Introduction

Semiconductor-based photocatalysis is a promising candidate for various applications in environmental pollution mediation and solar energy conversion (Hoffmann et al. 1995; Asahi et al. 2001; Maeda et al. 2006; Tokunaga et al. 2001; Tsuji et al. 2005; Khan et al. 2002) due to its photocatalytic activity and excellent physical and chemical properties (Mor et al. 2006; O’regan and Grfitzeli 1991; Han et al. 2011). However, titanium dioxide (TiO2) has a wide band energy (3.2 eV) and small UV fraction (ca. 4 %) of the solar light used in photocatalysis (Chen et al. 2010). Being confronted with organic pollutants and the energy crisis, the development of semiconductor-based photocatalysis with a high level of activity under visible or solar light is a major topic. In recent years, several visible light-driven photocatalytic semiconductors such as TiO2-xNx (Asahi et al. 2001), CaBi2O4 (Tang et al. 2004), BiVO4 (Kudo et al. 1998) and Silver orthophosphate (Ag3PO4) have been reported. Among them, Ag3PO4 is a new type of photocatalyst with a body-centered cubic structure, possessing an excellent photocatalytic activity to decompose organic pollution because of the highly dispersive band structure (Yi et al. 2010; Bi et al. 2011; Ge et al. 2012). Moreover, the size of Ag3PO4 remains relatively large, which might have an effect on the photcatalytic activity (Dinh et al. 2011). The higher surface area of smaller particle size is thought to be of benefit to the photocatalytic reaction, which mostly occurs on the catalyst surface (Linsebigler et al. 1995). To enhance the photocatalytic activity of Ag3PO4, construction of Ag3PO4 nanoparticles with highly uniform size is also an efficacious method to remove contaminants in solution. However, it is very difficult to remove nanoparticles from solutions, and this may cause secondary pollution to the environment. To solve this problem, hybridization of nanoparticles with renewable biomass macromolecules such as cellulose and chitin has been very impressive (Sehaqui et al. 2010; Caruso and Schattka 2000; Murray et al. 2005).

Cellulose is the most abundant and renewable natural polymer on earth as well as one of the raw materials for agriculture and industry. However, cellulose is the most intransigent macromolecule, being difficult to dissolve. In our laboratory, 4.6 wt% LiOH/15 wt% urea and 7 wt% NaOH/12 wt% urea aqueous solution precooled to −12 °C were developed to dissolve cellulose and yield a transparent cellulose solution (Cai et al. 2008b). Moreover, from the cellulose solution, regenerated cellulose hydrogels (Chang et al. 2009, a, b), films, fibers (Cai et al. 2007) and cellulose-based nanocomposites (Cai et al. 2012; Qi et al. 2009; Cai et al. 2008a; Liu et al. 2008; Dankovich and Gray 2011) have been fabricated, showing remarkable mechanical, optical and thermal properties. The portable photocatalyst using hydrogel as a matrix was conducive to water penetration and could be very efficient in degrading organic pollution. Therefore, photocatalyst/cellulose hydrogels were constructed for use in water treatment of organic pollution. In this article, we developed a strategy for cellulose-inorganic hybridization via in-suit reduction and oxidation to fabricate Ag3PO4 nanoparticles in regenerated cellulose hydrogels. The cellulose-based nanocomposites with finely distributed nanoparticles of Ag3PO4 were fabricated to achieve a maximum Ag3PO4 loading of about 26 %. Moreover, the Ag3PO4/cellulose nanocomposites as photocatalysts of decomposing organic dye (Rh B) under natural sunlight were evaluated. Our findings may provide a new and facile pathway for fabricating photocatalytic materials to solve the problem of organic pollution under natural sunlight and expand the application of cellulose.

Experimental section

Materials

Cotton linter pulp with an α-cellulose content of more than 95 % was provided by Hubei Chemical Fiber Co.n Ltd., China. Its viscosity-average molecular weight (M η) was determined to be 9.8 × 104 (Cai et al. 2006) by viscometry at 25 ± 0.05 °C in aqueous 4.6 wt% LiOH/15.0 wt% urea solution. Urea, LiOH·H2O, silver nitrate (AgNO3), 30 % hydrogen peroxide (H2O2) and other reagents were purchased from Hubei ShengShi Chemical Reagent Co., Ltd.

