Electrochemical sulfide removal by low-cost electrode materials in anaerobic digestion

Accepted Manuscript Electrochemical Sulfide Removal by Low-cost Electrode Materials in Anaerobic Digestion Hongjian Lin, Nicholas Williams, Amelia King, Bo Hu PII: DOI: Reference:

S1385-8947(16)30338-2 http://dx.doi.org/10.1016/j.cej.2016.03.086 CEJ 14937

To appear in:

Chemical Engineering Journal

Received Date: Revised Date: Accepted Date:

13 January 2016 14 March 2016 19 March 2016

Please cite this article as: H. Lin, N. Williams, A. King, B. Hu, Electrochemical Sulfide Removal by Low-cost Electrode Materials in Anaerobic Digestion, Chemical Engineering Journal (2016), doi: http://dx.doi.org/10.1016/ j.cej.2016.03.086

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title: Electrochemical Sulfide Removal by Low-cost Electrode Materials in Anaerobic Digestion Authors: Hongjian Lin, Nicholas Williams, Amelia King, and Bo Hu* Address: Department of Bioproducts and Biosystems Engineering, University of Minnesota, 1390 Eckles Ave., St. Paul, MN 55108, USA *

Corresponding author

Ph.D., Associate Professor Tel: 612-625-4215 (O) Fax: 612-624-3005 Email: [email protected]

Abstract Sulfide cleaning is usually a necessary but costly step following the anaerobic digestion (AD) in order to upgrade biogas quality. Mitigating biogas and digestate sulfide level simultaneously in AD process will save capital cost by eliminating standalone biogas upgrading facility. However, current in-situ remediation methods, for instance, dosing of magnesium/sodium hydroxide, oxygen gas, iron salts, nitrite or nitrate, usually cause the potential interference to biogas production and may intensively consume energy and chemicals. Here, an electrochemical remediation method was studied to use low-price electrode materials of carbon cloth and stainless steel AISI 304. These two electrode materials at 3 V showed complete removal of sulfide in 2 days in synthetic 10 mM sulfide solution. Operating conditions for carbon cloth and stainless steel electrodes in bench-scale digester fed with dairy manure were optimized to improve sulfide removal efficiency and to improve biogas production. Based on the results, it was concluded that carbon cloth at 2 V and 3 V, and stainless steel anodes at both 1 and 2 V have a potential of significantly removing hydrogen sulfide from biogas under continuous operation with sufficient electrode surface area. Intermittent 3 V voltage application (fifteen minutes per day) of stainless steel can remove most biogas hydrogen sulfide. The electrochemical sulfide oxidation and removal showed no/little negative effect on biogas and methane production, and therefore can be a promising technology for the AD industry to develop a cost-effective approach to produce sulfide free biogas.

Keywords: sulfide removal; electrochemical oxidation; carbon cloth; stainless steel; biogas cleaning

1 / 42

1. Introduction Anaerobic digestion (AD) is a commercialized technology to degrade organic wastes (e.g., animal manure, primary and activated sludge from wastewater treatment plant, organic fraction of municipal solid wastes, various types of food processing wastewater like vinasse and stillage) and generates biogas which mainly consists of methane and carbon dioxide. Sulfate (SO42-) is a common component for a wide range of substrates, resulted from protein hydrolysis and mineralization as well as pH adjustment using sulfuric acid during biological/chemical processing. Sulfide (S2-) is generated from sulfate in AD by sulfate-reducing bacteria in anaerobic respiration [1], and from sulfurcontaining amino acids (e.g., cysteine and methionine) and their intermediates by enzymatic degradation [2]. Hydrogen sulfide (H2S) is free to diffuse through cell membrane, and therefore a high concentration of sulfide tends to denature proteins inside the cytoplasm of methanogens by cross-linking polypeptide chains; in consequence, accumulated levels of sulfide/hydrogen sulfide may impose inhibition to methanogenesis for biogas generation, which has been frequently reported in literature [3]. The volatilized hydrogen sulfide in biogas or in ambient air causes irritating “rotten egg” odor, corrodes facility members, and poses health and safety threats to exposed humans and animals. It is a major nuisance in many biological processes where sulfate-reduction occurs. During the utilization of biogas in internal combustion engines (e.g., gas turbines), storage tanks, boilers, and fuel cells, biogas will need to be upgraded to remove hydrogen sulfide, in order to reduce the risk of equipment corrosion and toxicity to catalysts [4]. Therefore, controlling sulfide level in AD medium is beneficial to biogas production, maintenance of reactors and facilities, and biogas upgrade. With a well-controlled sulfide level, the

2 / 42

inhibition potentially caused by sulfide/hydrogen sulfide can be remediated, the threats of facilities corrosion can be decreased, and the subsequent requirement on biogas purification in terms of hydrogen sulfide removal may not be necessary and thus it reduces the overall cost of biogas production and application.

Most previous research focus more on biogas cleaning technology, which is to remove H2S from gas stream in a separate operation unit after AD [4]. For example, adsorption using iron oxide has long been used in industry. Biological processes, e.g., biofilters, biotrickling filters and bioscrubbers (fixed-film and suspended growth), are widely tested for biogas hydrogen sulfide removal [5, 6], in which defined sulfur-oxidation microorganisms grow in liquid phase or on packing media, and oxidize and remove sulfide from the inlet biogas. There are also various methods proposed for in situ sulfide/hydrogen sulfide removal from liquid media like sewer systems and digestate. Those methods include gas stripping using air or oxygen aeration to partially oxidize sulfide to elemental sulfur [7, 8], to dose nitrate and nitrite to couple sulfide oxidation for its removal. Iron salts dosing is another mitigation method that can scavenge sulfide to precipitate insoluble sulfide minerals [9, 10]. Intensive chemical consumption and cost, high energy input, and frequent waste disposal limit the application of those in situ methods. It will further be a concern if those methods are introduced to anaerobic digesters which have to produce biogas without compromise.

Electrochemical (anode oxidation) methods are recently introduced as an alternative method for sulfide control in in geothermal brines, sewage, and caustic solutions [11].

3 / 42

Sulfide species are electrochemically active, and the oxidation decreases sulfide species concentration by forming a wide spectrum of intermediates and final products, including disulfide (S22-), polysulfide (Sn2-), sulfite (SO32-), thiosulfate (S2O32-), and sulfate (SO42-) (Reactions 1-10, [12, 13]). For example, given the gold electrode tested in cyclic voltammetry [14], the oxidation of sulfide first forms mono-layer of elemental sulfur at low potential (-0.4 V vs. SHE) and then proceeds with the formation of multi-layer at higher potential (0.05 V vs. SHE). Polysulfide of mixed lengths is generated in the oxidation of sulfide as well as reduction of elemental sulfur during the reduction scan. Sulfide in geothermal brines was oxidized to elemental sulfur on graphite disk anode with controlled anode potential of 450 mV vs. saturated Ag/AgCl reference electrode [15]. It was reported that at this relatively low electrode potential, the deposited elemental sulfur then passivated the electrode surface and gradually decreased the rate of the sulfide oxidation. In fuel cell mode using potassium ferricyanide as electron acceptor, graphite granule electrodes removed sulfide in aqueous solution at a rate of 280 g-S/m3/day [16], while simultaneously producing electrical power with the maximum level of 166 W/m3. Sulfur was the predominant oxidation product, so the performance declined after 3-month operation as a result of excessive sulfur deposition on graphite granule anode [16].  ,ĂůĨƌĞĂĐƚŝŽŶƐ ^;ƐͿнϮĞͲнϮ,нͲͲх,Ϯ^ ^;ƐͿнϮĞͲн,нͲͲх,^Ͳ ^;ƐͿнϮĞͲͲͲх^ϮͲ ^KϰϮͲнϴĞͲнϵ,нͲͲх,^Ͳнϰ,ϮK ^KϰϮͲнϴĞͲнϭϬ,нͲͲх,Ϯ^нϰ,ϮK ^KϰϮͲнϲĞͲнϴ,нͲͲх^;ƐͿнϰ,ϮK ,Ϯ^KϯнϰĞͲнϰ,нͲͲх^;ƐͿнϯ,ϮK ^ϮKϯϮͲнϴĞͲнϴ,нͲͲхϮ^;ƐͿнϯ,ϮK Ŷ^;ƐͿнϮĞͲͲͲх^ŶϮͲ

4 / 42

ŽΖ sǀƐ͘^, ͲϬ͘ϮϳϮ ͲϬ͘Ϯϳϭ ͲϬ͘ϰϳϲ ͲϬ͘Ϯϭϯ ͲϬ͘Ϯϭϰ ͲϬ͘ϭϵϰ нϬ͘Ϭϯϲ ͲϬ͘Ϯϭϯ ͲϬ͘ϯϭϵ;ŶсϱͿ

ŽΖ sǀƐ͘ŐͬŐů ͲϬ͘ϰϳϵ ͲϬ͘ϰϳϴ ͲϬ͘ϲϴϯ ͲϬ͘ϰϮϬ ͲϬ͘ϰϮϭ ͲϬ͘ϰϬϭ ͲϬ͘ϭϳϭ ͲϬ͘ϰϮϬ ͲϬ͘ϱϮϲ;ŶсϱͿ

  ZĞĂĐƚŝŽŶϭ ZĞĂĐƚŝŽŶϮ ZĞĂĐƚŝŽŶϯ ZĞĂĐƚŝŽŶϰ ZĞĂĐƚŝŽŶϱ ZĞĂĐƚŝŽŶϲ ZĞĂĐƚŝŽŶϳ ZĞĂĐƚŝŽŶϴ ZĞĂĐƚŝŽŶϵ

^ŶϮͲн;ϮͲϮŶͿĞͲнŶ,нͲͲхŶ,^Ͳ

ͲϬ͘Ϯϱϱ;ŶсϱͿ

ͲϬ͘ϰϲϮ;ŶсϱͿ

ZĞĂĐƚŝŽŶϭϬ

To overcome the anode passivation by elemental sulfur, another study utilized the water electrolysis products of oxygen gas in anaerobic digester [17], with the anode made from titanium mesh with Ir-mixed metals oxides (MMO) coating and the cathode from stainless steel mesh. When the applied voltage was 3.5 V (35 mA for 3.5 L reactor), it turned out that biogas hydrogen sulfide concentration was reduced from between 200 and 3000 ppm to less than the detection limit of the method, and importantly, the methane production was improved by 10% to 25%. The resulting biogas contained an appreciable amount of hydrogen gas but almost all oxygen gas was biologically consumed. Their results also showed the possibility of sulfide oxidation in single-chamber reactor configuration, which is a simpler design and has smaller internal resistance compared with two-chamber configuration that has a separator or membrane between electrodes to avoid mixture of reactants and products. An issue in scale-up of this technology is that the cost of MMO electrode material is high because of the use of platinum-group noble metals like Ru, Ir and Pt oxides [18, 19]. The electrode configuration and implementation for well-distributed oxygen generation will be hindered by their limited availability and high price.

Less expensive electrode materials shall be offered for sulfide removal, and graphitic carbon cloth (CC) and stainless steels (AISI 304, 316, and 430) were identified for the current study. The main reasons for selecting these materials are as follows: first, both materials are potentially electrochemically active for oxygen evolution reaction [20, 21], which may also display sulfide oxidation performance and be manipulated for indirect

5 / 42

sulfide oxidation; second, both materials offer a flat-sheet surface structure, which may increase relative larger electrode surface area compared to graphite rods, discs and granules, and facilitate the potential recovery and cleaning of electrode surface with elemental sulfur or other deposits without impairing surface structure; third, both materials are in abundant supply and prices are relatively lower compared to MMO electrodes.

