Fa

Aa

b

c

h

••••

a

ARRAA

KBWTSH

1

idfbmsh[

a1

a

0h

Journal of Hazardous Materials 262 (2013) 554– 560

Contents lists available at ScienceDirect

Journal of Hazardous Materials

jou rn al hom epage: www.elsev ier .com/ locate / jhazmat

ungal permeable reactive barrier to remediate groundwater in anrtificial aquifer

lbert Folcha,∗,1, Marcel Vilaplanab,1, Leila Amadoa,b, Teresa Vicentb, Glòria Caminalc

Departament de Geologia, Universitat Autònoma de Barcelona (UAB), 08193 Bellaterra, Barcelona, SpainDepartament d’Enginyeria Química, Universitat Autònoma de Barcelona (UAB), 08193 Bellaterra, Barcelona, SpainInstitut de Química Avanc ada de Catalunya (IQAC) CSIC, Jordi Girona 18-26, 08034 Barcelona, Spain

i g h l i g h t s

A permeable reactive barrier using Trametes versicolor was carried out.Orange G dye was selected as tracer pollutant to degrade in an artificial aquifer.Continuous degradation over 85% was accomplished continually for more than 8 days.Fungus can potentially be used as a permeable reactive barrier in real aquifers.

r t i c l e i n f o

rticle history:eceived 27 May 2013eceived in revised form 2 August 2013ccepted 4 September 2013vailable online 13 September 2013

a b s t r a c t

Biobarriers, as permeable reactive barriers (PRBs), are a common technology that mainly uses bacteria toremediate groundwater in polluted aquifers. In this study, we propose to use Trametes versicolor, a white-rot fungus, as the reactive element because of its capacity to degrade a wide variety of highly recalcitrantand xenobiotic compounds. A laboratory-scale artificial aquifer was constructed to simulate groundwaterflow under real conditions in shallow aquifers. Orange G dye was chosen as a contaminant to visually

eywords:iobarrierhite-rot fungi

rametes versicolorubsurface flowydrogeology

monitor the hydrodynamic behaviour of the system and any degradation of the dye by the fungus. Batchexperiments at different pH values (6 and 7) and several temperatures (15 ◦C, 18 ◦C, 20 ◦C and 25 ◦C)were performed to select the appropriate residence time and glucose consumption rate required forcontinuous treatment. The maximum Orange G degradation was 97%. Continuous degradation over 85%was achieved for more than 8 days. Experimental results indicate for the first time that this fungus canpotentially be used as a permeable reactive barrier in real aquifers.

. Introduction

In recent years, permeable reactive barriers (PRBs) have becomemportant in situ treatments for polluted groundwater [1]. PRBs areefined as emplacements of reactive media situated in the subsur-ace and used to treat polluted groundwater flowing through themy transforming pollutants into less harmful compounds. Reactionechanisms used in PRBs are chemical or biological degradation,

orption and precipitation [2]. Previous studies of PRB technologyave focused on sorption, precipitation or chemical degradation1,3–6]. Although some biological barrier studies have been

∗ Corresponding author. Present address: GHS, Dept. of Geotechnical Engineeringnd Geo-Sciences, Universitat Politècnica de Catalunya-BarcelonaTech, Jordi Girona-3, Mòdul D-2, 08034 Barcelona, Spain. Tel.: +34 93 401 1859.

E-mail addresses: folch.hydro@gmail.com, afolch.water@gmail.com,lbertfolch@hotmail.com (A. Folch).1 Both the authors contributed equally to the experimental work.

304-3894/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jhazmat.2013.09.004

© 2013 Elsevier B.V. All rights reserved.

performed at field scale [7–9], most studies are at lab scale andmust be extrapolated to field conditions [10–15], among others.

All previously reported biobarriers have utilised bacterialprocesses. However, fungi have a high potential for utility in biore-mediation processes because they can degrade a wide varietyof organic compounds by co-metabolism under oxic conditionswithout any previous conditioning [16]. Specifically, the ability ofwhite-rot fungi (WRF) to degrade several xenobiotic compoundshas been previously reported [17–19]. WRF possess an extracellu-lar enzymatic system comprising laccase and various peroxidasesthat can degrade pollutants without internalising them. This adap-tation permits the degradation of compounds with low solubilitiesand increases the tolerance of the fungi to high concentrations ofpollutant [16]. Most WRF also have an intracellular degradationsystem, the cytochrome P450 enzymatic system, which is able to

degrade certain polycyclic aromatic hydrocarbons (PAHs) [20] andchloro-organic pollutants [21], among others.

Despite the ability of WRF to degrade multiple contaminantsand the current knowledge about biodegradation by fungi in soils

dous M

[hktasldtOFcAawaa(ta

2

2

2s([

A

2

(

2

cTP

wd

magnetic stirrer at 25 ◦C. The agitation speeds ranged from 500 to

F(

A. Folch et al. / Journal of Hazar

22,23], studies of PRBs in aquifers using fungi as a reactive mediumave not been undertaken. This study was performed to fill thisnowledge gap. The objective of this study is to demonstratehe suitability of groundwater remediation in a lab-scale artificialquifer using the fungus Trametes versicolor. This fungus was cho-en because it is able to degrade a wide variety of pollutants in bothiquid and soil media, such as dyes [24–27], PAHs [28,29], endocrineisruptors [30,31], chloro-organic pollutants [21] and pharmaceu-ical and personal care products [32,33], among others. The azo dyerange G was chosen as test contaminant for two different reasons.irst, previous experiments have demonstrated that T. versicoloran extensively degrade Orange G by co-metabolic processes [25].n initial pollutant concentration of 150 mg L−1 was degraded up to

percentage of 97% after 20 h of batch treatment at conditions thatere optimal for fungal degradation: 25 ◦C, pH 4.5, glucose addition

t a rate of 0.037 g glucose g−1 dry weight h−1 (carbon source), andmmonium chloride at a nitrogen/carbon (N/C) ratio of 0.85 mg g−1

nitrogen source) [25]. Second, the use of Orange G allows visualisa-ion of the evolution of the flow and the degradation of the pollutantlong the artificial aquifer.

