Home

Coupling Fluorescent molecules to nanophotonic structures

1. 24 Atwater H A amp Polman A Plasmonics for Improved Photovoltaic Devices Nature Materials 9 205 2010 25 Novotny L amp Stranick S J Near field Optical Microscopy and Spectroscopy with Pointed Probes Ann Rev of Phys Chem 57 303 2006 26 Schuller J A et al Plamonics for Extreme Light Concentration and Manipulation Nature Materials 9 193 2010 27 Crozier K B Sundaramurthy A Kino G S amp Quate C F Optical antennas Resonators for local field enhancement J Appl Phys 94 4632 4642 2003 22 28 Sundaramurthy A et al Field enhancement and gap dependent resonance in a system of two opposing tip to tip Au nanotriangles Phys Rev B 72 165409 2005 29 Boyd G T Yu Z H amp Shen Y R Photoinduced luminescence from the noble metals and its enhancement on roughened surfaces Phys Rev B 33 7923 7936 1986 30 Beversluis M R Bouhelier A amp Novotny L Continuum generation from single gold nanostructures through near field mediated intraband transitions Phys Rev B 68 115433 2003 31 Bouhelier A Beversluis M R amp Novotny L Characterization of nanoplasmonic structures by locally excited photoluminescence Appl Phys Lett 83 5041 5043 2003 32 Yablonovitch E Inhibited Spontaneous Emission in Solid State Physics and Electronics Phys Rev Lett 58 2059 1987 33 John S Strong Localization of Photons in Certain Disordered Dielectric Superl
2. walk the beam if it is coming in at an angle The mirror closest to the microscope M5 in Figure C 3 is the one that is used for angular adjustment of the beam and the previous mirror M4 in Figure C 3 is used for positional beam adjustment If further mirrors are used only the final two mirrors should be used for the final beam alignment Figure C 4 shows images of the beam in various states of alignment To accomplish this walk use the following method 143 Figure C 4 Alignment of the confocal beam using the Genwac CCD camera a The beam is centered in intensity but off of the ideal optical axis need to walk beam using both mirrors b Mirror M4 adjusted to move beam closer to ideal position but now is going in at an angle c Adjust Mirror M5 angle to fix angle and achieve properly aligned beam 1 Center beam intensity use CCD camera such as the Genwac Defocus the beam slightly The beam should have nice and symmetrical intensity Its position may or may not be in the center of the screen If a dichroic mirror is used especially with linearly polarized light the beam intensity appears to have a cloverleaf appearance This is an artifact of the polarization dependence of the coatings on the dichroic Don t worry about it the beam is actually quite clean The intensity should still have symmetric intensity across the mirror planes regardless of its shape This problem is not seen with a silvered mirr
3. Confocal Pinhole Collimating the Emission Signal Emission filters Aligning the Avalanche Photodiode APD Spectrometer Path C 4 Alignment of CCD Monochromator C 4 1 C 4 2 C 4 3 C 4 4 C 4 5 C 4 6 C 4 7 Introduction Input mirror Focusing lens Entrance slit Concave mirror Grating Focusing Concave Mirror xiii 132 133 134 135 138 138 139 141 143 144 144 144 146 147 148 148 148 149 149 149 150 150 151 151 152 153 C 4 8 Exit port C 4 9 Camera C 4 10 Final alignment C 4 11 Final comments C 5 Software C 5 1 Introduction C 5 2 Using Bin APD counts LabVIEW Program C 5 3 Topometrix Software for Confocal Scanning C 6 Scanning stages C 6 1 Piezoelectric Scanner C 6 2 Calibration and linearization of stages C 6 3 Hardware signals in out of ECU controller XIV 153 153 154 154 155 155 155 157 159 159 160 161 List of Figures Figure 1 1 Figure 1 2 Figure 1 3 Figure 1 4 Figure 1 5 Figure 1 6 Figure 1 7 Size mismatch between the diffraction limit and a nanoscale emitter 5 Surface plasmon polariton excited at a metal dielectric interface i Response of free electrons in a metal colloid to an AC electromagnetic field tuned to the particle s plasmon resonance 8 SEM scanning electron microscopy image of a gold bowtie nanoantenna fabricated with electron beam E beam lithograph
4. th Figure 3 7 Modeled enhancement of QE as a function of intrinsic QE a Theoretical predictions based on FDTD simulations for the change in intrinsic quantum efficiency 77 when a molecule is placed near a bowtie nanoantenna 77 The FDTD simulations provide f and fa and the curves show the values of Eqn 3 5 b Same data as in a this time plotting enhancement of quantum efficiency against the intrinsic quantum efficiency In both figures TPQDI s intrinsic quantum efficiency 7 2 5 is circled in red After Ref This number is in good agreement with the maximum experimentally measured enhancement factor of 1340 especially if one takes into account the experimental uncertainty in determining the exact location and orientation of the molecule Figure 3 6 also shows the enhancement factors at different positions In the vertical direction z direction the functions are relatively constant in the gap region and fall off quickly above the metal surface Figure 3 6b In the gap region the maximum total fluorescence enhancement occurs at the center Figure 3 6c and decreases closer to the metal tips because of lower quantum efficiencies from increased Ohmic losses cf Figure 3 6d radiative and Figure 3 6e nonradiative The same analysis above 64 indicates that a molecule with a high intrinsic quantum efficiency e g 7 gt 25 in fact would not have any quantum efficiency enhancement by the same anten
5. 137 optical images This platform serves near field and far field imaging experiments well C2 Input Optics Figure C 1 shows the microscope and its various components in its entirety which can be broken up into the following general components input optics microscope parts output optics and AFM parts The optical parts will be discussed here with AFM operation discussed later Figure C 2 shows the important confocal microscope input optics discussed here al iu Output optics ee Figure C 1 General microscope view Note several components including input optics output optics AFM head the microscope and the CCD Spectrometer 138 Figure C 2 Input optics for confocal microscopy including the single mode fiber SMF which is a spatial filter the collimating objective NA 0 18 the rear mirror Mp and the final mirror Mp C 1 1 Gaussian Beam Profile When creating a confocal excitation beam path it is important to have some sort of a spatial filter to produce a clean diffraction limited spot before entering the microscope Note that taking great care in producing a clean focal spot at the sample both lowers background noise and increases signal both of which are important for single molecule microscopy For laser diodes the emission is often a cat eye shape far from the ideal Gaussian beam shape The most widely used method for producing a clean beam is sending the beam through a
6. A Spectral tuning of plasmon enhanced silicon quantum dot luminescence Appl Phys Lett 88 131109 2006 12 Chen Y Munechika K amp Ginger D Dependence of Fluorescence Intensity on the Spectral Overlap between Fluorophores and Plasmon Resonant Single Silver Nanoparticles Nanolett 7 690 696 2007 13 Gerard D et al Nanoaperture enhanced fluorescence Towards higher detection rates with plasmonic metals Phys Rev B 77 045413 2008 14 Magde D Elson E amp Webb W W Thermodynamic Flucutations in a Reacting System Measurement by Fluorescence Correlation Spectroscopy Phys Rev Lett 28 705 1972 15 Eigen M amp Rigler R Sorting Single Molecules Application to Diagnostics and Evolutionary Biotechnology Proc Natl Acad Sci U S A 91 5740 5747 1994 88 16 Rigler R Fluorescence correlations single molecule detection and large number screening applications in biotechnology J Biotechnol 41 177 186 1995 17 Rigler R Elson E amp Elson E in Springer Series in Chemical Physics Vol 65 Fluorescence Correlation Spectroscopy Springer Series Chem Phys eds Schaefer F P Toennies J P amp Zinth W SpringerRigler R Berlin 2001 18 Widengren J Mets U amp Rigler R Fluorescence correlation spectroscopy of triplet states in solution a theoretical and experimental study J Phys Chem 99 13368 13379 1995 19 Hess S T Huang S Heikal A A a
7. Asano T Spontaneous Emission Control by Photonic Crystals and Nanocavities Nature Photonics 1 449 2007 2 Englund D et al Controlling Cavity Reflectivity with a Single Quantum Dot Nature 450 857 2007 3 Yoshie T et al Vacuum Rabi Splitting with a Single Quantum Dot in a Photonic Crystal Nanocavity Nature 432 200 2004 4 Altug H Englund D amp Vuckovic J Ultrafast Photonic Crystal Nanocavity Laser Nature Physics 2 484 2006 5 Barth M Nusse N Lochel B amp Benson O Controlled Coupling of a Single Diamond Nanocrystal to a Photonic Crystal Cavity Opt Exp 34 1108 2009 6 Barclay P Santori C Fu K Beausoleil R G amp Painter O Coherent Interference Effects in a Nano assembled Diamond NV Center Cavity QED System Opt Exp 17 8081 2009 7 Hennessy K et al Quantum Nature of a Strongly Coupled Single Quantum Dot Cavity System Nature 445 896 2007 8 Thon S M et al Strong Coupling through Optical Positioning of a Quantum Dot in a Photonic Crystal Cavity Appl Phys Lett 94 111115 2009 9 Rivoire K Faraon A amp Vuckovic J Gallium Phosphide Photonic Crystal Nanocavities in the Visible Appl Phys Lett 93 063103 2008 10 Choi Y et al GaN Blue Photonic Crystal Membrane Nanocavities Appl Phys Lett 87 243101 2005 11 Barth M Kouba J Stingl J Lochel B amp Benson O Modification of Visible Spontaneous Emission with Si
8. PMMA is spun onto the substrate at SNF using the Laurel spin coater To spin the resist 1 mL of 2 950k PMMA Microchem in anisole is pipetted onto the coverslip surface through a 450nm pore size filter The Laurel spin coater can be programmed to spin the coverslip at 3 speeds To achieve 60nm thick PMMA films 2 PMMA in anisole is first spun for 10s at 300RPM then 40s at 6 000RPM and finally 10s at 300RPM Resist films of varying thickness can be calibrated by first spinning onto Silicon pieces at different speeds and measuring the final thickness using SNF s Nanospec film thickness measurement tool Once the 35 resist is spun the coverslip is placed onto a 180 C hot plate for gt 2 minutes to bake out the resist and ensure the sample does not outgas under vacuum Now that the coverslip is coated in resist the sample is loaded into SNF s Raith 150 for e beam exposure Step 2 An abbreviated set of instructions for defining bowties using the Raith 150 can be found in Appendix A but this machine requires a week long training course in order to use In step 2 the sample is loaded into the Raith 150 and the electron beam is aligned so that it is in focus and stigmated properly This alignment ensures that the electron beam is focused to the smallest spot possible allowing the user to define the smallest gap bowties possible 10nm reproducibility The resist that is exposed to the electron beam becomes more soluble than the une
9. aluminum film with 80nm diameter holes 18 on a transparent substrate such as in Figure 1 11 The fact that illumination is now evanescently coupled means that excitation in confined in the propagation direction The production of fluorescence is also confined in the other two dimensions due to the metallic hole excluding molecules from the excitation region leaving a very small total excitation volume which allows for single molecule FCS or single molecule direct observations to be performed at much higher fluorophore concentrations in solution above the waveguide 10 uM This scheme has been extended to implement real time single molecule sequencing of DNA as well as real time translation of RNA into protein and is also being commercialized by the company Pacific Biosciences Chapter 4 concerns an extension of this concept to instead use bowtie nanoantennas for single molecule FCS at high fluorophore concentrations Glass 19 Figure 1 11 Zero mode waveguide geometry for high concentration FCS Yellow circles are molecules that occasionally enter the hole in the aluminum and emit fluorescence into the collection optics 1 5 4 Conclusions This chapter has outlined a number of different research concepts from dielectric photonic crystal cavities to plasmonic optical antennas to fluorescence This thesis combines these areas to show that photonic crystal cavities and gold bowtie nanoantennas can be used in powerf
