Journal of Hazardous Materials 317 (2016) 335–343

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Algae as an electron donor promoting sulfate reduction for the

bioremediation of acid rock drainage
Pedro Ayala-Parra, Reyes Sierra-Alvarez, Jim A. Field ∗
Department of Chemical and Environmental Engineering, The University of Arizona, P.O. Box 210011, Tucson, AZ, USA

h i g h l i g h t s

• Algal biomass can serve as an electron donor to drive reduction of sulfate to sulfide.
• Biogenic sulfide precipitates Cu2+ as stable sulfide mineral.
• Cu+2 removal in sulfidogenic bioreactors amended with algal biomass exceeded 99.5%.
• Acidity in synthetic acid rock drainage was consumed by sulfate reduction.

a r t i c l e i n f o a b s t r a c t

Article history: This study assessed bioremediation of acid rock drainage in simulated permeable reactive barriers (PRB)
Received 1 March 2016 using algae, Chlorella sorokiniana, as the sole electron donor for sulfate-reducing bacteria. Lipid extracted
Received in revised form 3 June 2016 algae (LEA), the residues of biodiesel production, were compared with whole cell algae (WCA) as an
Accepted 5 June 2016
electron donor to promote sulfate-reducing activity. Inoculated columns containing anaerobic granu-
Available online 6 June 2016
lar sludge were fed a synthetic medium containing H2 SO4 and Cu2+ . Sulfate, sulfide, Cu2+ and pH were
monitored throughout the experiment of 123 d. Cu recovered in the column packing at the end of the
Keywords:
experiment was evaluated using sequential extraction. Both WCA and LEA promoted 80% of sulfate
Heavy metal
Acid mine drainage
removal (12.7 mg SO4 2− d−1 ) enabling near complete Cu removal (>99.5%) and alkalinity generation rais-
Algae waste ing the effluent pH to 6.5. No noteworthy sulfate reduction, alkalinity formation and Cu2+ removal were
Biodiesel observed in the endogenous control. In algae amended-columns, Cu2+ was precipitated with biogenic
Permeable reactive barrier H2 S produced by sulfate reduction. Formation of CuS was evidenced by sequential extraction and X-ray
diffraction. LEA and WCA provided similar levels of electron donor based on the COD balance. The results
demonstrate an innovative passive remediation system using residual algae biomass from the biodiesel
industry.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction HM can reach surface waters and may accumulate to toxic levels
causing severe impacts on aquatic organisms [2].
Acid rock drainage (ARD) is produced by the contact of sulfide Acidity and HM impact human health and the environment. In
mineral residues of hard rock mining with moisture and air. ARD humans, exposure to high doses of Cu can cause headaches, dizzi-
is characterized by low pH and dissolved heavy metals (HM). Low ness, nausea, and diarrhea [3]. Cu is well known for its toxicity
pH, ranging from 2 to 6, is due to oxidation of sulfides and iron to aquatic life; lethal Cu concentrations (LC50 ) ranging from 5 to
generation of protons and sulfuric acid [1]. ARD samples from dif- 50 mg L−1 have been observed in several fish species [4]. Cu is toxic
ferent mining locations are shown in Table S1 of the Supplementary to green algae grown in fresh water at pH of 5.7–6.5; cell division
Data (SD). Acidity extracts and dissolves HM such as Cd, Pb and Cu. is affected at concentrations as low as 1 ␮g L−1 , and Cu is 20-fold
more toxic than uranium at the same concentrations [5]. Certain
HM, e.g., Pb and Cd, are toxic to humans. Pb is known to cause neu-
rotoxicity and affect the cardiovascular system and kidneys [6]. Cd
∗ Corresponding author at: Department of Chemical and Environmental Engineer-
is toxic to livers, lungs and kidneys and is a known carcinogen [7].
ing, University of Arizona, 1133 E James E. Rogers Way, Room 108, Tucson, AZ 85721,
USA. The main approaches to the remediation of ARD utilize alkali to
E-mail address: jimfi[email protected] (J.A. Field). precipitate HM and neutralize acidity or use the activity of sulfate-

http://dx.doi.org/10.1016/j.jhazmat.2016.06.011
0304-3894/© 2016 Elsevier B.V. All rights reserved.
336 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343

