- Email: [email protected]
Removal of cyanide from water by means of plasma discharge technology
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Removal of cyanide from water by means of plasma discharge technology Marı´a Hijosa-Valsero a,*, Ricardo Molina b, Hendrik Schikora c, Michael Mu¨ller c, Josep M. Bayona a a
Instituto de Diagno´stico Ambiental y Estudios del Agua (IDAEA), CID, CSIC, C/Jordi Girona 18-26, E-08034 Barcelona, Spain Instituto de Quı´mica Avanzada de Catalun˜a (IQAC), CID, CSIC, C/Jordi Girona 18-26, E-08034 Barcelona, Spain c Fraunhofer IGB, Nobelstraße 12, 70569 Stuttgart, Germany b
article info
abstract
Article history:
Two different nonthermal plasma reactors at atmospheric pressure were assessed for the
Received 26 July 2012
first time for cyanide removal (1 mg L1) from aqueous solutions (0.025 M NaHCO3/NaOH
Received in revised form
buffer, pH 11) at laboratory scale. Both devices were dielectric barrier discharge (DBD) re-
27 December 2012
actors; one of them was a conventional batch reactor (R1) and the other one was a coaxial
Accepted 1 January 2013
thin falling water film reactor (R2). A first-order degradation kinetics was proposed for both
Available online 9 January 2013
experiments, obtaining kR1 ¼ 0.5553 min1 and kR2 ¼ 0.7482 min1. The coaxial reactor R2 yielded a removal of 99% within only 3 min. Energy efficiencies (G) were calculated, yielding
Keywords:
1.74 mg kW1 h1 for R1 and 127.9 mg kW1 h1 for R2. When the treatment was applied to
Cyanide
industrial wastewaters, cyanide elimination was confirmed, although at a lower rate
Wastewater
(above 92% removal in 90 min with R2). Therefore, plasma reactors could be a relevant
Plasma
alternative to established advanced oxidation techniques (UV, H2O2, ozonation, etc.) for the
Dielectric barrier discharge
removal of cyanide from wastewaters with low organic loads or even drinking waters. ª 2013 Elsevier Ltd. All rights reserved.
1.
Introduction
The presence of cyanide in natural waters is a matter of great concern, since it can be harmful to aquatic ecosystems and to human health. In this respect, a future legislative toughening cannot be discarded (Directive 2000/60/EC; Directive 2008/105/ EC). In the European Union, the annual average concentration of total cyanides in surface continental waters cannot exceed 40 mg L1 (Royal Decree 60/2011). Cyanide can enter the natural waters through releases from the metal finishing industries, iron and steel mills, runoff from disposal of cyanide wastes in landfills, pesticides and the use of cyanide-containing road
salts (Dash et al., 2009). Due to its high toxicity, it is essential to find efficient and cost-effective technologies to reduce cyanide concentrations in industrial effluents. The most commonly used method for cyanidecontaminated effluents is the alkaline chlorination process, although many other technologies exist, like biological oxidation/biodegradation, hydrogen peroxide, SO2/air (INCO) process, ozonation, anodic oxidation, electrodialysis, reverse osmosis, electrowinning, hydrolysis/distillation, acidification/ volatilization and reneutralisation, flotation, iron cyanide precipitation, activated carbon, resin, catalytic oxidation, Caro’s acid or photolysis (Dash et al., 2009). Some of these
* Corresponding author. Tel.: þ34 93 400 61 00x1305; fax: þ34 93 204 59 04. E-mail addresses: [email protected], [email protected] (M. Hijosa-Valsero), [email protected] (R. Molina), [email protected] (H. Schikora), [email protected] (M. Mu¨ller), [email protected] (J.M. Bayona). 0043-1354/$ e see front matter ª 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.watres.2013.01.001
1702
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
processes are energy and/or reagent intensive, others generate problematic by-products, and others are not well established yet (Dash et al., 2009). High-voltage electrical discharges directly in water (electrohydraulic discharge) or in the gas phase above the water (nonthermal plasma) produce hydrogen peroxide, molecular oxygen and hydrogen, and hydroxyl, hydroperoxyl, hydrogen, oxygen and other radicals. In addition, shock waves and UV light may also be formed. These reactive species and physical conditions, in turn, have been shown to rapidly and efficiently degrade many organic compounds (Locke et al., 2006). Several ways of plasma treatments have been applied to evaluate the removal of many substances from water, especially volatile organic compounds (Gajewski et al., 2004), phenols (Lukes and Locke, 2005), organic dyes (Abdelmalek et al., 2006) and pharmaceuticals (Krause et al., 2009). The degradability of cyanide with a thermal plasma reactor at 10e20 kW has been demonstrated (Fortin et al., 2000; Soucy et al., 2001), but there are no reports about the removal of this compound with nonthermal plasma techniques. Since cyanide can be thermally degraded (thermal hydrolysis) from aqueous solutions, the use of nonthermal plasma reactors would allow the study of plasma effects separately. In the present study, the degradation of cyanide in aqueous solution (distilled water) by means of two different nonthermal plasma reactors at atmospheric pressure was assessed and compared for the first time. Both laboratory devices were based on dielectric barrier discharge (DBD) reactors; one of them was a conventional batch reactor and the other one was a coaxial thin falling water film reactor. Finally, the feasibility of the treatment was confirmed with industrial wastewater spiked with cyanide.
2.
Material and methods
2.1.
Chemicals and reagents
Water and methanol were HPLC and GC quality, respectively. Sodium cyanide, taurine and 2,3-naphtalenedialdehyde (2,3NDA) were provided by SigmaeAldrich (Steinheim, Germany), sodium hydroxide was bought from Carlo Erba (Rodano, Italy), citric acid monohydrate and di-sodium hydrogen phosphate dihydrate (Na2HPO4$2H2O) were purchased from Merck (Darmstadt, Germany), and hydrochloric acid 37% and sodium hydrogencarbonate (NaHCO3) were bought from Panreac (Castellar del Valle`s, Spain). Solid-phase microextraction (SPME) fibres, coated with 75 mm carboxenepolydimethylsiloxane, were obtained from Supelco (Bellefonte, PA, USA). An aqueous McIlvaine buffer (pH ¼ 8) was made with Na2HPO4$2H2O and citric acid. Then, a 5 mM taurine solution was prepared in that buffer. A 4 mM solution of 2,3-NDA in methanol was subsequently diluted with that buffer to obtain an aqueous solution of 1 mM 2,3-NDA.
2.2.
confirm the suitability of the plasma treatment for complex samples, a second experiment was performed with primarytreated wastewater from a pharmaceutical industry. This wastewater (initial conditions: COD z 18,000 mg L1, pH ¼ 4.80, electric conductivity ¼ 318 mS cm1), which originally contained no cyanide, was spiked with 1 mg L1 cyanide and alkalinised with KOH (final conditions: pH ¼ 10.92, electric conductivity ¼ 778 mS cm1). Warning: HCN is an extremely toxic gas; then, to avoid its formation, the pH of cyanide solutions must be kept 2 units above its pKa (9.2).
2.3.
Plasma reactors
2.3.1.