10. b FDTD simulation of electric field intensity of the fundamental cavity mode The mode is primarily y polarized c Schematic illustrating fabrication procedure 1 DNQDI PMMA is float coated over the entire structure i DNQDI PMMA is lithographically defined over cavity region d Bulk fluorescence emission spectrum of DNQDI when excited with a 633 nm HeNe laser measured with a confocal microscope and spectrometer The molecule has a peak in its absorption at this excitation wavelength e Chemical structure of DNQDI molecule After Ref 113 xxviii Figure 6 2 a Cross polarized reflectivity measurement of a cavity The box indicates fundamental cavity mode b Reflectivity spectrum of high quality factor fundamental cavity mode box in a Spectrum shows additional peaks at shorter wavelengths from higher order but lower Q cavity modes Solid line shows Lorentzian fit with quality factor 10 000 c Fluorescence collected using a confocal microscope approximately diffraction limited collection and spectrometer from the same photonic crystal cavity in a and b after molecules are deposited on cavity X polarized emission is shown in blue Y polarized emission is shown in red Inset Fluorescence measurements of fundamental cavity mode black box Line indicates Lorentzian fit with Q 10 000 d Quality factors measured for high Q cavity mode from reflectivity open circles before molecule deposition and fluorescence after
11. e For each of the 3 points focus and burn dots tick 1 tickbox go back to first burned spot click adjust go to next point e Repeat go through all points again until they are all in focus e Check that the sample is leveled well by attempting to burn a calibration dot somewhere in the intended write region If a dot cannot be burned easily redo this step 10 Set Doses e Exposure window gt Open dose calculator e Set doses area 110 line 300 dot 0 01 e Set stepsize 1 6nm e Calculate dwelltime by clicking on calculator e Note this needs to be done after measuring the beam current step 3 otherwise it will incorrectly calculate the dwelltime 11 Make sure the focus is set for the write e Adjust UVW Window gt Adjust W gt click Read then adjust 12 Write e Setup up position list with your patterns e On position list Scan all 128 Appendix B Focused lon Beam Lithography with FEI Strata This appendix contains instructions for running the FEI Strata dual beam FIB as well as additional tips for patterns with small features lt 5Onm Please note that these instructions are not a substitute for the required training sessions with a FIB trainer Check the latest training protocol to see current requirements for use of and training on the FEI Strata B 1 Start up 1 Log in e Log into CORAL with your CORAL id and password Enable the FIB e Log in to the FIB computer with your user id and pa
12. magenta curve in Figure 3 9a shows the fluorescence from bulk TPQDI in PMMA and yields tp 275 ps for molecules not coupled to a bowtie The green curve in Figure 3 9a corresponds to a SM with tp 78 ps while the red curve corresponds to a different SM with tp lt 10ps The red molecule therefore has a gt 28x decrease in Tr but this is due to changes in both the molecule s radiative and nonradiative rates Figure 3 9b is a scatter plot of SM decay lifetime versus fr for 73 molecules At low fr both small and large tr were observed in the data but all much smaller than the uncoupled value in magenta This result is expected since tr depends only upon the radiative and nonradiative rates while fr depends in addition upon local optical intensity and many combinations are possible for different molecule positions and orientations To achieve high fr the molecule s absorption and quantum efficiency must be significantly improved and this occurs only in the gap where both the radiative and nonradiative rates are faster which yields a greatly shortened lifetime Therefore only short lifetimes are to be expected for high fr molecules as observed in Figure 3 9b 2 A l 275 b 2 gt O15 80 a _ v 60 S afo fe pt 40 0 5 z d 20 Z 0m OL rd 0 05 1 0 200 400 600 800 Time Delay ns f Figure 3 9 Enhanced single molecule fluroescence time delay histograms a Magenta bulk
13. 2 6 Measurement of Enhanced Fields of Gold Bowtie Nanoantenna To finish characterizing the bowtie nanoantenna the enhanced fields when pumped at resonance were experimentally measured Gold has an intrinsic two photon photoluminescence TPPL that arises from transitions within the conduction band d band to sp band TPPL is proportional to the fourth power of the excitation electric field strength The ratio of the gold bowtie s TPPL to a smooth gold film s TPPL is then a measure of the enhanced fields of the bowtie nanoantenna In Ref the bowtie nanoantenna was pumped with a pulsed Ti Sapphire laser tuned to 830 nm resulting in TPPL emission TPPL emission from 460 nm to 700 nm was collected and compared to the emission from a smooth gold film of the same thickness in order to measure the local enhancement of fields due to the bowtie nanoantenna as a function of gap size Figure 1 7 It was found that as the gap of the bowtie gets smaller the electric field intensity E can reach values gt 10 times larger than the incident field intensity The two triangles of the bowtie require close proximity lt 50nm to be coupled so the field intensity quickly drops as the bowtie s gap size increases until reaching the field intensity expected for a single triangle beyond 50nm gap sizes 410 Gap vs JE i Enhancement E Enhancement 0 100 200 300 400 500 Gap nm Figure 1 7 Measurement of enhanced IEI fi
14. E is the electric field c is the speed of light is the frequency and w is the complex dielectric constant At high frequencies w is given by 2 e w 1 8 1 4 4mne 1 5 Wy ai where p is the plasma frequency n is the electron density e is the charge of an electron and m is the mass of an electron When p plasmons or coherent collective oscillations of the free electrons in the metal can be excited in the metal 1 2 3 Surface Plasmon Polaritons In real materials the penetration depth or skin depth of a metal when excited by visible or near IR light is a few tens of nm This means that only electrons near the surface of a metal are excited as shown in Figure 1 2 The condition necessary to excite a surface plasmon polariton SPP at the interface is found by applying appropriate boundary conditions for a metal dielectric interface to Maxwell s equations yielding the following dispersion relation k 2 lt 1 5 c Eqt m where k is the wavenumber while q and are the permittivities of the dielectric and metal respectively Notice that since q is real and positive for k to be real and thus yield a propagating wave the real part of m must be negative and larger than q This condition is satisfied at visible wavelengths for the metals silver and aluminum and in the near IR for gold and copper Dielectric ts Figure 1 2 Surface plas
15. Emission Time Delay Figure 2 2 Time tagging of photons is accomplished by measuring the time delay between a signal photon and the sync signal of a pulsed laser 2 3 Scattering Microscopy 2 3 1 Introduction Measuring the amount and wavelength of light that is scattered by a structure is essential for plasmonics research Plasmonic structures have resonances whereby they concentrate light but only at specific wavelengths that are dependent upon the material and shape of the structure Scattering measurements are often used to directly measure the plasmonic resonance of a nanostructure by illuminating the nanostructure with white broadband light The light scattered by the structure is collected and 31 dispersed onto a spectrometer alignment can be found in Appendix C Since the primary scattering signal is elastic and therefore at the same wavelength as the exciting light long pass filters cannot be used to discriminate between the two as is done in traditional fluorescence microscopy Instead scattering measurements are usually based upon methods which spatially separate the pumping light from the scattered light such as dark field microscopy or total internal reflection TIR microscopy the latter of which is used here Snell s law governs how light reflects off of a boundary and in particular predicts TIR since it shows that beyond a critical angle of incidence at a boundary between two materials of different indices all o
16. ICG and IR800cw in water and ethanol Measurements used IR800 phosphoramidite LiCor in methanol QE 15 as a QE reference 75 4 3 Bulk Bowtie Enhanced Fluorescence of Molecules in Solution Confocal measurements of concentrated dye solutions on bowtie nanoantennas were performed using the confocal microscope described in Chapter 2 Figure 4 2c f are confocal scans of IR800cw and ICG doped into 2 PVA in water solutions and spun to 30nm thick films on top of the bowtie surface These images reveal that the bowtie nanoantennas do enhance bulk fluorescence from these two molecules in rigid environments as was shown in Chapter 3 for TPQDI in PMMA Therefore these two molecules are good candidate molecules to look for enhanced fluorescence in solution Moving on to solution environments at 1uM concentration there are only 0 6molecules 100nm region The bowtie s enhanced region is only 20nm but even so it is easy to see fluorescence enhancement from bowties immersed in both ICG at luM concentration in water Figure 4 2d and IR800cw at 100nM concentration in ethanol Figure 4 2a It is as if the molecules linger longer in the enhanced region In fact the enhancement is actually due to molecules that are stuck to the substrate surface instead of floating in solution as will be shown below with several key pieces experimental evidence First the concentration dependence suggests that the surface is nearly saturated with sticking mo
17. Lett 95 017402 1 017402 4 2005 107 13 Zhou H et al Lithographically Defined Nano and Micro Sensors using Float Coating of Resist and Electron Beam Lithography J Vac Sci Technol B 18 3594 2000 108 Chapter 6 Lithographic Positioning of Fluorescent Molecules on High Q Photonic Crystal Cavities The research reported in this chapter has been previously published in K Rivoire A Kinkhabwala F Hatami W Ted Masselink Y Avlasevich K Miillen W E Moerner J Vu kovi Lithographic Positioning of Fluorescent Molecules on High Q Photonic Crystal Cavities Applied Physics Letters 95 123113 2009 published online September 23 2009 Experimental results were measured jointly by K Rivoire and A Kinkhabwala from Prof Vu kovic s and Prof Moerner s lab s respectively Gallium Phosphide samples were grown by F Hatami of Prof W Ted Masselink s group The DNQDI molecule was synthesized by Y Avlasevich of Prof K Miillen s group 6 1 Introduction Bowtie nanoantennas are not the only nanophotonic structures to alter fluorescence emission Photonic crystal cavities also act to confine light but they do so using total internal reflection and Bragg diffraction In this way photonic crystal nanocavities confine light into volumes smaller than a cubic optical wavelength with extremely high quality factor Q producing a strong interaction between light and 109 emitters located in or near
18. TPQDI in PMMA without bowtie nanoantenna Green SM on bowtie nanoantenna fp 271 lifetime 67 78 ps Red blue SM on bowtie nanoantenna excitation polarization parallel perpendicular to long axis Black instrument response function b Black Scatter plot of decay lifetime versus brightness enhancement for 73 SM s of TPQDI on bowtie nanoantennas Magenta Bulk TPQDI lifetime without bowtie nanoantenna present After Ref a1 3 7 Excitation Polarization Dependence Since the absorption and emission enhancements are decoupled the polarization of the excitation light should affect the fluorescence brightness enhancement which depends upon the absorption and emission enhancements but not the SM decay lifetime which only depends on the emission enhancement By changing the excitation polarization using an electro optic modulator every 1 5 s from parallel to perpendicular to the long axis of the bowtie 1 s for parallel 0 5 s for perpendicular fluorescence lifetimes of the same molecule as a function of excitation polarization can be measured Figure 3 10a shows a fluorescence time trace from a bowtie nanoantenna where red blue indicated parallel perpendicular excitation polarizations At 21 5 s there is a significant photobleaching step in the parallel excitation polarization channel while a much smaller drop in intensity can also be detected in the other channel This demonstrates that the total fluorescence enhancement is
