Introduction

Heavy metal contamination has become a considerable environmental concern throughout the world (Gryta et al. 2014; Liu et al. 2017). As the continuous development of traffic and industrial activities, the environmental emission of hazardous heavy metals has been multiplied (Al-Qodah 2006; Islam et al. 2015; Liao et al. 2014; Yang et al. 2015). In order to meet the ever-increasing legislative standards, the discharge of harmful heavy metals need to be reduced (Peng et al. 2017; Tang et al. 2018). The zinc ion, which is an essential nutrient for most organisms but can be toxic at high concentration, is one of the heavy metals that cause most immediate concern according to the report of World Health Organization (WHO), as discussed in Wei et al. (2017).

Zn2+ was discharged into waterbodies through industrial activities such as tannery, electroplating, and metallurgical, which caused the ever-increasing contamination of water environment (Betts et al. 2013). Conventional precipitation method for zinc removal from water is very expensive. Even worse, it generates toxic substances (i.e., metal sulfide precipitates). Adsorption, as a highly efficient way in zinc removal, receives considerable interest nowadays. Many studies have been carried out in order to find out the cost-effective adsorbents. Chen et al. (2011) used biochars produced by hardwood and corn straw to adsorb zinc. The adsorption capacities were recorded to be 11 mg g−1 for corn straw and 4.54 mg g−1 for hardwood. Veli and Alyüz (2007) applied natural clay to adsorb zinc, with the adsorption capacities of 80.64 mg g−1. Nasernejad et al. (2004) studied the removal of zinc by adsorption on carrot residues and the maximum adsorption capacity was found to be 29.61 mg g−1.

In biological wastewater treatment plants (WWTPs), sludge biomass is considered as a good bio-adsorbent for heavy metal removal. Under neutral pH conditions, the negatively charged sludge surfaces can easily interact with metal cations through electrostatic interaction. Since zinc is considered to be non-biodegradable, it could be accumulated in the activated sludge systems. Consequently, inhibition on biological removal processes would be accompanied. Generally, the effects of zinc on nitrification, denitrification, and anammox process have been studied by few researchers (Hu et al. 2004; Song et al. 2018; Zhang et al. 2015). It was reported that zinc in excess sludge of some municipal wastewater treatment plants can be as high as 2170.6 and 2841.2 mg kg−1 dry sludge (Zhang et al. 2015). Although zinc ions could be removed by bio-sludge, the adsorption capacity might be diverse from different types of bio-sludge, owing to their intrinsic physicochemical properties under different operational conditions. The aggregate morphology of sludge (flocs, granule, or biofilm) is an important factor affecting the heavy metal adsorption. Sludge floc has a loosely structures and poor solid–liquid separation. Previous studies are mainly focused on the zinc adsorption by conventional activated sludge flocs, and different adsorption efficiencies were reported (Remenárová et al. 2012; Sooan et al. 2010; Yang et al. 2010). In contrast to the sludge flocs, granular sludge has the different pore size and pore structure which enhances the mass transfer into the inner section (Song et al. 2017; Tang et al. 2017). In particular, granular sludge process has been considered as a promising heavy metal adsorption technology from wastewaters (Sun et al. 2011).

Extracellular polymeric substances (EPS), which have been considered as a key factor for metal adsorption, are complex mixtures synthesized by microbial cells and with a high molecular weight (Yang et al. 2015). Protein (PN) and polysaccharide (PS) are the main component of EPS, and the functional groups (like carboxyl, hydroxyl, phosphoric, sulfhydryl, and amide) of EPS make it capable of binding heavy metals. It is reported that two types of mechanisms are involved in the binding reactions between EPS and metals. The first is electrostatic interactions through the negatively charged groups on the EPS surface, and the other is complexing bonds due to the existed carboxylic and phosphoric groups in EPS (Sheng et al. 2013). This is an important mechanism for the inner biomass of sludge to prevent potential inhibition caused by heavy metals (Sajjad and Kim 2015). In addition, the bind interactions between EPS and heavy metals are in relation to these physicochemical processes including complexation, coordination, ion exchange, adsorption, and inorganic micro-precipitation (Aquino and Stuckey 2004). The forms of EPS that exist outside of cells usually are subdivided into three parts, soluble EPS (S-EPS), loosely bound EPS (LB-EPS), and tightly bound EPS (TB-EPS) (Sheng et al. 2010). The difference in chemical composition of these EPS may affect their metal adsorption abilities. The amount and composition of sludge EPS, and the available functional groups can also be affected by different cultivation conditions, which in turn affect the adsorption performance. The EPS secreted by sludges would differ from each other under different operational conditions including substrates and hydraulic loading. Indeed, the production of EPS was regarded as a response to the imposed stress or triggers among the operational conditions. These stress conditions can be summarized as physical shear, bacteriophage, organic contaminants, biocides, antibiotics, and heavy metals (Vu et al. 2009).

