Open Access
Add to BibSonomy
Abstract
We present an electrostatic-field-tunable ferroelectric template to produce photoreduced silver nanostructures for Raman scattering enhancement. The intensity and distribution of the surface electrostatic field in the ferroelectric template determine the morphology of the photoreduced silver nanostructures and thus the degree of the Raman signal enhancement. The surface electrostatic field is produced by periodically proton-exchanged (PPE) regions in LiNbO3 and is tuned by thermal annealing to obtain the favorable photoreduced silver nanostructure. The variation of surface electrostatic properties by thermal annealing is simulated using the finite element method and measured by electrostatic force microscopy. The mechanism of silver nanostructure formation affected by the electrostatic field distribution is discussed. The formed silver nanostructures are functionalized by R6G dye to enable Raman signal measurement. The proposed method is demonstrated to be effective in tuning the surface electrostatic field distribution and produces a 4.13 times higher silver nanostructure and a 2.51 times larger Raman intensity in comparison with the conventional PPE sample.
© 2017 Optical Society of America
1. Introduction
Surface enhanced Raman scattering (SERS) is a highly sensitive spectroscopic technique that can provide detailed molecular characterization information in chemical and biological analyses [1]. Its high-sensitivity and molecular-specificity abilities promote the development in scientific fields, such as diseases diagnosis [2], mutations in DNA sequences [3], pathogen quantification [4], explosive detection [5], and single-molecular detection [6]. The technical progress of SERS sensing comes from the advances in SERS signal enhancement by nanofabrication. SERS signal enhancement involves the stronger interaction between photons and metal nanostructure by the action of localized surface plasmon resonance (LSPR). Hence, producing ideal metal nanostructures to induce a strong localized electric field is significant for SERS signal enhancement.
Among the various nanostructure fabrication methods, ferroelectric lithography is promising owing to the features of low cost, simple fabrication, time effectiveness and reusable template. Ferroelectric materials have spontaneous polarization that can be reversed or tuned by electric poling [7–10] or chemical patterning [11–13]. The electrostatic field produced by positive/negative surface bound charges from the periodically modulated spontaneous polarization can be utilized to separate electron/hole pairs excited in ferroelectric materials under above-bandgap illumination. The photoelectrons moving to the substrate surface can produce photochemical silver reduction in an aqueous solution on the positive domain surface of ferroelectric materials to form the silver nanostructure. Ferroelectric materials, such as BaTiO3, PbZrxTi1-xO3(PZT), and LiNbO3 [14–16], have been adopted as ferroelectric templates. Among them, the LiNbO3 single crystal possesses the features of large spontaneous polarization and low defect concentration, and has the potential to produce high-quality metallic nanostructure for SERS signal enhancement.
The degree by which a SERS signal is enhanced is related to the properties of photoreduced silver nanostructures. Electric poling on LiNbO3 produces alternating spontaneous polarizations throughout the entire substrate thickness to form periodically poled lithium niobate (PPLN). The photoreduced silver nanostructures deposited using the PPLN template have the dependence on light wavelength [7], light polarization [7], light intensity [8], solution concentration [8], illumination time [9], and impurity level induced by photorefractive doping [10]. The first five factors affect the size and the distribution of the silver nanoparticles, and the last method can use below-bandgap illumination to produce photoelectrons. Chemical patterning on LiNbO3 by local proton exchange (PE) forms the periodically proton-exchanged (PPE) template and produces the electrostatic field near the substrate surface [11,12]. By comparing plasmonic properties of the silver nanostructures produced using PPLN and PPE templates, it can be seen that the silver nano-pattern arrays produced using PPE templates have stronger SERS signals than those using PPLN templates [13].
