Tx

Ma

b

h

•

•

•

•

•

a

ARRAA

KCbXPMW

h0

Journal of Hazardous Materials 317 (2016) 403–415

Contents lists available at ScienceDirect

Journal of Hazardous Materials

j o ur nal ho me pa ge: www.elsev ier .com/ locate / jhazmat

owards a continuous two-phase partitioning bioreactor forenobiotic removal

.Concetta Tomei a,∗, Domenica Mosca Angelucci a, Andrew J. Daugulis b

Water Research Institute, C.N.R., Via Salaria km 29.300, CP 10, 00015 Monterotondo Stazione, Rome, ItalyDepartment of Chemical Engineering, Queen’s University, Kingston, Ontario K7 L 3N6, Canada

i g h l i g h t s

A prototype of a continuous two-phase partitioning bioreactor wasinvestigated.The bioreactor contained coiled tub-ing of a selected extruded polymer,Hytrel 8206.Mass transfer and removal of axenobiotic, 4-cholorophenol, wereinvestigated.Removal efficiencies in the tub-ing wastewater stream werealways ≥ 96%.Presence of polymer tubing bufferedincreasing in organic load to thehybrid system.

g r a p h i c a l a b s t r a c t

r t i c l e i n f o

rticle history:eceived 30 March 2016eceived in revised form 8 May 2016ccepted 31 May 2016vailable online 31 May 2016

eywords:ontinuous two-phase partitioningioreactorenobiotics

a b s t r a c t

The removal of a xenobiotic (4-chlorophenol) from contaminated water was investigated in a simu-lated continuous two-phase partitioning bioreactor (C-TPPB), fitted with coiled tubing comprised of aspecifically-selected extruded polymer, Hytrel 8206. Wastewater flowed inside the tubing, the pollutantdiffused through the tubing wall, and was removed in the aqueous bioreactor phase at typical biologicalremoval rates in the C-TTPB simulated by varying aqueous phase throughput to the reactor. Operatingover a range of influent substrate concentrations (500–1500 mg L−1) and hydraulic retention times inthe tubing (4–8 h), overall mass transfer coefficients were 1.7–3.5 × 10−7 m s−1, with the highest valuecorresponding to the highest tubing flow rate. Corresponding mass transfer rates are of the same order asbiological removal rates, and thus do not limit the removal process. The C-TPPB showed good performance

olymer tubingass transferastewater treatment

over all organic and hydraulic loading ranges, with removal efficiencies of 4CP in the tubing wastewa-ter stream always ≥96%. Additionally, the presence of the Hytrel tubing was able to buffer increases inorganic loading to the hybrid system, enhancing overall process stability. Biological testing of the C-TPPB

t resu

confirmed the abiotic tes in the tubing stream.

∗ Corresponding author.E-mail address: tomei@irsa.cnr.it (M.Concetta Tomei).

ttp://dx.doi.org/10.1016/j.jhazmat.2016.05.092304-3894/© 2016 Elsevier B.V. All rights reserved.

lts demonstrating even higher 4-chlorophenol removal efficiency (∼99%)

© 2016 Elsevier B.V. All rights reserved.

1. Introduction

Solid-liquid two-phase partitioning bioreactors (TPPBs) havebeen demonstrated to be a powerful platform for xenobiotic

404 M.Concetta Tomei et al. / Journal of Hazard

Nomenclature

ATi Tubing internal area (m2)ATe Tubing external area (m2)CER Substrate concentration in the reactor effluent

(mg L−1)CET Substrate concentration in the tubing effluent

(mg L−1)C0 Substrate concentration in the tubing influent

(mg L−1)CP Substrate concentration within the polymer tubing

(mg L−1)CR Substrate concentration in the reactor (mg L−1)CT Substrate concentration in the liquid inside the tub-

ing (mg L−1)FT Flow rate in the tubing (L h−1)FR Flow rate to the reactor (L h−1)HRT Hydraulic retention timek0 Overall mass transfer coefficient (m s−1)kR Mass transfer coefficient (reactor side) (m s−1)kT Mass transfer coefficient (tubing side) (m s−1)P Polymer-water partition coefficientri Tubing internal radius (m)ro Tubing external radius (m)VER Volume of the collected reactor effluent (L)VET Volume of the collected tubing effluent (L)VP Volume of the polymer tubing (L)VR Reactor liquid volume (L)VT Internal tubing volume (L)

Subscriptin Influent

rfssfPcahauc

iohz3p

lt(botspie

out Effluent

emoval [1,2], and to be effective in reducing substrate toxicityor many classes of pollutants found in industrial wastewater. Theelf-regulating nature of substrate uptake and metabolic-drivenubstrate release characterizing their operation has been success-ully demonstrated for hydrocarbons [3], substituted phenols [4],CBs [5] and PAHs [6]. TPPB operation in the treatment of aqueousontaminants has mainly focused on batch and fed batch operation,nd granular commercial polymers typically in the form of beadsave been employed as the solid partitioning phase. One criticalspect of TPPBs is the flexibility in their operation provided by these of a polymer, as the type, amount, and shape of the polymeran be modified and adapted to each specific case.

In terms of shape, the recently proposed possibility of extrud-ng thermoplastic polymers into tubing opens new opportunities inperating TPPB bioreactors as continuous systems. This possibilityas been exploited in a recent study on the bioproduction of ben-aldehyde, which demonstrated the advantageous use of a Hytrel078 polymer tubing in a pervaporation system for continuousroduct recovery [7].

