CHAPTER 3 MATERIALS AND METHODS

CHAPTER 3 MATERIALS AND METHODS
CHAPTER 3
MATERIALS AND METHODS
3.1
CHEMICALS AND REAGENTS
All chemicals used were of analytical grade, and were purchased from Merck (South Africa),
unless stated otherwise. All gases of highest purity standard were purchased from African
Oxygen Limited (Afrox) (South Africa). Sr, Co and Cs stock solutions were prepared by
dissolving 3.0429g of SrCl2.6H2O, 4.9383g of Co(NO3)2.6H2O and 1.267g CsCl, respectively, in
10mL HNO3/distilled water solution (1:1) and then diluted to 1 L to give a final concentration of
1000 mg/L of each metal. Working concentrations and standard solutions were prepared by
diluting the stock solution with distilled water to give the desired concentration. All glassware
was soaked in 10% HNO3 and rinsed with distilled water prior to and after use.
3.2
MICROORGANISM
A mixed sulphate reducing bacteria (SRB) starter culture previously isolated from a coal mine
waste site was kindly supplied by Dr. Harma Greben (CSIR, Pretoria, South Africa). In previous
studies, a range of radionuclides, have been shown to be present in coal samples. In addition, low
to high radioactivity levels have also been detected around coal mines (Jamil et al., 1998;
Jasinska et al., 1998). Therefore, the use of the present SRB cultures for Sr, Co and Cs uptake
experiments was based on the hypothesis that the cultures are acclimatized to high metal and
radionuclide concentrations. Initial kinetic sorption performance parameters for Sr, Co and Cs
removal by growing SRB cultures were determined in batch anaerobic bioreactors (2L)
incubated at 25±0.5°C with continuous stirring at 120 rpm. Further batch experiments were
conducted in 100 mL serum bottles to evaluate the kinetic and equilibrium sorption parameters
of the SRB cultures under non-growth conditions.
33
3.3
MEDIA
SRB enrichments were performed according to a procedure by Butlin et al. (1949) using medium
B, a liquid medium containing sulphate, lactate and a trace of ferrous salt (Table 3.1). The media
was sterilized using a Tomin TM-323 autoclave (Durawell Co. Limited, Taiwan). The
microorganisms were incubated in the dark at 30±2°C with continuous stirring at 120 rpm. The
anaerobic growth vessel was regularly flushed with 99.9% N2 gas to maintain an anaerobic
atmosphere. When considerable blackening of the medium had occurred (due to FeS formation),
which took up to 28 days, the culture was examined to determine cell concentration using the
total bacteria count (TBC) method. The enrichment procedure was repeated until the desired
viable cell concentration was obtained. A 10 L stock culture bioreactor was started by aseptically
transferring a 10% (v/v) of the enriched cultures into medium C. The stock culture was subcultured every 3-4 weeks by transferring the culture to freshly prepared medium C. This medium
was selected on the basis of its lack of iron, as blackening of growth medium interferes with
analytical procedures for metal biosorption experiments.
Table 3.1 Composition of media for SRB enrichment and growth.
Component
Sodium lactate (60 % solution)
MgSO .7H O
4
2
Composition per litre
Medium B
Medium C
5 mL
6 mL
2.0g
0.1 g
Na SO
-
4.5 g
NH Cl
1.0g
1.0 g
CaSO
1.0g
-
-
0.06g
FeSO .(NH ) SO .6H O
0.5g
-
Yeast extract
K HPO
1.0 g
0.5 g
1.0 g
-
-
0.5g
0.03g
-
-
0.1 g
0.1 g
0.1 g
2
4
4
4
CaCl .6H O
2
2
4
2
4 2
4
4
KH PO
2
4
Na SO .7H O
2
3
2
Ascorbic acid
Sodium thioglycollate
Sodium citrate.2H O
2
2
34
3.4
BATCH SRB BIOREACTOR EXPERIMENTS
3.4.1 Bioreactor Configuration
The batch SRB bioreactor used for preliminary kinetic metal sorption experiments was a closed
anaerobic system, which was constructed from a 2L Pyrex glass Erlenmeyer flask (Figure 3.1).
Prior to bioreactor system assembly, the reaction vessel (bioreactor) was filled with 1.5 L
medium C spiked with the target metal, and autoclaved together with the magnetic stirrer bar.
The bioreactor was then immediately sealed with an airtight rubber stopper, and purged with
high purity (99.9%) nitrogen gas to maintain anaerobic conditions. A 10% (w/v) zinc acetate
solution was connected to reaction vessel to trap the hydrogen sulphide gas produced in the
bioreactor. Samples were withdrawn using a sterile hypodermic syringe attached to a glass tube.
Bioreactor mixing was achieved by means of an AM4 multiple heating magnetic stirrer (Velp
Scientifica, Labex Pty Ltd, South Africa) in conjunction with stirrer bars.
4
6
7
5
N2
gas
1
3
2
Figure 3.1: Schematic diagram of the laboratory-scale anaerobic SRB batch bioreactor. (1=
anaerobic bioreactor, 2= 10% zinc acetate trap solution, 3= magnetic stirrer, 4= sampling
syringe, 5= rubber stopper, 6= N2 gas inlet and 7= H2S gas outlet).
35
3.4.2 SRB Screening for Sr, Co and Cs Removal and Tolerance
Preliminary studies aimed at screening for selective Sr, Co and Cs removal and tolerance among
individual cultures in the bacterial consortium were performed in a series of batch SRB
bioreactors. Each of the batch bioreactors filled with growth medium supplemented with each
metal at an initial concentration of 75 mg/L, were inoculated with a 5 mL culture sample
containing 1.6 × 108 cells/mL of viable mixed SRB cells. All experiments were conducted in
duplicate. The suspensions were incubated under optimal SRB growth conditions for up to 28
days. At the end of the incubation period, samples were removed for monitoring changes in the
phylogeny of the original SRB as a result of exposure to the different metals. Prior to culture
isolation in agar plates, enrichments for each sample were prepared as described before. SRB
growth medium without metal but the bacterial inoculum (bacterial growth control) and medium
with metal but without bacterial inoculum (abiotic control) served as controls.
3.4.3 Kinetics of Sr2+, Co2+ and Cs+ Removal in the SRB Bioreactor
The removal of different initial Sr2+, Co2+ and Cs+ concentrations (25-500 mg/L) and their effect
on the growth and metabolism of the SRB consortium was also investigated. The suspensions
were incubated under optimal SRB growth conditions for up to 14 days. Sampling was carried
out daily over the entire incubation period, for sulphate, sulphide, residual metal concentration,
pH analysis, as well as total biomass count. Prior to analysis, the samples were clarified by
means of a Hermle Z 323 centrifuge (Memingen, Germany). Subsequent to centrifugation, the
microbial component of bacterial suspensions was separated using a 0.22µm (Millipore, USA)
filter paper. Results obtained were compared to with those from control experiments, consisting
of bioreactors inoculated with viable SRB cells in the absence of Sr, Co and Cs, and bioreactors
containing medium with metal but without bacterial inoculum (abiotic control). To determine the
nature of the mechanism of Sr2+, Cs+ and Co2+ removal from solution, metal-loaded SRB cells
were subjected to a five-step desorption procedure (Tessier et al., 1979). Obtained results were
compared with those obtained for Sr, Co and Cs removal by dead SRB cells (killed by
autoclaving at 121ºC and 15 psi for 20 minutes).
36
3.4.4 Model Formulations for SRB Bioreactor Processes
The reaction schemes, rate equations and kinetic constants used for the biological processes;
biomass population dynamics and sulphate reduction were adapted from literature. Different
models have been proposed to describe the growth kinetics of a microbial population growing
with a single limiting substrate. However, a study by Senn et al. (1994) has shown that out of
several proposed, the Monod model (Monod, 1949) has been used extensively as the model
parameters µ, µ max, S and Ks have a biological meaning and are experimentally accessible. The
Monod model describes the relationship between µ the specific growth rate) and S (substrate
concentration) by a type of saturation kinetics as shown in Equation 3.1.
µ = µ max
S
Ks + S
[3.1]
Where: µ is the specific growth rate (1/h), µ max the maximum specific growth rate (1/h), S the
substrate concentration (mg/L) and Ks is the substrate affinity, i.e. the substrate concentration at
which the cells grow at half maximum specific growth rate (mg/L).
To account for bacterial population growth, in the presence of a metal (inhibitor), a simple
Monod’s non-competitive inhibition kinetic model, incorporating an inhibition term, I, in which
Ki is the inhibition coefficient, was used to model bacteria growth (Equation 3.2). Parameters for
biological sulphate reduction in the presence of Sr2+, Co2+ and Cs+ were estimated using the Pirt
equation (Equation 3.3) as previously done by Kalyuzhnyi et al. (1998), and Robinson and
Tiedje (1983). The sulphate concentration levels were assumed to be non inhibitory. Inhibition
due to sulphide formation was assumed not significant as the sulphide concentration in the batch
system at any time was much lower than the reported inhibition concentration (300–550 mg/L of
sulphide is necessary to impart sulphide inhibition). The reactor was assumed to be completely
mixed, and pH was kept constant in a narrow range (pH 6.8–7.2, unless stated otherwise). Thus,
there was no inhibition associated acidic or alkaline conditions in the reactors. A second-order
rate equation (Equation 3.5), based on the solute concentration, and incorporating an active
biomass term was used to predict metal (Sr2+, Co2+ and Cs+) removal over time in the
bioreactors.
37
dX  µ max SX
=
dt  K s + S
−

 × I

[3.2]
dS
1  dX 
=


dt Yx / s  dt 
[3.3]
I=
−
Ki
Ki + C
dC
= k C CX
dt
[3.4]
[3.5]
where; µ max is the maximum specific growth rate coefficient (1/h), Ks is the half saturation
constant (mg/L), S is the concentration of substrate at time t (mg/L), Yx/s is the bacterial yield
coefficient (mg of biomass produced per mg substrate utilized), X is the concentration of biomass
at time t (mg/L), kC is the second-order rate coefficient (L/mg/h) and C is the metal concentration
at time t (mg/L).
3.4.5 Modelling Software
In this study, the computer program AQUASIM 2.0 (Reichert, 1998) was used for all bioreactor
processes modeling. AQUASIM is a computer program for the identification and simulation of
aquatic systems. The program can be used to calculate substrate removal in bioreactors for any
user specified microbial system. AQUASIM extends the capabilities of conventional
environmental simulation programs mainly in. two ways: (1) the program does not implement a
specific model, but allow users to specify models as freely as possible and (2) it offers system
identification tasks in addition to simulation. AQUASIM performs simulations by comparing
calculated results with measured data, and such simulations are then used to validate model
assumptions. Systematic deviations between calculated and measured values may suggest that
additional important processes are to be considered or corrections are needed in the formulation
38
of processes. Another important functionality of AQUASIM is automatic parameter estimation
for a given model structure using measured data. These estimations are done using the weighted
least-squares technique. This is not only important for getting neutral estimates of parameters,
but it is a prerequisite for efficiently comparing different models. Several calculations with
several target variables, and universal as well as calculation-specific parameters, can be
combined to a single parameter estimation process. This is a very important feature not present in
most of the statistical software. This software has been successfully used by a number of
researchers to describe SRB bioreactor processes (Kosseva and Hansford, 2001; Ristow and
Hansford, 2001). The program AQUASIM estimated the best fit by minimizing the χ2 values as
shown in Equation 3.6. However, the goodness of fit χ2 values may be subject to bias due to
multivariate non-normality. If data from the model are similar to the experimental data, χ2 will be
a small number; if they are different, χ2 will be a large number. For each metal set, the sensitivity
of the kinetic parameters related to the SRB bioreactor processes; biomass, substrate and metal
concentrations, was also computed using AQUASIM 2.0.
χ2 = Σ
(S exp - S mod )2
S mod
[3.6]
where: S,mod is the calculated (model) concentration at time t (mg/L), and S,exp is the experimental
concentration at time t (mg/L).
