We study the photoinduced deposition of Ag nanoparticles on Fe-doped LiNbO3 crystals by using an in situ probe. SEM and XPS analysis show that Ag metallic nanoparticles are close-packedly deposited on the crystal surface. Both 405 and 532nm laser are found effective for the photo-induced Ag deposition. The Ag deposition on the y-surface shows obvious anisotropy as compared to that on the + z-surface. Moreover, the Ag deposition on the + z-surface is found to start easily surrounding the focal point of the laser rather than at its center. Both photogalvanic and diffusion effects of photo-excited electrons are suggested to account for these features.
© 2014 Optical Society of America
Lithium niobate (LN) is considered as a promising substrate material for a medically used lab-on-chip platform because of its remarkable electro-optical, ferroelectric, piezoelectric and nonlinear optical properties. For example, the exploitation of surface acoustic wave (SAW) actuators built in LN substrates could increase the efficiency of microfluidic platforms in handling small sample volumes without the need for external pumping or different mechanisms that could cause modifications in the on-chip environment. Besides the fluidic actuator, the functionality of analyzing a single molecule  is also crucial to a LN-based microfluidic lab-on-chip platform. Nanopore-based sensors or surface enhanced Raman scattering (SERS) technology may be utilized for analyzing a single molecule, and silver (Ag) nanoparticles (NPs) are of considerable interest for application in biochemical sensors and SERS technology [2–4].
In the past few years, the selective deposition of Ag NPs was realized in LN, by exploiting the photoinduced reduction of Ag+ on a polarization-patterned or chemically patterned template [4–11]. In these works, the polarization or chemical patterns were firstly fabricated via electrical poling or proton exchange, and then Ag NPs were deposited selectively on the LN surface through photoinduced reduction according to the predefined pattern. Hanson et al. firstly reported the fabrication of silver NPs on a domain-patterned lithium niobate template by inducing photoinduced reduction in an aqueous solution . Then Sun et al. carefully studied the wavelength and intensity dependence of the photoinduced deposition (PID) of Ag NPs on nominally pure periodically poled lithium niobate (PPLN) [6,7]. They all suggested that the photo-excited electrons are driven, by the near-surface electric field stemming from the external screening of opposite polarization (or from the internal defect-related screening of the polarization), to the domain boundaries (or the domain surface) and enhance the formation of Ag NPs there [5–7]. The similar role of the electric field was also emphasized by Carville et al. for the selective deposition of Ag NPs on chemically patterned LN substrates. In their study, the LN was periodically proton-exchanged (PE), which leads to the modulation of the surface polarization and the enhanced electrical field at the boundary between the PE and LN regions [4,8]. Besides the electric field, the photogalvanic force was also found responsible for the migration of photo-excited electrons and the subsequent PID of Ag NPs. Liu et al. reported for the first time that visible light with a wavelength of 410 nm can induce Ag deposition on the + z-surface of Mn-doped LN [9,10]. Their results clearly showed that photorefractive doping may modify the photo-excitation energy required by the PID of Ag NPs, and in particular that, different transport mechanisms of photo-excited electrons should be considered for the PID of Ag NPs on LN substrates doped with photorefractive elements.
In this work, LN crystals, doped with another important photorefractive element Fe, are tried as the substrate for the PID of Ag NPs, and the selective deposition is realized through the focused laser control. Both of two below-band-gap excitation wavelengths (405 and 532nm) are found effective for the PID of Ag NPs on the Fe-doped LN surface. By using a real-time imaging system, the dynamic PID of Ag NPs is visualized. Some features associated with photogalvanic and diffusion effects of photo-excited electrons will be shown and explained.
2. Experimental procedures and results
In our experiment, 0.45mm-thick z-cut and 1.02mm-thick y-cut LN crystals doped with 0.03 wt% Fe2O3 were used. Figure 1(a) shows the typical absorption spectrum of Fe-doped LN samples. The dopant Fe introduces electron traps (FeLi2+/3+) in the LN lattice and they induce a strong absorption around 2.5 eV. Under the illumination (below 600 nm), the electrons could be easily photoexcited from Fe2+/3+ traps and transfer toward the + z-surface (i.e. photogalvanic effect). Figure 1(b) shows the scheme of the experimental setup. The laser beam (405 nm) was focused, through the droplet of silver nitrate (AgNO3) solution, onto the + z-surface of Fe-doped LN by a microscope objective (10 × ). White light was used as a background to ensure the in situ probe of the deposition process through the microscope system.
