A Method for 3D Nano-focusing of Optical Energy and its Application to Surface Enhanced Raman Spectroscopic Study by Kiang Wei Kho (2009) A Method for 3D Nano-focusing of Optical Energy and its Application to Surface Enhanced Raman Spectroscopic Study by Kiang Wei Kho SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY AT NATIONAL UNIVERSITY OF SINGAPORE Supervisors: Prof Frank Watt, National University of Singapore Prof Malini C Olivo, National University of Singapore Prof Shen ZeXiang, Nanyang Technological University of Singapore Table of Contents Acknowledgements Synopsis Acknowledgements Glossary List of Figures List of Tables Introduction Chapter Protein Layers Page 1.1 A Brief Discussion on Protein Layers Adsorbed at Solid-Liquid Interface 1.2 Study of Solid-Supported Protein Layers 1.2.1 Electronic Spectroscopy 1.2.2 Ellipsometry 1.2.3 Atomic Force Microscopy 1.2.4 Computer Simulation 1.2.5 Vibrational Spectroscopy 1.3 Surface-Sensitive Raman Spectrocopy: A New Paradigm for Protein-Layer Studies 1.4 Objective of Current Study 11 Chapter General Theory of Enhanced Raman Scattering and its Experimental Verifications 2.1 General Theory of Raman Scattering 13 2.2 Enhanced Raman Scattering 13 2.2.1 Chemical Raman Enhancement 14 2.2.2 Electromagnetic Raman Enhancement 16 i 2.3 Electromagnetic Raman Enhancement through Nano-focusing of Optical Energy 17 2.3.1 Nano-focusing of Optical Energy by Dielectric Microspheres 17 2.3.2 Nano-focusing of Optical Energy by Noble Metallic Nano-particles 18 2.4 Enhanced Raman Scattering by Dielectric Microspheres 21 2.5 Enhanced Raman Scattering by Noble Metal: Surface Enhanced Raman Scattering (SERS) 2.5.1 21 Evidence of SERS: Excitation- and Raman-wavelength Dependence of SERS from Metallic Nano-particles 2.5.2 Evidence of SERS: Dependence of Enhanced Raman Intensities on the Gap Separation within a Metallic Dimer 2.5.3 34 Evidence of SERS: Polarization Dependence of Enhanced Raman Intensities from Anisotropitc Metallic Substrates 2.6 25 38 Comparison of the Overall Enhancement Attainable by Above two Nano-focusuing Schemes 2.7 43 Conclusion 44 Chapter Limitations in Conventional Surface Enhanced Raman Scattering 3.1 Limited SERS Hot-spot Sizes 46 3.1.1 Distance Dependence of SERS from Si-Au-Shell-Core Structures 50 3.1.1.1 Fabrication of Silica-Shell Au-Core (Si-Au) SERS Nano-particles 51 3.1.1.2 Raman Microscopy System 52 3.1.1.3 TEM Analysis of Si-Au 54 3.1.1.4 SERS Scattering from Si-Au 56 3.1.1.5 Preparation of Silica Film graom on Au-Film Overcoated Nano-spheres (AuFON) array 60 3.1.1.6 Distance Dependence of SERS from Au-Film Overcoated Nano-spheres (AuFON) 61 ii 3.1.2 SERS Substrates for the Study of Protein Layer 63 3.2 Temporal Variation sin SERS signals as a Resulf of Surface Plasmonic Heating 65 3.2.1 Irreversible Signal Loss 65 3.2.1.1 Introduction 65 3.2.1.2 Experimental Procedure 67 3.2.1.3 Temporal Changes in SERS Signals 68 3.2.1.4 Changes in the Orientation or Chemical Structure of the Analyte Molecule 72 3.2.1.5 Desoprtion of Analyte Molecules 75 3.2.1.6 Themal-induced Lateral Diffusion of Au Nano-particles 79 3.2.2 A Stable substrates for Reproducible Results 86 3.3 Effects of SERS-active Metallic Sandwiches on Protein Structure 87 3.3.1 Structural Changes in Cytochrome-c Sandwiched in Between Nano-particles 87 3.3.1.1 Agarose Gel Analysis 88 3.3.1.2 TEM Analysis 89 3.3.1.3 Circular-Dichroism Analysis 90 3.3.1.4 SERRS of Cytochrome-c 92 3.3.2 Structural Study of Protein Adsorbed on SERS substrates 94 3.4 Conclusion 98 Chapter Electromagnetic Theory of 3D Nano-focusing of Surface Plasmon Polariton and its Application to Surface Enhanced Raman Scattering (SERS) 4.1 General Theory of Propagating Surface Plasmon on Smooth Surfaces 100 4.2 Excitation and Evidence of Propagating Surface Plasmons 105 4.3 Exciting Propagating Surface Plasmons Optically 108 4.3.1 Exciting Surface Plasmons via Kretschmann-Raether Configuration 108 4.3.2 Exciting Surface Plasmons via Otto Configuration 114 iii 4.4 Surface Plasmon Polariton Assisted Surface Enhanced Raman Scattering on Ultrasmooth Metal Surface 4.5 118 Electromagnetic Theory of Semi-infinite Metal Containing Dielectric