Introduction

Variety of pesticides and herbicides are generally applied to agricultural production, but the extensive applications result in serious contaminations of soil and water (Li et al. 2005). Glyphosate is one of the most widely used organophosphorus pesticides, which is a broad-spectrum, non-selective herbicide, and it is difficult to degrade naturally in the environment due to the high stability and durability (Khoury et al. 2010). Owing to the fact that the ingestion of highly toxic glyphosate contaminated drinking water could cause carcinogenic, teratogenic and genotoxic effects on humans, glyphosate has been classified as a probable carcinogenic by World Health Organization (Lin et al. 2000; Manas et al. 2009; Myers et al. 2016). Therefore, the removal of glyphosate from water resources should be taken seriously. So far, a number of methods are applied to remove glyphosate from water including advanced oxidation (Li et al. 2022; Skanes et al. 2021; Xue et al. 2021), photo-assisted degradation (Chen et al. 2022b; Tang et al. 2020; Tang et al. 2021), biodegradation (Chen et al. 2022a; Zhang et al. 2022), membrane filtration (Carneiro et al. 2015; Shen et al. 2014; Song et al. 2013), and adsorption (Diel et al. 2021; Liu et al. 2022; Ueda Yamaguchi et al. 2016). As a simple, highly efficient, and economic process, the adsorption method is more widely used than others (Dissanayake Herath et al. 2019; Rallet et al. 2022; Sen et al. 2021; Sittiwong et al. 2022).

Biochar is a carbon material prepared by pyrolysis of waste biomass at low temperature under anoxic or anaerobic conditions (Pan et al. 2021). Due to the high specific surface area, high mechanical strength and abundant surface functional groups, biochar has been widely used as adsorbents or the carrier (Fu et al. 2019; Shi et al. 2022). Meanwhile, using low cost, naturally biodegraded waste biomass as raw materials could cause less secondary pollution (Ma et al. 2021; Zhang et al. 2015). Water hyacinth is one of tropical plants which distributes widely in the southern watershed of our country. Due to the rapid growth, reproduction and strong adaptability to the environment, water hyacinth could block sunlight penetration, reduce the dissolved oxygen in water, and exclude other aquatic plants competitively and reduce the biodiversity (Huang et al. 2014; Yi et al. 2021). Composed of three carbon-based polymers—cellulose, hemicellulose, and lignin—water hyacinth is considered of a good source of biochar (González-García et al. 2013; Morales S et al. 2021). Compared with other biomass, water hyacinth possesses a higher pore size and specific surface area, and using water hyacinth as a precursor to prepare biochar can not only be used for wastewater treatment but also reduce ecological and environmental problems it caused (Chen et al. 2019; Xu et al. 2020). However, the non-ideal adsorption capacity, poor selectivity of bare biochar has limited its application (Jiang et al. 2018; Qu et al. 2021). Therefore, appropriately modification or load of biochar are necessary to improve the adsorption performance, especially the adsorption capacity and selectivity for glyphosate.

Many sorbents prepared based on metal hydroxide had been investigated in order to improve the better adsorption performance. Lu et al. reported the adsorption performance of ferrihydrite for Pb (II) and Cr (III) ions in the system contained microplastics (Lu et al. 2022). Jiang et al. investigated the adsorption properties of ferrihydrite on arsenate and phosphoric acid. The results showed that phosphoric acid adsorption on the surface of iron (hydrogen) oxide is not only a simple surface adsorption, but also includes surface complexation (Jiang et al. 2015). Antelo et al. prepared ferrihydrite nanoparticles and investigated the adsorption properties on arsenate and phosphate (Antelo et al. 2015). The adsorption performance of phosphate on ferrihydrite, goethite, and hematite were studied and analyzed by Liu et al. (2021). Zhou et al. prepared ferrihydrite-loaded bagasse and applied it to the removal of phosphate from aqueous solution (Zhou et al. 2020). Generally, there are numbers of work on the application of ferrihydrite in the removal of inorganic negative anions like phosphate. Therefore, it is hoped that ferrihydrite based adsorbent can be used in the removal of organophosphate compounds to expand the application in the treatment of industrial wastewater.

