Chemosphere xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Study of photocatalytic degradation of tributyltin, dibutylin and monobutyltin in water and marine sediments Stephan Brosillon a,⇑, Chrystelle Bancon-Montigny b, Julie Mendret a a b

IEM (Institut Européen des Membranes), UMR 5635 (CNRS-ENSCM-UM2), Université Montpellier 2, Place E. Bataillon, F-34095 Montpellier, France HydroSciences Montpellier, UMR 5569 (CNRS, IRD, UM1, UM2), Université Montpellier 2, Place E. Bataillon, F-34095 Montpellier, France

h i g h l i g h t s

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

 Organotins in aqueous solution are

completely removed by photocatalysis.  Organotins adsorbed on marine sediments are degraded by heteregeneous photocatalysis.  Photocatalytic organotins degradation is mainly due to hydroxyl radical attack.  Direct photolysis of organotins is the second mechanism of degradation.  AOP appears as a possible way for the remediation of marine sediments.

a r t i c l e

i n f o

Article history: Received 25 October 2013 Received in revised form 5 February 2014 Accepted 5 February 2014 Available online xxxx Keywords: Photocatalysis Organotin Tributyltin Water Marine sediment Dredge sediment treatment

a b s t r a c t This study reports on the first assessment of the treatment of sediments contaminated by organotin compounds using heterogeneous photocatalysis. Photocatalysis of organotins in water was carried out under realistic concentration conditions (lg L 1). Degradation compounds were analyzed by GC-ICP-MS; a quasi-complete degradation of tributyltin (TBT) in water (99.8%) was achieved after 30 min of photocatalytic treatment. The degradation by photolysis was about (10%) in the same conditions. For the first time decontamination of highly polluted marine sediments (certified reference material and harbor sediments) by photocatalysis proves that the use of UV and the production of hydroxyl radicals are an efficient way to treat organotins adsorbed onto marine sediment despite the complexity of the matrix. In sediment, TBT degradation yield ranged from 32% to 37% after only 2 h of irradiation (TiO2–UV) and the by-products: dibutyltin (DBT) and monobutyltin (MBT) were degraded very rapidly in comparison with TBT. It was shown that during photocatalysis of organotins in sediments, the hydroxyl radical attack and photolysis are the two ways for the degradation of adsorbed TBT. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Organotin compounds, especially tributyltin (TBT) compounds, was used in a wide range of applications including stabilizers in ⇑ Corresponding author. Tel.: +33 467143324. E-mail address: [email protected] (S. Brosillon).

the PVC industry, material protection (stone, leather, paper), industrial catalysts, algicides, fungicides, bactericides and wood preservatives (Hoch, 2001). In ports and marinas, TBT has been widely applied in the past as a biocide in antifouling paint marine coatings to prevent the growth and attachment of barnacles, mussels, tube worms, algae and other marine fouling organisms. TBT is perhaps the most toxic substance that has ever been

http://dx.doi.org/10.1016/j.chemosphere.2014.02.008 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Brosillon, S., et al. Study of photocatalytic degradation of tributyltin, dibutylin and monobutyltin in water and marine sediments. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.02.008