Various methods [7–10] have been proposed to increase the density of silver nanostructures on the PPLN template but rarely on the PPE template. The method of tuning the electrostatic field intensity and distribution has not been proposed on any ferroelectric template. In this work, we propose a novel electrostatic-field-tunable ferroelectric template by varying the proton concentration distribution using thermal annealing. In order to obtain the most suitable electrostatic field distribution for growing silver nanostructures, the proton concentration distribution is fine-tuned by varying the annealing time with the annealing temperature set near the PE temperature (240°C). It is different from the conventional PE annealing temperature at 300-400°C for waveguide production. The variations of surface and electrostatic properties of the PPE LiNbO3 with the annealing time are investigated by optical microscope (OM), atomic force microscope (AFM), and electrostatic force microscope (EFM). The morphology of deposited silver nanostructures is surveyed by AFM and scanning electron microscope (SEM). The dependence of spontaneous polarization distribution and electrostatic field distribution on the annealing time is simulated using the finite element method and is used to explain the experimental results. The dependence of Raman signal intensity on the annealing time is also measured and discussed.
2. Fabrication and measurement
We use the z-cut, optical grade, 1mm-thick LiNbO3 substrates to fabricate the PPE SERS substrate. The fabrication process is described as follows. First, the substrate is cleaned using ultrasonic baths in acetone and methanol. Subsequently, a 100nm-thick Cr film is deposited on the substrate surface by RF magnetron sputtering. After photolithography and wet etching, the formed Cr film used as the PE mask has an opening width of 5.6μm and a period of 10μm. The substrates are then immersed in benzoic acid at 200°C for 24 hrs. After the substrate clean and the removal of patterned Cr films, the substrates are annealed at 240°C for 1 hr, 4 hrs, 9 hrs, and 16 hrs. These annealing times are chosen because the diffusion distance is linearly proportional to the square root of the diffusion time [17]. The PPE LiNbO3 substrates are carefully cleaned and then the 0.01 M AgNO3 solution is dropped on the substrate surface. The UVC light with the wavelength centered at 254nm illuminates on the sample at a distance of 2 cm. Because the photon energy (~4.89 eV) is larger than the LiNbO3 bandgap (~3.9 eV), the electron-hole pairs are excited in LiNbO3 and separated by the electrostatic field produced by periodically modulated spontaneous polarization. The photoelectrons migrating to the substrate surface reduce the silver ions in the solution to form the silver nanostructures. After illumination for 30mins, the samples are immersed in de-ionized water and then blown dry by the N2 gas. The PPE LiNbO3 and the photoreduced silver nanostructure are characterized using an optical microscope (Carl Zeiss AxioLab), an atomic force microscope (JPK NanoWizard III), an electrostatic force microscope (JPK NanoWizard III), and a scanning electron microscope (Hitachi S-4700). The AFM images are measured in AC mode at the scan rate of 0.6 Hz under dry conditions. The highly doped silicon probe tip with a spring constant of 42 N/m and a resonant frequency of 330 kHz (Nanosensors) is used to dissipate static charge. In the transmission spectrum measurement, the light from a white light source is focused on the sample by a 10 × objective lens. The transmission light is collected by another 10 × objective lens and subsequently coupled to the 100 µm-core-diameter optical fiber connected to the spectrometer (Ocean Optics, HR2000). For the SERS measurement, the 10−5 M R6G dye solution is dropped on the sample surface for 2 hrs and the samples are then immersed in de-ionized water to remove excess dye molecules. The samples are excited by the CW semiconductor laser with the wavelength 532nm using a 100 × objective (NA = 0.9). The SERS spectra are measured by Ramboss 500i Micro Raman Measurement system with a spectral resolution of 0.9 cm−1.
3. Result and discussion
Figure 1 shows the surface images of the PPE LiNbO3 substrates for different annealing time obtained using OM and AFM. During the annealing process, the protons not only diffuse along the depth direction but also move along the lateral direction. The proton diffusion causes the reduction of the proton concentration and the expansion of the distribution range. Proton incorporation in LiNbO3 causes an extraordinary index increase, enhanced optical absorption, and volume swelling. In the OM image, the bright lines correspond to the PE regions. When the annealing time increases, the line width initially widens and then has no significant change for ta = 9 hrs and 16 hrs. The AFM tomography images show that the periodic swelling height gradually decreases with the annealing time. From the AFM line profiles, the swelling height is measured to be 10 nm, 7.2 nm, 3nm for ta = 0 hr, 1 hr, 4 hrs. When the annealing time is greater than 9hrs, the periodicity of the surface height becomes insignificant.
