Formation of nitrates in aqueous solutions treated with pulsed corona discharge: the impact of organic pollutants
Sergei Preis*, Iris Panorel, Sergi Llauger Coll, Iakov Kornev1
[email protected] Lappeenranta University of Technology, Skinnarilankatu 34 53850 Lappeenranta, Finland 1 Tomsk Polytechnic University, Tomsk, Russia
Abstract Pulsed corona discharge (PCD) in oxygen-nitrogen mixtures results in formation of nitrogen oxides, transformed to aqueous nitrates in contact with water. The experimental research into the impact of formate and oxalate to nitrate formation in aqueous solutions treated with PCD was undertaken. The impact of paracetamol, ibuprofen, indomethacin and their oxidation products to nitrate formation was also analyzed. Pharmaceuticals obstructed nitrate formation, while carboxylic anions and pharmaceuticals’ oxidation products noticeably improved nitrate formation in treated solutions as compared to water. The nitrate formation enhancement is explained by the aqueous ozone decomposition and hydroxyl radical formation known to be improved by carboxylic anions. Keywords: Ozone, advanced oxidation processes (AOP), anti-inflammatory drugs, hydroxyl radical, pharmaceuticals, water
1. Introduction One of the advanced technologies in water/wastewater treatment is the application of electric gas-phase discharges forming hydroxyl-radical (OH•), atomic oxygen (O), and also ozone (O3). Previous studies (Yavorovsky et al., 1997, Hoeben et al., 1999, Chauhan et al., 1999, Kornev et al. 2006) showed the intensification of gas-phase discharges at the gas-liquid interface, where the oxidants react with pollutants. Pulsed corona discharge (PCD) over the water surface (Aristova and Piskarev, 2002, Grabowski et al., 2006), and in water aerosol (Pokryvailo et al., 2006) showed the highest energy efficiency in oxidation of pollutants. The authors earlier proposed PCD applied to the treatment of water dispersed in gas in the form of droplets, jets and films sized up to a few millimetres forming a sufficient interface. Pulsed corona discharge showed the energy efficiency more than two times exceeding the one of traditional ozonation when applied to humic substances oxidation (Panorel, et al. 2011). Also, the advantage of PCD is relatively simple design of the equipment; PCD is insensitive towards the gas humidity; the reactor is a closed compartment with the gas transport necessary to only compensate oxygen losses in oxidation. 1
The use of air in PCD is often preferable for its simplicity. Electric discharge in oxygennitrogen mixtures, however, inevitably results in the formation of nitrogen oxides (Samoilovich et al., 1997, Fridman, 2008) forming aqueous nitrite- and nitrate-anions. Kornev et al. (2012) studied the regularities in the PCD treatment of aqueous solutions with different pH and electric conductivity and found that, unlike spark and dielectric barrier discharges, PCD resulted in formation of nitrates only. The impact of organic admixtures to the formation of nitrates in PCD treated aqueous solutions, however, remains unknown. This study is examining the formation of nitrates in a PCD-treated solutions of variable electric conductivity, pH and the content of slowly oxidized organic compounds such as antiinflammatory paracetamol, ibuprofen and indomethacin, the oxidation of which was described earlier (Panorel et al., 2013a, Panorel et al., 2013b), and also oxalate and formate as oxidation products of organic compounds. The possibility of forming nitro and nitroso derivatives of organic substrates reported, for example, by Yan et al. (2005) presents also a potentially important issue and deserves separate approach. 2. Materials and Methods Experimental device and materials The components of the device are a pulsed corona reactor, a high voltage pulse generator, and the reservoir containing 50 L of treated solution (Fig. 1). The pulse generator consists of a thyristor power switch circuit, followed by pulse step-up transformer and high-voltage magnetic compression stages, and a pulse compression block (Fig. 2). The reactor utilizes wire-plate corona geometry: horizontal electrode wires are concluded between vertical earthed plate electrodes. The electrodes geometry parameters determining the pulse characteristics were chosen for the maximum pulse energy as high voltage (HV) electrodes made of stainless steel wire of 0.5 mm diameter, positioned at 17 mm from the vertical grounded plate electrodes with the distance of 30 mm between the HV-electrodes. The total length of the HV-electrodes was 32 m in 0.5-m