Reusable Surface-Modified Bacterial Cellulose Based on Atom Transfer Radical Polymerization Technology with Excellent Catalytic Properties

The high catalytic activity of membrane-binding gold nanoparticles (AuNPs) makes its application in oxidation or reduction an attractive challenge. Herein, surface-functionalized bacterial cellulose (BC-poly(HEMA)) was successfully prepared with 2-hydroxyethyl methacrylate (HEMA) as monomers via the atom transfer radical polymerization (ATRP) method. BC-poly(HEMA) was further utilized as not only reducing agent but also carrier for uniform distribution of the AuNPs in the diameter of about 8 nm on the membrane surface during the synthesis stage. The synthesized AuNPs/BC-poly(HEMA) exhibited excellent catalytic activity and reusability for reducing 4-nitrophenol (4-NP) from NaBH4. The results proved that the catalytic performance of AuNPs/BC-poly(HEMA) was affected by the surrounding temperature and pH, and AuNPs/BC-poly(HEMA) maintained the extremely high catalytic activity of AuNPs/BC-poly(HEMA) even after 10 reuses. In addition, no 4-NP was detected in the degradation solution after being stored for 45 days. The reusable catalyst prepared by this work shows a potential industrial application prospect.


Introduction
Gold nanoparticles (AuNPs) are one of the most promising NPs compared with various metal and metal oxide NPs [1], owing to their excellent catalytic performance in a variety of chemical reactions including reductive catalysis of chlorinated [2], organic synthesis [3] and low-temperature CO oxidation [4]. However, AuNPs are prone to aggregation in reaction systems due to high surface energy, and then the catalytic activity is reduced [5]. In addition, AuNPs recycling is also a serious challenge [6]. Bioinorganic hybrid nanostructures can effectively combine the optical, electronic and mechanical properties of inorganic nanomaterial with good biocompatibility and low cost of natural biological materials, which have potential applications in optics [7], electronics [8], magnetism [9], catalysts [10], and battery materials [11]. Zawada et al. have reported a facile synthetic method of preparation of core-shell nanostructures using easy to obtain TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)-coated gold nanoparticles as precursor [12]. Therefore, polymers can be used as the carriers for AuNPs due to its abundant active sites, thus obtaining reusable and stable AuNPs [13].
Recently, biocompatible degradable polymers have been used as AuNP carriers [14]. Zhang et al. reported a novel Au/Bi 12 O 17 C l2 composite for methyl orange and phenol degradation [15]. Nguyen et 2.2.1. Initiation BC membrane was immersed into THF (150 mL) for 20 min. Then the TEA (10 mM, 210 µL) and 2-BIB (10 mM, 189 µL) were added into THF for 3.5 h at 35 • C. The prepared BC-Br initiator was extracted and stored in THF.

Surface Grafting
Prior to this reaction, HEMA (24 mL) and a mixture of HMTETA (400 µL) and DMF (24 mL) were deoxygenated through three freeze-pump-thaw cycles by using liquid nitrogen, and CuCl (100 mg) was then added to the mixed system. After the HMTETA/DMF mixture is magnetically stirred in a vacuum glove box for 2 h, BC-Br and HEMA are placed into the mixture system. The reaction system was stored in the glove box at 25 • C, and the ATRP reaction of HEMA was completed in 6 h. Finally, BC modified by HEMA was washed by ethanol and dried in air. According to the empirical results, the graft rate of HEMA monomer onto the BC-poly(HEMA) membrane was 21.7 wt%. The detailed procedure for making BC-poly(HEMA) is schematically depicted in Figure 1. BC was synthesized similar with a previous report [35].

Initiation
BC membrane was immersed into THF (150 mL) for 20 min. Then the TEA (10 mM, 210 µ L) and 2-BIB (10 mM, 189 µ L) were added into THF for 3.5 h at 35 °C. The prepared BC-Br initiator was extracted and stored in THF.