Nano-cavities 121 4.5.1 Maxwell Formulation 122 4.5.2 Computational Results 123 4.5.2.1 Field Enhancement by Isolated Dielectric Spherical Nano-cavities of Different Sizes 124 4.5.2.2 Field Enhancement by Isolated Dielectric Nano-cavities of Different Aspect-Ratios 128 4.5.2.3 Influence of the Dielectric Constant of an Isolated Dielectric Non-spherical Nanocavitites on the Field Enhancement 130 4.5.2.4 Influenece of the Cavity Depth on the Field Enhancement 132 4.5.2.5 Closely-packed Peridoic Arrays of Embedded Dielectric Nano-cavities 134 4.5.3 Discussion 137 4.6.1 SPP-Assisted Surface Enhanced Raman Scattering on Ultrasmooth Metal Surface with Embedded Dielectric Nano-cavity 4.7 144 Conclusion 146 Chapter Fabrication of Dielectric Spheroidal Nano-cavity Embedded Substrate for Surface Enhanced Raman Scattering 5.1 Fabrication Methodology 148 5.2 Fabrication of Dielectric Nano-cavity Embedded Metal Substrate 153 5.2.1 Formation of Ultrasmooth Metal Surface on a Mica Template 153 5.2.2 Ar+-Assisted Ion Beam Deposition of Au on Atomically Smooth Mica substrate 153 5.2.3 Fabrication of Nano-particle Monolayers 154 iv 5.2.3.1 Functionalization of Au surface 155 5.2.3.2 Fabrication of Spherical Nano-particle Monolayers by Convective Assembly 156 5.2.3.3 Fabrication of Oblate Nano-particle Monolayers by Convective Assembly 159 5.2.4 Electrodeposition 163 5.2.5 Attachement of Plated Samples to Solid Support and Peeling of Mica Template 163 5.3 Characterization of the nano-cavity Embedded Au Substrates 164 5.3.1 FE-SEM Cross-sectional Imaging of the Electroplated Spherical and Oblate Nanoparticle Array 164 5.3.2 Characterization of Exposed Au Surface by FE-SEM and AFM 167 5.3.3 Resistance Test Toward Thiol Functionalization and Buffer Solutions 168 5.3.4 Scanning Near-Field Imaging of Nano-cavity Embedded Au Substrates 170 5.4 Conclusion 173 Chapter Dielectric Nano-cavity Embedded SERS Substrates in the Study of a Model Protein 6.1 Setup of Measurement Systems 6.1.1 PDMS ATR Device for SERS and Surface Plasmon Response (SPR) Measurements 174 175 6.1.2 Setup for Measuring SPR Response of the Nano-cavity Embedded Au Substrates 182 6.1.3 Setup of Total Internal Reflection Raman System: Otto Setup 184 6.2 Preparation of Protein Sample and SERS Substrates 185 6.2.1 Functionalization of Nano-cavity Embedded Au Substrates 185 6.2.2 Preparation of Periodic Ag-Film Overcoated Nano-spheres Monolayer 186 6.3 SPR Response Nano-cavity Embedded au Surfaces 186 6.4 SERS from Nano-cavity Embedded Au Surfaces 189 6.5 SERS of a Protein Sample: Cytochrome-c 194 v 6.6 Conclusion 202 Chapter Conclusion and Future Direction Appendix I Appendix II Appendix III Appendix IV Appendix V Appendix VI Appendix VII Conference Abstracts and Publications References vi Synopsis Monolayers of large protein molecules self-assembled on a solid platform play a profoundly significant role in many fields ranging from biophysical sciences to biology Study of such a protein system permits insights into the interplay between protein–protein interactions, protein-surface interactions, as well as protein conformational changes, and thus allows for mechanistic understanding of a protein layer Unfortunately, conventional techniques such as X-ray crystallography and Nuclear Magnetic Resonance are not stuited for this purpose Hence, any progress made in advancing the technique for analyzing planar macromolecular systems is highly desirable The current thesis describes the development of a photonic device capable of nano-focusing light energy, and demonstrates its application to the Surface Enhanced Raman Spectroscopic (or SERS) study of a macromolecular layer Chapter provides a background on protein layers (PLs) Subsequently, various techniques previously used for PLs are introduced Their operating principles and limitations are discussed This is followed by a brief description on techniques previously developed for protein layer studies Lastly, the potential of a unique surface sensitive Raman scattering phenomenon known as the Surface Enhanced Raman Scattering (SERS) in probing conformational changes in molecular layers adsorbed on an interface is discussed In particular, it is stressed