In this study, ferrihydrite-loaded water hyacinth-derived biochar (FH/WHBC) was synthesized via in-situ precipitation method. The optimum preparation condition was explored, and the obtained adsorbent was characterized by SEM, XRD, FTIR, XPS, and zeta potential. The adsorption kinetics and isotherm of glyphosate on FH/WHBC were investigated. The effects of solution pH, co-existing ions, and regeneration on the adsorption of glyphosate by FH/WHBC were also studied. Moreover, to determine the application potential, the dynamic adsorption experiment and the waste water treatment were carried out. Additionally, the adsorption mechanism of glyphosate on FH/WHBC was analyzed.

Materials and methods

Materials

Water hyacinth was washed, dried, grinded, and sieved by 100–200 mesh before use. Chemical reagents including concentrated hydrochloric acid, sodium hydroxide, potassium hydroxide, iron nitrate nonahydrate, sodium chloride, sodium sulfate, sodium nitrate, potassium chloride, magnesium chloride, calcium chloride, glyphosate, and absolute ethyl alcohol with analytical grade were purchased from Sinopharm Chemical Reagent (Shanghai, China) and used without further purification.

Preparation and characterization of FH/WHBC-n

Fourteen grams of water hyacinth was treated with 500 mL of 1 mol·L-1 KOH to remove the soluble sugars and other impurities. The residues after centrifugation were washed with 500 mL of deionized water to remove the residual KOH, and 200 mL of absolute ethanol to remove unnecessary pigments successively. Finally, the residues were dried for later use. Water hyacinth biochar (WHBC) was obtained by heating the residues from 30 to 500°C at a rate of 5°C/min, and then calcinating for 1h in a tube furnace under nitrogen atmosphere. The obtained solid was washed with 1 mol·L-1 HCl solution and 90°C deionized water for 1h, respectively. The obtained WHBC were dried at 70 °C overnight before use.

The typical loading process of ferrihydrite (FH) onto WHBC was same as reported by previous work (Zhou et al. 2020), 0.10 g of WHBC was dispersed in 50 mL ferric nitrate solution with a concentration of 50.0 mmol·L-1 and stirred for 30 min at room temperature, and then, NaOH solution was added dropwise to adjust pH within 3.0–10.0. The mixtures were stirred for another 60 min and the prepared brown composite was washed with ethyl alcohol for three times to remove the residual ions. The FH/WHBC-n (where n represents the pH value of preparation solution) were dried at 70°C overnight before use.

The prepared sorbents were characterized by scanning electron microscope (SEM, Hitachi S4800, Japan). The crystal structure of the sorbents prepared at different pH were analyzed using an X-ray diffraction (XRD, Rigaku D/max2500, Japan) in 2θ 5°–80° with a scanning rate of 5°/min. The sorbents before and after glyphosate adsorption were characterized using Fourier transform infrared spectroscopy instrument (FTIR, Thermo Nicolet Corporation, America), the sorbents and KBr were mixed at mass ratio of about 1:100 and grinded evenly, and the grinding solid was pressed into disks for further determination at wavenumber range of 650–4000 cm−1. The sorbents before and after adsorption were also characterized by X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific Esca Lab 250Xi, America), and the binding energy (B.E.) values were corrected to the carbon peak C 1s at 284.8 eV before analyzing. The surface charge of prepared sorbent before and after glyphosate adsorption was measured by zeta potential analyzer (Zetasizer Nano Zs90, Malvern Instruments Ltd., UK) using 0.1 mol·L−1 NaCl as the supporting electrolyte. The concentration of glyphosate was determined by UV spectrophotometer (UV-6100 Double Beam Spectrophotometer, Mapada Instruments).