The main objective of the present study is to reveal the role of EPS in Zn2+ adsorption in different sludge systems. Four different types of sludge (activated sludge, denitrifying sludge, short-cut nitrification sludge, and anammox granular sludge) were applied. The adsorption equilibrium and the adsorption kinetics were studied and compared. Then, the variations of the EPS characteristics were analyzed to explore the detailed information of EPS on Zn2+ adsorption. The changes of the chemical structures in EPS were examined by comparing the FTIR spectroscopy, amount, and composition of EPS samples. The contribution of EPS fractions to zinc adsorption was also evaluated.

Materials and methods

Sludge samples

Four different kinds of sludge, including activated sludge, denitrifying sludge, short-cut nitrification sludge, and anammox granular sludge, were applied in this study. The activated sludge and denitrifying sludge were obtained from a full-scale aeration tank and anoxic tank at the Yanghu municipal wastewater treatment plant (100,000 m3 day−1) in Changsha, China. The short-cut nitrification sludge and anammox sludge were collected from a lab-scale glass-made upflow airlift reactor (ALR) and an upflow anaerobic sludge bed (UASB), respectively. Both reactors were stably operated for more than 2 years. The sludge samples were washed at least 3 times with buffer prior to experiments to remove residuals. Some physiochemical characteristics including the EPS contents of the four sludge samples are shown in Table 1, and images were taken by SEM-EDS to investigate their surface morphology after metal adsorption (Fig. S1).

Table 1 The properties of the tested four sludges
Full size table

Adsorption of Zn2+ by sludge

Zn2+ solution preparation

The aqueous zinc solution in this study was prepared using ZnSO4·7H2O (analytical grade, provided by Kermel). The stock solution of zinc ions (1 g L−1) was used to prepare the different degree of dilute solutions through the dilution with DI water. In order to prevent the formation of metal hydroxides, the stock solution was acidified with few drops of concentrated HCl.

Adsorption batch experiments

The adsorption batch experiments were carried out by adding a certain volume of sludge to 100 mL of metal ion solution in a 250-mL flask. The pH and temperature of the adsorption experiments were control at a constant (pH = 6, T = 25 °C) in the present study. The sample flasks were then stirred at 180 rpm and 25 °C in a water bath shaker. Batch samples were taken at different intervals to investigate the adsorption kinetics and maximum adsorption capacity. The effect of time on Zn2+ adsorption is studied at a constant sludge dosage of 5 g SS L−1 with an initial Zn2+ concentration of 50, 100, and 200 mg L−1. The effect of initial Zn2+ concentration is studied by varying Zn2+ concentration from 10 to 200 mg L−1 with sludge dosage of 5 g SS L−1. While the effect of sludge dose is conducted by using different sludge amounts from 1 to 10 g SS L−1 at a fixed Zn2+ concentration of 50 mg L−1.

Analysis of metal ions

The concentration of unabsorbed zinc ion in the batch solution was determined using inductively coupled plasma optical emission spectrometry (ICP-OES, Agilent 5100-VDV). In order to remove the sludge particles at each sampling, 3 mL of suspension was centrifuged at 5000 rpm for 5 min. Then, the supernatant was filtered using a 0.45-μm syringe filter and placed into a 2-mL centrifuge tube at 4 °C prior to determining the Zn2+ concentrations. Dilution of samples was made with 4% ultrahigh purity HNO3. In addition, sludge samples without metal ions were used as control to equilibrate the metal concentration in solutions. The mass balance equation (1) was used to calculate the amount of metal ion adsorbed by sludge:

$$ {q}_{\mathrm{e}}=\frac{\left({C}_0-{C}_{\mathrm{e}}\right)V}{W} $$
(1)

where qe (mg g SS−1) is the equilibrium adsorption capacity, C0 (mg L−1) and Ce (mg L−1) are the initial and equilibrium metal concentrations in solution, V (L) is the volume of solution, and W (g SS) is the weight of the suspended solids from sludge sample. All treatments were carried out with two replicates. The concentrations of adsorbed Zn2+ were the means of the duplicate experimental results.