that conventional SERS provides meaningful Raman spectra only if small adsorbates (< nm) are used This greatly limits its value in studying layers of macromolecules such as proteins, whose sizes are often larger than nm The objective of this study is thus to address this issue vii Chapter describes the general quantum theory of SERS It can be shown that nanofocusing of light fields by a roughened metal surface can lead to intensified surface fields capable of enhancing the Raman scattering of adsorbates Experimental evidence of metal surface induced SERS are then introduced, and shown to conform to the theory Chapter points out the limitations in conventional SERS In particular, a technique is developed to allow for measurements of the spatial extension of the SERS fields on roughened surface Additionally, various artifacts brought about by the SERS-active metal surface are stressed Chapter introduces the concept of surface-plasmon polariton assisted SERS on a smooth metal surface Subsequently, the use of a smooth metal surface with embedded dielectric spheroidal nano-cavities for SERS study is proposed Through a rigorous analysis of the Maxwell equations obtained for such a system, it is shown that nano-focusing of a light field is also possible on a smooth metal surface This is ascribed to a compression of the propagation vector of the surface plasmon polariton as a result of the embedded nano-cavity underneath the surface Chapter details the fabrication of the cavity system proposed in Chapter Different substrates bearing embedded nano-cavities of different shapes were fabricated Characterizations of the final substrates are given Chapter describes the various experimental setups for carrying out SERS experiments with the dielectric cavity-embedded substrates Comparison of results with theoretical predictions is discussed Lastly, SERS spectra of a test protein, Cytochrome c (2.9 viii (a) 14000 (b) 12000 Instantaneous Raman (count/s) C Co = 8837.8 10000 dS = −22 dt 8000 6000 4000 B 2000 Co = 1255 dS = −372 dt 0 50 100 150 200 250 Tim e (s ) Figure 23 (a) Logarithmic plot of integrated Raman intensity versus time Curve B and C were obtained with the 8-µm liquid layer geometry Curve B was derived from electrostaticallyimmobilised Au nanoparticle monolayer, while curve C from MPTMS-immobilised nanoparticles (b) A plot of the instantaneous Raman signal S (t ) versus time Curve annotation same as (a) 84 Table I Fitting values for d, S0, C0 and λ d (µm) So Co λ Sample A 170 9765 2790 0.015 Sample B 11346 1255.5 0.033 Sample C 88 3721 8837.8 0.0055 The explanation for the thermal-induced lateral diffusion of the Au nano-particles is hereby depicted in Figure 24 Initially, the Au nano-particles in the sample are immobilized and surrounded by a thin diffuse double layer (DDL) owing to the high ionic strength of the PBS solvent in which the CV is dissolved Upon laser irradiation at the coupled plasmon peak, nano-particles that are close enough (No–state) to undergo a strong electromagnetic coupling will begin to heat up (see Figure 24b) due to SPRH The increase in the particle’s temperature results in an increase in the DDL thickness, causing neighboring particles to repel each others (121) To “relax”, the closely-spaced particles undergo a lateral particle diffusion as shown in Figure 24c through which the AIG of these particles become widened (N1–state) Consequently, this reduces the coupling field and is eventually manifested as a decaying SERS signal in our data However, as the field decreases so does the SPRH as well as the rate of the diffusion Finally, the particles cease to move, and the SERS signal levels off to a plateau Though it is possible that optical forces might play a role in promoting the particle diffusion (122) Such effect, which is laser-intensity dependent, however cannot account for the effect of the LLT on the decay rate, since the laser intensity at the sample are essentially unaffected by the LLT due to the low NA of the microscope objective used in the current study 85 The dependence of the decay rate (or Ko) on the LLT, on the other hand, is likely due to a convective flow within the liquid layer induced