pubs.acs.org/Langmuir © 2009 American Chemical Society
Galvanic Deposition of Pt Clusters on Silicon: Effect of HF Concentration and Application as Catalyst for Silicon Nanowire Growth Marta Cerruti,* Gregory Doerk, Gail Hernandez, Carlo Carraro, and Roya Maboudian Department of Chemical Engineering, University of California at Berkeley, Berkeley, California 94720 Received June 5, 2009. Revised Manuscript Received October 1, 2009 We report on the galvanic deposition of Pt on Si from solutions containing PtCl2 and different concentrations of HF. The results show that for low [HF]/[Pt] ratios (e26), only a thin layer of PtSi is formed. The deposition rate of Pt increases with [HF] in the plating solution, up to [HF]/[Pt] ∼ 530; after this ratio, the morphology of the Pt film changes: larger clusters are formed, which cover the Si substrate less densely. Detailed atomic force microscopy and X-ray photoelectron spectroscopy analyses show that the deposited Pt layers do not completely cover the Si substrate. The Pt and PtSi films formed are able to catalyze the formation of Si nanowires (Si NWs) arrays formed via vapor-liquid-solid (VLS) process. By changing the immersion time in the Pt plating solution, Si NWs arrays with different density, diameter, and orientation are obtained.
I. Introduction Controlled deposition of Pt thin films and nanoparticles promises to have important implications for emerging nanotechnologies. For instance, while Au is typically used as the catalyst in VLS growth of Si NWs, Au incorporated into the Si lattice can introduce deep-level traps in the band gap that typically poison any optical or electronic devices.1-3 Pt on the other hand does not adversely affect silicon electronics as much as Au does.4 In addition, Pt and PtSi are used as contacts for Si devices and are considered CMOS compatible.5 Pt is thus considered an attractive alternative catalyst to Au for Si nanowire growth. Galvanic displacement involves the deposition of a noble metal from a solution containing the metal salt onto a substrate with a reduction potential less than that of the metal.6 Under such conditions, the deposition of the metal is spontaneously favored, at the expense of the substrate which, in turn, is oxidized. This process has several advantages over other metal deposition schemes such as evaporation or sputtering: it is fast, easy to perform, relatively inexpensive, and allows the deposition on surfaces with difficult access without the need for high temperature. Moreover, it is a selective process. As an example, it occurs only on Si and not SiO2 or silicon nitride, which is particularly attractive for postprocessing of micro- and nanoelectromechani*Corresponding author. (1) Chen, J. W.; Milnes, A. G. Annu. Rev. Mater. Sci., 1980, 10, 157-228. (2) Allen, J. E.; Hemesath, E. R.; Perea, D. E.; Lensch-Falk, K. L., Li, Z. Y.; Gass, M. H.; Wang, P.; Bleloch, A. L.; Palmer, R. E.; Lauhon, L. J. Nat. Nanotechnol. 2008, 3, 168-173. (3) Guichard, A. R.; Barsic, D. N.; Sharma, S.; Kamins, T. I.; Brongersma, M. L. Nano Lett. 2006, 6, 2140-2144. (4) Lisiak, K. P.; Milnes, A. G. J. Appl. Phys. 1975, 46, 5229-5235. (5) Baron, T.; Gordon, M.; Dhalluin, F.; Ternon, C.; Ferret, P.; Gentile, P. Appl. Phys. Lett. 2006, 89, 233111. (6) Carraro, C.; Maboudian, R.; Magagnin, L. Surf. Sci. Rep. 2007, 62 (12), 499-525. (7) Carraro, C. Magagnin, L.; Maboudian, R. Electrochim. Acta 2002, 47, 2583-2588. (8) Lombardi, I.; Marchionna, S.; Zangari, G.; Pizzini, S. Langmuir 2007, 23, 12413-12420. (9) Brito-Neto, J. A.; Araki, S.; Hayase, M. J. Electrochem. Soc. 2006, 153, C741-C746. (10) Peng, K.; Zhu, J. Electrochim. Acta 2004, 49, 2563-2568. (11) Gorostiza, P.; Diaz, R.; Servat, J.; Sanz, F.; Morante, J. R. J. Electrochem. Soc. 1997, 144, 909-914.