However, it is not yet known whether these materials would directly and selectively oxide sulfide and reduce back the oxidation products to sulfide again, especially in the circumstance of abundant presence of organic matter which is common in sulfidecontaining aqueous media of anaerobic digestate. It is especially unclear if the oxidation process with those electrode would interfere with biogas production in anaerobic digestion. The goal of this study therefore was to search for suitable electrochemical conditions for these low-price electrode materials (carbon cloth and stainless steel) that would remove sulfide/hydrogen sulfide from digestate/biogas but without compromise on biogas and methane production. To reach the goal, this study first tested sulfide removal rate from buffered synthetic sulfide media. The study then evaluated the effects of major electrochemical intermediates, including hydrogen and oxygen gas, and ferrous and ferric, on hydrogen sulfide removal and biogas production in liquid dairy manure-fed benchscale digesters in batch mode. When sulfide removal proved effective and methane production not largely affected, the study proceeded with inclusion of electrochemical oxidation in AD at different operating conditions for their performance in methane production and sulfide suppression from liquid dairy manure.

6 / 42

2. Materials and Methods 2.1. Chemicals and Substrate Sodium sulfide solution was prepared from sodium sulfide nonahydrate (Na2S·9H2O) solids of ACS reagent grade (Sigma-Aldrich, St. Louis, MO), and buffered at pH 7.0 by 100 mM phosphate buffer solution (PBS). Ferrous and ferric chloride were prepared from the solids of ACS reagent grade chemicals (Fisher Scientific, Pittsburgh, PA). Hydrogen gas and oxygen gas were produced by water electrolysis at 30 V using titanium wire as electrodes, and instilled via disposable syringe to headspace of serum bottle for gas addition. Manure samples as AD substrate were collected from the dairy barn in the Dairy Cattle Teaching & Research Facility at University of Minnesota, St Paul, MN. The manure sample was mixed with tap water at a manure/water ratio of 1:2, followed by 5 min manual homogenization. It was sequentially filtered through meshes with open-pore sizes of 2 mm and 0.295 mm. Inoculum anaerobic sludge was collected from a continuously maintained anaerobic digester in laboratory fed with liquid dairy manure.

2.2. Electrode materials and preparation Electrodes materials used in this study were plain carbon cloth (Fuel Cell Earth LLC, Stoneham, MA) and stainless steel meshes (AISI 304, 316 and 430 ) with a wire diameter of 0.8 mm (mesh #4, McMaster-Carr, Elmhurst, IL). Carbon cloth electrodes were 4.5 cm in length and 1.0 cm in width, and the effective surface area was 4.0×1.0 = 4 cm2. Stainless steel mesh electrodes were 4.5 cm in length and 0.7 cm in width with an projected surface area was 4.0×0.7 = 2.8 cm2. Electrodes were glued to titanium wire of

7 / 42

30 gauge (0.255-mm in diameter, Esslinger & Co., Saint Paul, MN) via conductive silver epoxy (Fisher Scientific, Pittsburgh, PA).

2.3. Reactor set-up and operation Serum bottles of 160 mL with a working volume of 100 mL, capped with viton stops and sealed with an aluminum cap, were used as reactor vessels. Titanium wire penetrated the viton stops for electrode insertion and the penetrated sites were sealed with marine glue. Cautions should be taken to not short circuit the electrodes, and to flush oxygen out from the bottles by nitrogen gas. Three electrode configurations inside the reactors were studied, including immersed (whole pieces of electrodes immersed in media), halfimmersed (half piece of electrode immersed in media), and suspended (electrodes suspended in headspace without contact with liquid). Voltage was applied by connecting the wire to power supplies at the pre-determined voltage levels of 1 V, 2 V, 3 V, or control without voltage. Serum bottles were controlled at two temperature levels of 25 oC and 35 oC in water baths.

For AD, the same types of bottles were used but fed with liquid dairy manure, and inoculated with anaerobic sludge at a volume ratio of 6:4. The headspace of each serum bottle was flushed with nitrogen gas for half a minute to remove remaining oxygen gas. For evaluating the effect of electrochemical intermediates, designated amount of ferrous chloride or ferric chloride solution was added to serum bottles; and designated volume of hydrogen or oxygen gas was added to headspace daily. Assuming 100% coubombic efficiency, the amounts of intermediates additions corresponded to a continuous current

8 / 42

level of up to 0.009 mA for ferrous and 0.013 mA for ferric, and 0.20 to 0.80 mA for gas. The serum bottle AD reactors were then incubated at either 25 oC of room temperature or at 35 oC in temperature controlled incubator with reciprocal shaker at a speed of 150 rpm.

For electrochemically assisted reactors, electrodes were inserted to serum bottles in a similar fashion as aforementioned. Reactors were incubated at either 25 oC of room temperature or at 35 oC in temperature controlled incubator with daily manure mixing. Electrical current was measured with multi-meter at 8 h intervals in the first 24 h. The produced biogas volumes were measured on the designated sampling dates by water displacement of hydrochloric acid solution of pH 2 in a calibrated glass cylinder manometer. Experimental duration was 7 days for continuous mode of voltage application, and 35 days for intermittent voltage application.

2.4. Gas composition analysis Biogas composition of methane, carbon dioxide, oxygen gas, hydrogen gas, and nitrogen gas was analyzed using a gas chromatograph (CP-4900 Micro-GC, Varian Inc., Palo Alto, CA) equipped with two columns of molecular sieve 5A and Porapak Q. Helium gas was used as the carrier gas. The temperatures for the sampling line, the injector, and columns were set at 50, 110, and 80 °C, respectively. Gas components were determined by a thermal conductivity detector (TCD) equipped to the micro-GC. Volumetric levels of hydrogen sulfide in biogas were determined using RAE Systems hydrogen sulfide tubes of varying ranges based on the real concentration. Dilution of biogas with air was necessary in some samples because of the high level of hydrogen sulfide. Cautions must

9 / 42

be taken during biogas sampling and analysis, which must be done in a ventilated environment or in fume hood to avoid its over-exposure.

2.5. Liquid analysis Liquid samples were collected from serum bottles after well mixing. Each digestate sample of 10 mL was centrifuged at 3000 × G for 15 min before the supernatant was used for sulfate, reactive phosphorus and iron analysis by colorimetric methods using commercial testing kits (TNTplus™, Hach Company, Loveland, CO) with a UV-Vis spectrophotometer (Hach DR 5000) following the protocol for each assay. Before sulfate analysis, high purity chitosan powder of 0.3 g was added to samples to flocculate remaining solids for lower turbidity. Sulfide concentration was determined by ionselective glass electrode (Cole Parmer, Chicago, IL) according to Standard Methods for the Examination of Water and Wastewater [22].

2.6. Electrochemical characterization Cyclic voltammetric measurements were carried out using a Reference 600 potentiostat (Gamry Instruments, Warminster, PA) in a three electrode configuration, consisting of a platinum wire counter electrode and Ag/AgCl (3 M NaCl) reference electrode (0.207 V versus a standard hydrogen electrode, SHE). The projected surface area of working electrodes subjected to test was 2.32 cm2. Tested electrode materials included carbon cloth, stainless steel AISI 304, 316, and 430. Before test, stainless steels were polished for its surface with sandpaper of 400 grit (22 µm) silicon carbide. The media for test was 100 mM phosphate buffer solution (pH = 7.2; oxygen gas stripped) with sodium sulfide

10 / 42

concentration of 0 and 5 mM, respectively. The scanning rate was between 10 mV/s to 2000 mV/s but only the results of 50 mV/s scan were given.

Electrochemical impedance spectroscopy (EIS) test was performed in liquid dairy manure using Reference 600 potentiostat to quantify the resistance of the whole electrolytic cells. The test was done in two-electrode configuration, with anode as the working electrode and cathode as the reference and counter electrodes [23]. The frequency for EIS analysis ranged from 1 MHz to 1 Hz at 20 points/decade. The AC amplitude was 1 mV.

3. Results and Discussion 3.1. Electrochemical removal of H2S in synthetic media At the applied voltage of 1 V on carbon cloth electrodes, the headspace hydrogen sulfide level continuously dropped in both 2 mM (64 mg-S2-/L) and 10 mM (320 mg-S2-/L) sulfide solutions (Figure 1 A and B). The headspace sulfide level was found to slightly increase in the first 24 h in the control groups, suggesting the equilibrating process of hydrogen sulfide from the dissolved phase in water to the volatilized phase in headspace. When electrodes were placed in headspace without contact with liquid medium, the removal was not distinguishable from the control, in which condition the sulfide oxidation was probably initiated by the remaining oxygen in either water or headspace to polysulfide. Autoxidation of aqueous sulfide was observed to be catalyzed by the intermediate of disulfide followed by chain growth to form polysulfide and then to polythiosulfite and gradually to thiosulfate and sulfite/sulfate [24]. The similar trends of sulfide removal between the configuration with electrodes in headspace and the control

11 / 42

indicated that the adsorption of hydrogen sulfide to carbon cloth did not cause an observable change of concentration, which distinguished the sorption property of the graphitic carbon fiber used here from that of the activated carbon [25] and its fiber [26]. Hydrogen sulfide was depleted by half-immersed and immersed electrodes after 3 or 4 days when the initial Na2S concentration was 2 mM. When the initial Na2S concentration increased to 10 mM, H2S was not depleted in 7 days but suppressed to between 2000 and 2500 ppm, which was substantially lower than that of the control (7000 ppm).

Figure 1.

The above results show that the anode oxidation of sulfide (or hydrogen sulfide electrolysis) on carbon cloth electrode at a voltage smaller than that of water electrolysis (1.23 V) was feasible. This oxidation was theoretically predicted feasible by the potential differences between redox couples of either the half reaction of sulfide/bisulfide oxidation to elemental sulfur (-271 mV vs. SHE, at pH=7 and 1 M) or even to sulfate (213 mV; but not like to occur due to kinetics limitation), and the other half reaction of water reduction to hydrogen gas (-414 mV vs. SHE at pH 7), which have relatively small potential differences of 0.143 V and 0.201 V, respectively. The cyclic voltammetry results showed the distinct voltammograms of carbon cloth electrode in PBS media with and without sulfide presence (Figure 2 A), demonstrating the extra oxidation peaks associated with sulfide oxidation at around -60 mV vs. Ag/AgCl (Ia1, or +147 mV vs. SHE; absent in the first scan of sulfide solution) and at around +400 mV (Ia2, or +607 mV vs. SHE). Since the peak Ia1 only started to appear after the first scan, it suggested to be a result of polysulfide oxidation. The peak Ia2 was more likely a result of combined sulfide 12 / 42

and polysulfide oxidation to elemental sulfur [14]. The current decrease from the first to the third scan tended more to be a result of sulfide concentration decrease at boundary rather than being deactivated by sulfur deposition, because the replenishment of sulfide solution completely recovered the current at the consequent scan. The further current increase from anodic sweep after Ia2 suggested the oxygen chemisorption and evolution reactions. The cathodic scan, both with and without sulfide presence, showed a reduction peak of Ic1 at -650 mV, and therefore, indicated the hydrogen adsorption followed by hydrogen evolution reaction. The lack of additional reduction peak indicated the irreversibility of sulfide oxidation at the tested potential range when carbon cloth was used as electrode. In fact, sulfide oxidation on graphite disk electrode was observed in an earlier study [15], and later this phenomenon was used to mitigate sulfide pollution in brine by using carbon felt [13] with further kinetics elucidation for graphite rod [27]. Catalysis to sulfide oxidation by carbon materials can be irreversible as reported in glassy carbon (or vitreous carbon) electrode [28]. Therefore, as reported in our results, it was feasible to effectively oxidize sulfide using graphitic carbon cloth in single chamber configuration without using platinum counter electrode or selective ion exchange membrane between electrodes. The use of the less expensive material of carbon cloth at low voltage application is promising because it may potentially increase coulombic efficiency and minimize intermediates generation (e.g., oxygen gas generation in mixed metal oxides electrodes) during sulfide removal.

