Appl. Phys. A (2016) 122:530 DOI 10.1007/s00339-016-0057-0
A new preparation of doped photocatalytic TiO2 anatase nanoparticles: a preliminary study for the removal of pollutants in confined museum areas Enrico Greco1 • Enrico Ciliberto1 • Antonio M. E. Cirino1 • Donatella Capitani2 Valeria Di Tullio2
•
Received: 1 August 2015 / Accepted: 15 April 2016 Springer-Verlag Berlin Heidelberg 2016
Abstract The use of nanotechnology in conservation is a relatively new concept. Usually, classical cleanup methods take into account the use of other chemicals: On the one hand they help the environment destroying pollutants, but on the other hand they often become new pollutants. Among the new oxidation methods called advanced oxidation processes, heterogeneous photocatalysis has appeared an emerging technology with several economic and environmental advantages. A new sol–gel method of synthesis of TiO2 anatase is reported in this work using lithium and cobalt (II) salts. The activation energy of the doped photocatalyst was analyzed by solid-state UV–Vis spectrophotometer. The mobility of Li ions on TiO2 NPs surface was studied by 7Li MAS NMR spectroscopy. Use of doped nanotitania is suggested from authors for the removal of pollutants in confined areas containing goods that must be preserved from decomposition and aging phenomena.
1 Introduction From the beginning of this century, many attempts have been undertaken to improve the environmental conditions using new photocatalytic nanomaterials like in Milan,
& Enrico Greco
[email protected] 1
Department of Chemical Sciences, University of Catania, viale A. Doria 6, 95125 Catania, Italy
2
Magnetic Resonance Laboratory ‘‘Annalaura Segre’’, Institute of Chemical Methodologies, CNR, Research Area of Rome 1, via Salaria km. 29.500, 00015 Monterotondo, Rome, Italy
Paris, Athens, Copenhagen, Tokio and Naples [1]. Several researches have been carried out for the development of innovative processes for the friendly environment purification of the atmosphere from industrial emissions and the polluted water using sunlight radiation since it is renewable and clean source of energy. In this light, we have studied new materials capable of degrading organic and inorganic air pollutants through the process of photocatalysis. This consists in the use of solid semiconductor able to oxidize harmful substances up to complete degradation. Currently, the titanium dioxide (TiO2) is one of the materials that has attracted major interest in this field. In the literature, the values of the band gap and the respective wavelengths corresponding to the maximum absorption for the active phases of TiO2 are reported. It has been found for the anatase a value of 3.20 eV (k = 384 nm), slightly higher than that reported for the rutile which is 3.03 eV (k = 410 nm) [2]. Despite the rutile that shows a ‘‘band gap,’’ in the case of a single crystal, lower than anatase and is therefore able to absorb in the range of wavelengths closer to the maximum intensity of the solar spectrum (around 500 nm), anatase shows the highest photocatalytic activity. This is due to the properties of anatase phase which entail a greater density of localized states, to a higher amount of hydroxyl radicals adsorbed on the surface and to a lower probability of recombination of electron– hole pair [3]. The mentioned gap values indicate that the two phases are excited when illuminated by electromagnetic radiation with a wavelength k B 410 nm for the rutile and k B 384 nm for the anatase and then from the UV portion of the electromagnetic spectrum. Since the spectrum of sunlight is made only 4 % of UV radiation [4], this is the biggest limitation of the use of TiO2 in applications that use solar energy. For this reason, it has been necessary for an
123
530
Page 2 of 6
enlargement of the region of absorption of the spectrum, restricting the value of the forbidden band energy in order to activate the TiO2 also with visible light. Therefore, we tried to face this problem by performing an appropriate doping which is able to generate levels within the gap in order to make the transition energies lower than the corresponding ones in the pure anatase. The effectiveness of the photocatalytic activity of TiO2 depends on many factors such as the type and the relative amount of crystalline phase present [5], the size of the particles and therefore their specific surface area, the type of materials to degrade [6], the crystallinity grade, the impurities, the density of the hydroxyl groups on the surface and the method of preparation [7]. Since the photocatalytic activity is expressed on the surface of the photocatalyst, the high surface–volume ratio that characterizes a nanomaterial, by increasing the availability of surface sites, greatly helps to increase the speed of the photodecomposition reactions [8]. The purpose of this work was to develop a synthesis method of nanoparticles of TiO2 doped with lithium and cobalt which fulfills the criteria of cost and which can be carried out without the use of organic solvents and hightemperature treatments. We have also paid attention to the role of lithium ions with respect to the ability to maintain in dispersion the particles for electrostatic repulsion and to their function within the nanoparticles. In this context, the prepared anatase nanoparticles were characterized by using techniques able to identify their size and their crystal phase: scanning electron microscopy (SEM), Raman spectroscopy and X-ray diffraction (XRD); in order to evaluate the absorption spectra in the visible light region, an UV–Vis spectrometer was used. The role and the mobility of the dopant Li? ions were finally investigated by 7 Li magic-angle spinning nuclear magnetic resonance spectroscopy (7Li MAS NMR).
