, or apramycin), depending on the mobilizable plasmid used for the mating. The single and double C. glutamicum arsenite permease mutants (ArsB1, ArsB2 and ArsB1-B2) were constructed as elsewhere (Ordo´n˜ez et al., 2005).
2.4.
Procedures
For the As retention procedure, the living bacterial cells (C. glutamicum strains) were cultivated in 50 ml of liquid MMC mixed with arsenic in the concentration range 0–200 ng ml1, using a 250-ml Erlenmeyer flasks for 5 h, as otherwise was noted, with continuous stirring (200 rev min1). The inoculated liquid cultures started with the same optical density (0.05) measured at l ¼ 598 nm (OD598 ¼ 0.05). After 5 h, the metal-biomass pellet was collected by centrifugation at 12 000 g, washed twice with TAB buffer, and treated with a 5.8 M nitric acid solution containing the optimum amount of palladium nitrate (2000 mg l1) as a chemical modifier. Then, maintaining this slurry by sonication, the amount of arsenic was determined by slurry sampling ETAAS. The concentration of arsenic in the sample can be evaluated from a calibration graph prepared by treating arsenic standards in the same way as samples via bacterial preconcentration.
3.
Results and discussion
3.1.
Optimization of the analytical procedure
As slurry ETAAS was though to be used for evaluating the retention process, those factors affecting the atomic absorption signal of arsenic injected as slurry were firstly optimized: the chemical modifier concentration, pryrolysis and atomization temperatures, and the slurry characteristics (the washing acid concentration and slurry concentration). All these experiments were carried out by using an As-biomass slurry,
ARTICLE IN PRESS 534
WA T E R R E S E A R C H
prepared after the retention of 50 ng ml1 of arsenic as As(III) by the C. glutamicum wild-type strain under standard culture conditions. Owing to the difficulties associated with the determination of arsenic by ETAAS in biological matrices, the use of a chemical modifier is mandatory. In order to optimize the suitable amount of the modifier, several studies working at the tentatively selected preliminary pyrolysis and atomization temperatures of 900 and 1800 1C, respectively, were carried out. The effect of the amount of two metals (Ni and Pd, as nitrates) on the stabilization of arsenic in the graphite tube was evaluated for platform atomization using the temperature programme shown in Table 1 and the results were shown in Fig. 1. From Fig. 1, it is possible to conclude that nickel and palladium concentrations of 2000 mg l1 in the final slurry were sufficient to stabilize the atomic absorption signal of arsenic. Nonetheless, the modifier selected for the subsequent comparative studies was palladium at the concentration of 2000 mg l1 in the final slurry. In order to find out the maximum allowable pyrolysis temperature for arsenic, introduced as a As-biomass slurry in the presence of palladium, a curve was deduced of the analyte signal (integrated absorbance) obtained at different temperatures for a pyrolysis step constituted of two pyrolysis temperatures. Pyrolysis 1 temperature was maintained fixed at 375 1C, while pyrolysis 2 temperature was modified from 400 to 1300 1C. In each case, the same atomization temperature (1800 1C) was used throughout. Pyrolysis curves were established for slurries prepared using biomass materials with and without modifier. Fig. 2 shows the effect of pyrolysis temperature on the atomic absorption signals of arsenic. When no modifier was used, arsenic was readily loss owing to its high volatility and the low absorption signal of arsenic decreased rapidly with pyrolysis temperature. By contrary, using chemical modifiers, higher and most stabilized arsenic atomic absorption signals were obtained. Nonetheless, as it is shown in Fig. 2, even in the presence of modifier, the atomic absorption signal of arsenic decreased rapidly at pyrolysis temperatures above 1100 1C. In conclusion, the pyrolysis
41 (2007) 531– 542
temperature value of 1100 1C was selected for further experiments. A second set of experiments exploring the relationship between the atomization temperature and the integrated absorbance of arsenic showed that the maximum and most stabilised atomic absorption signals (integrated absorbances) of arsenic were obtained at an atomization temperature of about 2100 1C when a 1100 1C pyrolysis 2 step was used. Using living cells, the removal of arsenic from the sample solution depends not only on the uptake rate but also on the growing rate of the bacterial cells, because the number of cells increases during cultivation and, consequently, the retention capacity for arsenic. However, this behaviour not necessarily should be observed in the atomic absorption signal of arsenic if analyte is introduced into the graphite tube as slurry. Biomass may affects on the atomization path of arsenic, but also it can alter the sample introduction stage by clogging the aspiration capillary. The amount of biomass injected into the tube affected on the atomic absorption signal of arsenic. Thus, the presence of organic material decreased the atomization efficiency of arsenic when the chemical modifier was absent, as the signals derived from arsenic aqueous solutions were slightly higher than those obtained for As-biomass slurry. However, in the presence of palladium as modifier, sensitivity was not only restored, but improved for the As-biomass slurries compared to Asaqueous solutions. Therefore, the bacterial mass favours atomization of arsenic, but only in the presence of palladium. The most likely explanation for this behaviour is that the organic matter may act as reducing agent for both palladium and arsenic, thus facilitating their interaction in reduced form. In consequence, the effect of slurry density was studied in the presence of the chemical modifier. For these situations, the atomic absorbance of arsenic increases gradually with the bacterial mass used in the retention process and, consequently, injected into the tube, but only till about 40 mg l1. Then, the atomic absorption signal of arsenic decreases suddenly. In conclusion, we have found that the best slurry density for further experiments was 40 mg l1.