Batch adsorption experiments

Adsorption experiments were carried out on a shaker operating at 250 rpm, 35°C. For the adsorption isotherm measurements, 0.010 g of the adsorbent was added into 40.00 mL of glyphosate solution (10–100 mg·L-1, pH = 3.4). The adsorption kinetics experiments were performed by adding 0.025 g of the adsorbent into 100.00 mL of glyphosate solution with concentration of 40 mg·L-1. In co-exist ions experiments, 0.010 g of the FH/WHBC-5.0 was added into 40.00 mL of glyphosate solution with concentration of 40 mg·L-1 containing coexisting cations (K+, Na+, Ca2+, Mg2+) and anions (Cl-, NO3-, SO42-) with initial concentration of 0, 50, 100, 150, 200 mg·L-1. In the wastewater treatment experiment, the simulated waste water was prepared with glyphosate and running water, and different amounts of FH/WHBC-5.0 were added to 40 mL glyphosate with initial concentration of 20 mg·L-1. Three adsorption-desorption cycles were carried out to evaluate the reusability of the prepared sorbents. For this work, 0.01 mol·L-1 NaOH was used for regeneration of sorbents, and then, the sorbents were washed with DI water until neutral.

The effects of pH on adsorption performance were investigated in the range of 1.0–12.0. To determine the glyphosate concentration before and after adsorption, glyphosate was oxidized to phosphate using potassium persulfate as oxidant, and the phosphate concentration were analyzed by ammonium molybdate spectrophotometry (Zhou et al. 2020). The specific determination principle is shown as follows: the yellow phosphorus-molybdenum-antimony complexes are generated through the reaction of phosphate and ammonium molybdate under acidic conditions, and then the complexes are reduced into blue compounds by ascorbic acid. The blue shade is proportional to the concentration of phosphate, and the concentration of glyphosate can be calculated indirectly. The amounts of glyphosate adsorbed were calculated by Eq. (1), where qe (mg·g-1) is the amount of glyphosate adsorbed per unit mass of the sorbent, V (mL) is the volume of glyphosate solution, C0 (mg·L-1) and Ce (mg·L-1) are the concentration of the glyphosate before and after adsorption, and m (g) is the weight of the adsorbent.

$${q}_e=\frac{V\left({C}_0-{C}_e\right)}{m}$$
(1)

Dynamic adsorption experiments of glyphosate adsorption on FH/WHBC-5.0

The dynamic adsorption experiments were conducted in a laboratory-scale fixed-bed column of inner diameter 1 cm and height 20 cm. Typically, 1.0 g of the FH/WHBC-5.0 was loaded into the column, and then the glyphosate solution (0.50, 0.75, 1.00 mmol·L-1) were pumped upward through the column at different flow rate (2.50, 3.75, 5.00 mL·min-1). In the waste water treatment experiment, the simulated waste water was prepared with glyphosate and running water, and the concentration and flow rate were set at 0.5 mmol·L-1 and 5 mL·min-1 respectively. Samples at different effluent times were collected and the concentrations of glyphosate were determined by the same method as in batch adsorption experiments.

Results and discussion

Preparation and characterization of FH/WHBC

Figure 1 shows the synthesis procedure of the FH/WHBC. Water hyacinth was firstly carbonized at 500°C, and then the obtained active carbon was dispersed into the Fe(NO3)3 solution. The iron-containing particles are loaded on the WHBC by adjusting the solution pH with NaOH. SEM images in Fig. 2a depict that WHBC had an obviously porous structure, however, that changed significantly after iron-containing particles were loaded. At pH 5.0, the precipitated nanoparticles were relatively uniformly distributed on the surface and in the pore channels of WHBC, while there was an obvious accumulation of nanoparticles in the pore structure when pH increased to higher than 7.0. At pH 10.0, iron-containing particles almost filled the pores. To determine the effect of the loaded particles on the adsorption capacity of the active carbon for glyphosate, adsorption experiments were carried out and results were shown in Fig. 2e. It was observed that the adsorption amount of glyphosate on the FH/WHBC prepared at pH 3.0, 4.5, 5.0, 7.0, and 10.0 was 42.7, 111.4, 116.8, 110.9, and 104.5 mg·g-1, respectively, indicating that the optimal pH range for preparing the adsorbent was from 4.5 to 10.0. Figure 2f shows the XRD pattern of the WHBC, FH/WHBC-5.0, FH/WHBC-7.0, and FH/WHBC-10.0. In the spectrum of the WHBC, a diffraction band ranged from 20 to 30° were present, demonstrating that WHBC is a typical amorphous carbon material. After loaded with ferric hydroxide, two new peaks at around 35° and 62° were present on all of FH/WHBC-5.0, FH/WHBC-7.0, and FH/WHBC-10.0 spectra. These two weak and broad peaks corresponding to the typical planes of (110) and (300), respectively (Liang et al. 2021), were the characteristic peaks of amorphous 2-line ferrihydrite, indicating that ferric hydroxide is deposited on the surface of WHBC in the form of ferrihydrite (Jiang et al. 2015; Zhou et al. 2020). Since FH/WHBC-5.0 had the highest adsorption capacity toward glyphosate, it was chosen as the sorbent in the following experiments. FTIR and zeta potential of FH/WHBC-5.0 before and after glyphosate adsorption was shown in Fig. 2g and Fig. 3a, respectively. The band at about 3383 cm-1 in Fig. 2g was corresponded to the stretching vibration of hydroxyl group from the ferrihydrite, and the peak at 1323 cm-1 was assigned to the characteristic peaks of Fe-O-H. Similar findings were reported by Liu et al. (2021) and Xu et al. (2016). Figure 3a shows that there was an obvious change for the zeta potential of the active carbon after ferrihydrite loading, and the isoelectric points increased from 2.11 to 11.56. The more positively charged surface was due to the introduction of ferrihydrite, which would provide more active adsorption sites for negatively charged molecule through electrostatic attraction.

