Introduction

Water is a resource that mankind depends on for survival. However, the water resources that people can use are becoming increasingly scarce. According to a WHO report, 2 billion people lacked access to safe drinking water services in 2020 and it is estimated that by 2025 half of the world’s population will live in places where water resources are lacking [1]. Therefore, it is necessary to improve the efficiency, productivity, and sustainable use of water resources. Industrial production produces a large amount of wastewater and reasonable treatment of it is conducive to the recycling of water resources. One of the challenges is the purification of industrial wastewater by the removal of organic and/or inorganic pollutants. At present, commonly used wastewater treatment technologies mainly include physical (adsorption [2,3,4], filtration [5, 6]), chemical (oxidation and reduction [7, 8], coagulation [9, 10]), and biodegradation methods (anaerobic treatment [11, 12], aerobic bioremediation [13]). The physical methods can effectively remove heavy metals, but are not suitable for the treatment of small molecular organic pollutants. The biodegradation methods are low-cost, but their treatment efficiency for toxic and non-biodegradable wastewater, such as various heavy metal ions, is relatively low. The Fenton reaction has attracted widespread attention among chemical methods because the decomposition of H2O2 can generate free hydroxyl groups (·OH) with a strong oxidizing ability, which can effectively decompose non-biodegradable compounds with high toxicity into non-toxic products. The entire process is relatively simple. However, existing catalysts for the Fenton reaction (e.g., Fe3O4 nanoparticles) usually need to operate in an acidic environment (pH≈3). Thus, both the adjusting of the pH value and the iron sludge by-product will cause secondary pollution and incur additional treatment costs [14]. Nevertheless, the Fenton reaction is an essential method for treating organic pollutants. Therefore, it is a great challenge to find a catalyst that is efficient, low-cost, non-polluting, and easily accessible.

Ge et al. synthesized a novel organic–inorganic hybrid nanoflower with higher catalytic activity and stability than free enzymes and most immobilized enzymes [15]. This means it is possible that biological enzymes with high efficiency and specificity can be used for catalytic oxidation in harsh environments such as organic wastewater. In addition, the organic–inorganic hybrid nanoflower synthesis method is mild, simple, and environmentally friendly. To date, different hybrid nanoflowers have been prepared with various enzymes and metal ions for applications in industrial catalysis [16,17,18], biomedicine [19, 20], energy production [21, 22], and wastewater treatment [23, 24]. In view of the advantages of organic–inorganic hybrid nanoflowers, we believe that horseradish catalase-Cu3(PO4)2 hybrid nanoflowers will exhibit higher activity, as well as excellent stability for recycling and cost reduction.

To facilitate the reasonable design of hybrid nanoflowers, it is critical to determine the origin of their increased catalytic activity. Huang et al. synthesized glucose oxidase (GOx)-Cu3(PO4)2 hybrid nanoflowers with peroxidase-like activity without the addition of horseradish peroxidase (HRP), thus realizing the one-step detection of glucose [25]. They proposed that the peroxidase-like activity only resulted from the conversion between Cu(I) and Cu(II). More recently, it has been found that amino acids [26], catecholamines [27], plant extracts [28], gallic acid [29], and Cu(II) can also form hybrid nanoflowers with peroxidase-like properties. Although these findings indirectly support those of Huang et al., it is necessary to confirm that the transformation between Cu(I) and Cu(II) can catalyze the oxidation of the substrate by hydrogen peroxide. More importantly, we believe that numerous active sites with an upper energy density in the nanostructure also contribute to the decomposition of hydrogen peroxide, generating reactive oxygen species, such as hydroxyl radicals (·OH), which are useful for the Fenton reaction. Further evidence is required to support this hypothesis.

To clarify the origin of nanoflowers’ peroxidase activity, bovine serum albumin (BSA)-Cu3(PO4)2 hybrid nanoflowers (NFs) were synthesized by biomineralization using BSA as the organic component and Cu(II) as the inorganic component. Scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), Brunauer–Emmett–Teller (BET), and X-ray diffraction (XRD) were used to evaluate its morphology, structure, and composition. We further elucidated the peroxidase-like activity of the NFs by comparing the catalytic activities of BSA, NFs, and HRP. Based on this, kinetic experiments were performed to assess the magnitude of the peroxidase-like activity of the NFs, demonstrating their practicality. Finally, Congo red (CR) was selected to simulate organic pollutants to evaluate the decolorization efficiency of the NFs.

