Introduction

In the current perspective, the heterogeneous photocatalysis has been considered as one of the economic, energy-efficient, and clean technology for removal of organic compounds such as amoxicillin, tetrodotoxin, ketamine, gentamicin, carbophenothion, ofloxacin, erythromycin, enrofloxacin, lidocaine, and ciprofloxacin from wastewater (Fakhri and Khakpour 2015; Fakhri and Naji 2017, 2016; Mohammadi et al. 2016; Fakhri and Kahi 2017; Fakhri et al. 2015, 2016; Fakhri and Behrouz 2015a, b; Hassani et al. 2018a, b, 2016). Heterogeneous photocatalysis utilizing titanium oxide (TiO2) semiconductor as a photocatalyst is more efficient environmental cleanup technique in comparison to conventional techniques. In this process, there is no residue after degradation of contaminant or pollutant, as it slowly breaks down the large contaminant molecules. Therefore, the process of sludge disposal to landfill is not anymore required. Photocatalysis is also suitable for decomposition of organic and inorganic compounds at very low concentrations ranging from 0.01 to 10 mg/L. The catalyst used in the photocatalytic process can be reused because it does not consume and remains unchanged (Hassani et al. 2018a, b, Byrne et al. 2018). Titanium oxide (TiO2) is of considerable industrial importance for photocatalytic applications due to its properties like low cost, chemical inertness, non-toxicity, and photo stability (Linsebigler et al. 1995). Its practical applications include self-cleaning surfaces, wastewater, and air purification, bacteria inactivation, H2 production via photocatalytic water splitting, and photo conversion of CO2 to methane or other lower hydrocarbons (Kumaravel et al. 2019; Hashimoto et al. 2005). However, its large-scale technological applications are limited because of its wide band gap (3.0–3.2 eV), low photonic yield, and high recombination rate of the electron-hole pair (Ward and Bard 1982). Due to the large band gap, this oxide can only utilize a small fraction of sunlight (> 5%) in ultra-violet (UV) region (λ < 387 nm). Whereas, visible light (λ > 400 nm) constitutes the major fraction (45%) of sunlight reaching to the earth surface (Ren et al. 2007). These limitations have opened a new domain of research for researchers as how to extend or shift the absorption spectra of TiO2 from UV to visible region. Many researchers have reported that any change or shift in the spectral response of TiO2 from UV to the visible region will also result in a significant increase in its photocatalytic activity (PCA).

Several approaches like encapsulation, functionalization, dye sensitization, complexation, doping, capping, surface modification, coupling with another semiconductor, and incorporation with other nanomaterials have been explored by researchers to extend the optical response of TiO2 (Shipra and Manoj 2011). Among them, doping is one of the efficient methods for inducing visible light activity (VLA) in TiO2. The very first attempt of doping in TiO2 with transition-metal was reported by Chio and co-workers in 1994 (Chio et al. 1994). However, metal doping has some limitations, because quite often, the metal dopant acts as an electron trap, and thus decreases the photocatalytic activity and sometimes causes thermal instability. As a result, many researchers have initiated to utilize non-metal dopants such as C, N, S, B, F, and I to improve photoactive properties of TiO2 (Ren et al. 2007; Viswanathan and Krishanmurthy 2012; Hamadanian et al. 2009; Xu et al. 2009; Park and Choi 2004; Hong et al. 2005). The addition of non-metal into TiO2 lattice results in mixing of the p states of the non-metal with the O-2p states that shifts the valence band edge upwards and thereby decreases the band gap. Due to this reason, the non-metal-doped TiO2 has proved to be a more efficient catalyst for photocatalytic degradation of organic pollutants.

Amongst non-metallic dopant, boron is quite an important dopant. Studies have reported that boron doping into TiO2 lattice has resulted in enhanced photocatalytic activity (Stengl et al. 2010; Zaleska et al. 2008). Zaleska et al. synthesized B-doped TiO2 catalysts through the sol-gel method and by grinding anatase powder with a dopant (Zaleska et al. 2008). They used boric acid triethyl ester and boric acid as boron sources and evaluated the photocatalytic activity of the prepared B-TiO2 catalysts through the degradation of phenol under UV and visible light irradiation. B-TiO2 prepared by grinding of anatase powder with boric acid triethyl ester exhibited better photoactivity under visible light as compared to that of pure TiO2. They further prepared B-TiO2 photocatalysts by surface impregnation method using boric acid triethyl ester as a boron precursor and observed that boron forms B-O-Ti species on the surface of TiO2 grains (Zaleska et al. 2009). Chen et al. (2006) observed that B-doped TiO2 has higher photocatalytic activity than pure TiO2 in a photocatalytic regeneration of nicotinamide adenine dinucleotide. They attributed this enhancement due to the formation of Ti3+, which facilitates the separation of excited electrons and holes pairs and also slows down their recombination rate. Sol-gel method is frequently used for in situ doping of B into TiO2 lattice. Lozano and co-workers prepared a series of boron-doped TiO2 photocatalysts (2% B-TiO2) with different water/alkoxide molar ratios through sol-gel method. Their study showed that degradation of Orange II azo dye increases with surface area, particle size, boron, and water content in photocatalyst (May-Lozano et al. 2014). A modified sol-gel method was used by Zhang et al. (2012) for the synthesis of boron-doped TiO2. The catalysts were calcined at a temperature ranging from 300 to 600 °C and the 3% B-TiO2 catalyst calcined at 400 °C showed 96.7% degradation of methyl orange under UV irradiation.

