Introduction

The cyanobacteria bloom has frequently occurred in various types of surface water and caused serious problems for drinking water supply (Henderson et al. 2008). The excessive growth of toxic cyanobacteria has become a notorious environmental phenomenon and posed a threat to aquatic ecosystems, due to their blooming deteriorated water quality subject to prevalent eutrophication (Min et al. 2008; Paerl and Paul 2012; Sun et al. 2012). As one of the typical toxin-producing and dominant cyanobacteria, Microcystis aeruginosa can excrete algal organic matters (AOM) during the exponential growth phase, such as intracellular organic matters (IOMs), extracellular organic matters (EOMs), precursors of disinfection by-products (DBPs), taste and odor substances, as well as microcystins (MCs). The AOM tended to release from the cells of M. aeruginosa under environmental stress responsible for cell lysis (Lei et al. 2012; Svrcek and Smith 2004). The accelerated mass release of AOM enhanced the stability of M. aeruginosa cell, making M. aeruginosa removal more difficult than an intact whole cell. Therefore, effective strategies are imperatively needed for enhancing the control of M. aeruginosa by decomposing the AOM without damaging cell integrity and causing metabolites releasing.

Oxidation and coagulation are common methods for removing cyanobacteria cells due to their high efficiency and excellent performance. The conventional coagulation has difficulty in complete removal of M. aeruginosa cells, because M. aeruginosa cells possess negatively charged structure nature with high mobility and diverse morphology (Ma et al. 2012a, b; Pieterse and Cloot 1997; Teixeira and Rosa 2006). The excretion of dissolved AOM from cell metabolism can form a protective shield to M. aeruginosa cells and make the removal of M. aeruginosa even more difficult (Ma et al. 2016). Furthermore, the addition of various oxidants prior to coagulation, such as chlorine, chlorine dioxide, ozone, potassium permanganate and potassium ferrate, has been widely applied for the removal of cyanobacterial cells (He and Wert 2016; Ma et al. 2012a, b; Zhao and Zhang 2011; Zhou et al. 2014). However, some oxidants (e.g., ozone, chlorine dioxide) may have negative effects on coagulation, due to the excessive oxidation, resulting in cyanobacterial cell lysis and IOM release. With the inevitable release of undesirable dissolved toxins, AOM degradation associated with forming low molecular weight (MW) compounds may make the coagulation process more challenging (Lei et al. 2012).

Recently, advanced oxidation processes (AOPs), such as hydroxyl radical (OH·) and sulfate radicals (\({\text{SO}}_{4}^{ \cdot - }\)), have been applied widely as promising options for degradation and mineralization of microorganic pollutants (Kwon et al. 2015). The \({\text{SO}}_{4}^{ \cdot - }\) can be formed by persulfate (e.g., PS, S2O82−) or peroxymonosulfate (e.g., PMS, HSO5−) in various activation ways, such as heat (Hisao et al. 2008), alkaline (Qi et al. 2016), UV (Mahdiahmed and Chiron 2014) and transition metal (Rodriguez et al. 2014). Compared to OH·, the selectivity, redox potential and half-life of \({\text{SO}}_{4}^{ \cdot - }\) were high (Ahmed et al. 2012; Wang and Wang 2018). In previous studies, UV-activated persulfate was found to significantly improve the removal efficiency of M. aeruginosa cells up to 98.2%, compared to employing UV inactivation alone, while at the same time it might cause complete M. aeruginosa cell lysis and damage, posing environmental risk for producing IOM, a secondary pollutant (Wang et al. 2016). Similarly, Jia et al. (2017) reported that UV/H2O2-generated OH· from pre-oxidation could enhance the coagulation of M. aeruginosa from Fe (II) with the removal efficiency of 94.7%. However, additional coagulants (e.g., Al3+, Fe3+) were also introduced to the coagulation process, resulting in sever cell lysis with secondary environmental pollutions. Those pre-oxidation processes are relatively independent and separate to subsequent coagulation process, which possess higher potential dangers owing to a longer contact time. Besides, it has been well documented that the cell damage was rarely observed in the coagulation process with potassium ferrate (VI); however, the removal efficiency of M. aeruginosa was rather poor (Zhou et al. 2014). Fe (II) /PMS has been proved to remove M. aeruginosa cell efficiently with a moderate cell lysis (Zhou et al. 2020); however, the instability of PMS in natural water impeded its wide application. Therefore, the strategies for enhancing the removal efficiency of M. aeruginosa without cell damage are highly desired in an environmental and economical way.