3.5
KINETICS OF Sr2+, Co2+ and Cs+ BIOSORPTION FROM AQUEOUS SOLUTION
3.5.1 Kinetic experiments
Kinetic sorption experiments for the effect of initial concentration, biomass concentration,
metabolic state and pH on the removal of Sr2+, Co2+ and Cs+ from single metal aqueous solutions
were performed in 100 mL rubber-sealed serum bottles. The bacterial consortium used for the
metal biosorption studies was reconstituted by combining all the bacterial isolates detected in
both non-contaminated and contaminated bioreactors. The initial metal concentration and pH
were varied between 25-500 mg/L and 2-9, respectively, and the biomass cell density was kept
constant at 0.5 g/L. For the biomass concentration studies, cell density was varied between 0.5-3
g/L, while the initial metal concentration and pH were kept constant at 75 mg/L and 4,
39
respectively. To determine the effect of bacterial metabolic state, live and heat-killed bacteria
cells were exposed to the different metals at an initial concentration of 75 mg/L and pH 4. SRB
cells suspended in aqueous solution but without metal and aqueous metal solution but without
bacterial inoculum (abiotic control) served as controls. In all experiments, samples (1 mL) were
removed at timed intervals by sterile syringes for residual metal analysis. Metal concentration
was determined by the AAS. The metal uptake capacity (qeq) was calculated as:
q eq =
(C
ini
-C eq
)
X
[3.7]
where: qeq = metal uptake capacity (mg/g), Cini = initial concentration of metal in solution
(mg/L), Ceq = equilibrium concentration of metal in solution (mg/L), and X = dry weight of SRB
added (g).
3.5.2 Kinetic Modelling
In an attempt to clarify the mechanism of Sr2+, Co2+ and Cs+ removal from aqueous solution by
SRB cells, and identify the main factors controlling sorption rate such as; mass transport, pore
diffusion and chemical reaction processes, the first-order (Lagergren, 1898), pseudo-secondorder (Ho and McKay, 1999), and external and intraparticle diffusion (Weber and Morris, 1963)
kinetic models were evaluated.
The linearised pseudo-first and second order models can be expressed as follows:
ln (q eq − qt ) = ln q eq − k1t
1
t
t
=
+
2
qt k 2 q eq q eq
[3.8]
[3.9]
where: qt = concentration of ion species in the sorbent at time t (mg/g), qeq = equilibrium
concentration of adsorbed ionic species in the sorbent (mg/g), k1= biosorption constant of
pseudo-first-order Lagergren equation (min-1) and k2 = biosorption constant of pseudo-secondorder Lagergren equation (min/ g/mg). The applicability of the pseudo-first and second order
40
models was determined by linear regression method, from which sorption parameters were
determined.
For a solid-liquid sorption process, the solute transfer is usually characterized by either external
mass transfer (boundary layer diffusion) or intraparticle diffusion or both. If external diffusion of
metal cations (within the diffuse layers outside the sorbent) is the rate-limiting step then it has
been shown that Equation 3.11 can be fitted into sorption data with some success.
ln
Ct
A
= −k f t
Cini
V
[3.11]
where: Cini = initial metal concentration (mg/L), Ct = metal concentration at time t (mg/L),
A=external sorption area (m2/g), V= total solution volume (L), and kf = the external diffusion
coefficient (cm/s). However, when intraparticle diffusion (transfer of metal cations from the
adsorbent surface to the internal active binding sites) is the rate-limiting step, then sorption data
can be described by Equation 3.12, as proposed by Weber and Morris in 1962;
q t = k i t 0.5
[3.12]
where: qt = concentration of ion species in the sorbent at time t (mg/g) and ki = the intraparticle
diffusion rate (mg.g.min0.5).
The applicability of the external and intraparticle diffusion model was determined by linear
regression method, from which sorption parameters were determined. For the external diffusion
model, if the experimental data (a plot of ln(Ct/Cini) against time t), conforms to a linear plot with
a high correlation coefficient (R2), then external diffusion is the rate-limiting step. The external
diffusion coefficient kf (cm/s) can be calculated from the slope of the straight line obtained from
Equation 3.10. On the other hand, if a plot of the experimental data (qt against t0.5) conforms to a
linear plot and the line passes through origin, then intraparticle diffusion is the rate-limiting step
(Webber and Morris, 1963). The intraparticle diffusion rate coefficient is given by the slope of
the line.
41
3.6
MAXIMUM BIOSORPTION CAPACITY OF Sr2+, Co2+ AND Cs+
3.6.1 Equilibrium Sorption Experiments
Batch equilibrium sorption experiments were carried out using standard batch methodology as
described by Volesky (1990). The equilibrium removal of Sr2+, Co2+ and Cs+ from single and
binary metal solutions was evaluated. For these studies, the bacterial consortium used was
reconstituted combination of all the bacterial isolates detected in both non-contaminated and
contaminated bioreactors. The initial concentration was varied between 25 and 1200 mg/L for
single metal isotherms. The competitive removal of metal ions from binary metal solutions was
carried out by adding equivalent initial concentrations of the target and competing metal ion
ranging from 25 to 1200 mg/L. In all experiments, pH and SRB cell density were kept constant
at 4 and 0.5 g/L, respectively. SRB cells suspended in aqueous solution metal but without metal
and aqueous metal solution but without bacterial inoculum (abiotic control) served as controls.
Samples (1mL) were removed by sterile syringes at equilibrium for residual metal analysis.
Metal concentration was determined by the AAS.
3.6.2 Equilibrium isotherm modeling
The applicability of well established empirical models, such as the classical Langmuir
(Langmuir, 1918) and Freundlich (Freundlich, 1907) equilibrium sorption isotherms in the
present sorption process are evaluated. The Langmuir equation can be expressed as:
q=
qmax bC eq
1 + bC eq
[3.13]
This models incorporates two constants that are easy to interpret; qmax, the maximum sorbate
uptake (mg/g), and b, the coefficient related to the affinity between the sorbent and sorbate. q =
sorption uptake (mg/g). On the hand, the Freundlich equation can be expressed as:
q = kC eq
(1 / n )
[3.14]
42
This model also incorporates two constants: k, which corresponds to the binding capacity; and n,
which characterises the affinity between the bacterial sorbent and metal. Ceq = equilibrium
concentration of the sorbate remaining in the solution (mg/L). In order to verify the good
compliance of the experimental data to the empirical (Langmuir and Freundlich isotherms)
models, a non linear regression method was used to construct model curves, from which
equilibrium parameters were estimated.
3.7
ADSORPTION OF PROTONS AND Sr2+, Co2+ AND Cs+ ONTO SRB CELLS
3.7.1 Preparation of Bacterial Adsorbent
A mixed SRB bacteria culture was obtained by combining all the bacterial isolates detected in
both non-contaminated and contaminated bioreactors. The microorganisms were grown in stock
in Postgate medium C for about 5-7 days, after which the actively growing bacteria cells were
harvested and aseptically transferred to 2L rubber-sealed batch anaerobic bioreactors and
allowed to grow in fresh Postgate Medium C. The cells were allowed to grow until midstationary phase (5-7 days). Equal volumes of liquid bacteria culture were dispensed into
centrifuge tubes, and the bacteria were harvested from the growth media by centrifugation (6000
×g, 15 minutes). Cells were repeatedly washed in deionized water to remove growth medium
impurities. Metal cations that maybe present on SRB cell wall surfaces were stripped by soaking
in 0.001M EDTA for 30 minutes, followed by intensive rinsing in deionized water. Finally cells
were then subjected to two final rinses in a NaNO3 electrolyte solution without an acid wash step
to prevent any damage to the outer membrane of the gram negative species. The ionic strength of
the electrolyte wash solution was varied according to that of the experiment in which the bacteria
were to be used.
3.7.2 Metal Adsorption Experiments
Experiments for Sr, Cs and Co adsorption onto SRB cells were conducted under anaerobic
conditions as a function of pH, ionic strength and temperature. Prior to use, the bottles were
cleaned by soaking in two successive acid baths (10% v/v HNO3) for 24 hours each, rinsed twice
in ultra-pure water, and then dried in an oven at 105ºC. For each metal, a stock bacterial
suspension consisting of a known mass of bacteria suspended in an electrolyte solution (0.1 M
NaNO3 or unless otherwise stated) was prepared. The stock bacterial suspension was then
43
supplemented with the target metal obtained from 1000 mg/L metal stock solution (prepared in
0.1 M NaNO3) to yield the required initial concentration. The initial concentrations for Sr, Cs
and Co were chosen based on calculations using MINTEQA2 (Allison et al., 1991) to circumvent
premature metal precipitation with increasing pH. After the addition of the metal, the pH of the
bacterial-metal stock suspension was recorded. The suspension was then divided into a series of
100 mL serum bottles, and the pH was adjusted to the desired value by adding small volumes of
concentrated HNO3 or NaOH solution. In all the experiments, the pH was allowed to drift, and
the final pH of the suspension was recorded at the end of the equilibration period. After pH
adjustment, the serum bottles were immediately sealed with airtight rubber stoppers, and
incubated in a Labcon SPL-MP 15 Orbital Shaker (Labcon Laboratory Services, South Africa) at
100 rpm, and allowed to equilibrate for 3 hours at 25±0.5°C. All experiments were performed
repeatedly in triplicates. At the end of the equilibration period samples were collected by sterile
syringes equipped with a needle, which was inserted into the rubber stoppers. The collected
sample was clarified by centrifugation. The clear supernatant was acidified to 1% v/v HNO3 and
stored at 4 °C before metal analysis.
3.7.3 Surface Complexation Modeling
In this study, FITMOD, a modified version of the computer program FITEQL 2.0 (Westall,
1982) was used to construct geochemical models describing proton interaction with the bacteria
(Daughney et al., 2004). FITMOD incorporates a number of models; both nonelectrostatic and
electrostatic double layer models. This program utilizes the proton balance approach to optimize
protonation constants of the various functional groups on the bacterial surface. SRB titration
experimental data obtained was plotted in terms of the concentration of deprotonated or
protonated sites per mass of SRB (mol/g), using Equation 3.15.
[H+] added/released = (Ca – Cb – [H+] + [OH-])/mb
[3.15]
where: Ca, Cb, [H+] and [OH-] are the molar concentrations of acid, base, H+ and OH- species,
respectively, and mb is the weight of SRB cells (mg/L) in the suspension.
44
The mass law relationship in conjunction with the mole balance expressions was used to define
the adsorptive process. A 1:1 metal/surface site stoichometry was used for all calculations, and
equilibrium constants for aqueous metal hydrolysis were obtained from Baes and Mesmer
(1979). This information was used to account for Sr, Co and Cs adsorption onto SRB cell
surfaces according to Equation 3.16. Under equilibrium conditions partitioning between the solid
surface and the aqueous phase is therefore quantified with the corresponding mass law (Equation
3.17).
M + + R − A − ⇔ R − AM
K ads =
[R − AM]
[M ]× [R − A ]
+
−
[3.16]
[3.17]
where: M+ = target metal, R-A- = deprotonated bacterial surface site, R-AM = metal-site
complex, Kads = thermodynamic equilibrium constant for the reaction, and brackets denote
concentrations of the specified species.
The goodness-of-fit of the different models to the titration data is indicated by the overall
variance, V(Y), which is calculated by FITMOD as shown in Equation 3.18. Values of V(Y)
between 1 and 20 generally indicate an acceptable fit to the data (Daughney et al., 2004).
V(Y ) =
Y
P 
S
YII  Y



2
n p • n II − nu
[3.18]
where: Y is the error in the mass balance calculations, SY is the default experimental error
calculated by FITMOD, np is the number of data points, nII is the number of chemical
components for which both total and free concentrations are known, and nu is the number of
adjustable parameters.
45
3.8
ANALYTICAL PROCEDURES
3.8.1 Bacterial Culture Characterization
Gram Staining
Gram staining of SRB was carried out as described by Shimeld and Rodgers (1999). At midstationary phase, a 0.5 mL sample of sulphate reducing bacteria culture was aseptically
withdrawn with a sterile syringe. The sample was smeared on a slide, soaked in a violet dye and
then treated with iodine. The slide was then rinsed with alcohol and counterstained with
safranine. The result was viewed on a phase microscope.