Surface Grafting
Prior to this reaction, HEMA (24 mL) and a mixture of HMTETA (400 µ L) and DMF (24 mL) were deoxygenated through three freeze-pump-thaw cycles by using liquid nitrogen, and CuCl (100 mg) was then added to the mixed system. After the HMTETA/DMF mixture is magnetically stirred in a vacuum glove box for 2 h, BC-Br and HEMA are placed into the mixture system. The reaction system was stored in the glove box at 25 °C, and the ATRP reaction of HEMA was completed in 6 h. Finally, BC modified by HEMA was washed by ethanol and dried in air. According to the empirical results, the graft rate of HEMA monomer onto the BC-poly(HEMA) membrane was 21.7 wt%. The detailed procedure for making BC-poly(HEMA) is schematically depicted in Figure 1.

Synthesis of Gold Nanoparticles (AuNPs)/BC-Poly(HEMA)
HAuCl4 (1.0 mM, 20 mL) aqueous solution in a 50 mL Erlenmeyer flask was heated to 110 °C. Then BC-poly(HEMA) was added to the boiling solution, and the reaction mixture gradually changed from light yellow to pink purple. Finally the membrane was taken out from the heated aqueous solution after 10 min, and afterwards washing it with DI water.

Catalytic Reduction of 4-Nitrophenol (4-NP)
The catalytic activity of AuNPs/BC-poly(HEMA) was examined by a model reaction of reducing 4-nitrophenol (4-NP) from NaBH4. Typically, 50 mL mixed aqueous solution of 4-NP (0.5 mM) and NaBH4 (0.25 M) were added into a reagent bottle, then 5 mg AuNPs/BC-poly(HEMA) was placed into the well-mixing solution. As the reaction proceeded, the color of the mixed solution gradually changed from yellow to colorless and the reduction process was monitored by ultraviolet (UV)visible spectrophotometer (U4100, Hitachi, Tokyo, Japan). The reduction mechanism of 4-NP by NaBH4 occurred on the surface of AuNPs/BC-poly(HEMA) was shown in Figure 2.

Synthesis of Gold Nanoparticles (AuNPs)/BC-Poly(HEMA)
HAuCl 4 (1.0 mM, 20 mL) aqueous solution in a 50 mL Erlenmeyer flask was heated to 110 • C. Then BC-poly(HEMA) was added to the boiling solution, and the reaction mixture gradually changed from light yellow to pink purple. Finally the membrane was taken out from the heated aqueous solution after 10 min, and afterwards washing it with DI water.

Catalytic Reduction of 4-Nitrophenol (4-NP)
The catalytic activity of AuNPs/BC-poly(HEMA) was examined by a model reaction of reducing 4-nitrophenol (4-NP) from NaBH 4 . Typically, 50 mL mixed aqueous solution of 4-NP (0.5 mM) and NaBH 4 (0.25 M) were added into a reagent bottle, then 5 mg AuNPs/BC-poly(HEMA) was placed into the well-mixing solution. As the reaction proceeded, the color of the mixed solution gradually changed from yellow to colorless and the reduction process was monitored by ultraviolet (UV)-visible spectrophotometer (U4100, Hitachi, Tokyo, Japan). The reduction mechanism of 4-NP by NaBH 4 occurred on the surface of AuNPs/BC-poly(HEMA) was shown in Figure 2

Materials Characterization
Surface morphologies of the BC, BC-poly(HEMA) were examined by a field emission scanning electron microscope (FE-SEM, Hitachi S4800, Tokyo, Japan). A transmission electron microscope (TEM, JEOL/JEM-2100, Hitachi, Tokyo, Japan) was used to analyze the structure and morphology of AuNPs, AuNPs/BC and AuNPs/BC-poly(HEMA) at the acceleration voltage of 100 kV. Fourier transform infrared spectroscopy (FT-IR, Nicolet Nexus, Thermo Electron Corporation, Waltham, MA, USA) was used to test the chemical structure of BC, BC-Br, BC-poly(HEMA). Using TGA (TGA-Q500, TA Instruments Corporation, New Castle, DE, USA) the thermal stability of samples were determined with the heating rate at 10 °C/min within the scope 30 to 800 °C under N2. The crystal structure of BC, AuNPs/BC-poly(HEMA) were measured by Powder D8 Advance X-ray diffraction (XRD, Bruker AXS D8, Leipzig, Germany). X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, Thermo Scientific Escalab, Thermo Fisher Scientific, Waltham, MA, USA) was also used to perform the elemental analysis of BC-Br and AuNPs/BC-poly(HEMA). The optical spectra and catalytic activity of the AuNPs/BC-poly(HEMA) was examined towards a Hitachi U4100 UV-visible spectrophotometer.