Integrated absorbance. s
0.065 0.060 0.055 0.050 0.045 Palladium
0.040
Nickel 0.035 0
500
1000 1500 2000 Modifier concentration, mg l−1
2500
Fig. 1 – Effect of the modifier concentration on the integrated absorbance of arsenic.
ARTICLE IN PRESS WAT E R R E S E A R C H
41 (20 07) 53 1 – 542
535
Integrated absorbance,s
0.07 0.06 0.05 0.04 0.03 As-biomass As-biomass As-biomass-Pd As-biomass-Pd
0.02 0.01 400
600
800
1000 1200 1400 1600 1800 2000 2200 2400 Temperature,°C
Fig. 2 – Pyrolysis (solid symbols) and atomization (open symbols) temperatures for As-biomass slurry alone (squares) and As-biomass slurry together with the optimum amount of palladium as modifier (circles).
Nitric acid usually has an effect on the atomic absorption signals of arsenic, mainly if injected as an organic matrix. Thus, arsenic integrated absorbance growths as proton concentration increases, but only up to 5.8 M nitric acid solutions. Above this proton concentration value, atomic absorption of arsenic decreases sharply, maybe as a result of deterioration of the graphite tube by nitrate anion. The 5.8 M nitric acid acts as a digestion agent breaking or destroying organic compounds of the biomass and releasing in part arsenic ions under ultrasonic agitation. The sensitivity of the determination is an average of 2.5 times higher when the slurry is prepared with 5.8 M nitric acid in comparison with the water-prepared slurry, suggesting that the nitric acid solution favours mainly the atomization rate of arsenic. For that reason, the aforementioned nitric acid concentration was used in all experiments. Arsenic was finally determined by the slurry sampling because integrated absorbance for this element was higher than that derived from the supernatant (the nitric acid washing solution) after separation by centrifugation. These results can be explained on the basis that arsenic species remain scavenged (at least partly) by biomass even after being treated with an acidic solution, or alternatively that biomass favours atomization of arsenic much more at relatively high nitric acid concentrations. The washing time over intervals up to 24 h (using 5.8 M nitric acid every time) did not show a strong effect on the atomic absorption signal of arsenic (50 ng ml1) by ETAAS. Hence, a washing time interval of 20 minutes was selected in all experiments. An additional advantage derived from the use of the slurry sampling procedure was to follow the retention of arsenic by bacteria in a direct way. Different elements could compete with arsenic on the retention process at the wall sites showing interferent effects on the final stage of the determination of arsenic by ETAAS, but the most plausible competitors for arsenic in the bacterial uptake processes, Bi(III) and Sb(III), did not interfere (o5%) on the As ETAAS when present in a 100-fold concentration.
Calibration lines were derived from As-biomass slurries with enough sensitivity to be used for analytical purposes. The slope derived for this calibration line of such situation was equal to 6.05 104 s (ng ml1)1 (r ¼ 0.9989). This means that the characteristic mass was 72.7 pg for 10 ml of the sample volume injected.
3.2.