Fig. 1
figure 1

Synthetic process of FH/WHBC

Full size image
Fig. 2
figure 2

SEM of a WHBC and FH/WHBC prepared at b pH 5.0, c pH 7.0, d pH 10.0, and e the effects of precipitation pH on the adsorption capacity of FH/WHBC, f XRD of WHBC, FH/WHBC-5.0, FH/WHBC-7.0, and FH/WHBC-10.0, g FTIR of WHBC, FH/WHBC-5.0 before and after glyphosate adsorption.

Full size image
Fig. 3
figure 3

Zeta potential of WHBC, FH/WHBC-5.0 before and after glyphosate adsorption (a), distribution coefficient diagram of glyphosate (b), adsorption isotherms (c), and kinetics of glyphosate on FH/WHBC-5.0 (d). Reaction condition: isothermal adsorption, 35°C, 250 rpm, 40 mL solution (10-100 mg·L-1), 0.01 g sorbent; kinetic adsorption, 100 mL solution (40 mg·L-1), 0.025 g sorbent.

Full size image

Adsorption isotherms and kinetics of glyphosate on FH/WHBC-5.0

Since glyphosate was a polybasic acid, it would exist in different form at different solution pH. Figure 3b shows the distribution coefficient diagram of glyphosate, and it mainly existed as HOOC-CH2-NH2+-CH2-PO3H2 (A) at pH 1.0, as HOOC-CH2-NH2+-CH2-PO3H- (B) at pH 2.0, as -OOC-CH2-NH2+-CH2-PO3H- (C) at pH 4.0, as -OOC-CH2-NH2+-CH2-PO3- (D) at pH 8.0, and as -OOC-CH2-NH-CH2-PO3- (E) at pH 12.0, respectively. To determine the adsorption performances of the FH/WHBC-5.0 for different species of glyphosate, isotherms and kinetics experiments were carried out at pH 1.0, 2.0, 4.0, 8.0, and 12.0, respectively.

