Effect of Fe-N Codoping on the Optical Properties of TiO2 for Use in Photoelectrolysis of Water.

TiO2 nanoparticles were synthesized by green chemistry where organic solvents are replaced by an aqueous extract solution of lemongrass leaves that act as a reducer and growth-stopper agent. The nanoparticles were codoped with N-Fe to modify the absorption range in the electromagnetic spectrum and were characterized by Fourier-transform infrared (FTIR), scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDS), and UV-vis/diffuse reflectance spectroscopy (DRS). The modified samples with Fe and N resulted in smaller nanoparticle size values than pure TiO2. Similarly, the band-gap energy for doped nanoparticles decreased to 2.22 eV in relation to the value of 3.09 eV for pure TiO2, due to the introduction of new energy levels.

Introduction

Titanium dioxide (TiO2) is characterized by being a photosensitive semiconductor, having good optical and electrochemical properties,[1] good dispersibility in organic solutions, and low toxicity.[2] These properties have led to numerous investigations directed to applications such as the removal of contaminants by photocatalysis[3−5] and photoelectrochemical devices for hydrogen generation.[6,7] TiO2 can present several different phases in the nanometric range at different temperatures, which are anatase, brookite, and rutile, though anatase has excellent physical and chemical properties for environmental remediation.[8] However, the successful application of TiO2 is still limited by its band gap energy[9] because the photoinduced reactions in TiO2 are restricted to the UV region, which comprises only 4% of the solar spectrum.[10] Therefore, recent research aims to improve the optical and morphological properties of TiO2, by codoping with different metal ion oxides in conjunction with nonmetals. This is due to the low rate of charge carrier recombination and the highly visible photocatalytic yield resulting from the synergistic effect of the codopant elements, compared to the results of doping with a single element.[11−14] Accordingly, in this work, TiO2 codoped with Fe–N is prepared; the ionic radius of the N atom is close to the O atom in the TiO2 lattice, resulting in the fusion of the N 2p orbital with the O 2p states, modifying the electronic structure of the valence band to easily transport load carriers.[15] Therefore, the simultaneous use of a metal and a nonmetal as codoping elements can be an effective modification. The aim of this research is to improve the optical and morphological properties of TiO2 nanoparticles by codoping iron and nitrogen for hydrogen generation using a photoelectrochemical cell. Anatase phase TiO2 nanoparticles were synthesized from a green synthesis mechanism by lemongrass extract; samples were prepared at different concentrations of Fe3+ and N, which have been characterized by scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDS), UV–vis, and Fourier-transform infrared (FTIR) spectroscopy.

Results and Discussion

FTIR Spectra Analysis

FT-IR spectra analysis has been carried out to determine the functional groups present in the prepared samples (Figure ). A strong absorption band is observed at the spectra of codoped TiO2 (Figure b–d) between 3200 and 3400 cm–1, corresponding to the stretching vibration mode of the hydroxyl bond (−OH), and the absorption band located around 1630 cm–1 can be assigned to the mode of bending of −OH vibration of absorbed water molecules in synthesized nanoparticles.[20] At low frequencies, a descent band in the range of 500–880 cm–1 has also been determined in all samples, corresponding to the Ti–O–Ti bond, indicating the formation of TiO2.[21][21] These patterns of vibrations are nearly similar to those identified in the unmodified TiO2 sample, which are shown in Figure a. However, for codoped samples, in addition to the vibrations mentioned above, a peak in the wavenumber of 1082 cm–1 is observed, which confirms the presence of a substituted N atom in the TiO2 lattice, corresponding to the Ti–N vibration.[22,23] Also at low frequencies, a peak in 500–800 cm–1 can be attributed to the symmetric Fe–O–Fe stretching vibration.[24] The bond vibrations of the samples are in accordance with that reported in literature. In addition, for all codoped samples, characteristic peaks were observed at 500–880, 500–800, and 1082 cm–1, confirming the presence of TiO2, Fe, and N, respectively. The addition of Fe in the TiO2 matrix results in changes that lead to the absorption of more amounts of OH groups.

