Preparation of Self-Releasing Carbon Bioﬁlm Carrier Based on Corncob and Denitriﬁcation Properties

: Wastewater with a low carbon/nitrogen (C/N) ratio is widespread and difﬁcult to treat. The addition of an external carbon source is an effective method for treating such wastewater. Therefore, we aimed to prepare a self-releasing carbon bioﬁlm carrier using agricultural waste (corncobs), polyvinyl alcohol, and sponge iron in various ratios to provide a carbon source that would facilitate denitriﬁcation, providing an optimal environment for microorganisms. We found that the carbon release of the MAC bioﬁlm carrier that accumulated over 60 d was 116.139 mg of chemical oxygen demand (COD) · g − 1 , whereas the accumulated total nitrogen release was approximately 0 mg · (g · d) − 1 . The NO 3 − -N removal rate after 24 h reached 98.1%, whereas the theoretical use rate of the carbon source (in terms of COD) was stable at 90.34%. In addition, the sum of the abundances of the denitrifying and cellulose-degrading bacteria was 49.89%. Furthermore, bioﬁlm carriers are used as functional carriers that contribute to cellulose degradation, a process in which sponge iron produces Fe 2+ to provide electron donors and shuttles for denitrifying bacteria and forms the iron cycle, thereby inducing an increase in microbial abundance; this increase then facilitates the microbial degradation of cellulose and synergistic denitriﬁcation through interspeciﬁc bacterial cooperation. This study provides a new and effective method for enhancing the denitriﬁcation of wastewater with low C/N ratios.


Introduction
Nitrogen pollution is a challenging global environmental issue that causes the eutrophication of water bodies and threatens freshwater ecosystem functions, aquaculture safety, and public health; therefore, the treatment of nitrogen-containing wastewater has become an important aspect of ensuring environmental sustainability [1,2]. Biological nitrogen removal technologies have been extensively developed in the field of water treatment because of their low-cost and pollution-free products, mainly including full nitrification, annamox, and partial nitrification [3], which have become the preferred processes for treating wastewater with a low carbon-to-nitrogen (C/N) ratio, owing to their economic and environmental advantages [4]. However, conventional biological denitrification technologies, such as A 2 /O and membrane biogroups, are inefficient [5], and the efficiency of nitrogen removal must be increased by adding additional carbon sources [6]. Traditional carbon sources include methanol, ethanol, sodium acetate, and glucose [7], but they are expensive, and controlling the dosage is difficult [8]. Therefore, combining a solid slow-release carbon source with a biofilm is one of the most efficient methods for managing wastewater with a low C/N ratio [9]. Common biofilm carriers include gravel, bioceramsite, and polyurethane sponges [10], but they do not release carbon. At present, solid carbon sources, such as agricultural waste and synthetic biodegradable polymers, are being considered Therefore, the addition of Fe-containing substances to carriers can improve denitrification [31]. Li et al. [32] loaded iron powder onto corncobs, and the maximum nitrate removal efficiency was 99.30%, reaching 81.73% without a carbon source. Therefore, a combination of corncobs, PVA, and sponge iron can enhance denitrification using this reasonably modified and combined method.
Biofilms are microbial aggregates embedded in a self-produced matrix of extracellular polymeric substances (EPSs) [33]. A mature biofilm formed on the carrier determines the overall performance of a bioreactor in the context of water treatment. To enhance denitrification, it is necessary to develop strategies that enhance biofilm formation on biofilm carriers [34]. Owing to their rough surface and large specific surface area, corncobs are excellent film-hanging carriers [35]. Ma et al. [36] observed that the porous surface structure of alkali-treated corncobs facilitated bacterial attachment. Furthermore, the bridging effect between metal ions and EPSs contributes to microbial aggregation [37]; Liang et al. [38] found that waste iron shavings promote the secretion of EPSs and L-EPSs in the system, processes that are conducive to biofilm formation and material transfer. Therefore, alkali-treated corncob structures and sponge iron accelerated biofilm formation and promoted denitrification in biofilms.
The application of biofilm carriers with self-releasing carbon capabilities in biofilm groups for wastewater treatment has not been extensively studied. Therefore, in this study, a spherical biofilm carrier with a self-releasing carbon capability was fabricated by loading sponge iron and PVA onto corncobs from agricultural waste, as the material and skele-Fermentation 2023, 9,796 3 of 18 ton, respectively. The biofilm carrier provided a carbon source for denitrification and a growth environment for microorganisms, whereas sponge iron provided and increased the number of electron donors and electron shuttles, promoted the activity of bacteria, and enriched the microbial community structure. The effects of the three pretreatment methods on the release performance of chemical oxygen demand (COD), total nitrogen (TN), static denitrification characteristics, and biofilm characteristics of the biofilm carriers were investigated, followed by an analysis of the microbial communities and a discussion of the nitrogen removal mechanism of the self-releasing carbon biofilm carriers. The biofilm carrier proposed in this study provides a new effective strategy for nitrogen removal from low C/N wastewater treated with biofilm groups, which will benefit future basic theoretical research and practical applications of nitrogen removal from self-releasing carbon biofilm carriers.