In this study, the oxidation with Fe2+ activated persulfate (i.e., PS, S2O82−) to generate \({\text{SO}}_{4}^{ \cdot - }\) was applied to remove M. aeruginosa without additional coagulants and serious cell lysis. The main objective of this study is to moderate the oxidation processes and simultaneously form in situ Fe (III) as coagulants to control AOM as well as M. aeruginosa. More specially, this study is aimed to: (1) determine the optimal concentration of FeSO4 and Na2S2O8 for M. aeruginosa removal and simultaneously maintaining cell integrity; (2) identify the cell structure and the characteristics of AOM; and (3) investigate the possible mechanisms involved in the pre-oxidation and coagulation process. This study was carried out in the laboratories of School of Environment and Energy, South China University of Technology, Guangzhou, China, in 2018.

Materials and methods

Materials and reagents

Microcystis aeruginosa FACHB-905 was obtained from the Institute of Hydrobiology, Chinese Academy of Science, and cultured in BG11 media. The cultures were incubated at 25 °C under illumination (2500 lx) with a 12 h light/12 h dark cycle in powder BG11 medium (Qingdao Haibo Institute of Biology) (Sheng et al. 2019). M. aeruginosa cells in late exponential growth phase were harvested and diluted by Milli-Q water to achieve a final cell density of 1.0 × 106 cells/mL (Gu et al. 2017). Afterward, the pH of M. aeruginosa solution was adjusted to 7.5 approaching ambient environment with 0.1 mmol/L H2SO4 and 0.1 mmol/L NaOH.

All the chemicals used were analytical grade. Ferrous sulfate (FeSO4), sodium sulfate (Na2S2O8), NaOH (≥ 98%), H2SO4 (≥ 99%), methanol (MeOH, ≥ 98%) and tert-butyl alcohol (TBA) were obtained from Aladdin Chemistry Co. Ltd., China. The stock solutions of FeSO4 were freshly prepared each time, and Na2S2O8 solution was stored for only one week.

Pre-oxidation and coagulation processes

The processes of generating \({\text{SO}}_{4}^{ \cdot - }\) in Fe2+/PS system are described as follows (Gu et al. 2017):

$${\text{S}}_{2} {\text{O}}_{8}^{2 - } + {\text{F}}_{{\text{e}}}^{2 + } \to {\text{F}}_{{\text{e}}}^{3 + } + 2{\text{SO}}_{4}^{ \cdot - } \quad k = 2.0 \times 10^{1} \;{\text{M}}^{ - 1} \,{\text{S}}^{ - 1}$$
(1)
$${\text{SO}}_{4}^{ \cdot - } + {\text{F}}_{{\text{e}}}^{2 + } \to {\text{F}}_{{\text{e}}}^{3 + } + {\text{SO}}_{4}^{2 - } \quad k = 4.6 \times 10^{9} \;{\text{M}}^{ - 1} \,{\text{S}}^{ - 1}$$
(2)
$${\text{S}}_{2} {\text{O}}_{8}^{2 - } + 2{\text{F}}_{{\text{e}}}^{2 + } \to 2{\text{F}}_{{\text{e}}}^{3 + } + 2{\text{SO}}_{4}^{2 - } \quad k = 3.1 \times 10^{4} \;{\text{M}}^{ - 1} \,{\text{S}}^{ - 1}$$
(3)

It is worth noting that Fe2+ has the dual-functional capability of activator and coagulant in this work that pre-oxidation and coagulation can be completed at the same time.

The prepared M. aeruginosa suspension was added with desired amount of Na2S2O8 and FeSO4. Then, the suspension was mixed at 225 rpm for 6 min and subsequently at 40 rpm for 20 min, enabling a complete oxidation and coagulation, as shown by the Eqs. (1)–(3) that \({\text{SO}}_{4}^{ \cdot - }\) for oxidation and Fe2+ for coagulation. Thereafter, the supernatants were taken out from 2 cm below the surface of water at specific time intervals and separated into a few of subsamples for further investigation of the reaction mechanisms. The cells settled down to the bottom were analyzed using scanning emission microscopy (SEM) and flow cytometry (Accuri C6, Becton Dickinson, USA). Given the limited effects on the M. aeruginosa cell removal, the impacts of co-existing anions (i.e., Cl−, NO3− and HCO3−) from BG-11 media would not be further discussed in the following sections (Wang et al. 2016). Additionally, the quenching reaction by the scavenger of MeOH and TBA was conducted to determine the functional radicals.

