- Email: [email protected]
Femtosecond pulsed laser deposition of nanostructured TiO2 films
Applied Surface Science 255 (2009) 5206–5210
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Femtosecond pulsed laser deposition of nanostructured TiO2 films Mikel Sanz a, Malgorzata Walczak a, Rebeca de Nalda a, Mohamed Oujja a, ˜ ares b, Marta Castillejo a,* Jose´ F. Marco a, Javier Rodriguez b, Jesu´s G. Izquierdo b, Luis Ban a b
Instituto de Quı´mica Fı´sica Rocasolano, CSIC, Serrano 119, 28006 Madrid, Spain Departamento de Quı´mica Fı´sica I, Facultad de Ciencias Quı´micas, Universidad, Complutense de Madrid, 28040 Madrid, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
Nanostructured deposits of TiO2 were grown on Si (1 0 0) substrates by laser ablating a TiO2 sintered target in vacuum or in oxygen using a Ti:sapphire laser delivering 80 fs pulses. The effect of the laser irradiation wavelength on the obtained nanostructures, was investigated using 800, 400 and 266 nm at different substrate temperatures and pressures of oxygen. The composition of the deposits was characterized using X-ray photoelectron spectroscopy (XPS) and the surface morphology was studied by environmental scanning electron microscopy (ESEM) and atomic force microscopy (AFM). Deposits are absent of microscopic droplets in all conditions explored. The best deposits, constituted by nanoparticles of an average diameter of 30 nm with a narrow size distribution, were obtained at the shorter laser wavelength of 266 nm under vacuum at substrate room temperature. ß 2008 Elsevier B.V. All rights reserved.
Available online 3 August 2008 Keywords: Femtosecond pulsed laser deposition Nanoparticles Nanostructured deposits TiO2
1. Introduction In recent years pulsed laser deposition (PLD) has proved to be an effective technique for the deposition of nanostructured films of a wide variety of materials [1] both for fundamental research and for technological applications [2]. The main drawback of this deposition method is the presence of microscopic particulates on the surfaces of the films which are present with high densities when the optical absorption coefficient of the target is small at the wavelength used for ablation. The origin of those particulates is not clear yet, but it is generally assumed that thermal processes, such as melting and boiling during the laser-target interaction, favour their ejection along with ionic and atomic species. Ultrashort femtosecond (fs) laser pulses offer a high material removal efficiency and high deposition rates of nanometer scale particles free of microscopic particulates and therefore fs PLD constitutes an attractive procedure for the fabrication of nanostructured deposits. However, it appears that the nature of nanoparticles (NPs) grown by fs PLD strongly depends on the material and deposition conditions and therefore various studies [3–10] have tried to elucidate the different processes involved, showing the possibility of using fs PLD as a general route to NP formation. Metal and semiconductor NPs have been synthesized by fs PLD [11–21]. However, the majority of these studies have been performed with light pulses centred around the peak wavelength
of Ti:sapphire laser at 800 nm and few reports have explored the effect of this parameter [16,20]. Analysis of the process over a broader range of wavelengths can provide important information about the NPs formation processes and serve as experimental tests for advanced theoretical models. TiO2 is a wide-band semiconductor with exceptional optical and electronic properties of wide use [22] as a photocatalytic air purification agent, for self-cleaning coatings, in environmental sanitation, and in photovoltaic devices for the fabrication of electrodes for solar energy conversion. Two of the three crystallographic phases of TiO2 namely anatase and rutile, with respective band gaps of 3.2 and 3.0 eV [23], are especially suitable for these applications where the performance of the material depends critically on its nanostructure. In this paper, we report on the growth of TiO2 deposits by PLD on Si (1 0 0) substrates from sintered TiO2 targets under different conditions. The effect of laser wavelength was studied under vacuum and in presence of oxygen at different substrate temperatures. Surface structures of prepared TiO2 films were observed using environmental scanning electron microscopy (ESEM) and atomic force microscopy (AFM). The chemical states of the material deposited were examined by X-ray photoelectron spectroscopy (XPS). To our knowledge, this is the first study of the dependence with laser wavelength of the properties of the TiO2 NPs created by fs PLD. 2. Experimental
* Corresponding author. Tel.: +34 91 5619400; fax: +34 91 5642431. E-mail address: [email protected] (M. Castillejo). 0169-4332/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2008.07.148
The experimental set up consists of a PLD stainless steel chamber evacuated by a turbomolecular pump down to
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