Self-Releasing Carbon Biofilm Carrier Preparation
Corncobs (60 mesh) and sponge iron (1-3 mm in diameter (95.0-98.0% iron content)) were collected from Lianyungang (Jiangsu, China). PVA (average degree of polymerization, 1799 ± 50) was purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). The dried corncobs were processed via the following 3 methods: (1) the corncobs were mechanically crushed and soaked in distilled water for 12 h; (2) the corncobs were mechanically crushed and soaked in 1% NaOH for 12 h; (3) the corncobs were mechanically crushed and biologically treated for 48 h (the fermenting bacteria were mainly Bacillus subtilis, Lactobacillus, etc., of which Bacillus subtilis were ≥5 × 107 CFU·g −1 ). The corncobs were mechanically crushed to a particle size of 0.5-1.0 mm, washed, dried, and prepared for use. The corncob powder treated via one of the above three methods was fixed with PVA, as a binder, and sponge iron to form a spherical biofilm carrier that was 15 ± 0.2 mm in diameter (corncob powder:PVA:sponge iron component ratio was 3:1:0.3), named the MC, MCA, and MCB biofilm carriers, respectively. A structural diagram of the biofilm carriers is shown in Figure S1.

Experimental Setup and Operation
In the static experiment of carbon release, the MC, MCA, or MCB biofilm carrier was added to a 250 mL brown bottle with 200 mL distilled water. The brown bottle remained at room temperature (20-25 • C). Parallel leach solution samples were removed from the brown bottle every three days, and the leached solution was completely replaced with 200 mL of distilled water. COD and TN were measured to study the release characteristics of the carbon sources. Dissolved organic matter (DOM) was analyzed using three-dimensional fluorescence spectroscopy.
We measured the water-quality-related metrics of real wastewater and used this real wastewater sample as a reference for synthetic wastewater. The synthetic wastewater comprised NaNO 3 and KH 2 PO 4 (N:P = 5:1). Trace elements were added to the synthetic wastewater samples. The mass concentration of NO 3 − -N and total phosphorus were 100 mg·L −1 and 20 mg·L −1 in the synthetic wastewater, respectively. Additional carbon sources in real wastewater can promote denitrification. However, this study lacked the ability to remove NH 4 + -N from real wastewater. We placed 150 mL of denitrifying sludge and 300 mL of domestic wastewater into a 500 mL brown bottle for a sludge domestication culture for 4 d with 100 mg·L −1 NO 3 − -N. The denitrifying sludge, at a concentration of 11,487.5 mg·L −1 , was taken from the secondary sedimentation tank of a sewage treatment plant in Lanzhou, China.
In the static denitrification experiment, the MC, MCA, or MCB biofilm carrier was added to a 500 mL brown bottle with 450 mL of synthetic wastewater and 50 mL of domesticated denitrified sludge and labelled as the MC, MCA, and MCB groups, respectively. The control group (CG) comprised bottles without carbon sources. We firmly sealed the vial and cultured it in an incubator at 20 • C and 60 r·min −1 . COD, TN, NO 3 − -N, NO 2 − -N, An alkaline potassium persulfate digestion spectrophotometric method (UVmini1240; Shimadzu, Kyoto, Japan) was used to test the TN. Thymol spectrophotometry (UVmini1240; Shimadzu, Japan) was used to measure the NO 3 − -N concentration. The NH 4 + -N concentration was determined using the nascent reagent photometric method (Prism Light 721G; Shanghai, China). The NO 2 − -N concentration was determined using the N-(1naphthalene)-diaminoethane colorimetric method (Prism Light 721G, Shanghai, China). The COD concentration was determined using a COD Fast Analyzer (Sheng Aohua 6B-200, Nanjing, China).
The fluorescence excitation-emission matrix (EEM) was measured using a fluorescence spectrophotometer (F7000, Hitachi Co., Tokyo, Japan) at an excitation (Ex) wavelength, emission (Em) wavelength, scanning interval, and a scanning speed of 200−400 nm, 250−550 nm, 5 nm, and 12,000 nm·min −1 , respectively. The morphologies of the MCA biofilm carrier surface and the biofilm attached to the MCA biofilm carrier were characterized using scanning electron microscopy (SEM; ZEISS Gemini 500, Oberkochen, Germany). The pretreatment protocol for the biofilm carrier samples that was applied is that described by Wang et al. [34]. X-ray diffraction (XRD; Ultimate 4, Rigaku, Japan) was used to analyze the properties of the functional groups and crystal structures.

Data Analysis
A second-order kinetic equation was used to characterize carbon release, as shown in Equation (1).
where C t is the accumulated COD at time t, mg·(g·L) −1 ; C m is the ultimate accumulated COD, mg·(g·L) −1 ; k 2 is the constant of carbon release, g·L·(d·mg) −1 ; K is the mass transfer coefficient, mg·(g·L·d) −1 ; and t 1/2 is the time needed for the carbon to release half of its maximum concentration. The Ritger-Peppas equation is as follows [22]: where C t is the accumulated COD at time t, mg·(g·L) −1 ; C m is the ultimate accumulated COD, mg·(g·L) −1 ; k is the constant of carbon release; and N is the carbon release index, which reflects the mechanism of carbon release. The diffusion process was dominant when N was <0.45, the diffusion and disintegration processes were dominant when N ranged from 0.45 to 0.89, and the disintegration process was dominant when N was >0.89.

Analysis of Dissolved Organic Matter
According to Wang et al. [39], 3D fluorescence spectra can be divided into six regions. Specifically, regions I and II represent the tyrosine, tryptophan, tyrosine, and tryptophan proteins, respectively, whereas regions III, IV, V, and VI represent polysaccharides, polycarboxylic acid, humic acid, polycyclic aromatic hydrocarbon humic acid, and fulvic acid, respectively.
The reduction of 1 mg of NO 3 − to NO 2 − requires 1.15 mg of COD, whereas the complete reduction to N 2 requires 2.86 mg of COD. The theoretical consumption and use of COD can be determined by calculating the mass of NO 3 − -N removed and amount of NO 2 − -N produced.