In the context of water pollution, 4-nitrophenol (4-NP) is a toxic, anthropogenic, inhibitory, and refractory organic compound. It is used extensively in chemical industries for the manufacturing of pesticides, pharmaceuticals, and synthetic dyes (Kavitha and Palanivelu 2005). As a result, it is one of the common pollutants present in industrial wastewater. According to the U.S. Environmental Protection Agency (U.S. EPA 1980), it is carcinogenic and can damage the central nervous system, liver, kidney, and blood of humans and animals. Various conventional methods like catalytic wet-air-oxidation, biological degradation, adsorption, UV-oxidation, nano-filtration, and coagulation–flocculation have been used for the treatment of wastewater, containing 4-NP. Most of these methods cannot degrade 4-NP, due to its high stability and solubility in water. Furthermore, most of these methods transfer the pollutant from one phase to another and do not provide an ultimate solution. However, the 4-NP can be reduced in the presence of suitable catalyst while using the strong reducing agent like sodium borohydride but the process is not economical as it requires the use of expensive reducing agent (Kamal et al. 2019a, b; Khan et al. 2017; Haider et al. 2016; Ali et al. 2017a, b, 2018). Thus, it is important to develop a more efficient and complete method for its remediation from wastewater before being discharged.

With this background, we believe that B-doped TiO2 can be a suitable material for photocatalytic degradation (PCD) of 4-NP. To the best of our knowledge, the potential of boron-doped TiO2 catalyst for photodegradation of 4-NP has not been studied so far. Therefore, in this work, we have performed the PCD studies of 4-NP using pure TiO2 and B-doped TiO2. The undoped and boron-doped TiO2 catalysts were synthesized through sol-gel route using titanium isopropoxide (TTIP) and boric acid (BA) as precursors for titanium oxide and boron respectively. The physicochemical and photophysical properties of the photocatalyst were investigated using various characterization techniques such as surface area and pore analysis, XRD, UV-Vis, FTIR, SEM-EDX, HRTEM, micro Raman spectroscopy, and XPS. The effect of various parameters like irradiation time, catalyst dose, and dopant concentration on PCD of 4-NP was studied. The kinetic modeling of photocatalytic degradation data was also performed to co- relates the effect of boron doping on PCD kinetics

Materials and methods

Synthesis of photocatalysts

The reagents used in this study are titanium isopropoxide (TTIP, Sigma-Aldrich, 99% purity), absolute ethanol (Merck, 99% purity), and boric acid (SRL, 99.5% purity). Pure commercial TiO2 (P25) was also purchased (Rankem, 99% purity) for comparison purpose and referred as TiO2C in the following text. The boron-doped TiO2 was prepared through sol-gel method, involving homogeneous hydrolysis of TTIP in ethanol and water as described elsewhere (Ellappan et al. 2014). Briefly, solution A was prepared by dispersing 8 ml of TTIP in 42 ml of ethanol followed by 30 min of stirring. Meanwhile, solution B was prepared by mixing ethanol (50 ml) and water (50 ml) (Merck Millipore) in a 1:1 v/v ratio. Then, 100 ml of solution B was added dropwise into solution A at 80 °C and continuously stirred for 60 min. After that, 10 ml of boric acid solution (1% w/v) in water was added dropwise to the mixture of solution A and B. The solution was further stirred for 30 min at 80 °C. The resulting milky suspension was kept aside for aging, for 24 h. After aging, the white precipitate was settled down at the bottom, which was dried in an oven at 120 °C for 12 h (or till complete dry). Finally, the dried material was ground and then calcined at 400 °C in air for 4 h using a heating rate of 3 °C/min. The so-obtained boron-doped TiO2 was ascribed as BT1 in the following experiments. Two more B-TiO2 samples were prepared similarly, by adding 5% and 7% w/v solutions of boric acid, and were denoted as BT5 and BT7, respectively. The TiO2 without doping was prepared in a similar manner but without adding boric acid and it is referred as undoped TiO2.