. Materials and methods

.1. Fungus and chemicals

T. versicolor (ATCC#42530) was maintained by subculturing on% malt extract agar slants (pH 4.5) at 25 ◦C. Subcultures weretarted every 30 days. Malt extract was obtained from ScharlauBarcelona, Spain). Pellets were produced as previously described34].

Orange G was obtained from Sigma–Aldrich (Barcelona, Spain).ll other chemicals used were analytical grade.

.2. Lab-scale aquifer design and assembly

The artificial aquifer was simulated using a glass tank1 m × 0.05 m × 0.15 m) with a total volume of 7.5 L (Fig. 1).

.2.1. Aquifer matrix and biobarrierThe porous aquifer matrix was simulated with silica sand (parti-

les 2–4 mm) (Fig. 1), which was first washed with distilled water.he tank was filled with the porous media following the Slurry

acking methodology until reaching a height of 0.12 m [35].

At the centre of the tank, a sand-free area was maintained inhich the biobarrier, with the fungus in the form of pellets (2 mmiameter approximately), was installed (Fig. 1). This space was

ig. 1. Schematic for a lab-scale aquifer containing a T. versicolor biobarrier for the reme2) air suppliers; (3) sampling points; (4) fungus biobarrier; (5) aquifer matrix (silica sand

aterials 262 (2013) 554– 560 555

delimited by two water-permeable plastic walls with a mesh poresize of 1 mm placed on each side of the biobarrier area in order tocontain the sand. In each experiment, the dimensions of the biobar-rier structure were selected to obtain the appropriate HRT to reachthe required reaction time in the PRB. Three sampling points wereused (Fig. 1): one in the aquifer inlet, a second one in the biobarrierand the final one in the aquifer outlet.

Polyurethane foam layers (1 cm wide) were placed between thematrix-free area and the sand to maximise fluid dispersion andguarantee complete mixing before and after circulation throughthe aquifer matrix.

2.2.2. Aquifer and biobarrier operating conditionsThe synthetic dye wastewater (SDW) consisted of a solution of

Orange G, glucose and ammonium chloride. In certain experiments,sodium bicarbonate was added to the SDW to generate a buffer. TheOrange G concentration was kept constant in all the experimentsat 150 mg L−1, while the glucose and ammonium chloride concen-trations were varied to satisfy the consumption rates of T. versicolor(active treatment) and the concentrations used are specified in thedescription of each experiment. When preparing the SDW, the pHwas adjusted as specified in the description of each experiment. Thesolution was then sterilised prior to being pumped into the non-sterile aquifer. The inflow rate was chosen to simulate typical flowconditions of shallow aquifers, which correspond to a flow velocityof less than 2 m day−1.

Air suppliers (Fig. 1) were used to introduce the required air flowinto the biobarrier to maintain the metabolic activity of the fungusand the pellet biomass in fluidisation (active treatment). The air wasdistributed by stone diffusers situated at the bottom of the tank inthe biobarrier.

2.3. Artificial aquifer characterisation

2.3.1. Orange G adsorption testA test was performed to determine the losses resulting from

the adsorption of the dye for each material present in the artifi-cial aquifer. The tested materials were the silica sand, the plasticsupports, the permeable plastic walls and the polyurethane foam.

Each material was added to a glass beaker containing 200 mLof an Orange G solution (150 mg L−1). The beaker was placed on a

1100 rpm. The time of contact of the materials with the dye solutionwas 48 h for all cases, except for the silica sand which was 71 h. Dyesolution samples (1 ml) were collected to determine the change inOrange G concentration over time.

diation of synthetic Orange G wastewater. Legend: (1) wastewater feeding vessel;); (6) Polyurethane foam layers; and (7) vessel collecting aquifer outflow.

5 dous M

2

sw

o

�

wtwppstwd

2

2

oeatcocvbwLfeSp1

t

2

icco1

tecwit

i

TPs

56 A. Folch et al. / Journal of Hazar

.3.2. PorosityThe porosity of the sand was determined by adding 255.5 g of

and to a 250 mL graduated cylinder that was filled with distilledater until the water level and the sand level were equal.

The effective porosity was determined by analysing the velocityf Orange G flow using the following equation [36]:

= t · q

L(1)

here � is the effective porosity; t is the transit time between thewo points being considered (h); q is the specific flow (cm h−1),hich is defined as the coefficient between the inflow and the per-endicular section; and L is the distance (cm) between the twooints being considered. The two points that define L are the initialilica sand placed at the entrance of the tank and the entrance ofhe biobarrier (Fig. 1). t is defined as the time between the time athich the Orange G contacts the silica sand and the time when theye arrives at the biobarrier space.