Analytical methods

Removal efficiency of OD680 and chlorophyll-a

As the M. aeruginosa cell concentration was positively correlated with the optical density (OD) at 680 nm, the M. aeruginosa density was determined at wavelength of 680 nm using the spectrophotometer (U-3900, Hitachi, Japan). The removal efficiency of OD680 was calculated as follows.

$$R_{{{\text{OD}}_{680} }} = \frac{{\left( {{\text{OD}}_{{680_{i} }} - {\text{OD}}_{{680_{t} }} } \right)}}{{{\text{OD}}_{{680_{i} }} }} \times 100\%$$
(4)

where \({\text{OD}}_{{680_{i} }}\) and \({\text{OD}}_{{680_{t} }}\) were the algal OD at initial and time t (min), respectively.

Microcystis aeruginosa cell density was also linearly related to the content of chlorophyll-a, reflecting the primary productivity and photosynthesis ability of algal (Ni et al. 2015). The extracted chlorophyll-a (Chl-a) was determined by measuring the optical value at 649 nm and 665 nm with the spectrophotometer (U-3900, Hitachi, Japan) using ethanol (95%) solution. The content of Chl-a was calculated as follows (Sheng et al. 2019):

$$C_{{\text{chl-a}}} = 13.95A_{665} - 6.88A_{649}$$
(5)

The removal rate of chlorophyll-a was calculated by the change of Chl-a concentration as follows:

$$R_{{{\text{chla}}}} = \frac{{({\text{chla}}_{i} - {\text{chla}}_{t} )}}{{{\text{chla}}_{i} }} \times 100\%$$
(6)

where chlai and chlat were the chlorophyll-a content at the initial time and time t (min), respectively.

Zeta potential and residual Fe

Zeta potential of M. aeruginosa suspensions was measured using the Zetasizer (Nano S90, Malvern, Britain) after slow mixing. The residual Fe was determined using an inductively coupled plasma optical emission spectrometer (ICP-OES, OPTIMA 8000, PerkinElmer, USA).

Dissolved organic carbon (DOC) and UV254

The M. aeruginosa supernatant was filtered by a 0.45-μm cellulose acetate membrane (JIN TENG Technology CO., Ltd, China). Then, the TOC analyzer (Multi N/C 2100, JENA, Germany) and ultraviolet spectrometer (U-3900, Hitachi, Japan) were used to analyze the dissolved organic carbon (DOC) and UV254 of the M. aeruginosa supernatant.

Three-dimensional fluorescence excitation emission matrix (EEM) spectra

The fluorescence variation of AOM was analyzed using the three-dimensional fluorescence excitation emission matrix (EEM) spectrophotometer (F-7000, Hitachi, Japan). The scanning parameter of emission (Em) and the excitation (Ex) spectra matrices were in the range of 250–550 nm and 220–450 nm, respectively. Scan rate was set to 2400 nm/min at a gradient of 5 nm (Wang et al. 2016).

K+ concentration and flow cytometer measurement

To evaluate the cell integrity, the concentration of K+ released from M. aeruginosa cell was determined using the atomic absorption spectrometer (Shimadzu AA-7000, Japan) (Zhou et al. 2014). To assess the cellular activity, chlorophyll fluorescence of the precipitated cells by integrated Fe2+/PS pre-oxidation and coagulation was determined using flow cytometry (Accuri C6, Becton Dickinson, USA). Yellow fluorescence in channel FL2 (585 nm) was detected for chlorophyll content, and the flow rate was 20 μL/min.

Scanning electron microscopy (SEM)

The M. aeruginosa cells in the different samples were obtained at 4000 rpm for 10 min using centrifuge. After discarding the supernatant, the precipitated M. aeruginosa cells were mounted on copper stubs, coated with gold and examined under a SEM (VEGA TS 5136 MM, TESCAN Brno s.r.o, Czech Republic).

Extracellular and intracellular MCs

The MCs concentrations were determined by the enzyme-linked immunosorbent assay kit (Beacon Analytical Systems Inc, USA). The supernatant samples were extracted and then filtered with 0.45-μm cellulose acetate membrane (JIN TENG Technology CO., Ltd, China) to determine extracellular MCs. The other identical samples were for the total MCs analysis described by the previous methods (Bi et al. 2019). The intracellular MCs were obtained by the difference between total MCs and extracellular MCs.

