Kinetics and mechanisms of ultrasonic degradation of

Ultrasonics Sonochemistry 9 (2002) 317–323 www.elsevier.com/locate/ultsonch

Kinetics and mechanisms of ultrasonic degradation of volatile chlorinated aromatics in aqueous solutions Yi Jiang a b

a,b

, Christian Petrier b, T. David Waite

a,*

School of Civil and Environmental Engineering, University of New South Wales (UNSW), Sydney, NSW 2052, Australia Laboratoire de Chimie Mol eculaire et Environnement, ESIGEC––Universit e de Savoie, 73376 Le Bourget du Lac, France Received 1 October 2001; received in revised form 16 January 2002; accepted 12 February 2002

Abstract Ultrasonic decompositions of chlorobenzene (ClBz), 1,4-dichlorobenzene and 1-chloronaphthalene were investigated at 500 kHz in order to gain insight into the kinetics and mechanisms of the decomposition process. The disappearance of ClBz on sonication is almost simultaneously accompanied by the release of chloride ions as a result of the rapid cleavage of carbon–chlorine bonds with a concomitant release of CO, C2 H2 , CH4 and CO2 . The intermediates resulting from attack of HO radicals were detected but in a quite low yield (less than 2 lM). The generation of H2 O2 on sonolysis is not significantly affected by the presence of aqueous ClBz while the generation of NO 2 and NO3 is inhibited initially due to the presence of ClBz which diffuses into the gas–bubble interfaces and inhibits the interactions between free radicals and nitrogen. Moreover, brown carbonaceous particles are present throughout the ultrasonic irradiation process, which are consistent with soot formation under pyrolytic conditions. These important features suggest that, at the relatively high initial substrate concentrations used in this study, ultrasonic degradation of ClBz takes place predominantly both within the bubbles and within the liquid–gas interfaces of bubbles where it undergoes high-temperature combustion. Under these conditions, the oxidation of ClBz by free radical HO outside of bubbles is a minor factor (though results of recent studies suggest that attack by HO is more important at lower initial substrate concentrations). The sonochemical decomposition of volatiles follows pseudo-first-order reaction kinetics but the degradation rates are affected by operating conditions, particularly initial substrate concentration and ultrasonic intensity. Ó 2002 Elsevier Science B.V. All rights reserved. Keywords: Sonochemistry; Chlorinated aromatics; Volatiles; Wastewater treatment

1. Introduction The sonolysis of a variety of volatile compounds in aqueous media is of great interest due to the growing need yet limited technologies available for the elimination of undesirable chemical contaminants from water. It has been suggested in the recent years that carbon tetrachloride [1–3], chlorobenzene [4–7], and a variety of other volatile contaminants [8–12] could be quickly destroyed by ultrasonic irradiation. Based on the experimental results of various investigators, ultrasonic degradation of volatile compounds probably occurs via oxidation at the hot bubble–liquid interfacial regions by highly reactive free radicals (HO ) or occurs inside col* Corresponding author. Tel.: +61-2-9385-5060; fax: +61-2-93856139. E-mail address: [email protected] (T. David Waite).

lapsing bubbles predominantly by direct pyrolysis. The destruction rates of volatile contaminants are expected to be dependent on chemical and physical conditions inside the bubbles and to be linked to their physical and chemical properties, particularly their hydrophobic and volatile nature. It is likely that volatile compounds will decompose within ‘‘hot-spots’’ where cavitation bubbles pulsate or collapse rather than taking place homogeneously. As such, the kinetics and mechanism of ultrasonic degradation of volatile compounds are expected to be rather complex. In this study, we investigate the ultrasonic degradation of three volatile chlorinated aromatics (chlorobenzene (ClBz), 1,4-dichlorobenzene (1,4-DClBz) and 1-chloronaphthalene (ClNt)) in air-equilibrated aqueous medium at 500 kHz. The yields of chloride ions resulting from sonication were detected and, for ClBz, the major

1350-4177/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 5 0 - 4 1 7 7 ( 0 2 ) 0 0 0 8 5 - 8

318

Y. Jiang et al. / Ultrasonics Sonochemistry 9 (2002) 317–323

intermediates and aqueous and gaseous products generated on sonolysis were analysed. The effects of both initial substrate concentration and ultrasonic power on kinetic rate constants were also examined in order to gain insight into the kinetics and mechanisms of the degradation of these volatile compounds.