19. Zoos N 0 4 5 S02 E A ane h 8o00 820 840 860 880 900 920 Wavelength nm Figure 4 4 a Spectra integrated over 10s from a 100nM concentration solution of IR800cw in ethanol with blue and without red a bowtie present as well as spectra from a 1uM concentration solution of ICG in water with green and without black a bowtie present Notice that none of the spectra contain Raman peaks b Normalized spectra from 100nM IR800cw with blue and without red a bowtie present Notice that the shape of the spectrum does not change depending on the bowtie s presence or 79 absence For both figures the laser filter cuts off emission 800nm and shorter causing aberrations in this spectral region particularly at 810 nm 4 5 FCS of Low Concentration Dye Solutions In a fluorescence correlation spectroscopy FCS experiment the fluorescence emission from a low concentration dye solution irradiated by a focused laser beam is analyzed by calculating the autocorrelation function _ 6h O Mblg t 0 GO TN 4 1 where I t is the fluorescence intensity on one of the two detectors at time t The autocorrelation asks the question on which time scales does the emission remain constant and which does it fluctuate The fluctuations can arise from diffusion as molecules move in and out of the focal volume or from internal dynamics of the emitter arising from triplet states other dark states or even the excited stat
20. as an easier approach to fabricating nanostructures when the Ga implantation inherent to FIB milling is acceptable 2 4 3 Float Coating EBL Resist Float coating is a useful technique for coating uneven substrates such as AFM tips and photonic crystal cavities described in this thesis with thin layers of PMMA for EBL optical experiments or other purposes Spin coating is the usual method for coating flat substrates with thin layers of PMMA but it fails when the surface is uneven Figure 2 6 due to buildup of resist around uneven features a b Resist Ges a Figure 2 6 A Spin coating resist onto a flat substrate yields a smooth even layer B Spin coating onto an uneven substrate leads to uneven coverage and buildup of resist at the base of features 40 Float coating is outlined in Figure 2 7 The first step is to place the uneven substrate pictured as an AFM tip into a Petri dish filled with water on top of a silicon wafer piece The silicon piece should be significantly lager than the sample but also smaller than the petri dish so that when the silicon piece is moved the sample can be moved within the petri dish In step 2 1 drop of a 1 PMMA solution in toluene is dropped onto the surface of the water bath The drop of polymer solution disperses on the water surface in step 3 The water bath should not be disturbed during this process to ensure the smoothest film possible Solvent choice is critical a 1 P
21. at 90 C for 30 minutes to remove any remaining water 98 Figure 5 4 SEM showing cantilever bending after float coating of E beam resist 99 Figure 5 5 E beam lithography process flow for nonconductive substrate a Deposit chrome onto float coated resist layer b Expose resist using Raith 150 E beam Lithography Tool c Etch chrome layer in CR14 chrome etchant to expose resist layer d Develop resist in 1 4 Methyl Isobutyl Ketone Isopropanol for 35 s and Isopropanol for 40 s e Deposit 4 nm titanium and 20 nm gold f Liftoff resist by various methods described below 100 XXV1 Figure 5 6 SEM s of best attempt at E beam bowtie fabrication on an AFM tip a SEM of an AFM tip after development and metal deposition An entire array of bowties were written on the cantilever not just on the tip apex so the white spots are bowtie shaped holes in the resist The red lines indicate the position of the bowtie that was targeted for the tip b SEM of the same tip after titanium gold deposition and liftoff The gold has peeled off of most of the cantilever and is now draped on top of the tip itself c SEM of one of the bowties written on the flat part of the cantilever next to the tip This bowtie is misshapen due to writing approximately 3 um out of focus 102 Figure 5 7 SEM of a bowtie on an AFM tip fabricated by Arvind Sundaramurthy using E beam lithography Scale bar 1 um 103 Figure 5 8 Schematic of FIB Process Flow a A 4
22. cavity By successfully 122 positioning molecules lithographically on top of the cavity region molecules are attractive candidate for future quantum optics experiments on single nanoscale emitters with well defined structure 7 2 Future Outlook The bowtie nanoantenna has been shown to be a very useful structure for enhancing optical emission whether it arises from fluorescence or Raman scattering This thesis has shown that bowtie nanoantennas can be used to perform single molecule experiments in crowded situations whether in solid or solution environments Many biological systems such as DNA replication and RNA translation require high fluorophore tagged substrate concentrations in order for the enzymes to perform well and to replicate biologically relevant behavior At the present time research groups and start up companies such as Pacific Biosciences use zero mode waveguides ZMG to sequence DNA but ZMG s have only been shown to enhance the fluorescence of single molecules by factors up to 25 Bowtie nanoantennas could potentially replace zero mode waveguides since they show much higher fluorescence enhancements When a fluorescently labeled substrate molecule binds to biological machinery e g an enzyme located within the gap of the bowtie nanoantenna then the fluorophore is significantly enhanced compared to the other very weakly emitting molecules not located in the gap region but still within a diffraction limit
23. confocal microscope Appropriate excitation and emission filters ensured that only TPQDI fluorescence reached the avalanche photodiode APD photon counting Si detector Figure 3 2a shows a confocal fluorescence scan from a dilute TPQDI concentration lt 1 molecule diffraction limited spot in PMMA without bowtie nanoantennas Essentially all fluorescent molecules irreversibly photobleach after a certain number of excitation cycles due to photodegradation e g photo oxidation so each spot in the image was observed until single step digital photobleaching occurred Figure 3 2b to ensure it corresponded to a single unenhanced TPQDI molecule Each molecule s dipole moment is randomly oriented with respect to the linear excitation field polarization causing each spot to emit with a different brightness The brightest molecules have their dipole moments aligned along the excitation polarization Each SM spot was fit to a 2D Gaussian to find the molecule s spatially integrated detected photons above background and only molecules that had intensities gt 60 counts 10 ms when exited with 79 kW cm were 53 considered 201 molecules Figure 3 2c is a histogram of the measured fluorescence from these unenhanced single molecules To estimate the photon emission rate expected from an unenhanced molecule with its dipole moment aligned along the polarization axis the emission rates of the brightest 5 were averaged together yielding 190 expected fl
24. crystal cavities and it was a pleasure to work with her to attempt to couple fluorescence molecules to the cavities she fabricated Above are mentioned my primary collaborators but I learned just as much from the other lab members of which there have been many I would like to thank Jaesuk Hwang Kallie Willets Stephanie Nishimura Kit Werley Hanshin Hwang Adam Cohen Marcelle Koenig Andrea Kurtz So Yeon Kim Jian Cui Nicole Tselentis Magnus Hsu Nick Conley Julie Biteen Sam Lord Randy Goldsmith Alex Fuerstenberg Majid Badieirostami Steven Lee Jianwei Liu Hsiao lu Lee Whitney Duim Lana Lau Yan Jiang Mike Thompson Sam Bockenhauer Quan Wang Marissa Lee Matt Lew and Yao Yue for making my time in the Moerner lab full of ideas and enjoyable vii I have so far only listed the people who have contributed to the science in this thesis but there are a great many more who have supported me outside of work I came to Stanford with very few connections and have since found a home in the Stanford community primarily due to the warmth and love from the friends I have made here of which there are too many to name here I thank everyone who has helped make my time here educational as well as fun In closing I d like to thank my family Amir Linda Yusuf Ali Amina and Yunus Kinkhabwala for their steady support throughout my entire life They encouraged me to study science and math at an early age which has stuck with me
25. ethanol solution but without a bowtie nanoantenna at 1 3 kW cm laser intensity b FCS curves from a are normalized to their value at t 100 ns and clearly show that the photobleaching time decreases as the laser intensity increases Fits to each curve using equation 4 1 are plotted with dashed black lines The FCS curve for a 10pM solution of IR800cw in the absence of a bowtie nanoantenna with 1 9MW cm laser intensity is plotted in solid black c e Fit parameters used for fit curves shown in b using equation 4 1 88 XXV Figure 5 1 Initial flattening of an AFM tip using FIB a Schematic of AFM tip before FIB processing A thin 4 nm layer of chrome is deposited uniformly on the tip to prevent charging during FIB milling and SEM imaging b After FIB milling the tip is flattened except for a short 30 nm post which will be used to protect the eventually fabricated bowtie nanoantenna during AFM imaging c SEM of Si3N4 AFM tip before FIB milling Scale bar 1 um d SEM of same Si3N4 AFM tip after FIB milling Scale bar 1 um 94 Figure 5 2 SEM of calibration marks milled into an AFM cantilever Scale bar 5 um 96 Figure 5 3 Float coating of resist onto an AFM tip a Tip is placed in a water bath b 1 drop of a 1 PMMA in toluene solution is dropped onto the water s surface A thin layer of PMMA forms as the toluene evaporates c Water is pipetted out letting the resist gently rest upon the AFM tip The tip is baked
26. expose resist and in this way define a pattern for etching or deposition FIB lithography differs in that the ions themselves ablate the material so that it can be sculpted in real time which eliminates the need for resist Appendix B is a detailed instruction manual for operating the FEI Strata in the Stanford Nanocharacterization Laboratory SNL The following discussion focuses only on strategies necessary for milling very small features There are several sample requirements to consider before choosing FIB lithography for a project First FIB lithography as in EBL requires a conductive sample This can be achieved for insulating samples by coating the sample with a thin conductive layer such as a 4nm thick layer of Chrome After the FIB milling is complete the chrome can be selectively removed by soaking the sample in Chrome Etchant CR14 for a few seconds Secondly the sample should not be magnetic otherwise the high resolution SEM mode UHR mode of imaging will pull the sample off of the sample holder and potentially scratch the E beam lens If the sample fulfills the above requirements then the effect of Gallium implantation on the sample s performance must be carefully weighed FIB 43 lithography often uses focused Ga ions in order to ablate material but this process also implants significant amounts of Ga into the sample surface There may even be some implantation into the unmilled edges of the nanostructure due
27. higher for the parallel orientation as expected Since all of the photons are time tagged the fluorescence lifetime can be determined for each polarization In Figure 3 9a the red blue curves is the SM time delay histogram for long short axis excitation polarization respectively yielding f s of 854 68 while both curves fit to lifetimes shorter than the IRF As expected the lifetime for the single molecule measured does not depend on the excitation polarization while the total 68 fluorescence intensity drops much more sharply for the perpendicular short axis polarization Figure 3 10b shows the same measurement for 20 more single molecules Red symbols show long axis excitation polarization while blue indicates short axis excitation polarization The fluorescence brightness enhancement was measured to change by up to a factor of 16 with different excitation polarization directions but the lifetime did not change significantly verifying our general interpretation even though the various molecules in the group had varying orientations relative to the bowtie long axis a aa 200 Wl la e re a i f 3 100 ey op mt yai diina ae i n es He 018 22 24 26 ot 20 40 60 rime s ioii Figure 3 10 Polarization dependence of single molecule enhanced fluorescence a Time trace for a single molecule with changing excitation polarization The polarization is changed from parallel red to perpendicular blue orientations with
