Complete dechlorination of pentachlorophenol using palladized bacterialcellulose in a rotating catalyst contact reactor

Upendra D. Patel, Sumathi Suresh ∗

Centre for Environmental Science and Engineering, Indian Institute of Technology-Bombay, Mumbai 400076, India

Abstract

A rotating catalyst contact reactor (RCCR) was developed which consisted of palladized bacterial cellulose immobilized on acrylic discs forhydrodechlorination of pentachlorophenol (PCP). More than 99% of 40 mg L−1 PCP was dechlorinated to phenol in the presence of hydrogen inbatch mode at initial pH values of 5.5 and 6.5 within 2 h of reaction with stoichiometric release of free chloride. The rate of PCP dechlorinationwas found to be independent of rotational speed of discs. PCP (40 mg L−1) hydrodechlorination experiments were also conducted using RCCR incontinuous flow mode at hydraulic retention times of 1 and 2 h. The average outlet PCP concentrations revealed that liquid phase in RCCR closelyresembled that of a continuous flow complete mix reactor (CFMR). Approximately 12 and 11 L of 40 mg L−1 PCP (pH 6.5) could be treated inRCCR with 99 and 80% efficiencies in batch and continuous flow modes, respectively without any appreciable loss of the catalytic activity. Theseresults suggested reusability of palladized bacterial cellulose which in turn is expected to substantially reduce the cost of treatment process. ThusRCCR seems to have high potential for treatment of ground water contaminated with chlorinated organic compounds. Dried palladized bacterialcellulose has been used as a material for electrodes in a fuel cell. However, its application as a hydrodechlorination catalyst in a reactor operatingunder room temperature and atmospheric pressure has not been reported to the best of our knowledge. Scanning electron microscopy, energydispersive X-ray spectroscopy and X-ray diffraction analyses suggested the irreversible deposition of palladium (Pd0) particles on the bacterialcellulose fibrils.

Keywords: Hydrodechlorination; Pentachlorophenol; Palladium; Bacterial cellulose; Rotating catalyst contact reactor

1. Introduction

Pentachlorophenol (PCP) is listed as one of the priority pol-lutants and identified as a carcinogen by US EPA [1,2]. Chronicexposure to PCP is known to damage kidney, liver, blood andnervous system [3]. PCP can be found in the air, water, and soil,and enters the environment through evaporation from treatedwood surfaces, industrial spills, and disposal at uncontrolledhazardous waste sites. PCP has been identified in 9 out of 65samples of paints used in children’s toys at a level of 100–2700 mg kg−1 [4]. The widespread presence of PCP in theenvironment necessitates development of techniques for its re-moval.

Many reports on the biodegradation of PCP are available inliterature [5–8]. Biological degradation of PCP often requireslong acclimation and retention times. In addition, requirementfor specific redox conditions, for anaerobic degradation [9] andaccumulation of more toxic chlorinated intermediates [4] fol-lowing the incomplete aerobic degradation processes restrictthe field application of biological systems for remediation ofwater contaminated with PCP. Advanced oxidation processeshave been reported for the effective oxidation of PCP [10].However, such processes are expensive, non-selective and at-tack chlorinated and non-chlorinated compounds with equalpotency. Abiotic processes such as adsorption on activated car-bon or other adsorbents, air stripping, reverse osmosis (RO)are expensive and merely result in phase transfer without de-struction of chlorophenols. Besides many abiotic and biologicalprocesses, zero-valent metal and bimetallic systems have beenutilized to dechlorinate chlorinated organic compounds. Reduc-

463

tive dechlorination by zero-valent metals such as Mg0 or Fe0 isvery slow and often incomplete. On the other hand, bimetallicsystems such as Fe0/palladium or Mg0/palladium can achievenear complete dechlorination of recalcitrant pollutants such asDDT, DDD, DDE and chlorophenols [11–13]. However, ad-vantages of bimetallic system such as high rate of reaction,efficiency, mild reaction conditions, and requirement for mini-mal follow-up treatment, will be defeated if reactor design doesnot permit the reuse and recovery of expensive catalysts suchas palladium through immobilization on suitable support ma-trices. Hydrodechlorination of chlorophenols using supportedpalladium has been reported profusely in the literature [14–19].Various types of support materials such as carbon, chitin, chi-tosan, amino acids, alumina, silica, zeolites, and metal oxidessuch as TiO2, MgO, ZrO2 have been used for the immobi-lization of palladium [14,18,20]. Some of the commerciallyavailable immobilized palladium catalysts are: Pd/C and Pd-alumina. However, their major limitations are:

• Expensive, which precludes their field scale application.• Support material undergoes unwanted ion-exchange reac-

tions with HCl, which is released as the by-product ofdechlorination reaction leading to substantial loss of cat-alytic activity [17,18].