2. Material and methods ClBz, 1,4-DClBz and ClNt, as well as 4-chlorophenol (4-CP) and 4-chlorocatechol (4-CC), were obtained from Aldrich, while phenol, hydroquinone (HQ) and benzoquinone (BQ) were obtained from Prolabo. All chemicals were reagent grade (at least 99% purity) and were used as received. Aqueous solutions were prepared by dissolving the compounds in ultra-pure Milli-Q deionized water. Ultrasonic irradiation was carried out in a cylindrical water-jacketed glass cell equipped with a Teflon holder. The 500 kHz ultrasonic transducer supplied by Electric Service was made up of a piezo-electric disc (diameter 4 cm) fixed on a titanium plate. The reactor was hermetically sealed and connected to a gas burette to ensure a constant pressure (1 atm). The temperature of the liquid was monitored using a thermocouple immersed in the reacting medium and was maintained at 20 1 °C. In all cases, 250 ml of aqueous solution was treated by sonication. The ultrasonic power dissipated into the reactors was adjusted and quantified by calorimetry in order to ensure comparative ultrasonic conditions. The volatile compounds and aqueous intermediates generated by ultrasonic irradiation were identified using a high performance liquid chromatograph (Waters model 600E) with an absorbance detector (Waters model 486) and equipped with a spherisorb ODS2 5 lm C18 column (250 mm 4:6 mm). The detection wavelength was set at 205 nm for ClBz, at 220 nm for 1,4DClBz and at 225 nm for ClNt. Specific mobile phases were used for each analysis, i.e., acetonitrile/water/acetic acid (80/20/1) for ClBz and DClBz and acetonitrile/ water/acetic acid (75/25/1) for ClNt. Samples were injected directly into the chromatograph. The identity of intermediates was confirmed by comparing retention times with those of known standards, and their concentration determined from calibration curve. Chloride ions, nitrite and nitrate ions were detected using an ion chromatograph (Waters model ILC-1) with a conductimetric detector (Waters model 430) and equipped with a Universal Anion column (150 mm 4:6 mm). The mobile phase was a benzoic acid aqueous solution (4 103 mol l1 ) at pH 6 (adjusted with LiOH). The calibration was performed using an aqueous sodium chloride solution. Carbon monoxide (CO), carbon dioxide (CO2 ), acetylene (C2 H2 ) and methane (CH4 ) were analysed using a

gas chromatograph (Intersmat model IGC 16). Separations were performed on a Porapak Q (2 mm 2:5 m) column and detection was achieved with a FID detector after hydrogenation. One hundred microlitres gaseous headspace was regularly sampled using a gas syringe and immediately injected. Calibration was realised with standard gaseous mixtures (Allteth Scotty II). Hydrogen peroxide (H2 O2 ) concentration was determined iodometrically via its ammonium molybdate decomposition reaction in a 10% potassium iodide solution.

3. Results and discussion 3.1. Kinetics of the sonolysis of volatile compounds in aqueous media As can be seen in Figs. 1 and 2, ultrasonic degradations of the three substrates were rapid and efficient at a frequency of 500 kHz and followed pseudo-first-order reaction kinetics. The rate constants obtained and yields of sonochemical degradation are calculated and given in Table 1. The release of chloride ions almost simultaneously accompanied the disappearance of the volatiles in solutions with more than 90% of chloride atoms in each case being recovered as chloride ions (Fig. 2). Based upon these experimental results and those from other investigators [4–7], it can be deduced that chlorine atoms are mineralised as chloride ions as a result of the

Fig. 1. Evolution of the relative concentration [log (Ct =Ci )] of the compounds versus sonication time (min) at 500 kHz with ultrasonic power of 25 W and temperature of 20 1 °C. Ct is the concentration of each compound at time t and Ci is the initial concentration of each compound.