Static Release Performance of Biofilm Carrier
The variation in COD concentration in the soaking solutions of the three types of biofilm carriers is shown in Figure 1a. The amount of carbon released by the three types of biofilm carriers showed the same variation trend, which can be divided into a rapid period of carbon release (1-3 days), a fast-release period (4-12 days), and a stable period of carbon release (13-60 days). Comparing the carbon release rates of the three biofilm carriers, we found MCA > MC > MCB; the cumulative amounts released by the three biofilm carriers were 116.139, 93.200, and 78.079 mg·g −1 , respectively. The MCA biofilm carrier showed a better carbon release effect. The rapid and fast release phases are due to the rapid dissolution of water-soluble organic matter on the surface of the corncob; the slow release phase is due to the slow hydrolysis of insoluble cellulose and hemicellulose into soluble organic matter by microorganisms when the small molecules inside the corncob are depleted and the carbon source is slowly released [40]. As a biofilm carrier skeleton, PVA equalizes the rate of the entire release process, prevents excessive carbon release in the initial stage, and guarantees sufficient carbon release in later stages [22]. Maize cobs are mechanically crushed and alkali pretreated to destroy their lignin structure, increase microbial availability, improve the COD release capacity of the corncob carbon source, and maintain the stability of the later carbon release phase [15]. Therefore, MCA biofilm carriers have a more stable carbon release capacity and can ensure the supply of the carbon source required by microorganisms. According to Wang et al. [39], 3D fluorescence spectra can be divided into six regions. Specifically, regions I and II represent the tyrosine, tryptophan, tyrosine, and tryptophan proteins, respectively, whereas regions III, IV, V, and VI represent polysaccharides, polycarboxylic acid, humic acid, polycyclic aromatic hydrocarbon humic acid, and fulvic acid, respectively.
The reduction of 1 mg of NO3 − to NO2 − requires 1.15 mg of COD, whereas the complete reduction to N2 requires 2.86 mg of COD. The theoretical consumption and use of COD can be determined by calculating the mass of NO3 − -N removed and amount of NO2 − -N produced.

Static Release Performance of Biofilm Carrier
The variation in COD concentration in the soaking solutions of the three types of biofilm carriers is shown in Figure 1a. The amount of carbon released by the three types of biofilm carriers showed the same variation trend, which can be divided into a rapid period of carbon release (1-3 days), a fast-release period (4-12 days), and a stable period of carbon release (13-60 days). Comparing the carbon release rates of the three biofilm carriers, we found MCA > MC > MCB; the cumulative amounts released by the three biofilm carriers were 116.139, 93.200, and 78.079 mg·g −1 , respectively. The MCA biofilm carrier showed a better carbon release effect. The rapid and fast release phases are due to the rapid dissolution of water-soluble organic matter on the surface of the corncob; the slow release phase is due to the slow hydrolysis of insoluble cellulose and hemicellulose into soluble organic matter by microorganisms when the small molecules inside the corncob are depleted and the carbon source is slowly released [40]. As a biofilm carrier skeleton, PVA equalizes the rate of the entire release process, prevents excessive carbon release in the initial stage, and guarantees sufficient carbon release in later stages [22]. Maize cobs are mechanically crushed and alkali pretreated to destroy their lignin structure, increase microbial availability, improve the COD release capacity of the corncob carbon source, and maintain the stability of the later carbon release phase [15]. Therefore, MCA biofilm carriers have a more stable carbon release capacity and can ensure the supply of the carbon source required by microorganisms. To understand the static carbon release mechanism of the three biofilm carriers, the kinetic fitting method of Lv [41] and Ritger [42] was used to fit the carbon release process of the three biofilm carriers (Table 1). In the analysis of the quasi-secondary kinetic fit results, the larger the Cm, the larger the total final release of COD from the biofilm carrier. The larger the K, the lower the mass transfer resistance of the biofilm carrier and the smaller the t1/2, indicating that the biofilm carrier releases a large amount of carbon in a short period [41]. The results showed that the minimum value of t1/2 for the MCB biofilm To understand the static carbon release mechanism of the three biofilm carriers, the kinetic fitting method of Lv [41] and Ritger [42] was used to fit the carbon release process of the three biofilm carriers (Table 1). In the analysis of the quasi-secondary kinetic fit results, the larger the C m , the larger the total final release of COD from the biofilm carrier. The larger the K, the lower the mass transfer resistance of the biofilm carrier and the smaller the t 1/2 , indicating that the biofilm carrier releases a large amount of carbon in a short period [41]. The results showed that the minimum value of t 1/2 for the MCB biofilm carrier was 3.15 d, indicating that the carbon release rate was too fast in a short time, and the saturation concentration was not reached. Therefore, the release of carbon sources during the later period may have been insufficient. The values of the indices for the MC biofilm carriers were intermediate compared to those of the other two. The C m and t 1/2 of the MCA biofilm carrier were greater, reaching 666.67 mg·(d·g·L) −1 and 9.45 d, respectively. The Ritger-Peppas-calculated N release indices of the three biofilm carriers were 0.279, 0.316, and 0.155, respectively, all less than 0.450. The carbon release occurred via a diffusion process, indicating that the dissolved organic matter in the three biofilm carriers could be leached through their pores of the biofilm carriers. The results showed that the MCA biofilm carrier contained the most soluble small molecules of organic matter and was the most stable during release, which is congruent with the carbon release results. The release of TN from the three biofilm carriers is shown in Figure 1b. TN showed a release process similar to that of carbon release, and all experienced a rapid release period (1-5 d) to reach the maximum value and then entered a stable release period (6-20 d). The release rates of the three biofilm carriers were close to 0 mg·(g·d) −1 after 20 d. In addition, the cumulative TN released by the three biofilm carriers was 0.064, 0.046, and 0.036 mg·g −1 for the MC, MCB, and MCA biofilm carriers, respectively. These values were similar to those reported by Zou et al. [40]. Proteins in cellulose can be dissolved in water, resulting in an increased nitrogen content in soaking solutions [43]. The hydrolysis of hard-to-degrade carbon sources reduced and stabilized the nitrogen content in the soaking solution during the late stage of carbon release [44], and biological and 1% NaOH alkali treatments also reduced the nitrogen in the corncob. The above results show that the accumulated TN release in the MCA biofilm carrier was the lowest.
The EEM of the soaking solutions for the three biofilm carriers are shown in Figure S2. Two distinct peaks, peak A (Ex/Em 220-225/310-320) and peak B (Ex/Em 275/320-330), representing tyrosine/tryptophan and tyrosine/tryptophan proteins, respectively, were observed in the EEM of the MC, MCA, and MCB biofilm carriers soaked in solutions for 13-15 days and 28-30 days. Tyrosine-and tryptophan-like substances are aromatic proteinlike substances that are easily biodegraded; the latter are easily used by microorganisms in the biodegradation process [45]. The results showed that the main component of corncob leachate was tryptophan (a protein-like substance), which can be used by denitrifying bacteria [46].
The integral standard volumes of the regions in which the four fluorescence peaks were displayed by EEM were calculated using the FRI method [45] (Table S1). With increasing time, the integral standard volume of the soaking solution area of the MC and MCB biofilm carriers decreased substantially, whereas that of the MCA biofilm carriers did not change considerably, indicating that the release of tyrosine/tryptophan-like substances from the MCA biofilm carriers was more uniform and stable, which is in agreement with the results of the COD release rate in the static carbon release experiment of the biofilm carriers. In summary, the carbon release capacity of the carbon source in the soaking solution of each biofilm carrier followed the order MCA > MC > MCB. Thus, MCA can be used as a carbon source.   Figure 2a shows the variations in the NO 3 − -N removal rates of the MC, MCA, MCB, and control groups. The MC, MCA, and MCB groups achieved more than 90% removal by the 30th day at an NO 3 − -N mass concentration of 100 mg·L −1 . The MC, MCA, and MCB biofilm carriers easily degrade organic matter to provide direct carbon sources for microorganisms [47]. Moreover, sponge iron produces iron ions that enhance heterotrophic microbial activity and provide electron donors and shuttles for denitrifying bacteria, further enhancing denitrification [29]. The removal of NO 3 − -N varied over time, with the removal of NO 3 − -N starting to decrease after 39 days in the MC and MCB groups and rapidly decreasing to 62.5% after 30 days in the CG. With the gradual maturation of the biofilm and stable use of the carbon source, denitrification in the MCA group entered a stable phase, and the removal rate continued to increase within 45 days. The highest NO 3 − -N removal rate was 98.1% after 45 d. This occurred because of the destruction of the lignin structure of the corncobs via mechanical shredding and alkali treatment, which was more easily degraded by the microorganisms attached to the surface of the biofilm carrier. This improves the carbon release capacity and provides a sufficient carbon source for denitrifying bacteria [36].