• Limited information with respect to their applicability forreactor development.

Survey of literature review revealed that a biogenic material,namely, bacterial cellulose (BC) has been used as a support ma-terial to immobilize palladium in zero-valent form (Pd0) [21].BC is made up of unbranched chains of 1,4-β-glucopyranosemolecules which are extruded in pure form by certain speciesof bacteria. Nascent chains of BC aggregate to form subfibrilsof ∼1.5 nm width and is one of the thinnest naturally occur-ring biopolymers. BC subfibrils are crystallized into microfib-rils, then into bundles, and subsequently into ribbons stabilizedby hydrogen bonds [22]. The presence of reductive terminalgroups of aldehyde hydrate [21,23] gives BC a capacity to re-duce a metal salt and impregnate the metal particles. The reticu-lated structure of BC provides high surface area, and capacity toabsorb ∼200 times of its own weight of water. Evans et al. [21]have shown that in addition to palladium, BC is capable of pre-cipitating gold and silver, respectively from hexachloroaurateand silver nitrate solutions. Although the presence of reduc-ing sugars is believed to catalyze reduction of a metal salt byBC, exact mechanism is yet to be fully understood. The con-ventional cellulose produced by plants is made up of ‘fringedmicelles’ which are held together by lignin and hemicellu-lose [22]. Chemical treatment at high temperature is essential topurify plant cellulose (PC) which involves the removal of ligninand hemicellulose bound to it. Moreover, the physical structureof PC is not as reticulated as that of BC. Thus chemical andphysical properties of PC do not allow it to reduce and incorpo-rate a metal. Other unique properties of BC are listed below.

• Can be grown to any required size and offers large surfacearea due to its sponge like structure.

• Production can be customized to suit any reactor config-uration (for example, it can be cut and used as cubes orimmobilized on a support material).

• An economical support material that can be produced froma wide variety of biodegradable organic substrates.

The above listed features of BC attracted our attention and ledus to explore its use for the reductive dehalogenation of recal-citrant pollutants such as pentachlorophenol (PCP). Evans etal. [21], demonstrated the use of palladized (metallized) bac-terial cellulose as an electrode. However its application as acatalyst for mediating reductive reactions such as hydrodechlo-rination in a reactor has not been reported to the best of ourknowledge. Thus, the specific objectives of our investigationswere: (1) to design and develop a reactor which employed pal-ladized bacterial cellulose, (2) to study the kinetics and extentof PCP dechlorination in a reactor operated in batch and con-tinuous flow modes, (3) to study the effects of pH and RPMon PCP dechlorination. Use of a palladium catalyst depositedon a biogenic material has been also reported by Mertens etal., [24]. The authors demonstrated reductive dechlorination oflindane by nano-scale Pd0 particles deposited on the surface ofShewanela oneidensis. However, the deposition of nano-scalePd0 particles and dechlorination required the supply of sodiumformate as an electron donor to the bacterial culture.

2. Materials and methods

2.1. Sources of chemicals

Potassium hexachloropalladate (K2PdCl6), potassium car-bonate, 1,1-dichloro-2,2-bis-(4-chlorophenyl) ethylene (DDE)were purchased from Sigma–Aldrich Chemical Company. Cy-clohexane, glacial acetic acid, phenol, sodium acetate, ferricnitrate, mercury(II) thiocyanate, and hydrochloric acid werepurchased from Merck India Ltd. Acetic anhydride was pur-chased from S.D. FineChem Ltd. Pentachlorophenol sodiumsalt (�90%) was purchased from Fluka. All chemicals wereof high purity and analytical grade. Potassium carbonate wasbaked overnight in furnace at 400 ◦C prior to use. No pretreat-ment was carried out with other chemicals. Deionized water ob-tained from Barnstead deionization system was used for makingreagents.

2.2. Design of rotating catalyst contact reactor (RCCR)

The major considerations for the design, development andoperation of a reactor were: (a) immobilization of bacterial cel-lulose on a suitable support material and in situ palladization,(b) provision to provide high surface area of palladized bacte-rial cellulose for dechlorination reactions and (c) requirementfor the exposure of immobilized palladium alternately to liquidphase containing the chlorinated pollutant (PCP) and gaseousphase containing molecular hydrogen [25]. Based on the abovelisted requirements we designed a RCCR that is similar to ro-tating biological contactor (RBC). It may be noted that a RBCtypically employs living microorganisms such as bacteria, fungi

464

Fig. 1. Schematic diagram of rotating catalyst contact reactor (RCCR) used for treating water containing pentachlorophenol (PCP). PdBC = palladized bacterialcellulose.

or algae, which are immobilized on the rotating discs. In ourinvestigations, the biogenic non-living matrix, bacterial cellu-lose was immobilized on circular acrylic discs. The rectangularreactor consisted of 7 circular plates of 100 mm diameter pro-viding a total surface area of 1100 cm2 (Fig. 1). The discs weremounted on a horizontal shaft that was rotated using a magneticdrive to avoid direct connection with the motor and to ensure aleak proof reaction chamber. The speed of rotation could bevaried between 1 and 50 rpm. Effective liquid volume of the re-actor was 650 mL. The space above the liquid was purged withhydrogen gas and maintained under atmospheric pressure. Ex-perimental details related to the production, immobilization ofbacterial cellulose on discs and its palladization are provided inthe following sub-sections.