Adsorption isotherms of glyphosate in Fig. 3c demonstrates that the amounts of glyphosate adsorbed on FH/WHBC-5.0 increased with the rise of the initial glyphosate concentration under the detected pH range. The greatest adsorption capacity for glyphosate on FH/WHBC-5.0 presents at pH 4.0. The data of adsorption isotherms were analyzed using Langmuir, Freundlich, and Temkin model (Eqs. (2)–(4)) (Guo et al. 2022; Rallet et al. 2022), and the results of mathematic fitting are shown in Table 1. The qe represents the equilibrium adsorption capacity of glyphosate (mg·g-1); qm is the maximum adsorption capacity (mg·g-1); Ce is the equilibrium concentration of glyphosate (mg·L-1); KL is the Langmuir adsorption constant (L·mg-1); KF and 1/n are Freundlich adsorption constants (mg·mg-1/n·L1/n·g-1), which are related to the adsorption intensity between adsorbent and adsorbate; AT (L·mg-1) is the maximum binding energy corresponding to the equilibrium binding constant; bT is the Temkin isothermal constant (mg-1·g-1); R and T represent the ideal gas constant (8.3145 J·mol-1·K-1) and temperature (K), respectively. According to the fitting parameters in Table 1, the Langmuir model shows a better fit to the data than Freundlich and Temkin model, indicating that the glyphosate adsorption on FH/WHBC-5.0 is a mono-layer adsorption. The predicted maximum adsorption capacities of FH/WHBC-5.0 for glyphosate at pH 1.0, 2.0, 4.0, 8.0, and 12.0 are 94.6, 101.2, 116.8, 103.3, and 9.0 mg·g-1, which were consistent with the experimental values. Under acidic conditions, the surface of FH/WHBC carried more positive charges, which was beneficial for the glyphosate adsorption via electrostatic interaction. However, the dissolution of ferrihydrite occurred at pH 1.0 and 2.0, and it was much more obvious at pH 1.0 according to the results in Fig. S1, which led to a lower removal efficiency. So, the adsorption amount of glyphosate onto FH/WHBC presented an upward trend with the increase of pH. But then, with the further increase of pH to an alkaline condition, hydroxyl would compete with glyphosate for the adsorption sites. Besides, the electrostatic repulsion between negatively charged surface of FH/WHBC and glyphosate was also not conductive to the adsorption. Therefore, the adsorption amount reduced sharply at pH 12.0. The fitted parameters also illustrated the great adsorption performance of FH/WHBC-5.0 for various forms of glyphosate at pH 1.0–8.0, demonstrating its application potential in treatment of acidic wastewater containing glyphosate. The comparation between several existing adsorbents in literature and FH/WHBC in this work was also summarized in Table S1. As can be seen that the developed FH/WHBC possessed a higher adsorption capacity than most of the reported materials, suggesting that it might be a potential candidate for the treatment of glyphosate containing wastewater.

$${q}_e=\frac{q_m{K}_L{C}_e}{1+{K}_L{C}_e}$$
(2)
$${q}_e={K}_F{C}_e^{1/n}$$
(3)
$${q}_e=\frac{RT}{b_T}\ln \left({A}_T{C}_e\right)$$
(4)
Table 1 Adsorption isotherm parameters fitted by Langmuir, Freundlich, and Temkin models
Full size table

Adsorption kinetics was also used to evaluate the adsorption performance of FH/WHBC-5.0, and the data are shown in Fig. 3d. It was observed that the adsorption amount of glyphosate increased rapidly with the increase of time at the initial stage, and higher than 90% of the glyphosate was adsorbed within the first 30 min. Then the adsorption rate slowed down, and the equilibrium was obtained at 120 min for all the adsorption curves, and the equilibrium amounts of glyphosate absorbed at different solution pH were in agreement with that obtained in the isotherm experiments. According to Fig. S2, the adsorption capacities of original biochar prepared at carbonization temperatures 400-700oC were all lower than 6 mg/g, demonstrating that the adsorption performance of FH/WHBC for glyphosate could be ascribed to the introduction of ferrihydrite. To further analyze the adsorption mechanism, the adsorption kinetic curves were fitted by pseudo-first-order and pseudo-second-order models (Eqs. 5–6) (Ueda Yamaguchi et al. 2016; Zhou et al. 2020), where qe (mg·g-1) and qt (mg·g-1) is the amount of glyphosate absorbed at equilibrium and time t, k1 (min-1) and k2 (g·mg -1·min-1) is the rate constant of pseudo-first-order and pseudo-second-order model, respectively.

$${q}_t={q}_e\left(1-{e}^{-{k}_1t}\right)$$
(5)
$${q}_t=\frac{k_2{q_e}^2t}{1+{k}_2{q}_et}$$
(6)

The fitting parameters in Table 2 shows that the adsorption kinetic process is well interpreted by the pseudo-second-order kinetic model due to its higher correlation coefficients. The predicted equilibrium adsorption capacities at different pH by pseudo-second-order model were consistent with the experimental values and the results estimated by Langmuir model.