Sample preparation and PID procedures are as follows: the crystal was sonicated for 20 min firstly in isopropanol, and then for 20 min in deionized water. After cleaning, the crystal was placed on a glass holder, and the + z-surface of LN was covered with a droplet of 0.01M AgNO3 solution. The 405-nm-laser beam was focused directly onto the + z-surface of the crystal. The diameter of focus spot was about 40 μm, which is roughly measured by 1/e maximum and the intensity at the focal region is about 1.56 × 107 mW/cm2. We note that, the surface region outside the focal point are not absolutely dark because of the scattering of light of the main beam, the reflection caused by the sample surfaces and the background white light for the in situ probe. However, the total light intensity outside the focal point is definitely far smaller than that at the focus, and the outside region could be considered approximately as dark. During the deposition the exposure time of the 405-nm-laser was controlled, depending on the expected size of the deposition. After the deposition the sample was immersed in deionized water for 1 minute and then blown dry with nitrogen.
In order to obtain large-size deposition and to make easier the characterization of the deposition, long-time (480 minutes) exposure of the 405-nm-laser was performed during the deposition process on the + z-surface of Fe-doped LN crystals. Figure 2 shows the topographic images of the deposition product. (a) and (b) are microscope images (Olympus STM6) while (c) is obtained by SEM (Hitachi S-4800). Figure 2(a) shows that the deposition happens in a round region indicating the deposition is controlled spatially by the focal shape of the laser. However, the deposition in the central part shown in Fig. 2(b) is quite particular, and a ring-like ridge can be observed. Figure 2(c) reveals the microscopic structure of the deposition: close-packed NPs, ranging from tens to hundreds of nanometers in diameter.
To analyze the chemical composition of the deposition product, XPS data was collected by Kratos Axis Ultra DLD multi-technique X-ray photoelectron spectroscopy. The X-rays were generated using a Mg anode Kα source with photon energy of 1253.68 eV. The XPS high-resolution spectrum obtained from the deposition product is shown in Fig. 2(d). Two peaks can be observed at BEs (Binding Energys) of 374 and 368 eV, with a difference of 6eV. According to the previous investigations [12,13], they can be assigned to the 3d peaks of Ag metal, indicating the NPs are composed of metallic Ag.
As the electrons at the doping defects (FeLi)2+/3+ are usually sensitive to green light , a 532nm-laser was also tried for the PID of Ag NPs on the + z-surface of Fe-doped LN. The image of 532nm-laser-induced Ag deposition with an exposure time of 20 min is shown in Fig. 3(a), where the deposition regions “a” and “b” are produced with the focal intensities of 2.8 × 106 and 5.6 × 106 mW/cm2, respectively. It can be seen from Fig. 3(a) that scattered NP clusters are formed under 532nm-illumination and their density increases with the focal intensity. The result reveals that for Fe-doped LN, photo energy of 2.3 eV is sufficient to induce the deposition of Ag NPs, and it is much lower than those required by Mn-doped LN and nominally pure LN. The lower NP density in Fig. 3(a) as compared with that in Fig. 2(a) is due to the shorter exposure time, lower focal intensity and photon energy.
The PID of Ag NPs was also performed, with 532nm-laser exposure time of 480 min and a focal intensity of 2.8 × 106 mW/cm2, on the y-surface of Fe-doped LN. The profile (see Fig. 3(b)) of Ag deposition in this case exhibits huge anisotropy as compared to that on the + z-surface of Fe-doped LN (see Fig. 3 (a)). The Ag NPs accumulate heavily on one side but scarcely distribute on the other side. This result seems unexpected but could be explained in the following paragraph from the distribution of photo-excited electrons in Fe-doped LN.
In order to study in detail the PID of Ag NPs on the + z-surface of Fe-doped LN, the dynamic deposition process was recorded visually by the in situ probe. Figure 4 shows six images taken at different times during the 405-nm-laser exposure. In Fig. 4(a), where the 405-nm-laser exposure is just performed, no clear deposition can be observed except a black bubble produced by the photochemical reaction connected with Ag reduction. From Fig. 4 (b) to (c), the Ag NPs (dark imprint) gradually spread on the + z-surface of the crystal. However, these NPs tend to form surrounding the focal point of the laser, leaving its central part untouched (see the hollow deposition pattern in Fig. 4(c)). In Fig. 4(d), the hollow deposition pattern becomes rather clear, and more interestingly, a small black bubble is forming at the boundary of the central hollow part. Figure 4(e) and (f) show the quick growth of the black bubble.
