SERS Activities of Green Synthesized Silver ...

12/22/2014

e.Proofing

SERS Activities of Green Synthesized Silver Nanoparticles M. R. Bindhu 1 V. G. Sathe 2 M. Umadevi 1,* Phone 04542241685 Email [email protected] 1 Department of Physics, Mother Teresa Women’s University, Kodaikanal, 624101 Tamil Nadu, India 2 UGCDAE Consortium for Scientific Research, University Campus, Khandwa Road, Indore, 452 017 India

Abstract Spherical silver nanoparticles with average particle size of 11 nm having surface plasmon resonance peak at 440 nm are synthesized using fruit extract of Ananas comosus as reducing agent. The bright circular spots in the selected area electron diffraction pattern and the peaks corresponding to (111), (200), (220) and (311) planes in the Xray diffraction pattern were evident for the crystallinity of face centered cubic structured nanoparticles. The surfaceenhanced Raman scattering (SERS) activities of prepared silver nanoparticles were found to be size dependent, the smaller nanoparticles showing higher SERS enhancement. The orientation of the pyridine molecule on the silver surface can be deduced from ring stretching vibrations, the ring breathing mode, inplane and outofplane vibrations and the SERS surface selection rule. The SERS spectrum indicates that the pyridine adsorbed on the silver surface in a standon orientation via its nitrogen lone pair electrons. It is used to indicate the advantage of this green method of preparing silver based SERS colloids.

Keywords http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

1/15

12/22/2014

e.Proofing

Ananas comosus Surface plasmon resonance Silver nanoparticles Surfaceenhanced Raman scattering Pyridine

Introduction Metal nanoparticles have been studied extensively because of their promising applications in areas such as optics, optoelectronics, catalysis, photography, nanostructure fabrication, photonics and surface enhanced Raman scattering. Among the known nanoparticles, silver has been widely studied because of its characteristic optical, spectroscopic, and catalytic and SERS properties. Silver nanoparticles can be synthesized using chemical and physical methods that involve the use of hazardous chemicals which may pose environmental risks. But green synthesis of nanoparticles using plants has received much interest to chemical and physical methods. Recently, green synthesis of silver nanoparticles using D. carrot [ 1 ], Moringa oleifera [ 2 ], S. lycopersicums [ 3 , 4 ], Hibiscus cannabinus [ 5 , 6 ] and Ananas comosus [ 7 ] have been reported. Bhosale et al. reported the synthesis of nanoparticles using A. comosus extract as reducing agent with kanamycin A and neomycin as stabilizing agents [ 8 ]. In this work, silver nanoparticles were synthesized using A. comosus fruit extract as reducing agent. A. comosus is a readily available fruit and it is a good source of citric acid, malic acid and ascorbic acid [ 9 ]. Surfaceenhanced Raman scattering (SERS) has been known and strongly studied in the past decades. SERS was of great interest because it constitutes an important tool for the highsensitivity detection of a broad kind of compounds. SERS activity of noble metal surfaces was a powerful method to understand the adsorption behaviour of molecules on metal surfaces. It reveals the orientation of molecules and the mechanism of interaction of the molecules with the surface. SERS was extensively used in biomolecules and in biomedical applications. The SERS activity of silver nanoparticles prepared by green method has been reported [ 10 ]. In the present study, the synthesis and characterization of silver nanoparticles using fruit extract of A. comosus as reducing agent is described. The SERS activity of the prepared silver nanoparticles is also http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

2/15

12/22/2014

e.Proofing

described.

Experimental Methods Ananas comosus fruit was collected from the local supermarket in Kodaikanal, Tamilnadu, India. Silver nitrate and pyridine were obtained from Sigma Aldrich Chemicals. All glasswares were properly washed with distilled water and dried in hot air oven before use. 10 ml of A. comosus extract was added to aqueous solution of AgNO3 (5 mM) and stirred for 5 min. The reaction was completed in 1 h and shows stable reddish brown colour of the silver colloid (P1). Similarly by adding 20 ml of fruit extract, silver colloid (P2) was prepared. UV–visible spectra of these solutions were recorded. Then the solutions were dried. The dried powders were taken for Xray diffraction (XRD), transmission electron microscope (TEM) and Energy Dispersive Xray Spectroscopy (EDX) measurements. The absorption spectra of the prepared nanoparticles were measured using a Shimadzu spectrophotometer (UV 1700) in 300–800 nm range. XRay Diffraction analysis of the prepared nanoparticles was done using PANalytical X’pert—PRO diffractometer with Cu Kα radiation operated at 40 kV/30 mA. transmission electron microscopic (TEM) analysis was done using a JEOL JEM 2100 High Resolution Transmission Electron Microscope, operating at 200 kV. The Raman spectral measurements were made using LABRAM HR800 spectrometer with 488 nm as excitation wavelength.