2 104 Pa. Deposition was carried out under vacuum or under a background pressure of oxygen of 5 Pa. TiO2 targets were prepared from a special grade rutile powder (Sigma Aldrich, particle size <5 mm) pelletized into disks. Targets were placed on a rotating sample holder to avoid crater formation. Pulses of 80 fs were produced by a Ti:sapphire amplified laser system (Spectra Physics) centred at 800 nm at a maximum repetition rate of 1 kHz. Pulses of 400 and 266 nm were generated by using a BBO-based frequency doubler/tripler. For the growing of the deposits, a frequency of 10 Hz has been used. The temporal profile of the pulses was diagnosed with a second harmonic generator autocorrelator. Energy control was performed by a half-wave plate/polarizer combination. The laser beam was focussed at 358 onto the target with a 20 cm focal length lens. The Si (1 0 0) substrates were ultrasonically degreased in acetone and methanol for 10 min. They were mounted on a heating element that allowed operating in the temperature range between room temperature (RT) and 700 8C and placed at a distance of 40 mm in front of the target. Deposits were grown in all cases with 7200 pulses at a repetition rate of 10 Hz. The modification of the irradiated area of the targets was studied by optical microscopy (Leica, S8APO) with a 160 microscope objective and equipped with a CCD camera. The surface morphology of the deposits was examined by ESEM (Philips XL30) and by AFM (Digital Instruments). All the AFM images were obtained in tapping mode and they were postprocessed using a third-order flattening routine. The deposits were also analyzed by XPS using a Leybold LHS10 XPS spectrometer, under an operating vacuum better than 1 106 Pa, using Al Ka radiation (130 W) and an analyser transmission energy of 200 and 50 eV for the wide and narrow scan spectra, respectively. The spectra were recorded at take-off angles of 908. All binding energy (BE) values were chargecorrected to the adventitious C 1s signal which was set at 284.6 eV and are accurate to 0.2 eV. Relative atomic concentrations were calculated using tabulated atomic sensitivity factors [24]. 3. Results and discussion 3.1. Modification thresholds In order to obtain the deposits under the same conditions for the different wavelengths of excitation, ablation of targets was performed at a laser fluence of 1.5 and 2 times the modification threshold fluence for each wavelength. For short pulse regime, the modification threshold fluence Fth depends on the material and the number of laser pulses N applied to the same spot [25]. For laser pulses with a Gaussian spatial beam profile, the maximum laser fluence, F, on the sample surface and the diameter, D, of the modified area are related by D2 = 2v0 ln(F/ Fth), where v0 is the 1/e2 radius of the beam distribution. The fluence at the target surface is the maximum calculated from the Gaussian beam radius and is related to the pulse energy by F ¼ E=pv20 . The diameter of the darkened modified area was determined trough observation by optical microscopy of the irradiated targets. From a plot of D2 versus ln E, Fth can be determined. As an example Fig. 1 shows the dependence of D2 with ln E upon irradiation with 10 pulses at the three different wavelengths. The obtained thresholds for 10 pulses are 140, 124, and 75 mJ/ cm2 for 800, 400 and 266 nm, respectively. The corresponding photon energies are respectively lower, around the same and larger than the optical bandgap of the target rutile material (3.0 eV, 413 nm) [23].
5207
Fig. 1. The squared diameter D2 of the modified (darkened) areas versus the logarithm of laser pulse energy, for 10 laser pulses upon irradiation at 266, 400 and 800 nm (squares, triangles and circles, respectively).
3.2. Surface structure The laser parameters, wavelength and fluence, gas atmosphere and substrate temperature were varied in order to determine their influence on the produced deposits. As observed by ESEM and AFM (see below) the deposited films contain principally NPs smaller than 50 nm and few larger particles with sizes ranging from 100 to 200 nm. This structure has also been observed in fs PLD with other materials [14,21]. The images do not show any droplets of micrometer size, as observed in the case of nanosecond (ns) PLD [26]. The thicknesses of the deposits are in the range of 20–60 nm, being considerably thinner than those obtained upon ns laser irradiation using the same number of pulses [26]. To characterize the density of particles we analyzed the deposits by ESEM (Fig. 2). For 266 nm (Fig. 2a), a high density is obtained, covering practically all the substrate. In the presence of oxygen (Fig. 2b) a similar density is produced but it decreases substantially when the substrate is heated up to 600 8C (Fig. 2c). It was also noted that in the range of laser fluences used to obtain the deposits (1.5 and 2 times the modification threshold fluence, corresponding to 112 and 150 mJ/cm2, respectively), changes in density or sizes of particles are not observed. Upon ablation at 400 nm, the number of particles with diameter 100 nm is higher than in the case of operation at 266 nm (Fig. 2d). When using oxygen (Fig. 2e) the total density of particles decreases while an increase is observed by deposition on a substrate heated at 600 8C (Fig. 2f), also showing a numerous population of particles with sizes in the range of 100–200 nm. For 800 nm, deposition in vacuum (Fig. 2g and h) leads to a lowdensity deposit. However, when the substrate is heated, their density increases considerably (Fig. 2i) as also observed at 400 nm. The surface morphology and topography was investigated in greater detail by AFM. As an example Fig. 3 shows an AFM image of a deposit grown by ablating the target at 266 nm in vacuum at RT. To obtain the characteristic sizes of the NPs in terms of the equivalent diameter, the images were analyzed by using Scanning Probe Image Processor program [27]. As reported with ESEM, analysis of AFM images revealed a very different influence of the gas atmosphere and substrate temperature depending on the excitation wavelength. For 266 nm, the most uniform deposits were obtained in vacuum at RT (Fig. 3). As mentioned before, a high density of nanoparticles is observed with a narrow size distribution, with average diameters of 30 nm and absence of particulates larger than 100 nm (Fig. 4a). If the deposits are grown in oxygen at
5208
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
Fig. 2. ESEM images of the surfaces of TiO2 thin films grown at the indicated deposition conditions. The bar size is 1 mm in all images.
