In view of the characteristics and risks of ammonia, its removal is important for industrial production and environmental safety. In this study, viscose-based activated carbon fiber (ACF) was used as a substrate and chemically modified by nitric acid impregnation to enhance the adsorption capacity of the adsorbent for ammonia. A series of modified ACF-based adsorbents were prepared and characterized using BET, FTIR, XPS, and Boehm titration. Isotherm tests (293.15 K, 303.15 K, 313.15 K) and dynamic adsorption experiments were performed. The characterization results showed that impregnation with low concentrations of nitric acid not only increased the surface acidic functional group content but also increased the specific surface area, while impregnation with high concentrations of nitric acid could be able to decrease the specific surface area. ACF-N-6 significantly increased the surface functional group content without destroying the physical structure of the activated carbon fibers. The experimental results showed that the highest adsorption of ammonia by ACFs was 14.08 mmol-L⁻¹ (ACF-N-6) at 293 K, and the adsorption capacity was increased by 165% compared with that of ACF-raw; by fitting the adsorption isotherm and calculating the equivalent heat of adsorption and thermodynamic parameters using the Langmuir–Freundlich model, the adsorption process could be found to exist simultaneously. Regarding physical adsorption and chemical adsorption, the results of the correlation analysis showed that the ammonia adsorption performance was strongly correlated with the carboxyl group content and positively correlated with the relative humidity (RH) of the inlet gas. This study contributes to the development of an efficient ammonia adsorption system with important applications in industrial production and environmental safety.

A composite packing material (CM-5) was prepared in this study, mainly consisting of compost with functional microorganisms, calcium carbonate (CaCO₃), perlite, cement and plant fiber. To get stronger compressive strength, mass ratios of these components were optimized based on single factor experiments, and finally adding amounts of perlite, cement, plant fiber, CaCO₃, compost and binder at 18%, 18%, 7%, 13%, 17% and 27%, respectively. According to the optimum proportion, CM-5 was extruded in cylindrical shape (12 mm in diameter and 20 mm in length) with a bulk density of 470 kg m⁻³, a moisture retention capacity of 49% and the microbial counts of × 10⁵ CFU g⁻¹ of packing material. The cumulative release rates of total organic carbon (TOC) and total nitrogen (TN) from CM-5 were 3.1% and 6.5%, respectively, after 19 times extraction in distilled water. To evaluate the H₂S removal capacity, CM-5 was compared with an organic (corncob) and an inorganic (ceramsite) packing material in three biofilters. The results showed that CM-5 had higher H₂S removal capacity compared with corncob and ceramsite. CM-5 could avoid the large fluctuation of pH value and pressure drop during the operation. The maximum H₂S removal capacity of CM-5 was 12.9 g m⁻³ h⁻¹ and the removal efficiency could maintain over 95.4% when the inlet H₂S loading rate was lower than 11.3 g m⁻³ h⁻¹ without any addition of nutrients and pH buffer substances. Besides, only 2–3 days were needed for the recovery of biofiltration performance after about two weeks of idle period.

The performances of two identical biofilters, filled with a new composite packing material (named CM-5) embedded with functional microorganisms or sterilized CM-5 without microorganisms, were investigated for H₂S treatment. Running parameters in terms of microbial counts, pressure drops, and inlet and outlet H₂S concentrations were measured. The results show that the microbial count of the CM-5 was approximately ×10⁵ CFU/g before being filled into the biofilter, while that of the sterilized CM-5 was negligible. The functional microorganisms embedded in CM-5 adapted to the environment containing H₂S quickly. In most cases, pressure drops of the CM-5 biofilter were slightly higher than those of the sterilized CM-5 biofilter when the gas flow rate was 0.6–2.5 m³/h. The maximum elimination capacity (EC) of the CM-5 biofilter in treating H₂S could reach up to 65 g/(m³·h) when the loading rate (LR) was approximately 80 g/(m³·h). If the LR was much higher, the measured EC showed a slight downward trend. The experimental ECs of biofilters were fitted by two typical dynamic models: the Michaelis-Menten model and the Haldane model. Compared with the Michaelis-Menten model, the Haldane model fit the experimental ECs better for the two biofilters because of the presence of the substrate inhibition behaviour.