Preparation of regenerated cellulose hydrogels

The 4.6 wt% LiOH/15.0 wt% urea aqueous solution was precooled to −12 °C, and the desired amount of cellulose pulp was immediately dispersed and stirred to obtain a transparent cellulose solution with 4 wt% concentration. The cellulose solution was subjected to centrifugation at 5,000 rpm for 15 min at 5 °C in order to carry out the degasification. The resulting transparent solution was cast on a glass plate to give a thickness of 0.5 mm and then immersed in a nonsolvent solution (5 wt% H2SO4 solution) for the desired time to form regenerated cellulose gel (RC). Regenerated cellulose gel film was washed thoroughly with deionized water to obtain cellulose hydrogel.

Preparation of Ag3PO4/cellulose composite hydrogel (SP)

The hydrothermal reduction of AgNO3 by cellulose was performed by immersing the regenerated cellulose hydrogel into AgNO3 aqueous solution at 80 °C for 24 h. The AgNO3 concentration was 0.05, 0.1 and 0.4 mol/l, respectively. Then, the cellulose hydrogels containing Ag nanoparticles were washed with deionized water and dipped into 0.2 M Na2HPO4/200 ml 30 % H2O2 solution for 0.5 h at room temperature to transform Ag3PO4. The resultant Ag3PO4/cellulose composite hydrogels were rinsed with deionized water for three times before using. The Ag3PO4/cellulose composite hydrogels prepared from AgNO3 solution with concentrations of 0.05, 0.1 and 0.4 mol/l were coded as SP05, SP10 and SP40, respectively. The aerogel samples were prepared by freeze-drying from the hydrogels for the measurement of SEM and X-ray diffraction (XRD).

Characterization

Scanning electron microscopy (SEM) observation was carried out on an FE-SEM (SIRION TMP, FEI) by using an accelerating voltage of 12 kV. The samples were coated with Pt for the SEM observation. Transmission electron microscopy (TEM) was observed on a JEOL JEM-2100 at an accelerating voltage of 200 kV. The samples were embedded using resin, and ultrathin sectioning was placed with a copper line. Attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy was measured on an FT-IR spectroscopy (model 1600, PerkinElmer Co., USA). Samples were cut into powder and dried in an oven at 60 °C for 24 h. X-ray diffraction (XRD) measurement was taken on an XRD diffractometer (D8-Advance, Bruker, USA) in a symmetric reflection mode. The patterns with Cu Kα radiation (λ = 0.15406 nm) at 40 kV and 40 mA were recorded in the 2θ range of 8 to 90°. All samples were cut into powder to remove the influence of the crystalline orientation. The crystallite size (D) of the sample for the (210) plane of Ag3PO4 nanoparticle was calculated by the Scherrer formula(Patterson 1939):

$$D = 0.9\lambda /(\beta \cdot \cos \theta )$$
(1)

where λ = 0.15406 nm, β is the corrected integral width, and θ is the Bragg angle. X-ray photoelectron spectroscopy (XPS) was recorded on an XSAM 800 Instrument (Kratos, UK). An MgKa target at 1,253.6 eV and 16 mA × 12.5 kv was used in the experiment. Regular light absorbance was measured at wavelengths from 200 to 800 nm using a UV-visible spectrometer (U-4100, Hitachi High-Technologies Corp.). The photocatalytic activities of Ag3PO4/cellulose nanocomposite hydrogel were tested by using a rhodamine B (Rh B) aqueous solution. A piece of (6.4 cm × 2.6 cm × 0.5 cm) Ag3PO4/cellulose nanocomposite hydrogel was immersed into Rh B aqueous solution (100 ml, 0.01 g/L) for 8 h in the dark. Then the Rh B aqueous solution including the sample was illuminated under sunlight. The change in the Rh B concentration was monitored using a UV-vis spectrophotometer at a wavelength of 552 nm. Thermal gravimetric analysis (TGA) was carried out by a Netzsch thermogravimetric analyzer (STA449C, 31MFC, G Jupiter, German). The dry samples were cut into powder in a crucible and heated from 30 to 600 °C at a heating rate of 10 K/min in nitrogen and air atmosphere, respectively. Tensile tests of the Ag3PO4/cellulose composite hydrogels were performed using a universal testing machine (CMT6503, Shenzhen SANS Test Machine Co., Ltd., Shenzhen, China) with a tensile rate of 5 mm min−1 to obtain tensile strength (\(\sigma_{\text{b}}\)) and elongation at break (\(\varepsilon_{\text{b}}\)). Rectangular-shaped specimen strips were 5 mm in length and 0.5 mm in width. Densities of the samples were calculated by weighing the samples and measuring the volumes.