RT, the obtained size distribution of NPs becomes wider and the average diameter increases slightly to 35 nm (Fig. 4b). Upon heating of the substrate to 600 8C in vacuum, no appreciable change is observed in the size distribution, although the ESEM results show a decrease in density. The observed NPs appear as islands of almost hemispherical shape, in contrast with the morphology obtained by 266 nm ns PLD of the same material [26]
where deposits are constituted by a nanostructured film with presence of microscopic particulates. The sizes of particulates grown at 400 and 800 nm are similar, with average diameter of 50 nm (Fig. 4c and d) although the particle density is higher at 400 nm. In vacuum, NPs of 20–50 nm are dominant, but some larger particles of 100 nm are also observed. The deposits obtained under oxygen atmosphere display
Fig. 3. AFM images of the nanostructures deposited under vacuum, at RT and 266 nm at F = 2Fth: (a) topography image, (b) phase image (c) 3D topography image.
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
5209
Fig. 4. Size histogram of the nanoparticles produced by fs laser ablation at F = 2Fth: (a) 266 nm in vacuum and RT, (b) 266 nm in oxygen and RT, (c) 400 nm in oxygen and RT and (d) 800 nm in vacuum, heating the substrate at 600 8C.
a richer content of small size particles (100 nm) (Fig. 4c), whereas by heating the substrate at 600 8C (Fig. 4d), the size distribution becomes wider, with particle diameters up to 200 nm. 3.3. Composition The Ti 2p and O 1s XPS spectra recorded from deposits obtained by ablating TiO2 targets are presented in Fig. 5. The Ti 2p spectrum corresponding to the film prepared at 266 nm shows two different spin-orbit doublets (Fig. 5a). The BE of the 2p3/2 and 2p1/2 core levels of the major doublet, which accounts for 86% of the total spectral area, 458.2 and 463.9 eV, respectively, are characteristic of TiO2 [28,29] and in fact, this component of the spectrum is identical to the Ti 2p spectrum recorded from the TiO2 target material (not shown). The minor doublet (14% of the total spectral area) is characterized by BEs of 456.1 and 461.8 eV of the Ti 2p3/2 and 2p1/2 core levels, respectively, and can be associated to the presence in the film of a Ti oxide in which Ti has an oxidation state lower than 4+ [29]. The O 1s spectrum recorded (not shown) from the film prepared at 266 nm shows four contributions corresponding to Ti–O bonds from the deposited material, Si–O bonds coming from the SiO2/Si substrate, and hydroxyl groups and water [28,29] which have been probably adsorbed on the surface of the film after exposure of the sample to the laboratory atmosphere. For excitation photon energies around or smaller than the bandgap (400 and 800 nm, respectively) the results are very similar. The Ti 2p spectrum shows (Fig. 5b) only one spin-orbit component, with BEs of the 2p3/2 and 2p1/2 core levels, which are, as mentioned above, characteristic of TiO2. The O 1s
spectrum recorded from these films displays the same four contributions as the case of 266 nm. Taking into account the relative areas of the main O 1s component (associated to Ti–O bonds) and the relative area of the main Ti 2p contribution, the corresponding O/Ti atomic ratio is 2.2, which is close, within the error of the experimental determination, to the expected stoichiometric ratio of 2.0. The results clearly show that the wavelength of the laser used for deposition has a strong effect on the characteristics of the NPs produced by fs PLD. The tendency of producing smaller NPs with a narrower size distribution at shorter wavelengths of the fs laser, as observed herein, has also been reported by other authors [16,20] and discussed in terms of the larger strain rate in the expansion of the plasma fluid into vacuum following irradiation of the target by UV pulses [16]. The photon excitation energy at 266 and 400 nm is sufficient to promote the electron to the conduction band of the semiconductor TiO2 material, while two photons are required at 800 nm. Upon 266 nm fs laser irradiation, and due to the high absorption coefficient at this wavelength (a 106 cm1 [30]), a thin layer below the surface of the target is suddenly heated up without changing its density. Irradiation proceeds under heat confinement conditions (the duration of the laser pulse is well below the heat diffusion time tpulse td = 1/a2Dh with Dh = 0.03 cm2 s1 the heat diffusion coefficient [31]) and once energy deposition is complete, the plasma expands adiabatically from an initially high temperature. As the plasma reaches densities close to the critical point in the phase diagram, phase decomposition takes place with high fragmentation leading to the observed small NP formation [5,9,14,17].
5210
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
about the processes of formation of nanoparticles as well as experimental tests for advanced theoretical models. Research continues to explore the crystalline quality of the deposits using micro X-ray diffraction dispersion with synchrotron radiation and to characterize the accompanying ablation plume by mass spectrometry and optical emission spectroscopy. Acknowledgements Funding from Spanish MEC (Projects CTQ2007-60177, CTQ2005-08493-C02-01) is gratefully acknowledged. M.W. thanks EU for a contract (MESTCT-2004-513915) and M.S. thanks CSIC for a I3P contract. We are grateful to D. Go´mez (Instituto de Ciencia y Tecnologı´a de Polı´meros, CSIC) for operating the ESEM and to F. Gonza´lez-Posada (Instituto de Sistemas Optoelectro´nicos y Microtecnologı´a, UPM) for AFM measurements. This research has been performed within the Unidad Asociada ‘‘Quı´mica Fı´sica Molecular’’ between Departamento de Quı´mica Fı´sica I of Universidad Complutense de Madrid (UCM) and CSIC. References
Fig. 5. Ti 2p XPS spectra recorded from the film deposited ablating the target with (a) 266 and (b) 800 nm. In all cases the deposits were produced at RT in oxygen.