Denitrification Performance
The NO3 − -N and TN removal rates and changes in the TN, NO3 − -N, NO2 − -N, and NH4 + -N mass concentrations of the groups with different biofilm carriers are shown in Figure 2. In this study, MCA, MC, MCB, and control groups were used to determine the suitability of various biofilm carriers. Figure 2a shows the variations in the NO3 − -N removal rates of the MC, MCA, MCB, and control groups. The MC, MCA, and MCB groups achieved more than 90% removal by the 30th day at an NO3 − -N mass concentration of 100 mg·L −1 . The MC, MCA, and MCB biofilm carriers easily degrade organic matter to provide direct carbon sources for microorganisms [47]. Moreover, sponge iron produces iron ions that enhance heterotrophic microbial activity and provide electron donors and shuttles for denitrifying bacteria, further enhancing denitrification [29]. The removal of NO3 − -N varied over time, with the removal of NO3 − -N starting to decrease after 39 days in the MC and MCB groups and rapidly decreasing to 62.5% after 30 days in the CG. With the gradual maturation of the biofilm and stable use of the carbon source, denitrification in the MCA group entered a stable phase, and the removal rate continued to increase within 45 days. The highest NO3 − -N removal rate was 98.1% after 45 d. This occurred because of the destruction of the lignin structure of the corncobs via mechanical shredding and alkali treatment, which was more easily degraded by the microorganisms attached to the surface of the biofilm carrier. This improves the carbon release capacity and provides a sufficient carbon source for denitrifying bacteria [36]. As shown in Figure 2b, the TN removal rate and mass concentration changed, exhibiting a removal trend similar to that of NO3 − -N. The removal rates of the MCA group on day 12 and the MC and MCB groups on day 30 were greater than 90%. The TN removal rates in the MC and MCB groups began to decrease after 39 days. The removal rate of the MCA group continued to increase; the removal rate of TN was stable at 45 d, and the removal efficiency reached 97.3% on the 45th d. The CG had a certain TN removal rate As shown in Figure 2b, the TN removal rate and mass concentration changed, exhibiting a removal trend similar to that of NO 3 − -N. The removal rates of the MCA group on day 12 and the MC and MCB groups on day 30 were greater than 90%. The TN removal rates in the MC and MCB groups began to decrease after 39 days. The removal rate of the MCA group continued to increase; the removal rate of TN was stable at 45 d, and the removal efficiency reached 97.3% on the 45th d. The CG had a certain TN removal rate within 45 d. The removal rate was 50.7% in the first three days, and the highest removal rate was 86.3% on the 30th day, after which the removal rate dropped rapidly to 62.5%. The decomposition of microorganisms in the later stages released nitrogen and P-containing substances, resulting in an increase in the TN mass concentration of the solution instead of a decrease. Both the group with the added biofilm carrier and CG showed a rapid decrease in the NO 3 − -N mass concentration and a small accumulation of NO 2 − -N from days 1 to 6, as shown in Figure 2c. The accumulation of NO 2 − -N in the group with added biofilm carriers gradually decreased after six days. The MCA group had the smallest average accumulation of NO 2 − -N at 0.037 mg·L −1 ; NO 2 − -N accumulation occurs rapidly during the initial phase of denitrification when carbon sources are insufficient, mainly due to the stronger competition for carbon sources by nitrate reductase than nitrite reductase [48]. As denitrification progressed, the decomposition and utilization of corncobs by microorganisms increased, the carbon source became relatively sufficient, and the NO 2 − -N concentration gradually decreased. The accumulation of NO 2 − -N began to appear again in the CG at 33-45 days. This was because no external electron donor was added to the control group, which easily caused the accumulation of NO 2 − -N when the electron donor was insufficient. The changes in the NH 4 + -N mass concentration during the experiment are shown in Figure 2d. The accumulation of NH 4 + -N occurred in the early stage, and the mass concentration of NH 4 + -N in the group with the addition of biofilm carriers was higher than that in the CG at days 1−9, up to 0.633 mg·L −1 , which was due to the growth and metabolism of the microorganisms. According to Kim and Chao [49], in the presence of Fe 2+ , nitrate is partially converted into ammonia nitrogen via abiotic and biotic reactions. The NH 4 + -N mass concentration tended to stabilize at approximately 0.073 mg·L −1 from day 10 to 45. The MCA group had the lowest NH 4 + -N mass concentration, indicating that the alkali treatment could reduce the amount of NH 4 + -N released from the corncobs. The CG had a much higher NH 4 + -N mass concentration than the group with biofilm carriers from days 10 to 45. This was because the CG had no external carbon source to feed the microorganisms, which may have led to microbial death and the release of NH 4 + -N. In summary, using the MCA biofilm carrier as a carbon source resulted in a more than 97% NO 3 − -N and TN removal after the 45th day. The average NO 2 − -N accumulation was 0.037 mg·L −1 , and the NH 4 + -N mass concentration was low, indicating that MCA, as a carbon source and biofilm carrier, can support the growth and metabolism of denitrifying bacteria and has a strong denitrification effect. Figure 3 shows the changes in COD during denitrification for different biofilm carriers and the COD use rate of the MCA group. As shown in Figure 3a, the COD of the effluent from the denitrification experiments of the groups with different biofilm carriers rapidly decreased from days 1 to 9, indicating that large quantities of small-molecule organic matter attached to the biofilm carriers were used by the microorganisms, thereby demonstrating a strong denitrification effect. The COD of the effluent from the denitrification experiments at 10-45 days was stable, and the mean COD of the effluent from the denitrification experiments for the MC, MCA, and MCB groups was 15.412, 29.690, and 19.877 mg·L −1 , respectively. After the rapid release of the carbon source, the organic matter in the biofilm carrier was slowly released through microbial degradation. Simultaneously, the released carbon source was used by various microorganisms attached to the surface of the corncob, resulting in low COD in the experimental effluent. Carbon release and denitrification were balanced, and the MCA biofilm carrier synthesized from mechanically crushed and alkali-treated corncob powder provided a sufficient carbon source for denitrification. The results of this study showed that MCA, as an additional carbon source and biofilm carrier, prevents additional pollution.  Hu et al. [50] used corn husk as a denitrifying solid slow-release carbon source; the average effluent COD was 20-50 mg·L −1 in their experiment. Tang et al. [51] used PCL and sulfur as a composite carbon source. The COD of the effluent was 40 mg·L −1 and it fluctuated greatly. The release of COD from denitrification was greater than that from static carbon, indicating that the cellulose and lignin in corncobs are easily released via microbial decomposition. The COD of the CG ranged from 0 to 20.271 mg·L −1 during the experimental cycle, which was due to the lack of an external carbon source to feed these microorganisms, resulting in some microorganisms cycling between death and degradation by other surviving microorganisms, thus causing the fluctuating COD. According to the carbon balance principle, the theoretical reduction of 1 mg of NO3 − -N requires the consumption of 2.86 mg of COD; the theoretical consumption of COD can be obtained by calculating the removal of NO3 − -N. As shown in Figure 3b, the theoretical use rate of COD in the MCA group increased from 53.21% to 94.25% from days 1 to 12, and was stable at 90.34% from days 12 to 45. This showed that the use of microorganisms released from the MCA biofilm carriers as the carbon source was high during denitrification, which reduced the residual COD, owing to insufficient use of the carbon source.