A similar process scheme could be applied to contaminatediquid streams containing groups of diverse types of pollu-ants including organic (potentially biodegradable) and inorganici.e. heavy metals, salts) components, which would normallye incompatible for treatment using a single approach asccurs in conventional treatment plants. The selective extrac-ion/partitioning of target organic substrates, mediated by carefully

elected polymer tubing, and separate targeted treatment, couldrovide a means of handling such diverse mixtures of contam-

nants. A similar strategy has been proposed and applied inxtractive membrane bioreactors (EMBs) [8]. The EMB principle

ous Materials 317 (2016) 403–415

of operation is based on the membrane extraction and perme-ation of the inhibitory/toxic organic compounds, which are thenbiologically removed in a dedicated bioreactor, while the residualwastewater stream can be treated with different, specific pro-cesses suitable for the inorganic pollutants [8,9]. The extractiondevice utilized in EMBs is a membrane of silicon-rubber permeableto the organic species (i.e. aliphatic or aromatic compounds) butnon–permeable to water and ionic species such as metals, providedin the form of parallel modules or continuous tubing, in the biore-actor [10]. The wastewater flows inside the tubing, and the organicsdiffuse through the membrane into the bioreactor, where they arebiodegraded by highly specialized microbial cultures under con-trolled conditions. In this way, the biotreatment environment is notnegatively affected by the characteristics of the original wastewa-ter (i.e. extremes of pH, ionic strength, salts, metals), which cannegatively affect biological process performance.

The same principle of operation could be applied in a contin-uous TPPB (C-TPPB) systems in which the tubing is comprised ofa specifically-selected extruded polymer, thereby combining theadvantages of high polymer affinity for target substrates, with thecontrolled reaction environments characterized by EMBs. The useof rationally-selected polymer tubing in C-TPPBs could extend theapplication to the treatment of broad classes of pollutants, arisingfrom the “tailoring” of the polymer for specific contaminants as hasrecently been demonstrated using first principles’ thermodynamicmethods [11]. Previous studies of EMBs have mainly been focusedon the removal of VOCs [12,13], and the bioreactors were operatedwith a silicone rubber membrane, which was effective for thesehydrophobic compounds, but could be limited by the solvationcapabilities of the membrane used [14], and the limited capacityof this single type of material (silicone rubber) to sorb other typesof important organic molecules such as phenols [15].

The complexity of C-TPPB operation, involving mass transfer inseries with biological removal, requires an accurate characteriza-tion of the system to determine the mass transfer coefficients andto evaluate the response to operating range of conditions expectedduring operation. Performing this first characterization step underabiotic conditions eliminates potential interferences caused bythe presence of biological materials, often possessing case-specificcharacteristics, which could increase the uncertainty of the results.Abiotic tests have been widely reported in the scientific literature tocharacterize the performance of different bioreactor configurationsprior to undertaking biotic testing, i.e. for granule-based biore-actors for autotrophic nitrogen removal [16], down flow jet loopbioreactors [17], photobioreactors for microalgae cultures [18], airlift bioreactors [19], biotrickling filters packed with polyurethanefoam for off gas treatment [20]. Abiotic characterization studieshave been also performed prior to biotic operation for differenttypes of membrane bioreactors i.e. submerged membrane biore-actors for hydrogen bioproduction [21], hollow fibers membranebioreactors for syngas fermentation to biofuel [22], and EMBs[23,24].

The aims of this study are to operate a prototype C-TPPB toinvestigate the partitioning of a target xenobiotic compound, 4-chlorophenol, by operating a bioreactor containing coiled Hytrel8206 tubing over a range of influent flowrates and substrateconcentrations within the tubing. 4CP was selected to be repre-sentative of chlorophenols, contaminants of serious environmentalconcern because of their widespread occurrence throughout theenvironment [25]. 4CP is often chosen as a model non-growthsubstrate combined with cometabolites such as phenol or readilydegradable substrates [26].

The first part of the study focused on tubing characterization interms of substrate-polymer partition coefficient and mass trans-fer properties, followed by removal tests performed with varyingflowrates through the external bioreactor to determine whether

Hazardous Materials 317 (2016) 403–415 405

ttptc

2

2

1p

2

autL1t

acp(

2

paitt

rfitt

2

tetsvf

olc

aatti

M.Concetta Tomei et al. / Journal of

he mass transfer properties of the tubing are potentially adequateo match simulated biodegradation rates. The feasibility of the pro-osed C-TPPB was confirmed with a biological test performed underhe same operating conditions as the abiotic ones with a microbialulture acclimatized to the target compound.

. Materials and methods

.1. Target compound

4-chlorophenol (4CP), (CAS number 106-48-9, molecular weight28.56 g mol−1 and purity >99%) was selected as the target com-ound and was purchased from Sigma-Aldrich (USA).

.2. Analysis

Measurements of 4CP concentrations were performed onqueous samples, after centrifugation (8 min at 13000 rpm), bysing a spectrophotometer (PerkinElmer, Lambda 25), accordingo the procedure suggested for chlorophenols by da Silva andaquipai [27]. From the UV-spectrum in the wavelength range of90–410 nm (data not shown) the peak for 4CP was at 279 nm, thushis wavelength was utilized for the analysis.

Volatile suspended solid (VSS) concentration was determinedccording to standard methods [28] as an estimate of the biomassoncentration in the bioreactor. Chloride concentration in the liquidhase of the bioreactor was measured with an ion chromatographDX-100 DIONEX) according to standard methods [28].

.3. Polymer tubing

Commercial polymer Hytrel 8206 was selected based on its highartition coefficient for 4CP reported for polymer beads, i.e. 303ccording to Tomei et al. [29], and on its polymer processing abil-ty for tubing extrusion. Du Pont, Canada kindly supplied extrudedubing, with internal and external diameters of 5 and 6 mm, respec-ively.

Multistep washing was carried out before use to remove anyesidual impurities from polymer fabrication, with the tubingrst washed with methanol under vigorous stirring conditions,

hen 5 subsequent washing steps with distilled water. After pre-reatment, the tubing was dried overnight in a fumehood.