DBD batch reactor (reactor R1)
The scheme of reactor R1 is shown in Fig. 1. Gas mass flow meter and controllers (Bronkhorst, Ruurlo, The Netherlands) were used in order to introduce helium gas (5 Ln min1) in the reactor chamber. A 100 kHz frequency was generated with a GF-855 function generator (Promax, L’Hospitalet de Llobregat, Spain) and connected to a linear amplifier AG-1012 (T&C Power Conversion Inc., Rochester, NY, USA). A matching network and two transformers (HR-Diemen S.A., Sant Hipo`lit de Voltrega`, Spain) were connected in series to the output of the amplifier in order to increase the voltage up to z20 kV. The input power in the plasma reactor was kept constant at 30 W. The experiments were performed at room temperature and the temperature inside the reactor did not exceed 40 C. A glass Petri dish (Steriplan, Duran Group, Wertheim, Germany) without its lid was placed inside the reactor. The distance between the Petri dish (ø ¼ 30 mm 10 mm height) and the upper part of the plasma reactor was kept constant at 2 mm. Four millilitres of a cyanide solution with a concentration of 1 mg L1 were filled into the Petri dish. The reactor was then closed, the high voltage was applied and the plasma was formed. In order to obtain an insight into the type of degradation kinetics, several samples (all of them of 1 mg L1) were subjected to individual treatments with increasing times (0 s, 10 s, 30 s, 2 min, 5 min and 15 min). The final volumes of the treated samples were measured in order to correct the
Cyanide solutions
For the experiments with aqueous solutions, cyanide was prepared in a buffer solution of 0.025 M NaHCO3/NaOH in distilled water (pH ¼ 11, conductivity 4 mS cm1). In order to
Fig. 1 e Scheme of the plasma reactor R1.
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
calculated cyanide concentration for water losses by evaporation in contact with the flowing helium.
2.3.2.
Coaxial thin film DBD reactor (reactor R2)
Reactor R2 is shown in Fig. 2. Its geometrical setup represents a coaxial DBD arrangement. In contrast to the setup used by Magureanu et al. (2008), the high voltage electrode (hot electrode) is a mesh of copper (120 cm2), which is wrapped at the outside of a glass vessel. The glass vessel represents the dielectric barrier. In an axial symmetric manner, the grounded electrode consists of a stainless steel tube, which is located in the axial centre of the vessel. The outer diameter of this metallic tube is adjusted in such a way, that there is only a small open slit of about 3.5 mm between the tube and the inner diameter of the glass vessel. A volume of 175 mL cyanide solution (1 mg L1) is pumped in a recirculating manner (Model M4230/1, Maprotec, Idstein, Germany) from the bottom side within the inner lumen of the metallic tube towards the top of the tube and then falling down in an axial symmetric manner (after correct vertical adjustment of the reactor) on the completely wetted outside surface of the tube to its bottom side. There, the solution is collected during the plasma treatment time and recirculated (1.2 L min1) by a gear pump into the inner lumen of the stainless steel tube (grounded electrode). At the lower base of the vessel 500 sccm helium (MKS Multi gas controller 647C in combination with an MKS Mass flow meter 1179B) is introduced and flows upwards within the open discoidal slit of about 2 mm between the falling water film and the inner wall of the glass vessel to a gas exit on the upper side of the glass vessel with a contact time of 0.5 s approximately. When DC pulses of high voltage (12 kV, ton 0.2 ms and toff 5 ms, repetition frequency 94 kHz, 24 W generated by a HV Generator G2000, Redline Technologies Elektronik, Baesweiler, Germany) is applied between the outer copper mesh and the inner metallic tube, a filamentous plasma discharge is formed within the discoidal slit and completely spread from the bottom to the upper end between the two concentric arranged electrode areas.
Fig. 2 e Scheme of the plasma reactor R2.
2.4.
1703
Analytical methodology
A fluorimetric determination of free cyanide in the aqueous samples (distilled water with NaHCO3/NaOH buffer) was performed by modifying the method described by Sano et al. (1992). Briefly, 1 mL of sample was mixed with 1 mL of taurine 5 mM and with 1 mL of 2,3-NDA 1 mM. A green fluorescent complex is formed. After 1 h, the fluorescence of the samples was measured with a Varian Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, Santa Clara, CA, USA) by adjusting the parameters at l excitation ¼ 441 nm, l emission ¼ 483 nm, voltage ¼ 480 V. The linearity range of the method was established from 0.01 mg L1 to 10 mg L1. Wastewater samples, whose matrix was more complex, were analysed by headspace SPME and gas chromatography with nitrogen phosphorus detector (GC-NPD), according to Boadas-Vaello et al. (2008). Briefly, 4 mL of sample were put into a 20 mL glass vial, which was immediately capped. By means of a syringe, 25 mL HCl was injected into the vial in order to decrease the solution pH with the double objective of releasing cyanide from possible complexes and of transforming cyanide into HCN (a gas available for headspace SPME). For the adsorption of HCN, the SPME fibre was introduced in the vial for 35 min at 37 C. The desorption was carried out during 5 min at 250 C in the injector of a Trace-GCultra coupled to a NPD and provided with a TriPlus autosampler (Thermo Scientific, Waltham, MA, USA). Injection was performed in the splitless mode (split flow 60 mL min1, splitless time 3 min). The carrier gas was helium at 1 mL min1. The GC separation was performed in a GasPro 60 m 0.32 mm column (J&W Scientific, Folsom, CA, USA). Oven temperature was programmed from 130 C (1 min), ramped at 2 C min1 to 150 C (0 min), and then at 8 C min1 to 260 C (5 min). The detector temperature was 250 C. The linearity range of the method was 0.08e10 mg L1 for wastewater samples.
3.
Results and discussion
3.1.
Removal of cyanide from aqueous solutions
Fig. 3 shows the concentration decrease of the cyanide solutions with increasing treatment times in both plasma reactors. Firstly, it must be said that the final cyanide concentration was below the detection limit (0.01 mg L1) for the samples treated during 15 min in R1 and during 3 min in R2 (these data were not included in Fig. 3). The removal attained in R1 was 99% after 15 min, whereas R2 obtained a removal of 99% after only 3 min. Fig. 3 shows that R2 has a slightly better performance than R1; which could be related to the greater surface of water exposed to the plasma irradiation in R2. The results of other innovative methods for cyanide removal are shown in Table 1, like established advanced oxidation processes (AOP) and thermal plasma. In those AOP systems, the removal of cyanide is probably due to photolysis or to the action of hydroxyl radicals and not to ozonation (Ford et al., 2006; Kepa et al., 2008). The disadvantage of the established AOP methods in comparison to plasma is the use of
1704
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
cyanide, but a great effort has still to be made to bring this biological method onto the full-scale operation (Dash et al., 2009).
3.1.1.
Degradation kinetics in aqueous solutions
The measured cyanide concentrations related to every treatment time were assessed to elucidate the order of the degradation kinetics. In both reactors, the reaction was fitted to an exponential decay equation (Fig. 3). Accordingly, a first-order degradation kinetics is proposed with values of k ¼ 0.5553 min1 for the batch reactor R1 and k ¼ 0.7482 min1 for the falling water film reactor R2 (for an initial concentration of 1 mg L1 cyanide). The calculated k values of the present work were higher than those reported in the literature (Table 1; this comparison refers only to k values expressed in the same units). Under the studied conditions, the energy efficiency (G) for a 90% removal in the studied plasma reactors was calculated by adapting Spinks and Woods’ formula (Spinks and Woods, 1976),
Fig. 3 e Removal of cyanide from distilled water (pH [ 11 with NaHCO3/NaOH buffer) in R1 and R2. Exponential decay equations are shown. Cyanide initial concentration 1 mg LL1.