reducing bacteria (SRB) to form biogenic sulfide and precipitate Table 1

Chemical composition of Chlorella sorokiniana biomass before and after lipid
HM. Typically, either NaOH or limestone is used to promote the
extraction.
chemical precipitation of HM. Metal hydroxides precipitate metals
due to an increase in pH, and metal carbonates precipitate using Composition WCA LEA
soluble carbonate (CO3 2− ) from limestone [8]. HM precipitation Total solids (g TS/g wet wt) 0.20 1.0a
with biogenic sulfide from SRB could be an economically attrac- Volatile solids (g VS/g dwt) 0.96 0.94
tive passive method for the treatment of ARD. A previous study has Total nitrogen (% dwt) 5.5 5.6
Total phosphorous (% dwt) 0.57 0.53
shown effective removal of Zn, Cd, Co and Ni from contaminated
Total carbon (% dwt) 51.0 48.4
groundwater using sulfate reduction [9]. Total lipids (% dwt) 9.4 ND
SRB are a group of anaerobic organisms that use organic com- a
LEA is dry powder.
pounds or molecular H2 as electron donor (e-donor) to reduce
sulfate (an external electron acceptor) to sulfide, a process known
as dissimilatory sulfate reduction [10–12] (Eq. (1)). SRB reduce sul- cellular polymeric substances (EPS) from algae, mainly composed
fate to sulfide, while organic matter is oxidized to CO2 and acidity by exopolysaccharides, as an e-donor for sulfate reduction [28,29].
is simultaneously consumed [13]. Due to the high content of complex cell wall polysaccharides in
algae, LEA is expected to degrade slowly and last a long time in PRB
SO4 2− + 2CH2 O + 2 H+ → H2 S + 2CO2 + 2 H2 O (1) as a slow-release e-donor, which is conducive to low cost and low
Formation of HM-sulfides (MeS) results from a chemical reac- maintenance. In this study, Chlorella sorokiniana-1412 is utilized as
tion between hydrogen sulfide produced by SRB and the soluble the e-donor. The objective of this study is to compare WCA with LEA
HM cation (Me2+ ), Eq. (2). and determine if algae biomass residues could serve as an e-donor
to support SRB in ARD remediation.
Me2+ + S2− → MeS ↓ (2)
The solubility constants (Ksp) of hydroxide, carbonate and sul- 2. Materials and methods
fide metals is in Table S2 of the SD. The Ksp values indicate the
lower solubility of MeS compared to metal hydroxides and carbon- This study uses WCA and LEA as an e-donor for SRB to remediate
ates [14]. Biogenic sulfide (H2 S) generated during sulfate reduction ARD. Columns were packed with glass beads and inoculated with
is an excellent ligand of HM and is extensively used for remediation sludge having sulfate-reducing activity to mimic PRBs. The exper-
of HMs in acidic effluents. iment compares an endogenous column (with no e-donor added)
E-donating compounds enable sulfate to convert sulfate to sul- with columns containing WCA or LEA.
fide. For the remediation of ARD, slow release e-donors such as
lignocellulosic materials and polysaccharide complex materials 2.1. Algal growth conditions
are generally preferred to enable passive treatment low mainte-
nance systems. Cellulosic wastes, cow manure, municipal compost Chlorella sorokiniana-1412 was cultivated on a BG-11 medium
and wood chips have been tested as e-donor for SRB [13,15–17]. in a photo-bioreactor [30]. The chemical composition of WCA and
The lower the lignin content of these materials, the higher their LEA is illustrated in Table 1.
biodegradability and capacity for developing bacterial activity [18].
Lignocellulosic materials are slow release e-donors, which are 2.2. Anaerobic inoculum
desirable in a long-lasting system, such as a permeable reactive
barrier (PRB). Sulfate-reducing anaerobic granular sludge was obtained from a
A PRB is capable of removing contaminants in a passive treat- containerboard mill wastewater treatment plant (RockTenn, Syra-
ment below-ground where a wall is installed to intercept an ARD cuse, NY). It had a volatile suspended solids (VSS) content of 0.094 g
plume, producing a clean effluent. PRBs are filled with pea-gravel VSS g−1 wet weight of pellet (after centrifugation). When unused,
and sand, mixed with reactive media. The sustainability of PRBs the sludge was stored anaerobically at 4 ◦ C.
is greatly impacted by the selection of reactive media. The goal in
biological PRBs is to use an organic carbon source with a life span 2.3. Chemicals
exceeding 5 years. In HM remediation, the reactive materials vary
from zeolite, limestone, zero-valent iron (ZVI) and local compost The main chemicals used were CuSO4 ·5H2 O, 98 +%, CAS 7758-
materials [19–21]. 998 (Sigma-Aldrich, St. Louis, MO) to add as heavy metal to the
Residual algae from the biofuels industry may provide large synthetic ARD medium. 18 M H2 SO4 , 98% v/v; CAS 7664-93-9 from
quantities of biomass in the future. Microalgae can play the dual Fisher Scientific (Fair Lawn, NJ) combined with cupric sulfate to
role of treating wastewater and generating biomass for biodiesel prepare 250 mg SO4 2− L−1 and lower the pH to 4. NaOH, ≥97%, CAS
production [22]. With current technology there is an excellent 1310-73-2, from Fisher Scientific (Fair Lawn, NJ) was used to adjust
economic outlook for algae-based biodiesel production [23]. Two the ARD medium to pH 4.
microalgal strains have emerged as production candidates from
recent U.S. Department of Energy research: Chlorella sorokini- 2.4. Chemical oxygen demand content
ana (DOE 1412) and Nannochloropsis gaditana (CCMP-1775) [24].
Chlorella sorokiniana-1412, a genetically modified microalga, has The chemical oxygen demand (COD) content of Chlorella
been widely tested for its biodiesel production potential. Lipid sorokiniana-1412 was quantified using standard methods [31], with
extracted algae (LEA) is the residual material left after lipid extrac- H2 SO4 -K2 Cr2 O7 at 150 ◦ C for 2 h and measured at 600 nm. The COD
tion from whole cell algae (WCA). The composition of algae cells (mg COD mg−1 dwt algae) was 1.43 for WCA and 1.32 for LEA.
lends itself as a slow release e-donor. LEA is expected to be a mas- Table 1 shows the algae composition.
sive waste product [25–27] that could be used as an e-donor for
bioremediation applications such as the treatment of HM in ARD 2.5. Lipid extraction from microalgae biomass
by sulfate reduction. The effectiveness of algal biomass as both car-
bon source and energy to drive the process of sulfide generation Lipids were extracted using 7.5-g dry algae biomass with 1.43 g
by SRB has not been studied except in two cases evaluating extra- COD g−1 dwt algae in a mixture of methanol and chloroform (3:1,
 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343 337

Table 3
Periods of column operation as defined by influent Cu2+ concentrations.

Period

a b c d e

Time (d) 0–26 27–59 60–74 75–100 101–123

Influent Cu2+ (mg L−1 ) 0 10 30 50 0

influent was kept at 24 ± 1 ◦ C and pumped upward using a peri-

staltic pump. The feed medium was pumped at a flow rate of
0.064 L day−1 or an empty bed hydraulic retention time (HRT) of
1.1 day, lasting 123 days. The sulfate and acidity of the synthetic
ARD medium (basal medium) were constant, while the Cu content
was gradually increased at different periods (Table 3). The influent
sulfate concentration was always 250 mg L−1 (supplied as sulfuric
acid). The pH of the medium was kept at 4.0 by partially neutral-
izing the sulfuric acid with NaOH. Cu2+ was added after the first
Fig. 1. A descriptive diagram of the three up-flow packed bed columns used to test 26 days, at three intervals. As illustrated in Table 3, Cu2+ was not
the algae biomass as electron donor for the SRB process. A) Endogenous column, B)
added during the first period (a). In subsequent periods (b–d) the
WCA column, and C) LEA column.
medium received 10, 30 and 50 mg L−1 Cu2+ . Cu was not added dur-
ing the last period (e) to test the sulfate-reducing activity and the
Table 2
Composition of the packed-bed columns.
absence of toxicity due to Cu.