.4. Experimental procedures

.4.1. Orange G degradation in batch experimentsThe objective of the experiments was the evaluation of the effect

f temperature and pH on Orange G degradation by fungus. Thesexperiments were performed in sterile conditions and in triplicatet different temperatures and pH. 250 mL Erlenmeyer flasks con-aining 150 mL of SDW (8 g L−1 glucose and 2.6 mg L−1 ammoniumhloride) were used. Pellets were inoculated at a concentrationf 66.7 g wet pellets L−1, which corresponds to the biomass con-entration used in previous dye degradation experiments by T.ersicolor [24]. The flasks were then incubated in a shaking waterath (Lauda, Lauda-Königshofen, Germany), and the temperatureas controlled by an RA120 cooling thermostat system (Lauda,

auda-Königshofen, Germany). The treatments were performed atour different temperatures (15, 18, 20 and 25 ◦C) and two differ-nt initial pH values (6 and 7). The experiment to degrade bufferedDW, which also contained 300 mg L−1 sodium bicarbonate, waserformed at 18 ◦C and initial pH 7. The orbital agitation speed was50 rpm in all the experiments.

One millilitre samples were taken at various times to determinehe change in Orange G and glucose concentrations over time.

.4.2. Continuous treatment in the lab-scale aquiferTwo experiments measuring SDW degradation were performed

n the artificial aquifer in a continuous mode and under non-sterileonditions (Table 1). To reduce the experimental time, the artifi-ial aquifer was first filled with SDW; then, the pump was turnedn to introduce SDW into the aquifer. The inflow rate was kept at4 mL h−1 in all of the experiments.

Experiment 1 was performed without temperature and pH con-rol. SDW (optimum pH for the fungi of 4.5, glucose concentrationquivalent to 0.037 g glucose g−1 dry weight h−1 and ammoniumhloride concentration equivalent to an N/C ratio of 0.85 mg g−1)as used as inflow and an HRT of 41.3 h was selected. Pellets were

noculated at the same concentration as in batch degradation (Sec-ion 2.4.1).

Experiment 2 required changes to the aquifer system usedn experiment 1. The volume of the biobarrier was increased

able 1arameter values for a PRB for orange G synthetic wastewater treatment in a lab-cale aquifer.

Experiment Biobarriervolume (L)

Pellets amount(g-dry weight)

HRTaquifer (h) HRTbiobarrier (h)

1 0.57 1.3 106 41.32 0.97 3.7 113 70

aterials 262 (2013) 554– 560

to 0.97 L to obtain an HRT of 70 h. HCO3− (300 mg L−1) in the

form of sodium bicarbonate was added to SDW. Wastewaterwas adjusted to pH 7 and a glucose concentration equivalent to0.013 g g−1 dry weight h−1. Ammonium chloride composition wasequivalent to an N/C ratio of 0.85 mg g−1. SDW glucose concentra-tion was varied during treatment to guarantee the stable metabolicactivity of the fungus. A RA120 cooling thermostat system (Lauda,Lauda-Königshofen, Germany) was used to maintain temperatureat 18 ◦C. The experimental system was equipped with a pH con-troller (Mettler Toledo, Barcelona, Spain) to measure and controlthe pH in the biobarrier with NaOH (0.075 M) and HNO3 (0.15 M).

For each experiment, 1 ml samples were taken daily at previ-ously indicated sampling points to measure Orange G and glucoseconcentrations. Temperature and pH were measured daily using apH 330i probe at the time of sampling (WTW, Weilheim, Germany).

Degradation percentage of Orange G was calculated accordingto the following equation:

% = Cin − Cout

Cin× 100 (2)

where Cin corresponds to Orange G concentration in the inflow andCout represents dye concentration in the outflow.

2.5. Analytical procedures

Colour was determined by spectrophotometric measurementsat the visible maximum absorbance, 478 nm, using a VarianUV/Visible Cary spectrophotometer (Palo Alto, CA, USA). Glucoseconcentration was measured using an YSI 2000 enzymatic analyser(Yellow Springs, OH, USA). Pellet dry weight was determined aftervacuum filtering the cultures through glass filters of known weight(Whatman GF/C, 47-mm-diameter, Maidstone, England). The fil-ters containing the biomass were placed in glass dishes and driedat 105 ◦C to a constant weight.

3. Results and discussion

3.1. Aquifer characterisation

Adsorption tests showed that Orange G was not adsorbedby tested materials. Previous studies had demonstrated that theOrange G is adsorbed by the fungus during the first hours ofthe degradation process causes an observable colouration of thebiomass surface and that this dye is subsequently degraded by thefungus. This degradation causes the colour to disappear from thefungal pellets without an observable increase in Orange G concen-tration in liquid phase [25].

Before beginning the degradation experiments, the ground-water flow in the artificial aquifer was checked using Orange G(150 mg L−1) as a tracer. After 30.1 h, Orange G reached the zoneof the biobarrier crossing it with a homogenous distribution dueto the air supplied to maintain fluidization. Between 77 and 95 hafter injection the dye concentration reached the outflow point.Using this information about the flow, the effective porosity wasestimated to be approximately 36% using Eq. (1). This value is con-sistent with the 36.5% total porosity that was determined using agraduated cylinder.

3.2. Lab-scale aquifer experiment at non-controlled conditions

Experiment 1 was performed in non-sterile conditions and

without controlling pH or temperature to simulate the con-ditions in a real aquifer. The glucose input was adjusted to0.037 g glucose g−1 dry weight h−1 to limit the growth and promotethe enzyme production of T. versicolor [25].

A. Folch et al. / Journal of Hazardous M

F(

orsGtalstetnaimrtpic

bsmh

da

time (Table 2). In addition, pH values at the end of the experi-

TOe

ig. 2. Time course of Orange G degradation by T. versicolor (�) and temperature©) in the biobarrier within a lab-scale aquifer under non-controlled conditions.