2.3. Production of bacterial cellulose in laboratory andpalladization

2.3.1. Isolation of a bacterial culture for the production ofcellulose

A bacterial culture capable of producing cellulose was iso-lated in the laboratory as per the tradition followed in Philip-pines except that orange residue was used instead of pineappleresidue [26]. Three cups of orange residue (after extraction ofjuice) was mixed with 6 cups of water and 1 cup of sugar. Themixture was left undisturbed in a wide mouthed plastic utensilcovered by a thin cloth. After 12 to 16 days of incubation, soft,jelly like BC was deposited on the sides of the utensil and onthe top of liquid phase containing the orange residue. Followingadequate deposition, the cellulose film was carefully removed,sliced into small pieces and used as the source of cellulose pro-ducing bacterial inoculum.

2.3.2. Development of bacterial cellulose (BC) film anddeposition of metallic palladium on RCCR discs

The medium for bacterial cellulose production was preparedas per the following method. Coconut water (∼650 mL) wasboiled, cooled, filtered through ordinary filter papers and intro-duced into RCCR. To the above medium 65 g L−1 of sucrose,25 g L−1 of acetic acid and bacterial cellulose producing bac-terial culture (preparation as described in the previous section)were added. The top of the reactor was covered with a tight lidand the discs were rotated at 7–8 rpm. Bacterial cellulose wasallowed to deposit on the discs until film thickness of ∼2 mmwas attained. Pre-treatment and palladization of BC was con-ducted according to the protocol described by Evans et al., [21].As a first step, the growth medium in RCCR was replaced with1% NaOH solution and the discs supporting bacterial cellulosewere rotated in it for 24 h. NaOH treated discs were washedthoroughly with water and incubated in 0.5 M sodium acetatesolution (pH adjusted to 4.5 using acetic acid). Finally the discswere rinsed with water and subsequently incubated in 500 mLof 5 mM solution of K2PdCl6 prepared in 50 mM sodium ac-etate solution (pH 4.5). The entire set up was maintained at70–80 ◦C until such time (5–6 h) that all the discs turned blackin color due to the deposition of metallic palladium. The pal-ladized bacterial cellulose discs were washed thoroughly in wa-ter to wash off the residual K2PdCl6 or unbound Pd0 deposits.The reactor was maintained by continuous slow rotation of thediscs in clean water until further use. Photographs of plain bac-terial cellulose and palladized bacterial cellulose (PdBC) de-posited on acrylic discs of RCCR are shown in Figs. 2 and 3,respectively.

465

Fig. 2. Bacterial cellulose immobilized on acrylic discs.

Fig. 3. Palladization of bacterial cellulose immobilized on acrylic discs.

2.4. Dechlorination of PCP using RCCR

2.4.1. Batch experimentsPCP dechlorination experiments were conducted in batch

mode by introducing 650 mL of 40 mg L−1 PCP solutions(prepared using PCPNa) in the reactor. Hydrogen was continu-ously bubbled (∼40–50 mL/min) at a depth of ∼30 mm belowthe liquid level at one of the corners of RCCR. The reactorwas maintained at atmospheric pressure and room temperature(26–28 ◦C) for all the dechlorination experiments. Experimentswere conducted using PCP solutions prepared in plain water af-ter the adjustment of pH using acid or alkali. Initial pH valuesof 5.5, 6.5 and 9.5 were chosen while the RPM values wereadjusted to 7, 15 or 25. An experiment was also conducted atRPM of 8 using PCP solution prepared in 10 mM phosphatebuffer at pH of 6.5. Rate of PCP dechlorination was moni-tored by analyzing the aliquots of reaction medium at varioustime points using GC-ECD. In addition, aliquots were also an-alyzed for the formation of end products using GC-ECD andGC-FID. Release of free chloride ions due to dechlorinationreaction was measured by mercury(II) thiocyanate spectropho-tometric method [27]. At the end of 2 h of reaction, RCCR wasemptied of its content and refilled with water and equilibratedfor 1 h to wash the palladized bacterial cellulose deposited ondiscs. The wash water was analyzed to quantify for sorbed PCP,dechlorinated products and free chloride ions.