319

3.1.1. Effects of initial substrate concentration on kinetics rate constant Although pyrolytic reactions are expected to follow simple first-order kinetics, the ultrasonic degradation of the volatile compounds examined here follows pseudofirst-order reaction kinetics with a significant effect of the initial substrate concentration on the decomposition rates of the target molecules. As can be seen in Fig. 3, the observed pseudo-first-order rate constant (kobs ) for sonolysis of ClBz solution was 0:020 0:001 min1 when ½ClBzi ¼ 500 lM, while kobs was found to increase to 0:035 0:001 min1 when ½ClBzi ¼ 40 lM. Similarly on sonolysis of 1,4-DClBz, the rate constant kobs was 0:022 0:001 min1 when ½1; 4 DClBzi ¼ 500 lM, and increased to 0:036 0:001 min1 when ½1; 4 DClBzi ¼ 50 lM. The finding that rate constant of ultrasonic degradation of volatiles decreases with increasing initial concentration is in agreement with previous observation of Drijvers and coworkers [7,12], Hoffmann et al. [8] and Zhang and Hua [10]. Decreasing rate constants of ultrasonic degradation at higher initial volatile concentrations can be explained by the influence of the volatile compounds on the average specific heat ratio c of the gas mixture in the cavitation bubbles where c is the ratio of constant pressure and constant volume heat capacities, i.e. Cp =Cv . The proportionality between the concentration of a volatile compound in the bubble and its concentration in the solution will influence the ultrasonic reaction rate as the temperature of bubble collapse is dependent upon the specific heat ratio (c) of the gas mixture. Tmax ¼ T0 ðc 1ÞPmax =Pmin

ð1Þ

where Tmax is the final temperature, Pmax the liquid pressure at collapse, Pmin the minimum pressure in vapour phase and T0 the ambient temperature of liquid. The specific heat ratio c depends on the composition of gas mixture in the cavitation bubble. If the mole fraction of volatile compounds in cavitation bubble is relatively small, then a linear relationship between the specific heat ratio (c) of the gas mixture and the volatile compound concentration (Ci ) in the liquid phase can be assumed [12]; i.e. c ¼ c0 KCi Fig. 2. Evolution of concentration versus sonication time for volatile compounds disappearance and chloride ion formation at 500 kHz with an ultrasonic power of 25 W and aqueous temperature of 20 1 °C.

rapid cleavage of carbon–chlorine bonds by high temperature combustion predominantly occurring both within the cavitation bubbles and at hot liquid–gas interfaces.

ð2Þ

Here, K is a proportionality constant (M1 ) and is positive since the specific heat ratio of the gas–water vapour mixture will be larger than the specific heat ratio of volatile compounds. As a result, the specific heat ratio of a gas mixture decreases with increasing volatile concentration resulting in a (relatively) lower temperature and pressure within the cavitation bubbles and thus a decrease in rate of sonochemical degradation of volatile compounds.

320


Table 1 Reaction rate constants (kobs ) observed on sonolysis of the three volatiles with different initial concentration (Ci ) under following operating conditions: ultrasonic power (Pw ): 25 W; frequency, 500 kHz; aqueous temperature: 20 1 °C Compound

Pw (W)

Ci (lM)

kobs (min1 )

Yield of Cl (%)

ClBz 1,4-DClBz ClNt

25 1 25 1 25 1

200 300 40

0:026 0:001 0:028 0:001 0:036 0:001

92.5 93.3 94.5

Fig. 3. Evolution of rate constant (kobs ) observed versus initial concentration of volatile compounds for the degradation of ClBz and 1,4DClBz at 500 kHz with an ultrasonic power of 25 W and aqueous temperature of 20 1 °C.