4. Conclusions TiO2 nanostructured deposits on Si (1 0 0) substrates were obtained from sintered rutile targets at the wavelengths of 800, 400 and 266 nm from a fs Ti:sapphire laser. The surface structure of the material deposited was observed by ESEM and AFM and the chemical states by XPS. The deposits consist of TiO2 nanoparticles with grain size down to 20 nm with overimposed larger particles in the 100–200 nm range. The best deposits, with the highest density of nanoparticles of around 30 nm diameter and negligible amount of larger particulates (>100 nm), were grown by ablation at 266 nm in vacuum at RT. At this wavelength, the use of oxygen during deposition and the heating of the substrate stimulate the growth of larger particles. The highest concentrations of particulates were observed at the longest laser wavelengths, 400 and 800 nm by heating the substrate. These results indicate the importance of wavelength in fs PLD and provide important clues
[1] D.B. Chrisey, G.K. Huber (Eds.), Pulsed Laser Deposition of Thin Films, Wiley, New York, 1994. [2] R. Eason (Ed.), Pulsed Laser Deposition of Thin Films: Applications-Led Growth of Functional Materials, Wiley, New York, 2006. [3] B.N. Chichkov, C. Momma, S. Nolte, F. von Alvensleben, A. Tu¨nnermann, Appl. Phys. A 63 (1996) 109. [4] J. Perrie`re, E. Millon, M. Chamarro, M. Morcrette, C. Andreazza, Appl. Phys. Lett. 78 (2001) 2949. [5] D. Perez, L.J. Lewis, Phys. Rev. Lett. 89 (2002) 255504. [6] D. Perez, L.J. Lewis, Phys. Rev. B 67 (2003) 184102. [7] L.V. Zhigilei, Appl. Phys. A 76 (2003) 339. [8] S. Amoruso, G. Ausanio, R. Bruzzese, M. Vitiello, X. Wang, Phys. Rev. B 71 (2005) 033406. [9] J. Perrie`re, C. Boulmer-Leborgne, R. Benzerga, S. Tricot, J. Phys. D: Appl. Phys. 40 (2007) 7069. [10] S. Amoruso, R. Bruzzese, X. Wang, N.N. Nedialkov, P.A. Atanasov, J. Phys. D: Appl. Phys. 40 (2007) 331. [11] O. Albert, S. Roger, Y. Glinec, J.C. Loulergue, J. Etchepare, C. Boumer-Leborgne, J. Perrie`re, E. Millon, Appl. Phys. A 76 (2003) 319. [12] R. Teghil, L. D’Alessio, A. De Bonis, A. Galasso, P. Villani, A. Santagata, Thin Solid Films 515 (2006) 1411. [13] S. Amoruso, R. Bruzzese, M. Vitiello, N.N. Nedialkov, P.A. Atanasov, J. Appl. Phys. 98 (2005) 044907. [14] S. Eliezer, N. Eliaz, E. Grossman, D. Fisher, I. Gouzman, Z. Henis, S. Pecker, Y. Horovitz, M. Fraenkel, S. Maman, Y. Lereah, Phys. Rev. B 69 (2004) 144119. [15] B. Liu, Z. Hu, Y. Che, Y. Chen, X. Pan, Appl. Phys. Lett. 90 (2007) 044103. [16] S. Amoruso, G. Ausanio, A.C. Barone, R. Bruzzese, C. Campana, X. Wang, Appl. Surf. Sci. 254 (2007) 1012. [17] S. Amoruso, R. Bruzzese, N. Spinelli, R. Velotta, M. Vitiello, X. Wang, G. Ausanio, V. Iannotti, L. Lanotte, Appl. Phys. Lett. 84 (2004) 4502. [18] T.W. Trelenberg, L.N. Dinh, C.K. Saw, B.C. Stuart, M. Balooch, Appl. Surf. Sci. 221 (2004) 364. [19] Y. Zhang, R.E. Russo, S.S. Mao, Appl. Phys. Lett. 87 (2005) 133115. [20] M. Okoshi, K. Higashikawa, M. Hanabusa, Jpn. J. Appl. Phys. 40 (2001) 1287. [21] X.L. Tong, D.S. Jiang, Y. Li, Z.M. Liu, M.Z. Luo, Physica B 382 (2006) 105. [22] X. Chen, S.S. Mao, Chem. Rev. 107 (2007) 2891. [23] L. Kavan, M. Gratzel, S.E. Gilbert, C. Klemenz, H.J. Scheel, J. Am. Chem. Soc. 118 (1996) 6716. [24] C.D. Wagner, L.E. Davis, M.V. Zeller, J.A. Taylor, R.M. Raymond, L.H. Gale, Surf. Interf. Anal. 3 (1981) 211. [25] J. Jandeleit, G. Urbasch, H. Hoffmann, H.G. Treusch, E. Kreutz, Appl. Phys. A 63 (1996) 117. [26] M. Walczak, M. Sanz, M. Oujja, J.F. Marco, M. Castillejo, Appl. Phys. A 98 (2008) 735. [27] SPIPTM, Image Metrology (see http://www.imagemet.com/WebHelp/spip.htm). [28] J.F. Marco, A. Cuesta, M. Gracia, J.R. Gancedo, P. Panjan, D. Hanzel, Thin Solid Films 492 (2005) 158. [29] P.M. Kumar, S. Badrinarayanan, M. Sastry, Thin Solid Films 358 (2000) 122. [30] G.E. Jellison Jr., F.A. Modine, L.A. Boatner, Opt. Lett. 22 (1997) 1808. [31] E. Halary-Wagner, F. Wagner, P. Hoffmann, J. Electrochem. Soc. 151 (2004) C571.