Table 2 Adsorption kinetic parameters fitted by two kinetic models
Full size table

Effects of co-ions on glyphosate adsorption on FH/WHBC-5.0

In practical wastewater, there were many co-existed anions and cations including Na+, K+, Ca2+, Mg2+, NO3-, Cl-, and SO42-, which might affect the glyphosate adsorption through competing for the active sites on the sorbent surface. Therefore, effects of different co-ions and their initial concentrations on glyphosate adsorption on FH/WHBC-5.0 were studied and results were shown in Fig. 4. For the co-existing cations, at the initial concentrations of 50, 100, 150, and 200 mg·L-1, amounts of glyphosate absorbed varied from 103.2 mg·g-1 to 110.1, 110.4, 113.1, and 105.7 mg·g-1 in the presence of Na+, to 114.1, 112.6, 108.8, and 118.1 mg·g-1 in the presence of K+, to 107.3, 117.0, 112.8, and 114.2 mg·g-1 in the presence of Ca2+, and to 114.1, 122.1, 112.0, and 118.6 mg·g-1 in the presence of Mg2+, respectively. These results illustrated that the co-existed cationic ions not only did not inhibit the glyphosate adsorption, but also promoted its adsorption to some extent. Similar results were obtained for the co-existing anions except SO42-. When the initial concentration of SO42- increased to 50, 100, 150, and 200 mg·L-1, amounts of glyphosate absorbed decreased from 103.2 mg·g-1 to 95.0, 95.4, 76.7, and 79.0 mg·g-1, respectively. Above results depicted that the common co-ions of Na+, K+, Ca2+, Mg2+, NO3-, and Cl- all promoted the adsorption of glyphosate on FH/WHBC-5.0. Similar results were reported by phosphate adsorption on nanoparticles and bentonite, and they ascribed it to the formation of inner-sphere complexes between the substrate and the absorbents (Ju et al. 2016; Lin et al. 2018).

Fig. 4
figure 4

Effects cations (a) and anions (b) on glyphosate adsorption on FH/WHBC-5.0. Reaction condition: 35°C, 250 rpm, 40 mL solution (40 mg·L-1), 0.01 g sorbents.

Full size image

Effects of dosage on glyphosate removal and the reusability of FH/WHBC-5.0

The effects of sorbent dosage on glyphosate removal were shown in Fig. 5a. At the dosage of 5, 10, 15, 20 mg, the removal efficiency of glyphosate increased from 33.7 % to 79.4, 90.6, and 97.1%. The concentration of glyphosate deceased from 20.7 to 0.6 mg·L-1, which was lower than the groundwater quality of class III mandated in China, and suitable for centralized drinking water source and industrial and agricultural water (Beijing 2018). The results of reusability of the sorbent in Fig. 5b showed that the adsorption capacity of sorbents decreased from 82.5 mg·g-1 to 75.4 and 68.3 mg·g-1. In this study, the FH/WHBC after glyphosate adsorption was regenerated using 0.01 mol·L-1 NaOH as desorption agent, during which some active adsorption sites might have been occupied by hydroxy groups, leading to a reduction in adsorption capacity of FH/WHBC in next cycle. Besides, the adsorbed glyphosate could not be completely desorbed in the regeneration process (Fig. S3), and the occupied active sites might also cause the decline in adsorption amount. Still, the effective regeneration methods need to be further studied in next research.

Fig. 5
figure 5

Effects of sorbent dosage on glyphosate removal (a) and the reusability of FH/WHBC-5.0 (b). Reaction condition: 35°C, 250 rpm; 40 mL solution (20 mg·L-1), dosage 5–20 mg; cycling experiments, 200 mL solution (20 mg·L-1), 0.05 g sorbent

Full size image

Dynamic adsorption of glyphosate on FH/WHBC-5.0

Dynamic adsorption is the most commonly method used to treat the practical water treatment, and flow rate and initial concentration were the key factors that affect the glyphosate adsorption behavior. Breakthrough curves of glyphosate on FH/WHBC-5.0 at different flow rate and initial concentration were determined and shown in Fig. 6a and c. It was observed that the breakthrough curves all presented a typical “S” shape, and Ct/C0 was close to zero at the initial stage and then increased sharply to 1.0 with the saturation of the column. Breakthrough time (Ct/C0 < 0.05) and saturated time both decreased with the increase of the flow rate and initial concentration. For example, the former decreased from 245.9 min to 121.3 and 96.4 min, and the latter decreased from 412.6 min to 321.5 and 224.9 min when the flow rate increased from 2.5 mL·min-1 to 5.0 mL·min-1. When the initial concentration increased from 0.5 mmol·L-1 to 0.75 and 1.0 mmol·L-1, breakthrough time decreased from 316.5 min to 223.1, and 115.2 min, and saturated time decreased from 422.7 min to 337.5, and 225.9 min, respectively.