Statistical analysis

Statistical analysis was performed using SPSS 19.0 (IBM, USA). One-way analysis of variance (ANOVA) was used to assess the variance with significance level of 5% (P < 0.05). The error bar indicates the standard error of triplicate measurements. Different capital letters indicate significant differences.

Results and discussion

The effect on M. aeruginosa removal in the integrated pre-oxidation and coagulation

The lifetime of \({\text{SO}}_{4}^{ \cdot - }\) radicals was 3–4 × 10−5 s, while that of the OH· radicals was only 2 × 10−8 s (Devi et al. 2016). As shown in Fig. 1, a two-stage reduction kinetics of M. aeruginosa in Fe2+/PS system was observed with a rapid change in the initial 30 min and a stationary phase thereafter. Therefore, the settling time of 30 min was regarded as the optimum time for future research, which not only curtailed the treatment process but also reduced the risk of releasing undesirable compounds of IOM.

Fig. 1
figure 1

Evolution of removal rate of OD680 over time by integrated Fe2+/PS pre-oxidation and coagulation treatments in different FeSO4 dosage systems. Data are presented as mean ± standard deviation (n = 3)

Full size image

Figure 2 presents the removal efficiency of M. aeruginosa after settling for 30 min at different FeSO4 concentrations. As FeSO4 concentrations changed from 0.25 to 0.15 mmol/L, the removal efficiencies of OD680 and Chl-a increased from 4.80 to 98.8% and from 8.21 to 99.1%, respectively (P < 0.05). However, after the FeSO4 concentration increased to 0.10 mmol/L, the removal efficiencies of OD680 and Chl-a changed slightly. This was attributed to the fact that the amount of available \({\text{SO}}_{4}^{ \cdot - }\) and OH· contents decreased, resulting from the competitive reaction of overdose Fe2+ ions shown as follows (Liang et al. 2004; Neppolian et al. 2009).

$${\text{Fe}}^{2 + } + {\text{SO}}_{4}^{ \cdot - } \to {\text{Fe}}^{3 + } + {\text{SO}}_{4}^{2 - }$$
(7)
$${\text{Fe}}^{2 + } + {\text{OH}}^{ \cdot } \to {\text{Fe}}^{3 + } + {\text{OH}}^{ - }$$
(8)
Fig. 2
figure 2

Residual Fe and removal rate of OD680 and Chl-a of the Microcystis suspensions after settling 30 min under different FeSO4 dosages. Data are presented as mean ± standard deviation (n = 3)

Full size image

Figure 2 also shows the change in the percentage of residual Fe. With the addition of FeSO4, the percentage of residual Fe firstly decreased and then increased slightly when the dosage of FeSO4 reached 0.10 mmol/L. The freshly produced in situ Fe (III) by hydrolyzation was significantly slower than the preformed Fe (III), leading to a superior reduction of M. aeruginosa (Ma et al. 2014). The insufficient generated Fe (III) hindered the massive aggregation of M. aeruginosa cells by adsorption bridging and charge neutralization. This finding corroborated the fact that M. aeruginosa was not removed efficiently with single 0.10 mmol/L Na2S2O8, nor 0.10 mmol/L FeSO4, as shown in Fig. S1. Therefore, the addition of 0.10 mmol/L FeSO4 could simultaneously achieve a relatively high removal efficiency of M. aeruginosa and produce low residual Fe (0.24 mg/L), which was below the standard threshold of Fe content in drinking water (0.3 mg/L, GB5749-2006). Additionally, the optimal Fe2+/PS mass ratio of 1:1 is observed in Fig. S2. The low Fe2+/PS mass ratio was less effective for AOM degradation, while the excessive oxidants would cause the IOM release. Therefore, the best optimal conditions for controlling M. aeruginosa was determined to be 0.10 mmol/L FeSO4 and 0.10 mmol/L Na2S2O8. Moreover, the initial dosage of 0.10 mmol/L Na2S2O8 (23.8 mg/L) had been far below the standard value of sulfate content in drinking water (250 mg/L, GB5749-2006).