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Femtosecond pulsed laser deposition of nanostructured TiO2 films Mikel Sanz a, Malgorzata Walczak a, Rebeca de Nalda a, Mohamed Oujja a, ˜ ares b, Marta Castillejo a,* Jose´ F. Marco a, Javier Rodriguez b, Jesu´s G. Izquierdo b, Luis Ban a b
Instituto de Quı´mica Fı´sica Rocasolano, CSIC, Serrano 119, 28006 Madrid, Spain Departamento de Quı´mica Fı´sica I, Facultad de Ciencias Quı´micas, Universidad, Complutense de Madrid, 28040 Madrid, Spain
A R T I C L E I N F O
A B S T R A C T
Article history:
Nanostructured deposits of TiO2 were grown on Si (1 0 0) substrates by laser ablating a TiO2 sintered target in vacuum or in oxygen using a Ti:sapphire laser delivering 80 fs pulses. The effect of the laser irradiation wavelength on the obtained nanostructures, was investigated using 800, 400 and 266 nm at different substrate temperatures and pressures of oxygen. The composition of the deposits was characterized using X-ray photoelectron spectroscopy (XPS) and the surface morphology was studied by environmental scanning electron microscopy (ESEM) and atomic force microscopy (AFM). Deposits are absent of microscopic droplets in all conditions explored. The best deposits, constituted by nanoparticles of an average diameter of 30 nm with a narrow size distribution, were obtained at the shorter laser wavelength of 266 nm under vacuum at substrate room temperature. ß 2008 Elsevier B.V. All rights reserved.
Available online 3 August 2008 Keywords: Femtosecond pulsed laser deposition Nanoparticles Nanostructured deposits TiO2
1. Introduction In recent years pulsed laser deposition (PLD) has proved to be an effective technique for the deposition of nanostructured films of a wide variety of materials [1] both for fundamental research and for technological applications [2]. The main drawback of this deposition method is the presence of microscopic particulates on the surfaces of the films which are present with high densities when the optical absorption coefficient of the target is small at the wavelength used for ablation. The origin of those particulates is not clear yet, but it is generally assumed that thermal processes, such as melting and boiling during the laser-target interaction, favour their ejection along with ionic and atomic species. Ultrashort femtosecond (fs) laser pulses offer a high material removal efficiency and high deposition rates of nanometer scale particles free of microscopic particulates and therefore fs PLD constitutes an attractive procedure for the fabrication of nanostructured deposits. However, it appears that the nature of nanoparticles (NPs) grown by fs PLD strongly depends on the material and deposition conditions and therefore various studies [3–10] have tried to elucidate the different processes involved, showing the possibility of using fs PLD as a general route to NP formation. Metal and semiconductor NPs have been synthesized by fs PLD [11–21]. However, the majority of these studies have been performed with light pulses centred around the peak wavelength
of Ti:sapphire laser at 800 nm and few reports have explored the effect of this parameter [16,20]. Analysis of the process over a broader range of wavelengths can provide important information about the NPs formation processes and serve as experimental tests for advanced theoretical models. TiO2 is a wide-band semiconductor with exceptional optical and electronic properties of wide use [22] as a photocatalytic air purification agent, for self-cleaning coatings, in environmental sanitation, and in photovoltaic devices for the fabrication of electrodes for solar energy conversion. Two of the three crystallographic phases of TiO2 namely anatase and rutile, with respective band gaps of 3.2 and 3.0 eV [23], are especially suitable for these applications where the performance of the material depends critically on its nanostructure. In this paper, we report on the growth of TiO2 deposits by PLD on Si (1 0 0) substrates from sintered TiO2 targets under different conditions. The effect of laser wavelength was studied under vacuum and in presence of oxygen at different substrate temperatures. Surface structures of prepared TiO2 films were observed using environmental scanning electron microscopy (ESEM) and atomic force microscopy (AFM). The chemical states of the material deposited were examined by X-ray photoelectron spectroscopy (XPS). To our knowledge, this is the first study of the dependence with laser wavelength of the properties of the TiO2 NPs created by fs PLD. 2. Experimental
* Corresponding author. Tel.: +34 91 5619400; fax: +34 91 5642431. E-mail address: [email protected] (M. Castillejo). 0169-4332/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2008.07.148
The experimental set up consists of a PLD stainless steel chamber evacuated by a turbomolecular pump down to
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
2 104 Pa. Deposition was carried out under vacuum or under a background pressure of oxygen of 5 Pa. TiO2 targets were prepared from a special grade rutile powder (Sigma Aldrich, particle size <5 mm) pelletized into disks. Targets were placed on a rotating sample holder to avoid crater formation. Pulses of 80 fs were produced by a Ti:sapphire amplified laser system (Spectra Physics) centred at 800 nm at a maximum repetition rate of 1 kHz. Pulses of 400 and 266 nm were generated by using a BBO-based frequency doubler/tripler. For the growing of the deposits, a frequency of 10 Hz has been used. The temporal profile of the pulses was diagnosed with a second harmonic generator autocorrelator. Energy control was performed by a half-wave plate/polarizer combination. The laser beam was focussed at 358 onto the target with a 20 cm focal length lens. The Si (1 0 0) substrates were ultrasonically degreased in acetone and methanol for 10 min. They were mounted on a heating element that allowed operating in the temperature range between room temperature (RT) and 700 8C and placed at a distance of 40 mm in front of the target. Deposits were grown in all cases with 7200 pulses at a repetition rate of 10 Hz. The modification of the irradiated area of the targets was studied by optical microscopy (Leica, S8APO) with a 160 microscope objective and equipped with a CCD camera. The surface morphology of the deposits was examined by ESEM (Philips XL30) and by AFM (Digital Instruments). All the AFM images were obtained in tapping mode and they were postprocessed using a third-order flattening routine. The deposits were also analyzed by XPS using a Leybold LHS10 XPS spectrometer, under an operating vacuum better than 1 106 Pa, using Al Ka radiation (130 W) and an analyser transmission energy of 200 and 50 eV for the wide and narrow scan spectra, respectively. The spectra were recorded at take-off angles of 908. All binding energy (BE) values were chargecorrected to the adventitious C 1s signal which was set at 284.6 eV and are accurate to 0.2 eV. Relative atomic concentrations were calculated using tabulated atomic sensitivity factors [24]. 