.4. Reactor

Mass transfer and removal tests were undertaken in the pro-otype reactor consisting of a glass vessel (working volume of 5 L)quipped with a magnetic stirrer and a thermostat for tempera-ure control (25 ± 0.5 ◦C). The reactor contained a support for thepiral-wounded tubing (length of 3.5 m), and the tubing/workingolume ratio was 2% (v/v), i.e. of the same order of magnitude asor granular polymers previously employed in TPPBs [4,30,31].

An aqueous solution of 4CP was fed to the tubing by meansf a peristaltic pump (Watson-Marlow, Cellai, Italy), suitable for

ow flow rates (in the range of 0.01–0.1 L h−1) and the effluent wasollected in a sealed glass flask

In the abiotic tests, biological removal rates were simulated bypplying different flow rates to the external reactor, and storing

nd analyzing samples of the reactor effluent. A biological kineticest was also conducted to complete the feasibility study of the pro-otype. A schematic representation of the experimental apparatuss given in Fig. 1.

Fig. 1. Schematic representation of experimental set up. VR = reactor volume;Fin = Influent flow rate; Fout = Effluent flow rate; S = storage tank. Subscripts:R = reactor; T = tubing.

2.5. Test plan

2.5.1. Partition testsInitially, small amounts of the polymer tubing were cut into

small pieces (average size of 1–2 mm) and different amounts(0.1–0.3 g) were added to 20 mL of a 4CP solution (initial concen-tration of ∼100 mg L−1) in flasks, and kept under mixed conditions(320 rpm) for 24 h.

Liquid samples were collected after 4 and 24 h, and analyzed for4CP. More details on the experimental procedure and data analysisare reported elsewhere [32].

Loaded polymer pieces, after the partitioning test, were addedto fresh water for desorption, and the liquid phase was monitoredfor 4CP with the same procedure described above.

2.5.2. Mass transfer testsIn the first test MT1, performed under static conditions, the

tubing (length = 3.5 m) was initially filled with a solution of 4CP(∼500 mg L−1) and the ends closed; then the reactor was filled withfresh water to completely immerse the tubing, and maintainedunder mixing conditions (320 rpm) for the next 2 days. The liquidconcentrations inside the tubing and in the reactor were regularlymonitored versus time. Only a few samples of the liquid inside thetubing were analyzed, in order to ensure negligible volume changein comparison to the internal tubing volume.

For the second test, MT2, the tubing was initially filled with asolution of 4CP (∼500 mg L−1), then it was continuously fed with thesame fresh 4CP solution at a flow rate of 0.013 L h−1 (correspondingto a HRT of ∼6 h). The 4CP concentration was regularly monitoredboth in the reactor and in the tubing, and the effluent from thetubing was collected and analyzed for the mass balance.

2.5.3. Abiotic removal testsStarting with the tubing filled with the feed solution (range

of nominal values 500–1500 mg L−1), a hypothetical biologicalremoval process was simulated by pumping a water stream to/from

the bioreactor at different flow rates to mimic biological degrada-tion of the target compound diffusing from the liquid inside thetubing through the polymer walls into the reactor, and being con-sumed (i.e. in this case removed from the bioreactor by aqueous

406 M.Concetta Tomei et al. / Journal of Hazard

Table 1Removal tests: overview of the operating conditions. FT = Flow rate flowing in thetubing; C0 = Influent tubing substrate concentration; FR = Flow rate flowing to thereactor, HRT = Hydraulic Retention Time within the tubing. Bold values indicate theoperating variable modified in the theree series of abiotic tests (A, B, C).

Run FT HRT (tubing) C0 FR

L h−1 h mg L−1 L h−1

A.1 0.013 6 523.8 0.7A.2 0.013 6 509.3 2.8A.3 0.013 6 510.3 6.2B.4 0.013 6 509.3 2.8B.5 0.013 6 1041.9 2.8B.6 0.013 6 1562.4 2.8C.7 0.0095 8 543.6 2.8

flfdrt

(f(scDec

2

aabwtoslottpTtf

2

i

wt

v

V

wt

C.8 0.013 6 509.3 2.8C.9 0.0203 4 507.6 2.8BIO 0.013 6 531.8 −

ow). Samples of the reactor and the tubing effluent were analyzedor 4CP concentration at time intervals of 15–30 min over the 6 huration of the experiments. Effluent streams from the tubing andeactor were analyzed for 4CP periodically and at the end of theest, and the data were utilized for mass balance calculations.

Removal tests were performed under fully mixed conditions320 rpm) at different reactor flow rates (FR) (runs A.1–3), dif-erent 4CP load (run B.4–6), and different tubing flow rates (FT)runs C.7–9), according to the test plan reported in Table 1. Theelected HRTs are in the range of values employed in biological pro-esses for wastewater treatment, which are of the order of hours.esignations A.2, B.4 and C.8 (in italics in Table 1) indicate the samexperiment, which is reported more times because its operatingonditions are of relevance in the three series of tests.

.5.4. Biotic removal testThe tubing was filled with a 4CP solution of ∼500 mg L−1, then

continuous flow of 0.013 L h−1 at the same concentration waspplied for 96 h. To ensure completely mixed conditions in theioreactor agitation in the presence of biomass, the agitation speedas increased to ∼ 400 rpm. An inoculum previously acclimated

o 4-chlorophenol [31] was added to the bioreactor, which wasperated at a biomass concentration of ∼2 gVSS L−1. A concentratedolution of a nutrient medium [33] was added daily to the externaliquid phase of the bioreactor to ensure the optimal C:N: P ratiof 100:5:1. Two replicates of the test were conducted according tohe operating conditions reported in Table 1 (run BIO), and withhe same procedure as the abiotic removal tests by withdrawingeriodic samples from the tubing, bioreactor and effluent tank.he biomass concentration was monitored daily, and, accordingo Caldeira et al. [34], biological removal of 4CP was evaluated byollowing the chloride concentration evolution in the bioreactor.