Optimization of the retention system
A comparison of the efficiency of four bacterial strains, C. glutamicum 13032 (wild type), C. glutamicum ArsB1, C. glutamicum ArsB2 and C. glutamicum ArsB1-B2, for the retention of As(III) and As(V) was performed (Fig. 3). In general, the genetically modified C. glutamicum bacteria showed higher retention capacity for both arsenic species, As(III) and As(V), than that derived with the wild-type strain. However, as a general rule and considering accumulation of arsenic as the analytical parameter, the best results, obtained as determined by ETAAS, were followed by the double-permease mutant (Fig. 3). The efficiency in the retention of arsenic by the mutant C. glutamicum ArsB1-B2 was particularly suited for As(III), which under the usual growth conditions showed a 35 times improved retention capacity compared to the wild-type strain. In conclusion, the following experiments were particularly stressed on the double-permease mutant. The effect of the growth/contact time on the retention of As(III) and As(V) at several concentrations mixed together with the liquid MMC is shown in Fig. 4 The maximum retention level of As(III) by the mutant C. glutamicum ArsB1B2, followed by discrete slurry sampling ETAAS determinations of arsenic, occurred after a growth/contact time of 5 h (Fig. 4A). The inserted derivative plot in Fig. 4 confirms this conclusion because it is in the initial stage of the growth process where the uptake rate shows the highest values. It may be concluded that arsenic-binding sites became saturated during the initial 5 h. Results for As(V) accumulation showed a poor retention capacity using the standard growth conditions (Fig. 4B). As a result of this, only the retention
ARTICLE IN PRESS 536
WA T E R R E S E A R C H
7
ng As / mg biomass
6
41 (2007) 531– 542
C. glutamicum ArsB1-B2
As(III) As(V)
5 4 C. glutamicum ArsB1 C. glutamicum ArsB2
3 2 1
C. glutamicum 13032 0 1
2
3
4
Bacterial strains Fig. 3 – Relative amounts of arsenic retained by four Corynebacterium strains from a liquid MMC standard culture containing 100 ng ml1 of As(III) and As(V).
process for As(III) was deeply evaluated. Nonetheless, we must to bear in mind that owing to the large number and assortment of different functional groups on the membrane surface, different types of adsorbent–adsorbate interactions may occur and any kinetics or mass transfer model is likely to be global.
3.3.
Retention kinetics
The retention of As(III) on the biomass may depend on two transport processes acting in parallel, i.e. film diffusion and intraparticle diffusion. In this regard, the Weber–Morris equation (Weber and Morris, 1963) was tested for kinetic data pffiffi (1) qt ¼ kd t, where qt is the amount of As(III) retained per mass unit of biomass at time t, and kd is the retention rate (ng mg1 h1/2). The values of the intraparticle diffusion rate (kd) for As(III) (Table 2) were derived from the plot of qt versus t1/2 for two concentration ranges. The plot for the lower concentration range was linear, passing very near through the origin, which means that intraparticle diffusion was one process taking relatively long contact time, while film diffusion contributes in a much smaller extension to the overall rate of transport. In this regard, film diffusion would represent an instantaneous retention stage for low concentrations, but not for increasing amount of analyte. Kinetic data were also treated with the pseudo-first-order kinetic model (Lagergren, 1898) using the integrated law for a pseudo-first-order reaction, which was rearranged to obtain the linear form logðqe qt Þ ¼ log qe
k1 t, 2:303
(2)
where qe and qt refer to the amount of arsenic retained at equilibrium and at any time t, respectively, and k1 is the rate
constant of pseudo-first-order sorption. Values of the rate constant, k1, and equilibrium retention capacity, qe, (Table 2) were calculated from the plots of log(qe–qt) versus t, for arsenic aqueous solutions. The heterogeneous nature of the active sites onto the biosorbent suggests that a metal retention mechanism with an order greater than one may occur. Then, the kinetic data of retention of arsenic were modelled (Fig. 5) using the pseudosecond-order rate equation developed by Ho and McKay (1998). The kinetic rate equation is dq1 ¼ k2 ðqe qt Þ2 , dt
(3)
where k2 (ng mg1 h1) is the rate constant of retention and qe and qt have the same meaning as above. So the rate of arsenic retention is proportional to the square of the number of remaining sites. Integrating this equation for the boundary conditions t ¼ 0 (qt ¼ 0) to te (qt ¼ qe) and then linearizing, it leads to t 1 1 ¼ þ t. qt k2 q2e qe
(4)
By plotting t/qt versus t, a straight line is obtained allowing the determination of qe and k2 (Table 2). The initial retention rate n0 (n0 ¼ k2 q2e ) was equal to 46.88 ng mg1 h1, while the half-retention time, t1=2 ¼ 1=k2 qe , at which the half of the maximum retention is reached was 0.23 h. This fast sorption kinetics is a key point when effluent treatment systems are designed. Success with the pseudo-second-order kinetics model suggests chemisorption as one of the rate-limiting steps (Ho and McKay, 2000). The correlation coefficient for the pseudo-first-order kinetic model was the lowest from the three kinetic models under study, suggesting that the retention system is better described by the pseudo-secondorder kinetic model, which provides a good correlation for the
ARTICLE IN PRESS
A 11
5
10 25 50 100 200
9 ng As/mg biomass
8
200 ng/ml
4 d(ng As/mg biomass)/dt
10
7
537
41 (20 07) 53 1 – 542
WAT E R R E S E A R C H
As(III)
3 2 100 ng/ml 1 0
50 ng/ml 25 ng/ml
10 ng/ml
−1
6
−2
5
0
2
4
6 Time, h
8
10
12
4 3 2 1 0 0
B 11
6 Growth time, h
8
0.5
8
10
12
50 ng/ml
0.4 0.3 d(ng As/mg biomass)/dt
9 ng As/mg biomass
4
10 25 50 100 200
10
7
2
As(V)
200 ng/ml
0.2 0.1
25 ng/ml
0.0 10 ng/ml
−0.1 −0.2
100 ng/ml
−0.3 −0.4
6
−0.5
5
0
2
4
6 Time, h
8
10
12
4 3 2 1 0 0
2
4
6 Growth time, h
8
10
12
Fig. 4 – Effect of the growth time on the retention of arsenic by the C. glutamicum ArsB1-B2 mutant strain for several concentrations of (A): As(III) and (B): As(V), 10, 25, 50, 100 and 200 ng ml1. The inserted figures represent the corresponding derivative plots.