2 Experimental In this work, we have used a modified method of sol–gel synthesis [9–13] which allows the direct synthesis of anatase TiO2 nanoparticles dispersed in water at low temperature, without the use of organic surfactants and treatments of calcination. The dispersion in solution is obtained by electrostatic way. The synthesis, carried out in excess of water, used titanium tetrabutoxide (TBot) as a precursor and a molar ratio between precursor, water, isopropanol and nitric acid equal to 1:155:1.5:0.1. After a careful washing and drying of the glasswares for the synthesis, 3 ml of the precursor TBot (Sigma-Aldrich, C97 %, MM = 340.32 g/mol, d = 1.00 g/ ml) was placed, together with 0.96 ml of 2-propanol (Lab-
123
E. Greco et al.
Scan, C99.7 %, d = 0.785 g/mL), in a dropping funnel. This was mounted above a 100-ml flask containing 24.7 ml of deionized water and 0.3 ml of nitric acid (Baker, 65 %, d = 1.40 g/ml) so as to obtain a pH of 1.3. The flask, containing inside a magnetic stirrer, was placed in an oil bath so as to keep the temperature constant during all the time necessary to the synthesis. After reaching the temperature of 80 C, it was opened the tap funnel containing the solution of the precursor that was dripped slowly (1 ml per minute) in the flask where the aqueous solution was contained and the system was kept under vigorous stirring for 24 h. At the end of the process, the colloidal solution of anatase was allowed to cool to room temperature. The particles remain dispersed in the solution for several months without showing noticeable agglomeration. To obtain dry particles, the colloidal solution was first subjected to cycles of centrifugation (Beckman Microfuge ETM) at 14,000 rpm for 30 min. At the end of each spin cycle, the supernatant solution was removed and replaced with ultrapure water to make the washings through other spin cycles. After scoring three washes, the obtained gel was dried under vacuum at 25 C so as to obtain the nanoparticles in powder form. In order to remove nitric acid from the synthesis, we doped the nanoparticles with lithium. The synthesis was then repeated with the same modalities and amounts used previously but without using the acid source of H?. In this case, the nitric acid has been replaced by a suitable amount of LiCl. In order to improve the photocatalytic properties of the particles of TiO2, we also realized a synthesis in which, besides the use of the lithium salt in place of nitric acid, CoCl2 was added as a dopant, in order to obtain a greater absorption in the visible light, fundamental for many technological applications. The experimental procedure is the same so far followed with the only difference that in the aqueous solution 0.84 g of CoCl2 was also added, equal to 5 % compared with TBot used. The XRD instrument used for diffractometric measurements was a Bruker-AXS D8. The spectra were obtained with an angle 2h between 20 and 70, using a Ka radiation ˚ ), with a scanning speed of 1/min. of Cu (1,542 A Raman spectra were recorded using a micro-Raman spectrometer made up of an argon laser source Spectra Physics with k = 514.5 nm, an optical microscope with objectives ranging from 10 to 1009 and a detector CCD Jobin–Yvon HR 460 cooled with liquid nitrogen. The images in electron microscopy have been recorded using the powdered sample by a microscope FEDSEM LEO Supra 55 VP, combined with a column GEMINI, Carl Zeiss, with the addition of an analysis tool Oxford EDX (energy-dispersive spectroscopy) to obtain information about quantitative distribution and amount of metal species in the samples.