Our results show that the PID of Ag NPs on the + z-surface has a ring-like-ridge topography (see Fig. 2 (b)). Moreover, further study of the dynamic PID reveals that the Ag NPs tend to form surrounding the focal point of the laser rather than at its center. Since it is generally accepted that during the PID of Ag NPs the Ag+ ions in solution are mainly reduced by the electrons photo-excited from the crystal, the features shown by our results are more connected with the spatial distribution of the photo-excited electrons. Concerning how the photo-excited electrons transfer and distribute during the PID of Ag NPs, two mechanisms have been proposed: the electrons are driven by the near-surface electric field [5–7] or by the photogalvanic force [9,10]. Since LN crystal samples used in our work are neither periodically poled nor treated through any ion implantation, they have a low density of defects near the surface resulting in a local weak electric field . Thus being driven by the photogalvanic force is more likely the dominant mechanism in Fe-doped LN. However, this mechanism cannot entirely explain the features shown by our results, and other mechanisms should be considered. As the light intensity used in our case is large (106 ~107 mW/cm2 at the focal point) as compared with that used in Ref. 4-10, the density of photo-excited electrons is much higher in our case and the diffusion of photo-excited electrons should be considered. The features regarding the PID of Ag NPs on the + z-surface of Fe-doped LN could be explained in the following way (see Fig. 5(a)): the photo-excited electrons firstly transfer along the z axis to the surface, mainly driven by the photogalvanic force; Then they migrate on the surface, governed by the diffusion mechanism, from the illuminated zone to the dark zone; Finally, they accumulate at the boundary of the illuminated and dark zone and react with the Ag+ ions in the solution. We note that, a small black bubble connected with the reaction between Ag+ ions and electrons is observed to form at the boundary of the central hollow and dark zone (see Fig. 4(e)-(f)), which is evidence for the diffusion of photo-excited electrons. It should be noted here that, the high intensity might not be necessary to observe Ag deposition, but it is indeed highly important to observe the diffusion-related dynamic PID process.
The PID of Ag NPs on the y-surface exhibits huge anisotropy as compared to that on the + z-surface of Fe-doped LN. This result reveals the dominant effect of photogalvanic force on the PID of Ag NPs. As plotted in Fig. 5(b), most of photo-excited electrons are driven by photogalvanic force to the + z-side, leaving the –z-side positively charged. As a result, the Ag Nps accumulate heavily on one side but distribute scarcely on the other side. In addition, some PID experiments were also tried on the -z-surface of Fe-doped LN. It is found that, under the same experimental condition, the deposition on the -z-surface is always much more difficult than that on the + z-surface, which is consistent with the dominant role of photogalvanic force in our case and with the experimental results reported in Ref. 6,7,9 and 10.
Besides the diffusion and photogalvanic effect, the pyroelectric effect may also contribute to the deposition of Ag NPs on the Fe-doped LN. As known, the LN doped with photorefractive elements shows a large pyroelectric effect. Liu et al. showed that through the pyroelectric effect the surface potential could be modified significantly by varying the local temperature . Moreover, Habicht et al. reported the deposition of charged PS micro-spheres on the particular domain of LN at elevated temperature . In our case, once Ag NPs form on the surface, the light reaching the surface is attenuated by Ag NPs and the optical excitation in the LN is reduced. However, the local surface may heat up due to the light absorption of Ag NPs, which leads to the decrease of spontaneous polarization charges (i.e. pyroelectric effect) and the overscreening of polarization charges, i.e. an excess of external negative screening charges. These excessive charges released from the LN surface may transfer into the Ag NPs and accelerate the growth of Ag NPs through the electrochemical reduction of the Ag+ ions in the solution. The growth of Ag NPs, the increase of light absorption, and the release of excessive negative screening charges build up a cycle and it may drive Ag NPs to grow until the local temperature reaches a constant. At the present time, a PID measurement with a monitor of the surface electrochemical potential is underway. The in situ probe of the electrochemical potential during the PID may bring some information about the photoexcited electron generation, the electron diffusion and charge transport during the PID.
We studied the photoinduced deposition of Ag nanoparticles on Fe-doped LN crystals. A real-time imaging system was developed to realize both the selective deposition of Ag NPs and the in situ probe of the deposition process under focused laser radiation. SEM and XPS analysis results show that Ag metallic nanoparticles, ranging from tens to hundreds of nanometers in diameter, are close-packedly deposited on the crystal surface. Both 405 and 532nm laser beams are found effective for photoinduced Ag deposition on the Fe-doped LN crystals. The dynamic deposition process was recorded visually. It is found that the Ag deposition on the y-surface shows obvious anisotropy as compared with that on the + z surface. Moreover, the deposition of Ag NPs on the + z-surface is found to start easily surrounding the focal point of the laser rather than its center. Both photogalvanic and diffusion effects of photo-excited electrons are suggested to account for the specific features of the photoinduced deposition of Ag NPs on the Fe-doped LN crystals.