Materials and methods

Chemicals and materials

HRP, BSA, copper sulfate pentahydrate (CuSO4·5H2O), phosphate-buffered saline (PBS) (10 mM, pH 7.4), hydrogen peroxide, and 3,3′,5,5′-tetramethylbenzidine (TMB) were purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd. The Bradford protein assay kit was purchased from Shanghai Sangon Biotech. Co., Ltd. CR was obtained from Aladdin Chemical Reagent Co., Ltd. All chemicals were used as received without any further purification. Deionized water was prepared using a molecular system (Chongqing, China).

Synthesis of NFs and characterization

BSA (1 mg) was dissolved in 1 mL of phosphate buffer (10 mM) to form a 1 mg/mL protein solution. Subsequently, 600 μL of the 1 mg/mL protein solution was diluted in 5.4 mL of phosphate buffer to form a 0.02 mg/mL protein solution. Then, 40 μL CuSO4 aqueous solution (120 mM) was added to 6 mL of the 0.02 mg/mL protein solution and stirred on a vortex for 30 s (1500 rpm), followed by reaction at 25 °C for 72 h. The supernatant of the solution was then collected to measure the encapsulation rate and the blue precipitate at the bottom was collected through a centrifuge (7000 rpm for 15 min). The precipitate was washed three times with deionized water and freeze-dried (the cold trap temperature was set to  −60 °C).

The surface morphology and structure of the NFs were characterized using SEM (Hitachi S-4800) and TEM (Talos F200S). The elemental distribution was obtained by energy-dispersive spectrometry (EDS) mapping (FEI Quanta 250 FEG), and XRD patterns were recorded using a Bruker D8 Advance X-ray diffractometer.

Entrapment efficiency of NFs

The encapsulation efficiency of an NF is equal to the ratio between the weight of the protein in the NF and the weight of the total protein. To plot the protein standard curve of BSA, 5 mg of BSA was added to 1 mL PBS to obtain a 5 mg mL−1 standard protein solution. Protein solutions of 10, 20, 40, 60, 80, 100, and 150 μg mL−1 were then prepared from the standard protein solution. Coomassie brilliant blue-G250 (0.25 mL) was added to detect the absorbance of different concentrations of protein at a wavelength of 595 nm on a microplate reader, and a standard curve was drawn. After 72 h of reaction, the supernatant in the solution was collected and its protein concentration was measured using a Bradford protein assay. Coomassie brilliant blue-G250 was added to the sample containing the supernatant and the absorbance of the sample was measured at a wavelength of 595 nm using a microplate reader. The sample concentration was calculated using the standard curve (see Electronic Supplementary Material Fig. S1). The equation for the encapsulation rate is as follows:

$$E=\frac{{G}_{T}-{G}_{S}}{{G}_{T}}\times 100\%$$
(1)

where E is the encapsulation efficiency, GT is the total protein weight, and GS is the protein weight of the supernatant.

Catalytic activity of NFs

To evaluate whether the NFs have peroxidase-like catalytic activity, TMB was used because it is gradually oxidized by hydrogen peroxide from colorless to blue. The blue product exhibits an absorption peak at 370 nm, which can be measured by ultraviolet spectrophotometry. In the reaction, 5 mM H2O2, 20 μg/mL NFs, 5 mM H2O2 (contain 20 μg/mL NFs), 5 mM H2O2 (contain 20 ng/mL HRP), and H2O (blank control) were each added to 0.8 mM TMB to form 3 mL solutions.

The catalytic mechanism of the NFs was studied using 0.8 mM TMB and 5 mM H2O2 as model substrates. In the reaction, BSA, Cu3(PO4)2, NFs, and H2O (blank control) were each added to 0.8 mM TMB and 5 mM H2O2 to form 3 mL solutions. The absorbance of the product (oxidation of TMB) was monitored at 370 nm using an ultraviolet–visible spectrophotometer.