Furthermore, it was noting that the optimal FeSO4 dosage of 0.10 mmol/L and the Fe2+/PS mass ratio of 1:1 in the present study was cost-effective compared to the previous studies (Wang et al. 2016; Jia et al. 2017). It has been reported that approximately 6 mmol/L sodium persulfate was consumed to achieve 98.2% removal efficiency of M. aeruginosa in the UV/persulfate process (Wang et al. 2016), while in the UV/H2O2 process, with the dosage of 0.375 mmol/L H2O2 and 0.125 mmol/L FeSO4, a high M. aeruginosa removal efficiency of 94.7% was observed (Jia et al. 2017). However, the high cost of UV process would no doubt impede its wide application. Additionally, compared to Fe2+/PS, other conventional coagulation methods also required high dosage of chemicals with high expenditure, such as 100 mg/L FeCl3 (Li et al. 2015), 7.31 mg/L Chitosan (Pei et al. 2014) and 15 mg/L AlCl3 (Sun et al. 2012). PACl has been reported to be a cost-effective coagulant with the dosage of 4 mg/L; however, a serious cell damage along with secondary environmental pollutions was observed (Sun et al. 2013). Therefore, Fe2+/PS process was a promising method for removing M. aeruginosa and controlling cyanobacteria blooms in an environmentally friendly and cost- effective way.

The change of pH and zeta potential in the integrated pre-oxidation and coagulation

Table S1 presents the pH and zeta potentials of M. aeruginosa suspension under different FeSO4 concentrations. The untreated M. aeruginosa cell possessed a relatively high negatively charged zeta potential of − 38.9 mV by the massive AOM. Subsequently, the zeta potentials of the cell increased to 3.27 mV along with the FeSO4 dosage constantly changing from 0.25 to 0.15 mmol/L, and the pH value decreased. In comparison, the zeta potentials increased to − 30.15 mV and − 17.76 mV, respectively, when using 0.10 mmol/L Na2S2O8 or 0.10 mmol/L FeSO4 alone, and the pH changed slightly (Table S2), while a much higher value of − 0.814 mV was observed in the Fe2+/PS system. The higher zeta potential observed in the Fe2+/PS system was because the generated \({\text{SO}}_{4}^{ \cdot - }\) oxidized negatively charged AOM and the formed in situ Fe (III), which was conducive to aggregating AOM and M. aeruginosa cells (Liu et al. 2017; Ma et al. 2012a, b). It was confirmed that the increased zeta potential of M. aeruginosa was the crucial reason for coagulation enhancement (Xie et al. 2016). Moreover, Liu et al. (2017) found that the zeta potential of M. aeruginosa cells was decreased significantly by pre-oxidation of ozone and permanganate, in which the negatively charged IOM was massively released due to cell rupture. However, this finding indicated that Fe2+/PS was not likely to affect the cell structure, since the decrease in zeta potential caused by the released IOM was not observed.

AOM removal in the integrated pre-oxidation and coagulation

It has been well documented that lower concentrations of AOM could enhance coagulation via adsorption bridging effect (Xie et al. 2016). However, the extremely high levels of AOM would increase the negative charge of algal cells, resulting in coagulation suppression (Pivokonsky et al. 2006). Therefore, the oxidation of AOM was considered as a crucial step for enhancing the removal of M. aeruginosa.

The DOC and UV254 decreased sharply (P < 0.05) along with FeSO4 addition, as shown in Fig. 3. The Fe2+ activated Na2S2O8 to generate \({\text{SO}}_{4}^{ \cdot - }\) for AOM degradation in the pre-oxidation, and simultaneously, the in situ formed Fe (III) aggregated M. aeruginosa cells as well as AOM, along with the changes in M. aeruginosa removal and residual Fe (Fig. 2). Furthermore, the declining trend of DOC indicated that Fe2+/PS caused limited damage to the M. aeruginosa cell walls. Else, the DOC level would increase significantly, due to the released IOM and internal metabolites caused by cell lysis (Zhou et al. 2014). The unsaturated carbon bonds and aromatic structures of the organic compound can be reflected by UV254 (Liu et al. 2017). The newly formed in situ Fe (III), derived from the pre-oxidation of Fe2+, showed better performance of AOM degradation compared to preformed Fe (III), particularly those with unsaturated structures (Ma et al. 2014). Therefore, the decline of UV254 in Fig. 3 affirmed the reaction between Fe (III) and organics with unsaturated carbon chains, such as humic acid (Silvia et al. 2010) or protein (Ma et al. 2016; Pivokonsky et al. 2012), which was pivotal step for M. aeruginosa removal.