Fig. 1 OM images (first row), AFM tomography images (second row) and line profiles (third row), and EFM images (fourth row) of the PPE LiNbO3 substrates with the annealing time of (a)0 hr; (b)1 hr; (c)4 hrs; (d)9 hrs; (e)16 hrs.
The surface electrostatic properties of the PPE LiNbO3 substrates are measured using EFM, as shown in the fourth row of Fig. 1. The EFM phase images are measured in the lift mode using a two-step process. The first step records the topography of the measured region in the AFM tapping mode. In the second step, the probe tip is lifted at a constant height above the surface according to the previous topography measurement. The phase shift between the signal of probe actuation and the laser signal reflected from cantilever is measured across the 20 μm × 20 μm region. The phase shift is proportional to the electrostatic force gradient and can be used as an indicator of the electrostatic field intensity. The EFM images are measured using a tip voltage of 3V and a lift height of 50 nm. The 50 nm lift height provides a sufficient buffer distance to avoid sudden probe impact on the substrate surface. It is found that the range and the value of the electrostatic force gradient decreases with the annealing time. Spontaneous polarization of LiNbO3 is caused by the displacement of Li and Nb ions from the plane of the three oxygen ions. When protons are exchanged with Li ions in LiNbO3, the spontaneous polarization intensity is reduced. When the annealing time increases, the proton diffusion causes a reduction of the proton concentration and thus a reduction of the spontaneous polarization. The contrast reduction of the spontaneous polarization results in the decrease of the electrostatic force gradient on the sample surface.
Figure 2 shows the SEM and AFM images of the silver nanoparticles on the PPE LiNbO3 substrates treated with different annealing time. For the un-annealed sample, silver nanoparticles are very small and their shapes are close to being spherical. When the annealing time increases from 1 hr to 9 hrs, not only the particle size becomes larger but also the nanoparticles' shape becomes irregular. Besides, the silver nanoparticle substantially increases in number. The high density nanoparticles distribution makes the calculation of particle count and particle diameter difficult. For the sample annealed for 16 hrs, the nanoparticle size slightly shrinks relative to that annealed for 9 hrs. It is noted from the SEM images in the second row of Fig. 2 that there exists a small amount of silver nanoparticles at the boundaries of PE regions or at the central part of non-PE regions, depending on the annealing time. For the samples treated with the annealing time 0-4 hrs, the regions with a small amount of silver nanoparticles appear at the boundaries of PE regions and become wider with the increase of the annealing time. This phenomenon is attributed to the proton lateral diffusion during the annealing process. When the annealing time increases beyond 9 hrs, the silver nanoparticles in the central part of the non-PE regions begin to appear. The corresponding nanoparticle density also increases with the annealing time. This phenomenon also can be observed in the AFM images in the third row in Fig. 2. It is inferred due to the overlapping of the proton lateral diffusion of PPE regions. The average height of the silver nanoparticles are 9.0 nm, 19.3 nm, 22.1 nm, 37.2 nm, and 32.5 nm for the samples with the annealing time ta = 0 hr, 1 hrs, 4 hrs, 9 hrs, and 16 hrs. The maximal nanoparticle height has a 4.13 times increase than that of the un-annealed sample. The ascending order of nanoparticle density is observed as: (ta = 0 hr) < (ta = 1 hr) < (ta = 4 hr) < (ta = 16 hr) < (ta = 9 hr). Because the lift height of the EFM (50 nm) is larger than the average heights of the nanoparticles, the real electrostatic field gradient which affects the formation of the silver nanoparticles is larger than the measured values. It is inferred from the EFM image of the PPE regions and the SEM and AFM images of the silver nanostructures that the suitable electrostatic field intensity and distribution near the substrate surface facilitates the formation of larger and denser silver nanostructures.