Morphology Analysis
The morphologies are shown in Figure 3 as carried out by FE-SEM and TEM. Figure 3a shows that the micro-pore structure of pure BC membrane with ultrafine nanofiber structure and nonwoven mechanically robust network, which provided high specific surface area and porosity. In addition, the pure BC presents a typical 3D network structure. The size distribution range of BC (inset image of Figure 3a) was 20 to 70 nm, with an average diameter of about 45 nm. After surface modification by ATRP, BC-poly(HEMA) (Figure 3b) showed no obvious damage, and still maintained similar morphology to that of BC. This is due to the fact that the surface modification of ATRP caused little damage to the BC membrane under non-violent conditions; in addition, BC can withstand a certain impact because of its excellent mechanical properties. In the one-step method synthesis of AuNPs, the characteristic peaks were observed at 524 nm ( Figure S1), which is consistent with the literature. Figure 3c depicts the successful formation of AuNPs/BC-poly(HEMA) with the average dimension of about 8 nm for AuNP, indicating that the BC-poly(HEMA) can act as the reducing agent in the preparation of gold nanoparticles. Additionally, the TEM images of pure BC and AuNPs/BC were depicted in Figure S2 and Figure

Materials Characterization
Surface morphologies of the BC, BC-poly(HEMA) were examined by a field emission scanning electron microscope (FE-SEM, Hitachi S4800, Tokyo, Japan). A transmission electron microscope (TEM, JEOL/JEM-2100, Hitachi, Tokyo, Japan) was used to analyze the structure and morphology of AuNPs, AuNPs/BC and AuNPs/BC-poly(HEMA) at the acceleration voltage of 100 kV. Fourier transform infrared spectroscopy (FT-IR, Nicolet Nexus, Thermo Electron Corporation, Waltham, MA, USA) was used to test the chemical structure of BC, BC-Br, BC-poly(HEMA). Using TGA (TGA-Q500, TA Instruments Corporation, New Castle, DE, USA) the thermal stability of samples were determined with the heating rate at 10 • C/min within the scope 30 to 800 • C under N 2 . The crystal structure of BC, AuNPs/BC-poly(HEMA) were measured by Powder D8 Advance X-ray diffraction (XRD, Bruker AXS D8, Leipzig, Germany). X-ray photoelectron spectroscopy (XPS, Escalab 250Xi, Thermo Scientific Escalab, Thermo Fisher Scientific, Waltham, MA, USA) was also used to perform the elemental analysis of BC-Br and AuNPs/BC-poly(HEMA). The optical spectra and catalytic activity of the AuNPs/BC-poly(HEMA) was examined towards a Hitachi U4100 UV-visible spectrophotometer.

Morphology Analysis
The morphologies are shown in Figure 3 as carried out by FE-SEM and TEM. Figure 3a shows that the micro-pore structure of pure BC membrane with ultrafine nanofiber structure and non-woven mechanically robust network, which provided high specific surface area and porosity. In addition, the pure BC presents a typical 3D network structure. The size distribution range of BC (inset image of Figure 3a) was 20 to 70 nm, with an average diameter of about 45 nm. After surface modification by ATRP, BC-poly(HEMA) (Figure 3b) showed no obvious damage, and still maintained similar morphology to that of BC. This is due to the fact that the surface modification of ATRP caused little damage to the BC membrane under non-violent conditions; in addition, BC can withstand a certain impact because of its excellent mechanical properties. In the one-step method synthesis of AuNPs, the characteristic peaks were observed at 524 nm ( Figure S1), which is consistent with the literature.

Characterization of BC-Poly(HEMA)
Figure 4a reveals that the SEM-energy-dispersive X-ray spectroscopy (EDS) mapping images contained C, O and Br elements on the surface of BC-Br initiator, thus indicating their uniform distribution in the BC membrane. Additionally, Figure 4b shows the observation of C, O and Br peaks on EDS spectrum, and the relative concentration of the above elements were about 37.49, 53.31 and 9.20 wt%, respectively. Furthermore, XPS of the BC-Br membrane are shown in Figure 4c, with the inset shown the magnified images at the binding energy range from 80 to 60 eV. The survey spectrum indicated the presence of C, O and Br elements, where the large peaks observed at 533 and 282 eV correspond to O1s and C1s, respectively. The high-resolution spectrum of Br 3d located at 68 eV confirms the binding of the Br-initiator to the BC membrane. Figure 4d depicts FT-IR spectra at the wavenumber from 4000 to 500 cm −1 acquired from BC, BC-Br and BC-poly(HEMA) to evaluate the effect of ATRP modification. As shown in Figure 4d, BC membrane has two characteristic bands centered at the wavenumber of 3347 cm -1 and 2905 cm −1 , which were attributed to O-H stretching and C-H stretching, respectively. After initiation by 2-BIB, -H on the hydroxyl group was replaced by -Br, as the evidenced of the significant reduction of O-H stretching at the wavenumber of 3347 cm −1 , and a new absorption band at the wavenumber of 1720 cm -1 was attributed to C=O stretching. By further modification of ATRP, the two FT-IR bands centered at 3347 cm -1 (O-H) and 1720 cm −1 (C=O) were obviously increased, due to the successful surface-grafting with HEMA.

Characterization of BC-Poly(HEMA)
Figure 4a reveals that the SEM-energy-dispersive X-ray spectroscopy (EDS) mapping images contained C, O and Br elements on the surface of BC-Br initiator, thus indicating their uniform distribution in the BC membrane. Additionally, Figure 4b shows the observation of C, O and Br peaks on EDS spectrum, and the relative concentration of the above elements were about 37.49, 53.31 and 9.20 wt%, respectively. Furthermore, XPS of the BC-Br membrane are shown in Figure 4c, with the inset shown the magnified images at the binding energy range from 80 to 60 eV. The survey spectrum indicated the presence of C, O and Br elements, where the large peaks observed at 533 and 282 eV correspond to O1s and C1s, respectively. The high-resolution spectrum of Br 3d located at 68 eV confirms the binding of the Br-initiator to the BC membrane. Figure 4d depicts FT-IR spectra at the wavenumber from 4000 to 500 cm −1 acquired from BC, BC-Br and BC-poly(HEMA) to evaluate the effect of ATRP modification. As shown in Figure 4d, BC membrane has two characteristic bands centered at the wavenumber of 3347 cm -1 and 2905 cm −1 , which were attributed to O-H stretching and C-H stretching, respectively. After initiation by 2-BIB, -H on the hydroxyl group was replaced by -Br, as the evidenced of the significant reduction of O-H stretching at the wavenumber of 3347 cm −1 , and a new absorption band at the wavenumber of 1720 cm -1 was attributed to C=O stretching. By further modification of ATRP, the two FT-IR bands centered at 3347 cm -1 (O-H) and 1720 cm −1 (C=O) were obviously increased, due to the successful surface-grafting with HEMA.

Characterization of AuNPs/BC-Poly(HEMA)
SEM-EDS mapping images of AuNPs/BC-poly(HEMA) contained C, O and Au elements were verified as revealed in Figure 5a, indicated the uniform distribution of Au on the suface of membrane. Moreover, it can be seen that the relative concentration of C, O, Au, Na and Cl elements on EDS spectra in Figure 5b were about 43.73, 49.46, 5.87, 0.80 and 0.14 wt%, respectively, while the corresponding element content of AuNPs/BC were about 46.78, 50.65, 2.36, 0.13 and 0.07 wt% ( Figure  5c), which indicated that the BC-poly(HEMA) had higher loading of AuNPs than pure BC.
TGA was used to study the thermal stability of BC, AuNPs/BC, AuNPs/BC-poly(HEMA).
The results shown in Figure 5d depicts three stage of weight loss. The first stage from ambient to 100 °C corresponds to the evaporation of water, while the other two stages of 100 °C to 350 °C and 350 °C to 600 °C are due to the collapse of a cellulose skeleton [36,37]. The total weight loss rates of BC, AuNPs/BC, AuNPs/BC-poly(HEMA) were 93 wt%, 86 wt% and 82 wt%, respectively, which indicated the improvement of thermal stability by protection from AuNPs. Meanwhile, Au content in the samples of AuNPs/BC and AuNPs/BC-poly(HEMA) determined on the basis of TGA were 7% and 11%, respectively, which indicated the higher loading of AuNPs on AuNPs/BC-poly(HEMA).
In addition, the Au oxidation state in the as-prepared AuNPs/BC-poly(HEMA) was obtained by XPS analysis as indicated in Figure 5c [38], with the inset showing the magnified images at the binding energy range from 95 to 77 eV. It can be seen that the strong peaks at 533 and 284 eV were attributed to the BC, and the fitted Au 4f peaks consisting of Au 4f 5/2 and Au 4f 7/2 distinctively showed on the magnified images at peaks of 87.4 and 83.7 eV, respectively. The results are in accordance with the literature values reported for Au (0) [39]. Au(III) peaks at 86.9 and 90.6 eV are absent in the XPS spectrum of the AuNPs/BC-poly(HEMA), which could be ascribed to the small

Characterization of AuNPs/BC-Poly(HEMA)
SEM-EDS mapping images of AuNPs/BC-poly(HEMA) contained C, O and Au elements were verified as revealed in Figure 5a, indicated the uniform distribution of Au on the suface of membrane. Moreover, it can be seen that the relative concentration of C, O, Au, Na and Cl elements on EDS spectra in Figure 5b were about 43.73, 49.46, 5.87, 0.80 and 0.14 wt%, respectively, while the corresponding element content of AuNPs/BC were about 46.78, 50.65, 2.36, 0.13 and 0.07 wt% (Figure 5c), which indicated that the BC-poly(HEMA) had higher loading of AuNPs than pure BC.
TGA was used to study the thermal stability of BC, AuNPs/BC, AuNPs/BC-poly(HEMA). The results shown in Figure 5d depicts three stage of weight loss. The first stage from ambient to 100 • C corresponds to the evaporation of water, while the other two stages of 100 • C to 350 • C and 350 • C to 600 • C are due to the collapse of a cellulose skeleton [36,37]. The total weight loss rates of BC, AuNPs/BC, AuNPs/BC-poly(HEMA) were 93 wt%, 86 wt% and 82 wt%, respectively, which indicated the improvement of thermal stability by protection from AuNPs. Meanwhile, Au content in the samples of AuNPs/BC and AuNPs/BC-poly(HEMA) determined on the basis of TGA were 7% and 11%, respectively, which indicated the higher loading of AuNPs on AuNPs/BC-poly(HEMA).
In addition, the Au oxidation state in the as-prepared AuNPs/BC-poly(HEMA) was obtained by XPS analysis as indicated in Figure 5c [38], with the inset showing the magnified images at the binding energy range from 95 to 77 eV. It can be seen that the strong peaks at 533 and 284 eV were attributed to the BC, and the fitted Au 4f peaks consisting of Au 4f 5/2 and Au 4f 7/2 distinctively showed on the magnified images at peaks of 87.4 and 83.7 eV, respectively. The results are in accordance with the literature values reported for Au (0) [39]. Au(III) peaks at 86.9 and 90.6 eV are absent in the XPS spectrum of the AuNPs/BC-poly(HEMA), which could be ascribed to the small amount of Au(III) in the reaction system after the hydrothermal reaction and confirmed the almost completed reduction of Au(III) to Au(0). Additionally, the XPS atomic percent values of Au atoms on AuNPs/BC-poly(HEMA) were 7.3%, while Au atoms on AuNPs/BC were only 3.1%. Figure 5d shows the XRD patterns of as-prepared BC and AuNPs/BC-poly(HEMA) from 10 • to 90 • (2θ). Several diffraction peaks centered at 2θ of 14.  [40], also consistent with previous SEAD pattern of Au. XRD results confirmed the successful synthesis of AuNPs. Meanwhile, this also indicated that the presence of AuNPs had no obvious effect on the crystallography of BC. Nanomaterials 2019, 9, x FOR PEER REVIEW 7 of 12 amount of Au(III) in the reaction system after the hydrothermal reaction and confirmed the almost completed reduction of Au(III) to Au(0). Additionally, the XPS atomic percent values of Au atoms on AuNPs/BC-poly(HEMA) were 7.3%, while Au atoms on AuNPs/BC were only 3.1%. Figure 5d shows the XRD patterns of as-prepared BC and AuNPs/BC-poly(HEMA) from 10° to 90° (2θ). Several diffraction peaks centered at 2θ of 14.2°, 16.9° and 22.7° correspond to the (100), (010) and (002) of BC. On the XRD pattern of AuNPs/BC-poly(HEMA), several peaks located at 2θ of 38.2°, 44.1°, 65.0°, 78.1°, and 83.4° appear, corresponding to the diffraction of (111), (200), (220), (311), and (222) planes of Au, respectively [40], also consistent with previous SEAD pattern of Au. XRD results confirmed the successful synthesis of AuNPs. Meanwhile, this also indicated that the presence of AuNPs had no obvious effect on the crystallography of BC.

Catalytic Activity of AuNPs/BC-Poly(HEMA)
As shown in Figure 6, the color of the reaction system faded from yellow to transparent. The reduction process was monitored by UV-visible spectroscopy. The yellow color of 4-NP solution changed deeper while added NaBH4, resulting in the maximum absorption peak of mixed solution

Catalytic Activity of AuNPs/BC-Poly(HEMA)
As shown in Figure 6, the color of the reaction system faded from yellow to transparent. The reduction process was monitored by UV-visible spectroscopy. The yellow color of 4-NP solution changed deeper while added NaBH 4 , resulting in the maximum absorption peak of mixed solution transferred from 317 nm to 400 nm due to the formation of 4-nitrophenolate ion. Figure 7a shows the UV spectrum of 4-NP/NaBH 4 with BC-poly(HEMA) and no obvious changes in the absorbance at 400 nm can be detected within 60 min. The results confirmed that BC-poly(HEMA) had almost no catalytic activity for the model reaction After replacing BC-poly(HEMA) with AuNPs/BC-poly(HEMA), the mixed aqueous solution showed a very different phenomenon with small bubbles continually released from AuNPs/BC-poly(HEMA), indicating the reduction reaction in the system. Figure 7b showed the absorbance intensity of the reaction system at 400 nm had a marked decline and a new peak presented at 300 nm on the typical time-dependent UV-vis spectra, which caused by the reduction of 4-NP to 4-AP. The reducing reaction was fast and the 4-NP can be almost completely reduced within 50 min. Nanomaterials 2019, 9, x FOR PEER REVIEW 8 of 12 transferred from 317 nm to 400 nm due to the formation of 4-nitrophenolate ion. Figure 7a shows the UV spectrum of 4-NP/NaBH4 with BC-poly(HEMA) and no obvious changes in the absorbance at 400 nm can be detected within 60 min. The results confirmed that BC-poly(HEMA) had almost no catalytic activity for the model reaction After replacing BC-poly(HEMA) with AuNPs/BCpoly(HEMA), the mixed aqueous solution showed a very different phenomenon with small bubbles continually released from AuNPs/BC-poly(HEMA), indicating the reduction reaction in the system. Figure 7b showed the absorbance intensity of the reaction system at 400 nm had a marked decline and a new peak presented at 300 nm on the typical time-dependent UV-vis spectra, which caused by the reduction of 4-NP to 4-AP. The reducing reaction was fast and the 4-NP can be almost completely reduced within 50 min. The relationship between absorbance at 400 nm on the UV-vis spectrum and reaction time was plotted to reflect the reaction kinetics and estimate the catalytic activity of AuNPs/BC-poly(HEMA). Figure 7c displays the effect of pH on the catalytic activity of AuNPs/BC-poly(HEMA) toward the 4-NP reduction by using the model of the plot of ln(At/A0) versus reaction time, where At is the absorbance at reaction time t and A0 is the initial absorbance. The inset image shows the apparent reaction rate constant kapp (min −1 ), which was obtained by fitting the experimental results. The result depicts that ln(At/A0) decreases linearly with reaction time and the reduction of 4-NP by AuNPs/BCpoly(HEMA) was closely dependent on pH value in the system. As shown in the inset image of Figure  7c, the kapp for 4-NP reduction is 5.28 × 10 −2 , 4.16 × 10 −2 and 2.42 × 10 −2 min −1 at pH 3, 7 and 9, respectively, and it is evident that low pH is conducive to high reduction efficiency. The reduction of 4-NP over AuNPs/BC-poly(HEMA) in the presence of NaBH4 follows the Langmuir-Hinshelwood kinetics and the step of adsorption is the important part of reduction reaction [41]; therefore, the negatively charged BH 4− could be easily adsorbed on the surface of AuNPs/BC-poly(HEMA) at low pH values, resulting in an improvement of rate and efficiency for 4-NP reduction under acidic conditions. Figure 7d shows the ln(At/A0) versus reduction time at different temperatures to investigate the relationship between catalytic activity and temperature of AuNPs/BC-poly(HEMA). As shown in the inset image of Figure 7d, with the increase of temperature at the range of 25 °C to 50 °C the kapp for reduction reaction was in the range of 3.39 × 10 −2 to 9.14 × 10 −2 min −1 . Meanwhile, kapp reached the highest value at 45 °C, and showed no obvious increase with the further increase of temperature. The heterogeneous catalytic reactions customarily obey the key molecular principle of catalysis, which corresponded to the Sabatier principle [42] stating that the substrate molecules must be adsorbed onto the catalyst and then activated, whilst the product molecules must also be desorbed. On the basis of this principle, the rate of catalytic reaction reaches a maximum when the rate of activation and product desorption were balanced. In this work, the balance was calculated at around 45 °C, where the maximum volcanic point was reached. The relationship between absorbance at 400 nm on the UV-vis spectrum and reaction time was plotted to reflect the reaction kinetics and estimate the catalytic activity of AuNPs/BC-poly(HEMA). Figure 7c displays the effect of pH on the catalytic activity of AuNPs/BC-poly(HEMA) toward the 4-NP reduction by using the model of the plot of ln(A t/ A 0 ) versus reaction time, where A t is the absorbance at reaction time t and A 0 is the initial absorbance. The inset image shows the apparent reaction rate constant k app (min −1 ), which was obtained by fitting the experimental results. The result depicts that ln(A t /A 0 ) decreases linearly with reaction time and the reduction of 4-NP by AuNPs/BC-poly(HEMA) was closely dependent on pH value in the system. As shown in the inset image of Figure 7c, the k app for 4-NP reduction is 5.28 × 10 −2 , 4.16 × 10 −2 and 2.42 × 10 −2 min −1 at pH 3, 7 and 9, respectively, and it is evident that low pH is conducive to high reduction efficiency. The reduction of 4-NP over AuNPs/BC-poly(HEMA) in the presence of NaBH 4 follows the Langmuir-Hinshelwood kinetics and the step of adsorption is the important part of reduction reaction [41]; therefore, the negatively charged BH 4− could be easily adsorbed on the surface of AuNPs/BC-poly(HEMA) at low pH values, resulting in an improvement of rate and efficiency for 4-NP reduction under acidic conditions. Figure 7d shows the ln(A t /A 0 ) versus reduction time at different temperatures to investigate the relationship between catalytic activity and temperature of AuNPs/BC-poly(HEMA). As shown in the inset image of Figure 7d, with the increase of temperature at the range of 25 • C to 50 • C the k app for reduction reaction was in the range of 3.39 × 10 −2 to 9.14 × 10 −2 min −1 . Meanwhile, k app reached the highest value at 45 • C, and showed no obvious increase with the further increase of temperature. The heterogeneous catalytic reactions customarily obey the key molecular principle of catalysis, which corresponded to the Sabatier principle [42] stating that the substrate molecules must be adsorbed onto the catalyst and then activated, whilst the product molecules must also be desorbed. On the basis of this principle, the rate of catalytic reaction reaches a maximum when the rate of activation and product desorption were balanced. In this work, the balance was calculated at around 45 • C, where the maximum volcanic point was reached.

Reusable Stability of AuNPs/BC-Poly(HEMA)
The reusability must be considered in evaluating the engineering application of the catalyst. In this work, the reusability of AuNPs/BC-poly(HEMA) was examined through degradation of 4-NP for 10 cycles. As shown in Figure 8, the removal of 4-NP by AuNPs/BC-poly(HEMA) after reusing for 10 times was still around 90%, almost the same as the removal rate for the first use. Moreover, the TEM image ( Figure S4) shows barely reduction in AuNPs on the surface of the AuNPs/BC-poly(HEMA) membrane after 10 reuses. Meanwhile, the comparison of catalytic activity by Au based nanocatalysts for the reduction of 4-NP was also shown in Table S1. These results clearly indicate that AuNPs/BCpoly(HEMA) is structurally stable with high catalysis efficiency and reusability.

Reusable Stability of AuNPs/BC-Poly(HEMA)
The reusability must be considered in evaluating the engineering application of the catalyst. In this work, the reusability of AuNPs/BC-poly(HEMA) was examined through degradation of 4-NP for 10 cycles. As shown in Figure 8, the removal of 4-NP by AuNPs/BC-poly(HEMA) after reusing for 10 times was still around 90%, almost the same as the removal rate for the first use. Moreover, the TEM image ( Figure S4) shows barely reduction in AuNPs on the surface of the AuNPs/BC-poly(HEMA) membrane after 10 reuses. Meanwhile, the comparison of catalytic activity by Au based nanocatalysts for the reduction of 4-NP was also shown in Table S1. These results clearly indicate that AuNPs/BC-poly(HEMA) is structurally stable with high catalysis efficiency and reusability.

Reusable Stability of AuNPs/BC-Poly(HEMA)
The reusability must be considered in evaluating the engineering application of the catalyst. In this work, the reusability of AuNPs/BC-poly(HEMA) was examined through degradation of 4-NP for 10 cycles. As shown in Figure 8, the removal of 4-NP by AuNPs/BC-poly(HEMA) after reusing for 10 times was still around 90%, almost the same as the removal rate for the first use. Moreover, the TEM image ( Figure S4) shows barely reduction in AuNPs on the surface of the AuNPs/BC-poly(HEMA) membrane after 10 reuses. Meanwhile, the comparison of catalytic activity by Au based nanocatalysts for the reduction of 4-NP was also shown in Table S1. These results clearly indicate that AuNPs/BCpoly(HEMA) is structurally stable with high catalysis efficiency and reusability.

Conclusions
Herein, we developed a highly recyclable novel nanocatalyst of AuNPs/BC-poly(HEMA) for the reduction of 4-NP. The monomer of HEMA was successfully grafted on the surface of BC membrane via the ATRP method and then the BC-poly(HEMA) obtained acted as not only the reducing agent but also the carrier for preparing AuNPs. The experimental results showed that the as-prepared AuNPs/BC-poly(HEMA) was an excellent catalyst in the model reaction. The reduction efficiency and rate was highly dependent on environmental conditions, including pH and temperature. Moreover, the catalytic activity of AuNPs/BC-poly(HEMA) could be reused for 10 cycles without considerable loss. In addition, no 4-NP was detected in the degradation solution after being stored for 45 days, indicating that the AuNPs/BC-poly(HEMA) was a reliable green material and has a broad application prospect of catalytic transformation in the field of environmental remediation.
Author Contributions: X.L. and Q.F. conceived and designed the experiments; X.L. performed the experiments; X.L. and D.L. analyzed the data; X.L. wrote the paper. N.C. and Q.W. advised the paper. H.K. and Q.W. supervised the project. All authors reviewed the paper.