.6. Mass balance

Mass balances allow quantification of the 4CP distribution dur-ng the test. At any time t, the compound is distributed in:

- the liquid phase in the reactor

VRCR(1)

here VR is the reactor volume and CR is the 4CP concentration inhe reactor at time t

- the liquid inside the tubing (evaluated assuming the averagealue of in and out concentrations at time t)

TCT (2)

here VT is the internal tubing volume and CT is the 4CP concen-ration at time t

ous Materials 317 (2016) 403–415

- the effluent from the tubing, which was collected during thetest from t = 0 and t = t

VETCET (3)

where VET is the stored effluent volume and CET the related 4CPconcentration at time t

- the effluent from the reactor, which was collected during thetest from t = 0 and t = t

VERCER (4)

where VER is the stored effluent volume and CER is the related 4CPconcentration at time t

The difference between the fed amount (including the initialmass in the tubing):

FTC0t + VTC0 (5)

and the amounts 1, 2, 3, 4 (all measured) gives the amountabsorbed into the polymer tubing �A in the time interval 0-t.

For the biological test, term (4) was replaced by the amount of4CP biologically degraded, estimated through the chloride balancein the liquid phase in the time interval 0-t.

2.7. Evaluation of mass transfer coefficients

2.7.1. Batch testsData from the batch test MT1 were correlated with a system

of differential equations describing the mass transfer among thethree phases in the system: i.e. the liquid phase in the tubing, thepolymer tubing, and the liquid phase in the reactor.

The mass transfer between the liquid phases in the tubing andin the reactor is modeled as a two-step in-series process describedas reported in the following.

Mass transfer tubing liquid phase–polymer:

dCT

dt= kT aTP

(Cp

P-CT

)(6)

Mass transfer polymer- reactor liquid phase:.

dCR

dt= kR aPR

(CP

P-CR

)(7)

Mass balance in the polymer:

dCP

dt= RTP

dCT

dt+RRP

dCR

dt(8)

where P is the 4CP partition coefficient polymer tubing − water,aTP = ATi/VT, the specific internal tubing surface area, aPR = ATe/VR,the specific tubing external area, RTP = VT/VP and RRP = VR/VP, andkT and kR are the mass transfer coefficients. Boundary conditionsfor integration are:

at t = 0 CR = 0 CP = 0 CT = C0Data fitting was performed by Scientist 3.0 for Windows (Micro-

math). The overall mass transfer coefficient kO (related to theinternal tubing area) from the tubing liquid phase to the reactorliquid phase can be estimated by:

1kOaTP

= 1kTaTP

+1

kRaPR(9)

2.7.2. Plug flow testsIn continuous fed tests, i.e. MT2 and all removal tests, the mass

transfer coefficient was evaluated assuming that the liquid flow-ing inside the tubing was in plug flow, and the bulk liquid in thereactor is perfectly mixed. According to Freitas Dos Santos and Liv-ingston [9], who operated with an extractive membrane bioreactor

M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415 407

Fig. 2. Partition coefficient data. (S, s = Sorption test; D, d = Desorption test; C0 = initial concentration at t = 0; C = concentration at time t; Mpol = polymer mass; Vw = watervolume; � = polymer density).

F profil(

tf

k

3

3

tp

ig. 3. Mass transfer tests MT1: experimental and calculated (c) 4CP concentrationp). R1 and R2 indicate the two replicates.

o detoxify wastewater containing VOCs, the resulting expressionor kO is:

O =FTln

(CT,in-CR

CT,out-CR

)

2�riL(10)

. Results and discussion

.1. Tubing characterization: partition coefficient

Results are shown in Fig. 2 and reported as the linearized form ofhe mass balance referred to 4 and 24 h. Details on the linearizationrocedure are reported elsewhere [32], and partition coefficients

es in the liquid phase in the tubing (t), in the reactor (r) and in the polymer tubing

were determined both in sorption and desorption tests, with highcorrelation coefficients (R2 > 0.992). It is interesting to observe thatequilibrium was almost complete after 4 h, and only a small vari-ation of P (from 328 to 349 in sorption tests and from 381 to 313in desorption tests) is detected after 24 h. The comparison with thegranular Hytrel 8206 shows very close P values (i.e. 303 vs 349 and313) for beads and tubing pieces, respectively.

3.2. Tubing characterization: mass transfer properties

Mass transfer tests were carried out in two operation modeswith respect to the state of liquid in the tubing, with the first (MT1)in static conditions i.e. there is no liquid flowing inside the tubingand the second (MT2) in dynamic conditions with a liquid flow

408 M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415

he liq

ifiirt

mtbttl

oi(kg

t

Fig. 4. Mass transfer tests MT2: 4CP concentration profiles in t

nside the tubing. The second can be considered as a reference testor the following substrate removal tests, where, in addition to thenternal tubing flow, an external liquid flow on the bioreactor sides applied. The results, reported in Fig. 3 and Fig. 4 for MT1 and MT2,espectively, show the change vs time of the 4CP concentration inhe liquid phase inside the tubing and in the reactor.

In Fig. 3 the concentration in the polymer, evaluated from theass balance, is also reported. The data show that, after 30 h, less

han 1% of the initial amount of 4CP remains inside the tubing foroth replicates. The residual 4CP amount is distributed betweenhe liquid phase in the reactor (53%), and the amount absorbed inhe polymer (46%), with only a very low fraction (∼1%) still in theiquid inside the tubing.