3.1.2. Effects of ultrasonic power on kinetics rate constant The rate constants of ultrasonic degradation for 1,4DClBz and ClNt solutions increase with increasing ultrasonic intensity. A linearly relationship between the ultrasonic power and reaction rate constant, at least between 10 and 50 W, was observed on sonolysis of 200 lM 1,4-DClBz and ClNt solutions (Fig. 4). The observed kinetics constant (kobs ) of 1,4-DClBz degradation is 0:054 0:001 min1 at 50 W compared to 0:009 0:001 min1 at 10 W. Similarly, the kobs of ClNt degradation is 0:038 0:001 min1 at 50 W while it is 0:008 0:001 min1 at 10 W. Increasing the ultrasonic power will increase the energy of cavitation, lower the threshold limit of cavitation, and enhance the quantity of the cavitation bubbles. The destruction of volatile compounds more likely occurs inside the cavitation bubbles thus the rate of sonochemical degradation should be related to the number of bubbles present if each bubble releases enough energy to ‘‘burn’’ the volatile pollutant. Above the cavitation threshold, the number of cavitation bubbles increases with increasing ultrasonic intensity; consequently, the ultrasonic degradation rate of both 1,4-DClBz and ClNt is higher with increasing ultrasonic intensity. An optimum of ultrasonic energy has been reported by various investigators [13,14] with no sonochemical reaction occurring below the cavitation threshold, in-

Fig. 4. Evolution of rate constant (kobs ) versus ultrasonic power for the degradation of 1,4-DClBz and ClNt at 500 kHz with aqueous temperature of 20 1 °C.

creasing rate on increasing ultrasonic energy for energies greater than the cavitation threshold then, at high energies, slowing down significantly. Under the operating conditions used here, no sonochemical activity occurs for ultrasonic power 2000 K) and pressure gradients exist at the interface of the gas–bubbles [18–20] where thermolytic decomposition of volatile compounds is likely to occur. In addition, the possible reaction of ClBz with OH radicals at the liquid–gas interface would result in the formation of chlorophenyl radicals (reaction (11)) [9,15]: C6 H5 Cl þ OH ! C6 H4 Cl þ H2 O

Fig. 6. Evolution of concentration versus sonication time for ClBz and gaseous products at 500 kHz with an ultrasonic power of 25 W and aqueous temperature of 20 1 °C.

C6 H5 Cl ! C6 H4 Cl þ H

ð4Þ

Dissociation by reaction (3) is energetically more favourable than by reaction (4) due to the lower bond dissociation energy of C–Cl (400 kJ/mol) compared to that of C–H (463 kJ/mol) [15]. Most of the ClBz will thermally decompose to phenyl radicals and chloride radicals. In addition, the solution becomes increasingly brown and cloudy on continued sonolysis. Brown carbonaceous particles were observed visually to appear after approximately 50 min of sonication. It is deduced

ð10Þ

ð11Þ

Besides direct degradation by dissociation and reaction with OH radicals, ClBz is thought to degrade by radicals formed as a result of the thermal dissociation of other ClBz molecules as well. For example, reaction with Cl leads to chlorophenyl radicals and HCl (reaction (12)) [15]: C6 H5 Cl þ Cl ! C6 H4 Cl þ HCl

ð12Þ

Therefore, the formation of most intermediate products identified during sonolysis of ClBz could be explained by reactions between ClBz, acetylene and phenyl, chlorophenyl, Cl and C4 H3 radicals. With regard to gaseous products, acetylene (C2 H2 ), methane (CH4 ), carbon monoxide (CO), and carbon dioxide (CO2 ) were all generated through the degradation process (Fig. 6). Hart et al. indicated [17] that C2 H2 and CH4 are formed by a thermolytic ring cleavage of phenyl radicals such as reaction (5), (7), (9) while CO constitutes the first product in the multi-step thermal degradation (reaction (8)) of volatile compounds. In the experiments here, the concentration of carbon dioxide

322


(CO2 ) in gaseous phase increased with increasing sonication time as a consequence of CH4 , C2 H2 and CO being further oxidised through high temperature combustion. The experimental results reported here demonstrate that the concentration of total C (carbon) detected in the gaseous phase rises as a function of sonication time, and represents about 45% (based on C content) of the starting ClBz concentration in aqueous solution (Fig. 6). Although the total carbon mass balance was not obtained due to lack of appropriate standards for the brown carbonaceous particles formed, such an analysis would be of value in future. The generation of hydrogen peroxide (H2 O2 ) in these systems is of particular interest. The quantity of H2 O2 formed on sonolysis of water increases with increasing sonication time (Fig. 5). It has been reported [21,22] that hydrogen peroxide forms as a consequence of hydroxyl (including hydroperoxyl) radical recombination external to the cavitation bubbles (reaction (13) and (14)): 2HO ! H2 O2