A new preparation of doped photocatalytic TiO2 anatase nanoparticles: a preliminary study…
Optical absorption spectra were recorded at room temperature with a Jasco V-650 spectrophotometer UV–Vis– NIR. 7 Li MAS NMR spectroscopy—Samples for NMR analysis were prepared in a dry box in N2 atmosphere, inserted in 4-mm zirconia rotors with the available volume reduced to 20 ll and sealed with Kel-F caps. 7 Li magic-angle spinning (MAS) NMR (I = 3/2, 92.6 % abundance) spectra were recorded at 155.50 MHz on a Bruker Avance 400 spectrometer. The p/2 pulse width was 5.5 ls. Spectra were acquired with a time domain of 4 k data points and were zero filled and Fourier transformed with a size of 16 k data points. The spin rate was kept at 5500 Hz. The ppm scale was externally referenced to LiCl used as a reference. The longitudinal relaxation time T1 was measured using the saturation recovery pulse sequence [14]. The measurement of the longitudinal relaxation times allowed the optimization of the recycle delay to be used for collecting 7Li spectra. The transverse relaxation time T2 was determined using a modified quadrupolar Carr–Purcell–Meiboom–Gill (QCPMG) pulse sequence [15].
3 Discussion The X-ray diffraction has been used to verify that the crystalline phase is anatase, as designed in the synthesis step. In Fig. 1, the diffractogram of a lithium-doped sample is reported; the most intense peak is at a value of 2h = 25.32 and was assigned to the lattice plane (101) of anatase. The peak at a value of 2h = 31 indicates the presence of a small percentage of rutile, probably formed during the synthesis phase or through conversion of the anatase phase due to exposure to the light. With these measures, the size of the nanoparticles was calculated by Scherrer’s equation [16].
Fig. 1 Diffractogram of Li–Co-doped TiO2. For comparison is also reported the position of the 100 % peaks of the anatase and missing rutile phase
Page 3 of 6
530
X-ray diffraction was also carried out on the samples doped with Co. Figure 1 shows that the diffraction pattern of TiO2 doped is almost coincident with pure TiO2. The ionic radius is one of the most important factors that determine the possibility of the dopant to enter into the lattice of the crystal of TiO2 in order to form a stable lattice. If the ionic radius of the dopant is too large or too ˚ ), the input of the dopant small compared with Ti4? (0.64 A in the crystal leads to a distortion of the same. This does not happen in this case because the ionic radius of Co2? ˚ ) is comparable with Ti4? and, in fact, it was not (0.65 A noticed any appreciable differences between the two diffractograms. In Fig. 2, the Raman spectrum of the undoped TiO2 is reported. From the comparison with the spectra of pure phases of rutile and anatase, it is clear that the synthesis produced a sample of nanoparticles whose signals are comparable to those of the anatase phase. In fact, the bands of our sample are located at 417, 527 and 645 cm-1. All the peaks appear very wide; in particular, in the peak at 645 cm-1 we can notice a raising of the baseline probably due to the overlap with the signal of rutile that falls to 610 cm-1. The presence of a small amount of rutile does not involve any problems because it has been shown that the best catalytic activity is obtained by using a mixture of anatase–rutile 4:1, as in the commercial product Degussa P25. The more complex structure of the undoped sample spectrum can be explained by supposing a little content of amorphous titania (459 and 612 cm-1 [17]) whose spectrum could overlap with the anatase one. On the contrary, the spectrum of the Li–Co-doped species fits quite well the Raman spectrum of the anatase standard.