3. Results and discussion 3.1. Modification thresholds In order to obtain the deposits under the same conditions for the different wavelengths of excitation, ablation of targets was performed at a laser fluence of 1.5 and 2 times the modification threshold fluence for each wavelength. For short pulse regime, the modification threshold fluence Fth depends on the material and the number of laser pulses N applied to the same spot [25]. For laser pulses with a Gaussian spatial beam profile, the maximum laser fluence, F, on the sample surface and the diameter, D, of the modified area are related by D2 = 2v0 ln(F/ Fth), where v0 is the 1/e2 radius of the beam distribution. The fluence at the target surface is the maximum calculated from the Gaussian beam radius and is related to the pulse energy by F ¼ E=pv20 . The diameter of the darkened modified area was determined trough observation by optical microscopy of the irradiated targets. From a plot of D2 versus ln E, Fth can be determined. As an example Fig. 1 shows the dependence of D2 with ln E upon irradiation with 10 pulses at the three different wavelengths. The obtained thresholds for 10 pulses are 140, 124, and 75 mJ/ cm2 for 800, 400 and 266 nm, respectively. The corresponding photon energies are respectively lower, around the same and larger than the optical bandgap of the target rutile material (3.0 eV, 413 nm) [23].
5207
Fig. 1. The squared diameter D2 of the modified (darkened) areas versus the logarithm of laser pulse energy, for 10 laser pulses upon irradiation at 266, 400 and 800 nm (squares, triangles and circles, respectively).
3.2. Surface structure The laser parameters, wavelength and fluence, gas atmosphere and substrate temperature were varied in order to determine their influence on the produced deposits. As observed by ESEM and AFM (see below) the deposited films contain principally NPs smaller than 50 nm and few larger particles with sizes ranging from 100 to 200 nm. This structure has also been observed in fs PLD with other materials [14,21]. The images do not show any droplets of micrometer size, as observed in the case of nanosecond (ns) PLD [26]. The thicknesses of the deposits are in the range of 20–60 nm, being considerably thinner than those obtained upon ns laser irradiation using the same number of pulses [26]. To characterize the density of particles we analyzed the deposits by ESEM (Fig. 2). For 266 nm (Fig. 2a), a high density is obtained, covering practically all the substrate. In the presence of oxygen (Fig. 2b) a similar density is produced but it decreases substantially when the substrate is heated up to 600 8C (Fig. 2c). It was also noted that in the range of laser fluences used to obtain the deposits (1.5 and 2 times the modification threshold fluence, corresponding to 112 and 150 mJ/cm2, respectively), changes in density or sizes of particles are not observed. Upon ablation at 400 nm, the number of particles with diameter 100 nm is higher than in the case of operation at 266 nm (Fig. 2d). When using oxygen (Fig. 2e) the total density of particles decreases while an increase is observed by deposition on a substrate heated at 600 8C (Fig. 2f), also showing a numerous population of particles with sizes in the range of 100–200 nm. For 800 nm, deposition in vacuum (Fig. 2g and h) leads to a lowdensity deposit. However, when the substrate is heated, their density increases considerably (Fig. 2i) as also observed at 400 nm. The surface morphology and topography was investigated in greater detail by AFM. As an example Fig. 3 shows an AFM image of a deposit grown by ablating the target at 266 nm in vacuum at RT. To obtain the characteristic sizes of the NPs in terms of the equivalent diameter, the images were analyzed by using Scanning Probe Image Processor program [27]. As reported with ESEM, analysis of AFM images revealed a very different influence of the gas atmosphere and substrate temperature depending on the excitation wavelength. For 266 nm, the most uniform deposits were obtained in vacuum at RT (Fig. 3). As mentioned before, a high density of nanoparticles is observed with a narrow size distribution, with average diameters of 30 nm and absence of particulates larger than 100 nm (Fig. 4a). If the deposits are grown in oxygen at
5208
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
Fig. 2. ESEM images of the surfaces of TiO2 thin films grown at the indicated deposition conditions. The bar size is 1 mm in all images.
RT, the obtained size distribution of NPs becomes wider and the average diameter increases slightly to 35 nm (Fig. 4b). Upon heating of the substrate to 600 8C in vacuum, no appreciable change is observed in the size distribution, although the ESEM results show a decrease in density. The observed NPs appear as islands of almost hemispherical shape, in contrast with the morphology obtained by 266 nm ns PLD of the same material [26]
where deposits are constituted by a nanostructured film with presence of microscopic particulates. The sizes of particulates grown at 400 and 800 nm are similar, with average diameter of 50 nm (Fig. 4c and d) although the particle density is higher at 400 nm. In vacuum, NPs of 20–50 nm are dominant, but some larger particles of 100 nm are also observed. The deposits obtained under oxygen atmosphere display
Fig. 3. AFM images of the nanostructures deposited under vacuum, at RT and 266 nm at F = 2Fth: (a) topography image, (b) phase image (c) 3D topography image.