ð13Þ

2HOO ! H2 O2 þ O2

ð14Þ

The finding that generation of H2 O2 is not significantly affected by the presence of aqueous ClBz demonstrates again that oxidation of ClBz by free radical HO outside of the bubbles (even at the interfaces of liquid–gas bubbles) is a minor factor. The observation that intermediates resulting from attacks of highly reactive radicals (HO , HOO ) on ClBz are too low to be detected is supportive of the above finding. It is therefore deduced that highly reactive HO radicals are involved to only a minor extent in the ultrasonic degradation of ClBz, at least under the conditions of this study. The results of recent studies [23,24] by Drijvers et al. and Dewulf et al. support the conclusion that, at initial substrate concentrations of 0.2–0.5 mM, degradation via combustion will dominate over that induced by hydroxyl radical attack. These investigators show however that at lower initial substrate concentrations (in the 1–10 lM range), hydroxyl radical induced degradation is more important than pyrolysis. The generation of nitrate and nitrite ions (shown in Fig. 5) also suggests that ClBz decomposition occurs predominantly inside the cavitation bubbles. It has been reported [25,26] that NO 2 are the primary products on sonolysis of air-equilibrated water with the generation of NO 3 resulting from the oxidation of NO2 by hydrogen peroxide. Hart and Henglein indicated [25] that in the first instance, nitrogen is oxidised to nitrogen oxide and nitrous oxide by reaction with HO and O inside the cavitation bubble. Further NO and N2 O oxidation with HO and molecular oxygen leads to nitrous acid (reaction (15)–(19)). The reaction rate is dependent on solution pH, which decreases rapidly with sonication time [26].

N2 þ HO ! N2 O þ Hþ

ð15Þ

N2 þ O ! N2 O

ð16Þ

N2 þ O ! NO þ N

ð17Þ

N2 O þ O ! 2NO

ð18Þ

NO þ ðO; HO Þ ! HNO2

ð19Þ

It is therefore expected that nitrate and nitrite ions would be generated as soon as an air-equilibrated solution is exposed to ultrasonic irradiation. However, the generation of NO 2 and NO3 is observed to be inhibited in the initial stages of sonication (seen in Fig. 5) presumably as a consequence of the presence of ClBz in solution. It is surmised that volatile ClBz diffuses to the gas–bubble interfaces and inhibits the interaction between free radicals (HO , HOO etc.) and nitrogen. Hence, increased yields of NO 2 and NO3 were detected after ultrasonic irradiation of approximately 60 min after which time the concentration of ClBz in solution was relatively low. The nitrates are further oxidized to nitrites by highly reactive free radicals (HO , HOO ) and by hydrogen peroxide (H2 O2 ) as well. These important features suggest that, under the conditions of this study, ultrasonic degradation of volatile ClBz takes place predominantly inside cavitation bubbles or more likely at the interfaces of liquid–gas bubbles where it undergoes high-temperature combustion. The results of other recent studies [23,24] suggest that hydroxyl radical attack in bulk solution may be more important than combustion at lower initial substrate concentrations. The degradation mechanisms of ClBz under ultrasonic irradiation are however quite complex and a comprehensive and quantitative analysis of the products remaining in solution, including the brown carbonaceous soots, would assist in gaining further insights as would more detailed kinetic modelling of both the pyrolysis and hydroxyl radical degradation pathways.

4. Conclusion Ultrasonic degradations of ClBz, 1,4-DClBz and ClNt in air-equilibrated aqueous solutions are rapid and efficient and follow pseudo-first-order reaction kinetics. The cavitation bubbles are the centre of chemical activity induced by acoustic perturbation. The disappearance of the three volatile compounds by sonication is simultaneously accompanied by release of chloride ions as the chlorine atoms are rapidly mineralised. More than 90% of the chlorine is recovered as chloride ions with CO, C2 H2 , CH4 and CO2 also identified as the major gaseous products of ClBz degradation. Intermediates resulting from attack of HO radicals were detected but