Adsorption isotherm study

Adsorption isotherm experiments were carried out at a constant sludge dosage of 5 g SS L−1. The initial concentrations of Zn2+ ranged from 10 to 200 mg L−1. The initial pH was controlled at 6, and the final pH when the equilibrium reached was measured. The sample flasks were then stirred at 180 rpm and 25 °C in a water bath shaker. In this study, two of the most common adsorption models, that is, Langmuir model (2) and Freundlich model (3), were applied to fit the experimental data, which is represented by the following equations:

$$ {q}_{\mathrm{e}}=\frac{q_{\mathrm{m}}b{C}_{\mathrm{e}}}{1+b{C}_{\mathrm{e}}} $$
(2)
$$ {q}_{\mathrm{e}}={K}_{\mathrm{F}}{C_{\mathrm{e}}}^{1/n} $$
(3)

where qm is the maximum metal uptake under the given conditions and b is a Langmuir constant related to the affinity between the adsorbent and the adsorbate. KF and n are Freundlich constants, which are correlated to the maximum adsorption capacity and adsorption intensity, respectively. The essential features of the Langmuir model can be expressed in terms of a dimensionless constant separation factor, RL, which is defined by Hall et al. (1966) as:

$$ {R}_{\boldsymbol{L}}=\frac{\mathbf{1}}{1+b{C}_0} $$
(4)

where b (1 mg−1) is the Langmuir isotherm constant and C0 (mg L−1) is the highest initial metal concentration. The value of RL can provide information as to whether the isotherms are irreversible (RL = 0), favorable (0 < RL < 1), linear favorable (RL = 1), or unfavorable (RL > 1).

Kinetic modeling

The adsorption kinetics characteristics of Zn2+ by different sludges were analyzed using pseudo-first-order kinetic model (5), pseudo-second-order kinetic model (6), and intra-particle diffusion model (7), which can be expressed as follows:

$$ \ln \left({q}_{\mathrm{e}}-{q}_t\right)=\ln\ {q}_{\mathrm{e}}-{k}_1t $$
(5)
$$ \frac{t}{q_t}=\frac{1}{q_{\mathrm{e}}^2{k}_2}+\frac{t}{q_{\mathrm{e}}} $$
(6)
$$ {q}_t={k}_{\mathrm{i}}\left({t}^{1/2}\right)+C $$
(7)

where qe and qt (mg g SS−1) are the amounts of Zn2+ adsorbed at equilibrium and at time t, respectively. k1 (min−1), k2 (g SS mg−1 min−1), and ki (mg g SS−1 min−0.5) are the corresponding rate constant of each model, respectively.

Other analytical methods

The suspended solids (SS) of the sludge were measured by gravimetric analysis. The pH was recorded using a digital pH meter (pHS-9V). For sludge flocs, zone settling velocity was used to represent their settling velocity (SV) values according to Tang et al. (2011). For anammox granules, the SV value was determined by averaging the time taken for an individual granule to settle at a certain height in a glass column filled with tap water. The zeta potential of each sludge samples was measured using a Zetasizer Nano ZS (Malvern, UK) at 25 °C. Sludge suspension was controlled at a concentration of approximately of 0.5 g SS L−1 and adjusted to an appropriate pH value. All the data were the average of three measurements and the standard deviation was considered as the error range. The surface elemental composition of the sludge samples was measured as the previous study (Ali et al. 2016). The samples coated with gold were observed using a scanning electron microscope (SEM, JSM-IT300LA, JEOL, Japan).

The EPS of four sludge samples were extracted using a modified “heating” extraction method according to Li et al. (2017) and Zhang et al. (2017). Sludge samples were washed three times with phosphate-buffered solution (pH 7.0), and then the sludge in the tube was re-suspended in 10 mL of PBS solution to extract the soluble EPS, loosely bound EPS (LB-EPS), and tightly bound EPS (TB-EPS). Details of the extraction steps were mentioned in supplementary materials (SM). Each phase of EPS samples was filtered using a 0.45-μm syringe filter prior to analysis. The contents of protein (PN) and polysaccharide (PS) were analyzed according to Song et al. (2018).

The characteristic absorption peaks of chemical bond in EPS samples and their interaction with Zn2+ were confirmed by a Fourier transform infrared (FTIR) spectrometer (PerkinElmer, Shanghai, China). The three parts of EPS extractions for each types of sludge were mixed and freeze-dried for 48 into powder according to Mu et al. (2012). One milligram of freeze-dried EPS was mixed with 100 mg of potassium bromide (KBr, IR grade) and then ground and homogenized. After compressed to form a KBr window at 20,000 psi, the samples were scanned 32 times from 4000 to 800 cm−1 with a resolution of 4 cm−1 (Liu et al. 2018).