Characterization of photocatalysts

The structural characterization of prepared catalysts was carried out using powder X-ray diffraction instrument (XRD- Philips X’Pert diffractometer) equipped with Cu-Kα radiation, operating in 2θ mode. The samples were analyzed in the range of 20°–80°. The Raman spectra of prepared samples were recorded using a Jobin Yvon Lab Ram HR 800 setup equipped with a 488 nm Ar ion laser excitation line. The band gap of prepared samples was calculated with UV-Vis absorption spectroscopic measurements, and carried out on a UV-Vis diffuse reflectance spectrophotometer (Shimadzu UV-2550) in the range of 200–800 nm, where the absorption spectra were referenced to BaSO4. The Brunauer-Emmett-Teller (BET) surface area of the samples (degassed for 60 min at 150 °C) was determined from low-temperature nitrogen adsorption-desorption isotherms using Micromeritics instrument (ASAP-2020 model). The pore size distribution (pore diameter and pore volume of samples) was determined by the Barrett-Joyner-Halenda (BJH) method. Fourier-transform infrared (FTIR) spectra of the samples were recorded on a spectrometer WQF-410 (Bruker), where the sample powders were mixed with KBr to prepare pellets and then spectra were recorded in the wavenumber range of 4000–400 cm−1. The morphology of the samples was observed using scanning electron microscopy (SEM-FEI Quanta 200FSEM, Hillsboro), that was equipped with energy-dispersive X-ray (EDX). The internal structure and d-spacing of the samples were analyzed out using high-resolution transmission electron microscopy (TEM- JEOL TEM Model 2100). For TEM analysis, the powdered samples are dispersed in isopropanol solvent for homogeneous dispersion and then the copper grids are dipped into it and dried. Then, the grids are placed on the specimen holder, which is to be loaded inside the TEM instrument for imaging. The chemical states of prepared materials were examined using an X-ray photoelectron spectrometer (XPS-Thermo Scientific Theta Probe Escalab 250, a monochromatic Al Kα X-ray source). Detailed scans were recorded for B1s, O1s, and Ti2p and all the binding energies were referenced to the C1s peak (284.6 eV).

Photocatalytic degradation studies

The photocatalytic efficiency of doped and undoped photocatalyst was evaluated by degradation of 4-NP under UV light irradiation. The photocatalysis experiments were performed in a cylindrical batch reactor (500 mL), made up of quartz glass (Lelesil Innovative Systems Ltd. Mumbai, India). The reactor was equipped with a water jacket to control the temperature at about 20 °C. The photo-irradiation was performed with a 125-W mercury lamp, emitting a wavelength of 354 nm as the UV light source.

In every experiment, before irradiation, 150 mL of 4-NP aqueous solution of variable concentration, containing the required amount of catalyst, was stirred continuously in the dark for 30 min to reach adsorption equilibrium. In order to determine the adsorption capacity of the catalyst, about 5 ml of sample was withdrawn at that time by syringe for analysis. Subsequently, the reaction mixture was continuously stirred under UV light, and 5 ml of aliquots was withdrawn at fixed intervals of time up to 3 h. The aliquots were filtered through a Millipore syringe filter (pore size 0.45 μm) to remove the photocatalyst. The concentration of remaining 4-NP in filtered solution was determined by measuring the aborbance using UV-Vis spectrophotometer (Thermo Scientific, Evolution 201) at λmax = 316 nm. Throughout this photocatalytic study, the experiments were carried out in duplicates, and wherever required, the experiments were repeated up to three times. The uncertainty of the experiments as estimated by the repeatability of the experiments was found below ± 2% and had no significance influence on the interpretation of the results.

Results and discussion

Structural characterization

X-ray diffraction (XRD) technique was used to analyze the alteration in crystallographic phases, before and after boron doping. The XRD pattern of the samples is presented in Fig. 1a. The average crystallite sizes of the samples were determined using following Scherrer equation (Scherrer 1918):

$$ \mathrm{D}=\frac{K\lambda}{\beta Cos\theta} $$
(1)
Fig. 1
figure 1

a XRD pattern of undoped TiO2 and B-doped TiO2 samples. b Raman spectra of undoped TiO2 and B-doped-TiO2 samples. c FT-IR spectra of undoped TiO2 and B-doped-TiO2 samples

Full size image

Where, K = 0.89 is shape factor, λ is the wavelength of X-ray irradiation, Cu-Kα of wavelength 1.54 Ǻ, β is the full-width half-maximum (FWHM) of the diffraction peak, and θ is the X-ray diffraction angle. The average crystallite sizes of the samples are given in Table 1.