Orange G degradation percentage by the fungus under continu-us treatment was expected to increase until a steady state waseached, after which point the degradation would remain con-tant. However, the evolution of degradation percentage of Orange

degradation was different from the expected value. After 26 h,he degradation percentage probably started to decline because of

decrease in temperature (Fig. 2). The degradation became neg-igible when aquifer temperature dropped to 13.3 ◦C. This resulthows that the artificial aquifer must be maintained at a constantemperature to simulate the conditions found in real aquifers. Thexperiment was continued to obtain additional information abouthe degradation process. The effect of low temperature was elimi-ated by installing a thermostat-controlled heater near the aquifert 255 h. The heater increased the temperature and maintainedt at a value between 16 and 22 ◦C until the end of the experi-

ent. As a result, the degradation in the biobarrier increased until iteached a maximum value of 60.5% between 350 and 400 h. Then,he degradation started to decrease, notably after the 400 h timeoint probably because the pH had decreased below 4. The exper-

ment was halted at 500 h because the degradation percentageontinuously decreased.

Orange G concentrations measured simultaneously in the bio-arrier and at the aquifer outlet were very similar when the steadytate was reached. This finding suggests that the potential growth oficroorganisms in the aquifer because of the non-sterile conditions

ad no significant effect on Orange G degradation.

Glucose consumption was relatively low; glucose concentration

ecreased from 0.037 to 0.026 ± 0.01 g glucose g−1 dry weight h−1

cross the biobarrier, depending on temperature.

able 2range G degradation percentage, final pH, reaction time, degradation constant (includixperiments performed in triplicate at varying temperature for each initial pH.

Temperature (◦C)(initial pH)

Orange G degradation (%) Final pH Reacti

15 (pH 6) 91.9 ± 5.1 3.58 ± 0.01 60.25

15 (pH 7) 89.2 ± 2.8 3.60 ± 0.02 60.25

15 (pH 6–7)a

18 (pH 6) 96.1 ± 1.4 3.63 ± 0.02 47.75

18 (pH 7) 93.2 ± 4.5 3.69 ± 0.02 47.75

18 (pH 6–7)a

20 (pH 6) 94.1 ± 4.8 4.01 ± 0.02 23

20 (pH 7) 96.3 ± 0.6 3.96 ± 0.06 23

20 (pH 6–7)a

25 (pH 6) 98.2 ± 0.1 4.01 ± 0.12 12.25

25 (pH 7) 98.0 ± 0.3 4.37 ± 0.39 12.25

25 (pH 6–7)a

a Means average value.

aterials 262 (2013) 554– 560 557

Temperature and pH affect the degradation percentage ofOrange G and glucose consumption by the fungus. In aquifers, thetemperature remains relatively constant throughout the year [37],but it is lower than 25 ◦C, which is the optimal temperature forthe degradation activity of T. versicolor [38]. Therefore, a studyto evaluate the dye degradation rates and glucose consumptionrates at different constant temperatures and initial pH values wasperformed. Moreover, the glucose concentration at the biobarrieroutlet must be almost zero to avoid an increase in the aquifer chem-ical oxygen demand (COD), which would promote the growth ofmicroorganisms present in the aquifer.

3.3. Batch degradation experiments at different temperatures

The batch experiments designed to measure the degradation ofSDW were performed at 15, 18, 20 and 25 ◦C and initial pHs of themedium of 6 and 7 for each temperature. The chosen temperatureand pH values were in the range observed for groundwater [39].

The results of the degradation experiments show that T. ver-sicolor is able to degrade significant quantities of Orange G at alltemperatures and achieve final degradation percentages higherthan 89% in all cases (Table 2). These results are the first evidenceof the ability of this fungus to degrade this pollutant at tempera-tures below 25 ◦C. Previous studies of Orange G degradation by thisfungus were only performed at 25 ◦C [25]. Decreasing the tempera-ture increases the time required to reach a degradation percentageover 89% (Table 2). Other studies have also demonstrated the abil-ity of WRF to degrade pentachlorophenols and PAHs present insoil at temperatures between 15 and 17 ◦C [17]; but in solid phasetreatment that is the fungus’ natural habitat.

Initially, the dye concentration decreased slowly over a longperiod of time (Fig. 3); in this case, the time prior to 49 h could beconsidered a lag period. The main cause for this phenomenon is thehigh initial pH compared to the pH of 4.5, which is optimal for themetabolic activity of T. versicolor [25]. This difference in pH slowsthe degradation reaction until the fungus acidifies the medium toa value close to the optimum pH. The medium acidification ratecaused by the fungus varies with the temperature; this variation isthe main reason that the lag period for degradation increases as thetemperature decreases. At 49 h, when pH is approximately equal tothe optimal value, Orange G degradation increased notably, and afinal degradation percentage of 89.2% was achieved 10 h later.