The balance of Zn in synthetic wastewater, EPS extractions (three parts), and cells were investigated under 50 mg L−1 of Zn2+ and 5 g SS L−1 of sludge. At the end of adsorption experiment, the sludge particle was separated from the liquid phase by vacuum filtration through a 0.45-μm cellulose acetate membrane, settled in the tube, and washed with PBS. After that, the extraction protocol was conducted to extract soluble EPS, LB-EPS, and TB-EPS. The EPS extractions were digested with mixed acid according to EPA method 3052 (Song et al. 2018); then, the Zn content was analyzed. The difference between the total amount and the sum of the measured amount in synthetic wastewater and EPS extractions was considered to be the immobilized Zn by inner cells.

Results and discussion

Adsorption performance by different sludges

Effect of contact time

Firstly, the effect of contact time on the Zn2+ adsorption was investigated at a sludge dosage of 5 g SS L−1 with an initial Zn2+ concentration of 50, 100, and 200 mg L−1. The relationship between the removal efficiency and adsorption time is shown in Fig. S2. When the initial Zn2+ concentration was 50 mg L−1, it was found that larger amounts of Zn2+ were rapidly removed in 10 min, and the plateau was reached within 30–60 min for all the sludges applied. Further increase in contact time to 480 min did not show significant increase in adsorption. Then, by increasing the initial Zn2+concentration to 100 and 200 mg L−1, a high removal of Zn2+ also can be observed at the beginning 10 min, and the time to equilibrium slightly shifted longer but still gradually attained in about 100 min. The initial fast metal uptake step is probably due to high initial Zn2+ concentration and empty metal binding sites on the sludge surface. While the slower subsequent phase is likely due to the saturation of metal binding sites. The data obtained from this experiment was further used to evaluate the adsorption kinetics. Figure S2 also shows that the bio-adsorption capacity for sludge is limited. Because, even at a relatively lower initial concentration (50 mg L−1of Zn2+), the equilibrium Zn2+ concentration among the sludge samples is still in a range from 10.8 to 20.2 mg L−1, which is far beyond the safe range for practical use of water.

Effect of initial Zn2+ concentration

The adsorption of zinc by the four sludges was investigated under various initial concentrations of Zn2+ (25 °C, 180 rpm, 5 g SS L−1, and pH = 6.0).

Results showed that for all the sludges, removal efficiency of Zn2+ decreased with the increase of initial Zn2+ concentration, whereas the adsorption capacity increased accordingly. At the low initial concentrations of Zn2+ (10 to 30 mg L−1), the removal efficiency was very high for all types of sludge (77.0–86.2%) and the adsorption capacity increased linearly with initial Zn2+ concentration (Fig. 1). This can be explained by the fact that the adsorption was not saturated at low initial concentrations. When the initial Zn2+ concentration increased continually, the removal efficiency decreased gradually and maintained at a relatively stable value. However, the adsorption capacity increased significantly with the increase of Zn2+ concentrations (Fig. 1). Under the tested concentrations (up to 200 mg L−1), the adsorption capacity increased in a rapid way without saturation, suggesting that all the sludges possessed a relatively high zinc adsorption capacity. Due to the practical factor that the Zn2+ concentration in activated sludge system will be not so high, so further increase of Zn2+ concentration was not applied in the present study.

Fig. 1
figure 1

Removal efficiency and adsorption capacity of Zn2+ using different types of sludge at initial concentrations of 10–200 mg L−1

Full size image

Among the four sludges, the autotrophic short-cut nitrification sludge showed the best adsorption performance, followed by autotrophic anammox granules (Fig. 1). It was surprising that the heterotrophic activated sludge and denitrifying sludge taken from real domestic wastewater treatment plant (WWTP) showed the weakest performance. Especially, the adsorption capacity of the short-cut nitrification sludge exhibited a sharper linear increase in the initial metal concentration range (10–200 mg L−1). The detected maximum adsorption capacity (22.3 mg g SS−1) seemed far lower than the saturation value, as shown in Fig. 1, which indicated that the short-cut nitrification sludge had a higher maximum adsorption capacity than the other three sludges.

Effect of sludge dose

The adsorption capacity was also significantly affected by the dosage of sludge. It can be seen from Fig. 2 that the removal efficiency of Zn2+ increased while the corresponding adsorption capacity decreased with the increase of sludge dosage for all types of sludge. The results demonstrated that the higher sludge dosage would induce the higher removal amount of metals from aqueous solutions. The highest adsorption capacity of the sludge was found at 1 g SS L−1. The reason for the higher metal uptake at lower sludge dose might be due to the higher metal to sludge ratio, which means more adsorption sites were available and metal could easily interact with these sites. From Fig. 2, it can also be found that (1) short-cut nitrification sludge still showed the highest adsorption capacity and metal removal efficiency among the four types of sludge applied, followed by the sequence of anammox granules, denitrifying sludge, and activated sludge, and (2) the increment of Zn2+ removal efficiency showed an obvious slowdown when the sludge dosage exceeded 3 g SS L−1, which might due to the overlapping or aggregation of adsorption sites at higher sludge dose. However, the removal efficiency still showed a rapid increase for the four sludges when sludge concentration was increased to 10 g SS L−1, suggesting that zinc removal could be further increased if the sludge dosage was further elevated.