Fig. 2 SEM images (first and second row), AFM tomography images (third row) and line profiles (fourth row) of photoreduced silver nanoparticles on the PPE LiNbO3 substrates treated with the annealing time of (a)0 hr; (b)1 hr; (c)4 hrs; (d)9 hrs; (e)16 hrs.
The proton concentration distribution in the PPE LiNbO3 substrate after annealing is helpful for understanding the dependence of the nanoparticles density on the electrostatic field distribution. However, the proton concentration distribution is difficult to be measured using material analytical instruments. X-ray photoelectron spectroscopy (XPS) cannot detect protons owing to its absence of core electron and very small photoionization cross section [18]. Since the width and the period of the PPE regions are only 5.6 μm and 10 μm, secondary ion mass spectroscopy (SIMS) is unable to provide sufficient resolution due to its large primary beam diameter. In order to understand the variation of the proton concentration distribution and thus the spontaneous polarization distribution with the annealing time, we use the finite element method to model the annealing process and calculate the corresponding electrostatic field distribution induced by the degraded spontaneous polarization. The main effect of the PE is assumed to erase its ferroelectric property (i.e. Ps = 0.78 C/m2→0) [19]. Besides, the reduced amount of spontaneous polarization is proportional to the normalized proton concentration. Figure 3(a)-3(e) shows the spontaneous polarization distribution in the range of 0~0.8 C/m2 and the vectorial electrostatic field distribution of the PPE LiNbO3 substrates treated with different annealing time. The approximate semicircular zone is the PE region in LiNbO3. For ta = 0 hr, the PE regions have the maximal proton concentration and thus have zero spontaneous polarization. When the annealing time increases, the proton diffusion causes the decrease of proton concentration and thus recovery of spontaneous polarization, such that the color of PE regions is changed from green to bright red. The induced electrostatic fields inside and outside the LiNbO3 substrate have different directions and a separate effect during the photoreduction process. Because electrons move in the reverse direction of the electric field, the electrostatic field inside the PE region can separate the photo-generated electron-hole pairs and drive the photoelectrons toward the substrate surface. The electrostatic field above the substrate surface is directed toward the PE region. It can bring the Ag+ ions in the solution to the surface of the PE region to react with photoelectrons for silver photoreduction. This can explain why the photoreduced Ag nanostructures mainly appear on the surface of the PE region. When the annealing time increases, the electrostatic field strength above the surface is obviously weakened. This simulation result is consistent with the EFM measurement results shown in the fourth row of Fig. 1.
Fig. 3 (a-e) Simulated spontaneous polarization distribution and vectorial electrostatic filed distribution of the PPE LiNbO3 substrates treated with the annealing time of 0 hr, 1 hr, 4 hrs, 9 hrs, and 16 hrs. The normalized distribution of (f) the in-plane component (Ex), (g) the out-of-plane component (Ez), and (h) the field value of the electrostatic field at the depth 10nm below the + z surface.
The electrostatic field distribution inside the substrate is shown in Fig. 3(f)-3(h). The Ex component is directed toward the PE region and has a maximal value at the boundary of the PE region. Its maximal value decreases with the increase of the annealing time. The peak electrostatic field at the two-side boundary of the PE regions can explain the high-density silver nanoparticles shown as the white lines in the AFM tomography images of Fig. 2(a). Since the PE causes the reduction of spontaneous polarization, the maximal Ez component appears outside the PE region. With the increase of the annealing time, the recovery of spontaneous polarization enhances the Ez component inside the PE region. The electrostatic field distribution, shown in Fig. 3(h), presents a noticeable phenomenon related to the formation of silver nanostructures. For the sample with ta = 0 hr, although there exists a maximal field value at the boundary of PE regions, the electrostatic field strength in the PE region is weak such that the number of photoelectrons moving to the surface is small. With the increase of the annealing time, although the maximal field value is reduced, the field strength in the PE region is significantly enhanced. Moreover, the regions with sufficient electrostatic field strengths become larger. Both factors facilitate the increase of the number of photoelectrons arriving at the substrate surface.