Kinetic parameter

To further study the catalytic potential of the NFs, their kinetic parameters were calculated. The kinetic constants of the NFs were determined by adjusting the concentration of the substrate H2O2 in the range of 0.05–1 mM. The concentration of TMB and the NFs was 0.8 mM and 0.02 mg/mL, respectively. Similarly, the steady-state kinetics of NFs were measured when TMB was used as a substrate. The concentration of H2O2 and the NFs was 1 mM and 0.02 mg/mL, respectively. Km and Vmax were determined using the following equation:

$$\frac{1}{{V}_{0}}=\frac{{K}_{\mathrm{m}}}{{V}_{\mathrm{max}}}\times \frac{1}{\left[S\right]}+\frac{1}{{V}_{\mathrm{max}}}$$
(2)

where V0 (mM min−1) is the initial catalytic rate, Vmax (mM min−1) is the maximum reaction velocity, [S] (mM) is the initial substrate concentration, and Km (mM) is the Michaelis–Menten constant.

Decolorization of Congo red

The amount of printing and dyeing wastewater shows a growing trend and it is becoming increasingly difficult to treat. CR is widely used in the printing and dyeing industries and easily enters the water system, and causes significant ecological damage. It is one of the most common organic pollutants in dye wastewater [30]. Therefore, it was selected as the target pollutant in this experiment. The decolorization of CR by NFs was performed at room temperature. NFs (60 μg) and 150 μg CR were added to 3 mL of the reaction solution containing 100 mM H2O2. The absorbance of the mixture was measured at 490 nm after incubation at room temperature for 3 h. The decolorization efficiency (\(D\)) is defined as:

$$D=\frac{{C}_{0}-{C}_{\mathrm{T}}}{{C}_{0}}\times 100\%$$
(3)

where \({C}_{0}\) and \({C}_{\mathrm{T}}\) are the initial and final absorbance of the CR solution at time t, respectively.

Results and discussion

Characterization of NFs

In this study, we constructed a protein-inorganic hybrid NF by a 3-day reaction between BSA and copper phosphate. Figure S2 (see Electronic Supplementary Material) shows the blue powder of the NFs. In the SEM image of the NFs (Fig. 1), we can see that they have a uniform hierarchical and flower-like structure. This porous structure is confirmed by the TEM images of the NFs, shown in Fig. 2A, B. This is of great value for catalytic engineering and drug delivery systems. The optical microscopy (Fig. 1A–C) and SEM (Fig. 1D–F) results indicate the obvious influence of the protein concentration on the morphology of the NFs. When the concentration of BSA is 0.5 mg/mL, the individual NFs appear as compact spheres composed of hundreds of nanosheets (Fig. 1D). The diameter of the NFs is approximately 5 μm and the pores are difficult to distinguish. In contrast, when the concentration of BSA is 0.02 mg/mL, the hierarchical structure of the NFs is very scattered and the whole NF appears as a blooming peony with a diameter of 20 μm. The size of the NF pores is inhomogeneous, with the largest size being approximately 2.5 μm (Fig. 1F). However, when the concentration of BSA is 0.1 mg/mL, the nanosheets on the surface of the NFs become thicker crystals, similar to a tangle of twine, which is significantly different from their shape when the concentration of BSA is 0.5 or 0.02 mg/mL. Accordingly, the pore sizes for 0.1 mg/mL BSA are between those for 0.5 and 0.02 mg/mL BSA (Fig. 1E). It is worth noting that Koley et al. [31] and Li et al. [32] reported similar results. We believe that the most critical factor affecting the morphology of the NFs is the ratio of copper ions to amide groups on the protein skeleton. When the ratio of protein to Cu2+ is relatively large, the amount of Cu2+ is not sufficient to cover all binding sites on the surface of the protein. In this case, the complexes of protein and Cu2+ can only induce a thin layer of Cu3(PO4)2 crystals on their surface, resulting in compact spheres. When the ratio of protein to Cu2+ is relatively small, the amount of Cu2+ is sufficient to form a scattered hierarchical structure after coordinating with all of the binding sites. A similar mechanism was proposed by Wei et al. [33]; however, more experimental proof is required.