Table 1 Surface properties and XRD parameters of undoped TiO2 and B-TiO2
Full size table

The XRD pattern of undoped TiO2 shows diffraction peaks at 25.42°, 37.76°, 48.28°, 54.7°, and 62.97° (PDF 86-1157) that correspond to (101), (004), (200), (105), and (204) planes of anatase TiO2. The XRD pattern of BT1 sample also shows similar peaks, which confirmed that it contains only anatase phase. However, the XRD pattern of BT5 and BT7 sample (doped with 5% and 7% boric acid, respectively) exhibits a peak at 28.11° (PDF 86-0147), which corresponds to (110) plane of rutile phase of TiO2. This observation is consistent with the previously published report, where rutile phase was observed on increasing the concentration of boron dopant (Chen et al. 2006). Furthermore, it has been reported that rutile content increases when there is an increase in the average crystallite size of the boron-doped samples (Table 1). The appearance of the rutile phase at a higher concentration of boron can be explained due to the structural rearrangement of the catalyst. The anatase to rutile rearrangement involves breaking of two of the six Ti-O bonds in anatase phase (Linsebigler et al. 1995). During the drying process, boron causes the Ti-O-Ti network to weaken, and thus facilitates the Ti-O bond cleavage. Consequently, structural rearrangement may lead to the formation of the rutile phase.

In BT7, a characteristic peak of titanium borate (TiBO3) at 54.08° (PDF 85-0168) was observed, which corresponds to (018) plane. The increase in intensity and sharpening of Bragg peak (101) in BT7 is due to increase in crystal size and improved crystallinity. These results indicate that crystallinity increases due to boron doping. Also, the anatase peak of (101) plane of BT7 shifted to towards lower angle (from 25.50° in TiO2 to 25.45° in BT7), and thus leads to increase in d-spacing value (Table 1). This suggests that the structure has been modified due to intercalation of boron into the TiO2 crystal lattice. By doping, the boron may either replace an oxygen atom or occupy an interstitial position in the TiO2 matrix. The probability that a dopant will substitute oxygen atom depends on its electronegativity. Here, the electronegativity of B (2.04) is much smaller than O (3.44); hence, the possibility of substitution of O with B is ruled out. Likewise, the possibility of substitution of Ti4+ with B3+ is also very low, because the radius of B3+ (0.023 nm) is much smaller than that of Ti4+ (0.068 nm). Therefore, it would be difficult for B3+ to replace the Ti4+ from its site (Chen et al. 2006). Thus, it indicates that boron may occupy the interstitial sites in TiO2 crystal lattice, thereby modifying the crystal lattice, and hence, variation in d-spacing is observed.

Micro-Raman spectroscopy is one of the most powerful techniques used to investigate the crystalline phase transformations, lattice distortions, oxygen defects, and also the presence of any defect states associated with boron doping into the TiO2 crystal lattice. Figure 1b shows the micro-Raman spectra recorded in the range of 100–1000 cm−1. The pure TiO2 (anatase) exhibited six Raman-active modes at wavenumbers 142.5 (Eg), 196 (Eg), 396 (B1g), 515 (B1g + A1g), and 640 (Eg) cm−1, respectively (Zhang et al. 2000). It is seen from the Fig. 1b that the Eg band of undoped TiO2 appeared at 146.2 cm−1. However, the Eg mode is blue shifted (148.3 cm−1) after boron doping at 1% and 5%. This shift after B-doping may be due to the stress induced in the TiO2 structure upon occupation of interstitial sites by boron, as explained in the XRD studies. However, at 7% boron doping, the peak has again shifted towards lower value (146.2 cm−1). This change in the peak shift may be because of release of stress on increasing the dopant concentration. This indicates that notable modification in structure occurs after the B doping. A blue shift in the anatase Eg peak can be attributed to phonon confinement and oxygen deficiencies (Kafizas and Parkin 2011). Phonon confinement can be caused by several factors, including an expansion in the unit cell volume caused by the introduction of a dopant. Thus, doping results in modification in TiO2 lattice, which causes broadening of most intense peak in BT7 (inset of Fig. 1b), may contribute to increase in crystallinity. Further, the line shape broadening can be due to tensile stress or defects arise due to doping.

Figure 1c shows the FT-IR spectra of undoped and boron-doped TiO2. The broad absorption peaks at 3100–3600 cm-1 are assigned to stretching vibrations of molecular water and OH groups (Wood et al. 1983). Bands for Ti-O and Ti-O-Ti bonds are observed in the 800–400 cm-1 region (Sigaev et al. 2001). The vibrational modes of the borate network are mainly due to the asymmetric stretching relaxation of the B-O bonds of trigonal BO3 units. Such type of vibrational modes occurs in the range 1200–1750 cm-1 with doped samples only (Sharma et al. 2006). Peaks at 1620–1630 cm-1 are due to vibrations of hydroxyl groups. In the IR spectra, another peak appears at 1384 cm−1, which can be ascribed to the vibration of tri-coordinated boron (Feng et al. 2011).