Phylogenetic Analysis
Culture isolation was carried out using a method by Butlin et al. (1949) in medium A, solidified
by 2.3% agar. An aliquot of 9 mL of cooled liquid agar medium was mixed with 1 mL of culture
dilution. The suspension was then incubated at 30ºC in an anaerobic jar. Individual colonies that
developed were picked and cultured in the corresponding culture medium. The process of
isolation was repeated several times until isolates were deemed to be pure. Genomic DNA was
extracted according to the protocol described for the Wizard Genomic DNA purification kit
(Promega).
16S
rRNA
genes
were
amplified
by
using
primers
Fd1
(5’-
AGAGTTTGATCCTGGCTCAG-3’) and Rd1 (5’-AAGGAGGTGATCCAGCC-3’) and by
using the following reaction conditions: 1 min at 94ºC, 30 cycles of 30 s at 94ºC, 1 min at 50ºC
and 2 min at 72ºC, and a final extension step of 10 min at 72ºC.
PCR fragments were then cloned into pGEM-T-easy (Promega). Recombinant clones, with
inserts
of
the
correct
length,
were
ATTTAGGTGACACTATAGAA-3’)
and
sequenced
T7
by
using
primers
SP6
(5’-
(5’-TAATACGACTCACTATAGGG-3’)
(Genome Express). The nucleotide sequences of the 16S rRNA genes were compared with
reference sequences from the GenBank database. The 16S rRNA gene sequences of the strains
were aligned with reference sequences of various Desulfovibrio species using programs provided
by the Ribosomal Database Project II. Sequence alignment was verified manually using the
program BIOEDIT. Positions of sequence and alignment uncertainty were omitted from the
analysis. Pairwise evolutionary distances based on an unambiguous stretch of 1274 bp were
computed by using the Jukes and Cantor (1969) method. The dendrogram was constructed by
46
using the neighbour-joining method. Confidence in the tree topology was determined by
bootstrap analysis based on 100 re-samplings.
Scanning Electron Microscopy
Surface morphology of the present SRB culture was studied by the scanning electron microscope
(SEM). SRB culture samples were filtered through a 0.22µm filter to obtain the cells. Filtered
SRB cells were first fixed overnight in 2.5 % glutaraldehyde in 0.1M phosphate buffer (pH 7.0)
solution. The fixative solution was decanted off and cells were then washed twice in phosphate
buffer (0.1M, pH 7.0) for 10 minutes. Thereafter, the cells were dehydrated in a series of ethanol
solutions, 30%, 50%, 70%, 80%, 90%, and twice in absolute ethanol, and each step lasted for 15
minutes. The samples were then dried in liquid CO2 for 2 hours at critical point (31.1º and 73
atmospheres). Pieces of the membranes containing dried cells were cut into small squares and
then mounted on stubs with double-sided tape, and then gold-coated for 30 minutes in a Large
Desk II Cold Sputter Etch Coater. The prepared samples were then observed under a JEOL-JSM840 scanning electron microscope.
Potentiometric titrations
Acid/base properties of SRB cells with regard to H+ and OH- ions were studied by potentiometric
titrations. The titrations were performed using an automated titration system comprising of a
burette system, glass electrode and a pH meter (Metrohm 718 STAT-Titrino model) according
previously published procedures (Fein et al., 1997; Cox et al., 1999; Martinez et al., 2002;
Ngwenya et al., 2003; Borrok and Fein, 2005; Johnson et al., 2007). Prior to titration, 0.3g (wet
weight) SRB cells were suspended in 25 mL of the appropriate electrolyte solution which had
been purged by N2 bubbling for 60 minutes to eliminate CO2. The suspension was immediately
placed into a sealed titration vessel with continuous stirring at 140 rpm under N2. Titrations were
carried out by the gradual addition of pre-set small volumes of 0.1M NaOH or 0.1M HCl titrant
(standardized against reagent grade KC8H4O4H and Na2CO3 respectively). The titrator was set to
add successive base or acid after a stability of 0.1Mv/sec was attained. Three sets of duplicate
SRB suspensions were first titrated with 0.1M HCl to about pH 4 and then titrated up to about
pH 10 with 0.1M NaOH. SRB titration experiments were conducted at different ionic strengths
(0.01M, 0.1M and 0.5M NaNO3) and temperatures (5°C, 50°C and 75°C). A total organic carbon
47
(TOC) analysis was conducted on the bacterial suspensions to confirm the presence or absence of
organic exudates, due to cell lysis. Blank titrations devoid of SRB cells were also performed for
base and acid titration. Reversibility of the SRB acid/base behaviour was established by
performing reverse titrations.
Fourier Transform Infrared Spectroscopy
The acid/base characterization was complemented by Fourier Transform Infra-Red (FTIR)
spectroscopy analysis to confirm the presence of different cell wall functional groups. FTIR
analysis of SRB cells was conducted according to a method described by Ojeda et al. (2008). The
SRB cells equilibrated with a neutral (pH 6.5) electrolyte solution 0.1M NaNO3. Triplicate SRB
cell suspensions (in minimal volume of the electrolyte solution) were initially frozen at -20ºC
and then freeze-dried. FTIR spectra of the freeze-dried samples of SRB were recorded on KBr
pellets at room temperature using a Perkin Elmer FTIR GX spectrum 2000 (PerkinElmer, USA).
The sample compartment was flushed with dry air to reduce interference of H2O and CO2. Data
analysis focused on the 600-4000 cm-1 region.
3.8.2 Total Organic Carbon (TOC) Analysis
Total organic carbon analyses of samples were performed in accordance with Standard Methods
of Examination of Water and Wastewater (APHA, 1994). Bacterial suspensions (100 mL) at
different pH values (2-11), were centrifuged and the supernatant passed through 0.22µm filters to
remove the microbial component. To prevent any organic compound losses, the clear extracts
were stored in the dark at −20°C for not longer than 7 days until analysis time. The analysis of
dissolved organic carbon was performed by means of a Shimadzu TOC-5000/5050 analyzer
coupled to an ASI-V autosampler (Shimadzu Scientific Instruments, Inc., Tokyo, Japan). TOC
analysis was performed using the ‘acidify and sparge’ method, otherwise known as NonPurgeable Organic Carbon (NPOC) method. An overview of the instrument specifications and
basic operation conditions used for TOC analysis is shown in Appendix 1. During the NPOC
analysis, the sample is automatically injected into the total carbon reactor along with an
oxidising agent (1.69 M Na2S2O8) and, with 5% H3PO4 (v/v) to adjust the sample pH to 2-3.
After pH adjustment, sparge gas is bubbled through the samples for 3 minutes to eliminate the
inorganic carbon component. The total carbon remaining in the sample after sparging is
48
measured to determine the total organic carbon. The TOC concentration of the samples was
estimated for a calibration curve constructed using potassium hydrogen biphthalate (KHP)
working standard solutions with an organic carbon range of 2-12 mg/L. The zero standards
(blanks) were prepared using Ultrapure Mill-Q water. All samples were analysed in duplicate.
3.8.3 Solid Phase Sr, Co And Cs Precipitates Analysis
Speciation Analysis
To determine the species distribution of the solid phase metal species, the bound metals were
released from the SRB biomass using different desorbing agents of varying strengths according
to a method by Tessier et al. (1979). SRB cells in aqueous solution not spiked with a metal
served as a control.
Step I: To extract exchangeable species, 8 mL of 1M MgCl2 solution (pH 6) was added to 2 g of
the pellet obtained by centrifugation (7000g × 15 minutes). The mixture was stirred at 25ºC for 1
hour. The mixture was then centrifuged and the supernatant was used for metal analysis.
Step II: To extract species bound to carbonates, 10 mL of 0.1M sodium acetate adjusted to pH 5
using acetic acid was added to the residue obtained at step I. The mixture was stirred at 25ºC for
3 hours, and then centrifuged and the supernatant was used for metal analysis.
Step III: To extract species bound to oxides, 10 mL of 0.1M Hydroxyl ammonium chloride (pH
2) was added to the residue obtained from step II. The mixture was also stirred at 96ºC for 3
hours, and then centrifuged and the supernatant was used for metal analysis.
Step IV: To extract species bound to sulphide, 5 mL of 30% hydrogen peroxide adjusted to pH 2
with HNO3 and 3 mL 0.02M HNO3 was added to the residue obtained from step III. The mixture
was stirred for 2 hours at 85ºC. This was followed by the addition of 5 mL aliquot of 3.2M
ammonium acetate (pH 2) in 20% (v/v) HNO3 to avoid adsorption of the extracted metal into the
oxidized fraction. The mixture was stirred for a further 30 minutes at 85ºC and then centrifuged
and the supernatant was used for metal analysis.
49
Step V: To extract residual strontium species, 2 mL each of H2O2, HNO3 and Hydrofluoric acid
were added to the residue obtained from step IV. The mixture was stirred at 96ºC for 3 hours,
and then centrifuged and the supernatant was used for metal analysis.
3.8.4 Metal Concentration Analysis
Strontium
Total, solid-phase and dissolved Sr2+ concentrations were determined separately. Raw samples
were centrifuged at 10 000 g for 20 minutes to obtain the dissolved and solid phase fractions. To
determine the dissolved Sr2+ concentration, 2.5mL of the supernatant was dispensed into 10mL
acid washed tubes to which 0.5mL of 30% H2O2 and 0.1mL of 70% trace metal grade HNO3
were added and incubated at 60°C overnight. The resulting pellet was also subjected to the same
acid digestion treatment. Following digestion or extraction, the volume of the solution was made
up to 10mL, giving a 4× dilution factor. Strontium concentrations were determined at a
wavelength of 460.7 nm in a nitrous oxide-acetylene flame using an AAnalyst400 Perkin Elmer
AAS (Perkin Elmer, Shelton, USA) equipped with a 10 mA hollow cathode Sr lamp. Prior to
sample analysis, the instrument was calibrated against a calibration and reagent blank (5% nitric
acid solution in ultrapure water), as well as Sr2+ standards containing 0.2, 0.4, 0.6, 0.8, and 1.0
mg Sr/L). Linear calibration graphs were obtained over the concentration ranges 0–1.0 mg/L ,
with the correlation coefficient set at not less than 0.995 (R2≥0.995). Strontium ionization in this
flame was suppressed by the addition of a potassium chloride solution to give a final
concentration of 2 mg/L K+ in all samples including the standards and blank. The lower detection
limit for this procedure is estimated to be about 20 µg/L.
Cobalt
Total, solid-phase and dissolved Co2+ concentrations were also determined as described
previously for Sr2+. Cobalt concentrations were determined at a wavelength of 240.7nm in an airacetylene flame using an AAnalyst400 Perkin Elmer AAS (Perkin Elmer, Shelton, USA)
equipped with a 10mA hollow cathode Co lamp. Prior to sample analysis, the instrument was
calibrated against a calibration and reagent blank (5% nitric acid solution in ultrapure water), as
well as Co2+ standards of known concentrations. Linear calibration graphs were obtained over
50
the standard concentration ranges, with the correlation coefficient set at not less than 0.995
(R2≥0.995).
Cesium
Total, solid-phase and dissolved Cs+ concentrations were also determined as previously
described for Sr2+. Cesium concentrations were determined at a wavelength of 852.1 nm in an
air-acetylene flame using an AAnalyst400 Perkin Elmer AAS (Perkin Elmer, Shelton, USA)
equipped with a 10 mA hollow cathode Cs lamp. Prior to sample analysis, the instrument was
calibrated against a calibration and reagent blank (5% nitric acid solution in ultrapure water), as
well as Cs+ standards of known concentrations. Cesium ionization was also suppressed by the
addition of a potassium chloride solution to give a final concentration of 2 mg/L K+ in all
samples including the standards and blank. Linear calibration graphs were obtained over the
standard concentration ranges, with the correlation coefficient set at not less than 0.995
(R2≥0.995). The lower detection limit for this procedure is estimated to be about 0.005µg/mL.