Fig. 1
figure 1

Hybrid nanoflowers made from BSA and Cu3(PO4)2. Optical microscope image of the nanoflowers achieved with different concentrations of BSA, 0.5 mg/mL (A), 0.1 mg/mL (B), 0.02 mg/mL (C); SEM image of the nanoflowers achieved with different concentrations of BSA, 0.5 mg/mL (D, inset: enlarged view of D), 0.1 mg/mL (E, inset: enlarged view of E), 0.02 mg/mL (F, inset: enlarged view of F). All copper ion concentrations are 120 mM

Full size image
Fig. 2
figure 2

TEM images (A), high-resolution TEM image (B), and XRD patterns (C) of NFs

Full size image

As described above, the dispersed hierarchical structure facilitates mass transfer and provides more active sites. Hence, we conclude that the NFs prepared from 0.02 mg/mL BSA have better enzymatic activity and were selected for the following study. XRD and SEM-energy-dispersive X-ray (EDX) analyses were performed to investigate the NF components. From the XRD analysis of as-prepared NFs, characteristic diffraction peaks were observed at 13, 18, 21, 30, 34, 54.62, and 68° (Fig. 2C). These correspond to the peaks in the JPSCD standard (00–022-0548). As shown in Fig. 3, the SEM–EDX analysis suggests that the NFs contain carbon, nitrogen, oxygen, sulfur, phosphorus, and copper. These elements correspond to the basic components of proteins and copper phosphate crystals. It is noted that the carbon, nitrogen, and sulfur quantities were relatively low. This might be due to two reasons: (i) the protein is located at the center of the NFs and wrapped by the Cu3(PO4)2 crystal shell, and (ii) the protein intrinsically lacks amino acids containing sulfur, that is, methionine and cysteine. The central location of the protein components has already been verified by Ge’s experiments [14].

Fig. 3
figure 3

A SEM images of NFs. The EDS element mapping of B C, C O, D P, E N, and F Cu with scale bar of 5 µm. G EDX pattern of the NFs

Full size image

X-ray photoelectron spectroscopy was used to study the surface chemical elemental composition of NFs and the chemical valence states of copper. As shown in Fig. 4A, the survey spectrum shows the presence of Cu, O, N, P, Cl, Na, and C elements in the NFs, where the chemical elements of Na and Cl were the essential components of phosphate-buffered solution and Cu, O, N, P, and C can correspond to elements in BSA and Cu3(PO4)2, further proving the successful synthesis of the materials. The chemical valence states of Cu were studied by the spectroscopy of Cu 2p. As shown in Fig. 4E, the peaks at 935.28 eV and 955.48 eV correspond to the Cu 2p3/2 and Cu 2p1/2 electrons of Cu2+, respectively [34]. The peaks at 934.58 eV and 954.08 eV correspond to the Cu 2p3/2 and Cu 2p1/2 electrons of Cu+, respectively. The peaks at 939.98 eV, 942.08 eV, and 944.08 eV are satellite peaks of Cu2+. Meanwhile, the content ratio of Cu+ to Cu2+ in NFs is 1:1.9, which is calculated according to peak area at 935.28 eV and 934.58 eV. On the basis of these results, we can speculate that the copper ions in the synthesized NFs indeed exhibit two valence states (i.e., Cu+ and Cu2+). And the conversion between them induces NFs’ peroxide-like properties.

Fig. 4
figure 4

XPS spectra and N2 adsorption–desorption curve of NFs: A survey spectra, B carbon, C phosphorus, D oxygen, E copper, and F N2 adsorption–desorption curve and the pore size (inset part), where the concentrations of BSA and CuSO4 are 0.1 mg/mL and 120 mM in the preparation processes, respectively

Full size image

The BET test was also conducted to explore the porosity of NFs. The results show that the surface area, pore volume, and average pore diameter are 59.65 m2 g−1, 0.44 cm3 g−1, and 29.53 nm respectively, as shown in Fig. 4F and Table 1. The high surface-to-volume area and the porous structure indicated that the prepared NFs can alleviate mass-transfer limitations and provide more active sites.