HRTEM analysis was performed on undoped TiO2 and B-TiO2 (BT7), shown in Fig. 2. Clear high-resolution lattice fringes (Fig. 2a) with a d-spacing of 0.357 nm can be ascribed to the most thermally stable and major anatase (101) facets in undoped TiO2. Whereas, in BT7 (Fig. 2c), a small increase in d-spacing (0.366 nm) is observed, indicating the interstitial incorporation of boron in the TiO2 lattice, and hence modification in the structure of TiO2. TEM images show small pore mesoporous materials with uniform rotundity. After boron doping, both the morphology and crystal structure of TiO2 are well retained, but the crystallinity is further improved compared to undoped TiO2. The observations of HRTEM analysis are in agreement with the XRD and Raman data.

Fig. 2
figure 2

a TEM image of TiO2 (high resolution) and low resolution [inset]. b The distance between the lattice fringes (red parallel lines) pointed by the red arrows is 0.357 nm for TiO2. c TEM images of BT7 (high resolution) and low resolution [inset]. d The distance between the lattice fringes (red parallel lines) pointed by the red arrows is 0.366 nm for BT7

Full size image

Photo-physical properties

The photocatalytic activity of semiconductor material is related to its band gap (Eg). The Eg of prepared materials was evaluated with UV-Vis diffuse reflectance spectroscopy. Figure 3 shows the UV-Vis absorption spectra of the undoped TiO2 and boron-doped TiO2 samples. The band gap energy was calculated using following Tauc’s relation (Tauc 1970):

$$ {\left(\alpha h\nu \right)}^{\frac{1}{2}}=\mathrm{A}\ \left( h\nu \hbox{--} {\mathrm{E}}_{\mathrm{g}}\right) $$
(2)
Fig. 3
figure 3

a UV-Vis absorption spectra. b Tauc’s plot of undoped TiO2 and B-doped-TiO2 samples

Full size image

Where α is absorption coefficient, A is a constant, hν is photon energy, and Eg is the optical band gap energy of the material. Figure 1d illustrates the plot of (αhν)1/2 versus photon energy (hν) for pure and boron-doped TiO2 samples. The Eg values were obtained by extrapolating the linear portion to the photon energy axis. The optical band gap energies of 2.98 eV, 2.98 eV, 2.96 eV, and 2.95 eV were obtained for pure TiO2, BT1, BT5, and BT7, respectively. These band gap energies were lower than those of commercial TiO2, which is P25 (3.2 eV). The optical absorption edge of B-doped TiO2 (BT5 and BT7) is significantly shifted towards visible region when compared to undoped TiO2. However, no significant differences were found in UV-Vis absorption spectra of BT1 (1 % B-doped TiO2) as compared to undoped TiO2. The band-gap values decrease in BT5 and BT7 samples due to the incorporation of B3+ ions into TiO2 crystal lattice and also because of sp-d exchange interactions (Singh et al. 2009). These results suggest that low B content is not sufficient to decrease band gap, however, 5% or higher B content is required to observe optical changes.

Surface/interface properties

The results of surface properties like specific surface area, pore volume, and average pore width of undoped and boron-doped TiO2 are summarized in Table 1. The specific surface area and pore volume of TiO2 decreased with increase in boron content. This may be due to the structure modification and partial pore blockage, caused by boron loading. XRD results indicated that boron occupies the interstitial sites in TiO2 lattice, and hence, there was an increase in particle size. Moreover, thus, the surface area decreases with increase in particle size (Table 1). These results are in good agreement with observations, as reported by Stengl et al. (2010).

The low-temperature N2 adsorption-desorption isotherm of undoped and boron-doped TiO2 is shown in Fig. 4. The samples exhibited a type IV isotherm with an H1-type hysteresis loop associated with open-ended cylindrical pores (Sing et al. 1985). This type of isotherm is a characteristic of mesoporous materials. The inset in Fig. 4 shows the pore size distribution of all the materials. The pore size distribution of BT1 and BT5 shows the presence of mesopores with pore diameter in the range of 2–12 nm, whereas the pore size distributions of TiO2 and BT7 samples are broad in range (2–15 nm).

Fig. 4
figure 4

Low-temperature N2 adsorption-desorption isotherms and pore size distribution (inset) of undoped TiO2 and B-doped-TiO2 samples

Full size image

The SEM micrographs of TiO2 and B-doped samples are shown in supporting information as Fig. SI 1. The SEM images showed changes in morphology of TiO2 after incorporation of boron. For instance, particles of regular size and shape were observed for undoped TiO2, whereas boron-doped TiO2 consists of particles of irregular shape and size with a fairly rough surface. In doped samples, the surface with a slightly whitish portion was observed which indicate the deposition of boron dopant. The EDX spectra of microstructures were recorded during SEM imaging and are shown in Fig. 5. It gave both qualitative and quantitative information about the elemental and atomic percentages in undoped TiO2 and B-doped TiO2 samples, as presented in Table 2. From EDX results, it is inferred that carbon atom also exists as a dopant in boron-doped sample. It may be due to the reaction of boric acid with carbon present in titanium isopropoxide, forming a complex carbonate species. The EDX results established in situ incorporation of B in TiO2 lattice with the desired percentage.

