Assistance of aminated straw on enhanced biodegradation of phenol through bacteria and potential prediction of machine learning for bioremediation

The pollution of phenol is urgent to effectively be solved, due to its high biological toxicity and wide pollution area which seriously threatens human health and the diversity of biological species (Barik 2021). The microbial method is considered as a promising treatment because of its economical and thoroughness of degradation features (Iqbal et al. 2018). There are many environmental microorganisms which are capable of degrading for phenol (Shahryari et al. 2018; Panigrahy et al. 2020; Zhao et al. 2021). Generally, those microorganisms face problems such as low tolerance concentration, long degradation period, degradation and survival ability decrease with the increase in external pollutant concentration (Long et al. 2019; Duraisamy et al. 2020; Nouri et al. 2020; Wen et al. 2020; Gong et al. 2021). In addition, the existence of complicated pH, high salt concentration and numerous metal ions in the wastewater inhibit the survival of microorganisms (Mei et al. 2019; Xu et al. 2021). Therefore, it is very necessary to develop strategies that can not only remove high concentrations of phenol but also maintain or improve microbial viability and degradation efficiency in harsh wastewater.

Recently, the immobilization of microorganisms was used to fix strains on the carriers to protect microorganisms from the toxicity of high concentration pollutants, so as to improve their ability to withstand harsh conditions. However, most studies focused on the concentration of phenol up to 1,500 mg/L, high concentrations have been rarely reported (Bera & Mohanty 2020; Dong et al 2020; Zhao et al 2020). Biomass and its derived materials have been applied as carriers of microorganisms, due to their economical features, excellent mechanical strength and structural stability. Cashew apple bagasse was applied to immobilize Candida tropicalis ATCC 750 to biodegrade phenol, showing that the adapted and immobilized strain was more resistant to higher phenol concentration than the free cell (Silva et al. 2019). Areca nut husks and luffa sponge fibers immobilizing bacterial consortium could degrade 1,000 mg/L phenol in 28 and 30 h, respectively (Bera & Mohanty 2020). Currently, most studies focus on the concentration of phenol at 250–1,500 mg/L, while high concentrations have rarely been reported. Due to the complexity of the actual wastewater, associated with various pH, heavy metals and salts, numerous verifications are required during the actual treatment process to determine the optimal conditions, which results in a waste of time and economic cost. Back propagation artificial neural network (BP-ANN), as a kind of machine learning method, could recognize complex non-linear relationships between input and output data sets, which further estimate output values referring to training and learning processes (Bao et al. 2019; Zhang et al. 2024). BP-ANN can not only solve the complicated biodegraded problems but can also be applied to predict the optimization of degraded conditions. BP-ANN has been developed for the prediction of soil heavy metals and persistent organic pollutants. However, few studies on immobilized microorganisms for biodegradation of phenol by the usage of BP-ANN have been reported, which is in great demand for savings of time and cost.

Herein, a kind of biomaterial was prepared by the electrostatic interaction between the negative Pseudomonas aeruginosa PAO1 and positive aminated straw, to apply on phenol degradation (Figure 1). This study aimed to (1) investigate the degraded rate and efficiency at high concentrations of phenol by immobilized PAO1; (2) verify the degraded ability of immobilized PAO1 in harsh environments; (3) determine the predictive ability of the BP-ANN model in predicting phenol biodegradation. This study provides an alternative for biodegradation of phenol wastewater, which offers predictions of phenol remediation in complicated surroundings.

The corn straw was from Jilin Agricultural University. Triethylamine (99%, v/v) and ethylenediamine (99%, v/v) were from Beijing Chemical Works. N, N-dimethylformamide (DMF) (99.5%, v/v), Cr₂(SO₄)₃ and K₂Cr₂O₇, were the products of Sinopharm Chemical Reagent Co. Ltd. Potassium ferricyanide (99.5%, w/w), CoCl₂ and HgCl₂ were the products of Tianjin Kermel Chemical Reagents Development Centre. CdCl₂, NiCl₂, Ag₂SO₄ and ZnCl₂ were from Shanghai Macklin Biochemical Co. Ltd. Phenol (99%, w/v), epichlorohydrin (99%, v/v), 4-aminoantipyrine (98.5%, w/w), MnCl₂, CuCl₂, and Pb(NO₃)₂ were bought from Tianjin Hengxing Chemical Preparation Co. Ltd, Xi'an Regent, Shanghai Yuanye Biotechnology Co. Ltd, Beijing Chemical Works and Xilong Chemical Co. Ltd, respectively. Pseudomonas aeruginosa PAO1 was provided by the Pathogenic Microbiology Team, College of Life Sciences, Northwest University, China.

After drying and crushing, the straw was screened with a particle size of about 75 μm, washed with distilled water until the filtrate was clarified, and dried at 65 °C in an oven for 24 h. Then, the straw was immersed into an excess of NaOH (1 mol/L), stirred for 60 min, followed by filtering and washing until the filtrate was neutral, and dried to constant weight to obtain alkali pretreatment straw. Straw (4 g), epiclorohydrin (20 mL, 99% wt) and DMF (20 mL, 99% wt) were placed together and heated to 85 °C with stirring for 60 min, followed with the addition of triethylamine (20 mL, 99% wt) for another 60 min. After cooling to room temperature, the straw was cleaned with anhydrous ethanol three times, and then washed with distilled water until the filtrate was clarified and neutral, and then dried at 80 °C to constant weight to obtain the amine-modified straw.