Fig. 2
figure 2

Effect of sludge dose on the absorption (contact time = 480 min, initial metal concentration = 50 mg L−1, and pH = 6)

Full size image

Adsorption isotherms

The adsorption isotherms study was performed to describe the adsorption characteristics of Zn2+ by the four biological sludges at initial pH value of 6, temperature of 25 °C, and agitation speed of 180 rpm. When the initial Zn2+ concentration increased (10 to 200 mg L−1), the equilibrium pH values were relatively stable at around 5.4~5.8 due to the pH buffering capacity of the four sludges. The decreased final pH might attribute to the deprotonation of functional groups on the sludge surface, and part of the H+ ions were released to the solution which involved an ion exchange mechanism (Reddad et al. 2002). Two common adsorption isotherms, Langmuir and Freundlich models, were chosen in the present study. The Langmuir model is usually applied for a monolayer adsorption which assumes that adsorption occurs at specific homogeneous sites; it can estimate the maximum adsorption capacity corresponding to complete monolayer coverage on the sludge surface. The Freundlich model is an empirical equation which describes the heterogeneous surface energies by multilayer adsorption, and the adsorption intensity of the sludge adsorbent can be estimated.

Table 2 and Fig. 3 exhibit the fitting results of Langmuir and Freundlich models for Zn2+ adsorption. Both isotherm models consider qe as a function of Ce, corresponding to the equilibrium distribution of ions between solid and aqueous phases as the increases of initial metal ion concentration. Results showed that both models satisfactorily fitted the experimental data of Zn2+ adsorption with R2 ranging 0.95–0.98 for Langmuir models and 0.92–0.98 for Freundlich models.

Table 2 A comparison of the Freundlich and Langmuir adsorption isotherm constants
Full size table
Fig. 3
figure 3

Adsorption isotherms of Zn2+ onto the four sludges at 25 °C

Full size image

The Freundlich constant, n is considered to reflect the average binding energy of the adsorption sites, and n > 1 means the favorability of adsorption. In this study, n was around 2–3 for all the sludges, which locates in the range of 2–10 (n−1 around 0.1~0.5 means easy to adsorb). Therefore, the adsorption of Zn2+ on biomass sludge was favorable.

The Langmuir constant (qm) of Zn2+ adsorption to short-cut nitrification sludge was found to be nearly 3 times of the other three sludges, suggesting that the short-cut nitrification sludge has the highest metal uptake capacity under the given conditions. The RL values (Table 2) ranged 0.051~0.649 for the four sludges, which means the adsorption behavior of biological sludge is favorable (0 < RL < 1). This was in accordance with the Freundlich modeling results.

The above results indicated that both Langmuir and Freundlich isotherms were valid for Zn2+ adsorption by using four biological sludges. This means that both monolayer adsorption and heterogeneous surface conditions existed in the present study (Gulnaz et al. 2005). Because the applicability of the Langmuir model implied that the Zn2+ sorbed from a monolayer coverage on the sludge surface, while the well-fitted equilibrium data with the Freundlich model actually revealed the real heterogeneous nature of the surface sites involved in the metal uptake (Reddad et al. 2002). So, the adsorption properties of biological sludges are more likely to involve multiple mechanisms.

Adsorption kinetics

The adsorption kinetics of Zn2+ by the four types of sludge was conducted to examine the adsorption reaction mechanism by using three kinetic models: pseudo-first-order kinetic model, pseudo-second-order kinetic model, and intra-particle diffusion kinetic model, all of which were applied under the following conditions: three initial Zn2+ concentrations (50, 100, 200 mg L−1), constant sludge dose of 5 g SS L−1, 25 °C, agitation speed 180 rpm, and initial pH of 6.0. The modeling results are shown in Table 3.