by the SPRH Thus a thicker layer implies a more efficient cooling of the SERS surface since there are more circulating liquid over it One thus expect samples with the 8-µm liquid layer to exhibit a higher λ value (i.e higher Ko, as a result of poor cooling) than that with the 170-µm liquid layer, as indicated in Table I (a) (b) + + + + + + + + + + - - - - + + + + + + + - - + + + + - + - - - + + + + + - - -+ - (c) + + + + + - - - +- - + + - - + - -+ - - + + + Figure 24 (a) Au nano-particles immobilised on an electro-statically charged glass surface with thin double diffuse layers (b)Thickness of the double diffuse layer increases as the surface plasmon related heating causes positive counterions to diffuse away from the particle surface Raman scattering from the molecule (the diamond) adsorbed in between particles is strong (c) Particle diffusion cease as the electromagnetic coupling weakens Raman scattering from the absorbed molecules reduces 3.2.2 A Stable Substrates for Reproducible Results The relationship between the Raman decay rate and the liquid layer thickness is detailed It was shown that the Raman decay obeys a three-state model Plasmon induced 86 particle diffusion is shown to be responsible for the irreversible loss of Raman signal A computer simulation has been carried out to validate the proposed hypothesis Finally, based on the three-state model, the product of λ and C derived from experimental data was shown to be a constant, in accordance with the model The current study indicates that changes in the average inter-particle gap separations due to plasmonic heating could lead to inevitable signal losses As such, D Keating‘s scheme of probing proteins with the inter-particle plasmon fields could lead to a lowered S/N 3.3 Effects of SERS-active Metallic Sandwiches on Protein Structure 3.3.1 Structural Changes in Cytochrome-c Sandwiched in Between Nanoparticles Beside irreversible signal loss, structural changes might also occur in a protein trapped between sandwiching metal nano-particles In this section, structural deformations in a test protein, Cytochrome C (Cyt-c) sandwiched in between Au nano-particles are studied in terms of Circular Dichroism (CD) and Surface Enhanced Raman Resonance Scattering (SERRS) analysis Cyt-c was chosen for the present study because of the large amount of previous reports on its vibrational spectra (both solution resonance Raman and resonant SERS) and its adsorption behavior at several types of surfaces (123-133) Preparation of the Cyt-c solution and the Au colloid is detailed in Appendix IV Preparation of Agarose Gel for Cyt-c binding studies is detailed in Appendix V Protocol for the SERRS experiments of Cytc and the Raman system setup can be found in Appendix VI TEM sample was prepared by flushing the analyte liquid droplet through a copper grid, followed by washing with distilled water before letting it to dry overnight in a 40 % humidity environment 87 3.3.1.1 Agarose Gel Analysis An Algarose Gel Electrophoresis was carried out to ascertain binding of the Cyt-c protein to the Au-nano-particle (15 nm in the current study) prior to a SERS experiment Electrophoresis bands corresponding to three Cyt-c-Au samples prepared with an initial Cytochrome-c concentration [Cyt-c] of 0.023, 0.047, 0.65 mg/ml respectively are shown in Figure 25 Band-positions are clearly different among these samples and shift upward with increasing incubating protein concentrations, indicating a lowering in mobility due to Cyt-c protein binding to the colloid The Cyt-c-Au sample prepared with 0.047 mg/ml Cyt-c was then arbritarily chosen for CD analysis and subsequent Raman experiments 0.065 mg/ml 0.047 mg/ml 0.023 mg/ml Figure 25 Electrophoresis bands for three Cyt-c-Au samples prepared with different Cyt-c concentrtation The arrow indicates the running direction 88 3.3.1.2 TEM Analysis Trapping of Cyt-c between Au nano-particles was achieved by adding 1M of NaCl in the Cyt-c-Au solution at a 1:10 ratio (v/v) followed by a 30-s incubation The resultant aggregated sample was subsequently analysed under TEM Top panel in Figure 26 shows a TEM micrograph of an Au cluster derived from the aggregated Cyt-c-Au