Figure 1

FT-IR spectra of (a) pure TiO2 nanoparticles and Fe-doped TiO2 nanoparticles at (b) 10%w/w N (c) 20%w/w N, and (d) 30%w/w N.

SEM/EDS Analysis

The morphology of pure and Fe–N codoped TiO2 nanoparticles has been determined through SEM images. Figure shows the surface of the synthesized nanoparticles; a nonuniform distribution is observed. In addition, there are agglomerations in some regions, which can be attributed to the calcination treatment to which the nanoparticles were subjected.[25]

Figure 2

SEM images of pure TiO2 and codoped TiO2. (a) Pure TiO2, (b) 1% Fe–10% N, (c) 0.7% Fe–10% N, (d) 0.5% Fe–10% N, (e) 1% Fe–20% N, (f) 0.7% Fe–20% N, (g) 0.5% Fe–20% N, (h) 1% Fe–30% N, (i) 0.7% Fe–30% N, and (j) 0.5% Fe–30% N.

Furthermore, using ImageJ Software, it has been found that with Fe–N codoping, the size of the nanoparticles is in the range of 37–58 nm, as shown in Table , which is lower than the particle size of unmodified TiO2 nanoparticles (70 nm),[18] whose SEM image is shown in Figure a. The decrease in particle size suggests that the codoping caused alterations in the structure of TiO2, since the growth of the particle size is obstructed, which can be associated with the incorporation of Fe3+ ions in the crystal structure of TiO2 due to differences in the atomic radius of Fe3+ and Ti4+, as determined in the investigations of Othman and co-workers.[26,27] However, a significant trend of increasing concentration in nanoparticle size is not observed; the particle sizes reported by Realpe Jimenez et al.[18] were smaller when they were doped with Cu. The analysis shows that codoping radically affects the size of TiO2 nanoparticles when is compared to the nondoped sample.

Table 1

Particle Size of Synthesized Samples

	pure TiO2 and codoped TiO2
	pure TiO2	1% Fe–10% N	0.7% Fe–10% N	0.5% Fe–10% N	1% Fe–20% N	0.7% Fe–20% N	0.5% Fe–20% N	1% Fe–30% N	0.7% Fe–30% N	0.5% Fe–30% N
size (nm)	70	52	44	54	58	38	56	37	38	41

Peaks corresponding to O, Cl, Ti, and Fe have been found with the elemental chemical analysis EDS, as shown in Figure , indicating the formation of TiO2.

Figure 3

EDS spectra of TiO2 nanoparticles codoped at 1% Fe–10% N.

Although the presence of Cl corresponds to ammonium chloride (the nitrogen precursor), no peaks have been detected for N, due to the detection limit of the EDS analysis for nitrogen, since there are interferences of lines of lighter elements superimposed with heavier elements.[28,29] The presence of Na, Mg, K, Ca, and V has also been identified; these lines are attributed to impurities or the equipment used.[25]

Table shows the percentage by mass of iron, over the total sample, measured by EDS analysis, which indicates that Fe is incorporated into the TiO2 support. A similar variation of the Fe concentration values given by EDS was observed by Kashale et al.[30]

Table 2

Percentage of Fe in Each Synthesized Sample Determined by EDS Analysis at Different Percentages of N and Compared to the Amount of Fe Initially Added

	amount of Fe measured by EDS at different N concentrations
amount of Fe initially added	10% N	20% N	30% N
1% Fe	0.9% Fe	1.2% Fe	1.21% Fe
0.7% Fe	0.94% Fe	0.9% Fe	0.83% Fe
0.5% Fe	0.61% Fe	0.33% Fe	0.66% Fe

UV–Vis Diffuse Reflectance Spectroscopy (UV–Vis/DRS)

Figure shows the optical properties of diffuse reflectance for pure TiO2 and codoped with N and Fe TiO2 in a wavelength range from 200 to 800 nm. For unmodified TiO2, a wide absorbance band for wavelengths lower than 400 nm can be observed, which indicates that its range of photoactivity is limited to the UV region of the spectrum. However, for codoped samples, there is a shift of the absorption band toward wavelengths greater than 400 nm, and this shift increases with increasing Fe3+ concentration while N keeping constant (Figure k). Therefore, the valence band of the modified samples can be excited with photons of lower energy. On the contrary, absorbance decreases with increasing N concentration keeping Fe constant.