Results and Discussion UV–Vis spectroscopy was used to characterize the optical properties of prepared silver nanoparticles. Surface plasmon resonance (SPR) absorption band was observed due to the combined oscillation of free conduction electrons of metal nanoparticles in resonance with light wave. The optical absorption spectra of silver colloids P1 and P2 were shown in Fig. 1 . In the present case, a steady increase in absorbance as increasing fruit extract concentration with a blue shift (445–440 nm). As the concentration of the fruit extract increases more number of biomolecules are available to reduce silver ion and forms large number of very small nanoparticles gives rise to sharp and intense SPR [ 11 , 12 ]. As the particles decrease in size, the absorption peak usually shifts toward the blue wavelengths, higher frequency and energies [ 13 ]. Generally the full width half maximum (fwhm) value supports the particle size and their distribution within the http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

3/15

12/22/2014

e.Proofing

medium. The plasmon peak and fwhm depends on the extent of colloid aggregation [ 14 ]. The fwhm value in observed SPR band was decreased from 237 to 222 nm with increase in the concentration of fruit extract. Here, the particle size of the prepared silver nanoparticles decreases with decreasing fwhm value [ 15 ]. The dispersibility of the prepared silver nanoparticles was found out by the SPR band. Absorption spectra of P1 showed the broadening of SPR peak at 445 nm indicating the formation of polydispersed nanoparticles with different size and shape. The absorption spectra of P2 showed the sharp SPR band at 440 nm confirming the formation of spherical nanoparticles. This result was confirmed by TEM image of Fig. 2 . Silver nanoparticles with sharp peaks are attractive for SERS applications. Fig. 1 UVvis spectra of silver nanoparticles a P1 and b P2

Fig. 2 TEM images at different magnifications (a–c) and SAED pattern d of P1 (i) and P2 (ii)

http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

4/15

12/22/2014

e.Proofing

The TEM images of P1 and P2 were shown in Fig. 2 i, ii. The TEM image of P1 revealed the formation of anisotropic nanoparticles with average size of 35 nm in ranging from 3 to 61 nm. As the concentration of fruit extract decreases lower quantities of reducing biomolecules were available to reduce silver nitrate ions and failed to protect the prepared nanoparticles from aggregation. The TEM image of P2 showed the formation of spherical nanoparticles in the range from 8 to 33 nm with an average of 11 nm. Here large number of biomolecules in the extract was available for reduction and formation of small nanoparticles. Strong interaction between biomolecules and surface of nanoparticles was sufficient to the formation of spherical nanoparticles preventing them from sintering. These results reveal that the concentration of fruit extract plays an important role in the formation of silver nanoparticles in varied sizes and shapes. Based on TEM images, the number bigger size particles are found less in number. This may possibly be due to the control in size of the silver nanoparticles as they are protected by the biomolecules, which provides steric hindrances between neighboring nanoparticles preventing aggregation by overcoming the Vander Waals of force of attraction between them. The spacing (Fig. 2 iic) between lattice planes reveals that the growth of silver nanoparticles occur on the (111) plane, which is in agreement with the XRD result. The observed selectedarea electron diffraction (SAED) pattern (Fig. 2 id, iid) corresponds to crystalline (FCC) nature of the nanoparticles. Xray diffraction (XRD) analysis was performed to find out the nature of http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