G¼
1 k C0 V ; 1 x P ln x
where G is expressed in mol J1, k is the first-order rate kinetic constant (s1), C0 is the initial concentration of the sample (M), V is the volume of the sample (L), P is the input power applied to the reactor (W) and (1 1/x) is the desired removal expressed as a fraction of one (in this case, for a 90% removal, this term
reagents, which make the treatment more expensive. In addition, the concentration of those reagents must always be optimised in order to avoid efficiency losses. In fact, Monteagudo et al. (2004) indicated that H2O2 can act as a radical scavenger. Biodegradation is also a feasible treatment for
Table 1 e Removal of cyanide with different AOP reactors. Treatment system
Plasma DBD, 30 W, He Coaxial thin film DBD, 24 W, He Coaxial thin film DBD, 24 W, He Thermal plasma torch inside water, 10 kW, Ar Thermal plasma torch inside water, 19 kW, Ar/N2 Other advanced oxidation UV/H2O2/O2 UV/H2O2 H2O2/O3 UV/O3/H2O2 O3/H2O2 H2O2/O3 TiO2/sunlight
Water type
C0 (mg L1)
Treatment time (min)
Removal (%)
First order kinetic constant, k (min1)
Reference
Pure water with NaHCO3/NaOH Pure water with NaHCO3/NaOH
1
15
99
0.555
This work
1
3
99
0.748
This work
Pharmaceutical industry wastewater
1
90
>92
e
This work
Aluminium industry
300
180
z80
0.010e0.012
Fortin et al., 2000
Aluminium industry
350
120
z90
0.022
Fortin et al., 2000
10
60
90
e
Dura´n et al., 2009
0.6
20
75
e
Ford et al., 2006
0.5 157 3
e e 30
92 99 >90
e e 0.053
Kepa et al., 2008 Kim et al., 2003 Monteagudo et al., 2004
250 5e62
117 e
99.9 e
e 0.001a
Mudliar et al., 2009 Augugliaro et al., 1999
processes (AOP) Thermoelectric power station Engine manufacturing facility Pure water Plating industry Thermoelectric power station Automobile industry Pure water
a The units in this particular case are M Einstein1 instead of min1.
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
1705
HOeC]N, O]CeNH2, OOC(O)NH2) and final by-products (NO3 , NO2 , NH4 þ , N2, HCO3 , CO3 2 , saponified compounds) (Monteagudo et al., 2004; Ford et al., 2006), whose toxicity is low. Plasma treatments are expected to lead to similar reactions to those of conventional AOPs.
3.2.
Fig. 4 e Temporal evolution of cyanide concentration and COD concentration in the industrial wastewater treated with R2. Initial concentrations: 1 mg LL1 cyanide, 18,000 mg O2 LL1 COD.
equals 0.9; hence, x ¼ 10); obtaining values of 1.85 1011 mol J1 (1.74 mg kW1 h1) for R1, and 1.37 109 mol J1 (127.9 mg kW1 h1) for R2. From these data, it can be clearly inferred that R2 is far more efficient than R1.
3.1.2.
Degradation by-products
Conventional AOPs are known to generate degradation byproducts, which sometimes have been identified as dangerous compounds (Vogna et al., 2004). In the case of the studied pollutant, cyanide ion is believed to react with O3, OH and H2O2, yielding intermediate compounds (OCN, NO2 ,
Removal of cyanide from industrial wastewater
The temporal evolution of cyanide and COD concentrations in the treated wastewater is shown in Fig. 4. Contrary to the observations in distilled water (Fig. 3), cyanide removal did not follow a clear exponential decay (k ¼ 0.2462; r ¼ 0.6583), and the kinetic reaction order could not be accurately established (Table 1). Cyanide concentration experienced a fast and steep decrease during the first minute of the treatment, but then the concentration remained almost constant until t ¼ 15 min. From that point, the concentration slowly decreased and after 90 min treatment it was below detection limits (Fig. 5), which represents a removal efficiency greater than 92%. This is a low elimination rate if we compare it with the distilled water experiment in the same reactor (99% removal after 3 min in R2, see Fig. 3). However, the great amount of organic matter present in this wastewater must be taken into account to evaluate the results. Normally, AOPs are conceived as tertiary treatments, and they should treat wastewaters with lower organic loads (approximately, COD below 150 mg L1). In this experiment, we used R2 as a secondary treatment; but it can be assumed that its cyanide removal performance would improve working as a tertiary treatment for more diluted wastewaters. On the other hand, COD removal from wastewater in the present experiment was low, attaining a decrease of 10% after 90 min treatment. This slow abatement
Fig. 5 e Chromatograms (GC-NPD) of the industrial wastewater treated with R2 during 0 min, 2 min, 15 min and 90 min. Hydrogen cyanide peak is marked with an arrow. Initial concentration: 1 mg LL1 cyanide.
1706
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
could be related both to the high initial COD concentration and to the nature of that organic matter. Other authors have studied COD removal with a different plasma technique (non-thermal gliding arc discharge, a highly energetic treatment), finding removals of about 90% in 3 h for an initial COD concentration of 80e150 mg L1 (Ghezzar et al., 2009) and 90% in 10 min for an initial concentration of 960 mg L1 (Njoyim-Tamungang et al., 2011). In less energetic processes, like conventional AOPs, a wide variety of COD removal efficiencies has been reported. In a homogeneous photocatalysis with UV/H2O2/O2, a COD elimination of 42% was attained in 60 min (initial concentration 998 mg L1) (Dura´n et al., 2009). Monteagudo et al. (2004) confirmed that UV radiation plays an important role in COD elimination, obtaining a removal of 75% in 40 min with O3/UV and O3/ H2O2/UV systems in a wastewater with an initial COD of 1500 mg L1, whereas lower efficiencies were recorded in the absence of UV light. Mahmud et al. (2012) proved that the efficiency of a photo-Fenton process for COD elimination in a landfill leachate decreased as the initial COD concentration increased (3000e10,000 mg L1). This supports the low COD removal performance observed in the present work, since the treated wastewater was heavily loaded with organic compounds. This clear difference between cyanide and COD removal efficiency from industrial wastewater in the DBD reactor must be associated with the complexity of the substances. The cyanide ion is a simple molecule, which is completely transformed and therefore eliminated after oxidation or cleavage reactions.
4.
Conclusions
The DBD plasma reactor at atmospheric pressure represents a promising technique to remove cyanide from water. Moreover, it avoids the input and consumption of additional reagents, which is indeed cost-effective and allows water reuse. The combination of electrical discharges, formation of reactive species and generation of UV radiation could have lead to the cleavage or substitution of the stable triple bond C^N. However, the efficiency of this kind of plasma reactor declines when the treated solution has a great concentration of dissolved organic matter, which can have a shielding effect or compete for the plasma-generated oxidant species. Therefore, the use of a coaxial DBD plasma reactor as a tertiary treatment for diluted wastewaters or even drinking waters is suggested. The application of the described treatment technology to large wastewater volumes implies a delicate scale-up stage to design a new reactor with the adequate dimensions and electric configuration. Some steps have already been taken in this direction, and the authors are working on the construction of a DBD full-scale reactor. Currently, the main limitation of this technology is the consumption of gases; therefore, some alternatives like gas recycling or the use of cheaper gases (like O2 or O3) should be carefully considered.