Fig. 3
figure 3

Mineralization of organic matters after settling 30 min under different FeSO4 dosage. Data are presented as mean ± standard deviation (n = 3)

Full size image

EEM spectra of extracellular AOM

Three-dimensional EEM was employed for further investigation of chemical composition of the extracellular AOM in Fe2+/PS process. Four fluorescence peaks were identified, as shown in Fig. 4a–e. Peak A at Ex/Em of 270/442 nm and peak C at 350/432 nm were associated with fulvic- and humic-like substances, respectively. Peak T1 (280/330 nm) and peak T2 (230/330 nm) were related to dissolved microbial metabolites and protein-like compounds, respectively (Coble 1996; Hudson et al. 2007). High intensities of Peak T1 and T2 were observed in the control (Fig. 4a), suggesting that AOM contained a higher concentration of dissolved microbial metabolites and protein-like substances (Huang et al. 2014; Ou et al. 2011). It was found that protein-like substances were relatively easy to be decomposed during pre-oxidation treatment process. Therefore, as shown in Fig. 4c, almost all protein-like components and a part of dissolved microbial metabolites were oxidized. The protein-like substances, with better hydrophilicity and poor aggregation ability, were preferentially oxidized by free radicals (Tian et al. 2018). However, as the concentrations of FeSO4 were increased to 0.125 mmol/L, the intensity of peaks A and C was significantly increased, along with the weakened intensity of peaks T1 and disappearance of peak T2 in contrast (Fig. 4e). The M. aeruginosa cells were under chemical or physiological stress during the coagulation process (Khatoon et al. 2018; Ma et al. 2016), which may accelerate the secretion of the fulvic- and humic-like substances and release IOM due to cell lysis. These results well explained the declining tendency of DOC and UV254 observed in Fig. 3 and demonstrated that the free radicals were relatively easy to react with protein-like substances and dissolved microbial metabolites containing greater unsaturated bonds. As shown in Fig. S3, the AOM changed slightly in single Na2S2O8 or FeSO4 system, resulting in poor M. aeruginosa removal. This finding indicated that a decisive process of removing M. aeruginosa was to generate free radical for pre-oxidation.

Fig. 4
figure 4

Fluorescence EEM spectra of extracellular AOM. a Control, b 0.05 mmol/L FeSO4, c 0.075 mmol/L FeSO4, d 0.10 mmol/L FeSO4, e 0.125 mmol/L FeSO4

Full size image

The role of free radicals in M. aeruginosa removal was further investigated. Table S3 shows the M. aeruginosa removal efficiencies with MeOH and TBA co-existed in Fe2+/PS system. This result showed that the inhibition degree caused by MeOH was greater than TBA, but the whole reaction was not completely suppressed. MeOH and TBA transformed the degradation and mineralization process of M. aeruginosa by quenching the useful free radical. Both \({\text{SO}}_{4}^{ \cdot - }\) and OH· ould be effectively inhibited by the scavenger of MeOH, while TBA was only effective for OH· inhibition OH (Wang et al. 2016). This phenomenon implied that both \({\text{SO}}_{4}^{ \cdot - }\) and OH· were responsible for oxidizing M. aeruginosa cells in this work.

Cell viability

Figure 5 shows the chlorophyll fluorescence intensity of the M. aeruginosa cell flocs measured by flow cytometry. M. aeruginosa possesses Chl-a auto-fluorescence, and its intensity is positively correlated with chlorophyll content, which can reflect the cellular activity (Hyka et al. 2013). Both the proportion of M. aeruginosa cells with higher activity and chlorophyll fluorescence intensity decreased with higher FeSO4 dosage (Fig. 5). Potassium (K) was one of the significant elements for maintaining cell osmotic pressure and reflected cell membrane integrity (Gu et al. 2017). As shown in Fig. S4, with increased FeSO4 concentration, the K+ released from intracellular fluid increased significantly (P < 0.05). Jia et al. (2017) reported that after H2O2 treatment, only small amounts of cells were damaged with approximately 50% K+ released. In the present study, under optimal dosage of 0.10 mmol/L FeSO4, only 42% K+ was released from M. aeruginosa, and consequently, the majority of M. aeruginosa cells remained intact. Additionally, under the optimal level, the cell percentage of high activity only decreased 15.64% compared to the controls, implying that only small amounts of M. aeruginosa cell membrane might have been impaired in Fe2+/PS process.