2

S. Brosillon et al. / Chemosphere xxx (2014) xxx–xxx

deliberately introduced into the marine environment by mankind (Goldberg, 1986). Marine snail and whelks imposex (hermaphroditism or sterility of females) occurs at low concentrations, in the order of ng L 1 (Alzieu, 1986, 1991, 1998, 2000). Owing to the high toxicity of TBT formulations in coastal ecosystems (Fent, 1996, 2004), organotins have been regulated and/or banned in antifouling paints, first in France in 1982 and later in many other countries. Studies have shown that despite the implementation of regulations or ban, high concentrations of TBT in water or in bottom sediments are still detected: in Spain (commercial and fishing harbors) (Diez et al. 2002), in France (Mediterranean marinas) (Michel and Averty, 1999; Cassi et al., 2008), in Poland (Filipkowska et al., 2011), Croatia (Furdek et al., 2012) and Argentina (Castro et al., 2012). Even with the enforcement of regulations or prohibition, it seems, therefore, that the problem of TBT pollution of harbor sediments is still a subject of concern, especially for the disposal of contaminated dredged sediments. After dredging, sediments are considered as waste and the dumping polluted sediments in the open sea is regulated by strict criteria (Mamindy-Pajany et al., 2012). The treatment of these highly contaminated sediments seems essential before their landfilling or reuse as building materials (construction of dams, barriers, etc.). There have been very few publications on the specific treatment of organotin in water and/or in sediment. Several techniques have been identified for the remediation of TBT from contaminated sediments: physical separation (hydrocycloning and froth flotation) (Reed et al., 2001), land deposition (Novak and Trapp, 2005), biodegradation (Saeki et al., 2007), thermal treatment (Mostofizadeh, 2001; Song et al., 2005), chemical oxidation (Mailhot et al., 1999; Pensaert et al., 2005) and electrochemical treatment (Stichnothe et al., 2001; Voulvoulis and Lester, 2006; Arevalo and Calmano, 2007). Anaerobic treatment of TBT adsorbed on sewage sludge was minimal (Voulvoulis and Lester, 2006) and this process appears to be not adapted for treating TBT. In this context, Advanced oxidation could be considered as a potential solution to decontaminate marine sediments. Among the advanced oxidative processes (AOP), heterogeneous photocatalysis appears as an interesting technique for the treatment of persistent organic pollution (Brosillon et al., 2011; Plantard et al., 2012). Indeed, TiO2 activation under UV irradiation (k < 390 nm) allows the generation of highly reactive OH free radicals from water or hydroxide ions. Navio et al. have studied the photocatalysis (UV/TiO2) of TBT (Navio et al., 1993, 1996) and triphenyltin (TPT) (Navio et al., 1997) in an aqueous solution at a very high concentration (5 mg L 1) in comparison with the levels found in raw water. TBT was partially degraded after 30 h of UV irradiation whereas dibutyltin (DBT) and monobutyltin (MBT) were under the limit of detection in less than 12 h. TPT (C0 = 2.5 mg L 1) was completely degraded after 2 h

30 min whereas diphenyltin and monophenyltin remain in solution as by-product of TPT degradation after 2 h 10 min. The goal of this study is to investigate the possibility to treat sediments contaminated with organotin compounds by heterogeneous photocatalysis. Firstly, photocatalysis of organotin in water was carried out in order to confirm previous findings (Navio et al., 1996, 1997). Unlike previous research, very low levels of organotin – close to those found in raw water – were studied: 5 lg(Sn) L 1 in this study vs 5 mg(Sn) L 1 in the literature (Navio et al., 1996; Bangkedphol et al., 2010). This experimental approach necessitated the ability to monitor low levels of organotin in both water and solid matrixes with an adapted analytical method. Decontamination of different marine sediments (certified reference material and harbor sediments) by photocatalysis was then examined.

2. Materials and methods 2.1. Reagents and environmental samples Chloride forms of MBT(95%), DBT(97%) and TBT(96%) were obtained from LGC Promochem (Molsheim, France). Tripropyltin chloride (TPrT, 98%) was obtained from Strem Chemicals (Bischeim, France). Stock organotin solutions containing 1000 mg(Sn) L 1 were prepared in methanol. When stored at +4 °C in the dark, they are stable for at least one year (Lespes, 1995). Methanol, sodium acetate, nitric and acetic acids and isooctane were purchased from Fisher Bioblock Scientific (Illkirch, France). Deionised MilliQ water (18.2 MX cm) was used. Tertbutanol (tButOH, purity 99.5%) was purchased from Sigma Aldrich (Saint-Quentin Fallavier, France). Sodium tetraethylborate (NaBEt4) used for derivatization was obtained from Strem Chemicals. NaBEt4 solutions (2% in deionised water) were prepared daily and stored at +4 °C in the dark. Glassware was decontaminated overnight in a 10% (v/v) nitric acid solution and rinsed thoroughly with deionised water prior to use (Bancon-Montigny et al., 1999, 2001). TiO2 (P 25 Degussa) was used as a photocatalyst in this study (BET: 54 m2 g 1; average particle diameter: 20 nm). Tests were performed on the PACS-2 marine sediment reference material (National Research Council of Canada-NRCC). This was collected from Esquimalt harbour, B.C, freeze dried, passed through a No. 120 (125 lm) screen, blended and bottled. Certified values and their uncertainties are reported as mass fractions (based on dry mass) in Table 1. Studied surface sediments were collected in February 2009 from a yachting harbor on the south-east French Mediterranean coast (Port-Camargue, Languedoc-Roussillon,) using a Shipek grab.