The MT1 data have been correlated with the differential systemf Eqs. (6)–(8), and the simulated concentration profiles, reported

n Fig. 3, show a very good prediction of the experimental dataR2 ≥ 0.98 for all the concentration profiles). The best fit data forT and kR are 2.6 × 10−7 and 1.7 × 10−5 m s−1, respectively, which

ives a k0 equal to 1.4 × 10−7 m s−1.

Analogous results have been obtained for the MT2 test, andhe evaluated mass transfer coefficient is 2.341 × 10−7 ms−1.

uid phase in the tubing and in the reactor; (b) is a zoom of (a).

The kO values for MT1 and MT2 are in the range of values(0.5–14.7 × 10−7 m s−1) reported for the selective extraction of phe-nolic compounds in tubular silicon rubber membranes immersedin an extracting solution [35], and therefore suitable for applicationfor extractive bioreactors. In Fig. 4 the zoom of the bottom part ofFig. 4a, reported in Fig. 4b shows that ∼24 h are sufficient to reachequal concentrations inside and outside the tubing. Previous stud-ies on biological removal of substituted phenols in conventionalsingle-phase and two-phase partitioning bioreactors demonstratedthat this time is compatible (i.e. not limiting) with the biologicalreaction characteristic times [36].

3.3. Removal tests

Operating conditions for the removal tests were determined byspecifying the external flow rate entering and leaving the biore-actor in order to achieve substrate removal rates in the range of

values from 0.7 to 8 mg gVSS

−1h−1 [31,37] obtained in single andtwo-phase fed batch bioreactors degrading 4CP with a biomassconcentration of 1.5–3 g L−1 This criterion allowed simulation ofan actual biological system.

M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415 409

centr

robFt∼wFiCtfl

Fig. 5. Variation of the external flow rate and biological removal test: 4CP con

In the first group of removal tests (A.1-3) the applied removalate was varied by modifying the external flow rate in the rangef 0.7–6.2 L h−1, i.e. over about one order of magnitude. Fig. 5(a-) shows the concentrations in the effluent from the tubing whileig. 5c displays the concentrations in the reactor. These concen-ration profiles suggest that steady state operation is reached after5 h. In the three tests no appreciable difference in concentrationas observed in the effluent from the tubing (see the zoom in

ig. 5b), thus, for the applied load, even the lowest removal rates sufficient to reach removal efficiencies in the range of 95–96%.

oncentrations in the reactor are always very low ≤3.5 mg L−1, andhe best performance is found, as expected, for the highest externalow rate i.e. the highest removal rate.

ation in the tubing effluent (a-b) and in the reactor (c) in A.1–3 and BIO tests.

In the second series of tests (B.1-3), the organic load was variedby modifying the influent concentration to the tubing from 500 to1500 mg L−1 (nominal values), with the results shown in Fig. 6. Alsoin this case, 5 h are sufficient to reach steady state conditions. The4CP concentration in the tubing effluent increases with the higherload from 20.7 to 47.3 mg L−1, but the related removal efficienciesare still very high in the range of 96–97%. Low concentrations inthe reactor (≤5 mg L−1) are observed even for the highest substrateload fed through the tubing. The system was able to respond toa three-fold increase in substrate load without a decrease in the

removal efficiency. This finding confirms previous results obtainedin the case of TPPBs using polymer beads to handle substrate surges[38,39].

410 M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415

the ef

ttaFa1ccrt

t

Fig. 6. Variation of the feed concentration: 4CP concentration in

In the third series of tests (C.7–9) the internal flow rate in theubing was varied from 0.0095 to 0.023 L h−1, which is equivalento modifying the applied organic load from 5.2 to 10.3 mg4CP h−1

nd to decrease the HRT from 8 to 4 h, with the results reported inig. 7. The time to reach steady state operation was always ∼5 h,nd very stable concentrations in the reactor were achieved after.5 h with values in the range of 1.9–3.5 mg L−1. The effluent tubingoncentrations varied from 17 to 21 mg L−1 with removal efficien-ies ≥96%. The data do not show any influence of the explored HRT

ange on the reactor performance and the system is able to managehe two-fold increase in hydraulic load.

Mass transfer coefficients were evaluated for all the removalests by applying Eq. (8) and the results are summarized in Fig. 8a.

fluent from the tubing (a-b) and in the reactor (c) in B.4–6 tests.

The kO values are in the range of 1.7–3.5·10−7 m s−1, and minimalvariations are observed for tests A and B, while in the third seriesof tests (C), as expected, an increase is observed with the increasedflow rate in tubing.

The mass transfer coefficients during the removal tests are com-parable to the values reported in Han et al. [35] for similar systemsoperated with silicon-rubber membranes. In the same Fig. 8(b),the calculated mass transfer rates for each test, and the biolog-ical removal rate evaluated at the steady state conditions from

−1

the biotic removal test (7.6 mg4CP h ), are reported. It is observedthat the mass transfer rates are always higher than the biologicalremoval rate, thus confirming that there is no limitation due to themass transfer resistance through the tubing walls relative to biolog-

M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415 411

the effl

i[pw

fottrep

Fig. 7. Variation of the internal flow rate: 4CP concentration in

cal reactions. Similar results have been reported in Pittman et al.40] in a study investigating the relative rates of mass transfer andhenol biodegradation in a conventional TPPB bioreactor operatingith Hytrel 8206 polymer beads.