3.8.5 Sulphate Concentration
The concentration of sulphate was determined using the turbidimetric method (APHA, 1994).
Prior to analysis, samples were centrifuged (10 000 g for 15 minutes) to remove suspended
solids. To a 5mL sample, 0.25mL of conditioning reagent (consisting of 50mL glycerol, 30mL
concentrated HCl, 75g NaCl, 100mL ethanol and 300mL deionized water) was added, and mixed
thoroughly. After that an excess amount of finely ground BaCl2 was added and mixture was
homogenized in a vortex for 1 minute. The absorbance of the mixture was measured at 420 nm.
Sulphate concentration was calculated from a calibration curve (R2≥0.995) obtained from
measurement of standard sulphate concentration in the range 2-20 mg/L, using a similar
procedure for sulphate analysis.
3.8.6 Biomass Concentration
Viable SRB population
The three-tube most probable number (MPN) method was used for estimating the viable SRB
population. Samples were first serially diluted by aseptically transferring 1 mL volumes into
rubber-sealed culture tubes containing 9mL Postgate medium C. The culture tubes were then
51
incubated in the dark at 37°C in a shaker at 200rpm for up to 28 days. Blackening of the medium
due to FeS formation was regarded as a positive growth result. After the 28 days of incubation,
the pattern of positive and negative tubes per dilution is noted, and then an MPN calculator was
used to determine the most probable number of organisms.
Total SRB cell counts
The total bacteria count (TBC) method was used to estimate the total bacterial population.
Bacteria were enumerated by direct counting using a Petroff-Hausser Counting Chamber
(Labotech, South Africa) on a phase contrast Zeiss Axioskop II microscope (Zeiss, Germany).
Bacterial density (g dry/L)
For dry biomass determination, bacterial suspension aliquots of 30 mL were passed through a
pre-weighed 0.22µm milli-pore filter. Prior to use, the filter was rinsed in ultrapure water to
remove soluble salts that might be present in the filter. The biomass accumulated on the filter
was dried at 80ºC for 48 hours and then weighed.
52
CHAPTER 4
Sr2+, Co2+ and Cs+ REMOVAL IN A BATCH SULPHIDOGENIC
BIOREACTOR
4.1
PROSPECTS OF RADIONUCLIDE REMEDIATION IN AN SRB BIOREACTOR
The present study evaluates the applicability of an SRB biomass as the reactive component in a
metal bioremediation system aimed at controlling radionuclide dispersion in the environment.
Due to their metabolic activities (which result in the production of biogenic ligands), reactive
surface properties and ubiquitous distribution, SRB are considered to be principal microbial
agents for controlling the dispersion of metals and radionuclides in the environment (Lovley and
Phillips, 1992; Lloyd et al., 1999; Smith and Gadd, 2000; Ngwenya and Whitely, 2006). While
the utilisation of SRB cultures for the bioremediation of water sources contaminated with Sr2+,
Co2+ and Cs+ as the ultimate solution is not completely fulfilling, however, it is the only practical
solution worth exploring, considering that these microorgansisms have been found to be capable
of surviving in irradiated environments (Bruhn et al., 2009). Engineered bioremediation (through
biostimulation and bioaugumentation) has, in some instances been proposed as a potential
strategy for the immobilization of various environmental contaminants in groundwater systems
(Brown et al., 2006). However, the application of active (or growing) microbial cultures to
contaminated sites is often complicated by unpredictable microbial activities due to the toxic
effects of the metals and radionuclides. In addition, fluctuations in environmental conditions
further present uncertainties in the performance of the culture in the contaminated environment.
Therefore, it is critical that the growth, metabolism and diversity dynamics of prospective metal
bioremediation bacteria is thoroughly evaluated under contaminated environmental settings. In
such cases, kinetic models for bacterial growth and biological substrate consumption become
valuable aids for the design, operation and control of the chosen bioremediation strategies.
Furthermore, the models become valuable research tools to improve our understanding of the
fundamental microbial processes and interactions during metal bioremediation. According to
53
Mazidji et al. (1992), the broad non-specific impact of metals on microbial activity is manifested
through two distinct modes of action; mortality of less tolerant bacterial species, leading to a
decrease in total number of bacterial population and diversity, and the reduction of metabolic
activities of the surviving bacterial culture. The former can be defined as the toxic effect of the
metal, while the decrease in metabolic rate reflects the inhibitory effect of the metal. The present
study presents findings on the toxic and inhibitory effects of Sr2+, Co2+ and Cs+ on an SRB
consortium by monitoring the changes in microbial diversity, biomass population and the
sulphate reduction rates.
4.2
SRB CHARACTERIZATION AND SCREENING
4.2.1 Partial Characterization of the Initial SRB Consortium
In this study, a number of approaches were used to characterize the SRB consortium before and
after exposure to increasing Sr, Co and Cs concentrations in a sulphidogenic bioreactor. Gram
staining results indicated that the consortium was predominantly Gram-negative with a few
Gram-positive species (Figure 1, Appendix 2). Surface morphology of SRB cells was studied
with the scanning electron microscope (SEM). Under the SEM mainly subspherical and rodshaped SRB cells occurring singly were observed, which is known for lactate as substrate
(Figure 2, Appendix 2). The cells were approximately 0.5-1 µm in diameter and 1.7 – 2.5 µm
long, which is in close agreement with literature values for SRB (Postgate, 1984; Motamedi and
Pedersen, 1998). Phylogenetic analysis using the 16S rRNA fingerprinting technique of the
original SRB consortium (control) indicated a dominance of only two genera, belonging to
Enterococcus and Staphylococcus (Figure 4.1). The genus Enterococcus belongs to a group of
bacteria that comprise of Gram-positive and produce lactic acid as the major metabolic endproduct of carbohydrate fermentation. The ability of Enterococcus sp. to reduce sulphate into
sulphide has been reported. The only difference between these microorganisms and SRB is that
Enterococcus sp. reduce sulphate to synthesize sulphur containing organic compounds (cystein
and methionin metabolism) and sulphide production is highly regulated due to its toxic effect
which may contribute to cell death. On the other hand, SRB use sulphate as an electron acceptor
to support growth under anaerobic conditions, a phenomenon known as sulphate respiration
(Barton, 1992).
54
68
Enterococcus faecium
Control 1 clone
87
Enterococcus villorum
69
Enterococcus hirae
Enterococcus mundtti
Enterococcus durans
Enterococcus ratti
Staphylococcus hominis subsp. novobiosepticus
93
Control 2 clone
Staphylococcus hominis subsp. hominis
79 Staphylococcus warneri
Staphylococcus haemolyticus
Staphylococcus sciuri subs. prodentium
Streptomyces griseus
Figure 4.1: Phylogenetic tree of Enterococcus and Staphylococcus sp. related clones obtained
from the original SRB culture (control). The cloned genes are named according to the isolate
number.
4.2.2 Microbial Diversity Analysis in the Sr, Co and Cs Bioreactors
A number of researchers have reported the microbial removal of Sr2+, Co2+ and Cs+ from
contaminated aqueous streams (Small et al., 1999; HaiLei et al., 2007; Gao et al., 2010).
However, there is lack of knowledge on the probable microbial diversity shifts that occur during
Sr2+, Co2+ and Cs+ bioremediation. Such results can help demonstrate the tolerance as well as
applicability of individual microorganisms for selective removal of these metal ions. Previous
results obtained from this study indicated that only two genera; Enterococcus and
Staphylococcus, were present in the initial SRB culture, and no significant changes were
observed in the SRB culture after a 28 day growth period. However, the presence of Sr2+, Co2+
and Cs+ in the growth media resulted in the emergence of new bacterial strains representing 3
different genera, belonging to Citrobacter, Paenibacillus and Stenotrophomonas spp.,
55
respectively. The phylogenetic affiliations and gram-staining results of these isolates obtained
from Sr2+, Co2+ and Cs+ contaminated enrichments with their nearest phylogenetic neighbour in
the GeneBank database are shown in Table 4.1. The emergence of these additional bacterial
species can be attributed to their tolerance and ability to adapt to conditions of toxic Sr2+, Co2+
and Cs+ concentrations. A detailed account on the growth and metabolism (sulphate reduction),
as well as ability of the bacterial consortia detected in the different bioreactors for Sr2+, Co2+ and
Cs+ removal is presented at a later stage in this chapter.
Table 4.1 Phylogeny of the SRB isolates from Sr2+, Co2+ and Cs+ contaminated bioreactors.
Clone
Source1
Closest Match in GeneBank2
Control 1
Control 2
Sr1
Sr2
Sr3
Sr4
Co1
Co2
Co3
Co4
Cs1
Cs2
Cs3
Coal mine wastewater
Coal mine wastewater
Sr bioreactor
Sr bioreactor
Sr bioreactor
Sr bioreactor
Co bioreactor
Co bioreactor
Co bioreactor
Co bioreactor
Cs bioreactor
Cs bioreactor
Cs bioreactor
Enterococcus faecium
Staphylococcus hominis
Citrobacter sp.
Citrobacter farmeri
Citrobacter farmeri
Citrobacter sp.
Enterococcus faecium
Paenibacillus motobuensis
Paenibacillus motubuensis
Paenibacillus motobuensis
Enterococcus sp.
Enterococcus faecium
Stenotrophomonas maltophilia
Similarity
(%)
98
100
100
97
100
100
99
99
99
99
99
99
99
Gram staining
result3
(+)
(-)
(-)
(-)
(-)
(-)
(+)
(+)
(+)
(+)
(+)
(+)
(-)
1
Direct from coal mine wastewater or Sr2+, Co2+ and Cs+ contaminated enrichments.
Based on partial 16S rRNA sequences. Bootstrap = 100.
3
Based on a method by Shimeld and Rodgers (1999), where (–) = Gram-negative and (+ ) = Gram-positive .
2
Microbial Diversity Shifts in the Presence of Strontium
Strontium exists in the environment in a variety of compounds, with the divalent cation, Sr2+, as
the most dominant species in contaminated groundwater systems (Brown et al., 2006). The
biosorption of Sr2+ by bacteria has been reported by Small et al. (1999) and Ngwenya and
Chirwa, (2010). However, this is the first study to demonstrate Sr2+ tolerance of individual
microorganisms in an SRB consortium. Results obtained from this study revealed that cloned
16S rRNA gene sequences of bacterial enrichments obtained from Sr2+ contaminated bioreactors
were closely affiliated to the marine SRB species belonging to genus Citrobacter. Figure 4.2
56
shows the phylogenetic affiliations of SRB clones isolated from Sr2+contaminated enrichments
and some known related bacteria based on partial 16S rDNA sequences. In particular,
Citrobacter farmeri is a member of family Enterobacteriaceae, Gram-negative, obligate
facultative and cocci-shaped anaerobe, which grows on lactate, acetate, pyruvate and hydrogen
when provided sulphate as an electron acceptor. The ability of the Citrobacter sp. for sulphur
and sulphate reduction has been reported (Odom and Singleton 1992; Sahrani et al., 2008; Qiu et
al., 2010). Citrobacter sp. have been isolated from broad range substrate ecologies such as
effluents from wastewater treatment plants, soil and aquatic habitats, as well as in microbial
communities associated to biofilms that cause corrosion of oil pipelines (Holt et al. 1994; Nada
et al. 2004; Neria-Gonzalez et al. 2006). These microorganisms have been used for the removal
of a range of metals, including Pb, Cu, Cd, An, Am, U, La, Th (Macaskie and Dean, 1982;
Macaskie et al., 1992; Sharma et al., 2009).
Citrobacter youngae
Escherichia albertii
63
Salmonella enterica subsp. houtenae
Citrobacter sedlakki
Sr1 clone
100
Sr4 clone
Citrobacter rodentium
Sr3 clone
100
Sr2 clone
Citrobacter farmeri
Enterobacter cloacea
Figure 4.2: Phylogenetic tree of Citrobacter sp. related clones obtained from cultures grown in
Sr2+ contaminated medium. The cloned genes are named according to the source/contaminating
metal and isolate number.