Fig. 5
figure 5

EDX spectra and SEM images (inset) of (a) TiO2, (b) BT1, (c) BT5, and (d) BT7

Full size image
Table 2 Catalyst composition TiO2 and B-TiO2 catalyst using EDX
Full size table

To investigate whether these boron atoms doped on the TiO2 surface, or elsewhere in the lattice structure, the samples were analyzed by XPS, which is a highly surface sensitive technique. The study revealed that B-TiO2 contains only Ti, O, B, and C elements. The C element can be ascribed to the residual carbon from the precursor solution (TTIP) and the adventitious hydrocarbon from the XPS instrument itself. The binding energies of B, Ti, and O were referenced to the C1s peak (284.6 eV). Figure 6a shows the XPS spectra of the B1s of all doped samples. Usually, B1s electron binding energy peak lies around 188–194 eV. The binding energy (BE) associated with the B1s signal depends on the extent of incorporation of boron atom on TiO2 surface (In et al. 2007; Gombac et al. 2007). Boron in TiB2 has BE of 187.5 eV (B–Ti bonds) and boron in B2O3 or in H3BO3 has BE of 193.3 eV (B–O bonds) (Zhu et al. 2007). The binding energy of B1s associated with B substituting O lies around 190.6 eV (Gombac et al. 2007). The peak at around 192 eV corresponds to B1s forming Ti–O–B linkage (Zaleska et al. 2009; Chen et al. 2006; Zhu et al. 2007). Figure 6b shows XPS spectra of Ti2p. The two major peaks at 457.54 eV (Ti 2p3/2) and 463.2 eV (Ti2p1/2) are observed for pure TiO2. The standard binding energy of Ti2p3/2 in TiO2 for Ti3+ is usually located at 457.7 eV, and that for Ti4+ is at 459.5 eV (Zhang et al. 2007). XPS peak at 457.54 eV can be ascribed to the formation of Ti3+ species, which is shifted to 457.67 eV in boron-doped TiO2. It indicates that the boron doping favors the formation of surface-layer Ti3+ species. Generally, the XPS peak of Ti2p is not broad, and no shoulder peak is observed. However, the TiO2 surface containing Ti3+ results into much broader peak, as seen in doped samples (Fig. 6) (Xiong et al. 2012). The XPS results suggested that B is incorporated at interstitial position in TiO2, facilitating the formation of Ti3+. Additionally, the existence of Ti3+ may result in decrease of recombination of charge carriers (electrons and holes), which influence photocatalytic properties of samples.

Fig. 6
figure 6

XPS pattern of (a) B1s peaks of B-TiO2, (b) Ti2p peaks of TiO2, and B-TiO2. c O1s peaks of TiO2 and B-TiO2

Full size image

Figure 6c shows XPS spectra of O1s. The main O1s component (at around 528 eV) can be ascribed to titanium dioxide (Ti-O), and the second component at around 530 eV which could be due to surface hydroxyl groups (-OH) (Chen et al. 2011). However, the broad O1s region of BT7 can be fitted by three peaks, which are Ti-O in TiO2, B-O bond, and hydroxyl groups respectively, which further confirms the presence of B-O bond (Chen et al. 2006). The decrease in optical band gap energies with increase in B loading could be due to formation of Ti-O-B-type linkage. Thus, the surface or interfacial properties confirmed the presence of boron into TiO2 lattice by forming a mesoporous type of nanostructures.

Photocatalytic studies

The effect of catalyst dose on photocatalytic degradation (after 3 h) of 4-NP was investigated to optimize the amount of catalyst. The dose was varied from 0.1 to 1.5 g/L, and the effect of a change in dose on degradation is presented in Fig. 7a. All the prepared materials showed a similar pattern of change in degradation with dose variation. As the concentration of the catalyst in solution was increased, the availability of adsorption sites or active sites was also increased. As a result, the adsorption of 4-NP molecules on the catalyst surface increased and consequently its degradation also get increased from 0.1 to 1.0 g/L of dose.

Fig. 7
figure 7

a Effect of catalyst dose on photocatalytic degradation of 4-NP (conc. of 4-NP = 5 mg/L). b Effect of boron loading on photocatalytic degradation of 4-NP (conc. of 4-NP = 1 mg/L). c Effect of the initial concentration of 4-NP on photodegradation with catalyst dose 1.0 g/L

Full size image

Further increase in dose did not improve the degradation which is possibly due to following two reasons: (1) aggregation of catalyst particles (Lea and Adesina 2001) and (2) increase in opacity of the solution (Gautam et al. 2006). These observations suggest that the 1.0 g/L dose of the catalyst is optimum dose for degradation of 4-NP. All further experiments were carried out at this dose for the optimization of other experimental variables.