To complete the data analysis, a mass balance has been per-ormed to evaluate the polymer tubing response under the testedperating conditions, and the results are reported in Table 2 inerms of 4CP distribution (amount and percent) at the end of each

est. For an easier visualization, the percent distribution is alsoeported in graphical form in Fig. 9a, which also shows the removalfficiencies in all tests. The absorbed amounts of substrate into theolymer highlight the high adaptability of the polymer tubing in

uent from the tubing (a-b) and in the reactor (c) in C.7–9 tests.

response to the different operating conditions. In tests A character-ized by increasing removal rate, we observe a proportional decreaseof the absorbed amount, while in tests B the absorbed amount pro-portionally increases by about one order of magnitude in responseto the increased organic load. In tests C, with an increased hydraulicload, we observe an increase in the amount of 4CP absorbed intothe polymer tubing followed by a decrease in C.9: this is consis-tent with the trend of the mass transfer coefficient, which in C.9

has the highest value. Higher mass transfer results in better sub-strate removal (see Table 2) with a consequent reduction of theamount of substrate absorbed by the polymer tubing. The achieved

412 M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415

Fig. 8. Mass transfer coefficient for the different tests (a) and mass transfer rates (b). The dashed line in (b) indicates the average biological removal rate of the BIO test.

Table 2Removal tests: 4CP distribution at the end of the tests (6 h). Bold values indicate the amounts in the mass balance.

Run Unit Absorbed intothe polymer

In the liquid phasewithin the tubing

In the effluentleaving the tubing

In the aqueous phasewithin the reactor

In the effluentleaving the reactor

MT2 (mg) 17.84 20.59 5.23 36.48 0.00(%) 22.3 25.7 6.5 45.5 0.0

A.1 (mg) 26.74 21.37 5.99 17.44 8.72(%) 33.3 26.6 7.5 21.7 10.9

A.2 (mg) 11.03 20.67 4.91 11.41 30.40(%) 14.1 26.4 6.3 14.6 38.8

A.3 (mg) 5.60 20.81 5.77 4.66 41.25(%) 7.2 26.7 7.4 6.0 52.8

B.4 (mg) 11.03 20.67 4.91 11.41 30.40(%) 14.1 26.4 6.3 14.6 38.8

B.5 (mg) 61.57 41.69 8.31 15.54 35.43(%) 37.9 25.6 5.1 9.6 21.8

B.6 (mg) 92.88 62.78 11.29 25.55 52.08(%) 38.0 25.7 4.6 10.4 21.3

C.7 (mg) 3.95 21.98 3.26 9.91 35.37(%) 5.3 29.5 4.4 13.3 47.5

C.8 (mg) 11.03 20.67 4.91 11.41 30.40

r(

o

(%) 14.1 26.4

C.9 (mg) 6.47 20.45

(%) 6.4 20.2

emoval efficiencies, referred to the tubing influent and effluenti.e. the treated stream), are in the range of 95–97% for all tests.

A biological run was also performed to verify that the resultsf the abiotic characterization are applicable to biological sys-

6.3 14.6 38.817.92 6.56 50.0517.7 6.5 49.3

tems, thus completing this first feasibility study of the C-TPPB.The employed culture was a microbial consortium acclimatized to4CP and the biotic system was operated for 96 h (16 HRTs) i.e. atime long enough to ensure steady state conditions. An overview

M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415 413

Fig. 9. 4CP distribution in removal test: at the end of the tests, after 6 h, for abiotic tests (a) and at different times during the BIO test (b). 4CP concentration profiles in thetubing and in the bioreactor and removal rates for the two replicates: bars indicate standard deviation between replicates (c).

414 M.Concetta Tomei et al. / Journal of Hazardous Materials 317 (2016) 403–415

Table 3Biological removal tests: 4CP distribution at different time during the test. Bold values indicate the amounts in the mass balance.

Time Unit Absorbed intothe polymer

In the liquid phasewithin the tubing

In the effluentleaving the tubing

In the aqueous phasewithin the bioreactor

Biologically removed

6 h (mg) 4.31 21.30 3.76 4.67 51.20(%) 5.1 24.9 4.4 5.5 60.1

24 h (mg) 6.64 20.86 4.57 6.20 175.65(%) 3.1 9.8 2.1 2.9 82.1

48 h (mg) 9.64 20.06 5.15 6.47 333.48(%) 2.6 5.4 1.4 1.7 88.9

72 h (mg) 27.15 20.10 5.65 6.93 473.531.16.50.9

ofirbicat(ttatcda8otwibcIrali

ta

4

bfatietdatctpiba

[

[

[

[

[

[

[

(%) 5.1 3.8

96 h (mg) 38.89 20.14

(%) 5.6 2.9

f the results is reported in Fig. 9c showing the concentration pro-les in the tubing effluent and in the bioreactor, and the calculatedemoval rates vs. time. Concentration profiles in the first 6 h of theiological removal test have been also reported in Fig. 5 for compar-

son with the A1-3 test series, performed under the same operatingonditions. Fig. 5a and b shows that tubing concentrations of 4CPre even lower than those measured in the abiotic tests, while inhe liquid phase of the bioreactor the 4CP concentration after 6 h∼ 1 mg L−1), is in the lower range of values observed for the abioticests (1–3 mg L−1). The mass balance at different times of the BIOest, allowed determination of the 4CP distribution in the system,nd the possibility of evaluating the biological removed amount,hrough the chloride measurement gives a precise estimate of theontribution of biodegradation to the observed 4CP removal. Theata reported in Table 3, show a biological removal efficiency > 80%fter 24 h, which increased in the following 90 h, reaching a value of9%. It is worth noting that stable performance of the bioreactor wasbserved during the experiment: the removal efficiencies referredo the tubing effluent, reported in Fig. 9b, are >99% after 24 h andere maintained for the entire experiment. It is also worth not-

ng that the removal efficiency can be improved by operating theioreactor with a recycle in the tubing or by increasing the biomassoncentration (cell recycle) but this may not always be advisable.n fact, too low concentrations in the bioreactors can give biologicalemoval rates so low as to be incompatible with acceptable biore-ctor volumes, and in common practice, when necessary due toow legal limits, the biological treatment of industrial wastewaters completed with a tertiary treatment step.