Fig. 6
figure 6

Effect of flow rate (a) and initial concentration (c) on the breakthrough curve and dynamic adsorption kinetics of glyphosate (b, d) on the fixed bed column. (Waste water treatment was operated at initial concentration of 0.5 mmol·L-1, flow rate of 5.0 mL·min-1.)

Full size image

The adsorption capacity of the fixed bed at different initial concentration and flow rate were calculated through the following equation (Eq. 7) and results were showed in Fig. 6b and d. The qe represents the adsorption capacity of fixed bed at different effluent time; Ct represents the effluent concentration of glyphosate at different time (mg·L-1); C0 is the initial concentration of glyphosate (mg·L-1); v is the flow rate (mL·min-1); m represents the mass of sorbents loaded into the column (g). It depicted that the amount of glyphosate increased linearly with the increase of time at first time and then it rose slowly to equilibrium values. The saturated adsorption capacity of the fixed bed column was 98.3, 99.8, and 101.5 mg·g-1 at flow rate of 2.5, 3.75, and 5.0 mL·min-1, and it was 108.6, 99.3, and 99.8 mg·g-1 at the initial concentration of 0.5, 0.75, and 1.0 mmol·L-1. Those results illustrated that the saturated adsorption capacity of the fixed bed column almost was not affected by those two factors, and that the prepared fixed bed column could be used to treat wastewater with high initial concentration at high flow rate.

$${q}_t=\frac{v\int_0^t\left({C}_0-{C}_t\right) dt}{m}$$
(7)

To further investigate the fixed bed adsorption process, the breakthrough curves were fitted using Bohart-Adams and Yoon-Nelson models (Eqs. 8–9) (Guo et al. 2022; Wang et al. 2021). The KAB is the rate constant of Bohart-Adams model (mL·mg-1·min-1); N0 is the absorptivity of fixed bed (mg·L-1); Z is the length of mass transfer zone (cm); V is the linear velocity (cm·min-1); KYN is the rate constant of Yoon-Nelson model (min-1); τ is the required time for half of glyphosate being adsorbed (min). The fitted parameters of the breakthrough curves at different initial concentration and flow rate by Bohart-Adams and Yoon-Nelson model are exhibited in Table 3.

$$\frac{C_t}{C_0}=\frac{e^{K_{AB}{C}_0t}}{e^{K_{AB}{N}_0Z/V}-1+{e}^{K_{AB}{C}_0t}}$$
(8)
$$\frac{C_t}{C_0}=\frac{1}{1+{e}^{K_{YN}\left(t-\tau \right)}}$$
(9)
Table 3 Fitted parameters of the breakthrough curves at different flow rates and initial concentrations
Full size table

The Yoon-Nelson fitting parameters in Table 3 showed that the rate constant of glyphosate (KYN) decreased from 0.08 to 0.07 min-1 as the flow rate increased from 2.5 to 5.0 mL·min-1. In addition, the half-penetration times predicted by Yoon-Nelson model at flow rates of 2.5, 3.75, and 5.0 mL·min-1 were 309.6, 210.6, and 139.9 min, respectively, indicating that the adsorption rate of glyphosate was positively correlated with the flow rate in column.