As compared to pure TiO2, the B-doped TiO2 catalysts exhibited a significant increase in 4-NP photodegradation efficiency. Figure 7b shows a variation of degradation (C/C0) as a function of irradiation time in min. It can be observed that the presence of boron species positively influences the photocatalytic activity of TiO2. All the doped samples showed to be more photoactive than undoped TiO2, where BT7 catalyst, having maximum boron content, showed the highest degradation of a solution containing 15 mg/L of 4-NP.

The pollutant concentration is a critical parameter in any water treatment process, so photocatalytic degradation kinetics was investigated by varying the initial concentration of 4-NP in the range of 1.0 to 15 mg/L. Figure 7c shows the effect of initial concentration of 4-NP solutions on the photocatalytic degradation of 4-NP after irradiation for a fixed duration (180 min). It is evident that with fixed catalyst dose, the increase in initial concentration leads to a gradual decrease in degradation efficiency (C/C0). It is likely due to the increase in pollutant to catalyst active site ratio. The increase in efficiency was observed after 10 mg/L of initial concentration, particularly with B-doped samples. This observation points towards the possibility of less soluble 4-NP molecules at high concentration. These less soluble molecules get localize on the catalyst surface and get degraded. It is one of the possible reason for an increase in PCD of 4-NP at higher concentration. This phenomenon was not observed in case of as-synthesized TiO2 (Fig. 7c), which indicates that only B-doped sites encourage the interaction of 4-NP to the surface, and hence, the PCD of TiO2 is low even at high initial concentration.

Modeling of photocatalytic data

Photocatalytic degradation of 4-NP in the presence of pure and doped TiO2 is shown in Fig. 8a. Before illumination, the amount of 4-NP adsorbed on the surface of pure TiO2 is significantly higher than that of B-TiO2, which is due to higher surface area of pure TiO2. However, after illumination, the B-TiO2 (BT7) degraded 90% of 4-NP as compared to 79% degradation by undoped TiO2. When a small amount of boron (1%) is doped into TiO2, the photocatalytic activity increases slightly (80%). On further increase in boron content, the activity of BT5 sample also increases (85%) as compared to BT1 and undoped TiO2. BT7 sample with the highest boron content showed the maximum photocatalytic degradation of 4-NP.

Fig. 8
figure 8

a Photocatalytic degradation of 4-NP in presence of undoped and B-TiO2. b First-order rate plot of photocatalytic degradation of 4-NP under UV irradiation in presence of undoped and B-TiO2 (catalyst dose 1.0 g/L, 4-NP conc. 1 mg/L)

Full size image

To further explore the kinetics of PCD, first-order rate equation was used, which is as follows:

$$ In\left(\frac{C_0}{C}\right)= kt $$
(3)

Where, C0 represents the initial 4-NP concentration in mg/L, C represents remaining 4-NP concentration in mg/L after time t (min) of irradiation, and k is first order rate constant (min-1).

The values of K were obtained from the slope of the graph of In(C0/C) versus time plot. The BT7 sample showed better photocatalytic activity as compared to all the other samples. Figure 8b gives a comparison of PCD kinetics with all the four samples. The steepness of slope is a measure of rate constant, and hence the photocatalytic activity. It is evident that doping caused an increase in photocatalytic activity with a higher k value. The k values obtained from Fig. 8b are 0.006 min-1,0.0109 min-1, 0.0133 min-1, and 0.0322 min-1 for TiO2, BT1, BT5, and BT7, respectively. The post doping increase in PCD is probably due to several factors which are discussed as follows.

Firstly, boron incorporation as a dopant into TiO2 crystal lattice occupying interstitial sites donates its three valence electrons to the 3d state of titanium ions, according to the process (Finazzi et al. 2009):

$$ B+3{Ti}^{4+}\to {B}^{3+}+3{Ti}^{3+} $$

The existence of a reduced form of Ti3+ on the surface of TiO2 (as evidence from XPS) is responsible for enhancement in photocatalytic activity. It might act as an active site to assist the adsorption of reactant and trap the photogenerated charge carriers (electrons and holes), and therefore may reduce their recombination. Secondly, it is well known that the specific surface area and the anatase phase plays a significant role in the photocatalytic activity of TiO2 (Zhou et al. 2005). A large surface area plays an important role in any photocatalytic degradation reactions, as a large number of organic molecules may get adsorbed on the catalyst surface, and thus promotes the reaction rate (Kim et al. 2009). However, powders with the large surface area are usually associated with more crystalline defects, which favor the possibility of recombination of photogenerated charge carriers, leading to a poor photocatalytic activity (Carp et al. 2004). Therefore, the sufficient surface area is a requirement, but not a determinant factor for a photocatalyst. The surface area of boron-doped samples is less than that of pure TiO2.