57
Microbial Diversity Shifts in the Presence of Cobalt
The toxic and inhibitory effects of cobalt on the growth sulphate reducers have been studied
before, as some SRB require the presence of a cobalt-containing, vitamin B12-methyltransferase
for carbon oxidation through the acetyl-CoA pathway (Weijma et al., 2000; Ekstrom and Morel,
2008). Results obtained in the present study indicated that the 16S rRNA gene sequences of SRB
clones isolated from Co-contaminated bioreactors were closely related to Paenibacillus sp. and
Enterococcus sp. Paenibacillus sp. are Gram-positive, facultative anaerobic bacteria that are
related to Bacilli but differ in the DNA encoding their 16S rRNA. They have been shown to be
capable of utilizing acetate as a carbon source (Nakamura, 1984). The phylogeny of
Paenibacillus sp. and some known related bacteria based on 16S rDNA sequences are shown in
Figure 4.3. The bacterium Paenibacillus has also been shown to be able to reduce colourless
potassium tellurite to black metallic tellurium (Chien and Han, 2009).
Similarly, the Co-
reducing capability of these microorganisms has been recently reported by Gao and coworkers
(2010).
Co1 clone
100
Co3 clone
Co2 clone
100
Paenibacillus motobuensis
Paenibacillus azoreducens
88
Paenibacillus cineris
98
Paenibacillus rhizospaerae
Paenibacillus favisporus
Paenibacillus curdlanolyticus
Figure 4.3: Phylogenetic tree of Paenibacillus sp. related clones obtained from cultures grown in
Co2+ contaminated medium. The cloned genes are named according to the source/contaminating
metal and isolate number.
58
Microbial Diversity Shifts in the Presence of Cesium
Phylogenetic analysis of bacterial isolates obtained from sulphidogenic bioreactors contaminated
with Cs, indicated that the medium favored the growth of two main strains; Enterococcus sp. and
Stenotrophomonas sp. Figure 4.4 shows the phylogenetic affiliations of the Enterococcus sp.
isolated from Cs+ contaminated enrichments and some known related bacteria based on partial
16S rDNA sequences. The biological effects of cesium on different living organisms are well
documented (Avery et al., 1993; Lloyd and Macaskie, 2000). However, limited studies have been
done on the resilience of individual microorganisms towards cesium. Therefore, results obtained
from this study are crucial in establishing the applicability of specific growing bacterial cultures
for cesium bioremediation. The observed results demonstrate that Enterococcus sp. (original
bacterial isolate) is more resilient to Cs+, compared to Staphylococcus sp.
Enterococcus faecium
Cs2 clone
Cs1 clone
68
Cs3 clone
Co1 clone
87
Enterococcus villorum
69
Enterococcus hirae
Enterococcus mundtti
Enterococcus durans
Enterococcus ratti
Figure 4.4: Phylogenetic tree of Enterococcus sp. related clones obtained from cultures grown in
Cs+ contaminated medium. The cloned genes are named according to the source/contaminating
metal and isolate number.
59
This genus belongs to a group of Gram-positive bacteria, which produce lactic acid as the major
metabolic end-product of carbohydrate fermentation. The metal removing capability of the
Enterococcus sp., particularly, E. faecium has been reported recently (Yilmaz et al., 2010). The
genus Stenotrophomonas, which is phylogenetically placed in the γ-Proteobacteriacea group, was
first described with the type species S. maltophilia1. The phylogenetic affiliations of the
Stenotrophomonas sp. isolated from Cs+ contaminated enrichments and some known related
bacteria based on partial 16S rDNA sequences are shown in Figure 4.5. S. maltofilia is found in
diverse environments, including water, soil debris, raw milk, frozen fish and disinfection
solutions used in hospitals (Garcia et al., 2002).
Pseudomonas geniculata
Cs clone
100
Stenotrophomonas maltophilia
62
Pseudomonas hibiscicola
Pseudomonas betelli
Stenotrophomonas nitritireducens
58
99
Stenotrophomonas terrae
Stenotrophomonas humi
Stenotrophomonas koreensis
Stenotrophomonas rhizophila
Pseudoxanthomonas kalamensis
Figure 4.5: Phylogenetic tree of Stenotrophomonas sp. related clones obtained from cultures
grown in Cs+ contaminated medium. The cloned genes are named according to the
source/contaminating metal and isolate number.
60
S. maltophilia, has extraordinary range of activities, including the breakdown of natural and
man-made pollutants that are central to bioremediation and phytoremediation strategies. Like
other Gram-negative bacilli, these microorganisms are tolerant to various toxic metals, such as
Ag, As, Cd, Co, Hg, Pb, Zn, U and selenite (Botes et al., 2007; Ryan et al., 2009). Members of
the genus Stenotrophomonas have also been reported to play an important ecological role in the
nitrogen and sulphur cycles (Banerjee and Yesmin, 2002).
4.2.3 Mechanisms of Sr2+, Co2+ and Cs+ Removal in the Bioreactors
Results obtained from this study showed no metal concentration loss in the absence of bacteria
(abiotic control). Experiments were also conducted in a closed bioreactor system, which
minimises metal losses/additions to and from the environment. Figure 4.6 shows the distribution
of Sr2+, Co2+ and Cs+ species on bacterial cells after exposure to a medium containing 75 mg/L
of each metal. Results obtained suggest that Sr2+ removal occurred mainly through biosorption as
about 68% of the solid phase Sr2+ was found to be occurring in the exchangeable fraction, gives
an indication of the amount of Sr2+ that is bound to cell surface functional groups by relatively
weak electrostatic interaction, and can be released by ion-exchange processes (Dahl et al., 2008).
Similarly, high metal concentrations were obtained in the exchangeable fractions of both soil
amended with sewage sludge and sulphide-rich tailings of a mine (Nyamagara, 1998; Carlsson et
al., 2002). The remainder of the solid-phase Sr2+ was a result of a chemical precipitation due to
the presence of ligands in the medium, including sulphate (resulting in the formation of insoluble
strontium sulphate), as well as products of both sulphate reduction and carbon oxidation. About
22% of the Sr2+ was bound to carbonates, 4% bound to oxides and hydroxides, 3% sulphides and
the remainder occurring as an immobile fraction. The mechanisms of Sr2+ removal by other
microorganisms, e.g. Aspergillus niger has been studied. The experimental data obtained showed
that immobilised A. niger was effective in removing Sr2+ from aqueous solution. Analysis of the
Sr2+ precipiates by FTIR analysis showed that Sr2+ removal was facilitated by the presence of
surface amide groups I and II (Pan et al., 2009). Regardless of the metabolic state of the SRB
cells, the results obtained clearly demonstrated that the different bacterial surface chemical
functional groups played a significant role in determining the mode and extent of Sr2+ removal in
the bioreactor. The unique and strong Sr2+ complexing ability organic ligands present on the SRB
cell surface is detailed in Chapter 5.
61
80
% Metal ion
60
Sr2+
Co2+
A
Cs+
40
20
0
F1
F2
F3
F4
F5
Fractions
70
60
Sr2+
Co2+
B
% Metal ion
50
Cs+
40
30
20
10
0
F1
F2
F3
F4
F5
Fractions
Figure 4.6: Partitioning of Sr2+, Co2+ and Cs+ species on viable (A) and dead (B) SRB biomass.
F1 = exchangeable, F2 = carbonates, F3 = oxides, F4 = sulphide/organics, and F5 = residual
fraction.
62
However, the metabolic state of the bacteria had a significant effect on the mechanism and fate
of Co2+ in the bioreactor. Under growth conditions, the Co2+ in the medium is ultimately
accumulated both inside (45%) and outside the cell (43%). The dual mechanism of metal
removal from solution by the bacterial cells has been reported before (Ho and McKay, 2000;
Aksu, 2001; Goncalves et al., 2007). Initially metal sorption occurs through a rapid sorption
phase, where the metal ions bind to the surface through the formation of ionic bonds, which is
followed by a slower, metabolism-dependent phase, where the metal is accumulated inside the
cell. The extent of Co2+ intracellular accumulation and adsorption is represented by the immobile
fraction (F5) and the exchangeable fraction (F1), respectively. Cobalt is a metabolically essential
metal required by most living organisms, including bacteria. Therefore, most Co-assimilating
organisms possess specialized energy-dependent systems, which transport the metal into the cell,
resulting in its intracellular accumulation (Ekstrom and Morel, 2008). Similarly, in the present
study, bioaccumulation accounted was the main mechanism of Co2+ removal by live cultures as
compared to the non-viable SRB cultures. The remainder of the Co2+ in the medium was
removed from solution through the formation of metal-ligand complexes, which was observed, in
both live and dead SRB cultures.
The accumulation of cesium by different life forms is well documented (Avery et al., 1993;
Lloyd and Macaskie, 2000). Cs+ is chemically similar to the biologically essential alkali cation
K+, and enters into the cells of biosorbents through K+ transporters. Alternatively, in the absence
of Cs+ transport across membranes, extracellular adsorption of the cation on the cell surface
occurs. Similarly, in the present study Cs+ uptake from solution by SRB biomass (dead or alive)
occurred manly through surface adsorption reactions, as most of the sorbed metal was released
by an ionic desorbing agent. The observed extracelluar adsorption of Cs+ exhibited by the
present culture can be attributed to the ability of the SRB potassium ion transporters to
discriminate between the toxic cesium cation and biologically essential potassium cation. The
resistance of live SRB cultures for Cs+ uptake are also be further demonstrated by the slight
differences in the percentage amount of Cs+ released by the different desorbing agents, as
compared to dead SRB cells. Similar results have been obtained in experiments for the
accumulation of cesium by the bacterium Thermus sp. TibetanG7 (HaiLei et al., 2007).
63
4.3
SIMULATION OF SULPHIDOGENIC BIOREACTOR PROCESSES
4.3.1 Modelling Approach
Kinetics of SRB growth and biological sulphate reduction
Batch experiments were conducted to estimate the kinetic parameters associated with SRB
growth and metabolism in the presence of increasing Sr2+, Co2+ and Cs+ concentrations. The
approach by Kalyuzhnyi et al. (1998) formed the basis for modeling the kinetics of SRB growth
and biological sulphate reduction (BSR) in the bioreactors. The Monod model provides a link
between bacterial growth and substrate utilization. According to this model, bacterial growth
phenomena can be described satisfactorily with four parameters; two kinetic parameters: the
maximum growth rate (µ max) and half velocity concentration (Ks), and two stoichiometric
parameters: the yield coefficient (Yx/s) and substrate (sulphate) concentration (Kovarova-Kovar
and Egli, 1998). Since it is anticipated that the presence of metal ions might have negative effects
on SRB cell activities, the Monod was modified by the addition of an inhibition term, I,
(Equation 3.3), to incorporate the toxic and inhibitory effects of Sr2+, Co2+ and Cs+. The present
model neglected cell decay because of the relatively short experimental durations.
Kinetics of Metal Uptake by Microbial Biomass
Results obtained in this study indicated that Sr2+, Co2+ and Cs+ removal occurred in the presence
of both active (alive) and non-active (heat-killed) SRB biomass, and metal ion removal occurred
mainly through biosorption processes, with the exception of Co2+ where additional metal uptake
through bioaccumulation also occurs. During metal uptake, the mobile metal ion (metal ion in
solution) concentration is in equilibrium with the immobile metal ion concentration on the
bacterial cells. Therefore, a linear relationship is assumed between the two. Thus, the decrease in
dissolved metal concentration in the bioreactors was accounted for by a simple pseudo secondorder rate law equation (Equation 3.4), incorporating a total biomass term (X) which accounts for
both the active (alive) and non-active (heat-killed) SRB biomass in the bioreactors. Since, the
sulphate concentration in the bioreactors was typically 2-7 orders of magnitude higher that the
initial metal concentration, the formation of insoluble SrSO4 at lower initial Sr2+ concentrations
was assumed to be minimal.