sample A TEM image of a NaCl-induced aggregated sample of un-coated Au colloid is also shown in the figure (bottom panel) A visible gap in the Cyt-c-Au dimer suggests successful trapping of the protein The gap separation is measured to be about 1.9 nm, which is less than the 2.9-nm size of a native Cyt-c protein (134) This is however not un-expected as the sample was dried on a copper grid prior to TEM imaging, which could have resulted in the two particles being drawn closer to each other due to the liquid capillary forces, and squeezing the trapped protein in the process (a) 89 (b) Figure 26 TEM image of aggregated samples (a) Aggregated sample of Cyt-c-Au complex (b) Aggregated sample of uncoated Au colloid Scale Bar = 50 nm 3.3.1.3 Circular-Dichroism Analysis Changes in the secondary structure of Cyt-c under aqueous condition following conjugation to Au nano-particles can be conveniently studied via CD spectroscopy (135) CD spectrum corresponding to the 0.042-mg/ml Cyt-c solution, the Cyt-c conjugated Au colloid (here denoted as Cyt-c-Au-Coll), and the NaCl-induced Cyt-c trapped Au aggregate (denoted as Cyt-c-Au-Aggr), was obtained at 25 ºC using a 1-mm path length quartz cuvette, and are shown in Figure 27 The sample chamber has been purged with N2 prior to the measurements Note that the CD contributions from the unbound Cyt-c has been subtracted from the Cyt-cAu-Coll and the Cyt-c-Au-Aggr CD spectra To determine the secondary structure content of the protein, these spectra were subjected to spectral deconvolution using the SOMCD algorithm (136), and the results are tabulated in Table II The resolved secondary structure agrees well with that previously reported for Cyt-c (135) Spectral deconvolution also shows a slight decrease in α-helicity in Cyt-c’s secondary structure upon conjugation with Au colloid from 35% to 33 %, which is likely due to minor reorganization of the internal 90 structures induced by the electrostatic interactions between the Lysine-rich patch around the heme group and the negative surface charge on the Au nano-particles (91, 137) This result agrees well with Aubin-Tam et al.’s report (135) Note that the observed CD signals originate solely from the Cyt-c and there is no contribution from the Au colloid for it is not optically active This has been verified by measuring the CD spectrum of the Au colloid in the absence of Cyt-c Cyt-c-Au-Aggr, on the other hand, exhibits a much larger conformational change with α-helicity reduced to only 12 % and a significant increase in the random-coil content from 29.2 % to 45 % CD spectra of Cyt-c 0.1 CD (1e4 deg cm2 dmol-1) -0.1200 210 220 230 240 250 260 -0.3 -0.5 Cyt-c-Au-Aggr -0.7 Cyt-c-Au-Coll -0.9 Cyt-c -1.1 -1.3 -1.5 Wavelength (nm) Figure 27 CD spectra of Cyt-c solution, Cyt-c conjugated Au colloid, and NaCl induced Cytc-Au aggregate 91 Table II Secondary structure of Cyt-c and Cyt-c conjugates Cyt-c Cyt-c-Au-Coll Cyt-c-Au-Aggr α-helix (%) 35.2 ± 6.4 33 ± 7.5 12 ± 2.2 β-sheet (%) 22.4 ± 12 20 ± 23 ± 0.3 γ-turns (%) 13 ± 2.2 14 ± 3.2 24 ± 0.2 Unordered (%) 29.2 ± 32 ± 5.5 40 ± 2.5 3.3.1.4 SERRS of Cytochrome C Structural changes in conjugated Cyt-c and Cyt-c trapped within Au aggergates can be infered from the SERRS spectra of the Heme group (137) A 514-nm excitation was used to match the Q-Band (500 – 580 nm) of the Heme in order to induce a resonant enhancement of the Raman scattering (91) At this wavelength, scattering mechanism of the porphryrin are of the B-term type based on the vibronic coupling between the excited states of the B and Q optical transitions (138-141) At 514-nm, Raman scatterings are predominantly derived from the A1g, A2g, B1g sysmetry modes (140) Figure 28 shows RRS and SERRS spectra of Cyt-c, Cyt-c-Au, and Cyt-c-Au-Aggr samples respectively obtained using a Jobin Yvon spectrometer Note all samples are prepared in a phosphate buffer with a pH of 7.0 Band assignment is shown in Table III The presence of 1370, 1588 and the 1638 cm-1 bands in Cyt-c RRS suggests a low-spin state with an oxidised iron Fe3+ as the prefered configuration of Cyt-c in the current solvent condition (140) Similar peak positions also appear in the SERRS of Cyt-c-Au