NO Removal in Continuous BioDeNOx Reactors: Fe(II)EDTA 2 Regeneration, Biomass Growth, and EDTA Degradation Peter van der Maas, Paula van den Brink, Sudarno Utomo, Bram Klapwijk, Piet Lens Sub-Department of Environmental Technology, Wageningen University, PO Box 8129, 6700 EV Wageningen, The Netherlands; telephone: þ31 317 483851; fax: þ31 317 482108; e-mail: piet.lens@wur.nl Received 10 August 2005; accepted 28 December 2005 Published online 4 April 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bit.20859 Abstract: BioDeNOx is a novel technique for NOx removal from industrial flue gases. In principle, BioDe- NOx is based on NO absorption into an aqueous Fe(II)EDTA 2 solution combined with biological regenera- tion of that scrubber liquor in a bioreactor. The technical and economical feasibility of the BioDeNOx concept is strongly determined by high rate biological regeneration of the aqueous Fe(II)EDTA 2 scrubber liquor and by EDTA degradation. This investigation deals with the Fe(II)EDTA 2 regeneration capacity and EDTA degrada- tion in a lab-scale BioDeNOx reactor (10–20 mM Fe(II)EDTA 2 , pH 7.2 0.2, 558C), treating an artificial flue gas (1.5 m 3 /h) containing 60–155 ppm NO and 3.5–3.9% O 2 . The results obtained show a contradiction between the optimal redox state of the aqueous FeEDTA solution for NO absorption and the biological regeneration. A low redox potential (below 150 mV vs. Ag/AgCl) is needed to obtain a maximal NO removal efficiency from the gas phase via Fe(II)EDTA 2 absorption. Fe(III)EDTA reduc- tion was found to be too slow to keep all FeEDTA in the reduced state. Stimulation of Fe(III)EDTA reduction via periodical sulfide additions (2 mM spikes twice a week for the conditions applied in this study) was found to be necessary to regenerate the Fe(II)EDTA 2 scrubber liquor and to achieve stable operation at redox potentials below 150 mV (pH 7.2 0.2). However, redox potentials of below 200 mV should be avoided since sulfide accumu- lation is unwanted because it is toxic for NO reduction. Very low values for biomass growth rate and yield, respectively, 0.043/d and 0.009 mg protein per mg ethanol, were observed. This might be due to substrate limitations, that is the electron acceptors NO and pre- sumably polysulfide, or to physiological stress conditions induced by the EDTA rich medium or by radicals formed in the scrubber upon the oxidation of Fe(II)EDTA 2 by oxygen present in the flue gas. Radicals possibly also induce EDTA degradation, which occurs at a substantial rate: 2.1 (0.1) mM/d under the conditions investigated. ß 2006 Wiley Periodicals, Inc. Keywords: nitric oxide; flue gas treatment; denitrifica- tion; iron reduction; bio-regeneration; EDTA INTRODUCTION Industrial flue gases are a major source of NOx emission to the atmosphere (Schnelle and Brown, 2002). At present, selective catalytic reduction (SCR) is the most common technology applied for NOx abatement. High energy consumption and costs, as well as ammonia emissions are, however, major drawbacks of the SCR technology. Biologi- cal NOx removal techniques may become attractive alter- natives, but the poor solubility of NO in water results into relatively high residence times of at least 1 min (Jin et al., 2005), and, therefore, large reactor volumes. To enhance the NO transfer from the gas to the liquid phase, aqueous solutions of ferrous chelates, for example Fe(II)EDTA 2 , can be applied as scrubber liquor, because ferrous EDTA reactively absorbs NO according to: NOðaqÞþ FeðIIÞEDTA 2 $ FeðIIÞEDTA NO 2 ð1Þ The BioDeNOx process (Buisman et al., 1999) utilises this principle of wet absorption of NO in an aqueous Fe(II)EDTA 2 solution and combines it with biological reduction of the sorbed NO according to the overall reactions: 6FeðIIÞEDTA NO 2 þ C 2 H 5 OH ! 6 FeðIIÞEDTA 2 þ 3N 2 þ 2 CO 2 þ 3H 2 O ð2Þ Since flue gases generally contain oxygen, part of the Fe(II)EDTA 2 is oxidized to Fe(III)EDTA (reaction 3). Therefore, reduction of EDTA chelated Fe(III) (overall reaction 4) is, besides NO reduction, a core reaction of the biological regeneration pathway in the BioDeNOx process. 4 FeðIIÞEDTA 2 þ O 2 þ 4H þ ! 4 FeðIIIÞEDTA þ 2H 2 O ð3Þ 12 FeðIIIÞEDTA þ C 2 H 5 OH þ 3H 2 O ! 12 FeðIIÞEDTA 2 þ 2 CO 2 þ 12 H þ ð4Þ ß 2006 Wiley Periodicals, Inc. Correspondence to: Dr. Piet Lens Contract grant sponsors: Dutch Foundation for Applied Sciences (STW- NWO); Biostar Development CV Contract grant number: STW WMK 4963