2.4.2. Continuous flow reactor studiesDechlorination of PCP (pH 6.5) was also studied in contin-

uous feed mode. 10 L of 40 mg L−1 PCP solution (preparedusing PCPNa) was passed through the reactor with continuoussparging of hydrogen. RPM was set at 7–8. Two flow rates;5.5 mL min−1 (HRT ∼2 h), and 11 mL min−1 (HRT ∼1 h),were used. Continuous inflow and outflow was maintained us-ing dual channel “Masterflex” (Cole-Parmar) peristaltic pump.

The effluent was pooled continuously at regular intervals oftime and analyzed for chloride, phenol, partially dechlorinatedproducts and residual PCP concentrations. At the end of batchand continuous flow experiments, palladized bacterial cellulosefrom one of the discs was scrapped off, cut to small pieces, andextracted using cyclohexane to check for the sorption of PCPand other dechlorinated products.

2.5. Gas chromatography (GC) analyses

Samples were prepared for GC analyses as per the pro-cedure described by NCASI 86.02 [28]. At designated timepoints during kinetic studies, 0.3 mL (for GC-ECD) or 3 mL(for GC-FID) aliquots were withdrawn, made up to final vol-ume of 3 mL and derivatized by adding 78 µL of 4.35 Mpotassium carbonate solution and 90 µL acetic anhydride. Sub-sequently, derivatized samples were extracted twice by mix-ing with 1.5 mL cyclohexane on a high-speed vortex mixerfor 45 s. 0.2 or 1 µL of pooled hexane extracts were in-jected for GC-ECD and GC-FID analyses, respectively. DDE(1,1,-dichloro-2,2-bis-(4-chlorophenyl) ethylene) was used asan internal standard to note the efficiency of solvent extrac-tion procedure. GC analyses were carried out using Agilent6890N model equipped with µECD and FID. HP-1 column(30 m × 0.25 mm ID × 0.25 µM film thickness) was used.Temperature programme used was: 1 min at 45 ◦C, temperatureramp 1: 45–100 ◦C @ 20 ◦C min−1, hold time 0.3 min, temper-ature ramp 2: 100–215 ◦C @ 4 ◦C min−1, final hold time 1 min.Injector and detector temperatures were set at 210 and 300 ◦C,respectively. Injections were performed in splitless mode usingnitrogen as the carrier gas (gas velocity 36 cm/s). Purge timeand flow were set at 0.5 min and 100 mL min−1, respectively.Calibration plots for chlorophenols were prepared in the con-centration range of interest and were found to be linear with R2

values >0.98.

2.6. Scanning electron microscopy (SEM), energy dispersiveX-ray spectroscopy (EDS) and X-ray diffraction (XRD)analyses

SEM and EDS analyses of dried palladized bacterial cel-lulose film were carried out using a Hitachi S3400N scan-ning electron microscope (Thermo Noran NSS200 instruments)equipped with a backscattered electron detector. An ultrathincoating of electrically conducting material, gold was depositedon the surface of PdBC for obtaining the high resolution backscattered scanning electron image. Powder XRD analysis wasperformed using a diffractometer system (model: Xpert-PROmake: Phillips, Netherland).

3. Results and discussion

3.1. Sorption of PCP onto palladized bacterial cellulose inRCCR

Removal of PCP in RCCR could occur through (a) non-reactive sorption of PCP on to palladized bacterial cellulose

466

Fig. 4. Time course profiles of PCP disappearance, emergence of chloride ionsand phenol in RCCR operated in batch mode. Solid and dotted lines representdata obtained at feed pH values of 5.5 and 6.5, respectively. Reaction condi-tions: reaction volume = 650 mL, initial PCP concentration = 40 mg L−1,initial pH values of PCP solution = 5.5 or 6.5, RPM = 8, HRT = 2 h, PCPloading = 23.6 µg cm−2.

mass or (b) sorption of PCP with concomitant reductive dechlo-rination. Control experiments conducted with RCCR (operatedin batch mode) in the absence of hydrogen revealed that asmuch as 60–70% of 40 mg L−1 of PCP was physically adsorbedonto palladized bacterial cellulose mass. This amount of sorbedPCP could be recovered through extraction in cyclohexane fol-lowing acetylation (since acetylated PCP is efficiently extractedinto hexane). In the next step we examined the contribution ofsorption towards removal of PCP following catalytic reaction.Hydrogen was sparged into RCCR containing PCP solution toinitiate reductive dechlorination reaction. Following 2 h of incu-bation period the palladized bacterial cellulose immobilized onone of the discs of RCCR was completely scrapped off and ex-tracted using cyclohexane following acetyl derivatization. GC-ECD analysis of this extract showed negligible presence of PCP.Thus it may be concluded that sorbed PCP was dechlorinatedfollowing reaction with the active species of hydrogen gener-ated on the surface of palladized bacterial cellulose.