Table 2 – Kinetics models and parameters Model Intraparticle diffusion Pseudo-first-order Pseudo-second-order
Rate constant
qe, ng mg1
r
kd, ng mg1 h1/2 (3.11–8.90) k1, h1 (1.748) k2, mg ng1 h1 (0.397)
10.70 10.87
0.9950 0.9789 0.9907
whole retention process. The intraparticle diffusion model fits better for an initial period of the retention process. Hence, it seems that the intraparticle diffusion would be rate limiting, followed by the pseudo-second-order kinetic model.
3.4.
Retention isotherms
The amount (q, in ng As mg1 bacteria) of the arsenic retained per mass unit (dry weight) of the bacterial cells increases with
ARTICLE IN PRESS 538
WA T E R R E S E A R C H
41 (2007) 531– 542
A 10
10
8 qt,ngmg−1
qe, ng mg−1
8
6
6 4 2
4
Weber-Morris equation
0 0.0
0.4
0.8
1.2
1.6
Time1/2, h1/2
2
Pseudo-second-order model Experimental results Pseudo-first-order model
0 0.0
B
0.5
1.0
2.5 Time, h
2.0
2.5
3.0
14 12
qe, ng mg−1
10 8 6 4
Langmuir isotherm Experimental results
2
Freundlich isotherm Dubinin-Radushkevich isotherm
0 0
20
40
60
80
100
120
Ce, ng ml−1 Fig. 5 – Plot of (A) the retained amount of As(III) versus time and (B) the three retention isotherms obtained using the linear method for the retention of As(III) onto living bacterial cells. The insert on (A) represents the Weber–Morris plot.
the concentration (C0, in ng As ml1) of arsenic in the original sample solution, showing the following relationship, q ¼ 0.003+0.109 C0, in the arsenic concentration range 0–50 ng ml1 (r ¼ 0.9989). This relationship means that the living bacteria could be successfully used as extractant from the analytical point of view. In order to obtain additional data to provide explanations about the retention mechanism, the following experiments were carried out. We used three isotherms, Langmuir, Freundlich and Dubin–Rudhesevich, which tend to mimic, respectively, three different retention processes: (i) monolayer retention, (ii) sorption on a heterogeneous surface, not limited by monolayer capacity and usually showing maximum retention for dilute solutions, and (iii) adsorption on microporous solids for which maximum adsorption capacity is usually analogous to monolayer capacity. Consequently, the Langmuir theoretical
model was used for the estimation of the maximum adsorption capacity corresponding to complete monolayer coverage on the biomass surface. The basic assumption of the Langmuir equation is a homogeneous surface of the outer membrane, averaging rather than specific description of the retention characterization. The Freundlich empirical model was chosen to estimate the adsorption intensity of the sorbate on the heterogeneous sorbent surface. On the other hand, the Dubinin–Radushkevich (DR) model was used to estimate the characteristic porosity, the maximum adsorption capacity, and the apparent free energy of adsorption. The retention isotherm of As(III) for C. glutamicum ArsB1-B2 (Fig. 5) suggest that the uptake of As(III) shows a Langmuirian shape, being fairly well fitted to a two-site Langmuir model, at least for a wider concentration range of As(III). That is, binding of As(III) decreases as As(III) loading of the bacterial
ARTICLE IN PRESS WAT E R R E S E A R C H
mass increases, which means that As(III) bind first to the ligands with the highest affinity and subsequently to those of lesser activity. It is also possible that for increasing analyte bacterial mass loads, the cell modifies the As(III) binding in a less extension or even a few of the earlier bound As(III) may be leached again. However, the results obtained, assuming a one-site Langmuirian sorption, can also explain the retention of As(III) with reasonable precision (Fig. 5). Thus, from the retention results of the bacterial cells with increments of As(III), we may derive equilibrium constant, KL, assuming a Langmuir linear plot (Langmuir, 1918) using the following equation: 1 1 1 ¼ þ , qe qm KL Ce qm
(5)