64
4.3.2 Model Calibration and Parameter Estimation
Experimental and model data sets (S0, S, X0, X, Cini, C and t) were obtained from a study of SRB
growth and metabolism in the presence of Sr2+, Co2+ and Cs+ at various concentrations (25-500
mg/L) in batch experiments, where sulphate, biomass and metal concentrations were measured
over time. The model was calibrated using sets of data for each metal at an initial metal
concentration of 75 mg/L. Model simulations of bacterial population, biological sulphate
reduction and metal removal in the bioreactors were implemented in the computer program for
simulation of aquatic systems, AQUASIM 2.0 (Reichert, 1998). The initial conditions and model
input values for parameters used for the simulation of SRB bioreactor processes are shown in
Appendix 3. Monod kinetic parameters were determined by fitting the experimental data to the
equations described in Section 3.5.4 by using a non-linear chi-square method, which minimizes
the residual sum of squares between the experimental data and calculated values.
4.3.3 Simulation of SRB Bioreactor Processes in the Presence of Strontium
Kinetics of SRB Growth in the Presence of Sr2+
Table 4.2 shows the optimised Monod kinetic parameters for SRB growth in the presence of
increasing initial Sr2+ concentrations. Fairly constant kinetic parameters were obtained,
indicating that the model was precise in simulating SRB growth in the presence of Sr2+ as an
inhibitor. Model fits improved with increasing inhibitor (initial Sr2+) concentration, suggesting
that the model is more applicable at higher initial Sr2+ concentrations where SRB growth
inhibition is well defined. Similarly, there a good agreement was observed between the
experimental data and model predictions (Figure 4.7). However, the model failed to capture the
initial short (24 hours) acclimitisation (lag) phase of the bacterial consortium.
Table 4.2 Optimised Monod parameters for SRB population growth in the presence of Sr2+.
Cini (mg/L)
X0
µ max (1/h)
Ks (mg/L)
Yx/s (mg/mg)
Ki (mg/L)
χ2
25
75*
100
300
500
46
48
47
56
48
0.981
0.987
0.989
0.986
0.982
466.9
466.9
462.4
466.8
466.1
0.0758
0.0735
0.0738
0.0749
0.0749
0.616
0.616
0.616
0.616
0.616
918.8
611.5
408.8
298.6
68.8
* = data used for model calibration and parameter estimation
65
400
SRB concentration (mg/L)
SRB concentration (mg/L)
400
25 m g/L Sr_exp
25 m g/L Sr_m odel
300
200
100
75 m g/L Sr_exp
75 m g/L Sr_m odel
300
200
100
0
0
0
50
100
150
200
250
300
0
350
50
150
200
250
300
350
Time (hours)
Time (hours)
400
400
100 m g/L Sr_exp
100 m g/L Sr_m odel
SRB concentration (mg/L)
SRB concentration (mg/L)
100
300
200
100
0
0
50
100
150
200
250
300
350
Time (hours)
300 m g/L Sr_exp
300 m g/L Sr_m odel
300
200
100
0
0
50
100
150
200
250
300
350
Time (hours)
SRB concentration (mg/L)
100
500 m g/L Sr_exp
500 m g/L Sr_m odel
80
60
40
0
50
100
150
200
250
300
350
Time (hours)
Figure 4.7: Experimental and model predicted growth of an SRB biomass in a batch bioreactor
in the presence of different initial Sr2+ concentrations.
66
Both the experimental observations and model simulations indicated that after the short
acclimitisation phase, there was a longer exponential phase (up to 250 hours) followed a
stationary phase. However, it was evident that increasing the initial Sr2+ concentration prolonged
the acclimatization phase and delayed the outset of the exponential phase. Generally, the
bacterial concentration multiplied remarkably (up to five times) in the presence of Sr
concentrations of ≤ 100 mg/L, after which an increase in initial concentration significantly
lowered microbial growth. In this case, the toxic and inhibitory effects of the metal were
inevitable as neither biosorption nor precipitation lowered the available Sr2+ concentration to a
level tolerable for SRB growth. Despite the presence of high Sr2+ concentrations, it should be
noted that despite positive SRB growth was maintained in all bioreactors. However, the SRB
biomass growth rate decreased with increasing Sr2+ concentrations. These results further
demonstrate the resilience of Citrobacter spp. towards Sr2+.
Kinetics of Biological Sulphate Reduction in the Presence of Sr2+
Table 4.3 shows the Monod kinetic parameters for biological sulphate reduction in the presence
of increasing initial Sr2+ concentrations and the corresponding initial SRB concentrations. The
obtained goodness of fit (χ2) values indicate that increasing the initial Sr2+concentration resulted
poor model fits. These parameters also suggest that the toxic (growth retardation) effect of Sr2+ is
well defined compared to its inhibitory effect, as lower inhibition coefficients were obtained for
the biological sulphate reduction process. Comparisons between experimental observations and
model simulations show marked model misfits (Figure 4.8), which can be attributed to the
unpredicted decrease in sulphate concentration due to the chemical precipitation of sulphate (Sr2+
reaction with sulphate), which is not accounted for by the present model.
Table 4.3 Optimised Monod parameters for biological sulphate reduction in the presence of Sr2+.
Cini
X0
S0 (mg/L) µ max (1/h) Ks (mg/L)
(mg/L)
(mg/L)
25
46
3034
0.982
466.9
75*
48
3027
0.985
466.6
0.989
466.9
100
47
3013
300
56
3114
0.988
467.0
500
44
2912
0.988
469.0
* = data used for model calibration and parameter estimation
Yx/s
(mg/mg)
0.0732
0.0730
0.0730
0.0735
0.0730
Ki (mg/L)
χ2
0.354
0.356
0.356
0.356
0.357
51864.8
27400.9
45181.3
99586.5
751165.6
67
3200
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
3200
25 m g/L Sr_exp
25 m g/L Sr_m odel
3000
2800
2600
2400
2200
2000
1800
1600
1400
75 m g/L Sr_exp
75 m g/L Sr_m odel
3000
2800
2600
2400
2200
2000
1800
1600
1400
1200
0
50
100
150
200
250
300
0
350
50
150
200
250
300
350
Tim e (hours)
Tim e (hours)
3200
3200
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
100
100 m g/L Sr_exp
100 m g/L Sr_m odel
3000
2800
2600
2400
2200
2000
1800
1600
0
50
100
150
200
250
300
350
Tim e (hours)
300 m g/L Sr_exp
300 m g/L Sr_m odel
3000
2800
2600
2400
2200
2000
0
50
100
150
200
250
300
350
Tim e (hours)
Sulphate concentration (mg/L)
3000
500 m g/L Sr_exp
500 m g/L Sr_m odel
2800
2600
2400
2200
2000
1800
1600
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.8: Experimental and model simulations of sulphate reduction in a batch SRB bioreactor
in the presence of different initial Sr2+ concentrations.
68
The obtained results also reveal that the efficiency of biological sulphate reduction (BSR) was
dependent on the initial Sr2+ and biomass concentration, where increasing the initial Sr2+
concentration resulted in the reduction viable biomass and consequently the metabolic (sulphate
reduction) potential of the surviving SRB cultures (previously identified as Citrobacter sp.). For
example, at an initial Sr2+ concentration of 25 mg/L, the sulphate concentration decreased from
3034 to 1501 mg/l, corresponding to a sulphate reduction rate of about 4.56 mg/L/h, compared to
a rate of 2.18 mg/L/h observed at an initial Sr2+ concentration of 500 mg/L. The sulphate
reduction capability of Citrobacter sp. is well documented (Odom and Singleton 1992; Barton
1995; Sahrani et al., 2008; Qiu et al., 2010).
Kinetics of Sr Removal in the Bioreactors
The kinetic parameters obtained for the removal of Sr2+ (25-500 mg/L) by a growing SRB
biomass at biomass initial concentration (X0) in the range 43.5-56.0 mg/L are shown in Table 4.4.
Results obtained indicate that the parameter kC (rate coefficient) is not a true constant as its value
changes with increasing initial concentration. This is not surprising considering that the rate of
metal removal from solution is dependent on both the metal and biomass concentration at any
given time. High Sr2+ concentrations retard SRB growth, consequently limiting Sr2+ removal
from solution. The model fits improved with decreasing initial metal concentration, implying that
the model predictions are more valid for Sr2+ uptake at lower initial concentrations. Figure 4.9
shows the uptake kinetics of by the growing SRB biomass, where a comparison is made between
the model predictions and experimental observations. The increased difficulties in model fits
with increasing initial Sr2+ concentrations is a direct result of the
premature chemical
precipitation of Sr2+ due to the formation of SrSO4, as previously reported.
Table 4.4 Optimised kinetic parameters for Sr2+ removal by growing SRB cells in a bioreactor.
Cini (mg/L)
X0 (mg/L)
kC (×10-3)
25
*75
100
300
500
46
48
47
56
44
9.73
9.88
2.44
0.104
0.0111
Metal removal
capacity (%)
100
100
100
64
40
χ2
4.73
122.4
1612.1
73940.1
95029.0
* = data used for model calibration and parameter estimation
69
30
Sr concentration (mg/L)
Sr concentration (mg/L)
100
25 m g/L Sr_exp
25 m g/L Sr_m odel
20
10
0
75 m g/L Sr_exp
75 m g/L Sr_m odel
80
60
40
20
0
0
50
100
150
200
250
300
350
0
50
Tim e (hours)
150
200
250
300
350
Tim e (hours)
120
350
100
100 m g/L Sr_exp
100 m g/L Sr_m odel
Sr concentration (mg/L)
Sr concentration (mg/L)
100
80
60
40
20
0
300
300 m g/L Sr_exp
300 m g/L Sr_m odel
250
200
150
100
50
0
0
50
100
150
200
250
300
350
Tim e (hours)
0
50
100
150
200
250
300
350
Tim e (hours)
Sr concentration (mg/L)
600
500
500 m g/L Sr_exp
500 m g/L Sr_m odel
400
300
200
100
0
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.9: Experimental and second-order model plot showing the removal of different Sr2+
initial concentrations by a growing SRB consortium in a batch bioreactor.
70
Generally, in all bioreactors experimental and model simulations an initial fast Sr2+ removal rate,
which then slows down with time in accordance with the outset of the stationary bacterial growth
phase, after which no additional metal removal occurs was observed. This is a common
phenomenon for most metal biosorption processes reported in literature (Ho and McKay, 1998;
Ho and McKay, 2000; Aksu, 2001). Results obtained also indicated that the overall Sr2+ uptake
(%) decreased with increasing initial concentration due to increased metal toxicity effects.
Complete Sr2+ uptake was attained in the first 12, 72 and 144 hours at initial concentrations of
25, 75 and 100 mg/L, respectively. This study is among the first to report on the removal of Sr2+
by growing bacterial cultures. Results obtained demonstrate that SRB bioreactors inoculated with
active bacterial species belonging to the genus Citrobacter hold a promise towards development
of in situ Sr2+ bioremediation technologies.
4.3.4 Simulation of SRB Bioreactor Processes in the Presence of Cobalt
Kinetics of SRB Growth in the Presence of Co2+
Results obtained show that Co2+ was less toxic to the present SRB culture compared to Sr2+, as
higher maximum growth rates (µ max) and lower growth inhibition coefficient (Ki) was observed
(Table 4.5). A good correlation between the experimental data and model simulations was
observed, and model fits slightly improved with increasing initial concentration. The effect of
cobalt on the growth and metabolism (sulphate reduction) of pure SRB cultures has been studied,
where decreased growth rates and metabolic activities were observed in the presence of high
Co2+ concentrations (Weijma et al., 2000; Ekstrom and Morel, 2008). Similarly, results obtained
from this study also show that initial Co2+ concentration ≥300 mg/L retarded the growth of the
SRB consortium (Figure 4.10).
Table 4.5 Optimised Monod parameters for SRB population growth in the presence of Co2+.