Voltammograms of the three types of stainless steel (AISI 304, 316, and 430) displayed common anodic and cathodic peaks as shown Figure 2 B-D. The presence of sulfide did

13 / 42

not insert any additional peak in the first scan compared to the voltammograms in PBS solution, but showed the same broad anodic peaks of Ia3 (between -230 and -210 mV) and Ia5 (>+500 mV), and cathodic peaks Ic2 of (+100 mV) and Ic3 (-380 mV). The broad anodic peak of Ia3 suggested the oxidation of iron to form passive film [29], and Ia5 indicated the oxidation of metal oxide to higher oxidation valence states, followed by oxygen evolution reaction thereafter. The reduction peak of Ic3 was suggested to relate to the reductive dissolution of hydrated ferric oxide to ferrous, and the peak Ic2 was attributed to the reduction of chromium oxide from six to three valence state based on a test for AISI 316 [29, 30]. Interestingly, the following scan on stainless steel, represented by the third scan, all showed an additional peak of Ia4 at -60 mV. After the first anodic scan, stainless steel surface was passivated with metal oxide which may display catalytic effect for sulfide oxidation to elemental sulfur [31], although the effect may be smaller than that of carbon cloth shown in the test. These results suggested that the direct anodic sulfide oxidation will be likely to occur for the three types of the tested stainless steel materials, and the indirect oxidation by reactive oxygen species and oxygen gas, and chemical binding to dissolved metal ions [32] can assist in scavenging sulfide from media. Based on these results, one of the stainless steel materials, AISI 304, cheaper than AISI 316 and with smaller tendency to corrosion than AISI 430, was chosen to use in the further experiment.

Figure 2.

When the applied voltage on carbon cloth was increased to 2 and 3 V, the headspace hydrogen sulfide expectedly decreased at a higher rate (Figure 3A and B), partially 14 / 42

because of sulfide being reacted at a faster rate caused by the increased driving force (overpotential) for oxidation which was indicated by higher anode potentials (Table 1), and partially because of intermediates of oxygen gas generation which assisted in sulfide oxidation. Since the tested pH value of the media was relatively stable during the operation in 0.1 M PBS buffer, the decrease of hydrogen sulfide in headspace indicated the removal of sulfide species in media rather than the redistribution (dissociation/association) between different sulfide species. When the voltage increased from 1 V to 2 V and 3 V at 25 oC, the calculated sulfide removal rate by carbon cloth was boosted from 53 g-S/d/m2-anode to 127 and >203 g-S/d/m2-anode in 2 mM Na2S, and from 177 g-S/d/m2-anode to 203 and 719 g-S/d/m2-anode in 10 mM Na2S (Table 2). For 3 V voltage application, hydrogen sulfide was not detectable after 0.65 day or 2 day depending on the initial concentration of sodium sulfide solution. Starting at 2 V, carbon cloth anode became dimensionally instable in sulfide solution and part of the graphite fiber consisting the carbon cloth disintegrated into the solution. This phenomenon would substantially affect the service life of the anode material and required the anode replacement over time. Meanwhile, water electrolysis produced appreciable amounts of oxygen and hydrogen gas when voltage was 2 V, and a substantial amount of gases when voltage was 3 V. Cyclic voltammetry results also suggested the occurrence of oxygen evolution reaction and the production of oxygen at the higher voltage may enhance the autoxidation of aqueous sulfide, which is a phenomenon of sulfide oxidation catalyzed by the intermediate of disulfide followed by chain growth to form polysulfide and then to polythiosulfite and gradually to thiosulfate and sulfite/sulfate [24].

15 / 42

Figure 3.

Table 1.

Table 2.

With carbon cloth electrodes treatment in 10 mM sulfide solution, the hydrogen sulfide level was observed to be much lower when temperature elevated from 25 to 35 oC at 2 and 3 V (Figure 3 C and D) and the calculated removal rate was larger at the higher temperature (Table 2), possibly resulted from an increased oxidation rate of sulfide. To the contrary, at 1 V, higher temperature condition (35 oC) yielded a final headspace hydrogen sulfide level comparable to that of the lower temperature (25 oC). The reason may be that despite a higher reaction rate at higher temperature, a higher temperature also gave rise to the Henry’s law constant of hydrogen sulfide thereafter its concentration in headspace as well [33]. Stainless steel electrode behaved similar to carbon cloth in headspace hydrogen sulfide removal at 3 V, but was less effective than carbon cloth at 2 V or lower (Figure 3 E and F). However, due to the small solubility product constants of iron phosphates (e.g., vivianite, Ksp = 10-36, and strengite, Ksp =10-26.4, retrieved from Mineql+ thermodynamic database), the competition for iron ions between sulfide and phosphate in 0.1 M PBS solution must occur and therefore reduce the sulfide removal efficiency. Meanwhile, temperature seemed to have played a less important role in the electrochemical oxidation rate of sulfide by stainless steel electrodes, and therefore headspace hydrogen sulfide was dominated by higher volatilization at higher temperature at lower oxidation rate at 1 and 2 V (Figure 3 E and F).

16 / 42

Figure 4.

The liquid analysis showed that after electrochemical oxidation, the sulfide concentration decreased and the sulfate concentration generally increased (Figure 4). In the media with the original 2 mM Na2S solution, 2 V voltage treatment increased the sulfate concentration from 0 to 0.71 and 0.57 mM, and 3 V treatment increased the concentration to 0.88 and 1.03 mM, at 25 oC and 35 oC, respectively. The sulfide level in the media decreased to a level less than the detection limit of the method (0.5 µM or 16 µg/L) except in the control reactors without voltage application. For the 10 mM sulfide media, the remaining sulfide level after about 8-day treatment was again substantially lower than that of the control reactors, and 3 V voltage completely removed sulfide at both electrode materials and temperatures. However, the sum of sulfide and sulfate in all reactors were not balanced with the original sulfide concentration, indicating the accumulation of a wide spectrum of sulfur intermediates and products during sulfide oxidation, including disulfide, polysulfide, and elemental sulfur at lower anode potential [34], and further to polythiosulfite, sulfite, and thiosulfate [24], and the remaining portion of hydrogen sulfide volatilized to headspace or removed by sampling for analysis. Accumulation of these intermediates and products were also widely reported in natural and artificial water samples [35, 36]. Nevertheless the complexity in oxidation products, conversion of sulfide to all of the other sulfur species is helpful in reducing hydrogen sulfide concentration in gas layer. Stainless steel electrode behaved similar to carbon cloth in sulfide removal when applied voltage was 3 V, but was less effective than carbon cloth at 1 or 2 V (Figure 4C). The sulfate concentrations with stainless steel treatments were correspondingly smaller than that with carbon cloth treatments of the same voltage level. 17 / 42

Since iron dissolution from anode was expected to occur, binding of ferrous with sulfide made up another important sulfide sink; however, the quantitative results were not obtained in this study.

Previous studies on hydrogen sulfide concentration in swine or dairy manure in anaerobic digestion usually observe peak hydrogen sulfide content in biogas/headspace gas within 5000 ppm (or 0.5% of gas volume) at the first few days of digestion [37, 38], which corresponds to a total sulfide content in digestate up to 185 mg-S/L (or 5.78 mM of sulfide) assuming pH = 8 and temperature of 35 oC, and dimensionless Henry’s law constant of 0.4143. Based on the observed results of sulfide removal rate in our study, it is possible to completely remove this sulfide level in one day by using between 2.3 cm2 and 42 cm2 of carbon cloth electrode per liter digestate, depending on the levels of applied voltage.

3.2. Effects of electrochemical intermediates addition on AD 3.2.1. H2S removal with electrochemical intermediates addition The direct anode oxidation of sulfide in media has a small tendency to interfere with AD; however, electrochemical intermediates (ferrous, ferric, hydrogen and oxygen gas) may be generated during electrolysis and may play a role in the indirect sulfide oxidation and removal, especially at a voltage overcoming water electrolysis requirement of 1.23 V. The results of hydrogen sulfide level with these intermediates additions during batch AD are shown in Figure 5. Ferrous (Fe2+) can be a product of anode oxidation from iron electrodes, and may be further oxidized to ferric (Fe3+) given oxygen gas generation.

18 / 42

Both ferrous and ferric (between 0.5 mM and 2 mM) consistently mitigated hydrogen sulfide level in biogas (Figure 5 A and B). During 25 days’ digestion, hydrogen sulfide concentration in the control reactor reached 1543 ppm on average. By adding 0.5 mM, 1 mM, and 2 mM of ferrous, hydrogen sulfide was decreased to 1271, 1184, and 796 ppm (up to 48% removal). The addition of corresponding level of ferric yielded hydrogen sulfide level of 1395, 1033, and 508 ppm (up to 67% removal). Ferric performed slightly better than ferrous, because ferric also exerted an additional oxidation capacity to sulfide before precipitating sulfide from liquid [39]. Water electrolysis products, i.e., hydrogen and oxygen gas, also showed impact on hydrogen sulfide level (Figure 5C and D). While the effect of hydrogen gas was small and can be either slight negative or positive, the effect of oxygen gas was unidirectional and a reduction of 23% to 39% of hydrogen sulfide concentration in biogas was observed within the tested addition range. The time profiles of hydrogen sulfide generation in biogas during the addition of intermediates are shown in Figure 6. These curves clearly demonstrated the slower rates of hydrogen sulfide generation and lower final levels when ferrous, ferric, or oxygen gas of different levels were added to AD feedstock. It is therefore reasonable to postulate the significant role of electrochemical intermediates at the tested addition levels on hydrogen sulfide control in AD through indirect oxidation of sulfide and precipitation [34] of different types of iron sulfide (ferrous sulfide, e.g., pyrrhotite and pyrite, solubility product constant Ksp of about 10-17, calculated from Mineql+ thermodynamic database [40]), which supported the possibility of combining AD with electrolytic cells for sulfide control that generate comparable concentrations of electrochemical intermediates as tested.

19 / 42

Figure 5.

Figure 6.

3.2.2. Methane production with electrochemical intermediates addition It was still necessary to evaluate the effects of electrochemical intermediates generation on biogas production because this electrochemical sulfide mitigation technology was aimed for use together with AD. The cumulative methane production in each treatment over time is presented in Figure 7. Both the time profile curves and maximum methane yields on day 25 were not significantly affected by ferrous and ferric addition at the tested concentration range. A slight methane yield decrease was observed, however, up to 1.6% decrease with ferrous addition, and 2.1% decrease with ferric addition. At the same time, methane percentage in biogas was marginally improved by ferric addition from 59.1% to between 59.3% and 59.6%. Therefore, within the tested range, both ferrous and ferric removed a large portion of hydrogen sulfide while their effects on methane yield was relatively small. Iron salts dosing is currently being used for sulfide control in sewer systems [41, 42] and also for avoiding precipitate scale in anaerobic digester pipes. Our results are in accordance with a previous study [9] that reported the improved or unimpaired methane production in sludge anaerobic digester with ferric salt addition in the range of 5-20 mg-Fe(III)/L. However, different effects of ferric and its salts were reported in literature on different types of substrates with widely different physiochemical properties. A study based on sewer biofilm observed an inhibited methane production rate by 80% at a dosing load of 21 mg-Fe(III)/L in sewage [43],

20 / 42

while another study dosing ferric chloride in wasted activated sludge achieved 79.6% of higher accumulated biogas production than the control [44]. The type of media therefore plays a very important role in the interaction of iron ions and biogas production.