Varying the initial pH of the medium between 6 and 7 did notaffect the degradation efficiency of the dye in terms of the reaction

ments were notably lower than the optimum value. In the lab-scaleaquifer, the pH value in the biobarrier is a key parameter affectingthe degradation efficiency. Therefore, a pH controller is necessary

ng an average value at each temperature) and glucose consumption rate for batch

on time (h) Degradation constant(k – g−1 dry weight h−1)

Glucose consumption rate(g glucose g−1 dry weight h−1)

0.042 ± 0.002 0.0150.042 ± 0.004 0.0170.042 ± 0.0010.084 ± 0.011 0.0120.099 ± 0.003 0.0130.091 ± 0.0110.114 ± 0.047 0.0410.152 ± 0.007 0.0460.133 ± 0.0270.246 ± 0.007 0.0450.292 ± 0.004 0.0260.269 ± 0.032

558 A. Folch et al. / Journal of Hazardous M

Fig. 3. Time course of the concentration of Orange G caused by T. versicolor degrada-tat

teppcpt

i2

odkttttktTat4a

tGptwoipd

k

w

k

w

w

w

ion [(�), (©) and (�)] in triplicate for the batch experiment (15 ◦C and initial pH 7)nd the corresponding fit of the first-order equation ( ) to values measured duringhe maximum degradation period.

o maintain the pH at approximately the optimal value during thexperiment and avoid excessive acidification of the medium. TheH decrease is caused by the production of different acidic com-ounds by the fungus, and pH values less than the optimal valuean reduce the metabolic activity of T. versicolor [40]. Fig. 5 (Sup-lementary material) shows the evolution of the degradation underhe other temperature conditions tested and initial pH levels tested.

Supplementary material related to this article can be found,n the online version, at http://dx.doi.org/10.1016/j.jhazmat.013.09.004.

The decrease in Orange G concentration can be fitted to a firstrder kinetic equation, as has been reported previously for theegradation of different dyes by WRF [26,41,42]. In all cases, theinetic equation was fitted only to dye concentration values duringhe period of maximum degradation (Fig. 3); the concentrations inhe initial lag period were not used. A good fit of the kinetic equationo the experimental values was obtained, resulting in a correla-ion coefficient (r2) higher than 0.96 in all cases. The first orderinetic constant (k) (Table 2) increased significantly with tempera-ure, reaching a maximum value at 25 ◦C for both initial pH values.he k values obtained for both pH values at the same temperaturere similar because the degradation of Orange G only starts whenhe pH of the medium has decreased significantly and is close to pH.5. Thus a constant average value for degradation can be obtainedt each temperature (Table 2).

The dependence of k on temperature (T) can be expressed usinghe Arrhenius equation, and the activation energy (Ea) for Orange

degradation by T. versicolor can be calculated for each initialH (Eqs. (3) and (4)). Obtaining the Arrhenius equation parame-ers allows determining the value of k at a certain temperature,hich can be useful to determine the HRT and the consumption

f glucose of a continuous PRB containing fungi. This calculations possible because when the system reaches the steady state, theH will approximate the optimal value, and the dye concentrationecrease should follow first order behaviour.

= 1.91 × 1021 · e(−124650/R·T),

hen the initial pH is 6 and r2 is 0.99. (3)

= 1.32 × 1023 · e(−134597/R·T),

2

hen the initial pH is 7 and r is 0.96. (4)

here R is 8.31 J K−1 mol−1.The difference between activation energies is less than 10%,

hich also demonstrates the marginal effect of the initial pH of

aterials 262 (2013) 554– 560

the medium (between 6 and 7) on the degradation process (Eqs.(3) and (4)).

Therefore, a lab-scale biobarrier operating on a continuousmode at a temperature between 15 and 25 ◦C is viable. However,if the temperature changes during the operation of the biobarrier,the fungus must adapt, which reduces the degradation efficiency,as was observed in experiment 1 (Fig. 2).

Despite the lack of a clear trend with increasing temperature(Table 2), differences between the glucose consumption rates areobserved for both pHs at high (20–25 ◦C) and low (15–18 ◦C) tem-peratures. These results show that it is necessary to calculate theglucose consumption rate at the temperature at which the treat-ment is performed when the steady state is reached. Accordingly,the addition of glucose can be matched to the rate of consumptionby the fungus, and accumulation in the biobarrier outflow can beavoided.

Once the ability of T. versicolor to degrade Orange G at differ-ent temperatures was demonstrated, a complementary experimentwas performed to investigate the effect of HCO3

− buffer. Thebicarbonate concentration was chosen to be 300 mg L−1 to matchthe concentration range of real groundwater. An experiment per-formed to measure the degradation of buffered SDW at 18 ◦C andan initial pH of 7 resulted in 97.1% degradation after 44.25 h oftreatment (Fig. 6 Supplementary material). These results indicatethat the presence of bicarbonate does not significantly affect thedegradation process; carbonate concentrations, as observed in realaquifers, do not have sufficient buffering strength to prevent thefungus from decreasing the groundwater pH.

Supplementary material related to this article can be found,in the online version, at http://dx.doi.org/10.1016/j.jhazmat.2013.09.004.

Results obtained from batch experiments at different temper-atures and pH values indicate that T. versicolor can potentially beused as a biobarrier. As shallow aquifers tend to have a temperaturesimilar to the annual average air temperature of the area in whichthey are located [37], T. versicolor can be used in many subtropicaland temperate zones [43]. Even in cooler, temperate areas, T. versi-color may be used if we consider that urban areas tend to increasethe average groundwater temperature [44].

3.4. A lab-scale aquifer experiment under controlled conditions

To verify the results obtained from the batch experiments,experiment 2 was performed under non-sterile conditions at aconstant temperature of 18 ◦C and a water composition similar tonatural groundwater (Section 2.4.2).

The degradation in the biobarrier increased from the beginningof the experiment until it reached a maximum value of 98% at 159 h(Fig. 4). At this point, the degradation began to decrease slowly butremained above 85% until 328 h. NaOH solution was used to controlthe pH in the biobarrier from the beginning of the experiment to328 h to maintain the pH above 4.5 in the biobarrier. Between 1and 138 h, a total amount of 0.002 g of NaOH L−1 wastewater h−1

was used. Subsequently, the pH reached a value slightly higher than4.5 without the addition of NaOH. Although the optimal pH for T.versicolor is 4.5, the pH was not modified because the degradationefficiency was high.