Fig. 5
figure 5

Chlorophyll fluorescence of the Microcystis after settling 30 min. a Control, b 0.025 mmol/L FeSO4, c 0.075 mmol/L FeSO4, d 0.10 mmol/L FeSO4

Full size image

Physiology morphology of M. aeruginosa cells

In order to further investigate the cell membranes integrity under optimum dosage of FeSO4 and Na2S2O8, the surface morphology of M. aeruginosa cells before and after Fe2+/PS treatment is shown in Fig. 6 by SEM. The M. aeruginosa cells in control systems (Fig. 6a) were smooth, and some secretions were observed, presumably due to the naturally released metabolites (Jia et al. 2017). After the addition of optimal Fe2+/PS dosage, a small amount of M. aeruginosa cells tended to wrinkle, and the surface morphologies were moderately altered without the leakage of cell inclusion (Fig. 6b), which was in accordance with the K+ release in Fig. S4. This might have attributed to the increase in the extracellular polymeric substances (EPSs) produced by M. aeruginosa, enhancing the formation of the protective shield for M. aeruginosa cells against the external stress (Zhao et al. 2011). However, if the Fe2+/PS dosage increased, the protective barrier provided by EPS would be weakened and diminished. This result indicated that only a small scale of algal cell membranes was damaged in the Fe2+/PS process, and the majority of M. aeruginosa cells, however, still remained integrated and spherical.

Fig. 6
figure 6

The SEM images of surface morphology of M. aeruginosa cells a before, b after 30-min settling by Fe2+/PS pre-oxidation and coagulation treatments

Full size image

The release of microcystins (MCs) in the integrated pre-oxidation and coagulation

Given the fact that the MCs are highly toxic, the change in microcystins is presented in Fig. 7. As FeSO4 dosage increased from 0.025 to 0.10 mmol/L, the extracellular MCs decreased significantly (P < 0.05), while the change in intracellular MCs was not obvious (P > 0.05), implying that most M. aeruginosa cells were intact. However, the intracellular MCs decreased remarkably and the extracellular MCs increased slightly, when the FeSO4 dosage reached 0.125 mmol/L. The M. aeruginosa cells lysis occurred, and numerous intracellular MCs were released from M. aeruginosa cells under this condition. However, the content of sulfate radical was high and high removal of MCs was observed. Compared to 0.10 mmol/L FeSO4, the concentrations of extracellular MCs declined slightly at the concentration of 0.15 mmol/L. This finding indicated that the degradation rate of MCs by sulfate radical was greater than the release rate of MCs from M. aeruginosa cells.

Fig. 7
figure 7

The concentrations of intracellular and extracellular MCs after 30-min settling under different FeSO4 dosages. Data are presented as mean ± standard deviation (n = 3)

Full size image

Wang et al. (2016) reported that the structure of M. aeruginosa cells was completely out of shape after the 120-min treatment by UV/PS systems. Slight damage of the spherical surface occurred during the UV/H2O2 pre-oxidation process. However, additional coagulants were needed along with the costs incurred. Although there was no evident alteration in M. aeruginosa cell morphology after ferrate (VI) oxidation, the removal efficiency of M. aeruginosa was relatively poor (Zhou et al. 2014). Therefore, compared to these previous studies, the Fe2+/PS process is not only effective in removing M. aeruginosa cells without additional coagulants, but also reliable in reducing the risk of secondary pollution, resulting from the better control of the microcystins.

Conclusion

This study demonstrated that the Fe2+/PS process was a highly effective method for removing M. aeruginosa cells and AOM via in situ Fe (III) formed. The generated \({\text{SO}}_{4}^{ \cdot - }\) and OH· were both responsible for the efficient degradation of AOM. Moreover, the dissolved microbial metabolites and protein-like substances seemed more susceptible to be oxidized, and subsequently, the coagulation process was significantly enhanced. Fe2+/PS process has been proved to be highly efficient in removing M. aeruginosa cells and AOM, due to in situ formed Fe (III) without the additional coagulants. SEM image indicated that the optimal Fe2+/PS ratio reduced the cyanobacterial cell damage and MCs had been well controlled, preventing from the risk of secondary pollution. Therefore, it is concluded that Fe2+/PS process is a promising and environmental-friendly solution to the problems associated with cyanobacteria blooming.