Table 1 Organotins, metals and organic carbon (POC) concentrations in PACS and surface sediments of Port-Camargue. The location in the harbor of the samples PC19, PC21, PC25 are given in Fig. 1 in Briant et al. (2013). (Italics: standard deviation) Organic carbone (%)

MBT

DBT

ng(Sn) g Certified reference material PACS 2

Certified values

3.3

Measured values

Port-Camargue marina

PC19

0.57

PC21

2.96

PC25

0.82

700

TBT

1

As

Co

mg kg

1

Cr

Cu

Mn

Ni

Pb

Ti

Zn

832 ±95 871 ±19

26.2 ±1.5

11.5 ±0.3

91 ±5

310 ±12

440 ±19

39.5 ±2.3

183 ±8

4400 ±300

364 ±23

579 ±28

1100 ±135 958 ±40

112 ±3 710 ±22 364 ±12

235 ±7 5066 ±127 1276 ±16

308 ±11 10738 ±161 3136 ±85

7.2 ±0.2 10.7 ±0.3 15.1 ±0.5

3.3 ±0.1 4.6 ±0.1 7.3 ±0.2

22 ±1 37 ±1 54 ±2

160 ±5 1497 ±45 321 ±10

202 ±6 206 ±6 353 ±11

11.4 ±0.3 17 ±1 29 ±1

29 ±1 94 ±3 66 ±2

1370 ±41 2253 ±68 2052 ±62

84 ±3 475 ±14 301 ±9

Please cite this article in press as: Brosillon, S., et al. Study of photocatalytic degradation of tributyltin, dibutylin and monobutyltin in water and marine sediments. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.02.008

3

S. Brosillon et al. / Chemosphere xxx (2014) xxx–xxx

2.2. Organotin extraction and analysis Extraction and analysis of organotin in sediments has been described previously in detail (Chahinian et al., 2012). Briefly, an extraction–derivatization step was first used to obtain complete alkylation of organotin compounds and to enable speciation analysis. Organotin compounds were determined using an inductively coupled plasma mass spectrometer as a detector (Thermo Electron ICPMS, XSeriesII) after capillary gas chromatography (Thermo Electron GC, Focus series) using a Thermo GC-ICP-MS transfer line (GPTR-AETE ‘‘Grand Plateau Technique pour la Recherche -Analyse des Eléments en Trace dans l’Environnement‘‘ analytical platform Montpellier2 University). All concentrations in the present paper are expressed with reference to the mass of tin. Quantification was performed using tripropyltin as an internal standard and by standard additions. The performance of X Series ICP-MS was tested, tuned and optimized as required for GC-ICP-MS analysis. An aqueous indium solution was aspirated continuously throughout the fully quantitative GC-ICP-MS analysis to allow correction for chromatographic baseline drift. The analytical method was validated with certified reference materials PACS-2 (harbor sediment). Comparison of determined concentrations and reference values has shown that experimental results conform well, and so the method can be considered as accurate. Blanks and certified reference material analyses were carried out systematically before each set of analyses. All analyses were done in triplicate. The limits of quantification range between 0.1 and 0.34 ng(Sn) g 1. 2.3. Dark adsorption experiments The adsorption isotherm of organotin on TiO2 was determined in the dark by mixing 50 mL of aqueous TBT solution with an initial concentrations C0  5000 ng(Sn) L 1 (in H2CO3/HCO3 buffer; pH = 6), in contact with 0.02–0.1 g TiO2. The temperature during the experiments was 25 ± 2 °C. The amount of TBT adsorbed onto the solid was estimated using the difference between the initial concentration of the solution and the concentration at equilibrium divided by the mass of adsorbent introduced. During the first experiments, samples were taken at increasing contact times (15 min, 30 min, 1, 6, 24 h) to determine the time necessary to reach equilibrium. 2.4. Photoreactor Irradiation was performed in a cylindrical batch reactor (volume: V = 500 mL), fitted with a 25 W low-pressure fluorescent lamp (Philips PLS9W10 4P, maximal emission wavelength 365 nm) placed vertically in a plugged tube. A quartz cylindrical jacket was placed around the plugging tube. In a typical experiment, the photocatalyst was mixed in the solution at a given mass. The radiant flux received by the solution was measured by means of a radiometer (UVA-365 Lutron Taiwan). The incident photon flux, Po, was estimated to (1.1 ± 0.1)  10 6 einstein s 1, corresponding to (23 ± 0.2) W m 2 at the outlet quartz jacket wall. Three hundred and fifty milliliters of solution (350 mL) were introduced into the photoreactor. The solution was mixed in the reactor by means of a magnetic stirrer. After an adsorption equilibrium was reached in the dark (30 min), the light was turned onto irradiate the solution and the first sample was taken (t = 0). A sketch of the reactor is given in the Fig. 1.