Between 328 and 352 h, the outlet tube was clogged because oftechnical problems, which modified the experimental conditionsin the biobarrier: the pH increased to 6.52, and the degradationdecreased to 63%. After 352 h, normal flow was restored, and the

pH was controlled with HNO3 to maintain the optimal pH of 4.5.After the restoration of the outflow and pH control, the degradationincreased again to 90% at 376 h. However, after this period of time,the degradation decreased to 42% at 544 h due to previous technical

A. Folch et al. / Journal of Hazardous M

Fi

pl

svba

tci

wfitTngmcp

bbBbideasa

4

atwtwb

mtat

[

[

[

ig. 4. Time course of the degradation of Orange G by T. versicolor (�) and pH (©)n the biobarrier within a lab-scale aquifer under controlled conditions.

roblems in the aquifer that altered the pH in the biobarrier for ateast one day. The experiment was ended at this time.

Similar to experiment 1, Orange G concentrations measuredimultaneously in the biobarrier and at the aquifer outlet wereery similar when the steady state was reached. Thus, the possi-le growth of microorganisms in the aquifer did not significantlyffect Orange G degradation.

The significant rate of Orange G degradation that was main-ained over several days in this lab-scale aquifer at non-sterileonditions (Fig. 4) validates the values obtained in the batch exper-ments.

One of the problems of using T. versicolor for real in situ ground-ater treatment is the reduction in pH caused by the activity of the

ungus. To improve the applicability of this fungus as a PRB, dur-ng experiment 2, the outflow water (pH 4.5) was forced to flowhrough a reservoir of carbonate sand with a residence time of 36 h.his treatment increased the pH to 7.9. When applying this tech-ology at the field scale, the PRB can be surrounded with carbonateravel at the outflow side instead of other geological materials toaintain the initial pH value of the aquifer. Alternatively, a basic

ompound (NaOH or carbonate salt) can be added to increase theH of the biobarrier outflow.

To perform long-term, continuous treatment at field scale, it wille necessary to develop a strategy of biomass renovation in theiobarrier to maintain stable fungal activity for as long as possible.lánquez et al. [45] determined a strategy to purge and renovate theiomass. Their method consisted of partially replacing the biomass

n a continuous treatment mode to degrade the Grey Lanaset Gye with T. versicolor in a bioreactor at 25 ◦C and pH 4.5. In thatxperiment, 1/3 of the biomass was replaced every 7 days to obtain

cellular retention time of 21 days. In the case of PRBs, a similartrategy will be suitable to assure successful in situ treatment overn extended period of time.

. Conclusions

A PRB using the fungus T. versicolor under non-sterile conditionsnd continuous flow in a lab-scale aquifer was shown to degradehe test pollutant Orange G. The maximum Orange G degradationas 97%. Continuous degradation over 85% was achieved for more

han 8 days. The pH, bicarbonate concentration and temperature athich the degradation occurred indicate that this technology can

e used in real aquifers.The glucose and ammonium chloride addition rates were

atched with the rates at which they were consumed by the funguso avoid an increase in the chemical oxygen demand (COD) of thequifer. In a real, field-scale PRB, glucose would be directly addedo the biobarrier. Sterile glucose would not be necessary because

[

aterials 262 (2013) 554– 560 559

of the large amount of fungi biomass present in the PRB, whichcould accommodate a low volume of highly concentrated glucosesolution. The high glucose concentration would inhibit microor-ganism growth during storage. Furthermore, as the fungal activityreduces the pH to approximately 4.5, the activity of autochthonousbacteria in the barrier would be low [46]. The addition of a basecompound, such as NaOH, directly to the biobarrier to maintain apH at 4.5 would allow keeping pollutant degradation efficiency ata high level for a longer period of time.

In contrast with other PRB technologies, the use of T. versicoloror other WRF would allow for several types of pollutants to betreated. Therefore, this technology, which only uses one organism,can potentially be applied to treat contamination sites containingdifferent pollutants.

More research is needed to determine the ability of the fun-gus to degrade a pollutant present in a real aquifer or how naturalhydrogeological conditions will affect the degradation efficiency(e.g., autochthonous bacteria and water composition). However, arecent study has demonstrated that T. versicolor is able to persistin non-sterilised sewage sludge for a long period of time [47]. Fur-thermore, the optimal pH for pollutant degradation by the fungi isin the acid range, which favours the fungi in their competition withautochthonous bacteria in the aquifer. In conclusion, this studyindicates that the application of WRF in a PRB is a viable strategythat can be tested at field scale in the near future.

Acknowledgments

This work was funded by projects CGL2011-29975-c04-01 and04/BTE and CTQ2010-21776-C02-01 from the Spanish Ministry ofEconomy and Competitiveness. The Department of Chemical Engi-neering of the Universitat Autònoma de Barcelona (UAB) is memberof the Xarxa de Referència en Biotecnologia de la Generalitat deCatalunya.

References

[1] R. Thiruvenkatachari, S. Vigneswaran, R. Naidu, Permeable reactive barrier forgroundwater remediation, J. Ind. Eng. Chem. 14 (2008) 145–156.

[2] K.E. Roehl, T. Meggyes, S.G. Simon, D.I. Stewart, Long-term Performance ofPermeable Reactive Barriers, Elsevier Publishers, Amsterdam, 2005.