Figure 4

UV–vis diffuse reflectance spectroscopy (UV–vis/DRS) for the unmodified TiO2 sample and TiO2 modified with Fe3+ and N. (a) Pure TiO2, (b) 1% Fe–10% N, (c) 0.7% Fe–10% N, (d) 0.5% Fe–10% N, (e) 1% Fe–20% N, (f) 0.7% Fe–20% N, (g) 0.5% Fe–20% N, (h) 1% Fe–30% N, (i) 0.7% Fe–30% N, (j) 0.5% Fe–30% N, and (k) N–Fe codoped TiO2 at different Fe concentrations keeping 10% N to easily observe the change in absorbance.

The band gaps of the modified samples were determined through the Tauc graphical method to analyze the optical properties of the nanoparticles, as shown in Figure using eq .where α is the absorption coefficient, A is constant, hv is the photon energy, Eg is the band gap, and n denotes the nature of the electronic transition interband. The variable n can have the values 1/2, 2, 3/2, and 3 corresponding to direct allowed, indirect allowed, direct forbidden, and indirect forbidden transitions, respectively. In this case, n = 2 for the indirect transition allowed to graph (αhv)1/2 vs hv.[31,32]Figure shows the extrapolation of the linear part on the energy axis, obtaining the band gap of the synthesized samples.

Figure 5

Band-gap TiO2 samples modified with 1Fe–10N using the Tauc method.

Figure shows the band gap for modified TiO2 nanoparticles; these results were lower compared to the unmodified TiO2 (3.09 eV). It is noted that when the percentage of Fe is kept constant, the band gap increases as the N concentration increased. On the other hand, when nitrogen percentage is kept constant, the band gap decreases as Fe percentage increased, which is consistent with the results found by Realpe Jimenez et al.[17] They worked with equal percentages (1, 0.7, and 0.5% w/w) of Fe, but performed only doping with Fe, concluding that the band gap decreases as the percentage of iron increases and obtaining their lowest band gap of 2.66 eV for 1.0% w/w Fe–TiO2. However, in the current work, the lowest band gap of 2.22 eV was also found for TiO2 codoping with 1.0% w/w Fe and 10% w/w N. This result supports the positive effect of doping with N. Furthermore, this entails that with doping there is a modification in the electronic structure of TiO2, so that additional electronic states can be provided through Fe within the TiO2 band gap.[33]

Figure 6

Band gap of the TiO2 samples at different concentrations of Fe and N obtained by the Tauc method.

As reported by Ali et al.,[34] doping with Fe3+ in a TiO2 lattice decreases the band gap due to the overlap of the conduction band due to the Ti (d-orbital) and metal (d-orbital) of the Fe3+ ions. Furthermore, the mechanism of the photocatalytic process in Fe-doped TiO2 proposes that the Fe3+ ions induce the formation of new electronic states (Fe4+ and Fe2+) that extend along with the TiO2 band separation.

These electronic states can act as electron trapping sites and holes, and ultimately improve photocatalytic activity.[34] On the other hand, the influence of nitrogen in the decrease of the band gap is due to the fact that nitrogen can lead to a mixture of the N 2p orbital with the O 2p orbitals to form intermediate energy levels and move the absorption edge toward the visible light region.[35]

Finally, it should be noted that codoping with nitrogen and iron causes a stronger impact on the decrease of the band gap in comparison to the samples doped only with N or Fe3+ or not doped at all. As shown in the results found by Ali et al.,[34] the band gap of TiO2 nanoparticles decreased when doped with Fe, but not down to the level achieved in this work. In other research, Nassoko et al.[35] performed N-doping, reaching a similar tendency of decreasing band-gap values per increment of N concentration. In addition, Grigorov et al.[36] found that doping N–TiO2 decreases the optical gap; however, a similar behavior was presented, since, with the lowest concentration of N, the lowest band gap was reached. Likewise, this increment did not reach the level of the codoping with Fe and N, showing that codoping is favorable compared to just a single element because it maximizes the absorption range up to visible light. When comparing with other dopants, such as KI and Cu/S-codoped TiO2,[37,38] it is observed that the band gap decreased more with Fe–N-doped TiO2, extending the absorption to the visible light region even more than the other dopants.