The decrease in the specific surface area and pore diameters due to the presence of boron was reported in the literature (Stengl et al. 2010). The BET surface area tends to decrease with increase in boron content, which indicated the presence of mesopores as evidenced by the pore size distribution. The photocatalytic activity of amorphous TiO2 is negligible, also indicates that crystallinity is another important requirement (Zhou et al. 2005; Anpo et al. 1987). So, a balance between specific surface area and crystallinity is a very important factor in determining the photocatalytic activity of the TiO2 material. The XRD, Raman, and BET results suggest that BT7 possess good anatase phase, crystallinity, and optimum surface area, which results in its better photocatalytic activity. In addition, narrowing of the band gap (2.95 eV) compared to undoped TiO2 (2.98 eV) caused increase in the photocatalytic activity of BT7. Also, the existence of Ti-O-B linkage, evidenced from XPS, has synergistic effect on the photocatalytic activity of the B-doped TiO2 (BT7). The photodegradation results of present study were compared with results of previous studies involving photo-catalytic degradation of 4-NP using different catalyst or conditions. Table 3 shows this comparison on the parameters like % degradation, band-gap, photo catalyst, and radiation type, concentration, and rate constants. It is evident, from this comparison, that photocatalytic efficiency of B-doped photo catalyst is comparable/and even better than some of the co-doped TiO2 catalyst. The rate constant of B-doped TiO2 based 4-NP degradation is reasonably high as compared to other photo catalysts.

Table 3 Comparison of prepared TiO2 and B-TiO2 catalyst with other photocatalysts for degradation of 4-NP under different conditions
Full size table

Mechanism of PCD of 4-NP

Figure 9 shows the absorption spectral changes as a function of time during photocatalytic degradation of 4-NP, in the presence of B-TiO2. The changes in absorption peaks of 4-NP were monitored with UV-Vis spectrophotometer. The UV-Vis spectra of reaction mixtures of 4-NP at different reaction times show observable changes as the photocatalytic degradation proceeds. Before degradation, 4-NP shows a sharp characteristic absorption band at λmax 316 nm, as evidenced in spectra of initial sample taken. After short irradiation times (15–30 min), this absorption band gets decreased in intensity. Then, after prolonged irradiation (120–180 min), the absorption band at 316 nm disappears along with emergence of two new bands at λmax 275 nm and 290 nm. The appearance of these bands is most likely due to the formation of intermediate degradation products, hydroquinone (λmax 290 nm) and 1,2,4-trihydroxy benzene (λmax 275 nm) (Edgar et al. 2012). In literature, previous studies also reported these degradation products as oxidation intermediates of 4-NP (Nevim et al. 2002; Won-Young et al. 2007). The photocatalytic reactions involve generation of .OH radicals, which are electrophilic in nature. The nitro group in 4-NP is a very good leaving group and can be eliminated easily by an electrophilic substitution of the .OH radicals at the para position with respect to the hydroxyl group. It further forms, hydroxylated products (hydroquinone and 1,2,4-trihydroxy benzene) by following two different pathways, through hydrogen abstraction and subsequent addition or substitution to the formed radicals as shown in Scheme 1. On the basis of absorption band spectra, it is expected that the degradation of 4-NP is following first pathway. Further, oxidation of degradation products leads to complete mineralization into carbon dioxide and water molecule (Di Paola et al. 2003).

Fig. 9
figure 9

Photocatalytic degradation of 4-NP monitored by UV-Vis spectroscopy in presence of B-TiO2

Full size image
Scheme 1
scheme 1

Expected chemical oxidation pathway of 4-NP degradation during illumination in presence of B-TiO2 [(i) benzoquinone, (ii) 4-nitro catechol and (iii) 1,2,4-trihydroxy benzene]

Full size image

Conclusions

Here, a single step, facile sol-gel synthesis of undoped and B-doped TiO2 nanostructures with various B loadings (1, 5, 7% w/v) is reported. XRD and HRTEM results showed that B doping has improved crystallinity, and doped catalysts contained a mixture of anatase and rutile phase. The 5% B (BT5) or higher amount (BT7) caused significant changes in the structure as evidenced by increase in average crystallite size and average pore width; however, not much appreciable change was observed with low B (1%) content. In addition, the band gap was decreased to 2.96 eV and 2.95 eV for BT5 and BT7, respectively as compared to undoped TiO2 (2.98 eV). XPS studies revealed that boron occupies interstitial sites in TiO2 crystal lattice by forming a Ti-O-B type linkage. Photocatalytic degradation of 4-NP with various B loadings was monitored and catalyst with highest B content (BT7) exhibited maximum degradation efficiency (90%) as compared to undoped TiO2 (79%). The formation of Ti-O-B type bond has synergistic effect on enhancing the photocatalytic activity. Thus, these results have clearly demonstrated that the photocatalytic properties of B-TiO2 catalysts depend on the amount of B loading and it is a deciding factor in determination of structural, interfacial, and photo-physical properties. The B-TiO2 catalysts could be a potential material for decomposition of organic contaminants from industrial wastewater samples.