3.2. Effect of pH on PCP dechlorination using RCCR in batchmode

Fig. 4 compares the time course profiles of PCP disappear-ance along with emergence profiles for free chloride ions andphenol in RCCR fed with influent at pH 5.5 and 6.5. More than99% of 40 mg L−1 of PCP was removed at initial pH valuesof 5.5 and 6.5 following 2 h of reaction. The first-order reac-tion rate constants were 0.056 and 0.07 min−1 at pH valuesof 5.5 and 6.5, respectively. The total (reaction phase + washwater) concentrations of phenol and free chloride were 5.6–6.0and 22–24 mg L−1, respectively. The lower yield of phenol (ascompared to the expected stoichiometric yield of 14 mg L−1

phenol from 40 mg L−1 PCP) may be explained on the basis ofhydrogenation and further degradation of phenol as reported inthe literature [15,16,29]. Roy et al., observed complete dechlo-rination of 0.5 mg PCP to cyclohexanone by Pd/alumina with ayield of 102% following 1 h of reaction while Hoke et al., re-ported dechlorination of 5 g of PCP to phenol by Pd/C with an

Fig. 5. Time course profiles of PCP disappearance in the RCCR operated at pHvalues of pH 6.5 (maintained throughout the reaction using 10 mM phosphatebuffer) and 9.5 (unbuffered). Reaction conditions: reaction volume = 650 mL,initial PCP concentration = 40 mg L−1, initial pH values of PCP solution =6.5 or 9.5, RPM = 8, HRT = 2 h, PCP loading = 23.6 µg cm−2.

overall yield of 98% in 83 h [16,19]. In the present investiga-tions almost stoichiometric release of chloride ions (expectedyield of 26 from 40 mg L−1 of PCP) was observed. pH of thereaction phase at the end of 2 h of reaction was found to be 3.2and 3.3, for PCP solutions fed to RCCR at pH values of 5.5 and6.5, respectively. Reduction of pH may be due to the generationof HCl as per the reaction shown below.

C6OHCl5 (PCP) + PdxHy (palladium hydride)→ C6H6O (phenol) + ring cleavage products (?)

+ xPd0 + 5HCl. (1)

However the catalytic activity of palladized bacterial cellulosewas not inhibited by the production of HCl. Several researchershave reported that dehydrodehalogenation reactions were inhib-ited due to the release of halides in form of mineral acid [17,18,29,30]. For example, Hoke et al., conducted hydrodechlorina-tion of chlorophenols in ethanol and 1:1 water:ethanol at 35 ◦Cand 2.45 kg cm−2 pressure using Pd/C catalyst [19]. They re-ported that the addition of a base (concentration ∼2 to 3 timeshigher with respect to acid production) such as sodium hy-droxide, sodium acetate, or ammonium hydroxide was essentialto maintain the pH of the reaction phase and that the cata-lyst (Pd/C) was rapidly deactivated in the absence of a base.It may be noted that Hoke et al. used very high concentra-tion of chlorophenol (e.g., ∼78 mM of 4-CP) and hence thecorresponding concentration of HCl released following com-plete dechlorination was very high (∼2.8 g L−1 HCl) [19].Calvo et al., studied hydrodechlorination of 100 mg L−1 of2-chlorophenol and 4-chlorophenol (stoichiometric HCl con-centration ∼28 mg L−1) using Pd/C catalyst [15]. Completedechlorination of chlorophenols to cyclohexanol was reportedby the authors without the addition of a base.

We also studied the dechlorination of PCP study under alka-line conditions since increase in pH from 5.5 to 6.5 increasedthe dechlorination rate. Time course profile of PCP disappear-ance at an initial pH of 9.5 is shown in Fig. 5. It is interesting tonote that PCP removal was incomplete and proceeded at a veryslow rate (first-order rate constant: 0.012 min−1). Moreover, itwas noted that the color of the aqueous solution turned blackish

467

due to the release of tiny fibers of palladized bacterial cellu-lose. The washing of the discs with distilled water following theexperimental run also resulted in substantial loss of palladizedbacterial cellulose in the form of fine fibers. Thus, alkaline con-ditions had detrimental effect on the performance of RCCR.

An additional experiment was also conducted using PCP so-lution made in phosphate buffer at pH 6.5 (Fig. 5). It is interest-ing to note that the rate (first order rate constant: 0.022 min−1)and extent of PCP dechlorination decreased substantially ascompared to those obtained at pH 6.5 in unbuffered reactionmedium (Fig. 4). Concentrations of phenol and chloride ionsgenerated from the dechlorination of PCP at initial pH valuesof 9.5 and 6.5 (buffered) were negligible and hence not shownin the figure. From the ongoing discussion it may be concludedthat unbuffered aqueous phase (for example, plain water) withan initial pH ranging from 5–6.5 is well suited for the completedechlorination of PCP in RCCR.