Transmission spectra of the SERS substrate treated with different annealing time are shown in Fig. 4(a). The transmittance T of the SERS substrate is expressed as optical density, which is defined as log10(1/T). The increase of optical density corresponds to the larger optical absorption induced by LSPR in the silver nanoparticles. For the non-annealed SERS substrate, no wavelength-dependent optical absorption is observed since only a very small amount of nanoparticles are formed. For the annealed SERS substrates, the obvious optical absorption in a certain wavelength range appears due to the occurrence of LSPR phenomenon. The LSPR wavelength is the wavelength corresponding to the peak optical density in the transmission spectrum. The dependence of the peak optical density and the LSPR wavelength on the square root of annealing time is shown in Fig. 4(b). The peak optical density initially increases with the annealing time but decreases when the sample is treated at ta = 16hr. There exists a maximal peak optical density for the sample annealed at ta = 9hr. It is noted that the trend of the peak optical density varied with the annealing time is the same as the nanoparticle density described in the above. This result shows that the nanoparticle density and distribution determine the LSPR excitation and thus the induced optical absorption. Besides, the LSPR wavelength also varies with the annealing time. The maximal LSPR wavelength appears at 526nm, which is close to the excitation wavelength in the SERS measurement. It is well known that the LSPR wavelength of the metallic nanostructure close to the excitation wavelength used for SERS is helpful to enhance the strength of the Raman signal.
Fig. 4 (a) Transmission spectra of the SERS substrates treated with the annealing time of 0 hr, 1 hr, 4 hrs, 9 hrs, and 16 hrs; (b) dependence of LSPR wavelength on the square root of the annealing time.
Figure 5(a) shows the SERS spectra of the R6G dye on the plain LiNbO3 substrate and the SERS substrates treated with different annealing time. The intensity peaks of the Raman signals of the R6G dye appears at the wave numbers 1184 cm−1, 1311 cm−1, 1360 cm−1, 1508 cm−1, and 1647 cm−1, which do not exist in the SERS spectrum of LiNbO3. When the annealing time increases, the Raman signal intensity becomes larger and the maximal intensity appears on the sample treated with the annealing time of 9 hrs, which has the maximal nanoparticle density among these samples as shown in Fig. 2(d). The dependence of the Raman signal intensity on the square root of the annealing time is shown in Fig. 5(b). Because annealing induces proton diffusion and diffusion distance is linearly proportional to square root of diffusion time, the Raman intensity is plotted as a function of the square root of annealing time to reveal the influence of proton diffusion. For the Raman peaks at the wave numbers 1360 cm−1, 1508 cm−1, and 1647 cm−1, the maximal intensity simultaneously appears at the same sample treated with ta = 9hrs. In comparison with the un-annealed sample, the Raman signal intensities for these wave numbers are magnified 2.32, 2.51, and 2.19 times. These Raman peaks correspond to specified vibrational modes of R6G dye, such as the motion of the xanthene ring, the motion of the two NHC2H5 groups, the motion of the two methyl groups, and so on [20]. The degree of Raman enhancement is related to the strength of the interaction between the vibration modes and the localized surface plasmon resonance. In general, the Raman signal enhancements can be attributed to the most favorable photoreduced silver nanostructure produced by using the electrostatic-field-tunable ferroelectric template.
Fig. 5 (a) R6G dye SERS spectra on the plain LiNbO3 substrates and the SERS substrates treated with the annealing time of 0 hr, 1 hr, 4 hrs, 9 hrs, and 16 hrs; (b) dependence of Raman intensity on the square root of the annealing time.