sections, i.e. 64 electrodes were positioned between two plates sized 0.5x2.0 m. The volume of the discharge zone of the reactor thus was 34 L. Water is fed to the top of reactor, where it is dispersed through a perforated plate producing jets, droplets and films. Water showers between electrodes to the zone of gas-phase PCD formation, where the treatment with oxidants takes place. The current and voltage waveforms (Fig. 3) were registered with Agilent 54622D Mixed Signal Oscilloscope using lowinductance resistive current sensor and high-voltage divider Tektronix P6015. The energy E released in the electrode systems was calculated by integrating the current and voltage oscillograms according to equation (1): 2
(1) where U is voltage, I is current and T is the duration of a pulse. The parameters of the discharge pulses are: voltage amplitude 18 to 20 kV, current amplitude 380 to 400 A at 100 ns pulse duration at its repetition frequency of 200 and 840 pulses per second (pps). The energy delivered to the reactor, calculated as an integral product of voltage and current peak areas, was 0.30 to 0.33 J per pulse. The energy consumption efficiency of the pulse generator was 67%. The power dissipated in the discharge at maximum pulse repetition frequency 840 pps was 250 W; 200 pps corresponds to 60 W of delivered power. Solution is pumped by the water pump (Oy Lohja AB), supplied with the frequency regulator Strömberg (SAMI Mini Star) controlling the pump engine rotation speed and thus the flow rate, to the top of reactor, where is dispersed with a perforated plate. Solution is showered between electrodes to the zone of gas-phase PCD formation, where the target compounds react with OH• radicals, ozone and other oxidants. The solution returns to the tank for recirculation. The flow rate was 15 L min-1 in all experiments. In the experiments with 50-L batches, the potable water was used with the initial conductivity from 20 to 200 µS cm-1 and TOC below 0.3 mg L-1. After dissolution of reactants and pH adjustment conductivity did not exceed 0.5 mS cm-1. The electric conductivity of treated solutions was adjusted to 10 mS cm-1 with sodium chloride. The 50-L batches were treated for 30 min at the pulse repetition frequency of 840 pps with sampling increment of 5 min, and for 2 h at 200 pps (20 min). The experiments were repeated twice with the deviations of the observed results from each other never exceeding 1%. Water
Perforated plate System of electrodes
Voltage pulse generator
Electric discharge reactor Sample feed port Input ports
Voltage and current monitoring
O2
N2
Storage tank Pump
Figure 1: Experimental setup outline 3
Gas cylinders
~ 380 V
VS1 Capacitor charger
C1
T1
Ls1 C2
Ls2 C3
Ls3 C4
Figure 2: The pulse generator circuit principle outline: С1 - С4 – storage capacitors, VS1 – thyristor switch, T1 – step-up transformer, Ls1 - Ls3 – saturable inductors Analytical methods The nitrate concentration was measured with the HACH photometer by brucine-sulphanilic acid method (Standard Methods, 1975). The nitrite formation was tested with the Griess reagent (Eaton et al., 1995). Total organic carbon (TOC) used in formate content analysis was analysed by Shimadzu 5050 TOC analyser. Electric conductivity of solutions was measured with conductivity meter HI 9033 (HANNA Instruments, USA). Gaseous ozone was determined iodometrically by bubbling 1 L of ozone-containing gas from the reactor tank through the Drexel bottle containing acidified 2% potassium iodide solution. Free iodine was titrated by sodium thiosulphate 0.1 N solution in presence of starch. Oxalate concentration was determined by titration with permanganate (Pohloudeck-Fabini and Beyrich, 1975). The application of HPLC to the analysis of pharmaceuticals was described in details by Panorel et al. (2013a and 2013b). Formic and oxalic acids were taken as supplied by Sigma-Aldrich with >99.9% purity. Brucine, sulphanilic acid, sodium hydroxide, sulphuric acid, potassium permanganate, starch, sodium thiosulphate (0.1 N fixanal), potassium iodide and sodium chloride were also taken as supplied by Sigma-Aldrich.
4
Voltage, kV 15 10 5 0 -5 0 .5
1 .0 Time, µs
1 .5
0 .5
1 .0 Time, µs
1 .5
Current, A 400
200
0
Figure 3: Voltage and current oscillograms of the pulse
3. Results 3.1 Formation of nitrates in absence of organic compounds The analysis of nitrite gave zero result in all experiments consistent with the previous observations (Kornev et al., 2012). Only nitrates were identified indicating oxidants sufficient to transform nitrite to nitrate.
Effect of frequency The treatment of low (