Fig. 2 Comparison of the Raman spectra of the undoped TiO2 (black), Li–Co-doped TiO2 (green), rutile standard (red) and anatase standard (blue)
123
530
Page 4 of 6
E. Greco et al.
Electron microscopy characterization allowed us to determine the size and the morphology of the nanoparticles produced with the various synthesis procedures. The measurements were performed on powder samples. The preparation of the samples was sometimes difficult because nanoparticles are well dispersed in aqueous solution and they tend to form aggregates when dried, and then it is very difficult to observe isolated nanoparticles. Figures 3 and 4 show an aggregate of non-doped TiO2 nanoparticles, fairly uniform in size with diameters of about 80 nm. As previously mentioned, although the synthesis has produced nanoparticles well dispersed in solution, these form agglomerates of varying sizes as a result of drying. Figure 5 shows 7Li MAS NMR spectra of LiCl (a), TiO2 nanoparticles synthesized with LiCl (b) and TiO2
Fig. 5 7Li MAS NMR spectra of a standard of LiCl (a), TiO2 nanoparticles synthesized with LiCl (b) and TiO2 nanoparticles synthesized with LiCl ? HNO3 (c). The deconvolution of the spectrum of TiO2 nanoparticles synthesized with LiCl ? HNO3 is reported in the inset
Fig. 3 SEM image of undoped TiO2 nanoparticles. The average dimension of the nanoparticles is around 80 nm
Fig. 4 SEM image of an aggregate of Li–Co-doped TiO2 dried under high vacuum
123
nanoparticles synthesized with LiCl ? HNO3 (c). The linewidth gives information on the lithium mobility. The comparison between the linewidth of 7Li in LiCl and the linewidth of 7Li in nanoparticles indicates the high mobility of lithium in nanoparticles. The spectrum of LiCl shows a peak at 0 ppm with a linewidth of about 287 Hz. The spectrum of TiO2 nanoparticles synthesized with LiCl shows a peak at -0.05 ppm with a linewidth of 10 Hz indicating the high mobility of lithium ions (Fig. 5b). The spectrum of TiO2 nanoparticles synthesized with LiCl ? HNO3 was deconvoluted to obtain a sharp and a broad peak centered at -0.08 and -0.06 ppm, respectively. The line width of the former peak was found to be 6 Hz, whereas that of the broad one was 37 Hz, see the inset in Fig. 5c. In this case, the broad peak was ascribed to Li? interacting with H? on the nanoparticles surface, whereas the sharp peak was ascribed to the other Li? moving on the nanoparticles surface and inside the nanoparticles porous matrix. To investigate Li? mobility, the longitudinal (T1) and transverse (T2) relaxation times were measured. Figure 6 shows the longitudinal relaxation decays of 7Li magnetization measured in LiCl and in Li-doped TiO2– NPs. In all samples, a mono-exponential decay of the longitudinal magnetization Mz(t) versus time was obtained. T1 values reported in Table 1 were obtained fitting experimental data to the following equation:
A new preparation of doped photocatalytic TiO2 anatase nanoparticles: a preliminary study…
Fig. 6 Longitudinal relaxation decay measured in LiCl (red circles) and Li-doped TiO2 nanoparticles (black circles)
Table 1 Longitudinal relaxation times measured in LiCl and in Lidoped TiO2 nanoparticles Sample
T1(s)
R2
Li-doped TiO2
2.160 ± 0.010
0.9999
Li-doped TiO2 ? HNO3
2.168 ± 0.018
0.9999
LiCl std
[268
0.9999
Mz ðtÞ ¼ Mz;eq Mz;eq Mz ð0Þ et=T1 It is worth noting that 7Li exhibited a definitely higher mobility in Li-doped TiO2–NPs (about 2.1 s) than in LiCl which was used as precursor ([268 s). To evaluate the presence of different lithium domains and their relative motions, transverse relaxation times were measured. With respect to T1, T2 yields a more detailed information on molecular motions occurring in domains with reduced dimensions. In fact, in the case of Li-doped TiO2–NPs, the decay of the longitudinal magnetization was averaged out to a mono-exponential decay, whereas the transverse magnetization showed a bi-exponential decay corresponding to two distinct values of the transverse relaxation time. T2 values reported in Table 2 were obtained fitting experimental data to the following equation. As Mxy ðtÞ ¼ M01 eðt=T21 Þ þ M02 eðt=T22 Þ where M01 and M02 are the spin densities of the two components corresponding to two lithium domains and T21