In the formation process of silver nanostructure, the quantities of photoelectrons and silver ions arriving at the surface of the PE region determine the density and the size of the silver nanostructures. When the photoelectrons are excited inside LiNbO3 by above-bandgap illumination, they migrate toward the surface of the PE region by the electrostatic field and the thermal diffusion. In order to successfully arrive at the substrate surface, they need to avoid recombination with holes and capture by defects. According to the simulation results, when the annealing time increases, the electrostatic field inside the PE region is significantly increased. Moreover, the electrostatic field above the substrate surface is dramatically reduced. The former can increase the number of photoelectrons arriving at the substrate surface but the latter reduces the number of silver ions arriving at the surface of the PE region. In addition to tuning the electrostatic field distribution, the annealing process also has the effect of repairing the lattice disorder caused by the PE and reducing the defect concentration. The stronger electrostatic field intensity in the PE regions can provide the photoelectrons with sufficient kinetic energy to reduce the possibility of being captured by the defects. Hence, increasing the annealing time has the effects of enhancing the electrostatic field inside the PE region, reducing the electrostatic field above the substrate, and reducing the defect concentration. The first and third items have a positive effect in growing silver nanoparticles, but the second item has an adverse influence. The densest silver nanostructure prepared using ta = 9 hrs is the result of a trade-off between these three factors. Hence, the electrostatic-field distribution most suitable for the formation of silver nanostructure can be produced by using the proposed electrostatic-field-tunable ferroelectric template.
4. Conclusion
We propose a novel electrostatic-field-tunable ferroelectric template to produce photoreduced silver nanostructures on LiNbO3 for Raman signal enhancement. The distribution of surface electrostatic field produced by periodic proton exchange on LiNbO3 is tuned by varying the proton concentration distribution using thermal annealing. The tuned electrostatic field distribution can effectively drive the photoelectrons excited by above-bandgap illumination on LiNbO3 to produce the most favorable silver nanostructures for Raman signal enhancement. The variation of electrostatic field distribution with the annealing time is characterized by numerical simulation and experimental measurement. With the annealing proceeding, the proton diffusion causes the contrast reduction of spontaneous polarization and the variation of the electrostatic field distribution. Its effects include the increase of the electrostatic field inside the PE regions, the decrease of the electrostatic field above the substrate surface, and the reduction of the defect concentration. These three factors affect the number of silver ions and photoelectrons arriving at the substrate surface for the formation of the silver nanostructure. Under the action of these three factors, the silver nanostructure most suitable for Raman signal enhancement occurs on the sample with the annealing time of 9 hrs. The produced PPE SERS substrate owns a 4.13 times higher silver nanostructure and an increase factor of 2.51 at the 1360 cm−1 Raman peak in comparison with the conventional PPE sample. The proposed electrostatic-field tunable ferroelectric template is proved to be an effective method to enhance the Raman signal intensity and has the features of low cost, simple fabrication, and high flexibility.
Funding
Ministry of Science and Technology of Taiwan (MOST 103-2112-M-019-003-MY3).
Acknowledgments
The authors acknowledge the EFM (JPK Instruments AG, Germany) usage from Research Center for Applied Sciences, Academia Sinica, Taipei, Taiwan, and the manufacturing assistance from the Nano-Electro-Mechanical-Systems (NEMS) Research Center, National Taiwan University, Taipei, Taiwan.