The very positive results obtained in the biotic test confirmedhe data of the simulated kinetics in the abiotic tests, and completed

first validation step.

. Conclusions

This study is a first demonstration, performed with abiotic andiotic tests, that a tubing-based C-TPPB can be successfully appliedor the partitioning and removal of inhibitory/toxic compounds,chieving residual 4-chlorophenol concentrations in the reactor inhe range of 1–5 mg L−1, even for the highest organic and hydraulicnfluent loads. This finding is explained in terms of the “buffer”ffect of the polymer tubing, which allowed adaptation of the sys-em in response to the increased organic load, as was previouslyemonstrated in the case of TPPBs using polymer beads. The mech-nism of uptake-release characterizing the polymer in responseo variations of the influent, already demonstrated effective in thease of polymer beads, is confirmed for the tubing. This feature andhe high affinity for the substrate demonstrated by the selected

olymer suggest applications to numerous other organic contam-

nants. Current work involves an extended characterization of theioreactor operated with enhanced microbial cultures specificallycclimatized to the target compounds to demonstrate integrated

[

1.3 88.70 8.17 617.07

1.2 89.4

transport/biodegradation over a wide spectrum of operating con-ditions, and the use of complex contaminant mixtures consisting oforganics and heavy metals to confirm process performance undersuch challenging conditions.

Acknowledgements

The authors gratefully acknowledge the contribution by DuPontCanada in providing the extruded Hytrel tubing.

References

[1] A.J. Daugulis, Two-phase partitioning bioreactors: a new technology platformfor destroying xenobiotics, Trends Biotechnol. 19 (2001) 457–462.

[2] G. Quijano, M. Hernandez, R. Munoz, S. Villaverde, Two-phase partitioningbioreactors in environmental biotechnology, Appl. Microbiol. Biotechnol. 84(2009) 829–846.

[3] G.P. Prpich, L. Rehmann, A.J. Daugulis, On the use and re-use, of polymers forthe treatment of hydrocarbon contaminatedwater via a solid-liquidpartitioning bioreactor, Biotechnol. Prog. 24 (2008) 839–844.

[4] M.C. Tomei, S. Rita, D. Mosca Angelucci, M.C. Annesini, A.J. Daugulis,Treatment of substituted phenol mixtures in single phase and two-phasesolid-liquid partitioning bioreactors, J. Hazard. Mater. 191 (2011) 190–195.

[5] L. Rehmann, A.J. Daugulis, Biodegradation of PCBs in two-phase partitioningbioreactors following solid extraction from soil, Biotechnol. Bioeng. 99 (2008)1273–1280.

[6] L. Rehmann, G.P. Prpich, A.J. Daugulis, Remediation of PAH contaminatedsoils: application of a solideliquid two phase partitioning bioreactor,Chemosphere 73 (2008) 798–804.

[7] T. Craig, A.J. Daugulis, Strategies for improved bioproduction of benzaldehydeby pichia pastoris and the use of hytrel as tubing material for integratedproduct removal by in situ pervaporation, Biochem. Eng. J. 82 (2014) 97–104.

[8] A.G. Livingston, J.P. Arcangeli, A.T. Boam, S. Zhang, M. Marangon, L.M. Freitasdos Santos, Extractive membrane bioreactors for detoxification of chemicalindustry wastes: process development, J. Membr. Sci. 151 (1998) 29–44.

[9] L.M. Freitas Dos Santos, A.G. Livingston, Novel membrane bioreactor fordetoxification of VOC wastewaters: biodegradation of 1,2-dichloroethane,Water Res. 29 (1995) 179–194.

10] A.G. Livingston, L.M. Freitas dos Santos, P. Pavasant, E.N. Pistikopoulos, L.F.Strachan, Detoxification of industrial wastewaters in an extractive membranebioreactor, Water Sci. Technol. 33 (1996) 1–8.

11] E.E. Poleo, A.J. Daugulis, Comparison of three first principles methods forpredicting solute-Polymer affinity, and the simultaneous biodegradation ofphenol and butyl acetate in a two-Phase partitioning bioreactor, J.Chem.Technol. Biotechnol. 89 (2014) 88–96.

12] C. Nicolella, I.C. Garcia Appendini, S.F. Zhang, A.G. Livingston, Control ofmembrane-attached biofilms in extractive membrane bioreactors, Water Sci.Technol 41 (2000) 227–234.

13] S.N. Mehdizadeh, M.R. Mehrnia, K. Abdi, M.H. Sarrafzadeh, Biologicaltreatment of toluene contaminated wastewater by Alcaligenese faecalis in anextractive membrane bioreactor; experiments and modeling, Water Sci. Tech.64 (2011) 1239–1246.

14] J.A. Howell, Future of membranes and membrane reactors in greentechnologies and for water reuse, Desalination 162 (2004) 1–11.

15] H. Fam, A.J. Daugulis, Substrate mass transport in two-Phase partitioningbioreactors employing liquid and solid non-Aqueous phases, Bioproc.Biosystem. Eng. 35 (2012) 1367–1374.

16] A.K. Vangsgaard, M. Mauricio-Iglesias, K.V. Gernaey, B.F. Smets, G. Sin,Sensitivity analysis of autotrophic N removal by a granule based bioreactor:

influence of mass transfer versus microbial kinetics, Bioresour. Technol. 123(2012) 230–241.

17] A.M. Jamshidi, M. Sohrabi, F. Vahabzadeh, B. Bonakdarpour, Hydrodynamicand mass transfer characterization of a down flow jet loop bioreactor,Biochem. Eng. J. 8 (2001) 241–250.