Hydrogen addition stimulated both methane yield and methane content, but oxygen gas addition suppressed methane yield and content in biogas. Hydrogen gas increased methane yield by up to 9.1% but oxygen gas reduced the yield by up to 12.9%. Therefore, oxygen gas should be cautiously controlled within the lowest addition level (29 mg-O2/ L-digestate/day) in terms of inhibition to methane production, which had no impact on methane yield but decreased hydrogen sulfide by 23% but further increase in oxygen would start to impair methane yield and percentage in biogas. The above results demonstrated that the electrochemical intermediates can be helpful in hydrogen sulfide control, while the impact on methane generation can be small given well-controlled electrochemical reactions.

Figure 7.

3.3. Electrochemical H2S removal with biogas production from dairy manure 3.3.1. Continuous voltage application The current in the electrolytic cells at 3 levels of voltage applications (1, 2, and 3 V) with different electrodes combinations (CC/CC, SS/CC, and SS/SS) at two levels of temperature (25 and 35 oC) is shown in Figure 8A. The results demonstrated the rapid increase in current when voltage increased from 1 V to 2 and 3 V. The drastic current increase when voltage increased from 1 V to 2 V and further to 3 V was because of the 21 / 42

overcome of overpotentials for both anode and cathode reactions for water electrolysis and anode corrosion, according to Butler-Volmer equation or Tafel equation which approximately predict exponential increases of current (density) [45]. The increase of the absolute values of overpotentials at both electrodes in dairy manure was confirmed by the measured electrode potentials at the three levels of voltage, following the same pattern as in synthetic media shown in Table 1. The current increase also accompanied with the decrease of overall resistance (both ohmic and polarization resistances of the whole system; polarization resistance included the charge transfer and diffusion resistances [46]), because of the occurrence of anode and cathode reactions (Figure 9). Visual inspection on the resistance information from EIS test also indicated that at 1 V, CC/CC had a lower resistance than SS/CC, while at 2 V and above, CC/CC had a higher resistance, suggesting a higher catalytic effect on anode reactions at 1 V than that of SS. The largest current of 6.29 mA occurred at the SS/CC (stainless steel anode and carbon cloth cathode) combination at 3 V at 35 oC. As a comparison, the addition of oxygen gas as an electrochemical intermediate at the level of 29, 57, and 114 mg/L/d corresponded to a continuous current level of 0.20, 0.40, and 0.80 mA calculated from Faraday’s law when assuming a 100% Coulombic efficiency. At 1 V voltage the water electrolysis hardly occurred, because either anode or cathode potential, or both electrode potentials, were not high enough to overcome overpotentials. As a results, the biogas tested after 7 days’ digestion was consisted of methane and carbon dioxide as that of the control reactor.

Figure 8.

Figure 9. 22 / 42

The biogas production and composition with the treatments at 3 V are shown in Figure 8B, C, and D. Because of the high current levels (between 1.3 mA and 6.29 mA depending on reaction conditions) at 3 V voltage application which continuously generated oxygen gas, total methane generation was negatively affected at both temperatures. The reason for the decrease in methane generation was the aerobic respiration with organic substrate consumption and the inhibition to methanogens when oxygen was introduced. The impact on methane generation was especially pronounced at 35 oC (Figure 8B) as a result of higher current and therefore oxygen generation rates. The biogas composition was consequently altered as well, although the amount of oxygen gas was small among all treatments, i.e., 4.9% for SS/SS at 35 oC, while in other treatments oxygen was less than 1.5%. At 25 oC, all electrochemical treatments resulted in a significant portion of hydrogen gas (26% and 59% of total biogas volume), and the methane content (40% to 47%) was much less than the control reactor (73%). The sum of methane and hydrogen gas accounted for 97.1% and 99.7% for the SS/CC and SS/SS treatments. This gas mixture may be used in combustion engine as hythane, which may enhance combustion rate, extend the lean limit of combustion of biogas, and improve brake thermal efficiency and brake power, as compared to methane alone [47]. At 35 oC, methane content of reactors with stainless steel anode was comparable to the control, while carbon dioxide content (2.9% and 10%) was much smaller than the control (21%). However, all 19 extracellular enzymes (phosphatases, esterases, lipases, proteinases, glucosidases, etc.) concentrations in digestate with SS anode at 3 V treatment, tested via API® Zyme strip, were found much lower than those of the control reactors, and CC

23 / 42

anode resulted in enzyme concentrations in between (Fig. A.1). These results showed a possibility of producing clean biogas ready for use in combustion engines without upgrading, by applying electrochemical systems in AD reactors, but some technical issues will have to be solved for long-term operation of a reactor, e.g., how to maintain high methane yields and prevent the loss of microbial and exoenzyme activities.

At 1 V voltage, hydrogen sulfide was removed with stainless steel anode, by 43% and 14% at 25 oC, and by 34% and 6% at 35 oC, for SS/CC and SS/SS combinations, respectively (Fig. A.2). Contrary to the result in synthetic media test, the sulfide removal effect of carbon cloth was not observed in this test, which can be a result of the small surface area of electrode and interferences caused by background chemical matrices which may compete with sulfide oxidation or shift the selection toward other electroactive species. In the future, experiment with larger carbon cloth electrode surface area will be conducted to signify the effect on sulfide removal. At 2 V and 3 V voltage treatment, hydrogen sulfide was completely removed when stainless steel was used as anode (Figure 8 and Fig. A.2). Carbon cloth anode removed 55% of hydrogen sulfide at 25 oC and 44% at 35 oC at 2 V, and removed 80% at 25 oC and 94% at 35 oC at 3 V. Note that the absolute levels of hydrogen sulfide in the control reactors of each parallel of experiment were different, which was most likely caused by the inoculum difference; therefore, the percentage removal data made more sense than the absolute removal amount in this circumstance. The sulfide oxidation to elemental sulfur (Reaction 1-3) or to sulfate (Reactions 4-6) occur at a potential of -0.271 V and -0.213 V vs. SHE at neutral pH [12], and all anode potentials in this experiment were sufficiently high to oxidize sulfide. However, the rate

24 / 42

of sulfur oxidation to sulfate is kinetically limited and occurs at a slower rate. Graphite granules as electrodes were tested in a previous study for sulfide oxidation [48]. In the potentiostatic condition with anode potential of 0.2 V vs. SHE, biofilm formed on electrode surface started to reduce sulfur with the present organic matter in media and gradually deteriorated sulfide removal efficiency. The issue may be avoided by applying higher voltages so that sulfur deposit will be oxidized and biomass may be stripped away via oxygen evolution, and by installing electrodes in supernant of digestate to minimize particulates and biomass contamination. The liquid analysis showed that under 1 V voltage, sulfate level in all reactors were similar. But at 3 V voltage, the sulfate level was significantly reduced in the electrode combination of SS/CC and SS/SS while the CC/CC had no substantial effect on sulfate level compared to the sulfate level of the control (Fig. A.2). With the release of ferrous ion from anode, precipitates of ferrous sulfide may also be present but this composition was not analyzed in this current study. Therefore, when SS anode was utilized, intermediate oxidation products of sulfide, or sulfide minerals sinks, together with a small portion of sulfate sink, took most part of sulfide from the digestate. Meanwhile, reactive phosphate level in digestate was completely removed by SS/CC and SS/SS combination under 3 V, with simultaneous increase in reactive iron concentration (Fig. A.2). Similar co-removal of sulfide and phosphate was observed in sewers originally aiming for sulfide removal with iron dosage [42].

3.3.2. Intermittent voltage application Intermittent voltage applications for 1 V, 2 V and 3 V voltage (15 min voltage application per day for 35 days) were then studied to improve the performance on hydrogen sulfide

25 / 42

control and biogas generation by generating less amount of electrochemical intermediates. Results showed again that at the same electrode surface area, the combination of SS/CC and SS/SS was more effective than carbon cloth anode in hydrogen sulfide mitigation. At 3 V voltage at 25 oC, the remaining hydrogen sulfide was only 3% and 7% of that of the control (from 427 to 12 and 30 ppmv), while at 35 oC, hydrogen sulfide level was further decreased to 2% of the control (from 339 to 7 and 6 ppm). The application of 2 V voltage was less effective, but it still removed H2S by 41% and 52% (250 and 206 ppmv) of at 25 o

C, and by 67% and 80% (67 and 110 ppmv) at the higher temperature (Table 3). The

results for biogas generation was averaged for 25 and 35 oC (the control reactors at both temperatures generated biogas at similar rate, due to the high (40%) inoculum volatile solids addition), and also showed that methanogenesis underwent normally without abrupt interference in methane production in SS anode treatments, with methane contents between 70% and 80% of biogas volume. It was especially promising that the combination of SS/SS removed majority part of hydrogen sulfide at 2 and 3 V, and retained the capability for methane generation at both voltage levels (Figure 10). However, CC anode treatment yielded less methane than the control, which may be explained by the reactive oxygen species (e.g., free hydroxyl) generation that may oxidize organic matter and can be toxic to microorganisms when accumulated [20]. Nevertheless, these results clearly demonstrated that by coupling AD with electrochemical oxidation process by suitable electrode materials, it will be feasible to mitigate biogas hydrogen sulfide level and meanwhile maintain methane production and yield in AD process.

26 / 42

Table 3.

Figure 10.

The commonly adopted processes for biogas hydrogen sulfide removal is via the installation and operation of standalone facilities for biofilter and biotrickling oxidation, iron oxide/hydroxide adsorption, and liquid absorption [49]. Adsorption using iron oxide has long been used in industry. This process is efficient but chemical intensive, and the adsorption media have to be replaced after utilization or after limited times of regeneration; otherwise, the performance starts declining due to elemental sulfur coverage on surface. The temperature of re-generation process needs to be well controlled to prevent ignition of the loading materials of wood chips because of the exothermic property of the conversion of iron sulfide to iron oxide. Additionally, the huge amount of spent solid waste that has to be processed and disposed is a concern [4]. Biological processes, e.g., biofilters, biotrickling filters and bioscrubbers (fixed-film and suspended growth), offer cheaper options than chemical adsorption methods, but there are strict conditions to maintain those biological reactor performance by controlling microorganism species, media nutrients level, pH, oxygen level, media recirculation, and gas and liquid flow velocity, which all add complexity to the upgrading systems [4].

Hydrogen sulfide removal from biogas is usually a prerequisite before biogas upgrading via processes such as pressure swing adsorption and water scrubbing because of its corrosive property and strong affinity to many adsorbents [50]. It will be ideal that hydrogen sulfide is not emitted to biogas phase or is being mitigated during anaerobic

27 / 42

digestion, so that biogas upgrading in terms of carbon dioxide will not be affected by the presence of hydrogen sulfide, not to mention the protective effects to digesters and pipes which are constantly subjected to the threats of sulfide. Therefore, different methods were proposed for in situ sulfide/hydrogen sulfide removal from reactor media. One benefit of the in situ sulfide removal is that no installation of an additional standalone cleaning unit for hydrogen sulfide is required, thus potentially reducing capital cost of overall AD facilities. Nitrate and nitrite dosing are used to couple sulfide oxidation, and iron salts dosing is another mitigation method that can scavenge sulfide to precipitate insoluble sulfide minerals. There are several studies reported the performances of the application of micro-aeration as a very promising mitigation method in anaerobic digesters [51, 52]. Electrochemical oxidation method used in this study share some similarity to microaeration (e.g., production and release of oxygen gas) and can be a good supplementary. Despite comparable hydrogen sulfide level after treatment by the two methods [53, 54], electrochemical oxidation fulfills direct sulfide oxidation at anodes, potentially generates pure oxygen gas in situ, and does not introduce nitrogen to gas phase. The spatial distribution of electrodes via the adoption of low-cost materials can be designed in a way to decrease the mixing requirement as in micro-aeration. In high rate sulfide generation reactors where biological sulfide oxidation via oxygen gas does not go fast enough to remove sulfide, the release of metal ions from anode will have additional mitigating effects by rapidly binding sulfide to insoluble metal sulfide so that the mitigation effect will not be impaired. Future research is needed to elucidate those benefits, and it is also necessary to scale up the electrochemical oxidation method in order to assess the cost of electrode material maintenance and possible replacement.