28. molecule coupled to a gold bowtie nanoantenna by Zongfu Yu used a custom three dimensional finite difference time domain FDTD method developed in the Fan laboratory to solve Maxwell s equations The frequency dependent refractive index of gold and titanium were modeled by a fit to tabulated experimental data using the method of complex conjugate pole residue 28 pairs 60 To simulate the excitation process plane waves polarized in the x direction are incident from SiO side The optical intensity enhancement factor driving increased absorption rate ff is then obtained by comparing the electric field intensity with and without the metallic bowtie Figure 3 6a shows the spectrum of fg at the center of the antenna gap region 10 nm above ITO layer At a wavelength of 780 nm the enhancement is f 181in the center of the bowtie gap while the maximum field enhancement occurs closer to the two gold tips Figure 3 1d For simulation of the emission process we place a point current source in the gap region In the presence of the bowtie antenna the radiated power P into the far field and the power dissipated in the metal P are calculated The enhancement factors are then obtained by normalization with respect to the radiated power P of the same point current source in the absence of the antenna As a result for a point current source polarized in the x direction at the center of the gap emitting at 820 nm the radiati
29. molecule deposition for structures with lattice constant a and hole radius r a tuned so that the fundamental cavity resonance shifts across the fluorescence spectrum of the molecule Blue open circles indicate reflectivity measurements for the cavities that were also measured in fluorescence blue closed circles After Ref 115 Figure 6 3 Aligning molecules to a photonic crystal cavity a Scanning confocal image of fluorescence from DNQDI doped PMMA float coated onto a photonic crystal membrane Pixel size is 200nm and scale bar indicates 2 um b Scanning confocal image of DNQDI fluorescence after E beam lithography is used to remove all molecules except for the ones coating XX X Figure C 1 Figure C 2 Figure C 3 Figure C 4 the cavity region at the center The same imaging laser power as in a was used Pixel size is 80nm and scale bar indicates 2 um c Fluorescence spectrum from the fundamental mode of photonic crystal cavity after selective removal of molecules by E beam lithography d Atomic force microscopy image showing localization of DNQDI doped PMMA to the cavity region PMMA thickness is 12nm Scale bar indicates 500nm After Ref 119 General microscope view Note several components including input optics output optics AFM head the microscope and the CCD Spectrometer 139 Input optics for confocal microscopy including the single mode fiber SMP which is a spatial filter the collimati
30. nm titanium sticking layer Figure 5 9 followed by a 20 nm gold layer are deposited by Tom Carver in the Ginzton cleanroom b The FIB is used to mill away gold in the pattern of a bowtie nanoantenna 104 SEM of a FIB BOAT fabricated on Raith s ionLiNE FIB tool Scale bar is 200 nm 105 Figure 5 10 Scattering study of FIB milled bowties a SEM of FIB bowtie nanoantenna on a flat quartz substrate with 20nm gap b Comparison between scattering spectra for E beam and FIB fabricated bowties on quartz substrates with similar gap sizes 106 xxvii Figure 5 11 Schematic of setup used to test for enhancement of bulk TPQDI fluorescence using a FIB bowtie on an AFM tip Blue circles are bulk high concentration TPQDI molecules 107 Figure 5 12 Fluorescence enhancement attempt with FIB bowtie on an AFM tip and sharpened gold AFM tip a schematic of the FIB bowtie AFM tip b Schematic of sharp gold coated AFM tip c FIB bowtie AFM tip was scanned over a bulk TPQDI in PMMA sample The sample remained fixed while the tip was scanned thus imaging the enhancement of fluorescence as a function of tip position When the bowtie is positioned over the objective focus the fluorescence is quenched Scale bar lum d When a sharpened gold coated AFM tip is scanned over the sample an enhancement of fluorescence is measured Scale bar lum 108 Figure 6 1 a SEM image of a fabricated photonic crystal cavity in GaP Scale bar indicates 200nm
31. radio wave antennas to optical frequencies proved to be a challenge Optical frequency antennas are interesting because light cannot be focused to an infinitesimally small point with normal lenses instead it is limited by diffraction to d 1 1 where A is the wavelength of light and NA is the numerical aperture of the optics used For visible wavelengths the diffraction limit is 200 300 nm much larger than many objects of interest such as single molecules which are typically just a few nm in size Figure 1 1 An antenna can help concentrate light to a smaller area and decrease the mismatch in size between light and the nanoscale emitter Diffraction Limit A 2NA Figure 1 1 Size mismatch between the diffraction limit and a nanoscale emitter 1 2 2 The Drude Model At radio frequencies the free electrons within an antenna oscillate with the AC electromagnetic field of the passing wave At high frequencies such as optical frequencies this simple model of free electrons moving instantaneously in response to an electromagnetic field fails The Drude model is a slightly more complicated model of electromagnetic fields in bulk metals In this model the metal is composed of fixed positively charged nuclei surrounded by free unbound conduction electrons This model predicts that when a metal is excited by an electromagnetic field then the field in the metal must satisfy the wave equation 2 4 E e w E 1 2 where
32. spot and raster scanned over a surface or the sample is moved relative to the fixed laser spot Fluorescence emission from the sample is focused through a pinhole located at one of the microscope s conjugate plane s which prevents emission from above and below the focal plane of the laser from passing through and reaching the detector The fluorescence is measured from each spot and an image is built from this information pixel by pixel The main advantage to confocal microscopy is Z sectioning which allows for fluorescence to be 25 detected only from a thin z section of sample thus allowing sensitive imaging of thick biological samples 2 2 2 Optical Setup Detailed information on setting up a confocal imaging pathway can be found in Appendix C thus only the main ideas behind confocal microscopy will be mentioned here Figure 2 1 shows a typical setup for the excitation and emission pathways in a confocal imaging setup for single molecule imaging For excitation Figure 2 1a a collimated laser beam is focused to a diffraction limited spot on the sample using a high numerical aperture NA objective Diffraction limits the size of the focus to A 2NA in the X and Y directions but note that the laser can excite fluorescence in a large range of Z positions A Excitation B Emission Sample Sample Objective Objective Dichroic Dichroic Mirror Mirror Pinhole A 26 Figure 2 1 a Schematic of typical excitation pathwa
33. taken from the cavity Figure 6 3a lt 10 of the emission is from molecules coupled to the cavity region but spectra of these coupled molecules are still visible over background from uncoupled molecules since molecules coupled to the cavity emit only at a few wavelengths The quality factor of the fundamental mode is measured to be 10 000 indicating that deposition of molecules onto the membrane does not degrade the properties of the cavity in agreement with finite difference time domain simulations for a thin lt 40 nm thick layer of PMMA It is worth noting that even though the bulk emission spectrum of the molecule does not show much emission at the longest wavelengths nevertheless there are molecules emitting there an effect which can be observed when the molecular emission is coupled to the cavity 115 After deposition of molecules we observe a small several nm red shift in the cavity resonance as expected from simulations With no DNQDI PMMA present only background counts are detectable over the entire spectral range We vary the spatial periodicity of the photonic crystal holes and hole radius to tune the fundamental cavity resonance through the fluorescence spectrum of the molecule We measure high cavity Q factors up to 12 000 via fluorescence Figure 6 2d across a range of more than 100 nm from 735 nm 860 nm The cavity Q is higher at longer wavelengths where we fabricate most of our cavities as fabrication imperfections a
34. the center mouse button and focus on that 126 e Move to magnification high enough to see 100nm features e Burn anew calibration dot e Turn on focus wobble e Optional use reduced raster scan area often helpful in stigmation e Click aperture align e Ifthe aperture is misaligned the dot will move as the focus is wobbled in and out of focus Move the x and y scrollbars until image stays centered in focus wobble it will still streak though due to astigmation but the center of mass of the spot will stay in the same position e Turn off focus wobble Setup the electron beam Astigmation e Click Stigmation e Burn anew calibration dot e Move the x and y scrollbars until the spot appears small and sharp You will need to alternate between x stigmation y stigmation focus and burning new calibration dots in order to get a lt 20nm calibration dot Align Writefield e Burn a bright spot in a recognizable place away from any other calibration dots e Open new positionlist e Go to Scan Manager Box gt Align write field procedures gt manual gt 1OOWF Manual Sum mark and drag to new positionlist e Add 10 15 to contrast on right computer e Start scan e CTRL left click will drag the mark onto the burned spot Do this 4 times e Accept new align writefield values Level the sample using three points sample leveling e Pick 3 points close to the write area type coordinates for 3 points into U V boxes 127
35. the number of counts per bin time Depending on what other processes are running on the computer this program may run slowly To check whether the program is running slowly watch the Available samples output number If this number is zero then the computer is reading the 155 output from the counters as quickly as they are being generated but if this is a non zero value then the computer is not updating fast enough Avoid using this computer for other tasks while collecting data and this program will work well The most important input to this program is the Frequency Hz input The user can choose any bin time desired using this input In Figure C 7 it is currently set for 100Hz which corresponds to 10ms bin times and is the fastest this computer can handle reliably The program then takes the number of photons for a specific time interval and divides by 100 This number is then converted into an analog voltage that is sent to the AFM controller in order to build up an image Note that when the image is formed by the AFM software it will plot the voltages it receives which are a factor of 100 smaller than the actual number of photons collected by the APD Since 10 V is the greatest voltage that can be sent the user should ensure that the number of photons does not exceed a value that will produce a gt 10V signal Any value 10V will just be plotted at 10 V When using an APD and 10ms bin times this limits the photon rate to 1
36. to this day Finally I d like to thank my partner for the last 5 5 years David Press who has always been there to provide love and support whenever I needed it viii Contents Abstract Chapter 1 Introduction 1 1 Overview 1 2 Optical Plasmonic Nanoantennas 1 2 1 Motivation 1 2 2 The Drude Model 1 2 3 Surface Plasmon Polaritons 1 2 4 Localized Surface Plasmon Resonance 1 2 5 Gold Bowtie Nanoantenna Plasmon Resonance 1 2 6 Measurement of Enhanced Fields of Gold Bowtie Nanoantenna 1 3 Photonic Crystals 1 3 1 Motivation 1 3 2 Planar Photonic Crystal Cavities 1 4 Fluorescence 1 4 1 Motivation 1 4 2 Fundamentals 1 4 3 Single Molecule Fluorescence 1 5 Fluorescence Correlation Spectroscopy 1 5 1 Motivation 1 5 2 Fundamentals 1 5 3 Zero Mode Waveguides for High Concentration FCS 1X iv 11 12 12 13 15 15 15 17 18 18 20 21 1 5 4 Conclusions 22 Chapter 2 Experimental Methods 25 2 1 Introduction 26 2 2 Confocal Microscopy 27 2 2 1 Introduction 27 2 2 2 Optical Setup 27 2 2 3 Technical Issues for Single Molecule Imaging 29 2 2 4 Time Correlated Single Photon Counting 32 2 3 Scattering Microscopy 33 2 3 1 Introduction 33 2 3 2 Optical Setup 34 2 4 Nanofabrication Techniques 36 2 4 1 Introduction 36 2 4 2 Electron Beam Lithography 36 2 4 3 Float Coating EBL Resist 41 2 4 4 Focused Ion Beam Lithography 44 2 5 Apertureless Near Field Optical Microscopy 46 2 5 1 Introduction 46 2 5 2 At