where the term qe represents the amount of analyte bound to bacteria (ng mg1), and Ce is the equilibrium concentration (ng ml1), while KL is the instability equilibrium constant (ml ng1) of the retention process, and qm is the amount of metal retained at saturation (ng mg1). The equilibrium constant, KL, obtained from the slope of the plot 1/qe versus 1/Ce in the Langmuir isotherm changes slightly with the concentration of arsenic. Such equilibrium constants for As(III) in the concentration range 10–200 ng ml1 are shown in Table 3, suggesting that the affinity of the bacterial cells with As(III) is relatively large, but lower than that for other metals and bacteria (Aller et al., 1996; Aller and Robles, 1998). Nonetheless, in order to predict the retention efficiency of the process, the dimensionless equilibrium parameter (RL) (Hall et al., 1966) was determined from Langmuir’s isotherm using the following equation: 1 , (6) RL ¼ 1 þ KL C0 where C0 is the As(III) concentration (ng ml1) in the initial sample solution and KL is the Langmuir isotherm constant. All values of RL were found to be lower than unity and in the range 0.05–0.5, depending on the initial concentration of arsenic, which means favourable retention. Results were also derived by Freundlich’s isotherm (Freundlich, 1906), using the following Freundlich’s equation: 1 log Ce , (7) log qe ¼ log KF þ n where KF and 1/n are the constants representing the retention capacity (mg g1) and intensity of the retention, respectively, with the other terms having the same meaning as above. The reported values for KF and 1/n (Table 3) were obtained from the slope and intercept of the plot of log qe versus log Ce.
539
41 (20 07) 53 1 – 542
Langmuir and Freundlich isotherms do not provide information about the retention type, and equilibrium data were also applied to the DR isotherm (Dubinin and Radushkevich, 1947), for which the amount retained is usually assumed to be a Gaussian function of the Polanyi potential (Hasany and Saeed, 1992), Lnqe ¼ Lnqm k2
(8)
with 1 ðPolanyi potentialÞ ¼ RT Ln 1 þ Ce
(9)
where, k is a constant related to retention energy, R the gas constant (kJ mol1 K1), T the absolute temperature (K), and assuming the prevalent meaning for the other terms. The DR isotherm constants, k and qm, were calculated from the slope and intercept of the plot of Ln qe versus e2 (Table 3). One additional advantage of the DR isotherm is that the k parameter can be used to estimate an average free energy value unique to the analyte–sorbent system. The mean free energy of retention (E), defined as the free energy change when 1 mol of ion is transferred from infinity in solution to the surface of the cellular membrane, was calculated from the k-value using the following equation (Hobson, 1969): 1 E ¼ pffiffiffiffiffiffi . 2k
(10)
It is known that magnitude of E is useful for estimating the retention type. Thus, if values for E are between 8 and 16 kJ mol1, the retention type can be explained by an ion exchange (chemisorption) process (Saeed et al., 2003). However, the calculated value for E ( ¼ 1.45 kJ mol1) in this study is within the energy range of physical retention (Eo8 kJ mol1) mainly attributed to the weak van der Waals forces. The results from Table 3 show that retention data fit reasonably well in the three models, but for a different analyte concentration range. In any case, a better fit in terms of r was obtained with Langmuir isotherm, giving a better representation of the experimental isotherm data in the whole concentration range studied. This may be due to the relatively low analyte concentrations used, which fit better to the monolayer condition generally recognized as for the Langmuir model. The decrease in the curvature of the isotherm, tending to a monolayer considerably increasing the Ce values for a small increase in qe, may be explained as due to the less active sites being available at the end of the adsorption process, arsenic species partially covering the surface sites. According to the DR isotherm, the relatively low qm value (Table 3) indicates that the biomass had a small
Table 3 – Retention isotherms and parameters Isotherm
qm, ng mg1
Isotherm constant
r
Langmuir Freundlich
15.4
0.9926 0.9865
Dubinin–Radushkevich
5.75
KL ¼ 0.10 KF ¼ 1.53 ng mg1 n ¼ 2.39 k ¼ 0.237 mol kJ2
0.9798
ARTICLE IN PRESS 540
WA T E R R E S E A R C H
affinity for arsenic. Hence, sorption of arsenic by biomass may not be significant also assuming that the porosity factor, k, was found to be less than unity. The correlation coefficient for Freundlich equation takes an intermediate value between than the other two isotherms. The KF value hardly above unity suggests a favourable but not great retention tendency towards the biomass. The value of 1/n was found to be smaller than unity indicating that no significant adsorption may take place. The isotherms rise sharply in the initial stage for low Ce and qe values, which indicates that there are plenty of readily accessible sites, but eventually a plateau is reached, particularly for the DR isotherm, indicating that the adsorbent (porosity in the DR case) is saturated at this level. The above results suggest that in the arsenic concentration range studied, the living cells of the mutant C. glutamicum ArsB1-B2 are long way from the saturation relating to their capacity of assimilation by metabolic processes. So, we deduced the maximum effective retention capacity, defined
41 (2007) 531– 542
as the amount of As(III) retained by mass unit of the bacterial cells corresponding to the maximum atomic absorption signal of As(III) when it is introduced as slurry. The maximum retention of arsenic was about 12 ng of arsenic per mg of bacteria mass (dry weight) for the best experimental conditions. This value is a little smaller than that derived from the Langmuir isotherm for the amount adsorbed at saturation, and higher than those deduced from the Freundlich and DR isotherms. In order to investigate the stability constants of binding sites of the outer membrane with As(III) species, the Scatchard equation (Scatchard, 1949) was also tested, written in the following form: qe ¼ KS ðqm qe Þ, Ce
(11)
where KS is the stability constant of As(III) species on the biomass, and the other terms taking the current meaning.
A 14 As(V) 1/20 As(V) 1/10 As(V) 1/1
ng As / mg biomass
12 10 8 6 4 2 0 0
50
100 As concentration, ng ml−1
150
200
100 As concentration, ng ml-1
150
200
B 14 As(III) 1/1 As(III) 1/10 As(III) 1/20
ng As / mg biomass
12 10 8 6 4 2 0 0
50
Fig. 6 – Effect of the amount of phosphate present in the liquid MMC on the retention of several concentrations of (A) As(V) and (B) As(III) by the mutant C. glutamicum ArsB1-B2. Concentration of phosphate was as in the standard culture medium (1/1), and diluted 1:10 (1/10) and 1:20 (1/20) times.
ARTICLE IN PRESS WAT E R R E S E A R C H
The plot of qe/Ce versus qe showed a curvature, which means that more than one class of complexes has been formed and each complex has its own unique formation constant, KS. The resultant curve was suitably resolved into two linear portions that are obviously due to the presence of two different types of binding site on biomass, which agrees with that found from Langmuir’s isotherm. The stability constants, KS1 ¼ 0.676 and KS2 ¼ 0.085, for the retained As(III) species were deduced from the respective slopes, whereas the values of the parameters qm1 ( ¼ 3.9670.05 ng mg1) and qm2 ( ¼ 10.7870.05 ng mg1) were derived from the corresponding intercepts. The values of the stability constants indicate that both bonding sites of biomass are not highly active towards As(III) species. Since this model involves only macroscopic parameters, therefore, it does not give any information on the molecular structure of the retained complex and only provides apparent conditional stability constants, which are not in accord with either the stoichiometric or molecular configuration.
3.5.
Having a look at speciation
Fig. 3 shows that the amount of As(III) retained by the mutant C. glutamicum ArsB1-B2 was higher than that of As(V). This means that this mutant could be used for speciation of As(III) in the presence of As(V), as for similar concentrations of both ions in the original sample solution As(III) was largely retained (15 fold) than As(V). However, as it was widely described that the uptake of As(V) follows the phosphate route (Cervantes et al., 1994), we evaluated the effect of the phosphate concentration of the culture medium on the retention of inorganic arsenic species by the mutant C. glutamicum ArsB1-B2 (Fig. 6). Fig. 6 shows that increased retention of As(V) by C. glutamicum ArsB1-B2 resulted from decreased phosphate concentrations in the culture media; however, retention of As(III) by this strain seems also to be slightly and inversely dependent on the phosphate concentration. Therefore, the mutant C. glutamicum ArsB1-B2 seems to be more adequate to preconcentrate As(III) than As(V) when cultured under standard conditions (liquid MMC). However, this suitability would be taken turns by changing the growth medium composition, and both inorganic arsenic species could be simultaneously retained by selecting properly the culture medium.
4.