Acknowledgements The authors are very grateful to the pharmaceutical industry LEBSA for providing the wastewater. We thank J. Carretero-
Ariza and Dr. J. Garcı´a (Universitat Polite`cnica de Catalunya) for COD analyses. This research was funded by the European Commission under the Seventh Framework Program for Research and Technological Development of the European Union (FP7-SME-2010-1 e Grant agreement no.: 2262033WATERPLASMA).
references
Abdelmalek, F., Ghezzar, M.R., Belhadj, M., Addou, A., Brisset, J.L., 2006. Bleaching and degradation of textile dyes by nonthermal plasma process at atmospheric pressure. Industrial & Engineering Chemistry Research 45, 23e29. Augugliaro, V., Blanco-Ga´lvez, J., Ca´ceres-Va´squez, J., Garcı´aLo´pez, E., Loddo, V., Lo´pez-Mun˜oz, M.J., Malato-Rodrı´guez, S., Marcı`, G., Palmisano, L., Schiavello, M., Soria-Ruiz, J., 1999. Photocatalytic oxidation of cyanide in aqueous TiO2 suspensions irradiated by sunlight in mild and strong oxidant conditions. Catalysis Today 54, 245e253. Boadas-Vaello, P., Jover, E., Llorens, J., Bayona, J.M., 2008. Determination of cyanide and volatile alkylnitriles in whole blood by headspace solid-phase microextraction and gas chromatography with nitrogen phosphorus detection. Journal of Chromatography B 870, 17e21. Dash, R.R., Gaur, A., Balomajumder, C., 2009. Cyanide in industrial wastewaters and its removal: a review on biotreatment. Journal of Hazardous Materials 163, 1e11. Directive 2000/60/EC of the European Parliament and of the Council. Official Journal of the European Communities L 327/1, 22 Dec 2000. http://eur-lex.europa.eu/LexUriServ/LexUriServ. do?uri¼CELEX:32000L0060:EN:NOT. (accessed July 2012). Directive 2008/105/EC of the European Parliament and of the Council. Official Journal of the European Union L 348/84, 24 Dec 2008. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do? uri¼CELEX:32008L0105:EN:NOT. (accessed July 2012). Dura´n, A., Monteagudo, J.M., Sanmartı´n, I., Garcı´a-Pen˜a, F., Coca, P., 2009. Treatment of IGCC power station effluents by physico-chemical and advanced oxidation processes. Journal of Environmental Management 90, 1370e1376. Ford, J., Hernandez, R., Zappi, M., 2006. Bench-scale evaluation of advanced oxidation processes for treatment of a cyanidecontaminated wastewater from an engine manufacturing facility. Environmental Progress 25, 32e38. Fortin, L., Kasireddy, V.K., Soucy, G., Bernier, J.-L., 2000. Novel reactor for cyanide solution treatment. Canadian Journal of Chemical Engineering 78, 643e649. ski, J., ny, D., Fleszyn Gajewski, J.B., Ma˛czka, T., Nowak-Woz niak, R., 2004. Some hydrocarbon derivative compounds Poz decomposition using the HV impulse discharges. Journal of Electrostatics 62, 269e275. Ghezzar, M.R., Abdelmalek, F., Belhadj, M., Benderdouche, N., Addou, A., 2009. Enhancement of the bleaching and degradation of textile wastewaters by gliding arc discharge plasma in the presence of TiO2 catalyst. Journal of Hazardous Materials 164, 1266e1274. Kepa, U., Stanczyk-Mazanek, E., Stepniak, L., 2008. The use of the advanced oxidation process in the ozone þ hydrogen peroxide system for the removal of cyanide from water. Desalination 223, 187e193. Kim, Y.-J., Qureshi, T.I., Min, K.-S., 2003. Application of advanced oxidation processes for the treatment of cyanide containing effluent. Environmental Technology 24, 1269e1276. Krause, H., Schweiger, B., Schuhmacher, J., Scholl, S., Steinfeld, U., 2009. Degradation of the endocrine disrupting chemicals (EDCs) carbamazepine, clofibric acid, and
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
iopromide by corona discharge over water. Chemosphere 75, 163e168. Locke, B.R., Sato, M., Sunka, P., Hoffmann, M.R., Chang, J.-S., 2006. Electrohydraulic discharge and nonthermal plasma for water treatment. Industrial & Engineering Chemistry Research 45, 882e905. Lukes, P., Locke, B.R., 2005. Degradation of substituted phenols in a hybrid gaseliquid electrical discharge reactor. Industrial & Engineering Chemistry Research 44, 2921e2930. Magureanu, M., Piroi, D., Mandache, N.B., Parvulescu, V., 2008. Decomposition of methylene blue in water using a dielectric barrier discharge: optimization of the operating parameters. Journal of Applied Physics 104, 103306-1e103306-7. Mahmud, K., Hossain, M.D., Shams, S., 2012. Different treatment strategies for highly polluted landfill leachate in developing countries. Waste Management 32, 2096e2105. Monteagudo, J.M., Rodrı´guez, L., Villasen˜or, J., 2004. Advanced oxidation processes for destruction of cyanide from thermoelectric power station waste waters. Journal of Chemical Technology and Biotechnology 79, 117e125. Mudliar, R., Umare, S.S., Ramteke, D.S., Wate, S.R., 2009. Energy efficient e advanced oxidation process for treatment of cyanide containing automobile industry wastewater. Journal of Hazardous Materials 164, 1474e1479.