Table 3 Adsorption kinetic models parameters for adsorption of Zn2+ onto sludges
Full size table

Pseudo-first-order kinetic model

As the most widely used kinetic model for bio-adsorption, the pseudo-first-order considers that the rate of occupation of adsorption sites is proportional to the number of unoccupied sites. The rate constant k1 and the calculated equilibrium capacities (qe cal) are calculated from the slope and intercept of the plot of ln(qe − qt) versus t, which should be a straight line to confirm the applicability of the first-order kinetic model. Results indicated that the correlation coefficients (R2) for all the sludge samples were low, and the qe cal values did not match the qe exp values (Table 3). This is because pseudo-first-order equation is not a true first-order equation (Gupta and Babu 2009). For the constant k1, it does not represent the total amount of available sites, and for the adjustable parameter ln(qe), it usually not equals to the intercept of the plot of ln(qe − qt) versus t. So, the bio-adsorption of Zn2+ cannot be predicted by pseudo-first-order kinetic in the present study.

Pseudo-second-order kinetic model

As a result of the inapplicability of pseudo-first-order kinetic model, the pseudo-second-order kinetic model was applied to describe the adsorption kinetics. It is assumed that chemisorption is the rate-controlling mechanism. The equilibrium capacities for the four sludges calculated by the pseudo-second-order kinetic model showed a good agreement with the experimental data (Table 3). Further, all the coefficients (R2) were greater than 0.998, which suggested that the pseudo-second-order kinetic model well described the adsorption of Zn2+ onto biomass sludge. Consequently, the overall rate of the Zn2+ bio-adsorption process was controlled by chemical process which involves valent forces through sharing or the exchange of electrons between the bio-adsorbent and metal ions (Gulnaz et al. 2005; Li et al. 2011). It has been reported that a lower k2 means a faster adsorption process. The values of k2 decreased with increase of the initial Zn2+ concentration (Table 3). It meant that the higher initial concentration of Zn2+, the faster adsorption rate. Besides, k2 values of the autotrophic short-cut nitrification sludge and anammox granules were much lower (Table 3), further verifying a more rapid adsorption process over the two heterotrophic sludges from domestic WWTP.

Intra-particle diffusion model

According to Wu et al. (2009), the form of intra-particle diffusion can be represented as multi-linearity in qt versus t1/2 plot. The whole process usually involved two or three steps. The first step is the external surface adsorption or instantaneous adsorption stage. The second step is the gradual adsorption stage where intra-particle diffusion controls the adsorption rate. The third step is the final equilibrium stage, where the solute moves slowly from larger pores to micro-pores and adsorption rate starts slow (Gulnaz et al. 2005; Ru et al. 2007). The adsorption process was simulated by intra-particle diffusion model, and the intra-particle diffusion rate constants (ki1, ki2, and ki3), qe exp, and qe cal are listed in Table 3. It is clear that Zn2+ adsorption occurred onto the external surface and the adsorption can be soon reached saturation (Fig. S3 B). Then, the Zn2+ entered into the pore of the sludge and adsorbed by the interior surface of sludge, which occurred at the second phase of the plot in Fig. S3 B. Due to the increased diffusion resistance, the diffusion rate (ki2) in this stage decreased significantly. With decrease of the Zn2+ concentration in the solution, ki2 became lower. Finally, the diffusion processes reached the final equilibrium stage. It should be pointed out again that the short-cut nitrification sludge had the highest diffusion rate (ki1) in the first stage over the other three sludges, which is in accordance with the results of k2 value in pseudo-second-order kinetic modeling. Gulnaz et al. (2005) reported that the diffusion rate would decrease with the increase of sludge particle size. However, the present result did not follow such hypothesis, as the ki1 of anammox granules was close to those of the flocculent activated sludge and denitrifying sludge.

The role of EPS for Zn2+ adsorption

As reported, EPS can significantly affect the bio-flocculation and bio-granulation of sludge systems, because the physicochemical properties and the special location of EPS make them an important part for maintaining microbial aggregate’s structure and function (Liu et al. 2010). EPS also play a crucial role in bio-adsorption of heavy metals due to the abundant binding sites and negatively charged organic functional groups on the surface and pores of EPS (Wei et al. 2017). Many researchers have indicated that EPS were in contact with cationic metal ions (Li et al. 2017; Wang et al. 2014; Yin et al. 2011; Zhang et al. 2014). As EPS were important constitutes of sludge flocs and granules, the compositions of EPS and their functional groups from different sludges before and after metal adsorption (50 mg L−1 of Zn2+ and 5 g SS L−1 of sludge) were subsequently analyzed.

Amount and composition changes after adsorption

The protein (PN) and polysaccharide (PS) are the major components of sludge EPS, and these two contents in the four sludge samples are shown in Fig. 4. Clearly, both PN and PS contents of all sludge samples decreased in certain degree after 480-min adsorption. The anammox granules contained the highest EPS content (110.33 mg g SS−1) with PN of 86.38 mg g SS−1 and PS of 23.95 mg g SS−1, respectively. After adsorption, the PN and PS decreased to 61.59 and 10.36 mg g SS−1, with 28.6% and 56.7% decrease in PN and PS, respectively. The decreased EPS content probably indicated the inhibition of biomass activity by the high Zn2+ concentration during the exposure or the hydrolysis of EPS. Madoni et al. (1999) reported that the activated sludge microbes were inhibited by zinc ions. In addition, some recent researches reported that zinc ions which released from its metal or metal oxide NPs could decrease the amount of EPS in biological sludge at a short term (Mu et al. 2012; Zheng et al. 2011).