37. water but not in ethanol Figure 4 3 further supporting the conclusion that the enhanced signal is from a concentrated layer of molecules sticking to the ITO b 900 soo 600 300 200 e f p 150 250 100 50 200 Figure 4 2 Confocal images of an array of bowties in the presence of a 100nM IR800cw in ethanol CK 1 10 100 IR800cw concentration in EtOH uM 109W cm imaging intensity b 100uM IR800cw in ethanol 3W cm imaging intensity c 30nm thick PVA film doped with IR800cw 36W cm imaging intensity d 1M ICG in water 1 2kW cm 77 imaging intensity e 14M ICG in ethanol 600W cm imaging intensity f 30nm thick PVA film doped with ICG 1 2kW cm imaging intensity g Signal to background ratio of bulk enhanced fluorescence from 25 bowtie nanoantennas and different IR800cw concentrations _ 9 Normalized counts 10ms 05 10 20 30 40 50 60 Time s Figure 4 3 Photobleaching curves from cleaned ITO interfaces immersed in different dye solutions without bowties Blue 1uM ICG in ethanol Red luM ICG in water Black 1 uM IR800cw in ethanol Green 1uM IR800cw in water If photobleaching drop in signal is measured beyond the first 10ms bin then molecules must be sticking to the surface and cannot be replaced since molecules only remain in the focal volume for no more than I ms unless they are stuck to the surface Therefore the only solution that did not show sticking i
38. 000 Quality factors up to 2 5 x 10 are achievable in silicon but these cavities have resonances further in the infrared 11 In Chapter 6 of this thesis fluorescent molecules are coupled to a photonic crystal cavity using a lithographic approach and optical experiments demonstrate successful coupling between the molecules and photonic crystal cavity modes 1 4 Fluorescence 1 4 1 Motivation Bowtie nanoantennas and photonic crystal cavities are interesting structures because they both alter local electromagnetic fields This thesis measures the effect these structures have on molecular fluorescence an effect that depends heavily on local electromagnetic fields Fluorescence is a photophysical process whereby light of one wavelength excites a molecule which in turn emits light of a lower energy or longer wavelength Fluorescent dyes have been commonly used as markers in biology a molecule protein area of interest in a cell is covalently linked to a fluorescent molecule and the fluorescent molecule s intensity can report upon the location and environment of the target 1 4 2 Fundamentals The simplest model to describe emission of light by a molecule is the two level system with each of the two primary electronic states are decorated with a progression of vibrational modes Fluorescence consists of the absorption of a photon by a molecule followed by the emission of a lower energy photon red shifted wavelength as see
39. 000 photons 10ms or 100 000 photons s which is a good limit to keep in mind anyway since significantly more photons than this will damage the APD There are two ways to stop the program By pushing the red stop sign located near the white arrow at the top of the page the program stops and no data is recorded this is useful for aligning taking confocal scans or anytime that the time trace does not need to be kept Alternately there is a red STOP button located towards the lower left corner of the screen If this button is pressed the program stops and asks where to save the acquired time trace 156 C 4 3 Topometrix Software for Confocal Scanning The previous section described collecting data from the APD but now the user still needs to create a confocal scan which is achieved using the Topometrix AFM head program This program is opened from the desktop of WEM12 and is called SPMLab602 When this program is opened click the AFM tip button that is set off from the other buttons in the top right corner of the screen This will open a dialog box that allows you to choose a stage For confocal scanning the stage not the AFM tip should be scanned so select scanner LX149707 and press ok The computer will ask if you want to energize high voltage for the 50um Tripod scanner so press Engage Now the stage is active remember the stage is VERY delicate Do NOT press down on the stage with any force or you wil
40. 407 491 493 2000 30 Fromm D P Sundaramurthy A Schuck P J Kino G S amp Moerner W E Gap dependent optical coupling of single bowtie nanoantennas resonant in the visible Nano Lett 4 957 961 2004 31 A Kinkhabwala Z Yu S Fan Y Avlasevich K Miillen and W E Moerner Large Single Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna Nature Photonics 3 654 657 2009 published online October 18 2009 12 Chapter 4 Fluorescence Correlation Spectroscopy at High Concentrations using Gold Bowtie Nanoantennas 4 1 Introduction As shown in Chapter 3 gold bowtie nanoantennas are able to enhance a single molecule s fluorescence by factors up to 1300 So far this property has only been demonstrated for molecules fixed in position and orientation in a thin polymer film The goal of this chapter is to extend the polymer work to molecules in solution by immersing bowtie nanoantennas in a liquid containing a high concentration of fluorescent dye molecules and studying dynamics in the fluorescence signal from these molecules Figure 4 la Importantly I observed that fluorescence enhancement from single molecules transiently sticking to the ITO surface near the bowtie nanoantenna could be measured over the background of nearby molecules in solution within the laser pumping volume This experiment shows that the bowtie produces extreme contrast enhancements and thus opens the way to f
41. 47 substrate a 50 nm layer of ITO n 2 and a 30 nm layer of XX PMMA n 1 49 The gold bowtie structure is 20 nm thick on a 4 nm layer of titanium b Radiative red and non radiative green enhancement factors along the center of the gap for wavelength 820 nm z measures the distance above the ITO PMMA interface Black dashed line shows the enhancement factor for electric field intensity at 780 nm Blue curve shows the fluorescence enhancement factor for quantum efficiency 2 5 molecules and grey dash line for quantum efficiency 100 molecules c e Illustration of the simulated structure side view section through the two triangle tips showing regions of fluorescence Blue radiative Red and non radiative Green enhancement factors for a molecule emitting at 820 nm wavelength After Ref o 64 Figure 3 7 Modeled enhancement of QE as a function of intrinsic QE a Theoretical predictions based on FDTD simulations for the change in intrinsic quantum efficiency 7 when a molecule is placed near a bowtie nanoantenna 7 The FDTD simulations provide f and fa and the curves show the values of Eqn 3 5 b Same data as in a this time plotting enhancement of quantum efficiency against the intrinsic quantum efficiency In both figures TPQDI s intrinsic quantum efficiency 7 2 5 is circled in red After Ref 66 Figure 3 8 Measuring excited state lifetime from a single molecule coupled to bowtie nanoantenna a Tim
42. 72 tp where N is the average number of molecules in the observation volume a is a geometrical factor dependent upon the shape of the excitation volume z is the time lag and Tp is the diffusion time The factor a is typically calibrated for a specific microscope using a dye with known properties so that the only free parameters are N and tp Notice that if the concentration of the solution N is known then tp can be measured or conversely if tp is known then N can be measured Clearly the contrast is highest for small values of N 1 5 3 Zero Mode Waveguides for High Concentration FCS FCS is usually performed at low concentrations 0 01 1 nM because in order to measure the largest fluctuations in the fluorescence there needs to be on average less than one molecule in the illumination volume at any one time However many biological studies such as DNA replication must occur at high concentrations One solution to this problem is to use a zero mode waveguide to confine the illumination volume much further than is possible with normal diffraction limited confocal microscopy Zero mode waveguides are simply holes in metal films that are smaller than the diffraction limit Electromagnetic waves cannot propagate through subwavelength holes so there is only a weak penetration of evanescent waves into these apertures restricting illumination to a few 10 s of nm from the substrate A typical geometry consists of a thin 100 nm
43. 88 4 Mohammadi A Sandoghdar V amp Agio M Gold Copper Silver and Aluminum Nanoantennas to Enhance Spontaneous Emission J Comput Theor Nanosci 6 2024 2009 5 Mie G Contributions to the Optics of Turbid Media Particularly of Colloidal Metal Solutions Annalen Der Physik 25 377 1908 6 Novotny L Effective Wavelength Scaling for Optical Antennas Phys Rev Lett 98 266802 2007 7 Bryant G Garcia de Abajo F J amp Aizpurua J Mapping the Plasmon Resonances of Metallic Nanoantennas Nano Lett 8 631 2008 8 Ditlbacher H et al Silver Nanowires as Surface Plasmon Resonators Phys Rev Lett 95 257403 2005 9 Taminiau T H Stefani F D Segerink F B amp van Hulst N Optical Antennas Direct Single Molecule Emission Nat Phot 2 234 237 2008 10 Sondergaard T amp Bozhevolnyi S I Slow Plasmon Resonant Nanostructures Scattering and Field Enhancements Phys Rev B 75 073402 2007 11 Sondergaard T amp Bozhevolnyi S I Metal Nano strip Optical Resonators Opt Express 15 4198 2007 12 Stockman M I Nanofocusing of Optical Energy in Tapered Plasmonic Waveguides Phys Rev Lett 93 137404 2004 13 Verhagen E Spasenovic M Polman A amp Kuipers L Nanowire Plasmon Excitation by Adiabatic Mode Transformation Phys Rev Lett 192 203904 2009 21 14 Fromm D P et al Exploring the chemical enhancement for surface enhanced Raman scatt
44. COUPLING FLUORESCENT MOLECULES TO NANOPHOTONIC STRUCTURES A DISSERTATION SUBMITTED TO THE DEPARTMENT OF APPLIED PHYSICS AND THE COMMITTEE ON GRADUATE STUDIES OF STANFORD UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Anika Amir Kinkhabwala June 2010 2010 by Anika Amir Kinkhabwala All Rights Reserved Re distributed by Stanford University under license with the author This work is licensed under a Creative Commons Attribution Noncommercial 3 0 United States License http creativecommons org licenses by nc 3 0 us This dissertation is online at http purl stanford edu mf049qp 1902 I certify that I have read this dissertation and that in my opinion it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy William Moerner Primary Adviser I certify that I have read this dissertation and that in my opinion it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy Mark Brongersma I certify that I have read this dissertation and that in my opinion it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy Gordon Kino Approved for the Stanford University Committee on Graduate Studies Patricia J Gumport Vice Provost Graduate Education This signature page was generated electronically upon submission of this dissertation in electroni
45. D has better quantum efficiency QE in the infrared but has poor timing resolution instrument response function width 500ps than compared to the MPD PDM series APD instrument response function width 50ps Currently a PE APD is installed on 8A back and when properly shielded it has approximately 200 300 dark counts sec C 2 5 Spectrometer Path This path is activated by flipping up a remote mirror Currently there is a mirror mounted in this flipper but a 90 10BS can be inserted which sends 10 of the light to the APD fast time resolution information and 90 of the light to the spectrometer slow spectral information The next section is devoted to spectrometer alignment C 3 Alignment of CCD Monochromator C 3 1 Introduction This section describes the proper alignment and optics used for the CCD Spectrometer attachment shown in Figure C 6 The monochromator is a Jarrell Ash Monospec 18 This device is chosen because it is easy to use and has high throughput 50 is possible with certain gratings This spectrometer sacrifices absolute resolution for transmission There is also an Acton spectrometer available in the lab which is not currently in use This newer spectrometer is very similar to the Monospec 18 but since it is made by Acton it is easier to use and computer controlled Also since Acton actively supports this spectrometer new gratings can easily be purchased 148 C 3 2 Input mirror The first optic i