28 / 42

4. Conclusions Simultaneously removing hydrogen sulfide from digestate and producing methane by implementing carbon cloth or stainless steel electrodes in AD was made possible in this study. In synthetic sulfide solution with sulfide concentration (10 mM or 320 mgsulfide/L) reaching the maximum level in AD digestate, both electrode materials at 3 V completely removed sulfide in solution and hydrogen sulfide in headspace gas. By adding 2 mM of ferrous and ferric to dairy manure in AD, hydrogen sulfide in biogas was decreased by 48% and 67%, respectively. The oxygen addition removed 39% of hydrogen sulfide in biogas. Meanwhile, a marginally methane yield decrease was observed with ferrous (by 1.6% reduction) and ferric (by 2.1%). Hydrogen gas addition increased methane yield by up to 9.1% but oxygen gas reduced the yield by up to 12.9%. During 35 days’ AD, the intermittent 3 V voltage application (15 min per day) on stainless steel electrodes showed a close to complete biogas sulfide removal (less than 10 ppmv) and maintained similar methane yields to control reactors, while this operating mode saved cost compared to continuous voltage application. Based on those results, assuming an increased electrode surface area, we concluded that at 1 and 2 V, stainless steel anodes have a potential of completely removing hydrogen sulfide from biogas under continuous operation; that at 2 V, carbon cloth has a potential of completely removing hydrogen sulfide from biogas under continuous operation; and that at 3 V, intermittent operation of both electrodes will be sufficient to remove all hydrogen sulfide. This method of electrochemical oxidation for sulfide has a potential to save energy and chemical costs compared to micro-aeration and chemical dosing methods, to avoid

29 / 42

possible inhibition caused by introducing too much chemical dosing reagents, and to better arrange electrode configuration in AD for well control of oxygen gas distribution because of the lower material cost compared to mixed metal oxide electrodes. Future studies will be carried out to evaluate the long-term effects of electrochemical oxidation on biogas production, sulfide removal, and microbial community responses.

Acknowledgements The authors greatly appreciate Drs. Carlos Zamalloa and Jing Gan, in the Department of Bioproducts and Biosystems Engineering at University of Minnesota, for their helpful comments on the manuscript. The authors also greatly appreciate the funding supports by MnDRIVE Bioremediation Project, MnDRIVE Undergraduate Scholar Program, and UMN UROP Program.

30 / 42

References [1] Q. Mahmood, P. Zheng, J. Cai, Y. Hayat, M.J. Hassan, D.-l. Wu, B.-l. Hu, Sources of sulfide in waste streams and current biotechnologies for its removal, Journal of Zhejiang University Science A 8 (2007) 1126-1140. [2] M.J. Higgins, Y.-C. Chen, D.P. Yarosz, S.N. Murthy, N.A. Maas, D. Glindemann, J.T. Novak, Cycling of volatile organic sulfur compounds in anaerobically digested biosolids and its implications for odors, Water Environment Research (2006) 243-252. [3] J.L. Chen, R. Ortiz, T.W. Steele, D.C. Stuckey, Toxicants inhibiting anaerobic digestion: A review, Biotechnology advances 32 (2014) 1523-1534. [4] E. Ryckebosch, M. Drouillon, H. Vervaeren, Techniques for transformation of biogas to biomethane, Biomass and bioenergy 35 (2011) 1633-1645. [5] S. Potivichayanon, P. Pokethitiyook, M. Kruatrachue, Hydrogen sulfide removal by a novel fixed-film bioscrubber system, Process Biochemistry 41 (2006) 708-715. [6] M. Hirai, M. Ohtake, M. Shoda, Removal kinetics of hydrogen sulfide, methanethiol and dimethyl sulfide by peat biofilters, Journal of Fermentation and Bioengineering 70 (1990) 334-339. [7] T. Duangmanee, S. Kumar, S. Sung, Micro-Aeration for Sulfide Removal in Anaerobic Treatment of High-Solid Wastewater: A Pilot-Scale Study, Proceedings of the Water Environment Federation 2007 (2007) 2748-2760. [8] D.G. Cirne, F.P. Van Der Zee, M. Fernandez-Polanco, F. Fernandez-Polanco, Control of sulphide during anaerobic treatment of S-containing wastewaters by adding limited amounts of oxygen or nitrate, Reviews in Environmental Science and Bio/Technology 7 (2008) 93-105. [9] H. Ge, L. Zhang, D.J. Batstone, J. Keller, Z. Yuan, Impact of iron salt dosage to sewers on downstream anaerobic sludge digesters: sulfide control and methane production, Journal of Environmental Engineering 139 (2012) 594-601. [10] C.M. Park, J.T. Novak, The effect of direct addition of iron (III) on anaerobic digestion efficiency and odor causing compounds, Water Sci. Technol 68 (2013) 23912396. [11] I. Pikaar, E.M. Likosova, S. Freguia, J. Keller, K. Rabaey, Z. Yuan, Electrochemical abatement of hydrogen sulfide from waste streams, Critical Reviews in Environmental Science and Technology 45 (2015) 1555-1578. [12] P.K. Dutta, J. Keller, Z. Yuan, R.A. Rozendal, K. Rabaey, Conversion of sulfur species in bioelectrochemical systems, in: K. Rabaey, L. Angenent, U. Schröder, J. Keller (Eds.) Bioelectrochemical systems: from extracellular electron transfer to biotechnological application, IWA Publishing, London. UK, 2010, pp. 243-262. [13] B. Ateya, F. Al-Kharafi, R. Abdallah, A. Al-Azab, Electrochemical removal of hydrogen sulfide from polluted brines using porous flow through electrodes, Journal of applied electrochemistry 35 (2005) 297-303. [14] G. Kelsall, I. Thompson, Redox chemistry of H2S oxidation by the British Gas Stretford Process Part. II: Electrochemical behaviour of aqueous hydrosulphide (HS−) solutions, Journal of applied electrochemistry 23 (1993) 287-295. [15] B. Ateya, F. AlKharafi, A. Al-Azab, Electrodeposition of sulfur from sulfide contaminated brines, Electrochemical and solid-state letters 6 (2003) C137-C140.

31 / 42

[16] P.K. Dutta, K. Rabaey, Z. Yuan, J. Keller, Spontaneous electrochemical removal of aqueous sulfide, Water research 42 (2008) 4965-4975. [17] B. Tartakovsky, P. Mehta, J.-S. Bourque, S. Guiot, Electrolysis-enhanced anaerobic digestion of wastewater, Bioresource technology 102 (2011) 5685-5691. [18] J. Desloover, J. De Vrieze, M. Van de Vijver, J. Mortelmans, R. Rozendal, K. Rabaey, Electrochemical nutrient recovery enables ammonia toxicity control and biogas desulfurization in anaerobic digestion, Environmental science & technology 49 (2015) 948-955. [19] I. Pikaar, R.A. Rozendal, Z. Yuan, J. Keller, K. Rabaey, Electrochemical sulfide oxidation from domestic wastewater using mixed metal-coated titanium electrodes, Water research 45 (2011) 5381-5388. [20] M. Panizza, G. Cerisola, Direct and mediated anodic oxidation of organic pollutants, Chemical reviews 109 (2009) 6541-6569. [21] H. Schäfer, S.M. Beladi-Mousavi, L. Walder, J. Wollschläger, O. Kuschel, S. Ichilmann, S. Sadaf, M. Steinhart, K. Küpper, L. Schneider, Surface Oxidation of Stainless Steel: Oxygen Evolution Electrocatalysts with High Catalytic Activity, ACS Catalysis 5 (2015) 2671-2680. [22] APHA, AWWA, WEF, Standard Methods for the Examination of Water and Wastewater, 21st ed., American Public Health Association, American Water work Association and Water Federation, DC2005. [23] A.P. Borole, D. Aaron, C.Y. Hamilton, C. Tsouris, Understanding long-term changes in microbial fuel cell performance using electrochemical impedance spectroscopy, Environmental science & technology 44 (2010) 2740-2745. [24] H. Fischer, G. Schulz!Ekloff, D. Wöhrle, Oxidation of aqueous sulfide solutions by dioxygen part I: autoxidation reaction, Chemical engineering & technology 20 (1997) 462-468. [25] J. Li, F.P. Chen, G.-P. Jin, X.-S. Feng, X.-X. Li, Removals of aqueous sulfur dioxide and hydrogen sulfide using CeO 2-NiAl-LDHs coating activated carbon and its mix with carbon nano-tubes, Colloids and Surfaces A: Physicochemical and Engineering Aspects 476 (2015) 90-97. [26] L. Le Leuch, A. Subrenat, P. Le Cloirec, Hydrogen sulfide adsorption and oxidation onto activated carbon cloths: applications to odorous gaseous emission treatments, Langmuir 19 (2003) 10869-10877. [27] B. Ateya, F. AlKharafi, A. Alazab, A. Abdullah, Kinetics of the electrochemical deposition of sulfur from sulfide polluted brines, Journal of applied electrochemistry 37 (2007) 395-404. [28] I.G. Casella, M.R. Guascito, E. Desimoni, Sulfide measurements by flow injection analysis and ion chromatography with electrochemical detection, Analytica chimica acta 409 (2000) 27-34. [29] L. Veleva, M.A. Alpuche-Aviles, M.K. Graves-Brook, D.O. Wipf, Comparative cyclic voltammetry and surface analysis of passive films grown on stainless steel 316 in concrete pore model solutions, Journal of Electroanalytical Chemistry 537 (2002) 85-93. [30] N. Ramasubramanian, N. Preocanin, R. Davidson, Analysis of passive films on stainless steel by cyclic voltammetry and Auger spectroscopy, Journal of The Electrochemical Society 132 (1985) 793-798.

32 / 42

[31] D. Shoesmith, P. Taylor, M. Bailey, B. Ikeda, Electrochemical behaviour of iron in alkaline sulphide solutions, Electrochimica Acta 23 (1978) 903-916. [32] C. Kustermans, In situ electrochemical production of iron and alkalinity for the removal of sulfide from sewer systems, University of Ghent, 2015. [33] J.J. Carroll, A.E. Mather, The solubility of hydrogen sulphide in water from 0 to 90 C and pressures to 1 MPa, Geochimica et Cosmochimica Acta 53 (1989) 1163-1170. [34] B. Zheng, H. Liu, Electrochemical Sulfide Removal on Carbon Electrode in Sulfate Reducing Bacteria Microbiological Fuel Cell, ECS Transactions 58 (2014) 1-9. [35] A. Kamyshny Jr, M. Zilberbrand, I. Ekeltchik, T. Voitsekovski, J. Gun, O. Lev, Speciation of polysulfides and zerovalent sulfur in sulfide-rich water wells in southern and central Israel, Aquatic Geochemistry 14 (2008) 171-192. [36] P.L. van den Bosch, D.Y. Sorokin, C.J. Buisman, A.J. Janssen, The effect of pH on thiosulfate formation in a biotechnological process for the removal of hydrogen sulfide from gas streams, Environmental science & technology 42 (2008) 2637-2642. [37] A.J. Belle, S. Lansing, W. Mulbry, R.R. Weil, Methane and hydrogen sulfide production during co-digestion of forage radish and dairy manure, Biomass and Bioenergy 80 (2015) 44-51. [38] F.J. Andriamanohiarisoamanana, Y. Sakamoto, T. Yamashiro, S. Yasui, M. Iwasaki, I. Ihara, O. Tsuji, K. Umetsu, Effects of handling parameters on hydrogen sulfide emission from stored dairy manure, Journal of environmental management 154 (2015) 110-116. [39] E.M. Likosova, J. Keller, R. Rozendal, Y. Poussade, S. Freguia, Understanding colloidal FeSx formation from iron phosphate precipitation sludge for optimal phosphorus recovery, Journal of colloid and interface science 403 (2013) 16-21. [40] MINEQL+, a chemical equilibrium modeling system, Environmental Research Software, Hallowell, ME, 2008. [41] R. Ganigue, O. Gutierrez, R. Rootsey, Z. Yuan, Chemical dosing for sulfide control in Australia: an industry survey, Water research 45 (2011) 6564-6574. [42] O. Gutierrez, D. Park, K.R. Sharma, Z. Yuan, Iron salts dosage for sulfide control in sewers induces chemical phosphorus removal during wastewater treatment, Water research 44 (2010) 3467-3475. [43] L. Zhang, J. Keller, Z. Yuan, Inhibition of sulfate-reducing and methanogenic activities of anaerobic sewer biofilms by ferric iron dosing, Water research 43 (2009) 4123-4132. [44] B. Yu, Z. Lou, D. Zhang, A. Shan, H. Yuan, N. Zhu, K. Zhang, Variations of organic matters and microbial community in thermophilic anaerobic digestion of waste activated sludge with the addition of ferric salts, Bioresource technology 179 (2015) 291298. [45] O.A. Petrii, R.R. Nazmutdinov, M.D. Bronshtein, G.A. Tsirlina, Life of the Tafel equation: Current understanding and prospects for the second century, Electrochimica acta 52 (2007) 3493-3504. [46] Z. He, N. Wagner, S.D. Minteer, L.T. Angenent, An upflow microbial fuel cell with an interior cathode: assessment of the internal resistance by impedance spectroscopy, Environmental science & technology 40 (2006) 5212-5217.

33 / 42

[47] E. Porpatham, A. Ramesh, B. Nagalingam, Effect of hydrogen addition on the performance of a biogas fuelled spark ignition engine, International Journal of Hydrogen Energy 32 (2007) 2057-2065. [48] P.K. Dutta, K. Rabaey, Z. Yuan, R.A. Rozendal, J. Keller, Electrochemical sulfide removal and recovery from paper mill anaerobic treatment effluent, Water research 44 (2010) 2563-2571. [49] R. Muñoz, L. Meier, I. Diaz, D. Jeison, A review on the state-of-the-art of physical/chemical and biological technologies for biogas upgrading, Reviews in Environmental Science and Bio/Technology 14 (2015) 727-759. [50] Q. Sun, H. Li, J. Yan, L. Liu, Z. Yu, X. Yu, Selection of appropriate biogas upgrading technology-a review of biogas cleaning, upgrading and utilisation, Renewable and Sustainable Energy Reviews 51 (2015) 521-532. [51] S.K. Khanal, J.-C. Huang, ORP-based oxygenation for sulfide control in anaerobic treatment of high-sulfate wastewater, Water Research 37 (2003) 2053-2062. [52] L. Krayzelova, J. Bartacek, I. Díaz, D. Jeison, E.I. Volcke, P. Jenicek, Microaeration for hydrogen sulfide removal during anaerobic treatment: a review, Reviews in Environmental Science and Bio/Technology 14 (2015) 703-725. [53] L. Krayzelova, J. Bartacek, N. Kolesarova, P. Jenicek, Microaeration for hydrogen sulfide removal in UASB reactor, Bioresource technology 172 (2014) 297-302. [54] L.D. Nghiem, P. Manassa, M. Dawson, S.K. Fitzgerald, Oxidation reduction potential as a parameter to regulate micro-oxygen injection into anaerobic digester for reducing hydrogen sulphide concentration in biogas, Bioresource technology 173 (2014) 443-447.

34 / 42

Title: Electrochemical Sulfide Removal by Low-cost Electrode Materials in Anaerobic Digestion Authors: Hongjian Lin, Nicholas Williams, Amelia King, and Bo Hu* Address: Department of Bioproducts and Biosystems Engineering, University of Minnesota, 1390 Eckles Ave., St. Paul, MN 55108, USA *

Corresponding author

Ph.D., Associate Professor Tel: 612-625-4215 (O) Fax: 612-624-3005 Email: [email protected]

1 / 11

ϰϬϬϬ ϯϬϬϬ ϮϬϬϬ ϭϬϬϬ Ϭ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

ŽŶƚƌŽů ůĞŚĞĂĚƐƉĂĐĞ ůĞŚĂůĨͲŝŵŵĞƌ ůĞŝŵŵĞƌƐĞĚ



ϱϬϬϬ

ŽŶƚƌŽů ůĞŚĞĂĚƐƉĂĐĞ ůĞŚĂůĨͲŝŵŵĞƌ ůĞŝŵŵĞƌƐĞĚ



ϭϲϬϬϬ ϭϮϬϬϬ ϴϬϬϬ ϰϬϬϬ Ϭ

Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ ƵƌĂƚŝŽŶ͕ĚĂLJ

ϲ͘Ϭ

ϴ͘Ϭ

Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ ƵƌĂƚŝŽŶ͕ĚĂLJ

ϲ͘Ϭ

ϴ͘Ϭ

Figure 1. Time profiles of H2S level in synthetic media during electrochemical oxidation with electrodes at different positions. The applied voltage was 1 V and bottles were kept at room temperature (25 ± 2oC). A, carbon cloth electrodes in 2 mM Na2S media; and B, carbon cloth electrodes in 10 mM Na2S media.

2 / 11

ϭ͘ϱ ϭ͘Ϭ



Ϭ͘ϱ Ϭ͘Ϭ ͲϬ͘ϱͲϭϬϬϬ Ͳϭ͘Ϭ

ƵƌƌĞŶƚ͕ŵ

ϭ͘Ϭ

Ϭ ϱϬϬ WŽƚĞŶƚŝĂůǀƐ͘ŐͬŐů͕ŵs

ͲϱϬϬ

^^ϯϬϰŝŶW^͕ϭƐƚƐĐĂŶ ^^ϯϬϰŝŶƐƵůĨŝĚĞнW^͕ϭƐƚƐĐĂŶ ^^ϯϬϰŝŶW^͕ϯƌĚƐĐĂŶ ^^ϯϬϰŝŶƐƵůĨŝĚĞнW^͕ϯƌĚƐĐĂŶ

ϭ͘ϱ ϭ͘Ϭ Ϭ͘ϱ



Ϭ͘Ϭ ͲϭϬϬϬ ͲϬ͘ϱ

ϭϬϬϬ

Ͳϭ͘Ϭ

/Ăϰ

/Ăϯ ͲϱϬϬ

/Ăϱ

Ϭ

ϱϬϬ

ϭϬϬϬ

/ĐϮ

/Đϯ

WŽƚĞŶƚŝĂůǀƐ͘ŐͬŐů͕ŵs

Ͳϭ͘ϱ



Ϭ͘Ϭ ͲϭϬϬϬ ͲϬ͘ϱ

Ͳϭ͘ϱ

ͲϱϬϬ

^^ϯϭϲŝŶW^͕ϭƐƚƐĐĂŶ ^^ϯϭϲŝŶƐƵůĨŝĚĞнW^͕ϭƐƚƐĐĂŶ ^^ϯϭϲŝŶW^͕ϯƌĚƐĐĂŶ ^^ϯϭϲŝŶƐƵůĨŝĚĞнW^͕ϯƌĚƐĐĂŶ

ϭ͘ϱ

Ͳϭ͘Ϭ

/Ăϭ

/Đϭ

Ͳϭ͘ϱ

Ϭ͘ϱ

ƵƌƌĞŶƚ͕ŵ

Ϯ͘Ϭ

/ĂϮ

ƵƌƌĞŶƚ͕ŵ

ƵƌƌĞŶƚ͕ŵ

ŝŶW^͕ϭƐƚƐĐĂŶ ŝŶƐƵůĨŝĚĞнW^͕ϭƐƚƐĐĂŶ ŝŶW^͕ϯƌĚƐĐĂŶ ŝŶƐƵůĨŝĚĞнW^͕ϯƌĚƐĐĂŶ

Ϯ͘ϱ

Ϭ

ϱϬϬ

WŽƚĞŶƚŝĂůǀƐ͘ŐͬŐů͕ŵs

ϭϬϬϬ

^^ϰϯϬŝŶW^͕ϭƐƚƐĐĂŶ ^^ϰϯϬŝŶƐƵůĨŝĚĞнW^͕ϭƐƚƐĐĂŶ ^^ϰϯϬŝŶW^͕ϯƌĚƐĐĂŶ ^^ϰϯϬŝŶƐƵůĨŝĚĞнW^͕ϯƌĚƐĐĂŶ

Ϯ͘Ϭ ϭ͘ϱ ϭ͘Ϭ Ϭ͘ϱ



Ϭ͘Ϭ ͲϬ͘ϱͲϭϬϬϬ Ͳϭ͘Ϭ Ͳϭ͘ϱ

ͲϱϬϬ

Ϭ

ϱϬϬ

ϭϬϬϬ

WŽƚĞŶƚŝĂůǀƐ͘ŐͬŐů͕ŵs

ͲϮ͘Ϭ

Figure 2. Cyclic voltammograms of electrode materials in 100 mM phosphate buffer solution with and without 5 mM sodium sulfide. A, carbon cloth; B, stainless steel 304; C, stainless steel 316; and D, stainless steel 430. The scan rate was 50 mV/s.

3 / 11

ϯϬϬϬ ϮϬϬϬ ϭϬϬϬ Ϭ Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ

ϲ͘Ϭ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

ŽŶƚƌŽů͕ϮϱŽ ϭs͕ϮϱŽ Ϯs͕ϮϱŽ ϯs͕ϮϱŽ



ϰϬϬϬ

ϴ͘Ϭ

ϰϬϬϬ ϯϬϬϬ ϮϬϬϬ ϭϬϬϬ Ϭ Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ



ϭϱϬϬϬ

ŽŶƚƌŽů͕ϮϱŽ ϭs͕ϮϱŽ Ϯs͕ϮϱŽ ϯs͕ϮϱŽ

ϭϬϬϬϬ ϱϬϬϬ Ϭ Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ

ϲ͘Ϭ

ϮϬϬϬϬ ϭϱϬϬϬ ϭϬϬϬϬ ϱϬϬϬ Ϭ

ϴ͘Ϭ

Ϭ͘Ϭ

Ϯ͘Ϭ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

ϮϬϬϬϬ ϭϱϬϬϬ

ŽŶƚƌŽů͕ϮϱŽ ϭs͕ϮϱŽ Ϯs͕ϮϱŽ ϯs͕ϮϱŽ

ϱϬϬϬ Ϭ Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ ƵƌĂƚŝŽŶ͕ĚĂLJ

ϰ͘Ϭ

ϲ͘Ϭ

ϴ͘Ϭ

ƵƌĂƚŝŽŶ͕ĚĂLJ



ϭϬϬϬϬ

ϴ͘Ϭ

ŽŶƚƌŽů͕ϯϱŽ ϭs͕ϯϱŽ Ϯs͕ϯϱŽ ϯs͕ϯϱŽ



ϮϱϬϬϬ

ƵƌĂƚŝŽŶ͕ĚĂLJ

ϮϱϬϬϬ

ϲ͘Ϭ

ƵƌĂƚŝŽŶ͕ĚĂLJ ,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

,Ϯ^ŝŶŚĞĂĚƐƉĂĐĞ͕ƉƉŵǀ

ƵƌĂƚŝŽŶ͕ĚĂLJ

ϮϬϬϬϬ

ŽŶƚƌŽů͕ϯϱŽ ϭs͕ϯϱŽ Ϯs͕ϯϱŽ ϯs͕ϯϱŽ



ϲ͘Ϭ

ϴ͘Ϭ

&

ϮϱϬϬϬ ϮϬϬϬϬ ϭϱϬϬϬ

ŽŶƚƌŽů͕ϯϱŽ ϭs͕ϯϱŽ Ϯs͕ϯϱŽ ϯs͕ϯϱŽ

ϭϬϬϬϬ ϱϬϬϬ Ϭ Ϭ͘Ϭ

Ϯ͘Ϭ

ϰ͘Ϭ

ϲ͘Ϭ

ϴ͘Ϭ

ƵƌĂƚŝŽŶ͕ĚĂLJ

Figure 3. Time profiles of headspace H2S levels in synthetic media during electrochemical oxidation at 1, 2 and 3 V voltage at different temperatures and electrode materials. A and B, in 2 mM Na2S media, carbon cloth electrodes; C and D, in 10 mM Na2S media, carbon cloth electrodes; and E and F, in 10 mM Na2S media, stainless steel electrodes.

4 / 11

&ŝŶĂůĐŽŶĐĞŶƚƌĂƚŝŽŶ͕ŵD

ϭ͘ϰϬ



^ƵůĨŝĚĞ ^ƵůĨĂƚĞ

Ϭ͘ϴϴ

ϭ͘ϮϬ

ϭ͘Ϭϯ

ϭ͘ϬϬ Ϭ͘ϳϭ

Ϭ͘ϴϬ

Ϭ͘ϱϳ

Ϭ͘ϲϬ Ϭ͘ϰϬ Ϭ͘ϮϬ

Ϭ͘Ϭϲ Ϭ͘ϭϰ Ϭ͘Ϭϱ Ϭ͘ϬϬ Ϭ͘ϬϬ

Ϭ͘ϭϯ Ϭ͘ϬϬ

Ϭ͘Ϭϭ

Ϭ͘Ϭϵ Ϭ͘ϬϬ Ϭ͘ϬϬ

Ϭ͘ϬϬ

&ŝŶĂůĐŽŶĐĞŶƚƌĂƚŝŽŶ͕ŵD

Ϭ͘ϬϬ

ϲ͘ϬϬ

^ƵůĨŝĚĞ ^ƵůĨĂƚĞ

ϰ͘ϱϯ



ϱ͘Ϯϴ

ϱ͘ϬϬ ϰ͘ϬϬ

ϯ͘ϯϳ

ϯ͘ϯϳ Ϯ͘ϯϬ

ϯ͘ϬϬ

ϭ͘ϲϯ

Ϯ͘ϬϬ ϭ͘ϬϬ

Ϭ͘ϵϳ Ϭ͘ϲϰ

Ϭ͘ϴϳ Ϭ͘ϳϵ Ϭ͘ϬϬ

Ϭ͘ϯϵ

Ϭ͘ϳϮ

ϭ͘ϯϵ Ϭ͘ϳϰ Ϭ͘ϬϬ

&ŝŶĂůĐŽŶĐĞŶƚƌĂƚŝŽŶ͕ŵD

Ϭ͘ϬϬ

ϲ͘ϬϬ

^ƵůĨŝĚĞ ^ƵůĨĂƚĞ



ϱ͘ϬϬ ϰ͘ϬϬ

ϯ͘ϯϳ

ϯ͘ϯϳ Ϯ͘ϰϰ

ϯ͘ϬϬ Ϯ͘ϬϬ ϭ͘ϬϬ

Ϯ͘ϳϱ ϭ͘ϵϮ Ϭ͘ϵϲ

Ϭ͘ϲϰ

Ϯ͘ϲϱ

ϭ͘ϯϴ Ϭ͘ϵϲ

Ϭ͘ϱϯ Ϭ͘ϬϬ

Ϭ͘ϯϵ

Ϭ͘ϰϰ

Ϭ͘ϯϮ

Ϭ͘Ϭϭ

Ϭ͘ϬϬ

Figure 4. Sulfide and sulfate concentrations in final synthetic media after electrochemical oxidation. A, in 2 mM Na2S media, carbon cloth electrodes; B, in 10 mM Na2S media, carbon cloth electrodes; and C, in 10 mM Na2S media, stainless steel electrodes. The concentration of 0.00 indicated sulfide concentrations less than the detection limit of 0.5 µM (or 0.016 mg/L).

5 / 11

ϭ͕ϬϵϬ ϴϮϴ ϱϭϰ

ϱϬϬ

ϮϬϬϬ



ϭ͕ϵϬϳ ϭ͕ϵϵϳ ϭ͕ϵϭϵ ϭ͕ϴϱϰ ϭ͕ϱϭϲ

ϭϱϬϬ ϭϬϬϬ

ϭ͕ϯϬϵ

ϭ͕ϭϵϴ

ƵŵƵůĂƚŝǀĞ,Ϯ^ůĞǀĞů ŽŶĚĂLJϮϱ͕ƉƉŵ

ϭϱϬϬ ϭϬϬϬ



ϭ͕ϰϵϮ

ϭ͕ϯϰϵ ϭ͕Ϯϰϯ

ƵŵƵůĂƚŝǀĞ,Ϯ^ůĞǀĞů ŽŶĚĂLJϮϱ͕ƉƉŵ

ƵŵƵůĂƚŝǀĞ,Ϯ^ůĞǀĞůŽŶ ĚĂLJϭϮ͕ƉƉŵ ƵŵƵůĂƚŝǀĞ,Ϯ^ůĞǀĞů ŽŶĚĂLJϭϮ͕ƉƉŵ

ϮϬϬϬ ϭ͕ϲϰϮ

ϭϵϬϬ ϭ͕ϱϰϯ ϭϰϬϬ

ϭ͕Ϯϳϭ



ϭ͕ϯϵϱ ϭ͕ϭϴϰ

ϭ͕Ϭϯϯ

ϳϵϲ ϵϬϬ

ϱϬϴ

ϰϬϬ

ϮϬϬϬ ϭϱϬϬ

ϭ͕ϴϬϱ ϭ͕ϴϲϵ ϭ͕ϳϳϯ

 ϭ͕ϲϲϱ ϭ͕ϯϴϯ

ϭ͕Ϯϯϭ

ϭ͕ϭϬϱ

ϭϬϬϬ

Figure 5. Cumulative H2S in biogas with ferrous and ferric addition (A, day 12; B, day 25), and hydrogen and oxygen addition (C, day 12; D, day 25).

6 / 11

ƵŵŵƵůĂƚŝǀĞ,Ϯ^ ƉƌŽĚƵĐƚŝŽŶ͕ŵ>

ŽŶƚƌŽů &ĞůϮͺϬ͘ϱŵD &ĞůϮͺϭŵD &ĞůϮͺϮŵD &ĞůϯͺϬ͘ϱŵD &ĞůϯͺϭŵD &ĞůϯͺϮŵD



Ϭ͘ϲ Ϭ͘ϱ Ϭ͘ϰ Ϭ͘ϯ Ϭ͘Ϯ Ϭ͘ϭ Ϭ͘Ϭ Ϭ

ϭϬ

ϮϬ

ϯϬ

ƵŵŵƵůĂƚŝǀĞ,Ϯ^ ƉƌŽĚƵĐƚŝŽŶ͕ŵ>

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ ŽŶƚƌŽů ,Ϯͺϯ͘ϯŵŐͬ>ͬĚ ,Ϯͺϲ͘ϱŵŐͬ>ͬĚ ,ϮͺϭϯŵŐͬ>ͬĚ KϮͺϮϲŵŐͬ>ͬĚ KϮͺϱϮŵŐͬ>ͬĚ KϮͺϭϬϱŵŐͬ>ͬĚ



Ϭ͘ϳ Ϭ͘ϲ Ϭ͘ϱ Ϭ͘ϰ Ϭ͘ϯ Ϭ͘Ϯ Ϭ͘ϭ Ϭ͘Ϭ Ϭ

ϭϬ

ϮϬ

ϯϬ

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ

Figure 6. Time profiles of cumulative hydrogen sulfide production with ferrous or ferric ions addition (A), and hydrogen or oxygen gas (B) addition. Data points were based on three replicates of experiments.

7 / 11

Ϭ͘ϯ Ϭ͘Ϯ

ŽŶƚƌŽů &ĞůϮͺϬ͘ϱŵD &ĞůϮͺϭŵD &ĞůϮͺϮŵD

Ϭ͘ϭ Ϭ Ϭ

ϭϬ

ϮϬ

ƵŵƵůĂƚŝǀĞ ďŝŽŵĞƚŚĂŶĞ͕>ͬŐͲs^

ƵŵƵůĂƚŝǀĞ ďŝŽŵĞƚŚĂŶĞ͕>ͬŐͲs^



Ϭ͘ϰ



Ϭ͘ϰ Ϭ͘ϯ Ϭ͘Ϯ

ŽŶƚƌŽů &ĞůϯͺϬ͘ϱŵD &ĞůϯͺϭŵD &ĞůϯͺϮŵD

Ϭ͘ϭ Ϭ

ϯϬ

Ϭ

ϭϬ

Ϭ͘ϰ



Ϭ͘ϯ Ϭ͘Ϯ

ŽŶƚƌŽů ,Ϯͺϯ͘ϲŵŐͬ>ͬĚ ,Ϯͺϳ͘ϭŵŐͬ>ͬĚ ,ϮͺϭϰŵŐͬ>ͬĚ

Ϭ͘ϭ Ϭ Ϭ

ϭϬ

ϮϬ



Ϭ͘ϯ Ϭ͘Ϯ

ŽŶƚƌŽů KϮͺϮϵŵŐͬ>ͬĚ KϮͺϱϳŵŐͬ>ͬĚ KϮͺϭϭϰŵŐͬ>ͬĚ

Ϭ͘ϭ Ϭ

ϯϬ

Ϭ

ϭϬ

Ϭ͘ϯϬϬ

Ϭ͘Ϯϵϳ

 Ϭ͘Ϯϵϵ

Ϭ͘Ϯϵϲ

Ϭ͘Ϯϵϱ

Ϭ͘ϯϯϴ Ϭ͘ϯϭϬ

Ϭ͘ϯϭϴ Ϭ͘ϯϭϵ

' Ϭ͘ϯϬϳ Ϭ͘ϮϵϬ Ϭ͘ϮϳϬ

ƵŵŵƵůĂƚŝǀĞďŝŽŵĞƚŚĂŶĞ ƉĞƌĐĞŶƚĂŐĞďĞĨŽƌĞĚĂLJϮϱ͕ й

Ϭ͘Ϯϵϴ

ϮϬ

ϯϬ

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ

ƵŵŵƵůĂƚŝǀĞďŝŽŵĞƚŚĂŶĞ ƉĞƌĐĞŶƚĂŐĞďĞĨŽƌĞĚĂLJϮϱ͕ й

ƵŵƵůĂƚŝǀĞďŝŽŵĞƚŚĂŶĞ ďĞĨŽƌĞĚĂLJϮϱ͕>ͬŐͲs^ ƵŵƵůĂƚŝǀĞďŝŽŵĞƚŚĂŶĞ ďĞĨŽƌĞĚĂLJϮϱ͕>ͬŐͲs^

Ϭ͘ϯϲ Ϭ͘ϯϰ Ϭ͘ϯϮ Ϭ͘ϯ Ϭ͘Ϯϴ Ϭ͘Ϯϲ Ϭ͘Ϯϰ

Ϭ͘ϯϬϭ

ϯϬ

Ϭ͘ϰ

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ

Ϭ͘ϯϬϰ Ϭ͘ϯϬϮ Ϭ͘ϯ Ϭ͘Ϯϵϴ Ϭ͘Ϯϵϲ Ϭ͘Ϯϵϰ Ϭ͘ϮϵϮ

ϮϬ

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ ƵŵƵůĂƚŝǀĞ ďŝŽŵĞƚŚĂŶĞ͕>ͬŐͲs^