Cini (mg/L)
X0
µ max (1/h)
Ks (mg/L)
Yx/s (mg/mg)
Ki (mg/L)
χ2
25
75*
100
300
500
50
50
50
50
50
1.50
1.50
1.53
1.51
1.51
455.0
456.2
456.9
455.1
455.1
0.0772
0.0779
0.0780
0.0780
0.0778
0.352
0.358
0.358
0.356
0.358
1810.6
2182.3
2745.3
2955.5
5534.5
* = data used for model calibration and parameter estimation
71
300
25 m g/L Co_exp
25 m g/L Co_m odel
250
SRB concentration (mg/L)
SRB concentration (mg/L)
300
200
150
100
50
75 m g/L Co_exp
75 m g/L Co_m odel
250
200
150
100
50
0
0
0
50
100
150
200
250
300
0
350
50
150
200
250
300
350
Time (hours)
Time (hours)
300
140
SRB concentration (mg/L)
SRB concentration (mg/L)
100
100 m g/L Co_exp
100 m g/L Co_m odel
250
200
150
100
50
300 m g/L Co_exp
300 m g/L Co_m odel
120
100
80
60
40
20
0
0
50
100
150
200
250
300
350
Time (hours)
0
50
100
150
200
250
300
350
Time (hours)
SRB concentration (mg/L)
140
120
500 m g/L Co_exp
500 m g/L Co_m odel
100
80
60
40
20
0
0
50
100
150
200
250
300
350
Time (hours)
Figure 4.10: Experimental and model predicted growth of an SRB biomass in a batch bioreactor
in the presence of different initial Co2+ concentrations.
72
Experimental observations for SRB growth showed that the presence of Co2+ in the growth
media resulted in a brief initial decline in SRB concentration, which the present model failed to
simulate. This phase was followed by a longer exponential phase which lasted up to 250 hours,
and lastly a stationary phase, which the were both simulated well by the mode. Generally,
positive SRB growth was only recorded at initial Co2+ concentrations ≤100 mg/L, after which
increasing the initial concentration resulted in growth retardation. Therefore, the poor model
simulations observed at higher concentrations (≥300 mg/L) can be attributed to the inadequacy
of the model to simulate negative bacterial growth. These results demonstrate the sensitivity of
Paenibacillus sp. towards high initial Co2+ concentrations.
Kinetics of Biological Sulphate Reduction in the Presence of Co2+
Table 4.6 shows the optimised Monod kinetic parameters for biological sulphate reduction by
SRB cultures in the presence of increasing initial Co2+ concentrations. When the experimental
data was fit into the sulphate reduction equation, an improvement in model fit was observed with
increasing initial Co2+ concentration. The parameters obtained reflect that the inhibitory effect of
Co2+ on sulphate reduction was less compared to Sr2+, as indicated by the lower sulphate
reduction inhibition coefficients. Lower inhibition coefficients (Ki values) were obtained for
sulphate reduction, compared to those obtained for SRB growth, suggesting that the toxic
(growth retardation) effect of Co2+ is superior to its inhibitory effect. This remark is substantiated
by experimental sulphate reduction studies at higher initial Co2+ concentrations (≥300 mg/L),
where positive metabolic activity (sulphate reduction) was observed (Figure 4.11).
Table 4.6 Optimised Monod parameters for biological sulphate reduction in the presence of
Co2+.
Cini
X0 (mg/L) S0 (mg/L) µ max (1/h) Ks (mg/L)
(mg/L)
25
50
3030
1.50
457.0
1.50
456.8
75*
50
3045
100
50
3000
1.50
456.1
1.53
455.0
300
50
3030
500
50
3060
1.50
456.0
* = data used for model calibration and parameter estimation
Yx/s
(mg/mg)
0.0779
0.0778
0.0777
0.0776
0.0776
Ki
(mg/L)
0.172
0.173
0.172
0.172
0.172
χ2
44118.3
14408.7
10297.0
18262.5
16575.6
73
3200
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
3200
25 m g/L Co_exp
25 m g/L Co_m odel
3000
2800
2600
2400
2200
2000
1800
1600
1400
75 m g/L Co_exp
75 m g/L Co_m odel
3000
2800
2600
2400
2200
2000
1800
1200
0
50
100
150
200
250
300
0
350
50
150
200
250
300
350
Tim e (hours)
Tim e (hours)
3200
3100
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
100
100 m g/L Co_exp
100 m g/L Co_m odel
3000
2800
2600
2400
2200
0
50
100
150
200
250
300
350
300 m g/L Co_exp
300 m g/L Co_m odel
3000
2900
2800
2700
2600
2500
2400
0
50
Tim e (hours)
100
150
200
250
300
350
Tim e (hours)
Sulphate concentration (mg/L)
3100
500 m g/L Co_exp
500 m g/L Co_m odel
3000
2900
2800
2700
2600
2500
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.11: Experimental and model simulations of sulphate reduction in a batch SRB
bioreactor in the presence of different initial Co2+ concentrations.
74
With regards to metabolic rate efficiency of the consortium, sulphate reduction rates (mg/L/h) of
4.79, 3.93, 3.41, 2.80 and 1.34 were obtained in the presence of 25, 75, 100, 300 and 500 mg/L
Co2+, respectively. In view of these findings, it is not clear whether both Co-tolerant bacterial
species (Paenibacillus sp. and Enterococcus sp.) were involved in the observed gradual decrease
of sulphate in the bioreactors. While Paenibacillus sp. are known to be capable of utilizing
acetate as a carbon source (Nakamura, 1984), further studies are necessary to validate their
ability or inability to reduce sulphate into sulphide. On the other hand, Enterococcus sp., have
been shown to be capable of reducing sulphate into sulphide (Barton, 1992). In summary, the
underlying prerequisite for successful utilization of active microorganisms for effective Co2+
bioremediation is that the initial Co2+ should be kept at a minimum.
Kinetics of Co2+ Removal in the Bioreactors
Table 4.7 shows the optimised kinetic parameters for Co2+ removal by a growing mixed SRB
consortium. The goodness of fit (χ2 values) of the model gradually improved with decreasing
initial Co2+ concentration. The second-order rate coefficient (kC) in the range 1-9 was obtained
at initial Co2+ concentration ≤300 mg/L. However, increasing the initial Co2+ concentration to
500 mg/L decreased the rate coefficient, and poor model fit. Accordingly, a lower removal rate
was observed at an initial Co2+ concentration of 500 mg/L. Generally, the batch experiments
showed a 100%, 74%, 69%, 13% and 6% removal of the total Co2+ concentration at initial
concentrations of 25, 75, 100, 300 and 500 mgL-1, respectively. These results further emphasise
that low initial Co2+ concentrations have less negative effects on SRB growth and metabolism.
Similarly, the experimental data and model simulations also showed that the removal of Co2+ by
the cultures was dependant on the initial concentration (Figure 4.12).
Table 4.7 Optimised kinetic parameters for Co2+ removal by growing SRB cells in a bioreactor.
Cini (mg/L)
X0 (mg/L)
kC (×10-5)
25
*75
100
300
500
50
50
50
50
50
8.85
4.98
7.11
1.44
0.578
Metal removal
capacity (%)
100
74
69
13
6
χ2
135.7
212.6
354.5
303.0
494.7
* = data used for model calibration and parameter estimation
75
30
Co concentration (mg/L)
Co concentration (mg/L)
100
25 m g/L C o_exp
25 m g/L C o_m odel
20
10
0
75 m g/L Co_exp
75 m g/L Co_m odel
80
60
40
20
0
0
50
100
150
200
250
300
350
0
50
Tim e (hours)
200
250
300
350
340
120
Co concentration (mg/L)
Co concentration (mg/L)
150
Tim e (hours)
140
100 m g/L Co_exp
100 m g/L Co_m odel
100
80
60
40
20
0
0
50
100
150
200
250
300
350
Tim e (hours)
Co concentration (mg/L)
100
540
300 m g/L C o_exp
300 m g/L C o_m odel
320
300
280
260
240
220
0
50
100
150
200
250
300
350
Tim e (hours)
500 m g/L C o_exp
500 m g/L C o_m odel
520
500
480
460
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.12: Experimental and second-order model plot showing the removal of different Co2+
initial concentrations by a growing SRB consortium in a batch bioreactor.
76
It is evident that increasing the initial Co2+ concentration, the Co2+ removal capacity of the SRB
consortium was lowered to a level corresponding to the observed growth and metabolic activity
of the surviving cultures. The removal of Co2+ through enzymatic reduction and bioprecipitation
(as CoS and CoCO3) in the presence of SRB cultures has been reported (Krumholz et al., 2003).
Since SRB possess specialized Co2+ assimilation mechanisms, the metal is then ultimately
accumulated inside the cell (Weijma et al., 2000; Ekstrom and Morel, 2008; Gao et al., 2010). In
accordance with these observations, results obtained in this study indicated that a portion of the
Co2+ in solution was removed through bioaccumulation. However, in addition to
bioaccumulation and biopreciptation, results obtained in this study indicated that a significant
portion of the Co2+ in solution was removed by biosorption.
4.3.5 Simulation of SRB Bioreactor Processes in the Presence of Cesium
Kinetics of SRB Growth in the Presence of Cs+
Table 4.8 shows the optimized Monod parameters for SRB growth in the presence of increasing
initial Cs+ concentrations. While parameter optimization was successful, there was no clear trend
in the model fits with an increase/decrease in initial Cs+ concentration. However, the best model
fit was obtained at higher initial Cs+ concentration, suggesting that the present Monod inhibition
model is more valid at higher initial Cs+ concentration (≥100 mg/L) where the toxic effects can
be successfully simulated. Lower inhibition (Ki) coefficients were obtained in the presence of
Cs+ in the bioreactors, compared to Sr2+ and Co2+ suggesting that Cs+ is less toxic to the SRB
consortium. Accordingly, a high maximum specific growth rate (µ max) was observed. At lower
initial Cs+ concentrations (≤100 mg/L), the experimental data obtained suggested that the SRB
consortium did not reach stationary phase for the duration of the experiments, while the model
predicted that stationary phase was reached after 264 hours of exposure (Figure 4.13).
Table 4.8 Optimised Monod parameters for SRB population growth in the presence of Cs+.
Cini (mg/L)
X0 (mg/L)
µ max (1/h)
Ks (mg/L)
Yx/s (mg/mg)
Ki (mg/L)
χ2
25
75*
100
300
500
47
49
47
49
55
9.281
9.287
9.281
9.289
9.298
492.7
492.8
492.9
492.9
492.1
0.0626
0.0626
0.0627
0.0627
0.0627
0.0571
0.0571
0.0571
0.0571
0.0572
3008.9
720.5
5518.1
1254.0
250.1
* = data used for model calibration and parameter estimation
77
300
25 m g/L Cs_exp
25 m g/L Cs_m odel
250
SRB concentration (mg/L)
SRB concentration (mg/L)
300
200
150
100
50
75 m g/L Cs_exp
75 m g/L Cs_m odel
250
200
150
100
50
0
0
0
50
100
150
200
250
300
0
350
50
100
200
250
300
350
Tim e (hours)
Tim e (hours)
200
200
SRB concentration (mg/L)
SRB concentration (mg/L)
150
100 m g/L Cs_exp
100 m g/L Cs_m odel
150
100
50
300 m g/L Cs_exp
300 m g/L Cs_m odel
150
100
50
0
0
50
100
150
200
250
300
350
0
0
Tim e (hours)
50
100
150
200
250
300
350
Tim e (hours)
SRB concentration (mg/L)
200
500 m g/L Cs_exp
500 m g/L Cs_m odel
150
100
50
0
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.13: Experimental and model predicted growth of an SRB biomass in a batch bioreactor
in the presence of different initial Cs+ concentrations.
78
This discrepancy can be attributed to the low toxicity of Cs+, suggesting that it was not necessary
to modify the Monod model with an inhibition term. The low toxicity of Cs+ towards live
microbial cells has been reported. However, its presence in high concentrations can dramatically
decrease the uptake of essential growth micronutrients K+ and Na+, and these elements, thereby
retarding vital microbial processes and growth (Avery, 1995). Therefore, the increased Cs+
tolerance displayed by the present SRB consortium may result from sequestration of Cs+ in
vacuoles or changes in the activity and/or specificity of transport systems mediating Cs+ efflux,
rendering it less toxic to the microorganisms identified as belonging to Enterococcus sp. and
Stenotrophomonas sp.