Conclusion
The retention of arsenic ions from aqueous solutions using double mutants of coryneforms was studied here. The above results made evident that retention, and even speciation, of arsenite and arsenate can be possible by a proper selection of both the bacterial strain and the composition of the culture medium. The equilibrium retention data were well described by Langmuir isotherm. The intraparticle diffusion kinetics model was evidenced, although the retention rate can also be described satisfactorily by a pseudo-second-order model. Our results suggest that this bacteria type is a potential biomaterial removing efficiently arsenic from dilute wastewater streams. However, there still remain analytical challenges to
41 (20 07) 53 1 – 542
541
be solved to harness this class of bacteria in complex matrices. Nonetheless, the relative simplicity and low cost of this kind of material will surely contribute to further refinement of these new bio-extraction systems for analytical purposes.
Acknowledgements This work was supported by funding provided by grants from the Junta de Castilla y Leo´n, Consejerı´a de Educacio´n, Ref.: LE04/04, and the Spanish Ministerio de Educacio´n y Ciencia, Ref.: BIO2005-02723. One of the authors (MIM) is thankful to the Spanish Ministerio de Educacio´n y Ciencia for support. R E F E R E N C E S
Abdrashitova, S.A., Abdullina, G.G., Ilialetdinov, A.N., 1986. Role of arsenites in lipid peroxidation in Pseudomonas putida cells oxidizing arsenite. Mikrobiologiya 55, 212–216. Aller, A.J., Robles, L.C., 1998. Speciation of selenomethionine and selenourea using living bacterial cells. Analyst 123, 919–927. Aller, A.J., Lumbreras, J.M., Robles, L.C., Ferna´ndez, G.M., 1996. Stability of bacterium–mercury complexes and speciation of soluble inorganic mercury species. Anal. Chim. Acta 330, 89–105. Cai, J., DuBow, M.S., 1997. Use of a luminescent bacterial biosensor for biomonitoring and characterization of arsenic toxicity of chromated copper arsenate (CCA). Biodegradation 8, 105–111. Cervantes, C., Ji, G., Ramirez, J.L., Silver, S., 1994. Resistance to arsenic compounds in microorganisms. FEMS Microbiol. Rev. 15, 355–367. Corbisier, P., Ji, G., Nuyts, G., Mergeay, M., Silver, S., 1993. luxAB gene fusions with the arsenic and cadmium resistance operons of Staphylococcus aureus plasmid pI258. FEMS Microbiol. Lett. 110, 231–238. Daus, B., Wennrich, R., Weiss, H., 2004. Sorption materials for arsenic removal from water: a comparative study. Water Res. 38, 2948–2954. Dubinin, M.M., Radushkevich, L.V., 1947. The equation of the characteristic curve of activated charcoal. Proc. Acad. Sci. USSR Phys. Chem. Sect. 55, 331–333. Duker, A.A., Carranza, E.J.M., Hale, M., 2003. Arsenic geochemistry and health. Environ. Int. 31, 631–641. Fattorini, D., Alonso-Hernandez, C.M., Diaz-Asencio, M., MunozCaravaca, A., Pannacciulli, F.G., Tangherlini, M., Regoli, F., 2004. Chemical speciation of arsenic in different marine organisms: importance in monitoring studies. Mar. Environ. Res. 58, 845–850. ¨ ber die adsorption in lo¨sungen. Z. Phys. Freundlich, H., 1906. U Chem. (Leipzig) 57A, 385–470. Hall, K.R., Eagleton, L.C., Acrivos, A., Vermeulen, T., 1966. Poreand solid-diffusion kinetics in fixed-bed adsorption under constant-pattern conditions. Ind. Eng. Chem. Fund. 5 (2), 212–223. Hasany, S.M., Saeed, M.M., 1992. A kinetic and thermodynamic study of silver sorption onto manganese dioxide from acid solutions. Sep. Sci. Technol. 27, 1789–1800. Ho, Y.S., McKay, G., 1998. Sorption of dye from aqueous solution by peat. Chem. Eng. J. 70, 115–124. Ho, Y.S., McKay, G., 2000. Batch sorber design using equilibrium and control time data for the removal of lead. Air Soil Pollut. 124, 141–153.