Component PRB Columnsa 2.8. Analysis

DBDb Endogenous WCAc LEAd
g/cm3 g dwt/column Influent and effluent were analyzed to determine pH, S2− , SO4 2−
Glass beads (2 mm) 1.5 32 17.0 17.0 and Cu2+ . Sulfide was analyzed colorimetrically by the methylene
Sludgee 1.0 0.94 0.94 0.94 blue method [34]. Sulfate was measured by ion chromatography
Whole Cell Algae (WCA)a 0.42 0 4.9 0 with suppressed conductivity using a Dionex AS11-HC4 column
Lipid Extracted Algae (LEA)d 0.41 0 0 5.3
(Dionex, Sunnyvale, CA) and a conductivity detector. To avoid
a
Packing volume (0.07 L). sulfate formation from sulfide oxidation in sample handling and
b
DBD = Dry weight bulk density. storage, the liquid samples were acidified, followed by stripping
c
WCA (1.43 mg COD mg−1 dwt algae).
d
LEA (1.32 mg COD mg−1 dwt algae).
H2 S with N2 -CO2 , as recommended by Hughes and coworkers [35].
e
Sludge inoculum: 9 g wet sludge (0.094 g VSS g−1 wet sludge). Cu was determined by inductively coupled plasma-optical emis-
sion spectrometry (2100 Optima ICP-OES, Perkin Elmer, Waltham,
MA). The wavelength used for Cu determination was 327.3 nm and
v/v) and microwaved at 80 ◦ C [32]. The remaining solids were the detection limit was 1.0 ␮g L−1 .
washed with water 5 times and dried in a 160 mL bottle with nitro- At the end of the experiment, the packing of each column
gen at 3 psi over 24 h. The total lipids extracted amounted to 9.4%. was extracted, glass beads were removed and the remainder
The algal residue remaining is described as LEA, COD value of 1.32 g was homogenized. To compare the amount of Cu retained in the
COD g−1 dwt algae. columns against the Cu recovered from the columns, sequential
extractions were performed using water, 1 M HCl, and 3:1 HNO3 -
2.6. Media for bioreactors HCl (15.7 and 12 M, respectively) sequentially to extract the Cu
retained in the packing. A single sample was collected from each
The basal mineral medium used for sulfate reducing experimen- column and the precipitates were characterized by X-ray diffrac-
tation contained (in mg L−1 ): NH4 Cl (30); K2 HPO4 (200); KH2 PO4 tometer (XRD), from (PANalytical X’Pert Pro, Netherlands). To
(300); CaCl2 ·2 H2 O (20), MgCl2 ·6 H2 O (80), and 1 mL per L of trace estimate the washout of algal biomass, a daily sample was analyzed
element solution [33]. Cu was added as CuSO4 ·5 H2 O (50), gradually by the previously described COD method. The cumulative volume
increased from 10 to 50 mg L−1 , and additional sulfate was added and the dilution factor were used to calculate the total COD lost by
as H2 SO4 to reach 250 mg L−1 . The pH was adjusted by addition of algae washout. The COD was quantified using the whole solution
NaOH. of the algae amended columns and corrected by the endogenous
column.
2.7. PRB columns The total sulfide (TS) associated with the liquid and gas was
calculated by Eq. (3). An effluent sample (0.5 mL) was analyzed for
This study was designed for a SRB process using 250 mg L−1 of dissolved sulfide (DS), the sum of H2 S + HS− + S2− , by the methylene
SO4 2− as the electron acceptor and algae as the e-donor, with no blue method. The concentration of undissociated sulfide [H2 S] was
other exogenous source of e-donor. Fig. 1 illustrates three glass calculated based on the DS, accounting for pH and using the first
columns (0.07 L). The same amounts of granular sludge, equiva- dissociation constant (pKa1 = 6.98).
lent to 12 g VSS L−1 reactor , and 2-mm diameter glass beads were
TS = DS × (1 + ␣0 × H × F) (3)
used to pack the columns and increase permeability. The endoge-
nous column did not contain any algae. The WCA column was filled where:
with intact Chlorella biomass and the LEA column was filled with TS = total sulfide in system (gas & liquid) per liter liquid
Chlorella biomass after lipid extraction. The mass composition of DS = dissolved sulfide (measured with methylene blue method)
the packing and the algae COD content are described in Table 2. ␣0 = fraction H2 S of DS
All columns were operated in continuous parallel mode at H = dimensionless H constant Cg/Caq (0.4)
24 ± 1 ◦ C with the same source of influent (synthetic ARD). The F = headspace volume/liquid volume (stripping factor)
338 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343

= 1/(10(pH − pKa) + 1)
␣0
pKa = dissociation constant of H2 S (6.98 at 25 ◦ C).
Sulfide precipitated in the mineral fraction was estimated by the
decrease in Cu. This calculation used a molar ratio of Cu/S = 1.0 with
the formation of CuS as the only precipitate.
Methane (CH4 ) production was measured by volume, using a
1.0-L empty gas sampling bag Tedlar-SCV (Sigma-Aldrich, MO) con-
nected to the upper part of each column. The amount of methane
was significant only during the first three weeks; after that period
the gas production declined to zero. In calculating methane pro-
duced, it was assumed the gas was composed 70% of methane [36].
The total initial algae COD (CODt0 ) was 7 g. At the end of this
study, the accounting of COD considered four fractions making up
the total COD: cumulative H2 S-COD (COD in all forms of sulfide),
algae washout-COD, CH4 -COD, and the COD remaining in the reac-
tor, which was calculated as follows:

CODremaining = CODt0 − (H2 S-COD + washout-COD + CH4 -COD).