46. Eisler H Martin O J F Hecht B amp Pohl D W Resonant optical antennas Science 308 1607 1609 2005 3 Fischer H amp Martin O J P Engineering the Optical Response of Plasmonic Nanoantennas Opt Exp 16 9144 9154 2008 4 Grober R D Schoelkopf R J amp Prober D Optical Antenna Towards a Unity Efficiency Near Field Optical Probe Appl Phys Lett 70 1354 1356 1997 5 Willets K A amp Van Duyne R P Localized Surface Plasmon Resonance Spectroscopy and Sensing Annu Rev Phys Chem 58 267 297 2007 6 Hamann H F Kuno M Gallagher A amp Nesbitt D J Molecular fluorescence in the vicinity of a nanoscopic probe J Chem Phys 114 8596 8609 2001 7 Anger P Bharadwaj P amp Novotny L Enhancement and Quenching of Single Molecule Fluorescence Phys Rev Lett 96 113002 2006 70 8 Kuhn S Hakanson U Rogobete L amp Sandoghdar V Enhancement of Single Molecule Fluorescence Using a Gold Nanoparticle as an Optical Nanoantenna Phys Rev Lett 97 017402 2006 9 Schuck P J Fromm D P Sundaramurthy A Kino G S amp Moerner W E Improving the mismatch between light and nanoscale objects with gold bowtie nanoantennas Phys Rev Lett 94 017402 2005 10 Farahani J N Pohl D W Eisler H amp Hecht B Single Quantum Dot Coupled to a Scanning Optical Antenna A Tunable Superemitter Phys Rev Lett 95 017402 2005 11 Farahani J N e
47. M IR800cw in ethanol using 430W cm laser intensity b Fluorescence time trace binned to Ims for a bowtie immersed in 1uM ICG in water using 144kW cm laser intensity Notice that ICG in water has higher contrast between enhanced molecules compared to background than IR800cw in ethanol Figure 4 8b plots the normalized FCS curves collected on a single bowtie immersed in 1 uM ICG in water All FCS curves were taken from 5 minutes of time tagged fluorescence data Notice that at lower excitation intensities the timescale for the bowtie FCS curve decay is much longer than the FCS curve in the absence of the bowtie nanoantenna black solid curve This difference in time scale is consistent with the picture that molecules transiently stick to the surface near the bowtie nanoantenna and then photobleach The simplest model for photobleaching is that a molecule has a fixed probability of photobleaching during any excitation cycle and this does not change with excitation power This means that a molecule has a total number of photons that it tends to emit before photobleaching that remains constant for different excitation powers Therefore as the excitation intensity is increased the molecule will emit the same number of photons but in progressively shorter periods of 83 time causing the photobleaching time Tproio to shorten This behavior is seen in the bowtie FCS curves in Figure 4 8b Therefore the long time decay in the bowtie FCS curve
48. MMA in anisole solution was found to ball up and not disperse evenly over the surface but a 1 PMMA in toluene solution does spread evenly The molecular weight of the PMMA was not found to significantly affect results The sample should be left for 5 minutes to let the film form fully and for the toluene to evaporate Next the water is pipetted out of the dish to allow the film to fall gently onto the substrate step 4 The film is usually visible on the surface due to large wrinkles in the surface In fact the film is uneven over the entire surface on large spatial scales but I ve found that the film is locally very uniform over 100um distances For this reason the thickness of the final film cannot be controlled well and 10nm 100nm thick films will be deposited with the same solution If the process is performed several times or on different samples then eventually an acceptable thickness will be deposited Finally in step 5 the substrate is heated at 90 C for 30 minutes in order to bake out the remaining water that could not be pipetted away Note that this baking step is longer and at a lower temperature than the baking step for spin coating 180 C for 2 minutes The temperature needs to be lower because the melting point for PMMA is 140 150 C If 41 the PMMA coated uneven substrate is heated above the melting point then the PMMA may run down the sides of the uneven features Baking is still important to remove excess water partic
49. Nanoparticle Sandwiches Nano Lett 8 485 490 2008 21 Song J H Atay T Shi S F Urabe H amp Nurmikko A V Large enhancement of fluorescence efficiency from CdSe ZnS quantum dots induced by resonant coupling to spatially controlled surface plasmons Nano Lett 5 1557 1561 2005 22 Chen Y Munechika K amp Ginger D Dependence of Fluorescence Intensity on the Spectral Overlap between Fluorophores and Plasmon Resonant Single Silver Nanoparticles Nano Lett 7 690 696 2007 23 Ruppin R Decay of an excited molecule near a small metal sphere J Chem Phys 76 1681 1683 1982 24 Rogobete L Kaminski F Agio M amp Sandoghdar V Design of plasmonic nanoantennae for enhancing spontaneous emission Opt Lett 32 1623 1625 2007 25 Khurgin J B Sun G amp Soref R A Practical limits of absorption enhancement near metal nanoparticles Appl Phys Lett 94 071103 2009 26 Sun G Khurgin J B amp Soref R A Practical enhancement of photoluminescence by metal nanoparticles Appl Phys Lett 94 101103 2009 27 Palik E D in Handbook of Optical Constants Academic Press Orlando 1985 28 Han M Dutton R W amp Fan S Model dispersive media in finite difference time domain method with complex conjugate pole residue pairs IEEE Microw Wirel Comp Lett 16 119 121 2006 29 Lounis B amp Moerner W E Single Photons on Demand from a Single Molecule at Room Temperature Nature
50. P a I V semiconductor with n 3 5 and indirect band gap at 550 nm The high index of the material enables a large photonic band gap and cavities with high quality factor while the large electronic band gap prevents absorption in the near IR and part of the visible 111 c d ii O 100 i 400 gt 200 a fa Acal 850 700 750 800 850 Sean Wavelength nm Figure 6 1 a SEM image of a fabricated photonic crystal cavity in GaP Scale bar indicates 200nm b FDTD simulation of electric field intensity of the fundamental cavity mode The mode is primarily y polarized c Schematic illustrating fabrication procedure i DNQDI PMMaA is float coated over the entire structure ii DNQDI PMMA is lithographically defined over cavity region d Bulk fluorescence emission spectrum of DNQDI when excited with a 633 nm HeNe laser measured with a confocal microscope and spectrometer The molecule has a peak in its absorption at this excitation wavelength e Chemical structure of DNQDI molecule After Ref Our cavity is a linear three hole defect L3 fabricated in a 125 nm gallium phosphide membrane grown by gas source molecular beam epitaxy A scanning electron microscope SEM image of a fabricated cavity and the simulated electric field intensity of the fundamental high Q cavity mode are shown in Figure 6 1a b Cavities are fabricated as described in Ref The molecule we use is 112 dinaphthoquaterryle
51. a background of 1 000 unenhanced molecules By extending this experiment to iv molecules in solution dynamics of single molecules in concentrated solutions can also be measured While bowtie nanoantennas act to concentrate light light does not remain in the structure for long The photonic crystal cavity can be used to trap and store light which has interesting implications for molecular emitters located nearby This thesis will show that molecules can be lithographically positioned onto a photonic crystal cavity and that the molecule s fluorescence emission is coupled to the cavity modes Acknowledgements The research in this thesis was aided by a great many people First and foremost I must acknowledge my advisor Prof W E Moerner W E has not only made funding possible to support me throughout my graduate career but has more importantly been a steady guide in my research efforts My projects needed a long time to mature and he was always there to encourage me and provide helpful ideas at the most frustrating times In addition he has been an excellent role model as a scientist someone who always makes sure the science is correct and complete as possible before publishing it I d also like to thank the rest of my reading committee First Prof Gordon Kino who actually originally began the bowtie project in the infrared region of the spectrum and was a very helpful collaborator early in my graduate career Prof Mark Br
52. ale bar 4um b Fluorescence time trace of TPQDI PMMA coated bowtie nanoantenna shown in Fig 1c Blinking dynamics and eventual photobleaching are due to 1 molecule that has been enhanced by a factor of 1340 After Ref As is well known the local field enhancement is highly dependent on the bowtie gap size E beam lithography produces a distribution of bowtie gap sizes even when using the same lithographic pattern After all optical data were obtained scanning electron microscope SEM images were taken to measure the precise gap size of every bowtie Figure 3 4a shows histograms of the actual bowtie gap sizes for all bowties used in this study and shows that approximately equivalent numbers of 10 25 nm 40 60 nm and 65 90 nm gap bowties were used allowing for a gap dependence of the enhancement to be ascertained 56 0 20 40 60 80 Single Bowtie Gap Size nm Triangles 0 20 40 60 80 Single Bowtie Gap Size nm Triangles Figure 3 4 Measurement of fp for SMs as a function of bowtie gap size a Histogram of gap sizes of all bowties measured b Scatter plot of 129 SM fluorescence brightness enhancements fr as a function of bowtie gap size for all bowties measured in a After Ref Confocal scans were taken of each array of similarly sized gap bowties and the 5 brightest spots in any array were measured as a function of time to look for highly enhanced molecules as determined by significant single photobleac
53. an even lower maximum fluorescence enhancement Bowtie nanoantennas are fabricated on ITO coated glass substrates using E beam lithography as described in Chapter 2 In order to immerse the bowties in concentrated solutions of dye molecules a simple fluid cell is constructed from 2 coverslips one with the fabricated bowtie nanoantennas on the surface and the other unstructured and an o ring sandwiched in between the coverslips The coverslips and 74 o ring were first cleaned in water and then ozone cleaned for 10 minutes before adding the concentrated dye solutions t Eo d E t 700 750 800 850 900 Wavelength nm c d SO3Na ZX KN SS ann 5 jip LL P Oz x lt Phy pk et AY ae ar an YAY a l J wS A a i i l Ken Ong D cm O Nat o X NaQ S ons NaO C Figure 4 1 a Bowtie nanoantennas are immersed in concentrated dye solutions for FCS experiments b Blue absorption solid and emission dashed spectra of IR800cw in ethanol Red solid and emission dashed spectra of ICG in water Black plasmon resonance of a 10 nm gap Au bowtie i 5 nanoantenna Measured as in Ref 100 nm c ICG molecule d IR800cw molecule Inset SEM of a typical gold bowtie nanoantenna Scale bar Dye Solvent Quantum Efficiency ICG Water 2 4 ICG Ethanol 14 IR800cw Water 10 IR800cw Ethanol 28 Table 1 Table of quantum efficiencies QE for
54. are actually sums of a broad continuum of photobleaching times Further it is interesting to note that as the excitation intensity 84 increases f decreases which means that at higher excitation intensities there are more underlying exponentials than at lower excitation intensities a reasonable observation given the fact that more and more non optimally oriented and located molecules can contribute under these conditions D AN 10 10 10 10 10 10 10 10 10 10 10 10 z ms z ms c 8 d e 04 a Ta rene 6 a Xx 0 35 z E we 0 3 i 0 a a P 0 25 i Zz 2 a 4 0 2 I 0 1 2 3 F 0 1 2 3 0 15 0 1 2 Py 10 10 10 10 10 10 10 10 10 10 10 10 Laser Intensity kWiem Laser Intensity kWicm Laser Intensity kWiem Figure 4 8 a FCS curves for a bowtie immersed with 1uM ICG in water when illuminated with pump intensity 1 3 kW cm blue 4 6 kW cm red 14k W cm green 50kW cm pink 144kW cm cyan 362kW cm purple and 940kW cm yellow The grey curve indicates the FCS curve for the same 1uM ICG in water solution but without a bowtie nanoantenna at 110kW cm laser intensity b FCS curves from a are normalized to their value at t 100ns and clearly show that the photobleaching time Tphotos decreases as the laser intensity increases Fits to each curve using equation 4 3 are plotted with dashed black lines The FCS curve for a 10pM solution of ICG in the absence of a bowtie nanoante