MCA Biofilm Carrier Surface Characteristics
The surface structure of biofilm carriers strongly affects the attachment and growth of functional bacterial communities [44]. The results of the SEM analysis of the surface morphology of the MCA biofilm carriers before and after microbial degradation in Figure  4a,b show that the surface of the MCA biofilm carrier before use was relatively rough, with more pores, and the distribution of pores was relatively uniform. Mechanical crushing and alkali treatment effectively destroyed the closed structure of the original corncob surface, forming effective release pores, thereby promoting the consistent and steady release of carbon sources, making it more suitable for microbial attachment and growth. As shown in Figure 4c,d, the surface of the MCA biofilm carrier was destroyed after degradation by the microorganisms. The morphological characteristics of the biofilm and the degradation of biofilm carriers by microorganisms were observed. The biofilm contained abundant microbial species, including not only filamentous and short rod-shaped bacteria but also cocci and long rod-shaped bacteria. In addition, the corncobs maintained a stable physical structure, which could be related to components such as crude fiber or lignin, which are not easily biodegradable. These results indicate that the MCA biofilm carriers provided porous structures with more space for bacterial attachment and growth, effectively enhanced long-term organic carbon release, and promoted bacterial enrichment. Hu et al. [50] used corn husk as a denitrifying solid slow-release carbon source; the average effluent COD was 20-50 mg·L −1 in their experiment. Tang et al. [51] used PCL and sulfur as a composite carbon source. The COD of the effluent was 40 mg·L −1 and it fluctuated greatly. The release of COD from denitrification was greater than that from static carbon, indicating that the cellulose and lignin in corncobs are easily released via microbial decomposition. The COD of the CG ranged from 0 to 20.271 mg·L −1 during the experimental cycle, which was due to the lack of an external carbon source to feed these microorganisms, resulting in some microorganisms cycling between death and degradation by other surviving microorganisms, thus causing the fluctuating COD. According to the carbon balance principle, the theoretical reduction of 1 mg of NO 3 − -N requires the consumption of 2.86 mg of COD; the theoretical consumption of COD can be obtained by calculating the removal of NO 3 − -N. As shown in Figure 3b, the theoretical use rate of COD in the MCA group increased from 53.21% to 94.25% from days 1 to 12, and was stable at 90.34% from days 12 to 45. This showed that the use of microorganisms released from the MCA biofilm carriers as the carbon source was high during denitrification, which reduced the residual COD, owing to insufficient use of the carbon source.