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
5209
Fig. 4. Size histogram of the nanoparticles produced by fs laser ablation at F = 2Fth: (a) 266 nm in vacuum and RT, (b) 266 nm in oxygen and RT, (c) 400 nm in oxygen and RT and (d) 800 nm in vacuum, heating the substrate at 600 8C.
a richer content of small size particles (100 nm) (Fig. 4c), whereas by heating the substrate at 600 8C (Fig. 4d), the size distribution becomes wider, with particle diameters up to 200 nm. 3.3. Composition The Ti 2p and O 1s XPS spectra recorded from deposits obtained by ablating TiO2 targets are presented in Fig. 5. The Ti 2p spectrum corresponding to the film prepared at 266 nm shows two different spin-orbit doublets (Fig. 5a). The BE of the 2p3/2 and 2p1/2 core levels of the major doublet, which accounts for 86% of the total spectral area, 458.2 and 463.9 eV, respectively, are characteristic of TiO2 [28,29] and in fact, this component of the spectrum is identical to the Ti 2p spectrum recorded from the TiO2 target material (not shown). The minor doublet (14% of the total spectral area) is characterized by BEs of 456.1 and 461.8 eV of the Ti 2p3/2 and 2p1/2 core levels, respectively, and can be associated to the presence in the film of a Ti oxide in which Ti has an oxidation state lower than 4+ [29]. The O 1s spectrum recorded (not shown) from the film prepared at 266 nm shows four contributions corresponding to Ti–O bonds from the deposited material, Si–O bonds coming from the SiO2/Si substrate, and hydroxyl groups and water [28,29] which have been probably adsorbed on the surface of the film after exposure of the sample to the laboratory atmosphere. For excitation photon energies around or smaller than the bandgap (400 and 800 nm, respectively) the results are very similar. The Ti 2p spectrum shows (Fig. 5b) only one spin-orbit component, with BEs of the 2p3/2 and 2p1/2 core levels, which are, as mentioned above, characteristic of TiO2. The O 1s
spectrum recorded from these films displays the same four contributions as the case of 266 nm. Taking into account the relative areas of the main O 1s component (associated to Ti–O bonds) and the relative area of the main Ti 2p contribution, the corresponding O/Ti atomic ratio is 2.2, which is close, within the error of the experimental determination, to the expected stoichiometric ratio of 2.0. The results clearly show that the wavelength of the laser used for deposition has a strong effect on the characteristics of the NPs produced by fs PLD. The tendency of producing smaller NPs with a narrower size distribution at shorter wavelengths of the fs laser, as observed herein, has also been reported by other authors [16,20] and discussed in terms of the larger strain rate in the expansion of the plasma fluid into vacuum following irradiation of the target by UV pulses [16]. The photon excitation energy at 266 and 400 nm is sufficient to promote the electron to the conduction band of the semiconductor TiO2 material, while two photons are required at 800 nm. Upon 266 nm fs laser irradiation, and due to the high absorption coefficient at this wavelength (a 106 cm1 [30]), a thin layer below the surface of the target is suddenly heated up without changing its density. Irradiation proceeds under heat confinement conditions (the duration of the laser pulse is well below the heat diffusion time tpulse td = 1/a2Dh with Dh = 0.03 cm2 s1 the heat diffusion coefficient [31]) and once energy deposition is complete, the plasma expands adiabatically from an initially high temperature. As the plasma reaches densities close to the critical point in the phase diagram, phase decomposition takes place with high fragmentation leading to the observed small NP formation [5,9,14,17].
5210
M. Sanz et al. / Applied Surface Science 255 (2009) 5206–5210
about the processes of formation of nanoparticles as well as experimental tests for advanced theoretical models. Research continues to explore the crystalline quality of the deposits using micro X-ray diffraction dispersion with synchrotron radiation and to characterize the accompanying ablation plume by mass spectrometry and optical emission spectroscopy. Acknowledgements Funding from Spanish MEC (Projects CTQ2007-60177, CTQ2005-08493-C02-01) is gratefully acknowledged. M.W. thanks EU for a contract (MESTCT-2004-513915) and M.S. thanks CSIC for a I3P contract. We are grateful to D. Go´mez (Instituto de Ciencia y Tecnologı´a de Polı´meros, CSIC) for operating the ESEM and to F. Gonza´lez-Posada (Instituto de Sistemas Optoelectro´nicos y Microtecnologı´a, UPM) for AFM measurements. This research has been performed within the Unidad Asociada ‘‘Quı´mica Fı´sica Molecular’’ between Departamento de Quı´mica Fı´sica I of Universidad Complutense de Madrid (UCM) and CSIC. References
Fig. 5. Ti 2p XPS spectra recorded from the film deposited ablating the target with (a) 266 and (b) 800 nm. In all cases the deposits were produced at RT in oxygen.