Page 5 of 6
530
Fig. 7 Comparison of the absorption spectra of Degussa P25 (blue), of the Li-doped TiO2 (red) and the Li–Co-doped TiO2 (black)
and T22 are the transverse relaxation times of each component. Therefore, transverse relaxation time measurements allowed the detection of two Li? domains and the evaluation of the relative amount of Li? in each domain. The fast relaxing component was assigned to Li? on the surface of nanoparticles and the slow relaxing one to Li? ions confined into the porous nanoparticles. In the case of Li-doped TiO2 nanoparticles, the amount of Li? ions on the nanoparticles surface was found to be higher on the surface (76 %) and then inside the nanoparticles (24 %), whereas in the case of Li-doped TiO2 ? HNO3 nanoparticles the presence of H? ions on the nanoparticles surface possibly reduced the presence of Li? ions moving on the nanoparticles surface (46 %). UV–Vis spectroscopy was used to characterize the nanoparticles synthesized in order to check the absorption in the visible region. Figure 7 shows the comparison between the absorption spectra of UV–Vis commercial product Degussa P25 and samples we synthesized and doped with Li or Li and Co. As the graph shows, the Degussa P25 shows the maximum peak of absorption around 309 nm and decreases rapidly up to about 400 nm. This means that its activity is conducted mainly in the region of the near UV spectrum. Both the samples of doped TiO2 synthesized by us show different levels of absorption. In sample doped with Li, an absorption maximum around 379 nm shifted by about 70 nm with respect to Degussa P25. The sample also has absorption in the visible up to above 500 nm, which means
Table 2 Transverse relaxation times measured in Li-doped TiO2 nanoparticles Sample
M01
T21(s)
M02
T22(s)
R2
Li-doped TiO2
0.76
0.022 ± 0.001
0.24
0.111 ± 0.006
0.9993
Li-doped TiO2 ? HNO3
0.46
0.010 ± 0.010
0.54
0.069 ± 0.003
0.9970
123
530
Page 6 of 6
that its activity also takes place in the region of the yellow– green. The sample doped with Li and Co shows an absorption maximum around 353 nm, with a clear tail in the visible range up to about 730 nm. This range indicates that the doping with Co makes it possible to obtain extremely noticeable absorption in the range of yellow–green with a tail to red.
4 Conclusions A new synthesis of TiO2 nanoparticles in the anatase phase was realized using a modified sol–gel method. This has allowed us to obtain nanoparticles in the anatase phase well dispersed in water without any use of organic surfactants and treatments at high temperature. The particles thus obtained were doped with Li? to study the effectiveness of this ion in replacing the H? ion on the surface of the particles and maintain the same well dispersed in solution. It was shown that the lithium ion is able to maintain the nanoparticles in dispersion by electrostatic repulsion. Furthermore, by solid-state 7Li MAS NMR it was shown that the lithium ions have higher mobility than the H? ions by determining its capacity as the electrostatic repeller. In particular, the concentration of lithium ions at the surface appears to be higher than in the bulk of the particles. This behavior parallels the mobility of the ions at the surface itself where very short relaxation times were measured. The use of lithium and cobalt as dopants has been studied by UV–Vis absorption measurements. It was showed an increased absorption in the visible range for the doped samples compared with the sample of pure TiO2. So, this type of nanoparticles is a good candidate to be used in different sectors of technology of materials that require the ability of the photocatalytic TiO2 even in conditions of irradiation with visible light alone and in the absence of UV radiation. The catalysts are in the course of further studies, in particular the possibility of applying these nanoparticles on glass, ceramic materials and other inorganic materials in order to make them usable in the fields of sustainable architecture, eco-building and, as a filter, in indoor museum areas for the removal of atmospheric pollutants.