Results and discussions

Evidence of Ag3PO4 nanoparticles immobilized in the cellulose matrix

Figure 1a shows a schematic illustration for the preparation of the Ag3PO4/cellulose nanocomposites. Silver ions can be reduced to silver nanoparticle by hydroxyl groups of cellulose and reducing end groups (Kotel’nikova et al. 2003). Moreover silver nanoparticles were dispersed in cellulose matrix, and the size of the silver nanoparticles could also be controlled in the pore. When the cellulose hydrogel films were immersed into AgNO3 aqueous solution, Ag+ ions were tightly stabilized in the cellulose hydrogel by interacting with –OH groups of the cellulose hydrogel because of electrostatic interactions (Cai et al. 2008a; Liu et al. 2011a). Subsequently, the cellulose hydrogel with AgNO3 was heated. Ag ions were reduced by cellulose end aldehyde and alcohol groups, and Ag nanoparticels were stabilized in cellulose hydrogels. By using the oxidation ability of H2O2, Ag nanoparticles were transformed into Ag3PO4 nanoparticles by the reaction of Ag and HPO4 2−. Hydrogen peroxide of cellulose can oxidize Ag particle to Ag+ ions (Wang et al. 2012). When H2O2 has been introduced, the redox potential of H2O2/H2O (E θ = +1.776 V) was higher than that of Ag3PO4/Ag (E θ = +0.451 V) (Hu et al. 2013). Thus, the color of cellulose hydrogels (b) was transformed into a brown color (c) after hydrothermal reduction in AgNO3 solution and then immediately changed to yellow (d) when cellulose hydrogels were immersed into Na2HPO4 hydrogen peroxide solution, indicating the formation of Ag3PO4 nanoparticles.

Fig. 1
figure 1

Schematic illustration of the fabrication of Ag3PO4 nanoparticles in the cellulose hydrogel film (a, i Ag ions were reduced to Ag nanoparticle by cellulose at 80 °C; ii Ag3PO4 nanoparticles were fabricated in NaH2PO4 of 30 % H2O2 solution.) and the photograps of cellulose hydrogel (b), Ag/cellulose composite (c) and Ag3PO4/cellulose nanocomposite (d). (The scale bar is 2 cm)

Full size image

FT-IR spectra of RC and SP05, SP10, SP40 are shown in Fig. 2. The peak at 3,420 cm−1 is attributed to hydroxyl group (OH) stretching vibrations of cellulose in the composites. A broad peak at around 3,400 cm−1 corresponds to a large number of hydroxyl groups in the Ag3PO4 nanoparticle surface (Chen et al. 2007; Nguyen et al. 2009). The results indicated that Ag3PO4 nanoparticles retained the intermolecular hydrogen bond in the cellulose hydrogels. The two peaks at 1,021 and 550 cm−1 corresponded to P–O stretching vibrations of PO4 3− ions (Thomas et al. 2002). This confirmed that the Ag+ ions existed in the Ag3PO4 form in the composite hydrogels.