Fig. 4
figure 4

Changes of biochemical compositions of EPS after Zn2+ adsorption: proteins (a), polysaccharides (b)

Full size image

The total EPS content of autotrophic anammox granules (110.33 mg g SS−1) and short-cut nitrification sludge (96.07 mg g SS−1) was much higher than that of both the heterotrophic activated sludge (33.12 mg g SS−1) and denitrifying sludge (34.25 mg g SS−1). The results could be attributed to the long-term optimization of culture under laboratory scale, which led to the two autotrophic sludges possessed a rich EPS content. Besides, some previous researches reported that the autotrophic nitrogen removal bacteria can secrete more EPS than heterotrophic ones (Hou et al. 2015; Tang et al. 2011). Adsorption capacity of Zn2+ by the short-cut nitrification sludge and anammox granules was higher than that of the two heterotrophic sludges. It seems that a rich total EPS content contributes much to the capacity of metal adsorption. But, anammox granules, which possessed the highest total EPS content, exhibited a lower adsorption capacity than short-cut nitrification sludge. It is suggested that the total EPS content did not show a strong positively correlation with metal adsorption. So, further comparison between the compositions of PN and PS in soluble/bound EPS was conducted.

Although the total EPS content was similar between short-cut nitrification sludge and anammox granules, Fig. 4 clearly shows that the PS content and the soluble EPS of short-cut nitrification sludge were much higher than those of anammox granules. First of all, the PN of short-cut nitrification sludge (44.33 mg g SS−1) was much lower than that of anammox granules, but the PS was much higher (55.74 mg g SS−1). As reported, the hydroxyl groups of hexose or pentose molecules in neutral PS and carboxyl groups in anionic PS exhibited a great ability to bind cationic metal ions via adsorption, ion exchange, or covalent binding (Alpat et al. 2008; Brown and Lester 1982; Rosales et al. 2015). Thus, the adsorption of metal ions was enhanced under higher PS contents. This might be one of the causes for the higher Zn2+ adsorption capacity by the short-cut nitrification sludge.

Interestingly, it was also found that the soluble EPS was the primary form of the bound EPS (LB-EPS and TB-EPS) of short-cut nitrification sludge, especially in PS, which was quite different from the other sludges including anammox granules. The soluble PN and PS in short-cut nitrification sludge were 12.6 and 35.8 mg g SS−1, respectively, while the contents in other sludges were all lower than 5 mg g SS−1. According to Comte et al. (2006), the soluble EPS has a greater capacity to exchange protonic ions than bound EPS, and the number of binding sites of soluble EPS for metal ions is larger than that of bound EPS. This could be another reason for the higher Zn2+ adsorption by the short-cut nitrification sludge, including the higher qe and an obviously higher qm in isotherm calculations.

FTIR changes after adsorption

FTIR spectra of original and metal-loaded EPS extracted from the four sludges were performed to investigate the functional groups that dominated in EPS after metal adsorption. The infrared spectra (4000–800 cm−1) of the extracted EPS from the four sludges before and after adsorption of Zn2+ are presented in Fig. 5. It was significant that the position and number of FTIR peaks for the EPS of four sludges were mostly the same except for some intensity difference, indicating they have similar chemical structure.

Fig. 5
figure 5

FTIR spectrum of the EPS extractions after Zn2+ adsorption: anammox granules (a), short-cut nitrifying sludge (b), activated sludge (c), and denitrifying sludge (d)

Full size image

For the four EPS sludge samples, the broad and strong band around 3403–3430 cm−1 was assigned to the OH stretching of carboxylic groups. The band around 2925–2932 cm−1 was attributed to the C–H stretching which was related to aliphatic chains of fats, lipids, and humic acid. The distinct bands near 1654 cm−1 and 1542 cm−1 presented the significant contribution of peptidic bonds in proteins. The strong peak at 1031–1105 cm−1 could be attributed to C–O–C and C–O, the stretching vibration for polysaccharides. Its intensity was relatively higher with respect to the 1654 cm−1 (proteins) band in short-cut nitrification sludge EPS compared to other sludge EPS. This consequence is in agreement with the results in the “Amount and composition changes after adsorption” section that the PS content of short-cut nitrification sludge was much higher than that of other types of sludge.