123
E. Greco et al. Acknowledgments This research was supported by TECLA PON 03PE_00214_1.
References 1. L. Cassar, Photocatalysis of cementitious materials: clean buildings and clean air. MRS Bull. 29(05), 328–331 (2004) 2. W. Wunderlich, T. Oekermann, L. Miao et al., Electronic properties of Nano-porous TiO2- and ZnO-thin films-comparison of simulations and experiments. J. Ceram. Process Res. 5, 343–354 (2004) 3. K. Hadjiivanov, D.G. Klissurski, Chem. Soc. Rev. 25, 61 (1996) 4. S. Karvinene, R.-J. Lamminma¨ki, Preparation and characterization of mesoporous visible-light-active anatase. Solid State Sci. 5, 1159–1166 (2003) 5. T. Alapi, P. Sipos, I. Ilisz, G. Wittman, Z. Ambrus, I. Kiricsi, K. Mogyoro´sie, A. Dombi, Synthesis and characterization of titania photocatalysts: the influence of pretreatment on the activity. Appl. Catal. A Gen. 303, 1–8 (2006) 6. Q. Zhang, L. Gaoe, J. Guo, Effects of calcination on the photocatalytic properties of nanosized TiO2 powders prepared by TiCl4 hydrolysis. App. Catal. B Environ. 26, 207–215 (2000) 7. T. Kawahara, Y. Konishi, H. Tada, N. Tohge, J. Nishiie, S. Ito, A patterned TiO2 (anatase)/TiO2 (rutile) bilayer-type photocatalyst: effect of the anatase/rutile junction on the photocatalytic activity. Angew. Chem. Int. Edition 41, 2811–2813 (2002) 8. S.S. Hong, M.S. Lee, S.S. Park, G.D. Lee, Catal. Today 87, 99 (2003) 9. C. Xiaboo, S.S. Mao, Titanium dioxide nanomaterials synthesis, properties, modifications and application. Chem. Rev. 107, 2891–2959 (2007) 10. B. O’Regan, M. Gra¨tzel, Nature 353, 737 (1991) 11. J.H. Lee, Y.S. Yang, Mater. Chem. Phys. 93, 237 (2005) 12. X. Chen, S.S. Mao, Chem. Rev. 107(7), 2891–2959 (2007) 13. J. Jing, J. Feng, Low-temperature synthesis of water-dispersible anatase titanium dioxide nanoparticles for photocatalysis. J. Colloid Interface Sci. 396, 90–94 (2013) 14. T.C. Farrar, E.D. Becker, Pulse and Fourier transform NMR (Academic Press, New York, 1971) 15. F.H. Larsen, H.J. Jakobsen, P.D. Ellis, N.C. Nielsen, Sensitivityenhanced quadrupolar-echo NMR of half-integer quadrupolar nuclei. Magnitudes and relative orientation of chemical shielding and quadrupolar coupling. J. Phys. Chem. A 101, 8597–8606 (1997) 16. P. Scherrer, Nachrichten von der Gesellschaft der Wissenschaften zu Go¨ttingen, Mathematisch-Physikalische Klasse. 2, 98–100 (1918) 17. F.D. Hardcastle, Raman spectroscopy of titania (TiO2) nanotubular water-splitting catalysts. J. Ark. Acad. Sci. 65, 43–48 (2011)