5/15

12/22/2014

e.Proofing

the obtained silver nanoparticles. The XRD pattern for the dried powder of A. comosus was shown in Fig. 3 a. It indicates the amorphous nature of the A. comosus extract. Figure 3 b, c show that the XRD patterns of the prepared silver nanoparticles. The diffraction peaks were observed at 38.2°, 44.3°. 64.5° and 77.4° in the 2θ range 20–80° can be indexed to the (111), (200), (220) and (311) reflection planes of face centered cubic (FCC) structure of metallic silver with space group of Fm3 m (JCPDS 04 0783). No peaks of crystallographic impurities in the sample have been found. From the XRD pattern, the obtained broader diffraction peaks in P2 indicate smaller crystallite size, which reflect the effects of the experimental conditions on the nucleation and growth of the crystal nuclei [ 16 ]. This was also confirmed by the estimation of fwhm values by using the width of the (111) Bragg’s reflection and tabulated in Table 1 . It was found to be increasing fwhm value with decreasing particle size [ 17 ]. The average size of silver nanoparticles was determined from the Debye– Scherrer Eq. by using the width of the (111) Bragg reflection and tabulated in Table 1 . The ratio between the intensity of the (200) and (111) diffraction peaks was 0.31 for P1 and 0.29 for P2, which was lower than the conventional bulk intensity ratio 0.52, suggesting that the (111) plane was the predominant orientation as confirmed by highresolution TEM measurements. The calculated values of lattice constant and cell volume were in very good agreement with the standard value (JCPDS 040783). Fig. 3 Xray diffraction pattern of a A. comosus fruit extract b P1 and c P2

Table 1


6/15

12/22/2014

e.Proofing

Table 1 The fwhm, average particle size, lattice constant, cell volume and SSA of prepared silver nanoparticles Prepared AgNPs

fwhm β (2θ)

Particle size D (nm)

Lattice constant a (Å)

Cell volume V (Å3)

Specific surface area (SSA) (m2/g)

P1

0.7

11

4.0804

67.94

51.94

P2

1.1

8

4.0673

67.28

71.42

The surface area to volume ratio (SA:V) and specific surface area (SSA) of the prepared silver nanoparticles was calculated to determine the type and property of a material. The SSA is of particular importance in reactivity. It gives the rate at which the reaction will proceed. The values of SA:V and SSA of P1 and P2 were evaluated and tabulated in Table 1 . The particle size decreased and specific surface area (SSA) increased with the silver nanoparticles synthesized using higher concentration of fruit extract (P2). Crystallinity index of the prepared silver nanoparticles was evaluated by comparing the crystalline size obtained by XRD to TEM particle size determination. If Icry is close to 1, it is monocrystalline whereas if it is greater than 1, it is a polycrystalline in nature. The calculated values of crystallinity index of P1 and P2 were ~1 and ~3 respectively. This indicates that P2 shows monocrystalline whereas P1 shows polycrystalline nature (Table 2 ). Table 2 Vibrational assignments of pyridine and pyridine adsorbed on silver nanoparticles Pyridine (nRs)/cm−1

Py/A1 (SERS)/cm−1

Py/A3 (SERS)/cm−1

Vibrational assignments

1,688

1,616

1,620

C–C ring stretching

1,581

1,582

1,586

C=N and ring stretching

–

1,521

1,535

C=N and ring stretching

1,488

1,482

1,472

C–H inplane bending

1,399

1,411

1,413

C–N and ring stretching


7/15

12/22/2014

e.Proofing

–

1,300

1,300

C–N and ring stretching

1,210

1,193

1,191

C–N ring stretching

1,138

1,153

1,152

Inplane C–H def.

1,023

1,029

1,029

Trigonal ring breathing

994

1,005

1,005

Totally sym. ring breathing

875

880

881

Inplane C–H def.

–

–

718

Inplane ring def.

692

677

674

Inplane ring def.

597

602

576

Inplane ring distortion

482

496

506

Outofplane C–N def.

424

413

413


328

324

334


–

229

221

Ag–N stretching

–

130

127

Ag–N stretching

AQ1

SERS is an analytical technique with a high sensitivity applied to biomedically significant molecules to study structural functional properties. It was used to elucidate the information about the identification and orientation of adsorbed biomolecules on the metal surfaces. Figure 4 a shows a typical normal Raman spectrum (nRs) of pyridine (Py). Figure 4 b, c depicts the SERS spectrum of pyridine absorbed on the silver colloids Py/P1 and Py/P2 respectively. Based on the observed SERS spectrum, the orientation of an adsorbate can be inferred. Fig. 4 a Normal Raman spectrum of Pyridine (inset The molecular structure of pyridine), b SERS spectrum of Py/P1 and c SERS spectrum of Py/P2 http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

8/15

12/22/2014

e.Proofing

Pyridine (azabenzene) is an organic compound of the aromatic heterocyclic series. The molecular formula of pyridine is shown in Fig. 4 a (inset). Pyridine is assumed to be planar and belongs to molecular symmetry C2ν and it has 11 atoms and therefore 27 fundamental modes of vibrations. These vibrational modes have the following distribution: 10a1 + 9b2 + 3a2 + 5b1 (a1 is IR; Raman, polarized, b2 and b1 are IR; http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

9/15

12/22/2014

e.Proofing

Raman, depolarized and a2 is Raman and depolarized) [ 13 ]. Aromatic π electron sextet, electro negative elementnitrogen and unshared pair of electrons of the nitrogen in C–N bond are the binding sites in pyridine molecule. The orientation of the adsorbate deduced from these binding sites through which the interaction takes place. The observed vibrational modes were assigned according to the literature [ 18 – 24 ]. The SERS enhancement in the nanometal surfaces is explained through electromagnetic and chemical enhancement mechanisms. Chemical enhancement arises at charge transfer between a molecule and a metal surface resulting in an increase in the polaraizability of the molecules. When the molecules are chemisorbed, the molecular structure of the adsorbate will be modified by an overlapping of molecular and metal orbital. This can be inferred from the shift in vibrational wavenumber of the modes in SERS with respect to the corresponding modes in nRs. The possible orientation of the pyridine molecule on the silver surface were lying down (flaton) on the silver surface through bonding with the ring system or standing up (endon) with bonding through the lone pair electrons of the pyridine ring nitrogen atom with silver. The orientation of the pyridine molecule on the silver surface can be deduced from ring stretching vibrations, the ring breathing mode, inplane and outofplane vibrations and the SERS surface selection rule. In the present case, two ring breathing modes are observed in both nRs and SERS. The trigonal ring breathing mode occurs at 1,029 cm−1, in both Py/P2 and Py/P1, was also observed in nRs at 1,023 cm−1 as a strong band. Similarly the symmetric ring breathing mode of SERS occurs at 1,005 cm −1 was upshifted by about 11 cm−1 and the bandwidth was decreased, compared to 994 cm−1 for nRs. Here the ring breathing modes involve vibrational motion of the N atom along the Ag–N bond. In the SERS spectra, the metal–molecule interactions decrease the frequency of the ring breathing mode when compared to the spectrum of the “free” molecule in the liquid state, which was the behavior of the flatadsorbed benzene on silver surface [ 25 ]. But in our case, there was no shift to the lesser frequency due to the adsorption and it was clearly suggests that pyridine adsorbed on the silver surface in an endon orientation. Medium intense bands in the general region 1,600–1300 cm−1 were assigned to ring stretching vibration [ 19 ]. The ring stretching vibrations http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

10/15

12/22/2014

e.Proofing

were observed in the spectrum of pyridine. The ring stretching vibrations involve stretching and contraction of all the bonds in the ring and interaction between the stretching modes. Pyridine show an absorption band in the region 1,600–1,500 cm−1 due to the C=N ring stretching vibrations [ 20 ]. The C–N stretching vibration was a possibility to mixed with other bands and assigned in the region 1,400–1,200 cm−1 [ 21 ]. The pyridine molecule shows two C–N and one C=N stretching vibrations. The C=N stretching band at 1,581 cm−1 was upshifted in both Py/P2 and in Py/P1. The C–N stretching bands at 1,399 and 1,210 cm−1 in nRs was upshifted in both Py/P1 and Py/P2. The intensity of all ring stretching vibrations was highly enhanced in Py/P2. For Py/P2, the frequency difference between the nRs and SERS of the ring stretching vibration was not more than 5 cm−1 suggests that pyridine adsorbed via its nitrogen lone pair electrons. This confirms the orientation of pyridine on silver nanoparticles was standon configuration. SERS intensities, appearance of new peaks and broadening of peaks provide important information about the molecular orientation of the adsorbed molecules on the silver surface. In our case, the SERS intensity of Py/P2 was very high when compared to Py/P1. It was found that the small size (11 nm) of silver nanoparticles (P2) has larger surface area (71.42 m2/g) and provide high enhancement. The size of the metal nanoparticle decreased with increasing electromagnetic field. The electromagnetic field depends on the degree of roughness of the surface, and this was independent of particlemetal interaction. The enhancement of the electric field at the interface and the resonance with the surface plasmon absorption band due to the excitation of the conduction electrons localized at the silver surface were responsible for the observed high SERS intensity due to electromagnetic effect. The particle size increased with increasing electromagnetic enhancement, but the larger particles adsorb less light and scatter. When pyridine molecules adsorbed on silver nanoparticles in standon orientation, the polaraizability tensor of C–N bond will be normal to the surface [ 26 ]. This will result in observed high enhancement in Py/P2. In the case of aromatic molecules, the intensity of outofplane vibrational modes increases substantially relative to the inplane vibrational modes when the adsorbate orientation is flaton orientation and viceversa when it is perpendicular to the surface [ 27 ]. The inplane vibrational modes occur −1 http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

11/15

12/22/2014

e.Proofing

at 1,472, 1,152, 881, 674 and 576 cm−1 in SERS spectra. The intensity of these SERS bands was higher than that of the band in Raman spectrum. The C–N outofplane bending modes observed at 506, 413 and 334 cm−1. These bands possess strong polarizability tensors corresponding to outof plane vibrations was parallel to the silver surface. The intensity of these vibrational modes in SERS was decreases with respect to nRs. The intensity of aromatic CH outofplane modes with respect to the inplane bending mode decreases as the molecule is adsorbed standon the silver surface. This also confirms the standon orientation of pyridine adsorbed silver nanoparticles. In the SERS spectrum, new bands were observed at 221 and 127 cm−1 was due to the Ag–N stretching mode [ 24 ]. This indicates that the pyridine molecule was adsorbed on silver surface through the nitrogen atoms in standon orientation. Under the same experimental conditions, the Raman enhancement effect in SERS was experimentally measured by using a enhancement factor (EF) that gives a measure of the enhancement of the Raman signal per molecule adsorbed on the surface of a SERSactive species [ 28 ]. The EF value is calculated as EF = (ISERS /IRS )/(C RS /C SERS )

where ISERS and IRS are the measured SERS and Raman intensities, and CRS and CSERS are the concentrations in reference and enhanced samples [ 29 ]. The EF value was measured for the four most intense vibration modes of C–C stretching (155 for Py/P1 and 781 for Py/P2), C–N ring stretching (112 for Py/P1 and 395 for Py/P2), trigonal ring breathing (21 for Py/P1 and 105 for Py/P2) and the symmetric ring breathing mode (17 for Py/P1 and 62 for Py/P2). The values of EF were found to be increasing as decreasing particle size. To determine experimentally the effect of spherical silver nanoparticles size on SERS intensity for molecules adsorbed onto the surface of colloidal AgNPs, thereby determining an optimal size of AgNPs to maximize the SERS signal. The observed high SERS signal indicates that the prepared silver nanoparticles are good source for biomedical applications as SERS substrate.

Conclusion http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

12/15

12/22/2014

e.Proofing

Silver nanoparticles synthesized using A. Comosus as reducing agent reported in this study is simple, energyefficient and environmentally benign process. The prepared silver nanoparticles have been characterized by UV–Vis and TEM measurements to identify the size, shape of nanoparticles. Crystalline nature of the nanoparticles was evident from bright spots in the SAED pattern and peaks in the XRD pattern. The SERS spectral analysis indicates that the smaller silver nanoparticles prepared at higher fruit extract concentration reveal high SERS activity. The vibration features of C–N and Ag–N stretching modes and inplane CH bending modes suggest that the pyridine molecule adsorbed in a standon orientation on the silver surface.

Acknowledgments The authors are thankful to DSTCURIE New Delhi, UGCDAECSR Indore for financial assistance.