55. as Our goal was to fabricate the gold bowtie nanoantenna onto AFM tips since they have proven useful in fluorescence enhancement Two approaches to fabricating gold bowtie nanoantennas onto AFM tips in the Stanford Nanofabrication Facility SNF and in the Raith facility in Dortmund Germany will be discussed In both approaches the AFM tip is first flattened in order to provide a flat surface for the bowtie nanoantenna to be fabricated In the first 90 approach developed by Arvind Sundaramurthy E beam lithography is used to define the bowtie shape which requires float coating of an E beam resist in addition to the normal E beam processing steps for a non conductive substrate In the second approach a gold metal film is deposited on an AFM tip and FIB Focused Ion Beam milling is used sculpt a bowtie nanoantenna out of the gold The E beam approach was found to be too technically challenging while the FIB approach proved to be relatively simple in fabrication but yielded bowties that did not enhance molecular fluorescence While a scanning bowtie would be very useful in ANSOM it has proven difficult to fabricate a highly resonant structure onto an AFM tip 5 2 Initial Preparation of AFM Tip The fabrication for both the E beam and FIB approach begins the same way In both cases a Siz3N4 AFM tip Veeco is flattened using a FIB in order to have a flat area upon which to fabricate the bowtie nanoantenna Si3N4 contact mode AFM
56. attices Phys Rev Lett 58 2486 1987 34 Rivoire K et al Lithographic Position of Fluorescent Molecules on High Q Photonic Crystal Cavities Appl Phys Lett 95 123113 2009 35 Takahashi Y Hagino H Tanaka Y Asano T amp Noda S High Q Photonic Nanocavity with a 2 ns Photon Lifetime Conf on asers and Electro Optics QFO1 2008 36 Moerner W E amp Kador L Optical detection and spectroscopy of single molecules in a solid Phys Rev Lett 62 2535 2538 1989 37 Orrit M amp Bernard J Single pentacene molecules detected by fluorescence excitation in a p terphenyl crystal Phys Rev Lett 65 2716 2719 1990 38 Betzig E amp Chichester R J Single Molecules Observed by Near Field Scanning Optical Microscopy Science 262 1422 1425 1993 39 Lord S J et al Photophysical Properties of Acene DCDHF Fluorophores Long Wavelength Single Molecule Emitters Designed for Cellular Imaging J Phys Chem A 111 8934 8941 2007 40 Magde D Elson E amp Webb W W Thermodynamic Flucutations in a Reacting System Measurement by Fluorescence Correlation Spectroscopy Phys Rev Lett 28 705 1972 23 41 Eigen M amp Rigler R Sorting Single Molecules Application to Diagnostics and Evolutionary Biotechnology Proc Natl Acad Sci U S A 91 5740 5747 1994 42 Rigler R Fluorescence correlations single molecule detection and large number screening applications in biote
57. brication of high quality Q factor photonic crystal cavities up to 12 000 with resonances in the near IR from 735 nm 860 nm For this experiment a near IR fluorescent molecule doped polymer film is float coated on top of high quality photonic crystal 110 nanocavities A lithographic polymer photoresist is used so that the molecules can be selectively positioned onto the location of the cavity by using a lithographic technique to remove unwanted molecules Coupled photonic crystal cavity emitter systems studied so far are primarily based on gallium arsenide and silicon materials which absorb strongly at wavelengths shorter than the electronic band gap of the material This precludes the use of emitters such as organic molecules which typically have resonances at visible and near IR wavelengths Research in photonic crystals operating at these shorter wavelengths has focused on materials such as GaN and SisN These materials have a lower refractive index than GaAs or Si n 2 4 for GaN and n 2 0 for Siz3N4 compared to n 3 5 for GaAs and Si which limits the size of photonic band gap and has generally led to low experimental quality factors of up to a few thousand although designs with higher quality factors Q up to 1 million have been proposed It has previously been demonstrated that photonic crystal cavities with quality factors up to 1 700 limited by fabrication inaccuracy could be fabricated in gallium phosphide Ga
58. c format An original signed hard copy of the signature page is on file in University Archives iii Abstract Fluorescence imaging and spectroscopy is an important tool in many areas of research Biology has particularly benefitted from fluorescence techniques since a single molecule s position local environment and even activity can be studied in real time by tagging it with a fluorescent label It is therefore important to be able to understand and manipulate fluorescence One way to control fluorescence is to shape the local electromagnetic fields that excite the fluorescent molecule This thesis studies the interaction between fluorescent molecules and two nanophotonic structures that highly modify local electromagnetic fields the bowtie nanoantenna and the photonic crystal cavity The study of plasmons or coherent excitations of free electrons in a metal has led to the fabrication of antennas at optical frequencies In particular gold bowtie nanoantennas have been shown to concentrate light from the diffraction limit at 800 nm 300 nm down to 20 nm while also enhancing the local electric field intensity by a factor of 1 000 This huge change in the local field greatly alters the absorption and fluorescence emission of nearby molecules This thesis will show that the fluorescence from an initially poor single molecule emitter can be enhanced by a factor of 1 300 allowing for the measurement of one highly enhanced molecule over
59. chnology J Biotechnol 41 177 186 1995 43 Hess S T Huang S Heikal A A amp Webb W W Biological and chemical applications of fluorescence correlation spectroscopy a review Biochemistry 41 697 705 2002 44 Rigler R Elson E amp Elson E in Springer Series in Chemical Physics Vol 65 Fluorescence Correlation Spectroscopy Springer Series Chem Phys eds Schaefer F P Toennies J P amp Zinth W SpringerRigler R Berlin 2001 45 Levene M J et al Zero Mode Waveguides for Single Molecule Analysis at High Concentrations Science 299 682 686 2003 46 Eid J amp et al Real Time DNA Sequencing from Single Polymerase Molecules Science 323 133 138 2009 47 Uemura S et al Real time tRNA Transit on Single Translating Ribosomes at Codon Resolution Nature 464 1012 1017 2010 24 Chapter 2 Experimental Methods 2 1 Introduction This chapter contains detailed information for the optical nanofabrication and nanocharacterization techniques needed to complete the experiments described in this thesis While each technique is described in detail references should be consulted as they give additional information on each technique 2 2 Confocal Microscopy 2 2 1 Introduction Confocal microscopy is a highly sensitive low background method for recording fluorescence images particularly in biology In confocal microscopy a laser is tightly focused to a diffraction limited
60. clusions 5 4 Focused Ion Beam Process Flow 5 4 1 Introduction 5 4 2 Chrome Etch and Gold Deposition 5 4 3 Focused Ion Beam Milling 5 4 4 Scattering measurements on flat substrate FIB bowties 5 4 5 Optical Results from FIB Bowties on AFM tips 5 5 Conclusions 98 99 100 101 101 102 102 103 104 107 Chapter 6 Lithographic Positioning of Fluorescent Molecules on High Q Photonic Crystal Cavities 6 1 Introduction 6 2 Sample Fabrication and Preparation 6 3 Optical Characterization of High Q Cavity Modes 6 4 Fluorophore Cavity Coupled Fluorescence Emission Spectra 109 110 111 114 116 6 5 Lithographically Defining Molecule Position over Photonic Crystal Cavity Error Bookmark not defined 6 6 Conclusions Chapter 7 Conclusions 7 1 Conclusions 7 2 Future Outlook Appendix A EBL using Raith 150 A l Writing Bowtie nanoantennas with Raith 150 Appendix B Focused Ion Beam Lithography with FEI Strata B 1 Startup xii 119 121 121 123 125 125 129 129 B 2 Focusing and Stigmating the Electron and Ion Beams B 3 Milling with the Ion Beam B 4 Pt deposition with the Ion Beam B 5 Shutdown Appendix C Confocal Microscope Operation C 1 Introduction C 2 Input Optics C 2 1 C22 C23 C 2 4 C25 Gaussian Beam Profile Beam Size Excitation Filter Polarization Alignment into Microscope C 3 Output Optics C 3 1 C32 C 3 3 C 3 4 C35
61. d by some structure on the glass air interface such as the bowtie in Figure 2 3 yellow squares viewed from the side The scattered signal is then collected through a confocal emission beam path Figure 2 1b If the signal is sent onto an APD and the sample is scanned a confocal image of the scattering signal can be formed so that a particular object of interest can be found but then the light must be directed onto a spectrometer to measure spectral information about the scattering signal Tungsten lamps are black body emitters and thus have a wavelength dependence of their emission This means that the sample is not uniformly illuminated in frequency space so a background spectrum must be obtained to normalize the scattering data Background spectra can often be taken by just measuring light leakage at a bare spot on the sample that is slightly rough e g at a spot with no bowtie and normalizing the scattering data to the background spectrum 33 L1 Polarizer Dove Prism _ Tungsten Lamp A m U E Glass Coverslip e f Objective To Spectrometer Figure 2 3 Schematic of TIR optical setup used to measure scattering from plasmonic structures As seen above this scattering setup is based upon a traditional fluorescence microscopy pathway A simple excitation pathway is used above the sample plane to ensure TIR and all that is needed is a confocal microscope with a spectrometer in order to sensitively measure the scatterin
62. data were taken at 1 2 kW cm while the perpendicular data were taken at 5 9 kW cm but the parallel data is scaled here to 5 9 kW cm for easy comparison b Red Blue SM TPQDI excited with light polarized parallel perpendicular xxii to the long axis of the bowtie Black dashed lines connect measurements from the same molecule After Ref a5 71 Figure 4 1 a Bowtie nanoantennas are immersed in concentrated dye solutions for FCS experiments b Blue absorption solid and emission dashed spectra of IR800cw in ethanol Red solid and emission dashed spectra of ICG in water Black plasmon resonance of a 10 nm gap Au bowtie nanoantenna Measured as in Ref Inset SEM of a typical gold bowtie nanoantenna Scale bar 100 nm c ICG molecule d IR800cw molecule 76 Figure 4 2 Confocal images of an array of bowties in the presence of a 100nM Figure 4 3 IR800cw in ethanol 109W cm imaging intensity b 100uM IR800cw in ethanol 3W cm imaging intensity c 30nm thick PVA film doped with IR800cw 36W cm imaging intensity d 1uM ICG in water 1 2kW cm imaging intensity e 1uM ICG in ethanol 600W cm imaging intensity f 30nm thick PVA film doped with ICG 1 2kW cem imaging intensity g Signal to background ratio of bulk enhanced fluorescence from 25 bowtie nanoantennas and different IR800cw concentrations 79 Photobleaching curves from cleaned ITO interfaces immersed in different dye solutions wi
63. down correctly do not leave it pumping indefinitely or you will damage the turbopump Select SED to turn off the OM Log out of the system Clean up after yourself as you leave the room Disable the FIB from CORAL 136 Appendix C Confocal microscope operation This chapter is an updated version of a guide previously published by Dave Fromm in his Ph D thesis C 1 Introduction The following guide will provide instructions to enable a new microscope user to use the modified Topometrix microscope system in all of its modalities The user is encouraged to consult the Topometrix User s Manual as there is a wealth of information available there This appendix will attempt to clarify the manual where applicable and provide caveats learned throughout the development of this instrument Despite its supposedly commercial nature this instrument is not easy to use and only two or three of these units were ever manufactured Therefore this instrument should be thought of as completely home built Several components have been modified from original parts but as there is no service offered for this instrument this is not important If there are any serious issues the user is encouraged to talk to Dr Stefan Kaemmer at Veeco who is one of the systems original designers Despite these warnings this instrument is incredibly versatile and offers stable mechanical components that can produce excellent AFM images concurrently with