Kinetics of Biological Sulphate Reduction in the Presence of Cs+
Table 4.9 shows optimised Monod parameters for sulphate reduction in the presence of
increasing initial Cs+ concentrations. The obtained results indicate that Cs+ had a lesser effect on
the metabolic activities (sulphate reduction) of the SRB consortium, compared to Sr2+ and Co2+,
as lower inhibition coefficient (Ki) values were obtained. Model fits improved with increasing
initial concentration. However, visual comparisons between the experimental data and model
simulations showed a poor correlation, particularly at an initial concentration of 500 mg/L
(Figure 4.14). While the experimental data evidently displayed the inhibitory effect of Cs+ at this
concentration, the model predicted a less inhibitory effect. Since Cs+ is chemically similar to the
biologically essential alkali cation, K+, it enters into the cells of living organisms through the K+
transporter system. If present in high concentrations, the sequestered Cs+ can act as a
replacement for the essential alkali cation, K+, thus causing undesirable biological effects on the
bacterial cell (Avery, 1995).
Table 4.9 Optimised Monod parameters for biological sulphate reduction in the presence of Cs+.
Cini
X0
S0 (mg/L) µ max (1/h) Ks (mg/L)
(mg/L)
(mg/L)
25
47
3034
6.35
499.3
75*
49
3027
9.30
495.0
9.97
492.7
100
47
3054
300
49
3041
9.98
492.4
500
55
3120
9.96
495.7
* = data used for model calibration and parameter estimation
Yx/s
(mg/mg)
0.0628
0.0647
0.0618
0.0620
0.0622
Ki (mg/L)
χ2
0.0283
0.0282
0.0283
0.0283
0.0283
40115.4
17106.0
10909.7
9228.5
8815.4
79
3200
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
3200
25 m g/L Cs_exp
25 m g/L Cs_m odel
3000
2800
2600
2400
2200
2000
1800
1600
1400
75 m g/L Cs_exp
75 m g/L Cs_m odel
3000
2800
2600
2400
2200
2000
1800
1600
1400
1200
0
50
100
150
200
250
300
0
350
50
3200
200
250
300
350
3200
Sulphate concentration (mg/L)
Sulphate concentration (mg/L)
150
Time (hours)
Time (hours)
100 m g/L Cs_exp
100 m g/L Cs_m odel
3000
2800
2600
2400
2200
2000
1800
0
50
100
150
200
250
300
350
300 m g/L Cs_exp
300 m g/L Cs_m odel
3000
2800
2600
2400
2200
2000
0
Time (hours)
Sulphate concentration (mg/L)
100
3200
50
100
150
200
250
300
350
Time (hours)
500 m g/L Cs_exp
500 m g/L Cs_m odel
3000
2800
2600
2400
2200
0
50
100
150
200
250
300
350
Time (hours)
Figure 4.14: Experimental and model simulations of sulphate reduction in a batch SRB
bioreactor in the presence of different initial Cs+ concentrations.
80
The precise intracellular target(s) for Cs+-induced toxicity and inhibitory effects have yet to be
clearly defined, although certain internal structures, e.g. ribosomes, become unstable in the
presence of Cs+ and Cs+ is known to substitute poorly for K+ in the activation of many K+requiring enzymes (Avery, 1995). In summary, it can be assumed that the observed reduction of
sulphate in the bioreactors was a direct result of the metabolic activities of the two Cs-tolerant
isolates, Enterococcus sp. and Stenotrophomonas sp., due to low toxicity of Cs toxicity.
However, the sulphate reduction efficiency of the SRB consortium decreased with increasing
initial Cs+ concentration.
Kinetics of Cs+ Removal in the Bioreactors
Table 4.10 shows the optimized parameters for Cs+ uptake by a growing SRB consortium. From
this table it is clear that an increase in initial Cs+ concentration resulted in poor model fits. The
parameter, kC (second-order rate coefficient) and the corresponding Cs+ removal capacity of the
SRB biomass decreased with increasing initial Cs+ concentration. The low Cs+ removal
capacities at higher (≥300 mg/L) can both be attributed to both the toxic and inhibitory effects of
Cs+, which retard SRB growth and metabolism. Comparisons between the experimental data and
model plots indicate that the model satisfactorily simulated removal at initial Cs+ concentrations
≤75 mg/L (Figure 4.15). At initial concentration ≥300 mg/L, the model predicted a faster rate
than was observed in Cs+ removal experiments. Similarly, there was an evident dependence of
Cs+ removal on SRB concentration. At low initial Cs+ concentrations of 25 and 75 mg/L,
complete removal was observed within the first 24 hours and after 240 hours, respectively.
Table 4.10: Optimised kinetic parameters for Cs+ removal by growing SRB cells in a bioreactor.
Cini (mg/L)
X0 (mg/L)
kC (×10-4)
25
*75
100
300
500
47
49
47
49
55
8.00
1.33
0.783
0.304
0.117
Metal removal
capacity (%)
100
100
68
38
18
χ2
105.7
162.1
1003.4
3404.9
6447.5
* = data used for model calibration and parameter estimation
81
Cs concentration (mg/L)
Cs concentration (mg/L)
100
30
25 m g/L C s_exp
25 m g/L C s_m odel
20
10
0
75 m g/L Cs_exp
75 m g/L Cs_m odel
80
60
40
20
0
0
50
100
150
200
250
300
350
0
50
Tim e (hours)
150
200
250
300
350
Tim e (hours)
120
320
100 m g/L Cs_exp
100 m g/L Cs_m odel
100
Cs concentration (mg/L)
Cs concentration (mg/L)
100
80
60
40
20
0
50
100
150
200
250
300
350
Tim e (hours)
300
300 m g/L C s_exp
300 m g/L C s_m odel
280
260
240
220
200
180
160
0
50
100
150
200
250
300
350
Tim e (hours)
Cs concentration (mg/L)
520
500
500 m g/L C s_exp
500 m g/L C s_m odel
480
460
440
420
400
380
0
50
100
150
200
250
300
350
Tim e (hours)
Figure 4.15: Experimental and second-order model plot showing the removal of different Cs+
initial concentrations by a growing SRB consortium in a batch bioreactor.
82
Cs+ exists almost exclusively as the monovalent cation Cs+ in the natural environment. Although
Cs+ is a weak Lewis acid that exhibits a low tendency to form complexes with ligands, its
chemical similarity to the biologically essential alkali cation K+ facilitates high levels of
metabolism-dependent intracellular accumulation (Lloyd and Macaskie, 2000). However,
previous results obtained in this study showed that limited Cs+ removal bioreactors occurred
through bioaccumulation. The low microbial Cs+ bioaccumulation observed in this study can be
attributed to presence of competitive cations, e.g. K+, Na+, NH4+ and H+, in the medium, whose
presence have been reported to decrease its significantly (Avery, 2004). The distinct chemical
properties of Cs+, indicate that different approaches are required for biological Cs+ removal to
those which are generally adopted for other metals/radionuclides. However, its low toxicity
eliminates one potential problem in the use of live/growing cells for its removal. The inherent
differences in Cs+ uptake capacities of different microorganisms appear to be largely attributable
to differences in the affinity of monovalent cation transport systems for Cs+. The application of
rigorous screening procedures involving the use of autoradiography has great potential for
isolation of microorganisms with particularly high affinities for Cs+ (Avery, 1995).
4.3.6 Sensitivity Analysis
The sensitivity test, computed in AQUASIM 2.0 (Reichert, 1998), was performed to evaluate the
relative importance of parameters on the Monod model output. The parameters of interest were
the kinetic and stoichiometric parameters, including X, Cini, Ki, Ks, Yx/s and µ max. A 1% change in
a given model input parameter was applied, while the others were kept constant, and the effect of
the change on the model was evaluated. From the plot, it is clear that minor adjustments in the
parameters; Ki, Ks, Yx/s and µ max are critical to the model output. The model sensitivity to the
parameters Ki and µ max increases from zero and reaches a maximum at 240 hours then decreases
again to zero, exhibiting the behaviour of the absolute value of the sensitivity function (Reichert,
1998). On the other hand, the model sensitivity to the parameter Yx/s, remained at zero until after
240 hours, after which it increased reaching a maximum at 312 hours. Relatively, the model was
less sensitive to the parameter Ks, suggesting that this constant is of minor significance to reactor
performance. The dependence of the biomass concentration (X) on the initial metal concentration
(Cini) is identifiable, as such is an important aspect that has to be taken into consideration for
evaluating the performance of bioremediation bioreactors utilizing growing biomass.
83
200
150
Sensitivity (X - mg/L)
100
Ki
Ks
50
µmax
Yx/s
0
Xo
Co
-50
-100
-150
-200
0
100
200
300
400
Time (hours)
Figure 4.16: Time course of the sensitivity functions of SRB biomass concentration (X) at an
initial metal concentration of 75 mg/L with respect to other kinetic parameters.
4.4
SUMMARY
In the original SRB consortium, microorganisms belonging to two genera; Enterococcus and
Staphylococcus were detected. However, exposure to media containing Sr2+, Co2+ and Cs+ led to
the emergence of new bacterial strains; Citrobacter, Stenotrophomonas and Paenibacillus sp.
The presence of Sr2+ in the growth medium resulted in the complete eradication of the original
consortium members, and the emergence of Citrobacter sp. These microorganisms demonstrated
a unique high Sr2+ tolerance, which permitted positive bacterial growth and metabolism, as well
as high metal removal capacities in the presence of Sr2+ concentrations of up to 500 mg/L.
However, there was an obvious dependence of bacterial activities (growth and sulphate
reduction) on the initial Sr2+ concentration, where an increase in metal concentration
corresponded to a decrease in bacterial activities. Likewise, Sr2+ removal from solution was
84
dependent on the bacterial concentration, as complete Sr2+ removal was observed where the
toxic and inhibitory effects of the metal on the culture was low. This study is the first to
demonstrate high Sr tolerance among the Citrobacter sp., and it can be assumed this trait is
genetically intrinsic as the microorganisms were not pre-exposed to Sr. Results from this study
also provide strong evidence Paenibacillus sp. and Enterococcus sp., isolated from the Cocontaminated SRB bioreactors are resilient towards Co2+ and facilitate its removal through
biosorption and bioreduction. However, initial Co2+ concentration ≥300 mg/L had a detrimental
effect on the growth and metabolism of the SRB consortium. As a result, the metabolismdependent Co2+ removal (bioreduction) capacity of the cultures was also lowered. At such high
concentration it can be assumed that biosorption processes supersede bioreduction. On the other
hand, the presence of Cs+ resulted in the replacement of Staphylococcus sp. with
Stenotrophomonas sp., while Enterococcus sp were retained. The bacterial isolates obtained from
the Cs+-contaminated sulphidogenic bioreactors displayed a high degree of Cs+ tolerance, as a
prolonged exponential growth phase was observed.
Microbial Cs+ removal from solution occurred mainly through biosorption processes, and the
removal capacity decreased with increasing initial concentration. Although Cs+ is a weak Lewis
acid that exhibits a low tendency to form complexes with ligands, complete removal was
observed at initial concentration ≤75 mg/L, suggesting that the present SRB consortium may
possess extremely hydrophilic binding sites that promote the uptake of Cs+ onto bacterial cells.
In summary, through experimental and model simulations, we have been able to demonstrate the
interrelationships between SRB activities (growth, metabolism and diversity) and individual
fission product cations removal from solution. Results obtained indicate that satisfactory (up to
100%) metal removal can be attained, where the toxic and inhibitory effects of the metal on the
bacterial culture are minimal. Generally, the sensitivity of the original SRB culture towards the
different metals was in the order Sr>Co≥Cs. It is evident that the biosorption mechanism and cell
capability can also be influenced by chemical and physical properties of the target metal ion
itself, thereby determining cell viability (Veglio and Beolchini, 1997). Therefore, elucidation of
mechanisms active during bacterial metal sequestration is essential, in order to establish
consistency, and for successful exploitation of the phenomenon in realistic settings of industrial
wastewater treatment.
85
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement