Removal of hydrogen sulfide gas using coal fly ash – blast furnace cement composite


 The number of complaints regarding offensive odors from service industries, such as restaurants and garages, has recently increased. In this study, we aimed to develop an adsorbent for hydrogen sulfide gas derived from domestic wastewater and reveal the mechanisms of its removal. The adsorbent used for hydrogen sulfide gas removal was prepared by mixing coal fly ash and blast furnace cement with a mixing ratio of 87:13 by mass percentage. The optimum calcination temperature of the adsorbent was 700 °C to achieve a high removal performance for both dry and humid hydrogen sulfide gas. The X-ray absorption fine structure analysis revealed that hydrogen sulfide was removed on the adsorbent by oxidizing to sulfate. A pilot-scale experiment was conducted to evaluate the removal performance of hydrogen sulfide gas derived from domestic wastewater using the developed adsorbent. For a week, the average removal percentage of hydrogen sulfide gas derived from domestic wastewater remained at 99.1%. Therefore, the developed adsorbent for hydrogen sulfide gas is promising and cost-effective for promoting the recycling of coal fly ash.


INTRODUCTION
The four major offensive odors, ammonia, trimethylamine, methyl mercaptan, and hydrogen sulfide, are regulated under the Offensive Odor Control Law in Japan. Currently, the number of complaints regarding offensive odors from service industries, such as restaurants and garages, has risen to 8,000 per year (Kamigawara 2003). Offensive odors are also generated by limestabilized biosolids, such as sewage sludge (Krach et al. 2008), pet food, roofing tar, dead animals (Eltarkawe & Miller 2019), urban rivers (Sado-Inamura & Fukushi 2018), and Johkasou, such as a septic tank, a Japanese compact-scale advanced wastewater treatment system (Ichinari et al. 2008;Fajri et al. 2018).
This study focused on removing hydrogen sulfide (H 2 S) which is a highly toxic, colorless gas, with a rotten-egg-like odor. H 2 S is generated from hydrocarbon sources, such as natural gas, biogas, crude oil, wastewater treatment, and anoxic sediments (Nassar et al. 2016;Asaoka et al. 2018;Khabazipour & Anbia 2019). Because H 2 S is highly toxic and corrosive, it can cause corrosion of oil and gas pipelines and their devices (Nassar et al. 2016), poison catalysts (Khabazipour & Anbia 2019), and pose a serious threat to human health. The odor threshold for human exposure to H 2 S is 0.011 mg m À3 (Selene & Chou 2003). When H 2 S exposure exceeds 2.8 mg m À3 , it negatively impacts human health (Selene & Chou 2003). Therefore, to control pollution and protect human health, H 2 S must be removed.
In a previous study on the removal of H 2 S gas, porous metal oxides such as Fe 2 O 3 /Al 2 O 3 (Su et al. 2017), Co 3 O 4 (Wang et al. 2017a), three-dimensionally ordered macropore CuO (Wang et al. 2017b), ZnO (Tran 2016), and zeolite (Liu & Wang 2017;Yang et al. 2018), metal oxide-loaded mesoporous silicas (Montes et al. 2013;Wu et al. 2017;Cara et al. 2018), and Fenton reagent (Wang et al. 2019) were proposed. However, these adsorbents require significant preparation costs. A costeffective adsorbent for removing H 2 S gas was developed in this study by mixing coal fly ash, which is a by-product of coal electric power plants and blast furnace cement, to increase the specific surface area through pozzolanic reaction. A coal fly ash-blast furnace cement composite was previously reported to be capable of removing H 2 S from both fresh and seawater (Asaoka et al. 2012;Asaoka et al. 2014;Asaoka et al. 2017). However, no data on the removal of H 2 S gas by the coal fly ash-blast furnace cement composite are available. This study was aimed to optimize the granulated coal ash to remove H 2 S gas derived from domestic wastewater and reveal the mechanisms of H 2 S gas removal by the coal fly ash-blast furnace cement composite.

METHODS
Preparing the coal fly ash-blast furnace cement composite Coal fly ash provided by the Chugoku Electric Power Co., Inc., Japan, was mixed with blast furnace cement (Class B: Nippon Steel Blast Furnace Slag Cement Co., Ltd, Japan) at a mixing ratio of 87:13 by mass percentage. Water was added so that the mixture amounted to 20 wt.%. The mixture was then granulated using a rotary pan-type granulator. Every day for the first week, pure water was sprayed to wet the surface of the granulated composite. The granulated composite was then air-dried in a laboratory at 25°C for at least 4 months to complete the pozzolanic reaction between the coal fly ash and blast furnace cement. The granulated composite was mainly composed of 39.5% silicon, 13.3% carbonate, 12.6% Al 2 O 3 , 5.5% CaO, 2.3% Fe 2 O 3 , 0.8% MgO, 0.6% K 2 O, 0.6% TiO 2 , 0.3%Na 2 O, 0.2% P 2 O 5 , and other trace substances (Asaoka et al. 2012). The granulated composite was crushed into a diameter of approximately 5 mm and screened using a sieve into a diameter of 3-5 mm. The composite was calcined for 3 h at temperatures ranging from 300-1,000°C to determine the optimum calcination temperature. The Brunauer-Emmett-Teller (BET) specific surface areas of the composites were measured by the adoption of N 2 gas using an advanced micropore size and chemisorption analyzer (Autosorb iQ, Quantachrome Instruments). The granulated composite for thermogravimetry/differential thermal analysis (TG-DTA) was ground using a mortar, rinsed five times with acetone, and dried one day before analyses. TG-DTA was performed using a thermal analyzer (Thermo plus EVO II; Rigaku) at a heating rate of 10°C min À1 from 25°C to 1,000°C in an alumina pan under N 2 gas flow (800 mL min À1 ). The surface pH of the granulated composite was measured using a compact water quality meter (LAQUA twin pH; HORIBA). On the electrode, a CMF filter sheet (Y100A13; Advantec) was installed, and the granulated composite particles were placed on the filter. The surface pH of the granulated composite was then measured by dripping a few drops of ultrapure water on the granulated composite particles.

Removal experiments of H 2 S gas
Adsorption experiments were conducted under both dry and humid gas conditions. Under the dry gas condition, approximately 28 g of the granulated composite samples (hereafter adsorbent) was loaded into a polyvinyl chloride column (31 mm inner diameter Â 50 mm length). To simulate humid conditions in Johkasou (RH:100%), the adsorbent was immersed in water for 1 day and used in the adsorption experiments.
To simulate H 2 S derived from wastewater of multiple dwelling houses, convenience stores, restaurants, and food factories, H 2 S gas was generated as follows (Supplementary Material, Figure S1): 200 g of sodium sulfide nonahydrate (Kishida Chemical) was added to 400 mL of saturated sodium sulfide nonahydrate solution in 2 L of polyethylene vial. Five milliliters of 10% HCl (Guaranteed Reagent; Kishida Chemical) was dropped into the polyethylene vial every 3 min to generate H 2 S gas. Using an air pump, deoxidized air was introduced into the saturated sodium sulfide nonahydrate solution, and the generated H 2 S gas was introduced into the column at a rate of 0.3 L min À1 . Gas trap bags were installed at the front and rear stages of the column to measure the concentration of H 2 S gas before and after the column using detection tubes (4LL, 4 L, and 4HM; Gastech). The laboratory temperature was maintained at approximately 28°C and 1 atm. The generated H 2 S gas with an average concentration of 1,000-1,100 ppm was introduced into the column. The adsorption of H 2 S gas was calculated using Equation (1).
where V represents the volume of the feed stream of the H 2 S gas to the column (m 3 ), C i represents the initial concentration of H 2 S gas (g m À3 ), C represents the concentration of H 2 S gas after the column (g m À3 ), and W represents the weight of the coal fly ash-blast furnace cement composite samples loaded in the column (m 3 ). The adsorbent sample for X-ray absorption fine structure (XAFS) analyses to identify the sulfur species adsorbed on the adsorbent was vacuum-packed and stored in a freezer to prevent changes in the chemical species.

Identification of sulfur species adsorbed on the adsorbent
At the Hiroshima Synchrotron Research Center, XAFS analyses were performed to identify the sulfur species adsorbed on the adsorbent using BL11 (HiSOR; Asaoka et al. 2012Asaoka et al. , 2018. The sulfur K-edge spectra (range 2,460-2,490 eV) of the adsorbent were measured using X-ray fluorescence yield mode with an SDD detector (XR-100SDD; AMPTEK) under He atmosphere. The K-edge main peak of sulfate derived from CuSO 4 ·5H 2 O was calibrated to 2,481.6 eV. The angle between the incident X-ray and the sample surface was adjusted to 20°, and the X-ray fluorescence was detected from the direction normal to the incident beam in the plane of the electron orbit of the storage ring. The obtained XAFS spectra were processed using the XAFS processing software (REX2000 ver. 2.6: Rigaku Co. Ltd).
The Eh-pH diagram of sulfur was illustrated by the geochemical modeling software, Geochemist's Workbench 8.0 (Rock-Ware). The parameters used in this thermodynamic calculation were as follows: the activity of sulfate is set to 0.04 mmol L À1 . The pressure and temperature were set at 1.013 hPa and 25°C, respectively.

Pilot-scale experiment on the removal of H 2 S gas from domestic wastewater
A schematic of the pilot-scale experiment on the removal of H 2 S gas from domestic wastewater is shown in Supplementary Material, Figure S2. Domestic wastewater discharged from a condominium with 12 houses was introduced into 5.6 m 3 of an experimental tank 12 times per day to simulate domestic wastewater-derived H 2 S. The total volume of domestic wastewater discharged into the experimental tank and the hydrological residence time were 4 m 3 d À1 and 1.25 days, respectively. The domestic water composition was adjusted to biological oxygen demand: 200 mg L À1 , total nitrogen: 45 mg L À1 , total phosphorus: 5.0 mg L À1 , suspended solid: 160 mg L À1 , and alkalinity: 190 mg L À1 using urea, ethanol, sodium bicarbonate, and potassium dihydrogen phosphate (Guaranteed Reagent; Kishida Chemical). To control and adjust the concentration of suspended solid, toilet paper was added to domestic wastewater. The domestic wastewater-derived gas was introduced into a polyvinyl chloride column (100 mm inner diameter Â 385 mm length) filled with the adsorbent by a blower (EC-75S-R3A3: Showa Denki). The flow rate was controlled at a rate of 50 L min À1 using a flow monitor (Eggs DELTA, FLM30-10PCW: Oval). The data logger (GHS-8AT: Gastec) was installed at the front and rear stages of the column to measure the temperature and concentration of H 2 S gas derived from the domestic wastewater, and the temperature and concentration of the H 2 S gas at the front and rear of the column were measured every 10 min for a week. The average temperature and concentration of H 2 S gas derived from domestic wastewater were 14.4°C and 31.0 ppm, respectively.

Optimum calcination temperature
The calcination temperature of the adsorbent prepared in this study was optimized in the range of 300-1,000°C because the adsorption capacity of the adsorbent without calcination (raw material) for humid H 2 S gas decreased significantly to 59 g-H 2 S m À3 (Figure 1). Supplementary Material, Figure S3 shows the TG-DTA curves of the adsorbents. The decreases in the TG at approximately 120°C and 120-300°C correspond to hygroscopic water evaporation, and both crystallization of water and strong physically bonded water removal, respectively (Burduhos Nergis et al. 2020). Over 300°C, the TG decreases, which corresponds to the dehydroxylation of iron, silicon, and aluminum hydroxide groups (Landmesser et al. 1997;Fuji & Chikazawa 2000;Laskou et al. 2006;Burduhos Nergis et al. 2020;Wang et al. 2020). The decrease in TG at 550-750°C corresponds to CaCO 3 decompositions (Duan et al. 2017). Based on this, when the calcination temperature was greater than 300°C, the adsorption capacity of the adsorbent for H 2 S gas under humid conditions did not decrease when compared to that under dry conditions (Figure 1). This result indicates that the hydrophobicity of the adsorbent surface increased because of the formation of siloxane and aluminum oxide derived from silanol and alaninol group condensation. Hence, H 2 S could be adsorbed on the adsorbent without the adsorption interference of water. The adsorption capacity for H 2 S gas reached a peak under dry conditions at 600-700°C under dry conditions and at 700°C under humid conditions ( Figure 1). However, when the calcination temperature exceeded 700°C, the adsorption capacity for H 2 S gas significantly decreased. The significant decrease in the adsorption capacity for H 2 S gas was attributed to the decrease in the specific surface area of the adsorbent (Supplementary Material, Figure S4). Without calcination, the specific area of the adsorbent was 25.0 m 2 g À1 , and the values decreased gradually with increasing calcination temperature up to 700°C. At calcination temperatures over 700°C, the specific surface area of the adsorbent significantly decreased, corresponding to the decomposition of CaCO 3 (Supplementary Material, Figure S3). Thus, the optimized calcination temperature was 700°C, and the breakthrough curve of H 2 S gas obtained by the calcinated adsorbent at 700°C is also shown in Figure 2. The breakthrough capacity of the adsorbent was calculated to be 650 g-H 2 S m À3 under dry conditions and 730 g-H 2 S m À3 under humid conditions.