MCA Biofilm Carrier Surface Characteristics
The surface structure of biofilm carriers strongly affects the attachment and growth of functional bacterial communities [44]. The results of the SEM analysis of the surface morphology of the MCA biofilm carriers before and after microbial degradation in Figure 4a,b show that the surface of the MCA biofilm carrier before use was relatively rough, with more pores, and the distribution of pores was relatively uniform. Mechanical crushing and alkali treatment effectively destroyed the closed structure of the original corncob surface, forming effective release pores, thereby promoting the consistent and steady release of carbon sources, making it more suitable for microbial attachment and growth. As shown in Figure 4c,d, the surface of the MCA biofilm carrier was destroyed after degradation by the microorganisms. The morphological characteristics of the biofilm and the degradation of biofilm carriers by microorganisms were observed. The biofilm contained abundant microbial species, including not only filamentous and short rod-shaped bacteria but also cocci and long rod-shaped bacteria. In addition, the corncobs maintained a stable physical structure, which could be related to components such as crude fiber or lignin, which are not easily biodegradable. These results indicate that the MCA biofilm carriers provided porous structures with more space for bacterial attachment and growth, effectively enhanced long-term organic carbon release, and promoted bacterial enrichment. The X-ray diffraction patterns of the sponge iron in the MCA biofilm carrier before and after use are shown in Figure S3. The main components of sponge iron are Fe 0 and FeO, which continuously provide an Fe source for the biological denitrification process and promote denitrification [24]. The presence of FeO, Fe2O3, and FeCO3 after denitrification indicates that sponge iron is involved and promotes the denitrification intensity [52]. Fe 0 and Fe 2+ play important roles in promoting the reduction of NO3 − -N, facilitating electron transfer on the surface of the biocarrier, and improving denitrification efficiency [53]. The results show that MCA biofilm carriers have a strong biofilm attachment effect and can be used as denitrification filter fillers.

Microorganism Community Analysis
High-throughput sequencing was performed on five sample groups: MC, MCA, MCB, CG, and DS. The diversity and richness of the microbial communities in these groups were then analyzed. The Operational Taxonomic Unit (OTU) dilution curves for each sample are shown in Figure S4. The richness and diversity of the microbial communities in the different samples were calculated using the OTU method, and the results are shown in Table S2. The Chao, Ace, Shannon, and Simpson indices for the MC, MCA, MCB, CG, and DS groups indicated that the richness and diversity of microorganisms in the MC, MCA, and MCB groups were greater than those in the CG and DS. Figure 5a shows that the main phyla in each group were Proteobacteria, Bacteroidota, Actinobacteriota, and Firmicutes. The microbial community diversity of the groups with the biofilm carriers was higher than that of the CG and DS groups. Bacteroidota is a group of bacteria that accelerates the hydrolytic uptake of carbon sources [54]. Actinobacteriota is a group of denitrifying bacteria that promotes the removal of nitrate and nitrite [55]. Firmicutes perform metabolic activities by degrading large organic molecules, such as cellulose, proteins, and polysaccharides [47,56]. The relative abundances of Bacteroidota, Actinobacteriota, and Firmicutes increased the most in the MCA group compared with those in the CG group, with increases The X-ray diffraction patterns of the sponge iron in the MCA biofilm carrier before and after use are shown in Figure S3. The main components of sponge iron are Fe 0 and FeO, which continuously provide an Fe source for the biological denitrification process and promote denitrification [24]. The presence of FeO, Fe 2 O 3 , and FeCO 3 after denitrification indicates that sponge iron is involved and promotes the denitrification intensity [52]. Fe 0 and Fe 2+ play important roles in promoting the reduction of NO 3 − -N, facilitating electron transfer on the surface of the biocarrier, and improving denitrification efficiency [53]. The results show that MCA biofilm carriers have a strong biofilm attachment effect and can be used as denitrification filter fillers.