Luria-Bertani medium (LB: tryptone 10 g/L, NaCl 10 g/L and yeast extract 5 g/L) and minimal salt medium (MSM: MgSO₄·7H₂O 0.2 g/L, CaCl₂·2H₂O 0.02 g/L, KH₂PO₄ 1.0 g/L, K₂HPO₄ 1 g/L, NH₄NO₃ 1 g/L and FeCl₃ 0.05 g/L) were used for activation and acclimation, respectively. The above medium was adjusted by 6 mol/L NaOH or HCl to pH 7.0–7.2, then sterilized at 115 °C under high temperature and pressure for 20 min, and cooled to room temperature before use. When using solid medium, 1.5–2% agar can be added.

The strain was cultured in MSM with phenol as the sole carbon source, and the phenol concentration was continuously increased each time (250 mg/L) to achieve maximum biodegradation (about 1,500 mg/L). Then, the strains were inoculated (10%, v/v) into MSM (phenol, 1,500 mg/L), the above operations were repeated at 37 °C and 180 rpm to shorten the degradation time, until degradation was stable. The concentration of phenol was detected by the 4-aminoantipyrine method (Baird et al. 2017).

The fermentation broth of PAO1 was centrifuged at 5,000 rpm at 4 °C for 20 min. Then, the supernatant was discarded, washed with sterile water and centrifuged again, and repeated three times. After that, the strains were suspended with sterile water, with an optical density value of 1.0 ± 0.11. The linear relationship between the number of bacteria and the optical density (OD) value can be obtained by the plate dilution coating culture method, as shown in Supplementary material, Figure S1. PAO1 bacterial suspension and aminated straw were immobilized at the ratio of solid to liquid 1:100 at 30 °C and 120 rpm for 6 h and repeated three times. After fixation, 300-mesh nylon cloth was used for filtration, and the OD600 value of filtrate was determined. The fixed number of bacteria could be obtained according to Supplementary material, Figure S1 and the OD600 difference of bacterial suspension before and after fixation. After fixation, nylon cloth with 300-mesh was used for filtration. According to the solid–liquid ratio of 1:100, the immobilized material was added into the MSM containing phenol to perform the phenol degradation experiment at 37 °C and 180 rpm.

The sterilized MSM was adjusted to pH 5, 6, 7, 8, 9, 10 and 11, with 6 mol/L NaOH or HCl, followed by the addition of 1,000 mg/L phenol. The immobilized material was treated with the solid–liquid ratio 1:100, and cultured at 37 °C and 180 rpm.

The MSM with 1,000 mg/L phenol was added to 0, 1, 2, 3, 4, 5, 6 and 7% NaCl, respectively, and autoclaved at 121 °C for 20 min. The pH was adjusted to 7.0–7.2. The cultural conditions were the same as before.

Heavy metal ions Cr³⁺, Cr⁶⁺, Ag⁺, Co²⁺, Mn²⁺, Cu²⁺, Pb²⁺, Cd²⁺, Zn²⁺, Hg²⁺ and Ni²⁺ were added to MSM, respectively, where the concentration of heavy metal salt ions was 5 mmol/L. The cultural conditions were the same as before.

According to the factors affecting the biodegradation of phenol, including phenol concentration (1,000, 1,500, 1,750, 1,800, 1,900 and 2,000 mg/L), pH (5, 6, 7, 8, 9, 10 and 11), NaCl concentration (0, 1, 2, 3, 4, 5, 6 and 7%) and time, with three repetitions per group (total of 295 groups), a BP-ANN model was established to test and predict whether the expected value and empirical value was consistent. Every set of data contained the effective values of all variables in each ML model data set, which were excluded in the case of missing data. Here, the BP-ANN model used 286 sets of data and was randomly divided into a training group and a test group, according to a ratio of 9:1. In order to ensure the accuracy of the predicted data, the training group had 256 sets of data and the test group had 30 sets of data.

Fourier transform infrared (FTIR) data were performed on a Bruker IFS66 V FTIR spectrometer (32 scans), using KBr pellets, and the spectra were recorded with a resolution of 4 cm⁻¹. Zeta potential data were recorded on a Malvern Instruments Zetasizer Nano ZS. Element analysis (EA) was carried out on a Flash EA1112 analyzer from ThermoQuest Italia S.P.A

In conclusion, a kind of biomaterial was prepared through strain PAO1 immobilized on aminated straw to enhance the phenol removal rate and efficiency. The aminated straw assisted the strain in increasing the phenol degraded concentration and shortening the degraded time. Compared with free bacteria, aminated straw helped the strain to degrade phenol in the external harmful environment. Furthermore, the degradation efficiency of phenol by the immobilized strains under different conditions was also predicted by BP-ANN, which could be applied to complex wastewater in practical applications, reducing time and economic costs. The present study envisioned that this biomaterial has great potential for the bioremediation of organic pollution.

This research was supported by the Education Department of Jilin Province of China (JJKH20230375KJ).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.