1707
Njoyim-Tamungang, E., Laminsi, S., Ghogomu, P., Njopwouo, D., Brisset, J.-L., 2011. Pollution control of surface waters by coupling gliding discharge treatment with incorporated oyster shell powder. Chemical Engineering Journal 173, 303e308. Royal Decree 60/2011, 22 Jan 2011. About the rules of environmental quality in the field of water policy (in Spanish). Boletı´n Oficial del Estado 19, 6854e6870. Sec. I. http://www. boe.es/boe/dias/2011/01/22/pdfs/BOE-A-2011-1139.pdf (accessed July 2012). Sano, A., Takimoto, N., Takitani, S., 1992. High-performance liquid chromatographic determination of cyanide in human red blood cells by pre-column fluorescence derivatization. Journal of Chromatography 582, 131e135. Soucy, G., Fortin, L., Kasireddy, V.K., Bernier, J.-L., Kimmerle, F.M., 2001. Innovative cyanide solution treatment by thermal plasma. In: Young, C.A., Tidwell, L.G., Anderson, C.G. (Eds.), Cyanide: Social, Industrial and Economic Aspects. TMS, Warrendale, pp. 379e384. Spinks, J.W.T., Woods, R.J., 1976. An Introduction to Radiation Chemistry. Wiley, New York. Vogna, D., Marotta, R., Andreozzi, R., Napolitano, A., d’Ischia, M., 2004. Kinetic and chemical assessment of the UV/H2O2 treatment of antiepileptic drug carbamazepine. Chemosphere 54, 497e505.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
Removal of cyanide from water by means of plasma discharge technology Marı´a Hijosa-Valsero a,*, Ricardo Molina b, Hendrik Schikora c, Michael Mu¨ller c, Josep M. Bayona a a
Instituto de Diagno´stico Ambiental y Estudios del Agua (IDAEA), CID, CSIC, C/Jordi Girona 18-26, E-08034 Barcelona, Spain Instituto de Quı´mica Avanzada de Catalun˜a (IQAC), CID, CSIC, C/Jordi Girona 18-26, E-08034 Barcelona, Spain c Fraunhofer IGB, Nobelstraße 12, 70569 Stuttgart, Germany b
article info
abstract
Article history:
Two different nonthermal plasma reactors at atmospheric pressure were assessed for the
Received 26 July 2012
first time for cyanide removal (1 mg L1) from aqueous solutions (0.025 M NaHCO3/NaOH
Received in revised form
buffer, pH 11) at laboratory scale. Both devices were dielectric barrier discharge (DBD) re-
27 December 2012
actors; one of them was a conventional batch reactor (R1) and the other one was a coaxial
Accepted 1 January 2013
thin falling water film reactor (R2). A first-order degradation kinetics was proposed for both
Available online 9 January 2013
experiments, obtaining kR1 ¼ 0.5553 min1 and kR2 ¼ 0.7482 min1. The coaxial reactor R2 yielded a removal of 99% within only 3 min. Energy efficiencies (G) were calculated, yielding
Keywords:
1.74 mg kW1 h1 for R1 and 127.9 mg kW1 h1 for R2. When the treatment was applied to
Cyanide
industrial wastewaters, cyanide elimination was confirmed, although at a lower rate
Wastewater
(above 92% removal in 90 min with R2). Therefore, plasma reactors could be a relevant
Plasma
alternative to established advanced oxidation techniques (UV, H2O2, ozonation, etc.) for the
Dielectric barrier discharge
removal of cyanide from wastewaters with low organic loads or even drinking waters. ª 2013 Elsevier Ltd. All rights reserved.
1.
Introduction
The presence of cyanide in natural waters is a matter of great concern, since it can be harmful to aquatic ecosystems and to human health. In this respect, a future legislative toughening cannot be discarded (Directive 2000/60/EC; Directive 2008/105/ EC). In the European Union, the annual average concentration of total cyanides in surface continental waters cannot exceed 40 mg L1 (Royal Decree 60/2011). Cyanide can enter the natural waters through releases from the metal finishing industries, iron and steel mills, runoff from disposal of cyanide wastes in landfills, pesticides and the use of cyanide-containing road
salts (Dash et al., 2009). Due to its high toxicity, it is essential to find efficient and cost-effective technologies to reduce cyanide concentrations in industrial effluents. The most commonly used method for cyanidecontaminated effluents is the alkaline chlorination process, although many other technologies exist, like biological oxidation/biodegradation, hydrogen peroxide, SO2/air (INCO) process, ozonation, anodic oxidation, electrodialysis, reverse osmosis, electrowinning, hydrolysis/distillation, acidification/ volatilization and reneutralisation, flotation, iron cyanide precipitation, activated carbon, resin, catalytic oxidation, Caro’s acid or photolysis (Dash et al., 2009). Some of these
* Corresponding author. Tel.: þ34 93 400 61 00x1305; fax: þ34 93 204 59 04. E-mail addresses: [email protected], [email protected] (M. Hijosa-Valsero), [email protected] (R. Molina), [email protected] (H. Schikora), [email protected] (M. Mu¨ller), [email protected] (J.M. Bayona). 0043-1354/$ e see front matter ª 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.watres.2013.01.001
1702
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
processes are energy and/or reagent intensive, others generate problematic by-products, and others are not well established yet (Dash et al., 2009). High-voltage electrical discharges directly in water (electrohydraulic discharge) or in the gas phase above the water (nonthermal plasma) produce hydrogen peroxide, molecular oxygen and hydrogen, and hydroxyl, hydroperoxyl, hydrogen, oxygen and other radicals. In addition, shock waves and UV light may also be formed. These reactive species and physical conditions, in turn, have been shown to rapidly and efficiently degrade many organic compounds (Locke et al., 2006). Several ways of plasma treatments have been applied to evaluate the removal of many substances from water, especially volatile organic compounds (Gajewski et al., 2004), phenols (Lukes and Locke, 2005), organic dyes (Abdelmalek et al., 2006) and pharmaceuticals (Krause et al., 2009). The degradability of cyanide with a thermal plasma reactor at 10e20 kW has been demonstrated (Fortin et al., 2000; Soucy et al., 2001), but there are no reports about the removal of this compound with nonthermal plasma techniques. Since cyanide can be thermally degraded (thermal hydrolysis) from aqueous solutions, the use of nonthermal plasma reactors would allow the study of plasma effects separately. In the present study, the degradation of cyanide in aqueous solution (distilled water) by means of two different nonthermal plasma reactors at atmospheric pressure was assessed and compared for the first time. Both laboratory devices were based on dielectric barrier discharge (DBD) reactors; one of them was a conventional batch reactor and the other one was a coaxial thin falling water film reactor. Finally, the feasibility of the treatment was confirmed with industrial wastewater spiked with cyanide.
2.
Material and methods
2.1.
Chemicals and reagents
Water and methanol were HPLC and GC quality, respectively. Sodium cyanide, taurine and 2,3-naphtalenedialdehyde (2,3NDA) were provided by SigmaeAldrich (Steinheim, Germany), sodium hydroxide was bought from Carlo Erba (Rodano, Italy), citric acid monohydrate and di-sodium hydrogen phosphate dihydrate (Na2HPO4$2H2O) were purchased from Merck (Darmstadt, Germany), and hydrochloric acid 37% and sodium hydrogencarbonate (NaHCO3) were bought from Panreac (Castellar del Valle`s, Spain). Solid-phase microextraction (SPME) fibres, coated with 75 mm carboxenepolydimethylsiloxane, were obtained from Supelco (Bellefonte, PA, USA). An aqueous McIlvaine buffer (pH ¼ 8) was made with Na2HPO4$2H2O and citric acid. Then, a 5 mM taurine solution was prepared in that buffer. A 4 mM solution of 2,3-NDA in methanol was subsequently diluted with that buffer to obtain an aqueous solution of 1 mM 2,3-NDA.
2.2.
confirm the suitability of the plasma treatment for complex samples, a second experiment was performed with primarytreated wastewater from a pharmaceutical industry. This wastewater (initial conditions: COD z 18,000 mg L1, pH ¼ 4.80, electric conductivity ¼ 318 mS cm1), which originally contained no cyanide, was spiked with 1 mg L1 cyanide and alkalinised with KOH (final conditions: pH ¼ 10.92, electric conductivity ¼ 778 mS cm1). Warning: HCN is an extremely toxic gas; then, to avoid its formation, the pH of cyanide solutions must be kept 2 units above its pKa (9.2).
2.3.
Plasma reactors
2.3.1.