UV lamp

Cover

Quartz jacket

Organotin solution or sediment suspension

Magnetic stirrer

Fig. 1. Photoreactor set up.

or without TiO2 for varying times. The solution (buffer + sediment) was then filtered using a previously decontaminated PVDF 0.45 lm (Millipore) filter to collect the sediment. The sediment deposited on the filter then was analyzed as specified in the ‘‘Organotin extraction and analysis’’ section. 3. Results–discussions 3.1. Adsorption The concentrations measured after different times were quite similar during the adsorption experiments. However, it cannot be concluded that adsorption on TiO2 does not occur. Indeed, at a TBT concentration of 5195 ± 47 ng(Sn) L 1 and with 0.02–0.1 g TiO2, the change in concentration is probably too slight to be measured. Evolution of the concentration observed during the adsorption experiments thus corresponds to the standard deviation of the analytical method. As a consequence, in further experiments, a significant decrease in the TBT concentration could not be attributed to the adsorption of TBT onto TiO2 (even if adsorption occurs). 3.2. Photolysis Fig. 2 shows the decrease of the TBT and the time course of DBT and MBT in water during UV irradiation (UV only). It should be mentioned that very small concentrations of DBT and MBT (1%) are present in the initial TBT solution. The maximum absorption wavelength of butyltin compounds is within the UV region

2.5. Sediments photolysis–photocatalysis experiments Experiments on the photolysis–photocatalysis of the sediment were conducted in the following manner: the sediments were placed in the photoreactor with 350 mL of buffer carbonate with

Fig. 2. Evolution of TBT, DBT, MBT concentrations in aqueous solution during UV irradiation without TiO2: empty symbol (photolysis) and with TiO2: full symbol ([TiO2] = 0.8 g L 1 pH = 5.5; T = 20 °C).

Please cite this article in press as: Brosillon, S., et al. Study of photocatalytic degradation of tributyltin, dibutylin and monobutyltin in water and marine sediments. Chemosphere (2014), http://dx.doi.org/10.1016/j.chemosphere.2014.02.008

4

S. Brosillon et al. / Chemosphere xxx (2014) xxx–xxx

(190–290 nm) (Navio et al., 1996). The overlap of the UV–Visible adsorption spectra of TBT and the maximum emission wavelength (kmax = 365 nm) of the lamp results in the TBT compound being broken down. The photolysis experiments confirm the results observed by Navio et al. (1993)) in a UV photoreactor or those reported by Clark et al. (1988) in raw water irradiated by the sun. The UV light with a 365 nm wavelength has energy of about 339 kJ mol 1 which is enough to break the Sn–C bonds by photolysis. Indeed, the dissociation energy of these bonds is about 190–220 kJ mol 1 (Skinner, 1964). After a linear regression the first order rate constant was k = 12.4  10 2 h 1. A slow accumulation of DBT, during UV irradiation would confirm the splitting of the carbon-tin bond inducing debutylation of TBT as a consequence of the photolysis. The concentration of MBT is nearly constant and indicates that either the degradation of DBT is very slow or the degradation rate of MBT is similar to that of DBT and then no accumulation of MBT is observed. The second hypothesis is more likely since (Navio et al., 1993) have shown very similar kinetic constants during the photolysis degradation of DBT and MBT.

3.3. Photocatalysis of organotins in aqueous solution In order to confirm the previous finding (Navio et al., 1993), photocatalysis of TBT solutions was carried out, but with TBT concentrations which were 1000 times lower (5 lg(Sn) L 1 vs 5 mg(Sn) L 1). The degradation of TBT in an aqueous suspension of TiO2 irradiated by UV is very fast (Fig. 2). Indeed, the photocatalysis appears to be very effective in breaking down the TBT in solution. After 15 min of irradiation, the degradation yield of TBT in water is equal to 90% and a quasi-complete degradation of TBT is reached after 30 min (Y = 99.8%). The degradation yield using photolysis for the same duration (15 and 30 min) was equal to 6% and 10% respectively. Consequently the decrease in the concentration of TBT in the presence of TiO2 is mainly due to heterogeneous photocatalytic degradation. The dramatic decrease of TBT is only accompanied by a slight increase of DBT and MBT. This result is in accordance with the previously performed studies: no quantitative increase of DBT and MBT concentrations was found during photocatalysis degradation of TBT (Navio et al., 1996). Considering the kinetic constants found by the previously cited authors, it appears that the degradation rates of DBT and MBT by photocatalysis are 20 and 120 folds higher, respectively, than the TBT degradation rate. Hence, the fact that there was no accumulation of DBT and MBT could be explained by a very fast degradation of the DBT and MBT in comparison with TBT. This is in accordance with (Navio et al., 1993) who have observed that the order of degradation rate is TBT < DBT