Hazard

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[solid–liquid two phase partitioning bioreactor (TPPB): kinetics of absorption,desorption and biodegradation, New Biotechnol. 30 (2012) 2037–2042.

M.Concetta Tomei et al. / Journal of

18] A. Pirouzi, M. Nosrati, S.A. Shojaosadati, S. Shakhesi, Improvement of mixingtime mass transfer, and power consumption in an external loop airliftphotobioreactor for microalgae cultures, Biochem. Eng. J. 87 (2014) 25–32.

19] Y. Chisti, U.J. Jauregui-Haza, Oxygen transfer and mixing in mechanicallyagitated airlift bioreactors, Biochem. Eng. J. 10 (2002) 143–153.

20] J.M. Estrada, A. Dudek, R. Munoz, G. Quijano, Fundamental study ongas–liquid mass transfer in a biotrickling filter packed with polyurethanefoam, J. Chem. Technol. Biotechnol. 89 (2014) 1419–1424.

21] Z. Trad, C. Vial, J.P. Fontaine, C. Larroche, Modeling of hydrodynamics andmixing in a submerged membrane bioreactor, Chem. Eng. J. 282 (2015) 77–90.

22] P.C. Munasinghe, S.K. Khanal, Syngas fermentation to biofuel: evaluation ofcarbon monoxide mass transfer and analytical modeling using a compositehollow fiber (CHF) membrane bioreactor, Bioresour. Technol. 122 (2012)130–136.

23] P.R. Brookes, A.G. Livingston, Aqueous-aqueous extraction of organicpollutants through tubular silicone rubber membranes, J. Membr. Sci. 104(1995) 119–137.

24] U. Cocchini, C. Nicolella, A.G. Livingston, Braided silicone rubber membranesfor organic extraction from aqueous solutions II Application to containedliquid membranes, J. Membr. Sci. 199 (2002) 101–115.

25] A.O. Olaniran, E.O. Igbinosa, Chlorophenols and other related derivatives ofenvironmental concern: properties: distribution and microbial degradationprocesses, Chemosphere 83 (2011) 1297–1306.

26] E. Sahinkaya, F.B. Dilek, Effect of biogenic substrate concentration on4-chlorophenol degradation kinetics in sequencing batch reactors withinstantaneous feed, J. Hazard. Mater. 137 (2006) 282–287.

27] J.C.G.E. da Silva, M.C.P.O.M. Laquipai, Method for rapid screening ofchlorophenols using a reduced calibration set of UV spectra and multivariatecalibration techniques, Anal. Lett. 31 (1998) 2549–2563.

28] APHA (American Public Health Association), Standards Methods for theExamination of Water and Wastewater, 20th ed., APHA, Washington, DC,

1998.

29] M.C. Tomei, D. Mosca Angelucci, N. Ademollo, A.J. Daugulis, Rapid andeffective decontamination of chlorophenol-contaminated soils by sorptiononto commercial polymers and process modelling, J. Environ. Manag. 150(2015) 81–91.

[

ous Materials 317 (2016) 403–415 415

30] L. Rehman, B. Sun, A.J. Daugulis, Polymer selection for biphenyl degradation ina solid-liquid two-phase partitioning bioreactor, Biotechnol. Prog. 23 (2007)814–819.

31] D. Mosca Angelucci, M.C. Tomei, Regeneration strategies of polymersemployed in ex-situ remediation of contaminated soil: bioregenerationversus solvent extraction, J. Environ. Manag. 159 (2015) 169–177.

32] M.C. Tomei, D. Mosca Angelucci, A.J. Daugulis, The use of used automobiletyres in a partitioning bioreactor for the biodegradation of xenobioticmixtures, Environ. Technol. 35 (2014) 75–81.

33] T.M. Williams, R.F. Unz, The nutrition of Thiothrix type 021N, Beggiatoa andLeucothrix strains, Water Res. 23 (1989) 15–22.

34] M. Caldeira, S.C. Heald, M.F. Carvalho, I. Vasconcelos, A.T. Bull, P.M.L. Castro,4-Chlorophenol degradation by a bacterial consortium: development of agranular activated carbon biofilm reactor, Appl. Microbiol. Biotechnol. 52(1999) 722–729.

35] S. Han, F. Castelo Ferreira, A.J. Livingston, Membrane aromatic recoverysystem (MARS)- a new membrane process for the recovery of phenols fromwastewaters, J. Membr. Sci. 188 (2001) 219–233.

36] M.C. Tomei, M.C. Annesini, V. Piemonte, G.P. Prpich, A.J. Daugulis, Two-phasereactors applied to the removal of substituted phenols: comparison betweenliquid–liquid and liquid-solid systems, Water Sci. Technol. 62 (2010) 776–782.

37] D. Mosca Angelucci, Ex-situ Remediation of Contaminated Soils by AbsorptivePolymers and Their Bioregeneration in a Two-phase Partitioning Bioreactor,Ph.D Thesis, Sapienza University of Rome, Italy, 2015, January.

38] K.F.L. Hagesteijn, A.J. Daugulis, Passive/aggressive detoxification ofcontinuous flow biotreatment systems using absorptive polymers:partitioning bioreactors treating transient phenol loadings, Biotechnol. Lett.34 (2012) 1817–1824.

39] M.C. Tomei, M.C. Annesini, A.J. Daugulis, 2,4-Dichlorophenol removal in a

40] M.J. Pittman, M.W. Bodley, A.J. Daugulis, Mass transfer considerations insolid-Liquid two-Phase partitioning bioreactors: a polymer selection guide, J.Chem. Technol. Biotechnol. 90 (2015) 1391–1399.