ƵŵƵůĂƚŝǀĞ ďŝŽŵĞƚŚĂŶĞ͕>ͬŐͲs^

ZĞĂĐƚŝŽŶƚŝŵĞ;ĚͿ

ϲϵ ϲϳ ϲϱ ϲϯ ϲϭ ϱϵ ϱϳ ϱϱ

ϲϬ ϱϵ͘ϱ ϱϵ ϱϴ͘ϱ ϱϴ ϱϳ͘ϱ

ϱϵ͘ϭ

ϱϵ͘ϲ ϱϴ͘ϵ ϱϴ͘ϱ

ϱϵ͘ϰ

ϱϵ͘ϯ

,

ϲϳ͘Ϯ ϲϭ͘ϵ

ϲϮ͘ϵ

&

ϱϵ͘ϭ

ϲϰ͘ϯ ϲϮ͘Ϭ ϲϬ͘Ϯ ϱϲ͘ϴ

Figure 7. Time profiles of biomethane production in anaerobic digestion with ferrous addition (A), ferric addition (B), hydrogen addition (C), and oxygen addition (D); and cumulative biomethane production and percentage (E and F, ferrous and ferric addition; G and H, hydrogen and oxygen addition)

8 / 11

ϭϬϬй ϴϬй ϲϬй ϰϬй ϮϬй Ϭй

 ϯs

ϯs

ϲ͘Ϯϵ ϰ͘ϬϮ

Ϯ͘ϲϮ Ϭ͘ϭϲ Ϭ͘ϬϮ ϯϱŽ͕ ͬ

Ϭ͘ϳϵ Ϭ͘Ϭϰ

Ϭ͘ϱϵ Ϭ͘Ϭϯ

ŝŽŐĂƐĐŽŵƉŽƐŝƚŝŽŶ͕ŵ>

Ϯs

 ϯs

ϯϬϬ ϮϱϬ ϮϬϬ ϭϱϬ ϭϬϬ ϱϬ Ϭ

,ϰ KϮ ,Ϯ KϮ

ϯϱŽ͕ ϯϱŽ͕ ^^ͬ ^^ͬ^^

,ϰ KϮ ,Ϯ KϮ

,Ϯ^ĐŽŶĐĞŶƚƌĂƚŝŽŶŝŶ ďŝŽŐĂƐ͕ƉƉŵǀ

ƵƌƌĞŶƚ͕ŵ ŝŽŐĂƐĐŽŵƉŽƐŝƚŝŽŶ

ϴ  ϭs ϳ ϲ ϱ ϯ͘ϰϬ ϰ Ϯ͘ϳϮ ϯ ϭ͘ϯϬ Ϯ Ϭ͘ϴϵ Ϭ͘ϭϬ Ϭ͘ϲϰ ϭ Ϭ͘ϬϮ Ϭ͘Ϭϭ Ϭ͘ϬϬ Ϭ ϮϱŽ͕ ϮϱŽ͕ ϮϱŽ͕ ͬ ^^ͬ ^^ͬ^^

ϲϬϬ ϰϬϬ ϮϬϬ

ϱϬϬ

 ϯs

ϰϬϬ

ϭϬϬ Ϭ

Ϭ

Ϯϱ

Ϭ

Ϭ

Ϭ

Figure 8. Electrical current (A) at 1 V, 2 V, and 3 V voltage applications, biogas composition (B and C), and hydrogen sulfide concentration (D) with different electrodes combinations at 3 V. The concentration of 0 ppmv indicates less than the detection limit of the analysis method.

9 / 11

ͲͺŝŵĂŐ;ёͿ

ͲͺŝŵĂŐ;ёͿ



ϴϬϬ

ϲϬϬ



ϰϬϬ

ϯϬϬ

ŶŽĚĞͬĂƚŚŽĚĞ͗ͬ ϰϬϬ

ŶŽĚĞͬĂƚŚŽĚĞ͗^^ͬ

Ͳ ^^Ͳ ^^Ͳ^^

ϮϬϬ

ŶŽĚĞͬĂƚŚŽĚĞ͗^^ͬ^^ ϭϬϬ

ϮϬϬ

Ϭ

Ϭ Ϭ

ϮϬϬ

ϰϬϬ ϲϬϬ ͺƌĞĂů;ёͿ

ϴϬϬ

Ϭ

ϮϬϬ ͺƌĞĂů;ёͿ

ϰϬϬ

Figure 9. Complex plan (Nyquist) plots of electrochemical impedance spectroscopy (EIS) of different electrode combinations in dairy manure scanned at 1 V (A) and 2 V (B).

10 / 11

DĞƚŚĂŶĞLJŝĞůĚŝŶĚŝŐĞƐƚŝŽŶ͕ŵ>Ͳ,ϰͬŐͲs^

/ŶĚĂLJϭͲ ϳ

ϰϬϬ

/ŶĚĂLJϭͲ Ϯϭ

/ŶĚĂLJϭͲ ϯϱ

ϯϱϬ ϯϬϬ ϮϱϬ ϮϬϬ ϭϱϬ ϭϬϬ ϱϬ Ϭ ͬ ^^ͬ ^^ͬ^^ ͬ ^^ͬ ^^ͬ^^ ͬ ^^ͬ ^^ͬ^^ ŽŶƚƌŽů ϭs

Ϯs

ϯs

Ϭs

Figure 10. Cumulated methane yields on day 7, 21, and 35 in dairy manure anaerobic digesters assisted with different levels of voltage.

11 / 11

Title: Electrochemical Sulfide Removal by Low-cost Electrode Materials in Anaerobic Digestion Authors: Hongjian Lin, Nicholas Williams, Amelia King, and Bo Hu* Address: Department of Bioproducts and Biosystems Engineering, University of Minnesota, 1390 Eckles Ave., St. Paul, MN 55108, USA *

Corresponding author

Ph.D., Associate Professor Tel: 612-625-4215 (O) Fax: 612-624-3005 Email: [email protected]

35 / 42

Table 1. Measured electrode potentials when three levels of voltages were applied (25 oC)  ŶŽĚĞͲ ĂƚŚŽĚĞ

ůĞĐƚƌŝĐĂůƉŽƚĞŶƚŝĂů͕ʔ tĂƚĞƌ ĂƚŚŽĚĞ ŽdžŝĚĂƚŝŽŶ sǀƐ͘ sǀƐ͘ sǀƐ͘^, ^, ^, /ŶϬ͘ϭDƉŚŽƐƉŚĂƚĞďƵĨĨĞƌƐŽůƵƚŝŽŶ;Ɖ,Εϳ͘ϬͿ нϬ͘ϴϵ ͲϬ͘ϭϭ нϬ͘ϵϱ ͲϬ͘Ϭϱ нϭ͘Ϯϯ Ϭ͘Ϯϯ нϭ͘ϰϯ ͲϬ͘ϱϳ  нϬ͘ϴϭϰ нϭ͘ϰϵ ͲϬ͘ϱϭ  нϭ͘ϰϮ ͲϬ͘ϱϴ нϭ͘ϳϵ Ͳϭ͘Ϯϭ нϭ͘ϳϳ Ͳϭ͘Ϯϯ нϭ͘ϳϭ Ͳϭ͘Ϯϵ

sŽůƚĂŐĞ

s

ͲΎ ^^Ͳ ^^Ͳ^^ Ͳ ^^Ͳ ^^Ͳ^^ Ͳ ^^Ͳ ^^Ͳ^^

ϭ ϭ ϭ Ϯ Ϯ Ϯ ϯ ϯ ϯ

ŶŽĚĞ

tĂƚĞƌ ƌĞĚƵĐƚŝŽŶ sǀƐ͘ ^,

ͲϬ͘ϰϭϱ

* Through the paper, the following abbreviations were used: CC, carbon cloth; SS, stainless steel; CC/CC, carbon cloth anode and cathode; SS/CC, stainless steel anode and carbon cloth cathode; and SS/SS, stainless steel anode and cathode.

36 / 42

Table 2. Sulfide removal rate calculated from the results of within two days’ operation at different applied voltages with carbon cloth electrodes sŽůƚĂŐĞ ^ƵůĨŝĚĞƌĞŵŽǀĂůƌĂƚĞ /ŶŝƚŝĂůEĂϮ^ Ž ĐŽŶĐĞŶƚƌĂƚŝŽŶ s ϯϱŽ hŶŝƚ Ϯϱ  ϮŵD ϭ͘Ϭ ϱϯ;цϭͿ ϰϰΎ Ϯ͘Ϭ ϭϮϳ;цϭϳͿ ϭϮϴ;цϭϮͿ  ϯ͘Ϭ хϮϬϯ;цϲϭͿΎΎ хϮϬϯ;цϲϭͿΎΎ  ŐͲ^ͬĚͬŵϮ ϭϬŵD ϭ͘Ϭ ϭϳϳ;цϮϬϯͿ ϭϮϭΎ Ϯ͘Ϭ ϮϬϯ;цϵϴͿ Ϯϯϱ;цϭϳϯͿ  ϯ͘Ϭ ϳϭϵ;цϭϭϮͿ ϴϭϲ;цϭϭϳͿ  * These two conditions were conducted without replicates. ** All sulfide was removed in two days in these two conditions, therefore resulting in a lower limit of removal rate.

37 / 42

Table 3. Cumulated hydrogen sulfide content (ppmv) in biogas on day 7, 21, and 35 at different applied voltages, temperatures, and electrodes combination

  ϭs ĂLJϳ ĂLJϮϭ  ĂLJϯϱ  Ϯs ĂLJϳ ĂLJϮϭ  ĂLJϯϱ  ϯs ĂLJϳ ĂLJϮϭ  ĂLJϯϱ 

38 / 42

ϮϱŽ͕ ͬ ϳϬϬ ϯϱϲ ϯϰϬ ϯϱϬ Ϯϵϱ Ϯϴϳ ϰϬϬ Ϯϴϳ Ϯϱϱ

ϮϱŽ͕ ^^ͬ ϱϬϬ ϯϵϳ ϯϵϳ ϮϱϬ ϮϱϬ ϮϱϬ ϭϬ ϭϲ ϭϮ

ϮϱŽ͕ ^^ͬ^^ ϱϬϬ ϰϬϮ ϯϵϯ ϮϱϬ ϭϵϲ ϮϬϲ ϮϬ Ϯϱ ϯϬ

ϯϱŽ͕ ͬ ϯϬϬ ϮϴϬ Ϯϳϳ ϮϱϬ ϮϱϬ Ϯϱϴ ϮϬϬ ϭϴϮ ϭϳϴ

ϯϱŽ͕ ^^ͬ ϰϬϬ ϯϱϵ ϯϱϴ ϱϬ ϲϮ ϲϳ ϭϬ ϳ ϳ

ϯϱŽ͕ ^^ͬ^^ ϯϬϬ ϯϭϲ ϯϭϴ ϴϬ ϵϴ ϭϭϬ ϭϬ ϲ ϲ

ŽŶƚƌŽů

ϮϱŽ

ϯϱŽ

ĂLJϳ ĂLJϮϭ ĂLJϯϱ      

ϱϬϬ ϰϯϯ ϰϮϳ      

ϯϱϬ ϯϯϭ ϯϯϵ      

Highlights: Carbon cloth and stainless steel AISI 304 were evaluated for sulfide oxidation Both electrode materials showed complete sulfide removal of 10 mM at 3 V in 2 days Carbon cloth had a stronger catalytic effect for sulfide oxidation in synthetic media Ferrous and ferric marginally affected methane production, but mitigated sulfide Stainless steel electrodes removed most sulfide without interfering methane production

42 / 42