in a quite low yield (less than 2 lM) and disappeared on extended ultrasonic irradiation. The generation of H2 O2 on sonolysis is not significantly affected by the presence of aqueous ClBz while the generation of NO 2 and NO3 is inhibited initially due to the presence of ClBz which diffuses to the gas– bubble interfaces and inhibits the interactions between free radicals and nitrogen. In addition, brown carbonaceous particles are present throughout the ultrasonic irradiation process, an observation consistent with soot formation under pyrolytic conditions. These important features suggest that, at the relatively high initial substrate concentrations used in this study, ultrasonic degradation of ClBz take place predominantly both within the bubbles and at the liquid–gas interfaces of bubbles where it undergoes high-temperature combustion. Under these conditions, the oxidation of ClBz by free radical HO outside of bubbles appears to be a minor degradation pathway (though results of recent studies suggest that attack by HO is more important at lower initial substrate concentrations). The sonochemical decomposition of volatile compounds follows pseudofirst-order reaction kinetics, however the kinetics rate constant of sonochemical degradation is affected by operating conditions, particularly the initial substrate concentration and the ultrasonic intensity. In summary, ultrasonic treatment of volatiles in aqueous media is efficient but the opportunity would appear to exist for further optimisation of operating conditions. References [1] I. Hua, M.R. Hoffmann, Environ. Sci. Technol. 30 (1996) 864. [2] A. Francony, C. Petrier, Ultrason. Sonochem. 3 (1996) S77.

323

[3] C. Petrier, A. Francony, Ultrason. Sonochem. 4 (1997) 295. [4] M.F. Lamy, Ph.D. thesis, Savoie University, 1994. [5] C. Petrier, Y. Jiang, M.F. Lamy, Environ. Sci. Technol. 32 (1998) 1316. [6] P. Kruus, R.C. Burk, M.H. Entezari, R. Otson, Ultrason. Sonochem. 4 (1997) 229. [7] D. Drijvers, H. Van Langenhove, K. Vervaet, Ultrason. Sonochem. 5 (1998) 13. [8] M.R. Hoffmann, I. Hua, R. Hochemer, Ultrason. Sonochem. 3 (1996) 163. [9] D. Drijvers, H. Van Langenhove, L. Nguyen Thi Kim, L. Bray, Ultrason. Sonochem. 6 (1999) 115. [10] G. Zhang, I. Hua, Environ. Sci. Technol. 34 (2000) 1529. [11] F. Cataldo, Ultrason. Sonochem. 7 (2000) 35. [12] A. De Visscher, P. Van Eenoo, D. Drijvers, H. Van Langenhove, J. Phys. Chem. 100 (1996) 11636. [13] T.J. Mason, J.P. Lorimer, Sonochemistry: Theory, Applications and Uses of Ultrasound in Chemistry, Ellis Horrid, New York, 1988. [14] J.-L. Luche, Synthetic Organic Sonochemistry, Plenum Press, New York, 1998. [15] E.R. Ritter, J.W. Bozzelli, A.M. Dean, J. Phys. Chem. 94 (1990) 2493. [16] R.D. Kern, K. Hie, H. Chen, Combust. Sci. Technol. 85 (1992) 77. [17] E.J. Hart, C.H. Fisher, A. Henglein, J. Phys. Chem. 94 (1990) 284. [18] K.S. Suslick (Ed.), Ultrasound: Its Chemical Physical and Biological Effects, VCH Publishers, New York, 1988. [19] E.B. Flint, K.S. Suslick, J. Phys. Chem. 95 (1991) 1484. [20] P. Riesz, T. Kondo, Free Radical Biol. Med. 13 (1992) 247. [21] K. Makino, M.M. Mossoba, P. Riese, J. Phys. Chem. 87 (1982) 3537. [22] K. Makino, M.M. Mossoba, P. Riese, J. Phys. Chem. 104 (1983) 1369. [23] D. Drijvers, H. Van Langenhove, V. Herrygers, Ultrason. Sonochem. 7 (2000) 87. [24] J. Dewulf, H. Van Langenhove, A. De Visscher, S. Sabbe, Ultrason. Sonochem. 8 (2001) 143. [25] E.J. Hart, A. Henglein, J. Phys. Chem. 90 (1986) 5992. [26] Y.N. Lee, J.A. Lind, J. Phys. Chem. 91 (1986) 2793.