The sulfur balance was calculated considering: 1) soluble H2 S-

S (measured + stripped), 2) SO4 2− -S in the effluent, and 3) CuS-S
due to the sulfur precipitated and adsorbed in the column during
periods when Cu was added.

3. Results

3.1. Sulfur

Both WCA and LEA contributed to sulfide formation in the efflu-

ent of the laboratory PRBs (Fig. 2A). In the case of WCA, the highest
effluent sulfide concentration, 50 mg L−1 , was observed in the latter
half of period (b). In the case of LEA, the maximum sulfide concen-
tration of 55 mg L−1 occurred at the end of period (e). Consistently,
the algae-amended columns produced orders of magnitude higher
effluent sulfide concentrations compared to the endogenous col-
umn. In fact, the endogenous column produced only very low levels
of sulfide (maximally 5 mg L−1 ) from day 5–40. Thereafter, there
was no longer detectable sulfide in the effluent of the endogenous
Fig. 2. Panel A: Total sulfide in the effluent (sulfide measured + stripped): Endoge-
column. nous column (), WCA column (䊏), and LEA column (). The concentration of Cu2+
Further evidence of sulfate reduction is supported by the in the influent during periods (a–e) was 0, 10, 30, 50 and 0 mg L−1 , respectively. The
removal of sulfate in the algae-amended columns (Fig. 2B). The maximum H2 S concentration stripped (at pH 6.5) was 0.23 mg L−1 -effluent during
highest sulfate removal occurred from the latter part of period (b) week 3. This corresponds to 5.1% of the total H2 S. Sulfide stripped was only corrected
during the first 40 days of active gas production. Panel B: Sulfate removal (%).
to the end of period (d) in the WCA column, or from period (c)
through (e) in the LEA column, corresponding to 60–80% sulfate
removal. There was little to no sulfate removal in the endogenous excellent on any given day; the sum of sulfur species was very sim-
column. ilar to the inlet S quantity. The balance in the endogenous column
Cu did not appear to impact sulfate reduction. Period (d), which primarily demonstrated that the effluent was sulfate-S equal to the
had the highest Cu concentration, also had the highest sulfate input sulfate-S (Fig. S3 of SD), what one would expect if no reaction
removal (Fig. 2B) and thus Cu inhibition was not witnessed. Sul- had taken place.
fide levels dropped in periods (c) and (d) when Cu concentrations
increased (Fig. 2A), but the drop was due to copper sulfide precip- 3.2. pH
itates (Fig. 3) and not to toxicity.
In the beginning of period (e), a noteworthy difference in behav- Fig. 4 shows the pH evolution in the effluent of the reactors.
ior was observed between the WCA and the LEA columns. Both The algae-amended columns had effluent pH values near neutral,
sulfate removal and sulfide production increased at a faster pace in ranging from 6.0 to 6.5. This contrasts starkly with the pH values
the LEA column. This indicates a greater release of e-donor from LEA in the endogenous columns, i.e., approximately 4.0, from period
as compared to WCA in the final period as the experiment was end- (b) onwards. The pH values were correlated with sulfate reduc-
ing. Alternatively, the sudden removal of Cu at the start of period tion activity. This activity was consistently high in algae-amended
(e) caused sulfide inhibition of SRB, and the LEA column acclimated columns. The endogenous column, however, had low activity from
to the toxicity faster. These observations taken together indicate the start of reactor operation until day 40, at which time activity
that both algae e-donor sources were quite suitable as e-donor for decreased to zero and concomitantly as the pH dropped to 4.0.
sulfate reduction and there was no evidence of Cu inhibition.
Sulfur balances during the system operation are shown for the 3.3. Copper
algae containing column in Fig. 3A and B. The figures clearly show
that sulfate was converted to aqueous sulfide (with a minor frac- The influent and effluent Cu concentrations are shown in Fig. 5.
tion of stripped S) and copper sulfide minerals. The balances are Fig. 5A is zoomed out to show the influent and effluent concen-
 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343 339

Fig. 5. Effluent and influent copper concentration: Panel A: View of data in the
range from 0 to 60 mg L−1 . Panel B: Close up view of effluent data in the range of
Fig. 3. Sulfur balance in WCA (panel A) and LEA (panel B) columns: sulfide aque- 0–0.6 mg L−1 . Endogenous column (), WCA column (䊏) and LEA column (). The
ous + stripped (䊏), sulfur in CuS (䊐), and sulfate effluent ( ). The dash shows the concentration of Cu2+ in the influent during periods (a, b, c, d and e) was 0, 10, 30, 50
sulfate concentration in the influent (5.1 mg SO4 2− -S d−1 ). and 0 mg L−1 , respectively. The dash line represents the influent Cu2+ concentration
in each of the periods.

trations ranging from 0 to 50 mg L−1 . Fig. 5B is zoomed in to the

concentration range from 0 to 0.5 mg L−1 in order to appreciate
differences at very low effluent concentrations. The endogenous
column reached Cu breakthrough towards the end of period (b).
Thereafter, only partial or no Cu retention was observed. At the
end of period (d), 50 mg L−1 of Cu were discharged with the effluent
and this was equal to the influent. This behavior is consistent with
lack of sulfate reduction. Conversely, Cu was effectively retained
in the algae-amended columns. Cu was retained in the reactor
throughout the experiment as shown in Fig. 5A. The Cu removal
percentages were higher than 99.5 and 99.7% for the WCA- and LEA
columns, respectively. The effluent Cu concentrations from these
columns are shown in Fig. 5B. The effluent Cu concentrations ranged
from 50 to 400 ␮g L−1 in the WCA column, and from non-detect to
450 ␮g L−1 in the LEA column.
Cu was precipitated as CuS as evidenced by XRD and sequential
extraction measurements. XRD analyses showed clear evidence of
CuS crystals that was only observable in the solids from the two
algae-amended columns (Fig. 6). No such evidence was found in
Fig. 4. Effluent pH: Endogenous column (), WCA column (䊏), and LEA column (). the endogenous column.
The horizontal line represents the influent pH.
The cumulative Cu balance is presented in Table 4. Firstly, the
table shows that there was approximately 4× more Cu retention
340 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343

Table 4
Sequential extraction of total copper from the homogenized column packing at the end of the experiment by means of: water, 1 M HCl, and 1:3 ratio (v/v) of concentrated
HNO3 :HCl.