DBD batch reactor (reactor R1)
The scheme of reactor R1 is shown in Fig. 1. Gas mass flow meter and controllers (Bronkhorst, Ruurlo, The Netherlands) were used in order to introduce helium gas (5 Ln min1) in the reactor chamber. A 100 kHz frequency was generated with a GF-855 function generator (Promax, L’Hospitalet de Llobregat, Spain) and connected to a linear amplifier AG-1012 (T&C Power Conversion Inc., Rochester, NY, USA). A matching network and two transformers (HR-Diemen S.A., Sant Hipo`lit de Voltrega`, Spain) were connected in series to the output of the amplifier in order to increase the voltage up to z20 kV. The input power in the plasma reactor was kept constant at 30 W. The experiments were performed at room temperature and the temperature inside the reactor did not exceed 40 C. A glass Petri dish (Steriplan, Duran Group, Wertheim, Germany) without its lid was placed inside the reactor. The distance between the Petri dish (ø ¼ 30 mm 10 mm height) and the upper part of the plasma reactor was kept constant at 2 mm. Four millilitres of a cyanide solution with a concentration of 1 mg L1 were filled into the Petri dish. The reactor was then closed, the high voltage was applied and the plasma was formed. In order to obtain an insight into the type of degradation kinetics, several samples (all of them of 1 mg L1) were subjected to individual treatments with increasing times (0 s, 10 s, 30 s, 2 min, 5 min and 15 min). The final volumes of the treated samples were measured in order to correct the
Cyanide solutions
For the experiments with aqueous solutions, cyanide was prepared in a buffer solution of 0.025 M NaHCO3/NaOH in distilled water (pH ¼ 11, conductivity 4 mS cm1). In order to
Fig. 1 e Scheme of the plasma reactor R1.
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
calculated cyanide concentration for water losses by evaporation in contact with the flowing helium.
2.3.2.
Coaxial thin film DBD reactor (reactor R2)
Reactor R2 is shown in Fig. 2. Its geometrical setup represents a coaxial DBD arrangement. In contrast to the setup used by Magureanu et al. (2008), the high voltage electrode (hot electrode) is a mesh of copper (120 cm2), which is wrapped at the outside of a glass vessel. The glass vessel represents the dielectric barrier. In an axial symmetric manner, the grounded electrode consists of a stainless steel tube, which is located in the axial centre of the vessel. The outer diameter of this metallic tube is adjusted in such a way, that there is only a small open slit of about 3.5 mm between the tube and the inner diameter of the glass vessel. A volume of 175 mL cyanide solution (1 mg L1) is pumped in a recirculating manner (Model M4230/1, Maprotec, Idstein, Germany) from the bottom side within the inner lumen of the metallic tube towards the top of the tube and then falling down in an axial symmetric manner (after correct vertical adjustment of the reactor) on the completely wetted outside surface of the tube to its bottom side. There, the solution is collected during the plasma treatment time and recirculated (1.2 L min1) by a gear pump into the inner lumen of the stainless steel tube (grounded electrode). At the lower base of the vessel 500 sccm helium (MKS Multi gas controller 647C in combination with an MKS Mass flow meter 1179B) is introduced and flows upwards within the open discoidal slit of about 2 mm between the falling water film and the inner wall of the glass vessel to a gas exit on the upper side of the glass vessel with a contact time of 0.5 s approximately. When DC pulses of high voltage (12 kV, ton 0.2 ms and toff 5 ms, repetition frequency 94 kHz, 24 W generated by a HV Generator G2000, Redline Technologies Elektronik, Baesweiler, Germany) is applied between the outer copper mesh and the inner metallic tube, a filamentous plasma discharge is formed within the discoidal slit and completely spread from the bottom to the upper end between the two concentric arranged electrode areas.
Fig. 2 e Scheme of the plasma reactor R2.
2.4.
1703
Analytical methodology
A fluorimetric determination of free cyanide in the aqueous samples (distilled water with NaHCO3/NaOH buffer) was performed by modifying the method described by Sano et al. (1992). Briefly, 1 mL of sample was mixed with 1 mL of taurine 5 mM and with 1 mL of 2,3-NDA 1 mM. A green fluorescent complex is formed. After 1 h, the fluorescence of the samples was measured with a Varian Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, Santa Clara, CA, USA) by adjusting the parameters at l excitation ¼ 441 nm, l emission ¼ 483 nm, voltage ¼ 480 V. The linearity range of the method was established from 0.01 mg L1 to 10 mg L1. Wastewater samples, whose matrix was more complex, were analysed by headspace SPME and gas chromatography with nitrogen phosphorus detector (GC-NPD), according to Boadas-Vaello et al. (2008). Briefly, 4 mL of sample were put into a 20 mL glass vial, which was immediately capped. By means of a syringe, 25 mL HCl was injected into the vial in order to decrease the solution pH with the double objective of releasing cyanide from possible complexes and of transforming cyanide into HCN (a gas available for headspace SPME). For the adsorption of HCN, the SPME fibre was introduced in the vial for 35 min at 37 C. The desorption was carried out during 5 min at 250 C in the injector of a Trace-GCultra coupled to a NPD and provided with a TriPlus autosampler (Thermo Scientific, Waltham, MA, USA). Injection was performed in the splitless mode (split flow 60 mL min1, splitless time 3 min). The carrier gas was helium at 1 mL min1. The GC separation was performed in a GasPro 60 m 0.32 mm column (J&W Scientific, Folsom, CA, USA). Oven temperature was programmed from 130 C (1 min), ramped at 2 C min1 to 150 C (0 min), and then at 8 C min1 to 260 C (5 min). The detector temperature was 250 C. The linearity range of the method was 0.08e10 mg L1 for wastewater samples.
3.
Results and discussion
3.1.
Removal of cyanide from aqueous solutions
Fig. 3 shows the concentration decrease of the cyanide solutions with increasing treatment times in both plasma reactors. Firstly, it must be said that the final cyanide concentration was below the detection limit (0.01 mg L1) for the samples treated during 15 min in R1 and during 3 min in R2 (these data were not included in Fig. 3). The removal attained in R1 was 99% after 15 min, whereas R2 obtained a removal of 99% after only 3 min. Fig. 3 shows that R2 has a slightly better performance than R1; which could be related to the greater surface of water exposed to the plasma irradiation in R2. The results of other innovative methods for cyanide removal are shown in Table 1, like established advanced oxidation processes (AOP) and thermal plasma. In those AOP systems, the removal of cyanide is probably due to photolysis or to the action of hydroxyl radicals and not to ozonation (Ford et al., 2006; Kepa et al., 2008). The disadvantage of the established AOP methods in comparison to plasma is the use of
1704
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
cyanide, but a great effort has still to be made to bring this biological method onto the full-scale operation (Dash et al., 2009).
3.1.1.
Degradation kinetics in aqueous solutions
The measured cyanide concentrations related to every treatment time were assessed to elucidate the order of the degradation kinetics. In both reactors, the reaction was fitted to an exponential decay equation (Fig. 3). Accordingly, a first-order degradation kinetics is proposed with values of k ¼ 0.5553 min1 for the batch reactor R1 and k ¼ 0.7482 min1 for the falling water film reactor R2 (for an initial concentration of 1 mg L1 cyanide). The calculated k values of the present work were higher than those reported in the literature (Table 1; this comparison refers only to k values expressed in the same units). Under the studied conditions, the energy efficiency (G) for a 90% removal in the studied plasma reactors was calculated by adapting Spinks and Woods’ formula (Spinks and Woods, 1976),
Fig. 3 e Removal of cyanide from distilled water (pH [ 11 with NaHCO3/NaOH buffer) in R1 and R2. Exponential decay equations are shown. Cyanide initial concentration 1 mg LL1.