References 1. M. Umadevi, S. Shalini, and M. R. Bindhu (2012). Adv. Nat. Sci: Nanosci. Nanotechnol. 3, 025008, 1. 2. M. R. Bindhu, V. Sathe, and M. Umadevi (2013). Spectrochim. Acta A 115, 409. 3. M. Umadevi, M. R. Bindhu, and V. Sathe (2013). J. Mater. Sci. Technol. 29, 317. 4. M. Umadevi and M. R. Bindhu (2014). Spectrochim. Acta A 128, 37. 5. M. R. Bindhu and M. Umadevi (2013). Spectrochim Acta A 101, 184. 6. M. R. Bindhu and M. Umadevi (2014). J. Clust. Sci.. doi:10.1007/s1087601306798. 7. M. R. Bindhu and M. Umadevi (2014). Mater. Lett. 120, 184. 8. V. Santosh Nalage, V. Sidhanath Bhosale, and V. Sheshanath http://eproofing.springer.com/journals/printpage.php?token=11Q2Kw_lMvLwTC7LUDM5P6Sxo3zyROxJgJNT6TH297k

13/15

12/22/2014

e.Proofing

Bhosale (2011). ONJ 5, 78. 9. E. K. Nelson (1925). J. Am. Chem. Soc. 47, 1177. 10. W. Li, Y. Guo, and P. Zhang (2010). J. Phys. Chem. C 114, 6413. 11. V. Parashar, R. Parashar, B. Sharma, and A. C. Pandey (2009). Dig. J. Nanomater. Biostruct. 4, 45. 12. M. M. Ganesh Babu and P. Gunasekaran (2009). Colloids Surf. B 74, 191. 13. S. L. Smitha, K. M. Nissamudeen, D. Philip, and K. G. Gopchandran (2008). Spectrochim. Acta A. 71, 186. 14. C. F. Bohren and D. R. Huffman Adsorption and Scattering of Light by Small Particles (Wiley, New York, 1983). 15. K. R. Brown, D. G. Walter, and M. Natan (2000). Chem. Mater. 12, 306. 16. A. Irshad, Wani, A. Ganguly, J. Ahmed, and T. Ahmad (2011). Mat. Lett. 65, 520. 17. A. Becheri, M. Durr, P. L. Nostro, and P. Baglioni (2008). J. Nanopart. Res. 10, 679. 18. R. G. Newkome The Chemistry of Heterocyclic Compounds (Wiley, New York, 1984). 19. R. M. Silverstein and F. X. Webster Spectrometric Identification of Organic Compounds, 6th ed (Wiley, Asia, 2003). 20. G. Socrates Infrared and Raman Characteristics Group Frequencies Tables and Charts, 3rd ed (Wiley, Chichoster, 2001). 21. B. S. Yadav, I. Ali, P. Kumar, and P. Yadav (2007). Indian J Pure & App. Phys. 45, 979.


14/15

12/22/2014

e.Proofing

22. F. P. Urena, M. F. Gomez, J. J. L. Gonzalez, and E. M. Torres (2003). Spectrochim. Acta A 59, 2815. 23. C. Y. Panicker, H. T. Varghese, D. Philip, and I. S. H. Nogueira (2006). Spectrochim. Acta A 64, 744. 24. X. Y. Li, Q. J. Huang, V. I. Petrov, Y. T. Xie, Q. Luo, X. Yu, and Y. J. Yan (2005). J. Raman Spectrosc. 36, 555. 25. M. Moskovits and D. P. D. Lella (1980). J. Chem. Phys. 73, 6068. 26. J. A. Creighton in R. J. H. Clark and R. E. Hester (eds.), Advances in Spectroscopy, Spectroscopy of Surfaces, vol. 16 (Wiley, New York, 1988), p. 37. 27. M. Moskovits (1982). J. Chem. Phys. 77, 4408. 28. V. Tiwari, T. Oleg, G. Darbha, W. Hardy, J. Singh, and P. Ray (2007). Chem. Phys. Lett. 446, 77. 29. J. B. Jackson and N. J. Halas (2004). Proc. Natl. Acad. Sci. U.S.A. 101, 17930.


15/15