[3] S.H. Liang, C.M. Kao, Y.C. Kuo, K.F. Chen, B.M. Yang, In situ oxidation ofpetroleum-hydrocarbon contaminated groundwater using passive ISCO sys-tem, Water Res. 45 (2011) 2496–2506.

[4] X. Lu, M. Li, C.M. Tang, C.P. Feng, X. Liu, Electrochemical depassivation forrecovering Fe-0 reactivity by Cr(VI) removal with a permeable reactive barriersystem, J. Hazard. Mater. 213 (2012) 355–360.

[5] G. Bartzas, K. Komnitsas, Solid phase studies and geochemical modelling oflow-cost permeable reactive barriers, J. Hazard. Mater. 183 (2010) 301–308.

[6] O. Gibert, T. Rötting, J.L. Cortina, J. de Pablo, C. Ayorac, J. Carrera, J. Bolzicco,In-situ remediation of acid mine drainage using a permeable reactive barrierin Aznalcóllar (Spain), J. Hazard. Mater. 191 (2011) 287–295.

[7] S.G. Benner, D.W. Blowes, W.D. Gould, R.B. Herbert, C.J. Ptacek, Geochemistryof a permeable reactive barrier for metals and acid mine drainage, Environ. Sci.Technol. 33 (1999) 2793–2799.

[8] P.C. Johnson, C.L. Bruce, K.D. Miller, A practical approach to the design, mon-itoring, and optimization of in situ MTBE aerobic biobarriers, Ground WaterMonit. Remediat. 30 (2010) 58–66.

[9] L.A. Schipper, M. Vojvodic-Vukovic, Five years of nitrate removal, denitrifi-cation and carbon dynamics in a denitrification wall, Water Res. 35 (2001)3473–3477.

10] W.J. Hunter, L.D. Kuykendall, Removing selenite from groundwater with anin situ biobarrier: laboratory studies, Curr. Microbiol. 50 (2005) 145–150.

11] C.M. Kao, S.C. Chen, J.Y. Wang, Y.L. Chen, S.Z. Lee, Remediation of PCE-contaminated aquifer by an in situ two-layer biobarrier: laboratory batch andcolumn studies, Water Res. 37 (2003) 27–38.

12] Y.C. Lee, S.G. Woo, E.S. Choi, Y. Ahn, J. Park, M. Lee, J.W. Yang, Bench-scale exsitu diesel removal process using a biobarrier and surfactant flushing, J. Ind.Eng. Chem. 18 (2012) 882–887.

13] M.M. Michalsen, A.D. Peacock, A.N. Smithgal, D.C. White, A.M. Spain, Y. Sanchez-Rosario, L.R. Krumholz, S.D. Kelly, K.M. Kemner, J. McKinley, S.M. Heald, M.A.Bogle, D.B. Watson, J.D. Istok, Treatment of nitric acid-, U(VI)-, and Tc(VII)-contaminated groundwater in intermediate-scale physical models of an in situbiobarrier, Environ. Sci. Technol. 43 (2009) 1952–1961.

5 dous M

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

60 A. Folch et al. / Journal of Hazar

14] S. Saponaro, M. Negri, E. Sezenna, L. Bonomo, C. Sorlini, Groundwater remedia-tion by an in situ biobarrier: a bench scale feasibility test for methyl tert-butylether and other gasoline compounds, J. Hazard. Mater. 167 (2009) 545–552.

15] S. Viamajala, B.M. Peyton, R. Gerlach, V. Sivaswamy, W.A. Apel, J.N. Petersen,Permeable reactive biobarriers for in situ Cr(VI) reduction: bench scale testsusing Cellulomonas sp. Strain ES6, Biotechnol. Bioeng. 101 (2008) 1150–1162.

16] C.A. Reddy, Z. Mathew, Bioremediation potential of white rot fungi, in: G.MGadd (Ed.), Fungi in Bioremediation, Cambridge University Press, Cambridge,UK, 2001, pp. 52–78.

17] P. Baldrian, Wood-inhabiting ligninolytic basidiomycetes in soils: ecology andconstraints for applicability in bioremediation, Fungal Ecol. 1 (2008) 4–12.

18] D.W. Gao, L.N. Du, J.L. Yang, W.M. Wu, H. Liang, A critical review of theapplication of white rot fungus to environmental pollution control, Crit. Rev.Biotechnol. 30 (2010) 70–77.

19] S.B. Pointing, Feasibility of bioremediation by white-rot fungi, Appl. Microbiol.Biotechnol. 57 (2001) 20–33.

20] C.E. Cerniglia, Fungal metabolism of polycyclic aromatic hydrocarbons: past,present and future applications in bioremediation, J. Ind. Microbiol. Biotechnol.19 (1997) 324–333.

21] E. Marco-Urrea, X. Gabarrell, M. Sarra, G. Caminal, T. Vicent, C.A. Reddy, Novelaerobic perchloroethylene degradation by the white-rot fungus Trametes ver-sicolor, Environ. Sci. Technol. 40 (2006) 7796–7802.

22] E. Federici, M. Giubilei, G. Santi, G. Zanaroli, A. Negroni, F. Fava, M. Petruccioli,A. D’Annibale, Bioaugmentation of a historically contaminated soil by poly-chlorinated biphenyls with Lentinus tigrinus, Microbial Cell Factories 11 (2012)11–35.

23] T. Fester, Arbuscular mycorrhizal fungi in a wetland constructed for benzene-,methyl tert-butyl ether- and ammonia-contaminated groundwater bioreme-diation, Microb. Biotechnol. 6 (2013) 80–84.