Table 1 BET surface area and pore size of NFs
Full size table

A high entrapment efficiency implies that more proteins are loaded in the NFs, which contributes to a higher catalytic efficiency and lowers the cost. Protein encapsulation efficiency was calculated using Eq. 1. During the preparation of the NFs, the BSA concentrations were set to 0.02, 0.1, and 0.5 mg/mL, with the entrapment efficiencies of BSA in the NFs being 91.23 ± 2.75%, 78.94 ± 2.11%, and 23.97 ± 0.52%, respectively (Table 2).

Table 2 Entrapment efficiency for different BSA concentrations
Full size table

Catalytic activity of NFs

Figure 5A shows that different experimental groups ((1) 0.8 mM TMB; (2) 20 μg/mL NFs and 0.8 mM TMB; (3) 0.8 mM TMB and 5 mM H2O2; (4) 5 mM H2O2, 20 μg/mL NFs, and 0.8 mM TMB; (5) 20 ng/mL HRP and 0.8 mM TMB) give different degrees of peroxidase-like activity. No significant change is observed in the color of the mixed solutions of groups (1) and (2) (see the inset of Fig. 5A). Meanwhile, the absorbance at 370 nm in group (3) increases slightly. This indicates the slow oxidation of H2O2 to TMB without any catalyst. In contrast, the reaction rate of group (4) is greatly improved, while the rate of group (5) is only 51.75%. By comparing groups (3) and (4), we confirmed that the BSA-Cu3(PO4)2 NFs possess peroxidase-like properties. However, their catalytic activity is lower than that of free enzyme. Thus, proteins with a higher catalytic activity should be selected as the organic component of the NFs in practical applications, such as HRP. Huang [24] proposed that the peroxidase-like property originates from the transformation between Cu(II) and Cu(I), similar to the transformation between Fe(III) and Fe(II) in the Fenton reaction. Specifically, Cu(II) reacts with H2O2 to form Cu(I) and ·OOH, and the generated Cu(I) then reacts with H2O2 to produce Cu(II) and hydroxyl radicals (·OH) with a strong oxidation activity. We believe that the micro-/nanostructure on the surface of the NFs can also provide extra active sites, exhibiting a synergistic effect with the transformation between Cu(II) and Cu(I). Thus, they both contribute to the high peroxidase-like properties of the NFs.

Fig. 5
figure 5

A Absorbance curves of TMB reaction solutions, (1) H2O and 0.8 mM TMB, (2) NFs and 0.8 mM TMB, (3) H2O2 and 0.8 mM TMB, (4) H2O2, NFs and 0.8 mM TMB, (5) H2O2, HRP and 0.8 mM TMB; B the comparison of catalytic activity, (1) H2O, 0.8 mM TMB and 5 mM H2O2, (2) BSA, 0.8 mM TMB and 5 mM H2O2, (3) Cu3(PO4)2, 0.8 mM TMB and 5 mM H2O2, (4) NFs, 0.8 mM TMB and 5 mM H2O2; the inset is the photograph of corresponding solutions

Full size image

To further explore the origin of the NFs’ peroxidase-like property, four experiments were conducted ((1) H2O + 0.8 mM TMB + 5 mM H2O2; (2) BSA + 0.8 mM TMB + 5 mM H2O2; (3) Cu3(PO4)2 + 0.8 mM TMB + 5 mM H2O2; (4) BSA-NFs + 0.8 mM TMB + 5 mM H2O2). As shown in Fig. 5B, the absorbance at 370 nm of group (1) shows a small increase, primarily owing to the reaction between TMB and a few reactive oxygen species (ROS) from the slow decomposition of H2O2. Interestingly, the absorbance at 370 nm in group (2) decreases slightly compared with that of group (1). We concluded that the competitive combination of electrophilic carbon atoms in BSA to ROS inhibits the oxidation of H2O2 to TMB. It is worth noting that the absorbance at 370 nm of group (3) is much higher than that of group (2), which implies that the irregular Cu2(PO4)3 crystal contributes to the peroxidase-like property. Comparing the absorbance at 370 nm for groups (3) and (4), we found that for equal amounts of Cu2(PO4)3 (subtracting the amount of protein that is only related to the nucleation of NFs [35]), the absorbance of group (3) was relatively low, only 65.42% of that of group (4). This difference demonstrates another origin of the catalytic activity: the synergistic effects of micro-/nanostructures due to the high specific surface area and porosity.

Kinetic constants

The peroxidase activity of the NFs was further investigated by determining the apparent steady-state kinetic parameters of the reaction. Within the suitable range of H2O2 and TMB concentrations, typical Michaelis–Menten curves were obtained for the NFs (Fig. 6). As shown in Table 3, the maximum initial velocity (Vmax) and Michaelis–Menten constant (Km) were calculated using Lineweaver–Burk plots [36]. The Km value indicates the affinity between the enzyme and substrate, with a small Km value indicating a stronger affinity between the enzyme and the substrate. The Km of the NFs with H2O2 as the substrate is much lower than that of Fe3O4 nanoparticles, which was similar to HRP, indicating that the NFs have a higher affinity for H2O2 than Fe3O4 nanoparticles. Thus, this study shows that NFs with intrinsic peroxidase-like active enzymes provide a potential method for designing biomimetic catalysts with good stability and low cost.

Fig. 6
figure 6

Steady-state kinetic study of NFs with A varying H2O2 concentration and 0.8 mM TMB, and B varying TMB concentration and 1 mM H2O2; Lineweaver–Burk plot for NFs with C varying H2O2 concentration, and D varying TMB concentration. Error bars represent standard errors obtained from triplicate experiments. The velocity of the reaction (ν) was measured using 20 μg/mL in 200 μL of 10 mM buffer (pH 4.0)

Full size image
Table 3 The Michaelis–Menten constant (Km) and maximum reaction rate (Vmax) of as-prepared NFs, HRP, and Fe3O4 nanoparticles
Full size table

Decolorization of Congo red solution

We also investigated the potential of NFs in the field of dye decolorization. The NFs and CR were used as model catalysts and organic dyes, respectively. As shown in Fig. 7, neither the NFs alone nor H2O2 can decolorize CR. In previous studies, the decolorization and degradation of azo dyes were mostly achieved using fungi, free enzymes, and immobilized enzymes, with the decolorization efficiency of laccase for Congo red reaching 58% (see Table 4), which is not ideal [38]. As shown in Fig. 7A, with increasing time, the absorbance at 490 nm of the solution with NFs and H2O2 decreases significantly. The solution in the tube gradually becomes colorless, indicating that the NFs can decolorize CR. Using Eq. 2, the calculated decolorization rate of the CR solution by the NFs reached 96% after 3 h, which is much higher than the decolorization efficiency of free laccase.

Fig. 7
figure 7

A CR decolorization over time with different solutions, (1) H2O and CR, (2) H2O2 and CR, (3) NFs and CR, (4) NFs, H2O2, and CR. Inset was the typical photograph of corresponding solutions. B The absorption spectra of the CR solution after incubated with different components for 5 h, (1) H2O and CR, (2) H2O2 and CR, (3) NFs and CR, (4) NFs, H2O2, and CR

Full size image
Table 4 The decolorization efficiency for CR with different materials
Full size table

Stability of the NFs

The storability and reusability of NFs are crucial factors that affect their large-scale industrial production and application. The storability of the NFs was thoroughly investigated. They were stored for 1 month at room temperature, and their decolorization efficiencies were recorded and calculated every 24 h. As shown in Fig. 8A, the decolorization efficiencies of the NFs remain at 90% after 1 month, indicating that the NFs have good storability. In addition, the reusability of the NFs was determined. After every reaction, the NFs were recovered by centrifugation and washed with deionized water. As shown in Fig. 8B, the NFs have excellent reusability, and after five repeated reactions, they still exhibit 57% catalytic activity compared with the original NFs. The fact that the NFs lost a portion at each centrifugal collection may explain why the decolorization efficiency declined after the fifth cycle.

Fig. 8
figure 8

A Storage stability and B reusability of NFs, decolorization efficiency of NFs for CR at different pH C and temperature D. The concentration of CR and NFs was 50 mg/L and 20 μg/mL, respectively

Full size image

The chemical/physical property of printing and dyeing wastewater is complex, so it is necessary to consider the effect of pH (3~10, 40 °C) and temperature (25~70 °C, pH 7) changes on the decolorization efficiency of the NFs. As shown in Fig. 8C, the decolorization efficiency (which is consistent with absorption capacity) of NFs increases first and then decreases with the lowering of pH value. Interestingly, the NFs showed a higher decolorization efficiency in somewhat basic conditions, while that is lower in acidic solutions. When the pH value is 3, the decolorization efficiency is about 16%. This is significantly different from most nano-enzymes preferring to acidic conditions such as goethite (pH 3.5), magnetite (pH 3.0), suggesting a scarce supplement for applications in basic cases. Additionally, the pollutants in organic wastewater usually are N-containing industrial dyes, resulting in an alkaline condition [39], which indicates a broad prospect. We, in theory, believe that it is closely related to obviously different solubleness in basic or acidic solutions of Cu3(PO4)2 crystals serving as protective shell. On the other hand, the NFs has excellent decolorization efficiency over a wide temperature range (Fig. 8D). With the increase of temperature before 40 °C, the decolorization efficiency of NFs slightly increases, and the maximum of 96.05% is achieved at 40 °C. Then, the decolorization efficiency gradually decreases with the increase of temperature. When the temperature reaches 70 °C, the decolorization efficiency of NFs is still maintained at ~ 95%, demonstrating that the NF has excellent thermal stability.

The performance of NFs in decolorization of dyes

There are a wide variety of dyes in industrial wastewater. Here, we present the decolorization efficiency of NFs to eight common mimics of organic pollutants in dyeing wastewater to elaborate the broad-spectrum decolorization ability (or non-specificity) of NFs. As shown in Fig. S3 (see Electronic Supplementary Material) and Table 5, we found that NFs could decolorize 8 kinds of mimics with more than 90% decolorization efficiency. It demonstrates that the NF prepared is a kind of broad-spectrum decolorization material for most organic pollutant in wastewater because of its broad-spectrum oxidation from ROS and porous adsorption. This broad spectrum implies a wide range of applications.

Table 5 The decolorization efficiency of NFs as a catalyst for different dyes
Full size table

Actual sample analysis

In order to further investigate the practicability of NFs for treating of wastewater, it is necessary to conduct actual sample analysis. Owing to that practical samples of seriously polluted wastewater are difficult to access, we conducted decolorization experiments of CR in the water collected from a random drainage ditch. First, different concentrations of CR (50, 100, 150 mg/L) were respectively added into same amount of practical water in different beakers. Five-minute stirring was required for ensuring sufficient dissolving. Subsequently, same NF powder was added into each beaker. Decolorization occurred for 3 h. The absorbance of the water from drainage ditch, the absorbance of solution added with CR, and that of solution after decolorizing were record. As shown in Table 6, the decolorization efficiency of CR for three cases was 90.90% ± 3.04, 90.55% ± 1.64, and 90.65% ± 0.49 respectively, which suggests a good application prospect.

Table 6 The decolorization efficiency of NFs as a catalyst for different concentrations of CR in real waste water
Full size table

Conclusions

In summary, by using Cu3(PO4)2 as the inorganic component and BSA as the organic component, we synthesized BSA-Cu3(PO4)2 hybrid nanoflower with porous hierarchical structures. Experiments show that the concentration of protein has a significant effect on the morphologies of the NFs. The higher the protein concentration is between 0.02 and approximately 0.5 mg/mL, the more compact the surface morphologies of the NFs are. We believe that the molar ratio of Cu2+ to BSA is a key factor that affects the morphology. More importantly, our results demonstrate that the NFs possess peroxidase-like activity. The reaction catalyzed by irregular Cu3(PO4)2 crystals was faster than that without any catalyst, indicating that the irregular Cu3(PO4)2 crystals also have peroxidase-like activity. However, the intensity of this catalytic activity is only 65% of NFs. These results indicate that two factors are responsible for the peroxidase-like properties of the NFs: (i) the transformation between Cu2+ and Cu+ promotes the formation of hydroxyl radicals, and (ii) the micro-/nanostructures of the NFs provide larger specific surface areas, resulting in more active sites. Additionally, the decolorization rate of CR under the catalysis of BSA-Cu3(PO4)2 NFs is as high as to 96% after 3-h reactions. The high efficiency, low cost, broad spectrum, and recyclability of NFs suggest a huge potential for wastewater treatment.