Microorganism Community Analysis
High-throughput sequencing was performed on five sample groups: MC, MCA, MCB, CG, and DS. The diversity and richness of the microbial communities in these groups were then analyzed. The Operational Taxonomic Unit (OTU) dilution curves for each sample are shown in Figure S4. The richness and diversity of the microbial communities in the different samples were calculated using the OTU method, and the results are shown in Table S2. The Chao, Ace, Shannon, and Simpson indices for the MC, MCA, MCB, CG, and DS groups indicated that the richness and diversity of microorganisms in the MC, MCA, and MCB groups were greater than those in the CG and DS. Figure 5a shows that the main phyla in each group were Proteobacteria, Bacteroidota, Actinobacteriota, and Firmicutes. The microbial community diversity of the groups with the biofilm carriers was higher than that of the CG and DS groups. Bacteroidota is a group of bacteria that accelerates the hydrolytic uptake of carbon sources [54]. Actinobacteriota is a group of denitrifying bacteria that promotes the removal of nitrate and nitrite [55]. Firmicutes perform metabolic activities by degrading large organic molecules, such as cellulose, proteins, and polysaccharides [47,56]. The relative abundances of Bacteroidota, Actinobacteriota, and Firmicutes increased the most in the MCA group compared with those in the CG group, with increases of 5.43%, 1.06%, and 3.9%, respectively. This increase suggests that the carbon source released from the MCA biofilm carriers changed the microbial community structure and increased microbial abundance. The relative abundances of Proteobacteria, Bacteroidetes, and Firmicutes accounted for 80.63% of the total, indicating that the biofilm contributed to the high nitrogen removal rate by supplying biodegradable carbon sources, further verifying the effectiveness of MCA in facilitating the formation of films. The top 50 microbial taxa, in terms of relative abundance at the genus level, were clustered on average, as shown in Figure 5b. At the genus level, although the five samples could eventually be clustered into one group and were similar, we noted differences between the MC, MCA, and MCB samples, and the CG and DS samples. These similarities and differences were mainly manifested in the composition of the dominant bacterial genera in each sample. The dominant genera in the MC, MCA, and MCB samples were Thauera (17.42%, 13.15%, 11.25%), unclassified_f__Comamonadaceae (7.15%, 4.38%, 5.76%), Tessaracoccus (4.22%, 4.84%, 2.56%), Flavobacterium (3.68%, 2.92%, 2.92%), Azospira (2.67%, 1.25%, 1.93%), Dechloromonas (2.29%, 7.77%, 7.13%), Paludibacter (1.38%, 6.60%, 3.83%), and norank_f__Acidaminococcaceae (0.60%, 3.05%, 1.86%). Thauera was the dominant genus with the highest relative abundance in each sample, at 17.42%, 13.15%, 11.25%, 26.74%, and 37.93% in the five sets of samples, respectively. This genus is mostly rod-shaped, has a denitrification capacity, and is widely present in various wastewater treatment groups [57,58]. This genus was the most abundant in the group with biofilm carriers, demonstrating that corncobs released a sufficient carbon source for growth.
Substantial differences were observed in the bacterial species among the different biofilm carriers. The relative abundances of bacterial genera with denitrification capabilities in the MC, MCA, MCB, CG, and DS samples were 42.04%, 38.72%, 38.58%, 46.17%, and 48.43%, respectively. Bacterial genera with denitrification capabilities, such as Thermomonas, Dechloromonas, Arenimonas, Acidovorax, and norank_f__Acidaminococcaceae [2], were found to increase, the reason for which might be the competition among species for electron donors and acceptors (NO 3 − , NO 2 − , and NH 4 + ). The Thauera abundance was found to decrease, which could be attributed to the presence of NO 3 − and NO 2 − . Zeng et al. [59] also discovered a negative correlation between the abundance of Thauera spp. and the amounts of NO 3 − and NO 2 − . The abundance of Hydrogenophagafen and Tessaracoccus in the MCA group was 1% and 4.8%, respectively. Hydrogenophagafen is a hydrogen-using autotrophic denitrifying microorganism [60] that is strongly enriched in systems using in situ-produced H 2 as the electron donor for denitrification. Anaerobic iron dissolution produces H 2 via the reduction of protons to convert NO 3 − -N to N 2 [61]. The genus Simplicispira (1.57%), which can denitrify under both aerobic and anaerobic conditions, was abundant in the MCA [62].
Compared with the CG and DS samples, the relative abundances of unclassified_f__ Comamonadaceae and Dechloromonas in the biofilm carrier samples also substantially increased, both exceeding 4.0%, with MCA showing the highest abundance. Unclassi-fied_f__Comamonadaceae and Dechloromonas are solid-phase denitrifying bacteria that utilize lignocellulosic carriers as carbon sources [63]. They can convert macromolecular organic compounds, such as corncob cellulose, into soluble small-molecule organic substances [64]. In addition, Fe 2+ and unclassified_f__comamonadaceae cooperated to reduce NO 3 − -N under anoxic conditions [65]. This indicates that the solid denitrifying and fermenting bacteria were markedly enriched in the MCA group and that these bacteria considerably improved the denitrification effect, which is consistent with the previously reported denitrification effect. In addition, sponge iron produces large amounts of Fe 0 and Fe 2+ to provide electrons for denitrification, which is beneficial for increasing EPSs and promoting the formation and stability of biofilms, as confirmed by the XRD results. The abundances of Zoogloea, Azospira, and Sphaerotilus in the MCA group increased substantially. Zoogloea is a common EPSproducing strain that stimulates biofilms to produce EPSs in large quantities, contributing to biofilm formation and stabilization. The increase in Azospira abundance indicated that the presence of Fe 2+ generated from the sponge iron corrosion affected this bacterial genus. Li et al. [66] investigated the microbial community composition and diversity during the anaerobic nitrate reduction and Fe 2+ oxidation processes in paddy soils under neutral pH conditions, and the results were consistent. This confirms that Sphaerotilus is an aerobic genus with the capacity for Fe 2+ -autotrophic denitrification and Fe compound transfer [67].
The bacterial genera capable of degrading cellulose and polysaccharides showed increased relative abundances in the MC, MCA, and MCB samples, accounting for 5.28%, 11.17%, and 7.63%, respectively, including Trichococcus, Sphingobium, Nannocystis, unclassi-fied_o__Micrococcales, Sphaerotilus, and Paludibacter. Trichococcus is related to organic matter degradation and contributes to the use of carbon sources by heterotrophic denitrifying microorganisms [68]. Paludibacter [69] exhibited the highest increase in abundance in the MCA samples. It uses various sugars to produce acetic and propionic acids as the main fermentation end products. This suggested that the MCA biofilm carrier was more easily used by Paludibacter and subsequently decomposed by Thauera. The results of our analysis indicate that the biofilm carriers synthesized from corncobs and added to the MC, MCA, and MCB groups provided a growth substrate for these bacterial genera, promoting their growth and increasing their abundance. Moreover, corncobs mechanically ground and treated with 1% NaOH were more easily used by the bacterial genera capable of degrading cellulose and polysaccharides. When acid-alkali-pretreated corncobs were used to determine their effects on denitrification in a sequencing batch biofilm group, Liu et al. [15] observed a significant increase in cellulose-degrading bacterial genera.
Among the dominant bacterial genera in the MC, MCA, MCB, and CG groups, the MCA group showed the smallest differences in the proportion of bacterial genera with denitrification capabilities and those with the capability to degrade cellulose and polysaccharides. The combined proportion of these two functional bacterial genera was highest (49.89%) in this group. Considering the NO 3 − -N removal rate and effluent COD concentration, we speculated that a symbiotic relationship may exist between bacterial genera with denitrification capabilities and those with the capability to degrade cellulose and polysaccharides in the MCA biofilm carrier group.
A schematic of the postulated denitrification pathway is shown in Figure 6. Alkali treatment and microbial action decompose part of the cellulose and form a porous and rough corncob structure, providing a suitable place for organisms to attach and grow. The microbial degradation of cellulose produces a biodegradable carbon source that provides an electron donor for heterotrophic denitrifying bacteria, whereas sponge iron corrosion produces H 2 to provide an electron donor for autotrophic denitrifying bacteria and enriches the structure of the microbial community attached to the surface of the biofilm carrier. With the enrichment of the attached microbial community, bacteria with different functions work together to degrade the remaining cellulose into an easily used carbon source, part of which provides an electron donor for heterotrophic denitrifying bacteria, part of which is used by Fe 2+ and Fe 3+ . One part of the Fe 2+ and Fe 3+ was reduced to Fe 0 by Fe-reducing bacteria, which provided an electron donor for the autotrophic denitrifying bacteria, while the other part provided a bridge for carbon source use as an electron shuttle. Under these synergistic effects, nitrate is reduced to N 2 and released into the environment via a series of reduction reactions. The dynamic community relationships among the various functional microorganisms in the MCA biofilm carrier reflect their tendency for teamwork, thereby increasing the denitrification efficiency and biofilm performance. This is crucial for enhancing the denitrification efficiency under low C/N conditions. Fe-reducing bacteria, which provided an electron donor for the autotrophic denitrifying bacteria, while the other part provided a bridge for carbon source use as an electron shuttle. Under these synergistic effects, nitrate is reduced to N2 and released into the environment via a series of reduction reactions. The dynamic community relationships among the various functional microorganisms in the MCA biofilm carrier reflect their tendency for teamwork, thereby increasing the denitrification efficiency and biofilm performance. This is crucial for enhancing the denitrification efficiency under low C/N conditions.

