Efficient Process for Phenol Removal from Water by Generating Pulsed Positive Corona above the Surface

Key words: Modeling, Ozone Generation, Phenol Removal, Pulsed Positive Corona, Water Treatment

1. Introduction

High-energy electrons (~ 10 eV) form in the streamer channels of high voltage pulsed corona discharge (Veldhuizen, 1999). Such discharge channels with high electric fields (2*10⁷ V/m) at the streamer heads are propagating away from the powered anode. Very high voltages without breakdown between the electrodes are sustained by using short pulses (<1 ms). The energetic electrons and intense ultraviolet radiations associated with the pulsed discharge cause electrochemical reactions to occur, which are otherwise not possible under atmospheric pressures.

In air, the pulsed corona provides an efficient method for ozone generation and formation of chemically active species such as H, O, OH radicals and H₂O₂ molecules. When a layer of water is interspersed between the corona wire and the grounded electrode, ozone, OH and other oxidizing radicals formed in air diffuse in the water layer. These can efficiently remove water pollutants through oxidising reactions.

Typical lifetime of the reactive radicals is in the microsecond range, leaving primarily ozone to dominate the water cleaning process and, maybe, UV photons. For radicals to contribute to water cleaning, they should be generated close to the water layer or within the water itself. Since the pulsed corona discharge creates radicals and ozone in-situ it is expected to be superior to the conventional ozonation and it can be used in situations where biological processes fail.

2. Experimental Set-Up

A heterogeneous pulsed corona reactor with pulse energisation has been used to remove phenol from water (Grabowski et al, 2003). The reactor, shown in Fig. 1 has four parallel wire electrodes having length of 38 cm each.  High voltage pulses with 20 ns rise time are applied between the wires and a grounded electrode placed below the perpex reactor bottom. A thin layer of water having a thickness 8 mm is kept over the grounded electrode, this corresponds to a water volume of 250 ml.

Fig 1: Pulsed Corona Above Water reactor and power supply for water cleaning. Input impedance 50/4 ohm, output impedance 50x4 ohm

Phenol is added to water. Pulse energisation is accomplished by means of a transmission line transformer containing four stages. Pulse voltages up to 45 kV are applied to the reactor, which provides power levels over 2 MW. This power supply has a better impedance matching to the corona discharge used here than the spark gap switched capacitor used earlier (Hoeben et al, 2000). An additional advantage is that the voltage pulses have short duration (~100 ns). Ozone concentration in the air space above water layer is measured as a function of time using absorption spectroscopy.

3. Results and Discussion

3.1 Experimental

Fig 2: Temporal dependence of ozon concentration in an empty reactor, a reactor with plane water and a reactor with a 1mM phenol solution

It is seen from Fig. 3 that the ozone concentration rises to 4.5 x 10¹⁶ molecules per cm³ inside the air. Initially there is a sharp increase in the concentration followed by a saturating trend after nearly 5 minutes of pulsed corona operation. When a layer of pure water is present in the reactor first the ozone density in air is higher because the same amount of energy is dissipated in a smaller volume. Later a part of the ozone is taken up by the water. Water with phenol absorbs 25% more ozone than clean water as indicated in the lowest line of fig. 2. This indicates that the phenol really consumes the ozone thereby increasing the transport of ozone into the water. A decrease of the phenol concentration to 50% of its initial value of 1mM is observed in 30 minutes with a total power input of 2 10⁴ J/l. In an earlier experiment with a water layer of 5 cm thick 40% degradation was achieved at the same energy consumption but it took 180 minutes (Hoeben et al, 2000). This shows that the surface-to-volume ratio is important for the transfer of ozone into water.

3.2 Modeling

To understand and predict the electrochemical processes occurring under the corona condition it is modeled by taking into account the prominent reactions involving ozone. Generation and loss of ozone are also accounted for in the model. Three rate equations are formulated to describe the temporal evolution of ozone in air, ozone in water and phenol in water.

Fig 3: Simulink flow diagram for modeling phenol from water in pulsed corona reactor

The well-stirred reactor concept is adopted so constant concentrations are assumed with a negligibly thin transport layer between air and water, (Gurol and Singer, 1983). The model can be converted straightforward into a Matlab/Simulink block diagram, which makes it easy to solve the equations and change the parameter values. This block diagram is shown in Fig. 3. The rate constants for various reactions involved in the process are taken from (Gurol and Singer, 1983), but in this reference the production of ozone in air is not taken into account. The recombination constant for ozone in air and time constant obtained by putting the diffusion into water zero and fit the model curve to the measured ozone concentration in a reactor with only air. Figure 4 shows the result of the model when the parameter for diffusion of ozone is adapted to obtain 50% phenol degradation in 30 minutes. The model shows a lower ozone concentration in air as long as phenol is present in the water. This again shows the ozon consumption by phenol but the curve of fig. 4 is rather different from the measured curve of fig. 2.

Fig 4: Calculated concentrations of ozone and phenol (case [2])

The differences of phenol degradation as observed in (Grabowski, 2003) and (Hoeben et al, 2000) and mentioned in section 3.1 are also described by changing the parameter that governs the transport of ozon into water. The ratio of the water surface to its volume is 1000 cm²/l in the first case and 200 cm²/l in the second. This difference is roughly the same as the difference in treatment time. The value of the diffusion constant for ozon into water that is obtained when fitting these experimental data to the model is considerably lower than the value given in (Gurol and Singer, 1983).

4. Conclusion

A simplified model is developed to describe the behaviour of ozone in a corona reactor. The ozone is generated in air above a water surface. Recombination of ozone in air is a significant loss process. The ozone also diffuses into the water where it reacts with phenol. The consumption of ozone by phenol is confirmed by measurements of ozone concentration in air. It is shown that the ratio of surface to volume of the phenol solution is the governing parameter for the duration of phenol breakdown. So the diffusion of ozone in to the water is the rate limiting process.

This simplified model does give insight about the phenol breakdown; it does however not reproduce all the details of the ozone concentration in air as measured by absorption spectroscopy. The same is the case for the reactions in water. Other models are available which totally ignore ozone but explain the oxidation of phenol by OH radicals (Grymonpré, 2001).

Pulsed corona discharge is shown to be an efficient process for water cleaning. Though phenol removal from water is investigated in the present paper, the corona based process gives possibilities for cleaning different impurities from polluted water under diverse conditions.

Acknowledgements

One of the authors (Rani Devi) acknowledges financial support from the European Commission within the ytriD project (GRD1-2001-40374) [5].

References

***

Copyright © 2007, ECO Services International