References and links
1. S. Schlücker, “Surface-enhanced Raman spectroscopy: concepts and chemical applications,” Angew. Chem. Int. Ed. Engl. 53(19), 4756–4795 (2014). [CrossRef] [PubMed]
2. H. T. Beier, C. B. Cowan, I.-H. Chou, J. Pallikal, J. E. Henry, M. E. Benford, J. B. Jackson, T. A. Good, and G. L. Cote, “Application of Surface-enhanced Raman spectroscopy for detection of beta amyloid using nanoshells,” Plasmonics 2(2), 55–64 (2007). [CrossRef]
3. S. Mahajan, J. Richardson, T. Brown, and P. N. Bartlett, “SERS-melting: a new method for discriminating mutations in DNA sequences,” J. Am. Chem. Soc. 130(46), 15589–15601 (2008). [CrossRef] [PubMed]
4. K. Gracie, E. Correa, S. Mabbott, J. A. Dougan, D. Graham, R. Goodacre, and K. Faulds, “Simultaneous detection and quantification of three bacterial meningitis pathogens by SERS,” Chem. Sci. (Camb.) 5(3), 1030–1040 (2014). [CrossRef]
5. A. Chou, E. Jaatinen, R. Buividas, G. Seniutinas, S. Juodkazis, E. L. Izake, and P. M. Fredericks, “SERS substrate for detection of explosives,” Nanoscale 4(23), 7419–7424 (2012). [CrossRef] [PubMed]
6. S. Nie and S. R. Emory, “Probing single molecules and single nanoparticles by surface-enhanced Raman scattering,” Science 275(5303), 1102–1106 (1997). [CrossRef] [PubMed]
7. 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]
8. Y. Sun, B. Eller, and R. J. Nemanich, “Photo-induced Ag deposition on periodically poled lithium niobate: concentration and intensity dependence,” J. Appl. Phys. 110(8), 084303 (2011). [CrossRef]
9. X. Liu, K. Kitamura, Q. Yu, J. Xu, M. Osada, N. Takahiro, J. Li, and G. Cao, “Tunable and highly reproducible surface-enhanced Raman scattering substrates made from large-scale nanoparticle arrays based on periodically poled LiNbO3 templates,” Sci. Technol. Adv. Mater. 14(5), 055011 (2013). [CrossRef] [PubMed]
10. F. Jia, W. Yan, D. Wang, L. Zhang, L. Shi, A. Lin, G. Liang, M. Li, Y. Zhang, J. Zhang, H. Dong, G. Chen, and H. Chen, “Photoinduced Ag-nanoparticle deposition on Fe-doped lithium niobate crystals,” Opt. Mater. Express 4(2), 359–365 (2014). [CrossRef]
11. 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]
12. N. C. Carville, S. M. Neumayer, M. Manzo, K. Gallo, and B. J. Rodriguez, “Biocompatible gold nanoparticle arrays photodeposited on periodically proton exchanged lithium niobate,” ACS Biomater.-.Sci. Eng. 2(8), 1351–1356 (2016).
13. S. Damm, N. C. Carville, B. J. Rodriguez, M. Manzo, K. Gallo, and J. H. Rice, “Plasmon enhanced Raman from Ag nanopatterns made using periodically poled lithium niobate and periodically proton exchanged template methods,” J. Phys. Chem. C 116(50), 26543–26550 (2012). [CrossRef]
14. J. L. Giocondi and G. S. Rohrer, “Spatially selective photochemical reduction of silver on the surface of ferroelectric barium titanate,” Chem. Mater. 13(2), 241–242 (2001). [CrossRef]
15. S. Dunn, P. M. Jones, and D. E. Gallardo, “Photochemical growth of silver nanoparticles on c- and c+ domains on lead zirconate titanate thin films,” J. Am. Chem. Soc. 129(28), 8724–8728 (2007). [CrossRef] [PubMed]
16. X. Y. Liu, K. Kitamura, K. Terabe, H. Hatano, and N. Ohashi, “Photocatalytic nanoparticle deposition on LiNbO3 nano-domain patterns via photovoltaic effect,” Appl. Phys. Lett. 91(4), 044101 (2007). [CrossRef]
17. E. Y. B. Pun, K. K. Loi, and P. S. Chung, “Proton-exchanged optical waveguides in z-cut LiNbO3 using phosphoric acid,” IEEE Trans. Lightw. Technol. 11(2), 277–284 (1993). [CrossRef]
18. N. Stojilovic, “Why can’t we see hydrogen in X-ray photoelectron spectroscopy?” J. Chem. Educ. 89(10), 1331–1332 (2012). [CrossRef]
19. M. Manzo, F. Laurell, V. Pasiskevicius, and K. Gallo, “Electrostatic control of the domain switching dynamics in congruent LiNbO3 via periodic proton-exchange,” Appl. Phys. Lett. 98(12), 122910 (2011). [CrossRef]
20. H. Watanabe, N. Hayazawa, Y. Inouye, and S. Kawata, “DFT vibrational calculations of rhodamine 6G adsorbed on silver: analysis of tip-enhanced Raman spectroscopy,” J. Phys. Chem. B 109(11), 5012–5020 (2005). [CrossRef] [PubMed]