24] P. Blánquez, N. Casas, X. Font, X. Gabarrell, M. Sarra, G. Caminal, T. Vicent, Mech-anism of textile metal dye biotransformation by Trametes versicolor, Water Res.38 (2004) 2166–2172.

25] N. Casas, P. Blanquez, X. Gabarrell, T. Vicent, G. Caminal, M. Sarra, Degradationof orange G by laccase: fungal versus enzymatic process, Environ. Technol. 28(2007) 1103–1110.

26] M. Sukumar, A. Sivasamy, G. Swaminathan, In situ biodecolorization kinetics ofAcid Red 66 in aqueous solutions by Trametes versicolor, J. Hazard. Mater. 167(2009) 660–663.

27] N.K. Pazarlioglu, A. Akkaya, H.A. Akdogan, B. Gungor, Biodegradation of directblue 15 by free and immobilized Trametes versicolor, Water Environ. Res. 82(2010) 579–585.

28] E. Borràs, G. Caminal, M. Sarra, C. Novotny, Effect of soil bacteria on theability of polycyclic aromatic hydrocarbons (PAHs) removal by Trametes ver-sicolor and Irpex lacteus from contaminated soil, Soil Biol. Biochem. 42 (2010)2087–2093.

29] M.J. Han, H.T. Choi, H.G. Song, Degradation of phenanthrene by Trametes versi-color and its laccase, J. Microbiol. 42 (2004) 94–98.

30] A. Soares, B. Guieysse, B. Mattiasson, Influence of agitation on the removal ofnonylphenol by the white-rot fungi Trametes versicolor and Bjerkandera sp. BOL13, Biotechnol. Lett. 28 (2006) 139–143.

[

aterials 262 (2013) 554– 560

31] P. Blánquez, B. Guieysse, Continuous biodegradation of 17 beta-estradiol and17 alpha-ethynylestradiol by Trametes versicolor, J. Hazard. Mater. 150 (2008)459–462.

32] C. Cruz-Morató, C. Rodríguez-Rodríguez, E. Marco-Urrea, M. Sarrà, G. Cami-nal, T. Vicent, A. Jelic, M.J. Garcia-Galán, S. Pérez, M.S. Díaz-Cruz, M. Petrovic,D. Barceló, Biodegradation of pharmaceuticals by fungi and metabolites iden-tification, in: Emerging Organic Contaminants in Sludges: Analysis, Fate andBiological Treatment, Springer, Berlin/Heidelberg, Germany, 2012.

33] Y. Zhang, S.U. Gei�en, Elimination of carbamazepine in a non-sterile fungalbioreactor, Bioresour. Technol. 112 (2012) 221–227.

34] X. Font, G. Caminal, X. Gabarrell, S. Romero, M.T. Vicent, Black liquor detoxifi-cation by laccase of Trametes versicolor pellets, J. Chem. Technol. Biotechnol. 78(2003) 548–554.

35] J. Lewis, J. Sjostrom, Optimizing the experimental design of soil columns in sat-urated and unsaturated transport experiments, J. Contam. Hydrol. 115 (2010)1–13.

36] R.A. Freeze, J.A. Cherry, Groundwater, Prentice-Hall Inc., Englewood Cliffs, NJ,USA, 1979.

37] D. Banks, An Introduction to thermogeology ground source heating and cooling,Blackwell Publishing, Oxford, United Kingdom, 2008.

38] A.M.P. Tavares, M.A.Z. Coelho, M.A.Z. Agapito, J.A.P. Coutinho, A.M.R.B. Xavier,Optimization and modeling of laccase production by Trametes versicolor in abioreactor using statistical experimental design, Appl. Biochem. Biotechnol.134 (2006) 233–248.

39] J.D. Hem, Study and interpretation of the chemical characteristics of naturalwater, United States Geological Survey Water-Supply, 1985, pp. 263.

40] E. Borràs, P. Blanquez, M. Sarra, G. Caminal, T. Vicent, Trametes versicolor pelletsproduction: low-cost medium and scale-up, Biochem. Eng. J. 42 (2008) 61–66.

41] M.S.M. Annuar, S. Adnan, S. Vikineswary, Y. Chisti, Kinetics and energetics ofazo dye decolorization by pycnoporus sanguineus, Water Air Soil Pollut. 202(2009) 179–188.

42] N.K. Pazarlioglu, R.O. Urek, F. Ergun, Biodecolourization of Direct Blue 15by immobilized Phanerochaete chrysosporium, Process Biochem. 40 (2005)1923–1929.

43] EPA, Average shallow groundwater temperatures, 2012, http://www.epa.gov/athens/learn2model/part-two/onsite/ex/jne henrys map.html (accessed15.07.12).

44] A. Allen, D. Milenic, P. Sikora, Shallow gravel aquifers and the urban heat islandeffect: a source of low enthalpy geothermal energy, Geothermics 32 (2003)569–578.

45] P. Blánquez, M. Sarra, M.T. Vicent, Study of the cellular retention time andthe partial biomass renovation in a fungal decolourisation continuous process,Water Res. 40 (2006) 1650–1656.

46] G.L. Solomons, Production of biomass by filamentous fungi, in: ComprehensiveBiotechnology, Pergamon Press, Oxford, United Kingdom, 1985.

47] C.E. Rodríguez-Rodríguez, A. Jelic, A. Pereira, D.Z. Sousa, M. Petrovic, M.M.Alves, D. Barceló, G. Caminal, T. Vicent, Bioaugmentation of sewage sludge withTrametes versicolor in solid-phase biopiles produces degradation of pharma-ceuticals and affects microbial communities, Environ. Sci. Technol. 46 (2012)12012–12020.