conjugates, implying no major conformational changes upon binding to isolated Au colloid This result is consistent with the above CD analysis and also with the SERS data obtained by Christine D Keating and MacDonald for Cyt-c absorbed on Ag sols at high concentration of Cyt-c (91, 129) SERRS derived from the Cyt-c-Au-Aggr sample 92 however shows some dissimilarity from previously reported results In particular, the ν10 spin-marker band has down-shifted to 1627 cm-1 from 1638 cm-1, indicating a switch from a low-spin state to a high-spin one Such a transformation signifies that Fe has undergone a ligand loss, likely due to a severe conformational change in the trapped Cyt-c’s internal structure (124, 126, 142) This result also agrees with the CD spectrum discussed above for 1638 1627 1568 1406 1370 1588 the Cyt-c-Au-Aggr Resonance Raman of Cyt-c SERRS of Au-Cyt-c-Aggr SERRS of Au-Cyt-c Figure 28 SERRS and RRS spectra of Cytochrome-c Excitation wavelength = 514 nm Laser power = 150 mW Integration time = 120 s for Cyt-c; 60s for Cyt-c-Au; 100s for CytAu-Aggr 93 Table III Band Assignments for the SERRS spectra of Cyt-c Cyt-c-Au- Band Assignment V(pyr half-ring)sym, A1g (V4 Heme Cyt-c Cyt-c-Au Aggr 1370 1366 1369 1406 1404 1406 1568 1565 1567 1588 1588 1586 - - 1627 1638 1638 - Oxidation State Marker) V(pyr quarter-ring), A2g (V29) V(Ca-Cm)asym,A2g (V19 High-Spin Marker) V(Ca-Cm)asym,A2g (V19 Low-Spin Marker) V(Ca-Cm)asym,B1g (V10 High-Spin Marker) V(Ca-Cm)asym,B1g (V10 Low-Spin Marker) 3.3.2 Structural Study of Protein Adorbed on SERS Substrates The current experiment of SERRS on Cyt-c molecules trapped within metal colloidal aggregates serves to demonstrate the instability of a protein toward nano-particle agglomeration It should be pointed out that the current observations corroborates quite well with Hongxing Xu et al observations of structural damages in hemoglobin protein trapped inbetween two SERS-active Ag nano-particles (143), but unfortunately disagrees with Christine D Keating’s observation that Cyt-c remains intact within Ag/Au aggregates (91) While the protein studied by Christine D Keating and that in the current experiment is of the same species, both used different [Cyt-c]/[Au] ratios By way of centrifugation, it is estimated that 94 only about 23 % of the 0.042 mg/ml Cyt-c added is absorbed onto the Au nano-particles in the current experiment Assuming a radius of 15 nm for the Au nano-particles (see TEM image in Figure 26) and a particle concentration of × 1013 particles/ml as calculated through an optical extinction measurement, one obtains a surface coverage density of 60 molecules/particle, i.e only 58 % of the total Au surface is covered as opposed to a monolayer coverage in Christine D Keating’s case (91) As a consequence, the conformation of the bound Cyt-c molecules in the aggregates is not restricted (as in a well-packed monolayer configuration) and thus the protein is prone to denature under the compressing van der Waals forces exerted by the sandwiching Au nano-particles Although Cyt-c treated with twice as much [NaCl] (i.e 200 mM) has been shown to remain intact based on Yiqing Feng‘s 2D H1 NMR analysis (144), it is of worthy to further confirm that the protein denaturation observed above in Cyt-c-Au-Aggr is not a salt-induced effect To this end, a CD spectrum for a Cyt-c-Au-Aggr sample prepared with a higher initial [Cyt-c] of 0.12 mg/ml was obtained Same [NaCl] as that utilized for the Cyt-c-Au-Aggr sample that was discussed above was used here to induce particle agglomeration Figure 29 shows the CD spectra of the newly prepared Cyt-c-Au-Aggr as well as the un-coated Cyt-c No significant denaturation was observed in the current Cyt-c-Au-Aggr sample on the basis that the Cyt-c-Au-Aggr CD spectrum closely resembles that of Cyt-Au-Coll, even though the same amount of NaCl has been used to induce aggregation here This is attributed to a saturated Cyt-c layer around the Au nano-particles owing to the higher initial [Cyt-c] used to prepare the sample A crowded packing implies a constrained protein conformation, which prevents the protein from denaturation under the pressing forces exrted by the Au