Fig. 2
figure 2

FT-IR spectra of RC (a) and Ag3PO4/cellulose nanocomposite hydrogels (b SP05, c SP10, d SP40)

Full size image

The X-ray photoelectron spectroscopy (XPS) was carried out to investigate the surface composition and chemical state of the composite hydrogels. Figure 3 shows the XPS spectra for the Ag3PO4/cellulose composite hydrogel (SP10). The C1s spectra of the Ag3PO4/cellulose nanocomposite are shown in Fig. 3a. There is an individual peak at 286.3 eV attributed to the C–O bond of cellulose (Belgacem et al. 1995). The peaks at the binding energy of 374.3 and 368.3 eV are attributed to Ag3d 3/2 and Ag3d 5/2 binding energies (Ge et al. 2012; Zhang et al. 2008; Murray et al. 2005; Chen et al. 2006). The peaks at 133.2 and 532.7 eV correspond to the P5+ and O2− anion (Zheng et al. 2008; Yang et al. 2009; Zhao et al. 2009). The results from XPS indicated that the Ag3PO4 nanoparticles were successfully synthesized in situ in the cellulose hydrogels.

Fig. 3
figure 3

The XPS spectra for the Ag3PO4/cellulose nanocomposite hydrogel (SP10)

Full size image

The morphology of the cellulose and Ag3PO4/cellulose composite hydrogels were observed by SEM. The SEM images of the RC hydrogel and Ag3PO4/cellulose composite hydrogels are shown in Fig. 4. The RC hydrogels display a micro- and nanoporous structure, which was caused by the phase separation of the cellulose solution during the regenerating process. The SEM images of the Ag3PO4/cellulose composite hydrogels in Fig. 4b–d displayed that the Ag3PO4 nanoparticles were embedded in the cellulose matrix and were filled into the pore of the cellulose hydrogels. Clearly, the Ag3PO4 nanoparticles were uniformly distributed in the cellulose matrix. As shown in Fig. 4, the composite hydrogels of SP05, SP10 and SP40 exhibited a relatively denser surface than those of RC. It was confirmed that Ag3PO4 particles were synthesized in situ in the cellulose hydrogel and tightly hybridized with a cellulose backbone via hydrogen bond and electrostatic interactions. The porous structure of the cellulose hydrogels supplied not only cavities for the formation of the Ag3PO4 nanoparticles, but also a shell to protect their nanostructure.

Fig. 4
figure 4

The SEM images of cellulose hydrogel (RC) (a) and Ag3PO4/cellulose nanocomposites SP05 (b), SP10 (c) and SP40 (d)

Full size image

Figure 5 shows the TEM images of the Ag3PO4/cellulose composites and size histograms of the Ag3PO4 nanoparticles. As expected, Ag3PO4 nanoparticles at each AgNO3 concentration were synthesized and dispersed uniformly in the regenerated cellulose hydrogels without aggregation. The size histograms showed that the average diameter of the Ag3PO4 particle of SP05, SP10 and SP40 increased gradually from 3.1 ± 2.7 to 11 ± 4.5 nm with an increase in AgNO3 concentration. It has been reported that the growth and aggregation of nanoparticles can be controlled in cellulose hydrogel because the pores of the cellulose matrix can limit the growth of the nanoparticles (Luo et al. 2009; Pinto et al. 2008; Yu et al. 2012). Therefore, the microporous nature of cellulose hydrogels could play an important role in the in situ synthesis, dispersion and stabilization of Ag3PO4 inorganic nanoparticles.

Fig. 5
figure 5

HRTEM images (a, b, c) and particle size histograms (d, e, f) of Ag3PO4 nanoparticles corresponding to a-c. (All scale bars are 10 nm)

Full size image

The X-ray diffraction (XRD) patterns of RC, SP05, SP10 and SP40 are shown in Fig. 6. There are two peaks of RC at 2θ = 12.2°, 20.2°, corresponding to the \((1\bar{1}0)\), (110) crystal plane of cellulose II crystalline (Isogai et al. 1989). However, the X-ray diffraction patterns of SP05, SP10 and SP40 clearly showed that all peaks of the samples are a near-systematic superposition of those of pure cellulose and body-centered cubic structure of Ag3PO4 (JCPDS card no. 060505) (Yi et al. 2010). The peak intensity of Ag3PO4 intensified with the enhancement of the AgNO3 solution concentrations, which is a sign of the increasing of Ag3PO4 amounts in the composite hydrogels. The average crystal size of Ag3PO4 nanoparticles for SP05, SP10 and SP40 obtained using Scherrer’s formula (Formula 1) was 12.9, 15.5 and 19.3 nm, respectively. Therefore, the Ag3PO4 nanoparticles were successfully constructed via a facile and simple pathway, leading to the conformation of inorganic/cellulose composite materials.

Fig. 6
figure 6

XRD patterns of the RC (a), SP05 (b), SP10 (c) and SP40 (d), respectively

Full size image

Photocatalytic properties of Ag3PO4/cellulose composite under sunlight

Figure 7 shows the UV-visible spectra of Ag3PO4/cellulose composite (SP05, SP10 and SP40). The results indicated that the Ag3PO4/cellulose composites (SP05, SP10, SP40) can absorb energy with a wavelength shorter than ∽530 nm. Moreover, the absorption peak edge of the Ag3PO4/cellulose nanocomposite hydrogels shifted to higher wavelength with the increasing of Ag3PO4 concentration. The band gap (E g) of the semiconductor can be obtained from the equation below (Kim et al. 2008):

Fig. 7
figure 7

The UV-vis spectrum (a) of Ag3PO4/cellulose composites (SP05, SP10, SP40) and the plot of (αhv)2 versus hv (bottom) for the direct transition as a function of photon energy (b)

Full size image
$$\alpha hv = k(hv - E_{\text{g}} )^{n/2}$$
(2)

where α, v and k are the adsorption coefficient, light frequency and proportionality constant, respectively. Figure 7b shows the curves of (αhv)2 versus hv for a direct transition with n equal to 1. The direct band gap fit of the absorption edges were calculated to be 2.21, 2.27 and 2.31 eV for SP05, SP10 and SP40, respectively. It was not difficult to imagine that the Ag3PO4 nanoparticles in the cellulose hydrogel had a good photocatalytic ability as a result of an increase in the apparent bandgap energy.

To prove the narrower band gap to enhance photon energy utilization for comparison, the photocatalytic degradation of Rh B solution by Ag3PO4/cellulose hydrogels was evaluated under natural sunlight at room temperature. As shown in Fig. 8, all of the nanocomposite hydrogels exhibited good photocatalytic activities for the Rh B degradation reaction under sunlight. The concentration of the Ag3PO4 nanoparticles of all samples for photodegradation of the Rh B aqueous solution (0.1 g/l) was about 0.54, 1.4 and 3.55 mg/ml, respectively. It was noted that the photocatalytic activity of the sample (SP40) under sunlight was higher than that of SP10 and SP05. It was noted that SP40 decomposed 50 % Rh B aqueous solution within about 6.5 h under sunlight, whereas the other hyhdrogels (SP05, SP10) needed more time. This could result from the relatively high Ag3PO4 concentration of SP40. However, if we consider the degradation rate per unit mass of the Ag3PO4 particles, SP05 should possess higher degradation efficiency. Namely, the nanoparticle size was smaller and the degradation efficiency higher. Figure 8b shows the relationship between ln(C/C0) and reaction time with different Ag3PO4 concentrations. As shown in Fig. 8b, SP05 had a higher degradation efficiency. Therefore, it was suggested that the narrower band gap benefited enhanced photon energy utilization. Moreover, in Fig. 8c, the Rh B aqueous solution did not change under dark. This indicated the establishment of an adsorption-desorption equilibrium without light.

Fig. 8
figure 8

a Photodegradation of RhB over the cellulose hydrogel (RC) and Ag3PO4/cellulose nanocomposites (SP05, SP10, SP40) under natural sunlight. The inset photographs show the color changes of Rh B solution corresponding to the SP40. b The relationship between ln(C/C0) and reaction time with different concentrations of Ag3PO4. c Photodegradation of Rh B over the cellulose hydrogel (RC) and Ag3PO4/cellulose nanocomposites (SP05, SP10, SP40) at night and without light

Full size image

The cyclic photogradation curves of Rh B by SP40 composite hydrogel under sunlight are shown in Fig. 9. The composite exhibited remarkable photostability almost without any loss of photocatalytic activity after three cycles. Namely, the photocatalytic activity of the composite hydrogel was very stable. The Ag3PO4/cellulose hydrogels with excellent mechanical properties, dimensional stability and reusability can be used as portable photocatalysts.

Fig. 9
figure 9

Cyclic photodegradation curves of Rh B with the SP40 composites

Full size image

Thermal and mechanical properties of Ag3PO4/cellulose hydrogel

The satisfactory mechanical and thermal stability is essential for the successful application of the materials as a portable photocatalyst. The TGA and DTG curves for cellulose (RC) and Ag3PO4/cellulose nanocomposites (SP05, SP10, SP40) under air and nitrogen atmosphere are shown in Fig. 10. For TGA curves, the weight loss around 5 % below 150 °C resulted from the evaporation of absorbed water. The decomposition temperature of the cellulose was about 332 °C under air and nitrogen atmosphere, and the cellulose completely decomposed (about 100 %) at about 520 °C under air atmosphere and decomposed (about 80 %) at 600 °C under nitrogen atmosphere. Moreover, the decomposition temperature of nanocomposites shifted slightly to lower temperature in comparison with cellulose and shifted to lower temperature with increasing Ag3PO4 concentration under air and nitrogen atmosphere. Heating in air and nitrogen induced the decomposition of cellulose, with the first stage of the samples of SP05, SP10 and SP40 between 210 and 300 °C shifting slightly to lower temperature, probably because of the catalysis effect of Ag3PO4 particles. Moreover, the secondary stage of the sample decomposition appeared between 300 and 370 °C for SP05 and SP10, and the decomposition temperature of SP40 appeared between 300 and 600 °C. Importantly, the third stage of the samples of SP05 and SP10 appeared between 370 and 600 °C because of the burning of char (Liu et al. 2011b). Since the Ag3PO4 was relative stable, the Ag3PO4 content for SP05, SP10 and SP40 was estimated to be about 8, 16 and 26 wt%, respectively. In view of the above results, the thermal stability of the Ag3PO4/cellulose hydrogels was slightly lower than that of cellulose hydrogel, but the thermal resistance of composite hydrgels was better than that of the general synthetic plastic (Jeon et al. 2013; Soltani et al. 2013) and was sufficient for the application of portable photocatalyst.

Fig. 10
figure 10

TGA and DTG curves for RC, SP05, SP10 and SP40 under air (top) and nitrogen (bottom) atmosphere

Full size image

Figure 11 shows the typical stress-strain curves of RC, SP05, SP10 and SP40. The result indicated that the tensile strength of Ag3PO4/cellulose composite hydrogels increased slightly with an increase in Ag3PO4 content as a result of the strong interactions between the Ag3PO4 particles and cellulose hydrogel. Moreover, the Young’s modulus of Ag3PO4/cellulose composite hydrogels also increased slightly with increasing Ag3PO4 content, as determined by the rigidity of the inorganic nanoparticle. In our findings, the Ag3PO4 particles were tightly embedded in the cellulose hydrogel loading to enhance the mechanical properties of the composites. The physical properties of cellulose/Ag3PO4 are shown in Table 1. Therefore, the mechanical properties of portable photocatalytic hydrogel might be more excellent than those of general hydrogel (Nakayama et al. 2004; Xiang et al. 2006), and the portable photocatalytic hydrogels have a broader potential for photocatalytic application.

Fig. 11
figure 11

Typical stress-strain curves of RC, SP05, SP10 and SP40 at wet state

Full size image
Table 1 Mechanical properties of cellulose hydrogel and Ag3PO4/cellulose nanocomposite hydrogel
Full size table

Conclusion

Ag3PO4/cellulose nanocomposite hydrogels were successfully synthesized in situ in two steps using the cellulose reduction itself and H2O2 oxidation capacity. The Ag3PO4 nanoparticles at different AgNO3 concentrations were dispersed uniformly in the regenerated cellulose hydrogels, and their average diameter was in the range of 3.1 ± 2.7 to 11 ± 4.5 nm and slightly increased with an increase in Ag3PO4 concentration. The Ag3PO4/cellulose nanocomposite hydrogels had good degradation efficiency in the photocatalytic degradation on Rh B, and the Ag3PO4 nanoparticle size was smaller and the degradation efficiency higher. The Ag3PO4/cellulose nanocomposite hydrogels exhibited excellent mechanical properties and moderate thermal stability. This material has the potential for application in the field of visible light photocatalytic water treatment and solar energy conversion.