Conclusions

TiO2 nanoparticles codoped with Fe and N have been prepared at different concentrations by green chemistry using the lemongrass leave extract. The synthesized nanoparticles are explored for possible applications in photoelectrochemical cells. TiO2 codoping shows a reduction in the particle size from 70 to 38 nm and the band gap energy from 3.09 to 2.22 eV with respect to the undoped TiO2. Finally, the codoping method with Fe and N was successful, and FT-IR and EDS analyses reveal that these species are present in the samples. Therefore, the synthesis route of the codoped TiO2 is interesting for its simple methodology and potential to synthesize various other nanocomposite materials.

Materials and Experimental Section

Materials

The materials used for the synthesis of titanium dioxide nanoparticles were titanium isopropoxide (Ti[OCH(CH3)2]4, 95%, Alfa Aesar) as a titanium precursor and natural lemongrass extract as a reducing agent. Ethanol (C2H5OH, Chemi) was used to wash the nanoparticles. The codoping of the titanium dioxide nanoparticles was performed using ammonium chloride (NH4Cl, Chemi) as a nitrogen precursor and nonahydrated ferric nitrate (Fe(NO3)3·9H2O).

Experimental Section

To obtain titanium dioxide nanoparticles, the process was divided into two stages, which began with the production of the natural lemongrass extract and subsequent reduction synthesis by means of a chemical reaction. For the preparation of the reducing extract, fresh leaves of lemongrass (Cymbopogon citratus) were washed with abundant distilled water, cut and dried in an oven at 60 °C; then they were cut into smaller pieces and milled. The infusion was prepared by immersion of 100 g of ground leaves in 500 mL of distilled water (0.2 g/mL) at a temperature of 90 °C. This extract was filtered several times to leave no solid residue and then concentrated by evaporation at 70 °C to 100 mL of solution.[16]

The titanium dioxide nanoparticles were made from a reduction mechanism, using a green chemistry process in which organic solvents are replaced by natural extracts. The nanoparticles were synthesized through the reduction of titanium tetra-isopropoxide (TTIP) with natural lemongrass extract. An aqueous solution of 850 mL of titanium tetra-isopropoxide at 10 mM was subjected to ultrasonic agitation for 30 min; then, 100 mL of lemongrass extract was added and subjected to magnetic agitation for 24 h at room temperature. The nanoparticles were separated by centrifugation at 3500 rpm for 15 min, and then, they were washed with ethanol and submitted to the same centrifugation conditions to be finally washed with distilled water and calcined up to 550 °C for 3 h, as reported by Realpe Jimenez and co-workers.[17,18] The nanoparticles of titanium dioxide were codoped using the wet impregnation method;[19] this process was divided into two parts, initially Fe3+ doping was performed and then doping with N. An aqueous suspension of the synthesized nanoparticles was subjected to ultrasound agitation for 30 min, and then an aqueous solution of nonahydrated ferric nitrate was added and ultrasound shaken for 1 h and then magnetic agitated with heating to 80 °C to evaporate the solvent (water). Finally, Fe-doped nanoparticles were calcined in a muffle at 400 °C during 2 h. For the N codoping of the TiO2–Fe nanoparticles, the same method described above was used. The concentrations of both dopants were modified to determine their effect and interaction on the optical and charge-transfer capabilities of the photoelectrode. Thus, the factors studied are Fe doping, N doping, and the levels at which they are evaluated are the concentrations in % wt/wt with respect to the amount of TiO2.

Characterization

FT-IR spectra were performed to determine the functional groups present in the synthesized samples according to the characteristic peaks at different wavelengths between 500 and 4000 cm–1. The size and morphology of the synthesized nanoparticles were determined through an SEM analysis using a JEOL JSM 540 scanning electron microscope. Finally, the spectrum of UV–visible diffuse reflectance was measured in the wavelength of 200–800 nm using a Thermo Scientific model EVOLUTION 600 UV/VIS spectrophotometer. This analysis allows determining the absorption range of the nanoparticles.