4. Conclusions TiO2 nanostructured deposits on Si (1 0 0) substrates were obtained from sintered rutile targets at the wavelengths of 800, 400 and 266 nm from a fs Ti:sapphire laser. The surface structure of the material deposited was observed by ESEM and AFM and the chemical states by XPS. The deposits consist of TiO2 nanoparticles with grain size down to 20 nm with overimposed larger particles in the 100–200 nm range. The best deposits, with the highest density of nanoparticles of around 30 nm diameter and negligible amount of larger particulates (>100 nm), were grown by ablation at 266 nm in vacuum at RT. At this wavelength, the use of oxygen during deposition and the heating of the substrate stimulate the growth of larger particles. The highest concentrations of particulates were observed at the longest laser wavelengths, 400 and 800 nm by heating the substrate. These results indicate the importance of wavelength in fs PLD and provide important clues
[1] D.B. Chrisey, G.K. Huber (Eds.), Pulsed Laser Deposition of Thin Films, Wiley, New York, 1994. [2] R. Eason (Ed.), Pulsed Laser Deposition of Thin Films: Applications-Led Growth of Functional Materials, Wiley, New York, 2006. [3] B.N. Chichkov, C. Momma, S. Nolte, F. von Alvensleben, A. Tu¨nnermann, Appl. Phys. A 63 (1996) 109. [4] J. Perrie`re, E. Millon, M. Chamarro, M. Morcrette, C. Andreazza, Appl. Phys. Lett. 78 (2001) 2949. [5] D. Perez, L.J. Lewis, Phys. Rev. Lett. 89 (2002) 255504. [6] D. Perez, L.J. Lewis, Phys. Rev. B 67 (2003) 184102. [7] L.V. Zhigilei, Appl. Phys. A 76 (2003) 339. [8] S. Amoruso, G. Ausanio, R. Bruzzese, M. Vitiello, X. Wang, Phys. Rev. B 71 (2005) 033406. [9] J. Perrie`re, C. Boulmer-Leborgne, R. Benzerga, S. Tricot, J. Phys. D: Appl. Phys. 40 (2007) 7069. [10] S. Amoruso, R. Bruzzese, X. Wang, N.N. Nedialkov, P.A. Atanasov, J. Phys. D: Appl. Phys. 40 (2007) 331. [11] O. Albert, S. Roger, Y. Glinec, J.C. Loulergue, J. Etchepare, C. Boumer-Leborgne, J. Perrie`re, E. Millon, Appl. Phys. A 76 (2003) 319. [12] R. Teghil, L. D’Alessio, A. De Bonis, A. Galasso, P. Villani, A. Santagata, Thin Solid Films 515 (2006) 1411. [13] S. Amoruso, R. Bruzzese, M. Vitiello, N.N. Nedialkov, P.A. Atanasov, J. Appl. Phys. 98 (2005) 044907. [14] S. Eliezer, N. Eliaz, E. Grossman, D. Fisher, I. Gouzman, Z. Henis, S. Pecker, Y. Horovitz, M. Fraenkel, S. Maman, Y. Lereah, Phys. Rev. B 69 (2004) 144119. [15] B. Liu, Z. Hu, Y. Che, Y. Chen, X. Pan, Appl. Phys. Lett. 90 (2007) 044103. [16] S. Amoruso, G. Ausanio, A.C. Barone, R. Bruzzese, C. Campana, X. Wang, Appl. Surf. Sci. 254 (2007) 1012. [17] S. Amoruso, R. Bruzzese, N. Spinelli, R. Velotta, M. Vitiello, X. Wang, G. Ausanio, V. Iannotti, L. Lanotte, Appl. Phys. Lett. 84 (2004) 4502. [18] T.W. Trelenberg, L.N. Dinh, C.K. Saw, B.C. Stuart, M. Balooch, Appl. Surf. Sci. 221 (2004) 364. [19] Y. Zhang, R.E. Russo, S.S. Mao, Appl. Phys. Lett. 87 (2005) 133115. [20] M. Okoshi, K. Higashikawa, M. Hanabusa, Jpn. J. Appl. Phys. 40 (2001) 1287. [21] X.L. Tong, D.S. Jiang, Y. Li, Z.M. Liu, M.Z. Luo, Physica B 382 (2006) 105. [22] X. Chen, S.S. Mao, Chem. Rev. 107 (2007) 2891. [23] L. Kavan, M. Gratzel, S.E. Gilbert, C. Klemenz, H.J. Scheel, J. Am. Chem. Soc. 118 (1996) 6716. [24] C.D. Wagner, L.E. Davis, M.V. Zeller, J.A. Taylor, R.M. Raymond, L.H. Gale, Surf. Interf. Anal. 3 (1981) 211. [25] J. Jandeleit, G. Urbasch, H. Hoffmann, H.G. Treusch, E. Kreutz, Appl. Phys. A 63 (1996) 117. [26] M. Walczak, M. Sanz, M. Oujja, J.F. Marco, M. Castillejo, Appl. Phys. A 98 (2008) 735. [27] SPIPTM, Image Metrology (see http://www.imagemet.com/WebHelp/spip.htm). [28] J.F. Marco, A. Cuesta, M. Gracia, J.R. Gancedo, P. Panjan, D. Hanzel, Thin Solid Films 492 (2005) 158. [29] P.M. Kumar, S. Badrinarayanan, M. Sastry, Thin Solid Films 358 (2000) 122. [30] G.E. Jellison Jr., F.A. Modine, L.A. Boatner, Opt. Lett. 22 (1997) 1808. [31] E. Halary-Wagner, F. Wagner, P. Hoffmann, J. Electrochem. Soc. 151 (2004) C571.