ARTICLE IN PRESS 542
WA T E R R E S E A R C H
Hobson, J.P., 1969. Physical adsorption isotherms extending from ultrahigh vacuum to vapor pressure. J. Phys. Chem. 73, 2720–2727. Jain, C.K., Ali, I., 2000. Arsenic: occurrence, toxicity and speciation techniques. Water Res. 34, 4304–4312. Ji, G., Silver, S., 1992. Regulation and expression of the arsenic resistance operon from Staphylococcus aureus plasmid pI258. J. Bacteriol. 174, 3684–3694. Kaneko, H., Sakaguchi, K., 1979. Fusion of protoplasts and genetic recombination of Brevibacterium flavum. Agric. Biol. Chem. 43, 1007–1013. Kostal, J., Yang, R., Wu, C.H., Mulchandani, A., Chen, W., 2004. Enhanced arsenic accumulation in engineered bacterial cells expressing ArsR. Appl. Environ. Microbiol. 70, 4582–4587. Lagergren, S., 1898. About the theory of so-called adsorption of soluble substances. K. Seven. Vetenskapsakad. Handl. Band 24, 1–39. Langdon, C.J., Piearce, T.G., Meharg, A.A., Semple, K.T., 2003. Interactions between earthworms and arsenic in the soil environment: a review. Environ. Pollut. 124, 361–373. Langmuir, I., 1918. The adsorption of gases on plane surfaces of glass, mica and platinum. J. Am. Chem. Soc. 40, 1361–1403. Lin, T.F., Wu, J.K., 2001. Adsorption of arsenite and arsenate within activated alumina grains: equilibrium and kinetics. Water Res. 35, 2049–2057. Maity, S., Chakravarty, S., Bhattacharjee, S., Roy, B.C., 2005. A study on arsenic adsorption on polymetallic sea nodule in aqueous medium. Water Res. 39, 2579–2590. Mateos, L.M., Schafer, A., Kalinowski, J., Martin, J.F., Puhler, A., 1996. Integration of narrow-host-range vectors from Escherichia coli into the genomes of amino acid-producing corynebacteria after intergeneric conjugation. J. Bacteriol. 178, 5768–5775. Oliveira, V., Go´mez-Ariza, J.L., Sa´nchez-Rodas, D., 2005. Extraction procedures for chemical speciation of arsenic in atmospheric total suspended particles. Anal. Bioanal. Chem. 382, 335–340. Ordo´n˜ez, E., Letek, M., Valbuena, N., Gil, J.A., Mateos, L.M., 2005. Analysis of genes involved in arsenic resistance in corynebacterium glutamicum ATCC 13032. Appl. Environ. Microbiol. 71, 6206–6215.
41 (2007) 531– 542
Peta¨nen, T., Romantschuk, M., 2002. Use of bioluminescent bacterial sensors as an alternative method for measuring heavy metals in soil extracts. Anal. Chim. Acta 456, 55–61. Peta¨nen, T., Romantschuk, M., 2003. Toxicity and bioavailability to bacteria of particle-associated arsenite and mercury. Chemosphere 50, 409–413. Peta¨nen, T., Virta, M., Karp, M., Romantschuk, M., 2001. Construction and use of broad host range mercury and arsenite biosensor plasmids in the soil bacterium Pseudomonas fluorescens OS8. Microb. Ecol. 41, 360–368. Ramanathan, S., Shi, W., Rosen, B.P., Daunert, S., 1997. Sensing antimonite and arsenite at the subattomole level with genetically engineered bioluminescent bacteria. Anal. Chem. 69, 3380–3384. Roberto, F.F., Barnes, J.M., Bruhn, D.F., 2002. Evaluation of a GFP reporter gene construct for environmental arsenic detection. Talanta 58, 181–188. Rosen, A.L., Hieftje, G.M., 2004. Inductively coupled plasma mass spectrometry and electrospray mass spectrometry for speciation analysis: applications and instrumentation. Spectrochim. Acta B 59, 135–146. Saeed, M.M., Ahmed, M., Chaudary, M.H., Gaffar, A., 2003. Kinetics, thermodynamics, and sorption profile of Eu(III) and Tm(III) on 4-(2-pyridylazo) resorcinol (PAR) imbedded polyurethane foam. Solvent Extr. Ion Exchange 21, 881–898. Scatchard, G., 1949. The attractions of proteins for small molecules and ions. Ann. NY Acad. Sci. 51, 660–672. Scott, D.L., Ramanathan, S., Shi, W., Rosen, B.P., Daunert, S., 1997. Genetically engineered bacteria: electrochemical sensing systems for antimonite and arsenite. Anal. Chem. 69, 16–20. Silver, S., Phung, L.T., 1996. Bacterial heavy metal resistance: new surprises. Annu. Rev. Microbiol. 50, 753–789. Tauriainen, S., Karp, M., Chang, W., Virta, M., 1997. Recombinant luminescent bacteria for measuring bioavailable arsenite and antimonite. Appl. Environ. Microbiol. 63, 4456–4461. Tauriainen, A., Virta, M., Chang, W., Karp, M., 1999. Measurement of firefly luciferase reporter gene activity from cells and lysates using E. coli arsenite and mercury sensors. Anal. Biochem. 272, 191–198. Weber, W.J., Morris, S.C., 1963. Kinetics of adsorption on carbon from solution. J. Sanit. Eng. Div. Am. Soc. Civ. Eng. 89, 31–60.