G¼
1 k C0 V ; 1 x P ln x
where G is expressed in mol J1, k is the first-order rate kinetic constant (s1), C0 is the initial concentration of the sample (M), V is the volume of the sample (L), P is the input power applied to the reactor (W) and (1 1/x) is the desired removal expressed as a fraction of one (in this case, for a 90% removal, this term
reagents, which make the treatment more expensive. In addition, the concentration of those reagents must always be optimised in order to avoid efficiency losses. In fact, Monteagudo et al. (2004) indicated that H2O2 can act as a radical scavenger. Biodegradation is also a feasible treatment for
Table 1 e Removal of cyanide with different AOP reactors. Treatment system
Plasma DBD, 30 W, He Coaxial thin film DBD, 24 W, He Coaxial thin film DBD, 24 W, He Thermal plasma torch inside water, 10 kW, Ar Thermal plasma torch inside water, 19 kW, Ar/N2 Other advanced oxidation UV/H2O2/O2 UV/H2O2 H2O2/O3 UV/O3/H2O2 O3/H2O2 H2O2/O3 TiO2/sunlight
Water type
C0 (mg L1)
Treatment time (min)
Removal (%)
First order kinetic constant, k (min1)
Reference
Pure water with NaHCO3/NaOH Pure water with NaHCO3/NaOH
1
15
99
0.555
This work
1
3
99
0.748
This work
Pharmaceutical industry wastewater
1
90
>92
e
This work
Aluminium industry
300
180
z80
0.010e0.012
Fortin et al., 2000
Aluminium industry
350
120
z90
0.022
Fortin et al., 2000
10
60
90
e
Dura´n et al., 2009
0.6
20
75
e
Ford et al., 2006
0.5 157 3
e e 30
92 99 >90
e e 0.053
Kepa et al., 2008 Kim et al., 2003 Monteagudo et al., 2004
250 5e62
117 e
99.9 e
e 0.001a
Mudliar et al., 2009 Augugliaro et al., 1999
processes (AOP) Thermoelectric power station Engine manufacturing facility Pure water Plating industry Thermoelectric power station Automobile industry Pure water
a The units in this particular case are M Einstein1 instead of min1.
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
1705
HOeC]N, O]CeNH2, OOC(O)NH2) and final by-products (NO3 , NO2 , NH4 þ , N2, HCO3 , CO3 2 , saponified compounds) (Monteagudo et al., 2004; Ford et al., 2006), whose toxicity is low. Plasma treatments are expected to lead to similar reactions to those of conventional AOPs.
3.2.
Fig. 4 e Temporal evolution of cyanide concentration and COD concentration in the industrial wastewater treated with R2. Initial concentrations: 1 mg LL1 cyanide, 18,000 mg O2 LL1 COD.
equals 0.9; hence, x ¼ 10); obtaining values of 1.85 1011 mol J1 (1.74 mg kW1 h1) for R1, and 1.37 109 mol J1 (127.9 mg kW1 h1) for R2. From these data, it can be clearly inferred that R2 is far more efficient than R1.
3.1.2.
Degradation by-products
Conventional AOPs are known to generate degradation byproducts, which sometimes have been identified as dangerous compounds (Vogna et al., 2004). In the case of the studied pollutant, cyanide ion is believed to react with O3, OH and H2O2, yielding intermediate compounds (OCN, NO2 ,
Removal of cyanide from industrial wastewater
The temporal evolution of cyanide and COD concentrations in the treated wastewater is shown in Fig. 4. Contrary to the observations in distilled water (Fig. 3), cyanide removal did not follow a clear exponential decay (k ¼ 0.2462; r ¼ 0.6583), and the kinetic reaction order could not be accurately established (Table 1). Cyanide concentration experienced a fast and steep decrease during the first minute of the treatment, but then the concentration remained almost constant until t ¼ 15 min. From that point, the concentration slowly decreased and after 90 min treatment it was below detection limits (Fig. 5), which represents a removal efficiency greater than 92%. This is a low elimination rate if we compare it with the distilled water experiment in the same reactor (99% removal after 3 min in R2, see Fig. 3). However, the great amount of organic matter present in this wastewater must be taken into account to evaluate the results. Normally, AOPs are conceived as tertiary treatments, and they should treat wastewaters with lower organic loads (approximately, COD below 150 mg L1). In this experiment, we used R2 as a secondary treatment; but it can be assumed that its cyanide removal performance would improve working as a tertiary treatment for more diluted wastewaters. On the other hand, COD removal from wastewater in the present experiment was low, attaining a decrease of 10% after 90 min treatment. This slow abatement
Fig. 5 e Chromatograms (GC-NPD) of the industrial wastewater treated with R2 during 0 min, 2 min, 15 min and 90 min. Hydrogen cyanide peak is marked with an arrow. Initial concentration: 1 mg LL1 cyanide.
1706
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
could be related both to the high initial COD concentration and to the nature of that organic matter. Other authors have studied COD removal with a different plasma technique (non-thermal gliding arc discharge, a highly energetic treatment), finding removals of about 90% in 3 h for an initial COD concentration of 80e150 mg L1 (Ghezzar et al., 2009) and 90% in 10 min for an initial concentration of 960 mg L1 (Njoyim-Tamungang et al., 2011). In less energetic processes, like conventional AOPs, a wide variety of COD removal efficiencies has been reported. In a homogeneous photocatalysis with UV/H2O2/O2, a COD elimination of 42% was attained in 60 min (initial concentration 998 mg L1) (Dura´n et al., 2009). Monteagudo et al. (2004) confirmed that UV radiation plays an important role in COD elimination, obtaining a removal of 75% in 40 min with O3/UV and O3/ H2O2/UV systems in a wastewater with an initial COD of 1500 mg L1, whereas lower efficiencies were recorded in the absence of UV light. Mahmud et al. (2012) proved that the efficiency of a photo-Fenton process for COD elimination in a landfill leachate decreased as the initial COD concentration increased (3000e10,000 mg L1). This supports the low COD removal performance observed in the present work, since the treated wastewater was heavily loaded with organic compounds. This clear difference between cyanide and COD removal efficiency from industrial wastewater in the DBD reactor must be associated with the complexity of the substances. The cyanide ion is a simple molecule, which is completely transformed and therefore eliminated after oxidation or cleavage reactions.
4.
Conclusions
The DBD plasma reactor at atmospheric pressure represents a promising technique to remove cyanide from water. Moreover, it avoids the input and consumption of additional reagents, which is indeed cost-effective and allows water reuse. The combination of electrical discharges, formation of reactive species and generation of UV radiation could have lead to the cleavage or substitution of the stable triple bond C^N. However, the efficiency of this kind of plasma reactor declines when the treated solution has a great concentration of dissolved organic matter, which can have a shielding effect or compete for the plasma-generated oxidant species. Therefore, the use of a coaxial DBD plasma reactor as a tertiary treatment for diluted wastewaters or even drinking waters is suggested. The application of the described treatment technology to large wastewater volumes implies a delicate scale-up stage to design a new reactor with the adequate dimensions and electric configuration. Some steps have already been taken in this direction, and the authors are working on the construction of a DBD full-scale reactor. Currently, the main limitation of this technology is the consumption of gases; therefore, some alternatives like gas recycling or the use of cheaper gases (like O2 or O3) should be carefully considered.