3.3. Effect of RPM on PCP dechlorination using RCCR inbatch mode

Since the dechlorination of PCP in RCCR was facilitatedby the sequential exposure of rotating palladized bacterial cel-lulose discs to aqueous phase containing target pollutant andgaseous phase containing hydrogen, we decided to study theeffect of RPM of discs on the rate of PCP dechlorination. Com-parison of first-order kinetics plots for PCP removal data ob-tained from experiments conducted at various RPM revealedthat the speed of rotation did not have significant effect onthe rate or extent of PCP removal (plots not shown). One canhypothesize that the adsorption of hydrogen with concomitantreductive dechlorination of PCP was very rapid and hence in-dependent of RPM. This feature should be advantageous forthe field scale application of this system since it can reduce thepower requirement for the operation of reactor.

3.4. PCP dechlorination using RCCR in continuous flow mode

Fig. 6 presents results related to the dechlorination of PCP(pH 6.5) using the RCCR in continuous flow mode. It was ob-served that at any time point the concentration of PCP wasmore or less uniformly distributed in the reactor. Thus onecan assume that the liquid phase of RCCR was almost com-pletely mixed. For a completely mixed continuous flow reactor(CMFR) operating under steady state conditions, the relation-ship between the concentration of PCP in the reactor (or effluentconcentration) to that of in the reactor inlet can be given by theEq. (2) as shown below [31].

(2)Ce = C0

1 + kθ,

where, Ce, and C0 are the concentrations of PCP (mg L−1) inthe reactor/effluent and inlet to the reactor, respectively. θ =hydraulic retention time, min, and k = first-order reaction rateconstant, min−1.

Substituting the value of first-order reaction rate constant(k = 0.07 min−1) obtained from batch studies and C0 =

Fig. 6. PCP dechlorination using RCCR in continuous flow mode. Reactionconditions: reaction volume = 650 mL, initial PCP concentration = 40 mg L−1,initial pH of the PCP solution = 6.5, RPM = 8, HRT = 59 min and 118 min atflow rates of 11 and 5.5 mL min−1, respectively.

Fig. 7. Back scattered scanning electron image of dried palladized bacterialcellulose film following the completion of PCP dechlorination experiment.

40 mg L−1, the outlet concentrations of PCP (Ce) were cal-culated to be 4.3 and 7.8 mg L−1 for θ values of 2 h and1 h, respectively. Fig. 6 reveals that the range of concentra-tions of PCP in the effluent (determined experimentally) closelymatched with that of the computed values for corresponding θ

values. In addition there was concomitant release of phenoland free chloride in the effluent thereby suggesting completedechlorination of PCP to the hydrocarbon end product.

3.5. SEM, EDS and XRD analyses of palladized bacterialcellulose

SEM and EDS analyses of a dried palladized bacterial cellu-lose film were done after the completion of PCP dechlorinationexperiments (Figs. 7 and 8). Back scattered scanning electronimage of the dried PdBC revealed the presence of bright spher-ical Pd0 deposits interspersed with the BC fibrils (Fig. 7). At aresolution of 20 K, the average diameter of Pd0 deposits wasestimated to be in the range of 100 to 150 nm. QuantitativeEDS analyses further indicated that the element, palladium con-stituted ∼23% of the total dry weight of PdBC (Fig. 8 and

468

Fig. 8. Energy dispersive X-ray spectrum of dried palladized bacterial cellulosefilm following the completion of PCP dechlorination experiment.

Table 2). The powder XRD analysis of PdBC also showed thepresence of distinct peaks corresponding to palladium therebyindicating the deposition of the metal in the bacterial cellulosematrix (Fig. 9). The results obtained through SEM, EDS andXRD analyses clearly suggest the irreversible impregnation ofPd0 on the BC fibrils.

3.6. Recovery and leaching of palladium from palladizedbacterial cellulose

Following the completion of PCP dechlorination experi-ments in batch and continuous flow modes, the PdBC masswas scrapped off from all the rotating discs, dried at 105 ◦Covernight followed by ignition at 650 ◦C for 4 h. The weight ofthe remaining black solid matter, palladium metal was 61 mg.Based on these results; the loading of palladium metal was esti-mated to be 0.76 mg g−1 of wet weight of BC (0.055 mg cm−2

of disc area). Initial loading of Pd0 could not be determined

Fig. 9. Plot of data obtained through XRD analysis of dried palladized bacterialcellulose film following the completion of PCP dechlorination experiment.

since it would require the sacrifice of whole PdBC mass. PCPsolutions were analyzed for the presence of soluble species ofpalladium by ICP-AES (model: Ultima 2000, make: HoribaJobin Yvon) following treatment with PdBC in the presenceof hydrogen. For the range of experimental conditions de-scribed by us, the maximum concentration of Pd in the treatedPCP solution was found to be ∼85 ppb. Considering total vol-ume of 23 L (40 mg L−1 PCP) treated in RCCR, the totalloss of Pd was estimated to be ∼2 mg. Thus, initial and fi-nal palladium loadings may not differ significantly from eachother.