432 DOI: 10.1021/la902032x
cal systems.7 Galvanic deposition has been demonstrated using a number of metals, including Au, Ag, Cu, Ni, and Pt.8-18 During the galvanic deposition, the plating solution with the salt of the metal to be deposited needs to contain also an agent that can remove the oxidized substrate formed during the deposition, otherwise a passivated layer would prevent further metal deposition6. For galvanic displacement performed on Si, this is achieved with the addition of HF in the plating solution. As an example, the overall reaction describing galvanic displacement of Pt (II) on Si is: 2Pt2 þ ðaqÞ þ Si0 ðsÞ þ 6F- ðaqÞh2Pt0 ðsÞ þ SiF6 2 - ðaqÞ
ð1Þ
Reaction of Si with HF yields soluble silicon hexafluoride if a molar ratio of at least 6:2 HF:Pt ions is present in solution. It has been observed that the reaction rate is increased with increase in concentration of HF in solution.15,16 However, using large amounts of HF is detrimental to many applications, because it can lead to significant dissolution of device parts made out of SiO2 or silicon nitride. A few studies have been reported on the effect of HF concentration in solution on the morphology and rate of Pt galvanic deposition.8,11 However, no systematic study has attempted to elucidate the overall rate of Pt deposition with varying HF concentrations. In addition, the molar excess of HF to Pt used in the plating solution in the literature vary from 159 to 2700.11,14 In this paper, we study the effect of a wide range of concentrations of HF on the galvanic displacement of Pt. In particular, we analyze the Pt deposition from solutions with molar excess of HF to Pt precursor varying from 26 to ∼8000 at different immersion (12) Gorostiza, P.; Diaz, R.; Servat, J.; Sanz, F.; Morante, J. R. J. Electroanal. Chem. 1999, 469:48-52. (13) Gorostiza, P.; Diaz, R.; Morante, J. R. J. Electrochem. Soc. 1997, 144, 4119-4122. (14) Gorostiza, P.; Servat, J.; Morante, J. R.; Sanz, F. Thin Solid Films 1996, 275 (2), 12-17. (15) Srinivasan, R.; Suni, I. I. J. Electrochem. Soc. 1999, 146 (2), 570-573. (16) Nagahara, L. A.; Ohmori, T.; Hashimoto, K.; Fujishima, A. J. Vac. Sci. Technol. A 1993, 11, 763-767. (17) Magagnin, L.; Maboudian, R.; Carraro, C. J. Phys. Chem. B 2002, 106, 401-407. (18) daRosa, C.; Iglesia, E.; Maboudian, R. J. Electrochem. Soc. 2008, 155, E70-E78.
Published on Web 10/19/2009
Langmuir 2010, 26(1), 432–437
Cerruti et al.
Article
times. We then apply the knowledge obtained from this study to control the diameter and density of Si NWs grown using galvanically deposited Pt as catalyst.
II. Materials and Methods A. Sample Preparation. Silicon substrates are mechanically cleaved from p-doped Si(100) wafers, with resistivity values of 10-30 Ω cm; a typical piece has sides measuring 4-10 mm in length. The samples are first degreased in acetone and sonicated for 10-15 min, then quickly rinsed in deionized water and dried with compressed nitrogen gas. The samples are then decontaminated with UV-ozonolysis (UVO-Cleaner, model n.42, Jelight Company, Inc.) for 15 min to remove the adventitious hydrocarbons on the surface. To remove the silicon dioxide layer formed on the substrate in the previous step, the samples are submerged in concentrated HF for 2 min, and then again quickly rinsed in deionized water and dried with a stream of nitrogen gas. Platinum solutions are prepared by dissolving 5.7 mg of platinum(II) chloride (analytical grade PtCl2,Aldrich) in 6.0 mL of water and 80 µL of ammonium hydroxide (reagent grade NH4OH, Fischer) resulting in a 3.5 mM concentration of PtCl2. The mixture is then sonicated for one to two hours and/or stirred with a stir bar overnight until a homogeneous solution is obtained. Next, the solution is transferred in a Teflon beaker, and 20 mL of water and varying amounts of HF are added, producing the desired ratio of HF to Pt precursor concentrations ([HF]/[Pt]). In the final solutions, [Pt] is varied between ∼0.65 and 0.8 mM. [HF] is varied from 0.0212 to 5.17 M, yielding [HF]/[Pt] varying from 26 to 7917. In a few cases, some HCl is added in order to lower the pH of the plating solution. Immediately after the Si treatment steps described in the previous paragraph, the samples are immersed in this solution for a time duration ranging from 30 s to 1 h. The solution is not stirred during the deposition, with the beaker covered with parafilm. Following immersion, the silicon samples are rinsed and dried with water and nitrogen, respectively. B. Sample Characterization. Contact angle