Sequential extraction of Cu (mg)

Column Water 1 M HCl HNO3 :HCla Total extracted Cumulativeb retained (mg) % Cu Recoveredc

Endogenous 0.3 33.1 0.2 33.6 36.0 93.2

WCA 0.0 28.6 92.3 120.9 127.6 94.8
LEA 1.5 24.1 98.0 123.6 126.0 98.1
a
HNO3 -HCl, 3:1 (v/v); 15.7 M and 12 M, respectively.
b
Cumulative copper removal calculated from difference between flux in and flux out.
c
% Recovery = (Cu extracted from packing material/cumulative Cu retained) × 100.

sulfate-reducing activity during the first 40 days of the experiment

and then activity ceased. Thus, algae could serve as a long-term e-
donating substrate. Active sulfate reduction was linked to highly
efficient Cu removal and an increase in the pH of the treated ARD.

4.2. Algae as electron donor in anaerobic conditions

This study provides direct evidence that WCA and LEA

can serve as e-donors to drive sulfate reduction. Previously,
microbially-driven sulfate reduction in a constructed wetland was
hypothesized to be linked to the degradation of algal biomass
from two genera of naturally occurring algae in that wetland,
Scenedesmus and Carteria [28]. Additionally, extracellular polysac-
charides produced by a mixed culture of algae in a high-rate pond
were considered to be responsible for sulfate removal [29]. Studies
of a wastewater ponding system containing HM [37] demonstrated
that a combination of algae growing in the pond and co-disposed
Fig. 6. XRD spectrum of the column packing at the end of the column experiment.
Endogenous column (1), WCA column (2) and LEA column (3). Copper sulfide (cov- organic wastes served as a carbon source for sulfate reduction,
ellite) used as the reference. enabling Cu removal with an efficiency of 80% (from 500 mg L−1 ).
This study provides for the first time direct evidence that WCA
and LEA can be utilized as e-donor for sulfate reduction. In addi-
in the algae-amended columns compared to the endogenous col-
tion to algae, the use of phototrophic cyanobacterial biomass for
umn. Secondly, the retained Cu speciation was mostly in the form
sulfate reduction has also been reported. Sulfate reduction by a
of sulfides because most of the Cu could only be extracted by a
mixed culture utilized 31% COD from Spirulina sp. biomass as a
strong oxidative acid (HNO3 -HCl, 3:1). In the endogenous column,
sole carbon source [38]. The cell wall of Spirulina, similar to that
HNO3 -HCl extraction did not extract much copper because 99% of
of other cyanobacteria, has a high glucan content [39]. The cell
the Cu was extracted with water and 1 M HCl. Thirdly, the table
walls of Chlorella are rigid and more resistant to biodegradation
shows a good balance of Cu extracted versus cumulative Cu that
because they consist of a matrix containing glucosamine, the dom-
was retained in all columns (as estimated from inlet—outlet Cu con-
inant cell wall polymer, combined with hemicellulose with glucose
centrations). In the algae columns, the total recovery of Cu during
and mannose [40,41].
the extractions ranged from 95 to 98% of that retained.
Utilization of cellulosic and hemicellulosic materials as e-donor
by SRB has been demonstrated in several studies. A signifi-
3.4. Chemical oxygen demand cant sulfate reduction rate of 12.3 mg L−1 d−1 was observed in a
laboratory-scale study using rice straw as e-donor for the bioreme-
A cumulative COD balance is provided for the columns contain- diation of ARD (1.5 mg Cu L−1 ) at pH 2.0 [42]. The corresponding Cu
ing algae, WCA and LEA, in Fig. 7. During the initial 26 d of operation removal efficiency was 98%. In another study, grass cellulose was
(period (a)), biogas production provoked washout of the biomass, used as the carbon and energy source for microorganisms in rumen
accounting for the loss of 40.7 and 26.7% of the algae COD, respec- fluid, enabling sulfate removal efficiency of 86% [43].
tively. The COD metabolized to produce sulfide during the study While little is known about algae as a e-donor by sulfate-
corresponded to only 9.5 and 9.2% of the added algae-COD, respec- reducing consortia, it is well established that algal biomass can
tively. Only 3.5% and 1.3% of the WCA- and LEA-COD was converted serve as an e-donor for methane production. In large-scale algae
into methane-COD. cultures, anaerobic digestion is a necessary step to make microal-
gal biodiesel sustainable [44]. The COD conversion of algal biomass
4. Discussion from Chlorella sorokiniana and Chlorella vulgaris to methane by
anaerobic digestion ranged from 40–73% [45]. In a different study,
4.1. Main findings around 50% of the biomass of C. vulgaris was converted to methane
[46]. The methane yield of C. vulgaris biomass ranged from 189 to
Both WCA and LEA algae biomass served as e-donating 450 mL CH4 g−1 VS, Anaerobic digestion can also release nutrients
substrates for sulfate reduction. The biogenic sulfide formed pre- (nitrogen and phosphorus) from LEA [47], which are essential to
cipitated Cu2+ in the form of copper sulfide as evidenced by the make biodiesel production sustainable.
results of XRD and sequential extractions. Sulfate reduction was Sonication and thermal treatment enhance the methane yields
active in the algae-amended columns even as the experiment ended that can be obtained during anaerobic digestion of algal biomass.
on day 123. By comparison, the endogenous reactor exhibited low The solvent extraction step applied to remove the lipid extraction
 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343 341