Adsorption mechanisms of H 2 S gas
In this study, sulfur species adsorbed on the prepared adsorbent were identified to reveal the adsorption mechanisms of H 2 S gas by XAFS (Figure 3 and Supplementary Material, Figure S5). Sulfate and elemental sulfur were detected on the adsorbent without calcination and the 500°C -calcined adsorbent. In the case of the 700°C and 1,000°C -calcined adsorbents, sulfate was identified. Sulfur and sulfate are considered to be derived from the oxidation of H 2 S (Asaoka et al. 2012). When the calcination temperature exceeded 700°C, H 2 S was completely oxidized to sulfate, indicating that the adsorption capacity of H 2 S gas reached a peak at 700°C. This enhancement of oxidation is considered to be caused by an increase in the surface pH of the adsorbent. Supplementary Material, Figure S6 shows the surface pH of the adsorbent calcined at each temperature. The surface pH of the adsorbent without calcination and the adsorbent calcinated at 300-500°C ranged from 8.9 to 9.5. When the calcination temperature was 600°C, the surface pH of the adsorbent increased by 10.4, reaching a plateau (pH 11.4-12.1) at a calcination temperature of 700-1,000°C. The increase in the surface pH corresponded to the transformation of CaCO 3 in the Journal of Water, Sanitation and Hygiene for Development Vol 11 No 5,827 prepared adsorbent to CaO. According to the Eh-pH diagram of sulfur, the redox boundary between sulfate and sulfide thermodynamically shifts toward low Eh with increasing pH (Supplementary Material, Figure S7), indicating that the sulfide on the adsorbent calcined at 700-1,000°C was completely oxidized to sulfate.

Pilot-scale experiments on the removal of H 2 S gas from domestic wastewater
Pilot-scale experiments on the removal of H 2 S gas derived from domestic wastewater is shown in Figure 4. The H 2 S gas concentration in the control (before a column) ranged from ,0.1-93 ppm. During the entire experimental period, the average removal percentage of H 2 S gas was 99.1%. The H 2 S gas concentration after the adsorbent column ranged from ,0.1-0.9 ppm until the cumulative amount of adsorbed H 2 S reached 8.4 g at 98.7 h, which corresponded to 98-100% removal of H 2 S gas. Thereafter, the H 2 S gas concentration after the adsorbent column ranged from ,0.1-4.7 ppm, with a removal percentage was 91.1-100%. The H 2 S gas removal performance of the adsorbent is considered to have silently decreased as the amounts of H 2 S increased.

CONCLUSIONS
The adsorbent for the removal of H 2 S gas was developed by mixing coal fly ash and blast furnace cement in a ratio of 87:13 by mass percentage. To achieve a high removal performance for both dry and humid H 2 S gas, the optimum calcination  Journal of Water, Sanitation and Hygiene for Development Vol 11 No 5,828 temperature was 700°C. For a week, the average removal percentage of H 2 S gas derived from domestic wastewater remained at 99.1%. The mechanism of H 2 S removal is primarily oxidation to sulfate. Therefore, the developed adsorbent for H 2 S gas in this study is one of the most promising cost-effective adsorbents for promoting the recycling of coal fly ash.