Conclusions
An effective (i.e., in terms of stable carbon release performance, efficient denitrification, and strong microorganism attachment) and novel self-releasing carbon biofilm carrier (MCA) was prepared using corncobs and sponge iron with PVA as scaffolds for low C/N wastewater treatment. We found that the accumulated carbon release of the biofilm

Conclusions
An effective (i.e., in terms of stable carbon release performance, efficient denitrification, and strong microorganism attachment) and novel self-releasing carbon biofilm carrier (MCA) was prepared using corncobs and sponge iron with PVA as scaffolds for low C/N wastewater treatment. We found that the accumulated carbon release of the biofilm carriers was 116.139 mg COD·g −1 , whereas the NO 3 − -N removal rate of the MCA was 98.1%. The release of the carbon sources and nitrate removal were simultaneously achieved by the coexistence of the dominant denitrifying and cellulose-degrading bacterial genera in the MCA. Cellulose degradation and iron cycling by different bacteria contribute to synergistic denitrification. Compared to other experimental biofilm carriers, the biofilm carrier proposed in this study exhibits not only superior performance in the context of wastewater treatment but also a greater reduction in operational costs.  Table S1. Fluorescence region integral standard volume of soaking solutions from 13 to 15 days and 28 to 30 days for three biofilm carriers. Table S2. Alpha diversity parameter result statistics.