Fig. 7. Algae-COD balance in WCA and LEA columns: Remaining algae (䊏), S2− -COD (䊐), CH4 -COD ( ), and washed out algae ( ).

is by itself a pretreatment that can increase methane production which is much lower than the IC50 value reported in the literature
from LEA [48]. Thermal hydrolysis of Nannochloropsis biomass at for hydrogenotrophic SRB.
170 ◦ C enhanced the methane yield by 40% for WCA and 15% for
LEA [49]. Another study investigating the anaerobic digestion of 4.4. Implications of algae use as e-donor for AMD remediation in
Scenedesmus biomass demonstrated a 2-fold and 1.6-fold increase PRB
in the methane yield, compared with untreated biomass, when the
biomass was pretreated by sonication at 128.9 MJ Kg−1 and thermal The utilization of LEA biomass may have remarkable advantages
hydrolysis at 80 ◦ C, respectively [50]. The increase was attributed as a slow release e-donor to remediate ARD in PRBs. Additionally,
to cell wall disruption and COD solubilization. algae can sustain PRBs for years producing benefits to the envi-
ronment by removing HM and acidity from ARD. LEA utilization
could also add profitability to the biodiesel industry. The main chal-
lenge in using algae as a reactive material in PRBs, however, is
4.3. Copper and sulfide toxicity to SRB and methanogens how to reduce the washout of the suspended algae that we sus-
pect was exacerbated by biogas production in the initial period.
In this study, copper and sulfide had no apparent inhibitory Eventually SRB will outcompete methanogens under prolonged
effect on sulfate-reducing activity. The 50% Cu inhibiting concen- sulfate-reduction and low pH conditions in ARD plumes.
tration (IC50 ) to acetoclastic and hydrogenotrophic sulfate reducers
was 32.3 mg L−1 and over 200 mg L−1 , respectively. In the same
study, the IC50 values of Cu2+ for acetoclastic and hydrogenotrophic 5. Conclusions
methanogens were reported as 20.7 and 8.9 mg L−1 , respectively
• WCA and LEA biomass were both shown to be effective e-donors
[51]. Greater IC50 values for SRB and methanogens have been
reported for Cu2+ , 1136 and 130 mg L−1 , in a different study [52]. to drive sulfate reduction of ARD, enabling the precipitation and
The reported Cu2+ inhibitory values on SRB were higher than the removal of Cu2+ in laboratory-scale PRBs.
• Efficient sulfate reduction was maintained during the exper-
concentration used in this study of 50 mg L−1 , thus our findings are
consistent with the lack of toxicity expected. imental period of 4 months, using a low pH, synthetic ARD
Undissociated sulfide (H2 S) is the main toxic form of sulfide, as influent. Sufficient algae-COD remained to sustain the operation
only the neutral form permeates the cell membrane [53]. A previ- for another 20 months.
• The Cu+2 removal efficiency was greater than 99.5% in the WCA-
ous study considering the toxicity of Cu+2 towards methanogens
and SRB have shown that the average IC50 values of undissociated and LEA-amended columns.
• Due to metal sulfide precipitation, the SRB community was not
H2 S at pH of 6.8 towards acetoclastic and hydrogenotrophic SRB
were 272 and 299 mg L−1 , respectively, while the IC50 values deter- inhibited at 50 mg Cu L−1 in the influent.
• The precipitate retained in the columns was composed mostly of
mined for acetoclastic and hydrogenotrophic methanogens were
97 and 136 mg L−1 , respectively [54]. In a similar study, the IC50 val- insoluble copper sulfide formed from the biogenic sulfide.
ues determined for sulfide at pH of 6.8 in assays with acetoclastic
and hydrogenotrophic SRB were 270 and 380 mg L−1 undissociated Acknowledgements
H2 S, respectively, while the IC50 values reported for acetoclastic
and hydrogenotrophic methanogens were 160 and 220 mg L−1 of This work was funded by a grant from the National Institute
undissociated H2 S, respectively [55]. In our study, the pH of the of Environmental Health Sciences-supported Superfund Research
effluent typically ranged from 6.4–6.7. In the worst case scenario, Program (NIH ES-04940). P. Ayala-Parra was partly supported by
this corresponds to 62.5 mg L−1 of undissociated H2 S (at pH = 6.5), the Mexican National Council of Science and Technology (CONA-
342 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343