experiments were performed using a Rame-Hart contact angle goniometer (model 100-00-115) equipped with a CCD camera and analyzed with the Rame-Hart software. Static contact angles were measured by the sessile drop method, using deionized (DI) water with drop volumes of 2 µL. The results shown in this paper are the average of at least three measurements performed on different samples. Standard deviations on these measurements varied between 1 and 5. X-ray photoelectron spectroscopy was performed in an ultrahigh vacuum chamber with base pressure of 10-9 Torr, equipped with an Omicrometer EA125 electron energy analyzer and an Omicrometer DAR400 source of Al KR X-rays at an energy of 1486.6 eV. The detector angle was 0 to the surface normal. Spectral deconvolution was performed after background subtraction with the Shirley method. Energy calibration was performed according to the commonly employed method of using the C 1s peak of hydrocarbon contamination as a reference.19 Atomic force microscopy was used for characterization of the topography of the samples. The AFM instrument (Multimode Nanoscope IIIa with Extender Electronics 7 Module, Digital Instruments) was operated in tapping mode using a Ti-Pt coated silicon cantilever (MikroMasch) with force constant, resonant frequency, and in-air Q-factor values of 4.5 N/m, 150 kHz (nominal), and ∼175, respectively. A 2.5 V peak-to-peak AC voltage at the resonant frequency of the cantilever was applied between the probe tip and the sample, and scanned at frequencies ranging from 0.5 to 1.0 Hz. Scanning electron microscopy images were taken using a Leo 1550 Schottky field-emission SEM. (19) Clark, D. T.; 16:791-820.
Thomas, H. R.
Langmuir 2010, 26(1), 432–437
J. Polym. Sci., Polym. Chem. 1978,
Figure 1. Water contact angles measured on samples prepared by galvanic deposition of Pt for times ranging from 30 s to 60 min, and [HF]/[Pt] ranging from 26 to 8000, as explained in the legend.
C. Silicon Nanowire Growth. Silicon nanowires were grown in a manner similar to that used for Au-catalyzed nanowires, described previously.20 Samples prepared with galvanically displaced Pt were loaded without further treatment into a hot-wall atmospheric pressure reaction chamber and heated to 970 C under 270 sccm flow of 10% H2 in Ar. After temperature stabilization growth was initiated by introducing 6 sccm of SiCl4 via a bubbler held at 0 C with 10% H2 in Ar as the carrier gas. The growth time was set at 10 min. Growth was terminated by stopping the flow of SiCl4. Immediately after, the gas flow was switched to Ar only and the reactor was allowed to cool to room temperature at a rate of approximately 40 C/min.
III. Results and Discussion Figure 1 shows contact angle data as a function of immersion time using plating solutions with ratios of [HF]/[Pt] ranging from 26 to 7900. As a comparison, HF-treated Si samples have contact angles of ∼75.21 All samples show a very quick decrease in contact angle measured within the first 5 min of deposition. This indicates that the surface of the substrate rapidly changes from being hydrophobic, as expected due to the hydrogen termination of Si treated with HF, to being hydrophilic. A plateau is reached after a deposition time of ∼15 min. Contact angle for Pt has been reported to be ∼22,22 although the value strongly depends on such factors as the methodology used for cleaning the surface and the morphology of the films. Nevertheless, the value is consistent with the plateaus measured for samples prepared under [HF]/ [Pt] > 26. The plateau measured for the sample prepared at the lowest [HF]/[Pt] is higher (∼40), thus suggesting that a complete transformation of the Si substrate to Pt is not achieved for this sample. As discussed later in the paper, XPS analysis indicates the formation on platinum silicide layer in this case which could be responsible for the higher water contact angle. Figure 2A represents the XPS spectra of the Pt 4f region recorded on samples prepared with [HF]/[Pt] = 26, showing the Pt 4f7/2 and 4f5/2 doublet. The positions of the 4f peaks remain constant at ∼73 and 76 eV for all the samples shown in Figure 2A. These peak locations are indicative of the presence of PtSi.14 On the other hand, a different behavior is observed for depositions from solutions with higher [HF]/[Pt]. Figure 2B depicts the XPS spectra recorded on samples prepared with [HF]/[Pt] = 528. The spectra show a shift in the location of the Pt 4f peaks at longer deposition times. This behavior indicates that during the initial (