Acknowledgements The authors are very grateful to the pharmaceutical industry LEBSA for providing the wastewater. We thank J. Carretero-
Ariza and Dr. J. Garcı´a (Universitat Polite`cnica de Catalunya) for COD analyses. This research was funded by the European Commission under the Seventh Framework Program for Research and Technological Development of the European Union (FP7-SME-2010-1 e Grant agreement no.: 2262033WATERPLASMA).
references
Abdelmalek, F., Ghezzar, M.R., Belhadj, M., Addou, A., Brisset, J.L., 2006. Bleaching and degradation of textile dyes by nonthermal plasma process at atmospheric pressure. Industrial & Engineering Chemistry Research 45, 23e29. Augugliaro, V., Blanco-Ga´lvez, J., Ca´ceres-Va´squez, J., Garcı´aLo´pez, E., Loddo, V., Lo´pez-Mun˜oz, M.J., Malato-Rodrı´guez, S., Marcı`, G., Palmisano, L., Schiavello, M., Soria-Ruiz, J., 1999. Photocatalytic oxidation of cyanide in aqueous TiO2 suspensions irradiated by sunlight in mild and strong oxidant conditions. Catalysis Today 54, 245e253. Boadas-Vaello, P., Jover, E., Llorens, J., Bayona, J.M., 2008. Determination of cyanide and volatile alkylnitriles in whole blood by headspace solid-phase microextraction and gas chromatography with nitrogen phosphorus detection. Journal of Chromatography B 870, 17e21. Dash, R.R., Gaur, A., Balomajumder, C., 2009. Cyanide in industrial wastewaters and its removal: a review on biotreatment. Journal of Hazardous Materials 163, 1e11. Directive 2000/60/EC of the European Parliament and of the Council. Official Journal of the European Communities L 327/1, 22 Dec 2000. http://eur-lex.europa.eu/LexUriServ/LexUriServ. do?uri¼CELEX:32000L0060:EN:NOT. (accessed July 2012). Directive 2008/105/EC of the European Parliament and of the Council. Official Journal of the European Union L 348/84, 24 Dec 2008. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do? uri¼CELEX:32008L0105:EN:NOT. (accessed July 2012). Dura´n, A., Monteagudo, J.M., Sanmartı´n, I., Garcı´a-Pen˜a, F., Coca, P., 2009. Treatment of IGCC power station effluents by physico-chemical and advanced oxidation processes. Journal of Environmental Management 90, 1370e1376. Ford, J., Hernandez, R., Zappi, M., 2006. Bench-scale evaluation of advanced oxidation processes for treatment of a cyanidecontaminated wastewater from an engine manufacturing facility. Environmental Progress 25, 32e38. Fortin, L., Kasireddy, V.K., Soucy, G., Bernier, J.-L., 2000. Novel reactor for cyanide solution treatment. Canadian Journal of Chemical Engineering 78, 643e649. ski, J., ny, D., Fleszyn Gajewski, J.B., Ma˛czka, T., Nowak-Woz niak, R., 2004. Some hydrocarbon derivative compounds Poz decomposition using the HV impulse discharges. Journal of Electrostatics 62, 269e275. Ghezzar, M.R., Abdelmalek, F., Belhadj, M., Benderdouche, N., Addou, A., 2009. Enhancement of the bleaching and degradation of textile wastewaters by gliding arc discharge plasma in the presence of TiO2 catalyst. Journal of Hazardous Materials 164, 1266e1274. Kepa, U., Stanczyk-Mazanek, E., Stepniak, L., 2008. The use of the advanced oxidation process in the ozone þ hydrogen peroxide system for the removal of cyanide from water. Desalination 223, 187e193. Kim, Y.-J., Qureshi, T.I., Min, K.-S., 2003. Application of advanced oxidation processes for the treatment of cyanide containing effluent. Environmental Technology 24, 1269e1276. Krause, H., Schweiger, B., Schuhmacher, J., Scholl, S., Steinfeld, U., 2009. Degradation of the endocrine disrupting chemicals (EDCs) carbamazepine, clofibric acid, and
w a t e r r e s e a r c h 4 7 ( 2 0 1 3 ) 1 7 0 1 e1 7 0 7
iopromide by corona discharge over water. Chemosphere 75, 163e168. Locke, B.R., Sato, M., Sunka, P., Hoffmann, M.R., Chang, J.-S., 2006. Electrohydraulic discharge and nonthermal plasma for water treatment. Industrial & Engineering Chemistry Research 45, 882e905. Lukes, P., Locke, B.R., 2005. Degradation of substituted phenols in a hybrid gaseliquid electrical discharge reactor. Industrial & Engineering Chemistry Research 44, 2921e2930. Magureanu, M., Piroi, D., Mandache, N.B., Parvulescu, V., 2008. Decomposition of methylene blue in water using a dielectric barrier discharge: optimization of the operating parameters. Journal of Applied Physics 104, 103306-1e103306-7. Mahmud, K., Hossain, M.D., Shams, S., 2012. Different treatment strategies for highly polluted landfill leachate in developing countries. Waste Management 32, 2096e2105. Monteagudo, J.M., Rodrı´guez, L., Villasen˜or, J., 2004. Advanced oxidation processes for destruction of cyanide from thermoelectric power station waste waters. Journal of Chemical Technology and Biotechnology 79, 117e125. Mudliar, R., Umare, S.S., Ramteke, D.S., Wate, S.R., 2009. Energy efficient e advanced oxidation process for treatment of cyanide containing automobile industry wastewater. Journal of Hazardous Materials 164, 1474e1479.
1707
Njoyim-Tamungang, E., Laminsi, S., Ghogomu, P., Njopwouo, D., Brisset, J.-L., 2011. Pollution control of surface waters by coupling gliding discharge treatment with incorporated oyster shell powder. Chemical Engineering Journal 173, 303e308. Royal Decree 60/2011, 22 Jan 2011. About the rules of environmental quality in the field of water policy (in Spanish). Boletı´n Oficial del Estado 19, 6854e6870. Sec. I. http://www. boe.es/boe/dias/2011/01/22/pdfs/BOE-A-2011-1139.pdf (accessed July 2012). Sano, A., Takimoto, N., Takitani, S., 1992. High-performance liquid chromatographic determination of cyanide in human red blood cells by pre-column fluorescence derivatization. Journal of Chromatography 582, 131e135. Soucy, G., Fortin, L., Kasireddy, V.K., Bernier, J.-L., Kimmerle, F.M., 2001. Innovative cyanide solution treatment by thermal plasma. In: Young, C.A., Tidwell, L.G., Anderson, C.G. (Eds.), Cyanide: Social, Industrial and Economic Aspects. TMS, Warrendale, pp. 379e384. Spinks, J.W.T., Woods, R.J., 1976. An Introduction to Radiation Chemistry. Wiley, New York. Vogna, D., Marotta, R., Andreozzi, R., Napolitano, A., d’Ischia, M., 2004. Kinetic and chemical assessment of the UV/H2O2 treatment of antiepileptic drug carbamazepine. Chemosphere 54, 497e505.