RCCR was able to treat a large volume (∼23 L treated over aperiod of 2 months) of PCP solution with relatively small quan-tity of immobilized palladium. During this period the catalyticactivity of PdBC in the reactor remained consistent. Thus, ifconstructed and managed properly, RCCR may prove to be an

Table 1Comparison of the experimental conditions and results reported in the literature for the dechlorination of selected pollutants using immobilized palladium with thoseobserved in the present investigation for the dechlorination of PCP using palladized bacterial cellulose

Compoundstudied

Catalyticsystem

Reactionconditions

Reaction rates/% removal

End product(s)detected and yield

Reference

PCP Pd/C and hydrogen 35 ◦C, 2.5 atm pressure,completely anaerobicconditions, PCP in ethanolor 1:1 water:ethanol

Complete removal of 5 g ofPCP in 83 h, initial reactionrate 0.28 mmol min−1

Phenol with a yield of 98% [19]

PCP Pd-alumina and hydrogen Room temperature andpressure in 1:99methanol:water

Complete removal of0.5 mg PCP within 1 h ofreaction

Only product detected wascyclohexanone with a yieldof 102%

[16]

4 CP Pd/C and hydrogen Room temperature andpressure with triethylamineas an additive

Complete removal of0.53 mmol compound in 3 h

Phenol with a yield of 92%.Reaction was incompletewhen triethylamine wasomitted

[14]

Lindane Pd0 supported on cells ofShewanela oneidensis and680 mg L−1 of sodiumformate as the electrondonor

Atmospheric pressure,28 ◦C

Complete removal of50 mg L−1 lindanefollowing 24 h reactionperiod

Dichlorobenzene andbenzene in gaseous formwere detected as endproducts

[24]

PCP Palladized bacterialcellulose and hydrogen

Room temperature andpressure, no precautions toexclude oxygen

>99% removal of40 mg L−1 PCP in �2 h.1st order reaction rateconstant 0.07 min−1

Phenol detected with anoverall yield of 42%

Present study

469

Table 2Quantification of various elements present in palladized bacterial cellulose based on EDS analyses

Element Netcounts

ZAF Weight(%)

Weight error(%)

Atom(%)

Atom error(%)

C 885 2.683 25.50 ±0.92 42.87 ±1.55O 1802 4.568 29.22 ±1.14 36.88 ±1.43Na 1954 2.202 12.08 ±0.29 10.61 ±0.26Cl 999 1.115 4.17 ±0.28 2.38 ±0.16K 947 1.183 5.18 ±0.49 2.68 ±0.25Rb 212 1.461 1.04 ±0.22 0.25 ±0.05Pd 3356 1.288 22.81 ±1.60 4.33 ±0.30Total 100.00 100.00

economically viable option for the rapid treatment of water con-taminated with PCP. A concise comparison of the experimentalconditions and results reported in the literature for the dechlo-rination of selected pollutants such as PCP, 4-chlorophenol(4-CP) and lindane using immobilized palladium with those ob-served in the present investigation for the dechlorination of PCPusing PdBC is provided in Table 1.

4. Conclusions

RCCR seems to have high potential for remediation of largevolumes of water contaminated with PCP. RCCR is simple andflexible with respect to: (a) immobilization of bacterial cellu-lose onto rotating discs, and (b) in situ reduction and immobi-lization of palladium onto bacterial cellulose, and (c) operationin batch or continuous flow mode. Efficient dechlorination ofPCP to the hydrocarbon end products, such as phenol withinpH range of 5.5–6.5 suggested that bacterial cellulose is a suit-able support matrix for the reductive catalyst, palladium. SEM,EDS and XRD analyses suggested the irreversible depositionof Pd0 particles on the BC fibrils. Irreversible immobilizationof palladium in BC and its reusability is expected to reduce thecost of treatment process to a significant extent.

Acknowledgments

The authors wish to thank Council for Scientific and In-dustrial Research (CSIR), Government of India for providingfinancial assistance to execute this project. The authors alsowish to thank Prof. Satish Vitta (Department of MetallurgicalEngineering and Materials Science, IIT-Bombay) for his timelyhelp in performing SEM, EDS and XRD analyses.