CYT). We are grateful to Dr. Kimberly Ogden for the algal biomass [24] N. Sudasinghe, H. Reddy, N. Csakan, S. Deng, P. Lammers, T. Schaub,
provided and to Dr. Sue A. Roberts for XRD analysis. Temperature-dependent lipid conversion and nonlipid composition of
microalgal hydrothermal liquefaction oils monitored by Fourier transform ion
cyclotron resonance mass spectrometry, Bioenergy Res. 8 (2015) 1962–1972.
[25] A.E. Atabani, A.S. Silitonga, I.A. Badruddin, T.M.I. Mahlia, H.H. Masjuki, S.
Appendix A. Supplementary data
Mekhilef, A comprehensive review on biodiesel as an alternative energy
resource and its characteristics, Renew. Sust. Energ. Rev. 16 (2012)
Supplementary data associated with this article can be found, in 2070–2093.
the online version, at http://dx.doi.org/10.1016/j.jhazmat.2016.06. [26] A.E.M. Abdelaziz, G.B. Leite, P.C. Hallenbeck, Addressing the challenges for
sustainable production of algal biofuels: I. Algal strains and nutrient supply,
011. Environ. Technol. 34 (2013) 1783–1805.
[27] T.M. Mata, A.A. Martins, N.S. Caetano, Microalgae for biodiesel production and
other applications: a review, Renew. Sust. Energ. Rev. 14 (2010) 217–232.
References [28] R.A. Russell, P.J. Holden, K.L. Wilde, B.A. Neilan, Demonstration of the use of
Scenedesmus and Carteria biomass to drive bacterial sulfate reduction by
[1] D.B. Johnson, K.B. Hallberg, The microbiology of acidic mine waters, Res. Desulfovibrio alcoholovorans isolated from an artificial wetland,
Microbiol. 154 (2003) 466–473. Hydrometallurgy 71 (2003) 227–234.
[2] A. Giguere, P.G.C. Campbell, L. Hare, D.G. McDonald, J.B. Rasmussen, Influence [29] J.B. Molwantwa, N.P. Molipane, P.D. Rose, Biological sulfate reduction utilizing
of lake chemistry and fish age on cadmium, copper, and zinc concentrations algal extracellular products as a carbon source, in: WISA 2000 Biennal
in various organs of indigenous yellow perch (Perca flavescens), Can. J. Fish. Conference, Sun City, South Africa, 2000.
Aquat. Sci. 61 (2004) 1702–1716. [30] L.A. Jones, K.L. Ogden, F. Jia, Comparative study of biosorption of copper (II) by
[3] ATSDR, Toxicological Profile for Copper, Agency for Toxic Substances and lipid extracted and non-extracted Chlorella sorokiniana, Clean Soil Air Water
Disease Registry, U.S. Department of Health and Human Services, Atlanta, GA, 43 (2015) 73–78.
2004, pp. 314. [31] American Society for Testing and Materials, Standard Test Methods for
[4] S. Kousar, M. Javed, Evaluation of acute toxicity of copper to four fresh water Chemical Oxygen Demand (dichromate oxygen demand) of Water, in:
fish species, Int. J. Agric. Biol. 14 (2012) 801–804. D1252-95, Am. Soc. Test. Mater., Philadelphia, PA., 1995.
[5] N.M. Franklin, J.L. Stauber, S.J. Markich, R.P. Lim, pH-dependent toxicity of [32] J. Folch, M. Lees, G.H.S. Stanley, A Simple method for the isolation and
copper and uranium to a tropical freshwater alga (Chlorella sp.), Aquat. purification of the total lipides from animal tissues, J. Biol. Chem. 226 (1957)
Toxicol. 48 (2000) 275–289. 497–509.
[6] ATSDR, Toxicological Profile for Lead, Agency for Toxic Substances and [33] S. Karri, R. Sierra-Alvarez, J.A. Field, Zero valent iron as an electron-donor for
Disease Registry, U.S. Department of Health and Human Services, Atlanta, GA, methanogenesis and sulfate reduction in anaerobic sludge, Biotechnol.
2007 (pp. 582). Bioeng. 92 (2005) 810–819.
[7] ATSDR, Toxicological Profile for Cadmium, Agency for Toxic Substances and [34] H.G. Truper, H.G. Schlegel, Sulphur metabolism in thiorhodaceae. 1
Disease Registry, U.S. Department of Health and Human Services, Atlanta, GA, quantitative measurements on growing cells of Chromatium Okenii, Antonie
2012, pp. 487. Van Leeuwenhoek Microb. Serol. 30 (1964), 225-&.
[8] C.A. Cravotta, M.K. Trahan, Limestone drains to increase pH and remove [35] M.N. Hughes, M.N. Centelles, K.P. Moore, Making and working with hydrogen
dissolved metals from acidic mine drainage, Appl. Geochem. 14 (1999) sulfide the chemistry and generation of hydrogen sulfide in vitro and its
581–606. measurement in vivo: a review, Free Radic. Biol. Med. 47 (2009) 1346–1353.
[9] J. Geets, K. Vanbroekhoven, B. Borremans, J. Vangronsveld, L. Diels, D. Van der [36] A.P. Francese, G. Aboagye-Mathiesen, T. Olesen, P.R. Cordoba, F. Sineriz,
Lelie, Column experiments to assess the effects of electron donors on the Feeding approaches for biogas production from animal wastes and industrial
efficiency of in situ precipitation of Zn, Cd Co and Ni in contaminated effluents, World J. Microbiol. Biotechnol. 16 (2000) 147–150.
groundwater applying the biological sulfate removal technology, Environ. Sci. [37] P.D. Rose, G.A. Boshoff, R.P. van Hille, L.C.M. Wallace, K.M. Dunn, J.R. Duncan,
Pollut. Res. 13 (2006) 362–378. An integrated algal sulphate reducing high rate ponding process for the
[10] L.L. Barton, G.D. Fauque, Biochemistry, physiology and biotechnology of treatment of acid mine drainage wastewaters, Biodegradation 9 (1998)
sulfate-reducing bacteria, in: A.I. Laskin, S. Sariaslani, G.M. Gadd (Eds.), 247–257.
Advances in Applied Microbiology, vol. 68, Elsevier Academic Press Inc, San [38] G. Boshoff, J. Duncan, P.D. Rose, The use of micro-algal biomass as a carbon
Diego, 2009, pp. 41–98. source for biological sulphate reducing systems, Water Res. 38 (2004)
[11] A.H. Kaksonen, J.A. Puhakka, Sulfate reduction based bioprocesses for the 2659–2666.
treatment of acid mine drainage and the recovery of metals, Eng. Life Sci. 7 [39] C. Van Eykelenburg, A glucan from the cell wall of the Cyanobacterium