Adsorption mechanisms

Acidity and coexisting ion experiments showed that the adsorption of glyphosate by FH/WHBC-5.0 was mainly through electrostatic attraction and inner-sphere complexation. FH/WHBC-5.0 before and after adsorption were characterized by zeta potential, FTIR and XPS to determine the conclusions. The results of acidity experiments and zeta potential in Fig. 3 showed that the surface of the material was negatively charged at the solution pH of 12, which was higher than the isoelectric point of 11.5. The fact well explained that the equilibrium adsorption capacity of FH/WHBC-5.0 decreased significantly under strong alkaline condition. Meanwhile, isoelectric point of the material decreased after adsorption of glyphosate, indicating that the positive charged active site on the surface of adsorbents participated in the adsorption process through electrostatic interaction. Figure 2g depicts the FTIR spectrum of FH/WHBC-5.0 before and after glyphosate adsorption. The FTIR spectrum showed a wider and stronger absorption peak at 1047 cm-1 after adsorption, which could be attributed to the structure of -H2PO3- in glyphosate (Hao et al. 2019; Liu et al. 2022). The wide peak at 3000-3500 cm-1 and the characteristic peak of Fe-OH at 1323 cm-1 were significantly weakened, demonstrating that Fe-OH on the surface of the adsorbents was involved in the adsorption process (Zhou et al. 2020). In addition, the enhancement of absorption peaks at 2900 and 2970 cm-1 after adsorption might be related to the symmetric and asymmetric stretching vibration of C-H bonds in glyphosate attached to surface (Xu et al. 2016).

The XPS spectra of FH/WHBC-5.0 before and after glyphosate adsorption were shown in Fig. 7. The peaks of Fe 2p at 724.6 and 711.1 eV shifted to peaks with higher binding energy of 725.1 and 711.5eV after glyphosate uptake, indicating the possible electron transfer in the valence band of Fe 2p and the formation of Fe-O-P inner-sphere complexation (Liu et al. 2019; Zhou et al. 2020). The results were consistent with the work reported by Sittiwong et al., which pointed out that the metal ions on the surface of sorbents could interact with the P-O in glyphosate as Lewis acidic site (Sittiwong et al. 2022). The high-resolution XPS spectrum of P 2p in Fig. 6d showed a new peak at 133.5eV after adsorption, which could be attributed to the P-O-Fe bond, indicating the formation of inner-sphere complexation (Hao et al. 2019; Liu et al. 2019). The O 1s XPS spectrum in Fig. 6b showed that two peaks at binding energy of 532.1 and 530.3 eV attributed to C-O and Fe-O-Fe, respectively, became weakened due to the effect of electrostatic interaction and complexes after adsorption (Zhou et al. 2020).

Fig. 7
figure 7

XPS spectra of wide scan (a), O 1s (b), Fe 2p (c), P 2p (d) of FH/WHBC-5.0 before and after glyphosate adsorption and the possible mechanism adsorption of glyphosate by FH/WHBC-5.0 (e).

Full size image

Wastewater treatment

Since the prepared sorbent had good adsorption property, it was used to treat the wastewater containing glyphosate, and the other co-ions of Na+, K+, Ca2+, Cl-, and SO42-. Breakthrough curve and dynamic adsorption curves were shown in Fig. 6. The breakthrough time and saturated adsorption capacity for glyphosate were 274.6 min and 118.3 mg·g-1, respectively. After the treatment, concentration of glyphosate decreased significantly from 68.3 mg·L-1 to 0.06 mg·L-1, which was lower than the groundwater quality of class II mandated in China and suitable for various purposes (Beijing 2018). This experiment demonstrated that the prepared sorbent could be used in treating the practical waste water.

Conclusions

The adsorbent FH/WHBC with high adsorption capacity and selectivity for glyphosate was prepared by an in-situ precipitation method at pH 5.0. The adsorption process of glyphosate on FH/WHBC was consistent with Langmuir model and second-order kinetic model, and the adsorption process can be completed within 30 min, with a maximum adsorption capacity of 116.8 mg/g. The prepared FH/WHBC performed better under high acidity environment, and could maintain the adsorption performance after regeneration for three times. The presence of most co-existing ions could hardly affect the glyphosate adsorption onto FH/WHBC. Further, it still possessed excellent adsorption performance at high flow rate and high glyphosate concentration in the fixed bed column, and could remove glyphosate from wastewater efficiently to a final concentration of 0.06 mg·L-1, which was lower than the groundwater quality of class II mandated in China. The FTIR and XPS results showed that electrostatic attraction and inner-sphere complexation were the main driving forces contributing to the glyphosate adsorption on FH/WHBC. Based on the excellent adsorption performance and low cost, the ferrihydrite-loaded biochar materials have great prospect in the removal of glyphosate from industrial wastewater.