It has been reported that protein and polysaccharide in the EPS were able to react with metals (Mu et al. 2012; Wang et al. 2014; Wei et al. 2017). In the present study, the absorbance at 2925 cm−1, 1654 cm−1, 1542 cm−1, and 1031–1105 cm−1 decreased after adsorption of Zn2+, indicating that those C–H stretching, C=O-related carboxylic acids, and peptidic bond in proteins and polysaccharides indeed played an important role in zinc adsorption. Besides, compared to the spectra of original EPS of four sludges, the peak intensities for 1031–1105 cm−1 were significantly decreased by Zn2+-loaded EPS. This indicated polysaccharides of EPS had a higher binding capacity for Zn2+ than other groups. As evident in adsorption experiment, the short-cut nitrification sludge possesses a higher adsorption rate of Zn2+ than other sludges, which could be attributed to the abundant existence of polysaccharides in its EPS. Therefore, the FTIR analysis further confirmed that the adsorption of metal ions was enhanced under higher PS contents.

Metal distribution after adsorption

In order to further explore the contribution of EPS to the Zn2+ adsorption, the distribution of Zn in sludge was analyzed in details under 50 mg L−1 of Zn2+ and 5 g SS L−1 of sludge. The results are shown in Fig. 6. Generally, the Zn distribution could be divided into three parts, i.e., the water solution, EPS, and inner cells. Due to the different metal adsorption capacities, the residual Zn content in water solution differed in sludges. However, the Zn immobilization by cells did not vary much (Fig. 6). As shown in Fig. 6, about 50% of Zn was adsorbed by the sludge samples. The Zn distribution in EPS showed some new information. First of all, more Zn (17.8%) was trapped by soluble EPS in short-cut nitrification sludge which possessed a much higher content of soluble EPS than other sludges. Besides, it is clear that the content of soluble EPS was much lower than that of bound EPS in activated sludge, denitrifying sludge, and anammox sludge (as shown in Fig. 4). But Fig. 6 shows that the percentage of adsorbed Zn in soluble EPS was quite comparable to that in bound EPS. Further calculation indicated that the adsorbed Zn in soluble EPS was 127.6~211.5 mg g−1, while it was only 38.7~143.6 mg g−1 in bound EPS among the sludges. The results further confirmed the hypothesis aforementioned that soluble EPS contributed much to the cationic metal adsorption.

Fig. 6
figure 6

The distributions of Zn2+ among four types of sludge at the end of adsorption experiment (C0 = 50 mg L−1, dosage = 5 g SS L−1; temperature = 25 °C, agitation speed = 180 rpm, pH = 6)

Full size image

It was reported that EPS in sludge can protect the microbes against inhibitors including Zn2+ (Mu et al. 2012; Zhang et al. 2017). The present study confirmed that EPS can trap Zn2+ and prevent its diffusion into the microbial communities. Taking short-cut nitrification sludge as an example, nearly 30% of Zn was trapped by EPS. However, the protecting capability of EPS was actually very limited if the sludge concentration as well as EPS content was low. As calculated, 61.6–76.9% Zn escaped from the capture of EPS, which would pass through the EPS and be finally immobilized by bacterial cells. Consequently, a great toxic risk on microorganisms as well as the biological activated sludge system would be raised. So, it is necessary to remove metal ions if the activated sludge processes encounter high stress of heavy metals.

Conclusion

In this study, four different biological sludges, i.e., activated sludge, denitrification sludge, short-cut nitrification sludge, and anammox granules, were applied to evaluate the Zn2+ adsorption behavior and the variations of EPS characteristics at a neutral pH of 6 and room temperature (25 °C). We found that both Langmuir and Freundlich isotherms suitably described the adsorption process, which means both monolayer adsorption and heterogeneous surface conditions existed at the present experimental conditions. In addition, the short-cut nitrification sludge has the highest maximum adsorption capacities (36.4 mg g SS−1), which was much higher than other sludges (12.78–14.67 mg g SS−1). The adsorption kinetics study has highlighted a fast metal uptake by biological sludge, which might be attributed to the large availability of the functional groups in EPS. The FTIR spectroscopic analysis of metal-loaded EPS indicated the involvement of these functional groups in the Zn2+ adsorption process. Compared with the physicochemical properties of the four sludges, the soluble EPS, especially soluble PS, were found to play a more important role in binding cation metals, which was further confirmed by the analysis results of FTIR. It was also found that less than 30% of Zn2+ was trapped by EPS, while 61.6–71.9% could be harvested directly by cells, showing that the protecting capability from EPS trapping was limited.