(2007) 541–564. Spirulina platensis, Antonie Van Leeuwenhoek 44 (1978) 321–328.
[12] G. Muyzer, A.J.M. Stams, The ecology and biotechnology of sulphate-reducing [40] H. Takeda, Sugar composition of the cell-wall and the taxonomy of Chlorella
bacteria, Nat. Rev. Microbiol. 6 (2008) 441–454. (Chlorophyceae), J. Phycol. 27 (1991) 224–232.
[13] G.J. Zagury, V.I. Kulnieks, C.M. Neculita, Characterization and reactivity [41] H.G. Gerken, B. Donohoe, E.P. Knoshaug, Enzymatic cell wall degradation of
assessment of organic substrates for sulphate-reducing bacteria in acid mine Chlorella vulgaris and other microalgae for biofuels production, Planta 237
drainage treatment, Chemosphere 64 (2006) 944–954. (2013) 239–253.
[14] J.F. Blais, Z. Djedidi, R.B. Cheikh, R.D. Tyagi, G. Mercier, Metals precipitation [42] J. Lu, T.H. Chen, J. Wu, P.C. Wilson, X.Y. Hao, J.H. Qian, Acid tolerance of an acid
from effluents: review, Pract. Period. Hazard. Toxic Radioact. Waste Manag. mine drainage bioremediation system based on biological sulfate reduction,
12 (2008) 135–149. Bioresour. Technol. 102 (2011) 10401–10406.
[15] I.S. Chang, P.K. Shin, B.H. Kim, Biological treatment of acid mine drainage [43] H.A. Greben, J. Baloyi, J. Sigama, S.N. Venter, Bioremediation of sulphate rich
under sulphate-reducing conditions with solid waste materials as substrate, mine effluents using grass cuttings and rumen fluid microorganisms, J.
Water Res. 34 (2000) 1269–1277. Geochem. Explor. 100 (2009) 163–168.
[16] P. Kijjanapanich, A.P. Annachhatre, G. Esposito, P.N.L. Lens, Use of organic [44] B. Sialve, N. Bernet, O. Bernard, Anaerobic digestion of microalgae as a
substrates as electron donors for biological sulfate reduction in gypsiferous necessary step to make microalgal biodiesel sustainable, Biotechnol. Adv. 27
mine soils from Nakhon Si Thammarat (Thailand), Chemosphere 101 (2014) (2009) 409–416.
1–7. [45] G. Polakovicova, P. Kusnir, S. Nagyova, J. Mikulec, Process integration of algae
[17] K.R. Waybrant, D.W. Blowes, C.J. Ptacek, Selection of reactive mixtures for use production and anaerobic digestion, in: P.S. Varbanov, H.L. Lam, J.J. Klemes, S.
in permeable reactive walls for treatment of mine drainage, Environ. Sci. Pierucci (Eds.), Pres 2012: 15th International Conference on Process
Technol. 32 (1998) 1972–1979. Integration, Modelling and Optimisation for Energy Saving and Pollution
[18] O. Gibert, J. de Pablo, J.L. Cortina, C. Ayora, Chemical characterisation of Reduction, Aidic Servizi Srl, Milano, Italy, 2012, pp. 1129–1134.
natural organic substrates for biological mitigation of acid mine drainage, [46] M. Ras, L. Lardon, S. Bruno, N. Bernet, J.P. Steyer, Experimental study on a
Water Res. 38 (2004) 4186–4196. coupled process of production and anaerobic digestion of Chlorella vulgaris,
[19] F. Obiri-Nyarko, S.J. Grajales-Mesa, G. Malina, An overview of permeable Bioresour. Technol. 102 (2011) 200–206.
reactive barriers for in situ sustainable groundwater remediation, [47] A.J. Ward, D.M. Lewis, B. Green, Anaerobic digestion of algae biomass: a
Chemosphere 111 (2014) 243–259. review, Algal Res. 5 (2014) 204–214.
[20] J. Fronczyk, M. Radziemska, Z. Mazur, Copper removal from contaminated [48] V.T.C. d. Neves, E.A. Sales, L.W. Perelo, Influence of lipid extraction methods as
groundwater using natural and engineered limestone sand in permeable pre-treatment of microalgal biomass for biogas production, Renew. Sust.
reactive barriers, Fresenius Environ. Bull. 24 (2015) 228–234. Energy Rev. 59 (2016) 160–165.
[21] P. Ayala-Parra, R. Sierra-Alvarez, J.A. Field, Treatment of acid rock drainage [49] M.E. Alzate, R. Munoz, F. Rogalla, F. Fdz-Polanco, S.I. Perez-Elvira, Biochemical
using a sulfate-reducing bioreactor with zero-valent iron, J. Hazard. Mater. methane potential of microalgae biomass after lipid extraction, Chem. Eng. J.
308 (2016) 97–105. 243 (2014) 405–410.
[22] D.C. Kligerman, E.J. Bouwer, Prospects for biodiesel production from [50] C. Gonzalez-Fernandez, B. Sialve, N. Bernet, J.P. Steyer, Comparison of
algae-based wastewater treatment in Brazil: a review, Renew. Sust. Energy ultrasound and thermal pretreatment of Scenedesmus biomass on methane
Rev. 52 (2015) 1834–1846. production, Bioresour. Technol. 110 (2012) 610–616.
[23] J.W. Richardson, M.D. Johnson, R. Lacey, J. Oyler, S. Capareda, Harvesting and [51] S. Karri, R. Sierra-Alvarez, J.A. Field, Toxicity of copper to acetoclastic and
extraction technology contributions to algae biofuels economic viability, Algal hydrogenotrophic activities of methanogens and sulfate reducers in
Res. 5 (2014) 70–78. anaerobic sludge, Chemosphere 62 (2006) 121–127.
 P. Ayala-Parra et al. / Journal of Hazardous Materials 317 (2016) 335–343 343

[52] J.L. Chen, R. Ortiz, T.W.J. Steele, D.C. Stuckey, Toxicants inhibiting anaerobic kinetics and sulphide toxicity thresholds of a range of methanogenic,
digestion: a review, Biotechnol. Adv. 32 (2014) 1523–1534. syntrophic and sulphate-reducing bacteria, Process Biochem. 33 (1998)
[53] S.J.W.H. Oude Elferink, A. Visser, L.W. Hulshoff Pol, A.J.M. Stams, Sulfate 555–569.
reduction in methanogenic bioreactors, Fems Microbiol. Rev. 15 (1994) [55] T. Yamaguchi, H. Harata, T. Hisano, S. Yamazaki, I.C. Tseng, Process behavior of
119–136. UASB reactor treating a wastewater containing high strength sulfate, Water
[54] V. O’Flaherty, T. Mahony, R. O’Kennedy, E. Colleran, Effect of pH on growth Res. 33 (1999) 3182–3190.