References

[1] Proposition 65, Office of Environmental Health Hazard, EPA, State of Cal-ifornia, USA, 2005.

[2] L.H. Kieth, W.A. Telliard, Environ. Sci. Technol. 13 (1979) 416–423.[3] Agency for Toxic Substances and Disease Registry (ATSDR), Toxicolog-

ical Profile for Pentachlorophenol, 2001.[4] R.J. Irwin, M. VanMouwerik, L. Stevens, M.D. Seese, W. Basham, En-

vironmental Contaminants Encyclopedia, National Park Service, WaterResources Division, Fort Collins, CO, 1977.

[5] D.-S. Shen, X.-W. Liu, H.-J. Feng, J. Hazard. Mater. B 119 (2005) 239–243.

[6] C.M. Kao, C.T. Chai, J.K. Liu, T.Y. Yeh, K.F. Chen, S.C. Chen, WaterRes. 38 (2004) 663–672.

[7] S.K. Banerji, R.K. Bajpai, J. Hazard. Mater. B 39 (1994) 19–31.[8] M.D. Mikesell, S.A. Boyd, Appl. Environ. Microb. 52 (1986) 861–865.[9] B.-V. Chang, Y.-X. Zheng, S.-Y. Yuan, Chemosphere 33 (1996) 313–320.

[10] M. Pera-Titus, V. Garcia-Molina, M.A. Banos, J. Gimenez, S. Esplugas,Appl. Catal. B 47 (2004) 219–256.

[11] S.K. Gautam, S. Suresh, J. Hazard. Mater. B 139 (2007) 146–153.[12] S.K. Gautam, S. Suresh, Water Environ. Res. 79 (2007) 430–436.[13] U.D. Patel, S. Suresh, J. Hazard. Mater. B 147 (2007) 431–438.[14] Y. Monguchi, A. Kume, K. Hattori, T. Maegawa, H. Sajiki, Tetrahedron 62

(2006) 7926–7933.[15] L. Calvo, A.F. Mohedano, J.A. Casas, M.A. Gilarranz, J.J. Rodiguez, Car-

bon 42 (2004) 1377–1381.[16] H.M. Roy, C.M. Wai, T. Yuan, J.-K. Kim, W.D. Marshall, Appl. Catal.

A 271 (2004) 137–143.[17] G. Yuan, M.A. Keane, Catal. Today 88 (2003) 23–27.[18] F.J. Urbano, J.M. Marinas, J. Mol. Catal. A 173 (2001) 329–345.[19] J.B. Hoke, G.A. Gramiccloni, E.N. Balko, Appl. Catal. B 1 (1992) 285–

296.[20] N.V. Kramareva, A.Y. Stakheev, O.P. Tkachenko, K.V. Klementiev,

W. Grunert, E.D. Finashina, L.M. Kustov, J. Mol. Catal. A 209 (2004)97–106.

[21] B.R. Evans, H.M. O’Neill, V.P. Malyvanh, I. Lee, J. Woodward, Biosens.Bioelectron. 18 (2003) 917–923.

[22] S. Bielecki, A. Krystynowicz, M. Turkiewicz, H. Kalinowska, L. Poland,in: E.J. Vandamme, S.D. Baets, A. Steinbuchel (Eds.), Biopolymers,vol. 5, Wiley–VCH, 2004, p. 37.

[23] H.S. Barud, C. Barrios, T. Regiani, R.F.C. Marques, M. Verelst, J.D-Ghys, Y. Messaddeq, S.J.L. Ribeiro, Mater. Sci. Eng. C (2007),doi:10.1016/j.msec.2007.05.001.

[24] B. Mertens, C. Blothe, K. Windey, W.D. Windt, W. Verstraete, Chemo-sphere 66 (2007) 99–105.

[25] S. Suresh, U.D. Patel, Indian patent, application number: 691/MUM/2007,2007.

[26] P.K. Thampan, Handbook on Coconut Palm, Oxford & IBH Publishing,New Delhi, India, 1981.

[27] T.M. Florence, Y.J. Farrar, Anal. Chim. Acta 54 (1971) 373–377.[28] National Council of the Paper Industry for Air and Stream Improvement,

Inc. (NCASI), Methods Manual: Chlorinated Phenolics in Water by In-Situ Acetylation/GC/MS Determination (CP-86.02), 1997.

[29] G. Yuan, M.A. Keane, Appl. Catal. B 52 (2004) 301–314.[30] C. Menini, C. Park, E.-J. Shin, G. Tavoularis, M.A. Keane, Catal. To-

day 62 (2000) 355–366.[31] G. Tchobanoglous, F.L. Burton, H.D. Stensel, Wastewater Engineering,

Treatment and Reuse, forth ed., McGraw–Hill, New Delhi, India, 2003.