We thank the referees for their valuable comments. This work is partly supported by NSFC (No. 61108060), Excellent Young Researcher Foundation of HEBUT (No. 2011001) and of HeBei Education Dep. (No.YQ2013029), Key Project of Chinese Ministry of Education (No. 212016), Project-sponsored by SRF for ROCS of SEM (2012), Hebei NSF (No. F2013202153), and Hebei Foundation for the introduction of oversea scholars in 2013.
References and links
1. N. Souza, “Single-cell methods,” Nat. Methods 9(1), 35 (2011). [CrossRef]
2. E. J. Bjerneld, F. Svedberg, P. Johansson, and M. Fäll, “Direct observation of heterogeneous photochemistry on aggregated Ag nanocrystals using Raman spectroscopy: The case of photoinduced degradation of aromatic amino acids,” J. Phys. Chem. A 108(19), 4187–4193 (2004). [CrossRef]
3. N. Leopold and B. Lendl, “On-column silver substrate synthesis and surface-enhanced Raman detection in capillary electrophoresis,” Anal. Bioanal. Chem. 396(6), 2341–2348 (2010). [CrossRef] [PubMed]
4. N. C. Carville, M. Manzo, S. Damm, M. Castiella, L. Collins, D. Denning, S. A. Weber, K. Gallo, J. H. Rice, and B. J. Rodriguez, “Photoreduction of SERS-active metallic nanostructures on chemically patterned ferroelectric crystals,” ACS Nano 6(8), 7373–7380 (2012). [CrossRef] [PubMed]
5. J. N. Hanson, B. J. Rodriguez, R. J. Nemanich, and A. Gruverman, “Fabrication of metallic nanowires on a ferroelectric template via photochemical reaction,” Nanotechnology 17(19), 4946–4949 (2006). [CrossRef]
6. Y. Sun and R. J. Nemanich, “Photoinduced Ag deposition on periodically poled lithium niobate: wavelength and polarization screening dependence,” J. Appl. Phys. 109(10), 104302 (2011). [CrossRef]
7. Y. Sun, B. S. Eller, and R. J. Nemanich, “Photoinduced Ag deposition on periodically poled lithium niobate: Concentration and intensity dependence,” J. Appl. Phys. 110(8), 084303 (2011). [CrossRef]
8. L. Balobaid, N. Carville, M. Manzo, K. Callo, and B. Rodriguez, “Direct shape control of photoreduced nanostructures on proton exchanged ferroelectric templates,” Appl. Phys. Lett. 102(4), 042908 (2013). [CrossRef]
9. X. Liu, K. Kitamura, K. Terabe, H. Hatano, and N. Ohashi, “Photocatalytic nanoparticle deposition on LiNbO3 nanodomain patterns via photovoltaic effect,” Appl. Phys. Lett. 91(4), 044101 (2007). [CrossRef]
10. X. Liu, H. Hatano, S. Takekawa, F. Ohuchi, and K. Kitamura, “Patterning of silver nanoparticles on visible light-sensitive Mn-doped lithium niobate photogalvanic crystals,” Appl. Phys. Lett. 99(5), 053102 (2011). [CrossRef]
11. S. Dunn and D. Tiwari, “Influence of ferroelectricity on the photoelectric effect of LiNbO3,” Appl. Phys. Lett. 93(9), 092905 (2008). [CrossRef]
12. S. Hüfner, G. Wertheim, and J. Wernick, “XPS core line asymmetries in metals,” Solid State Commun. 17(4), 417–422 (1975). [CrossRef]
13. G. Schön, J. Tummavuori, B. Lindström, C. R. Enzell, C. R. Enzell, and C.-G. Swahn, “ESCA Studies of Ag, Ag2O and AgO,” Acta Chem. Scand. 27, 2623–2633 (1973). [CrossRef]
14. H. Kurz, E. Krätzia, W. Keune, H. Engelmann, U. Gonser, B. Dischler, and A. Räuber, “Photorefractive centers in LiNbO3, studied by optical-, Mössbauer- and EPR-methods,” Appl. Phys. (Berl.) 12(4), 355–368 (1977). [CrossRef]
15. W. C. Yang, B. J. Rodriguez, A. Gruverman, and R. J. Nemanich, “Polarization-dependent electron afﬁnity of LiNbO3 surfaces,” Appl. Phys. Lett. 85(12), 2316 (2004). [CrossRef]
16. X. Liu, K. Kitamura, and K. Terabe, “Surface potential imaging of nanoscale LiNbO3 domains investigated by electrostatic force microscopy,” Appl. Phys. Lett. 89(13), 132905 (2006). [CrossRef]