nanoparticles (145) The current observation thus serves as an evidence to indicate that the cause of the protein conformational changes depicted by the Cyt-c-Au-Aggr CD spectrum shown in Figure 28 is not the added NaCl but is the aggregating Au nano-particles A further SERRS 95 study (see Figure 30) on the current Cyt-c-Au-Aggr sample also collaborates with the CD data In particular, the Cyt-c-Au-Aggr SERRS spectrum resembles that for Cyt-c, and agrees well with Christine D Keating’s experimental results (91) More importantly, the band position of the spin marker at 1638 cm-1 indicates a low-spin Fe state, which suggests the absence of ligand loss Hence, the trapped protein molecules are intact only at a higher [Cytc], but not so when a lower initial [Cyt-c] is used CD spectra of Cyt-c 0.5 0.3 CD (1e4 deg cm2 dmol-1) 0.1 -0.1200 210 220 230 240 250 260 -0.3 -0.5 Cyt-c -0.7 -0.9 Cyt-c-Au-Aggr -1.1 -1.3 -1.5 Wavelength (nm) Figure 29 CD spectra of the newly prepared Cyt-c-Au-Aggr sample as well as the un-coated Cyt-c sample 96 1300 1200 Raman Intensity (arb unit) 1100 1000 900 Resonance Raman of Cyt-c 800 700 600 SERRS of Cyt-c-Au-Aggr 500 400 300 1350 1400 1450 1500 1550 1600 1650 Raman shift (cm-1) Figure 30 SERRS and RRS spectra of Cytochrome-c Cyt-c-Au-Aggr was prepared with an initial [Cyt-c] of 0.12 mg/ml Excitation wavelength = 514 nm Laser power = 150 mW Integration time = 120 s for Cyt-c; 100s for the Cyt-c-Au-Aggr The current study is imperative as it highlights the major problems concerning the application of SERS-active aggregates in protein studies (146) Particularly, it demonstrates that one must always practice caution when interpreting SERS protein data derived from metal aggregates; one must factor in the possibility of protein denaturation since the protein may be perturbed from its native state Secondly, care must be taken to ensure a monolayer coverage of proteins within the SERS-active aggregates if significant structural changes are to be avoided or minimized However, this may be difficult to achieve if the protein of interest is non-spherical (such as Integrin) or is lipid membrane-bound Furthermore, access to the active sites on the protein trapped between nano-particles is also severely limited, preventing the use of this scheme for the studies of ligand-protein interactions To this end, a smooth SERS active metal surface appears to be a better choice, and this is exactly the 97 objective of the current thesis; to study SERS of protein attached onto a smooth surface as will discussed later in Chapter and 3.4 Conclusion This chapter considered at first the limited spatial extension of hot-spot fields responsible for SERS Particularly, the hot-spot volumes on Si-Au nano-particles as well as on AuFON were measured and shown to be extremely confined in most cases to a region less than nm from the SERS-active surface This obviously could limit the values of these substrates in the study of macro-molecules of biological relevance Sandwiching of protein molecules in between SERS-active nano-particles was proposed by Keating and could potentially resolve the issue of confined hot-spot volumes because of the more uniform plasmon field distribution set up in the interstitial space within a dimer However, surface plasmon induced heating and possible protein denaturation as pointed out above (§ 3.2 and 3.3) could impede its use in the study of proteins using SERS It is thus the objective of this thesis to show how to resolve these issues by carrying out SERS on a smooth metal surface as will be discussed in the next chapter 98 ... Theory of Enhanced Raman Scattering and its Experimental Verifications 2 .1 General Theory of Raman Scattering 13 2.2 Enhanced Raman Scattering 13 2.2 .1 Chemical Raman Enhancement 14 2.2.2 Electromagnetic... Schematic diagram of SNOM system for phase measurements 17 0 17 1 Fig 15 A typical AFM and SNOM mappings of a 10 0-nm nano- cavity substrate with cavitysurface distance of (a and b) 10 nm, and (c and d).. .A Method for 3D Nano- focusing of Optical Energy and its Application to Surface Enhanced Raman Spectroscopic Study by Kiang Wei Kho SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE