Applied Surface Science 256 (2010) 3369–3373
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Optimization of porous silicon preparation technology for SERS applications M.V. Chursanova a,*, L.P. Germash a, V.O. Yukhymchuk b, V.M. Dzhagan b, I.A. Khodasevich c, D. Cojoc d a
National Technical University of Ukraine, ‘‘Kyiv Polytechnic Institute’’, 37 Prospect Peremohy, 03056 Kyiv, Ukraine V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine, 45 Prospect Nauky, 03028 Kyiv, Ukraine c B.I. Stepanov Institute of Physics of National Academy of Sciences of Belarus, 68 Nezalezhnasti Ave., 220072 Minsk, Belarus d CNR-INFM, Laboratorio Nazionale TASC, Area Science Park - Basovizza, Edificio MM, S.S. 14 km 163,5, 34012 Trieste, Italy b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 May 2009 Received in revised form 25 November 2009 Accepted 12 December 2009 Available online 21 December 2009
A series of porous silicon samples prepared at different etching parameters, namely etchant composition, etching time and current density, was investigated as substrates for surface-enhanced Raman scattering (SERS). Silver nanostructures were deposited on porous silicon by immersion plating method and Rhodamine 6G was used as analyte. The relation between the etching parameters, morphology of porous silicon surface and its SERS efficiency after silver deposition is examined. We show that a high HF content in the etchant allows the formation of a film with close-packed silver nanocrystals, which possess strong surface enhancement properties. ß 2009 Elsevier B.V. All rights reserved.
Keywords: SERS substrate Porous silicon Silver crystallites Rhodamine 6G
1. Introduction Raman spectroscopy is one of the most promising analytical methods for detection and identification of chemical and biological substances, due to the correspondence of the vibrational Raman frequencies to certain types of chemical bonds. However, due to the very low cross-section of the Raman scattering process, the detection of substances of low concentration is complicated without special enhancement. This is why techniques to enhance the Raman scattering, such as the surface-enhanced Raman scattering (SERS), have been attracting great attention in the last years. It was demonstrated recently that Raman signal amplification through SERS even gives the possibility of single molecule detection [1,2]. One of the most popular methods of SERSsubstrates preparation in our days is deposition of Au or Ag nanoparticles from colloids. However, this method does not allow to obtain stable substrates with reproducible particle size [1]. The large-scale applications in chemistry, biology and medicine require SERS-active substrates with optimal efficiency/price relation. One of the promising candidates in this respect is formation of metallic nanocrystals structures on porous template, for example porous silicon (por-Si). Silver is probably the best choice of the metal for this purpose, due to its broad plasmon resonance in visible-near IR spectral range, stability and simple preparation procedure [3].
* Corresponding author. E-mail address: afi
[email protected] (M.V. Chursanova). 0169-4332/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2009.12.036
Porous materials were used as bases for SERS-active substrates preparation in several recent works [4–10], namely porous silicon was investigated as a template for silver nanostructures deposition. Due to its large surface area and open porous structure, this semiconductor material allows to obtain highly sensitive SERS substrates. This can be achieved by silver nanostructures synthesis on porous silicon surface [4–8] or homogenous coating of pore walls with metal layer [9]. Morphology of the final SERS-active film depends on the porous silicon substrate structure, which, in its turn, is determined by silicon etching conditions. The process of porous silicon formation is very complex and still not completely understood and influenced by numerous factors. However, in previous studies usually only one of the technological parameters was varied during the porous silicon fabrication. Our present investigation is aimed at finding of the technological conditions of por-Si preparation optimal for application as SERS substrate. Immersion of porous silicon into solution containing metal ions such as Ag+ leads to a spontaneous formation of silver nanoparticles by means of silver ions reduction through Si–H bonds on silicon surface [5,8]: 2Siðsurf:Þ þ H2 O ! SiOSiðsurf:Þ þ 2Hþ ðsolutionÞ þ 2e ;
(1a)
2SiHðsurf:Þ þ H2 O ! SiOSiðsurf:Þ þ 4Hþ ðsolutionÞ þ 4e ;
(1b)
Agþ ðsolutionÞ þ e ! Agðsurf:Þ
(1c)
The morphology of por-Si provides a large density of appropriate sites for nucleation and growth of silver nanocrystals
3370
M.V. Chursanova et al. / Applied Surface Science 256 (2010) 3369–3373
[8]. The morphology and chemical composition of por-Si surface has strong influence on the parameters of silver nanostructures formed on it and, consequently, on the enhancement of Raman signal of adsorbed molecules. The por-Si surface morphology is determined by crystal orientation, doping degree and type, current density and etching time, additional illumination, temperature, electrolyte composition, etc. and therefore is quite difficult to trace the particular influence of each of them. Establishing dynamics of the relation between the parameters of porous silicon preparation and its surface enhancing properties after silver deposition represents the aim of our work, of which results are presented hereafter. 2. Material and methods Porous silicon samples were prepared by electrochemical etching of p-type Si(1 0 0) plates with specific resistance r = 10 V cm. Solution of 40% HF in ethanol in proportions HF:C2H5OH as 1:2 or 2:1 was used as etchant. The series of porSi samples was obtained at values of current density (J) of 5, 10, or 20 mA/cm2 and etching time (t) of 10, 20, or 30 min. Silver nanostructures were deposited on por-Si by immersion plating in 102M AgNO3 solution during t = 5, 10 or 15 min. The morphology of por-Si with Ag nanoformations was studied by scanning electron microscopy (SEM). The SERS efficiency of the obtained metallized por-Si samples was investigated with using 105M Rhodamine 6G as analyte. Raman spectra were measured with the Renishaw Ramanscope 2000, using 514 nm line of Ar+-ion laser for excitation. All the spectra were recorded at similar conditions (excitation laser power, accumulation time) and corrected for the luminescence background. 3. Results and discussion First, we have determined optimal silver deposition time on porous silicon samples. The variation of the silver deposition duration was performed in the range from 5 to 15 min. As it was shown in work [7], only single Ag nanoparticles form at deposition time about 5 min. But as time goes up to 10 min and longer, they tend to self-assemble into larger and more complicated nanostructures like nanoclusters or dendrites which provide strong SERS effect [4,7,8]. It was found in our investigation that the noticeable signal enhancement takes place when silver deposition time is around 10–15 min in 102M AgNO3 solution. The maximum enhancement is observed for 10 min (Fig. 1), deposition time of 15 min is also better than 5 min. This means that in order to form stable, welldeveloped nanostructures and achieve optimal sizes of nanoparticles immersion for about 10 min is required. This observation is in qualitative agreement with the 15 min optimal deposition time as observed in Ref. [8].
Fig. 1. Raman spectra of 105M Rh6G on porous silicon samples with silver nanoparticles deposited from 102M AgNO3 solution depending on deposition time t. Etchant HF:C2H5OH = 1:2, etching time t = 20 min, current density J = 10 mA/cm2.
Influence of different silicon etching factors is investigated next. Etching time defines mainly pore depth. SEM images show that sample etched during 10 min has numerous nanopores separated by narrow macropores (Fig. 2a). Macropores (macrocracks) become wider and their number grows as etching time increases to 20 (Fig. 2b) and 30 min (Fig. 2c) and silver nanoparticles of 50– 100 nm size can form on their walls. Fig. 3 shows that increase of etching time from 10 to 20 min leads to increase of the Raman scattering intensity from the analyte molecule. Therefore, we can conclude about the importance of macropores for effective surface enhancement. However, no noticeable difference in enhancement is observed between 20 and 30 min. This means that at given silver deposition conditions Ag nanostructures do not enter in depth of nanopores and form on the surface and macropore walls. Thus, the role of macrocracks is diminished at long duration times; probably the reason is the optimum size-to-surface area ratio of the cracks. Previous works [5,11] state that in the procedure of porous silicon etching increase of current density causes increase of sample porosity and average pore sizes. Current density defines the amount of holes (per area unit) directed to the silicon surface and, consequently, influences the pore width and enlarges total surface area of porous silicon. In our experiments we observe that sample with smaller current density (5 vs 10 mA/cm2) has much wider cracks, which can be related to the less stable structure of the porous layer obtained at 5 mA/cm2 (Fig. 4a and b). At the same time, the sample obtained at 10 mA/cm2 reveals better structural stability (Fig. 4a) and better SERS efficiency (Fig. 5). Raman spectra in Fig. 5 show that signal enhancement is higher for larger values of etching current density for both the etchant compositions used. HF concentration in the etchant is known to have a strong influence on the surface morphology of por-Si [11]. The porosity
Fig. 2. SEM images of the por-Si surface after Ag deposition. Etching time of por-Si was: 10 (a), 20 (b) and 30 min (c). J = 10 mA/cm2, t = 10 min, HF:C2H5OH = 1:2 in all cases.
M.V. Chursanova et al. / Applied Surface Science 256 (2010) 3369–3373
Fig. 3. Raman spectra of 105M Rh6G on samples etched during 10, 20 and 30 min. Silver nanoparticles were deposited from 102M AgNO3 solution on porous silicon, etchant HF:C2H5OH = 1:2, J = 10 mA/cm2, t = 10 min during 10 min for substrates.
decreases with increase of HF concentration and macropores are formed instead of nanopores. When concentration is above 25%, not only pores but also micrometer size cracks are present on the surface. As was noted above, the pores and cracks of micrometer scale appear to be of advantage for SERS-active substrates preparation because big enough silver nanoparticles can form on their walls. At the same time these silver nanoparticles do not block pores and effective surface of the sample increases. Samples prepared at different HF concentration, in our study, also show significant distinctions of the final morphology. Electrochemical treatment in electrolytes of different composition results in formation of different morphology and chemical structure of por-Si surface, which can strongly influence the process of Ag nanostructures formation. At low HF content (HF:C2H5OH = 1:2) only separate silver nanoaggregates form (Fig. 4b), while larger HF content (HF:C2H5OH = 2:1) in the etchant resulted in formation of a densely structured nanocrystal film (Fig. 4c). The reason may be the larger amount of Si–H bonds formed on the por-Si surface prepared at high HF concentration. The number of Si–H bonds plays an important role in silver reduction process during immersion of por-Si into AgNO3 solution. Also, as it is mentioned in Ref. [7], treatment in solution with high HF concentration causes decrease of por-Si surface energy, which leads to the fast formation of well-developed silver nanostructures on it. SEM-image of one of the samples prepared at high HF concentration (Fig. 4c) shows a film of closely packed silver crystallites of 0.1–2 mm size. Fig. 6 shows the Raman spectra of Rh6G on the structured silver surface on por-Si obtained at different etching conditions. No
3371
Fig. 5. Raman spectra of 105M Rh6G on samples obtained by silver deposition during t = 15 min. Etching parameters por-Si: J = 5 (thin curves) and 10 mA/cm2 (thick curves), etching time t = 20 min. Background level is shifted for visual clearness.
Fig. 6. Raman spectra of 105M Rh6G on samples obtained by silver deposition during t = 5 (thin curves) and 15 min (thick curves) at different etchant content: HF:C2H5OH = 1:2 for lower curves and HF:C2H5OH = 2:1 for upper curves. Etching parameters: J = 10 mA/cm2, t = 20 min. Background level is shifted for visual clearness.
signal of Rh6G on the initial (without Ag) por-Si was registered at the same concentration of analyte (spectra not shown). Similar to those shown in Fig. 4c, closely packed structure of silver nanocrystals has formed on the second series of por-Si samples which was etched in HF:C2H5OH = 2:1 etchant but stayed for about 2 years in air atmosphere and then refreshed right before silver deposition by putting into HF solution for several seconds (Fig. 7a). This shows stability of porous silicon substrate properties.
Fig. 4. SEM images of the por-Si surface after Ag deposition. The por-Si was prepared at (a) current density J = 10 mA/cm2 and HF:C2H5OH = 1:2; (b) J = 5 mA/cm2 and HF:C2H5OH = 1:2; (c) J = 5 mA/cm2 and etchant composition HF:C2H5OH = 2:1. Etching time t = 20 min in all cases.
3372
M.V. Chursanova et al. / Applied Surface Science 256 (2010) 3369–3373
Fig. 7. SEM images of the por-Si surface after Ag deposition. The por-Si was prepared at (a) etchant composition HF:C2H5OH = 2:1; (b) HF:C2H5OH = 1:2. Etching time t = 20 min and current density J = 10 mA/cm2 in both cases. Samples were stored for about 2 years in air atmosphere before silver deposition.
Fig. 8. Comparison of the morphology of the silver crystallites deposited onto fresh (a) and 2 years stored por-Si (b). The por-Si was prepared at etching time t = 20 min and current density J = 10 mA/cm2, etchant composition HF:C2H5OH = 2:1, silver deposition time t = 10 min in both cases.
Even after long-term storage por-Si prepared at high HF concentration allows to obtain highly sensitive SERS-active film of Ag nanocrystals by immersion plating, which is shown in Fig. 8. Raman spectra of 105M Rh6G (Fig. 9) measured on these samples also demonstrate strong surface enhancement by such nanocrystal film. It is to be mentioned that the strongest electric field enhancement occurs exactly between sharp faces of closely positioned metal particles [5]. Due to the dense arrangement of silver crystals, large local electric field concentrates in gaps between their faces, resulting in strong enhancement of Raman signal. The Ag crystallites cover por-Si surface with a continuous layer and ‘‘bulk’’ (nano-)porosity of the sample does not affect its properties noticeably. Probably, only hydride-reducing bonds on Si surface, which lie in the upper (outer) part of the pores and between the pores (i.e., actually cover the sample surface), play a role for such film formation, resulting in higher enhancement of Raman signal. Similar results were obtained in Ref. [5], where por-Si was etched in 3:1 HF:ethanol solution and silver dendrites were formed on its surface during immersion plating. In Ref. [7] it was demonstrated that at proper technological conditions silver assembles into the fractal-like structures which also provide strong SERS effect. Silver deposition process flew slowly for samples prepared in the etchant with low HF concentration. Only separate nanoparticles of 50–70 nm size were formed and positioned quite sparsely on the top surface and walls of the pores (Figs. 4b and 7b). This is in agreement with Ref. [8] where por-Si was etched at low HF concentration (HF:H2O:C2H5OH = 1:1:3) and only separate poly-
dispersed Ag nanoparticles aggregated on its surface after silver deposition. Such silver nanostructure parameters are not optimal for surface enhancement. Raman spectra measurements have shown that high HF concentration in etchant results in formation of substrates with stronger surface enhancement. Figs. 6 and 9 clearly display that at the same other preparation conditions (current density, etching time and duration of silver deposition)
Fig. 9. Raman spectra of 105M Rh6G on samples obtained by silver deposition during t = 10 min on 2 years old porous silicon at different etchant content: HF:C2H5OH = 1:2 for lower curve and HF:C2H5OH = 2:1 for upper curve. Etching parameters: J = 10 mA/cm2, t = 20 min.
M.V. Chursanova et al. / Applied Surface Science 256 (2010) 3369–3373
3373
Raman signal is several times stronger for samples produced in etchant HF:C2H5OH = 2:1 compared to those etched with HF:C2H5OH = 1:2.
Stepanov Institute of physics of National academy of sciences of Belarus) for assistance with samples preparation and fruitful discussion.
4. Conclusions
References
Our study reveals how por-Si preparation parameters, such as HF concentration in etchant, current density, duration of etching and silver deposition time, influence morphology of obtained por-Si/Ag substrates regarding to their SERS application. The high HF concentration (exceeding 25%) in etchant during por-Si preparation was found to result in formation of a densely packed film of silver crystallites. Por-Si samples preserve this property even at long-term storage before silver deposition. Resulting silver nanocrystal film appears to give extremely high surface enhancement of the Raman signal. Though the enhancement at lower HF concentration was somewhat smaller, those substrates are more mechanically stable and thus promising for multiple using. In the present report, we show the possibility to obtain silver micro- and nanocrystals of a compact faceted shape, revealing strong SERS properties. The enhancement of the Raman signal of the analyte molecules in the present case is supposed to be due to the high surface density of the crystallites and multiple edges at their surface.
[1] F. Yan, T. Vo-Dinh, Surface-enhanced Raman scattering detection of chemical and biological agents using a portable Raman integrated tunable sensor, Sensors and Actuators B 121 (2007) 61–66. [2] K. Kneipp, M. Moskovits, H. Kneipp (Eds.), Surface-Enhanced Raman Scattering: Physics and Applications, Springer, 2006. [3] M.V. Cacamares, J.V. Garcia-Ramos, S. Sanchez-Cortes, M. Castillejo, M. Oujja, Comparative SERS effectiveness of silver nanoparticles prepared by different methods: a study of the enhancement factor and the interfacial properties, Journal of Colloid and Interface Science 326 (2008) 103–109. [4] A.Yu. Panarin, V.S. Chirvony, K.I. Kholostov, P.-Y. Turpin, S.N. Terekhov, Formation of SERS-active silver structures on the surface of mesoporous silicon, Journal of Applied Spectroscopy 76 (2) (2009) 280. [5] H. Lin, J. Mock, D. Smith, T. Gao, M.J. Sailor, Surface-enhanced Raman scattering from silver-plated porous silicon, Journal of Physical Chemistry B 108 (31) (2004) 11654–11659. [6] F.A. Harraz, T. Tsuboi, J. Sasano, T. Sakka, Y.H. Ogata, Metal deposition onto porous silicon layer by immersion plating from aqueous and nonaqueous solutions, Journal of the Electrochemical Society 149 (9) (2002) C456–C463. [7] W. Ye, C. Shen, J. Tian, C. Wang, L. Bao, H. Gao, Self-assembled synthesis of SERSactive silver dendrites and photoluminescence properties of a thin porous silicon layer, Electrochemistry Communications 10 (2008) 625–629. [8] F. Giorgis, E. Descrovi, A. Chiodoni, E. Froner, M. Scarpa, A. Venturello, F. Geobaldo, Porous silicon as efficient surface-enhanced Raman scattering (SERS) substrate, Applied Surface Science 254 (2008) 7494–7497. [9] S. Chan, S. Kwon, T.-W. Koo, L.P. Lee, A.A. Berlin, Surface-enhanced Raman scattering of small molecules from silver-coated silicon nanopores, Advanced Materials 15 (19) (2003) 1595–1598. [10] Z. Pan, A. Zavalin, A. Ueda, M. Guo, M. Groza, A. Burger, R. Mu, S.H. Morgan, Surface-enhanced Raman spectroscopy using silver-coated porous glass-ceramic substrates, Applied Spectroscopy 59 (6) (2005) 782–786. [11] J. Dian, A. Macek, D. Niznansky, I. Nemec, V. Vrkoslav, T. Chvojka, I. Jelınek, SEM and HRTEM study of porous silicon—relationship between fabrication, morphology and optical properties, Applied Surface Science 238 (2004) 169–174.
Acknowledgements Authors express their gratitude to B.M. Bulakh for samples of por-Si preparation (V. Laskaryov Institute of semiconductor physics of NAS Ukraine), A.Yu. Panarin and S.N. Terekhov (B.I.