64. e higher the pitch the better the resolution and the smaller wavelength range available in the spectrum It is important to exercise great care when handling gratings If they are scratched or dirty they CANNOT be cleaned or touched in any way Keep them in their protective case 151 C 3 7 Focusing Concave Mirror This mirror FM in Figure C 6 focuses the collimated beam at the exit port of the monochromator C 3 8 Exit port Instead of a slit a CCD detector chip is placed here to create a real time image of all spectral lines simultaneously The WinSpec software sums all vertical pixels into a single pixel value making an array with 512 pixel values calibrated by using an appropriate pen lamp source e g Hg Xe and laser sources The CCD camera has a housing that slides back and forth in this mounting and is held in place with a set screw You must minimize the width of the calibration lines on the camera by changing the focal position and you should make the lines as symmetrical as possible by rotating the camera to properly align the input slit and the pixel array axes Any C mount camera can be attached to the exit port by using the black C mount adapter that I machined located in Rm 8a Popular calibration lines are given here Hg lamp visible Xe lamp IR 404 66 nm 823 2 nm 435 84 828 0 546 1 840 9 577 0 881 9 579 0 895 2 C 3 9 Camera CCD cameras vary widely in their perf
65. e lifetime This equation assumes that the emission is split between two detectors using a 50 50 beamsplitter as shown in Figure 4 5 to allow correlation information to be extracted at 18 19 short times below the dead time of the APD detectors Fluorescence Emission 50 50 BS Figure 4 5 In order to measure autocorrelations at short time scales the fluorescence emission is split onto two detectors using a cube 50 50 beam splitter 80 Figure 4 6 plots the measured FCS curve for a 10pM concentration of ICG in ethanol blue and for a 10pM concentration of IR800cw in ethanol red both in the absence of bowtie nanoantennas The autocorrelation function in each case was computed from the photon arrival times using a commercial package Symphotime Picoquant These curves have been normalized to the value of G 100 ns They can be fit with the following standard equation which includes contributions from diffusion through the focal volume as well as a short lived dark state and are plotted as black dashed lines in Figure 4 6 1 1 N a 2 14 gt Tp KTD where N relates to the average number of molecules in the focal volume Tp is the 1 D De taark 4 2 diffusion time is a shape factor that describes the asymmetry of the ellipsoidal Gaussian focus D is the amplitude of the dark state contribution and Taak is the dark state lifetime For ICG tp is 608 us which corresponds to the amount of time the m
66. e trace of fluorescence from a single bowtie nanoantenna Black and red lines indicate times before and after one molecule photobleaches b Time delay histograms from time trace in a xXi Figure 3 9 corresponding the before black and after red photobleaching step c Blue Normalized single molecule time delay histogram formed by subtracting the red from the black curves in b Green is the instrument response function The deconvolved lifetime for this curve was less than 10 ps the minimum value we were able to determine experimentally After Ref 68 Enhanced single molecule fluroescence time delay histograms a Magenta bulk TPQDI in PMMA without bowtie nanoantenna Green SM on bowtie nanoantenna fr 271 lifetime 78 ps Red blue SM on bowtie nanoantenna excitation polarization parallel perpendicular to long axis Black instrument response function b Black Scatter plot of decay lifetime versus brightness enhancement for 73 SM s of TPQDI on bowtie nanoantennas Magenta Bulk TPQDI lifetime without bowtie nanoantenna present After Ref 69 Figure 3 10 Polarization dependence of single molecule enhanced fluorescence a Time trace for a single molecule with changing excitation polarization The polarization is changed from parallel red to perpendicular blue orientations with respect to the long axis of the bowtie Due to differences in dichroic reflectivity the parallel orientation
67. ectly the overall dose was chosen to be high enough It is still unknown why liftoff works well on flat areas when using float coated resist but the resist on the tip cannot be lifted off 200nm Figure 5 6 SEM s of best attempt at E beam bowtie fabrication on an AFM tip a SEM of an AFM tip after development and metal deposition An entire array of bowties were written on the cantilever not just on the tip apex so the white spots are bowtie shaped holes in the resist The red lines indicate the position of the bowtie that was targeted for the tip b SEM of the same tip after titanium gold deposition and liftoff The gold has peeled off of most of the cantilever and is now draped on top of the tip itself c SEM of one of the bowties written on the flat part of the cantilever next to the tip This bowtie is misshapen due to writing approximately 3 um out of focus 5 3 8 E beam Fabrication Conclusions The two main problems in the E beam process are the bending of the cantilever during float coating and incomplete liftoff While the bending problem just makes the 99 eventual yield of bowtie AFM tips smaller the liftoff problem was never solved and thus no E beam bowtie AFM tips were fabricated A previous graduate student Arvind Sundaramurthy did successfully fabricate a few E beam bowtie AFM tips using this method Figure 5 7 so this fabrication is possible but very difficult to reproduce Figure 5 7 SEM of a bowt
68. ecules We probe cavity resonances using cross polarized normal incidence reflectivity with a tungsten halogen white light source The cross polarization configuration is used to obtain a sufficient signal to noise ratio to observe the cavity resonance above the 113 reflected background uncoupled to the cavity A typical reflectivity spectrum is shown in Figure 6 2a showing the multiple resonances of the L3 cavity the fundamental mode is denoted with a black box The spectrum of the fundamental mode Figure 6 2b is fit to a Lorentzian giving a quality factor of 10 000 The improvement in quality factor from Ref is due to better fabrication a b 500 0 wm 400 150 5 300 2 QO 200 O 190 100 5 i wre 760 780 800 820 840 8195 820 820 5 Wavelength nm Wavelength nm c d 10 2 080 10000 9 G amp o j Se 2 re 8 sh a 5 O a ee O4 Quam 8225 823 823 5 5000 2 2 Wavelength nm PL o Reflectivity 0 70 750 800 850 05o 80 850 Wavelength nm Wavelength nm Figure 6 2 a Cross polarized reflectivity measurement of a cavity The box indicates fundamental cavity mode b Reflectivity spectrum of high quality factor fundamental cavity mode box in a Spectrum shows additional peaks at shorter wavelengths from higher order but lower Q cavity modes Solid line shows Lorentzian fit with quality factor 10 000 c Fluorescence collected using a c
69. ed laser focal spot Thus the bowtie could provide a contrast enhancement device to monitor enzyme activity optically and in real time 123 A second possible use for a single molecule coupled to the bowtie nanoantenna is as a single photon source Single photon sources are single emitters that can only emit a single photon upon each absorption event By coupling a molecule to the bowtie nanoantenna the fluorescence from the molecule is increased up to a factor of 1 300 and the lifetime is shortened by a factor 20 This means that the bowtie molecule system could potentially emit photons very quickly because the excited state lifetime is shortened This system could be very useful in quantum cryptography applications where single photon sources ensure security of encrypted data In summary this thesis has shown that the bowtie nanoantenna and photonic crystal cavity are very useful structures in the field of fluorescence research especially in the single molecule regime Bowtie nanoantennas could potentially aid experiments that need single molecule sensitivity in crowded environments while the lithographic approach to defining molecule on a photonic crystal cavity gives a new approach to solve a problem that has plagued the photonic crystal community References 1 Kinkhabwala A et al Large Single Molecule Fluorescence Enhancements Produced by a Gold Bowtie Nanoantenna Nat Photonics 3 654 2009 2 Fromm D P et al Explorin
70. elds near a gold bowtie nanoantenna as a function of bowtie gap size using TPPL Figure from Ref 6 The previous work in the Moerner Lab described above characterized the bowtie nanoantenna by measuring its plasmon resonance enhancement of local field strength and SERS effects One goal of this thesis was to understand how these highly enhanced and confined fields affected a single molecule s fluorescence Chapter 3 4 1 3 Photonic Crystals 1 3 1 Motivation The bowtie nanoantenna highly confines and enhances local field strengths but it is not the only nanophotonic structure capable of altering local electromagnetic fields Multi dimensional photonic crystals first introduced by Eli Yablonovitch and Sanjeev John in 1987 are also able to trap and manipulate light Photonic crystals are formed by a periodic array of material regions with different dielectric constants in either one two or three dimensions The periodicity is often on the order of the light s wavelength Distributed scattering of light from the periodic interfaces leads to the formation of dispersive energy bands sometimes with energy band gaps Light within the energy gap experiences constructive interference on reflection and destructive interference on transmission and is therefore unable to propagate in the photonic crystal material The photonic crystal thus acts as a highly reflective mirror to frequencies within the energy gap One can then engineer op
71. eless microscopy optical imaging at 10 angstrom resolution Science 269 1083 1085 1995 5 Hamann H F Gallagher A amp Nesbitt D J Enhanced sensitivity in near field scanning optical microscopy Appl Phys Lett 73 1469 1471 1998 6 Hillenbrand R amp Keilmann F Material specific mapping of metal semiconductor dielectric nanosystems at 10 nm resolution by backscattering near field optical microscopy Appl Phys Lett 80 25 27 2002 7 Hartschuh A Sanchez E J Xie X S amp Novotny L High resolution near field Raman microscopy of single walled carbon nanotubes Phys Rev Lett 90 95503 2003 8 Bouhelier A Beversluis M R amp Novotny L Characterization of nanoplasmonic structures by locally excited photoluminescence Appl Phys Lett 83 5041 5043 2003 9 Hamann H F Kuno M Gallagher A amp Nesbitt D J Molecular fluorescence in the vicinity of a near field probe J Chem Phys 114 8596 8609 2001 10 Gerton J M Wade L A Lessard G A Ma Z amp Quake S R Tip Enhanced Fluorescence Microscopy at 10 Nanometer Resolution Phys Rev Lett 93 180801 1 2004 11 Farahani J N et al Bow tie optical antenna probes for single emitter scanning near field optical microscopy Nanotech 18 125506 125510 2007 12 Farahani J N Pohl D W Eisler H amp Hecht B Single Quantum Dot Coupled to a Scanning Optical Antenna A Tunable Superemitter Phys Rev
72. ence time trace binned to Ims for a bowtie immersed in 1uM Figure 4 8 IR800cw in ethanol using 430W cm laser intensity b Fluorescence time trace binned to Ims for a bowtie immersed in 1uM ICG in water using 144kW cm laser intensity Notice that ICG in water has higher contrast between enhanced molecules compared to background than IR800cw in ethanol 84 a FCS curves for a bowtie immersed with 1uM ICG in water when illuminated with pump intensity 1 3 kW cm blue 4 6 kW cm red XXIV 14kW cm green 50kW cm pink 144kW cm cyan 362kW cm purple and 940kW cm yellow The grey curve indicates the FCS curve for the same luM ICG in water solution but without a bowtie nanoantenna at 110kW cm laser intensity b FCS curves from a are normalized to their value at t 100ns and clearly show that the photobleaching time Tphoto decreases as the laser intensity increases Fits to each curve using equation 4 3 are plotted with dashed black lines The FCS curve for a 10pM solution of ICG in the absence of a bowtie nanoantenna with 2 9MW cm laser intensity is plotted in solid black c e Fit parameters used for fit curves shown in b using equation 4 3 87 Figure 4 9 a FCS curves for a bowtie immersed in 100nM IR800cw in ethanol when illuminated with 0 14 kW cm blue 0 47 kW cm red 1 3 kW cm green 4 6 kW em pink and 13 8 kW em cyan The grey curve indicates the FCS curve for the same 100nM IR800 in
73. ent marks to the flattened tip apex At 93 this point taking a quick SEM of the tip apex is necessary and should not harm the sample since there is no resist on the sample to expose This brief SEMing will deposit a small amount of carbon onto the sample but not enough to harm the bowtie that will be fabricated there Once the sample is loaded each of the alignment marks and AFM tip apex are found by moving the feature into the middle of the image and recording the stage coordinates This step can be done in local coordinates by setting the tip s flattened apex or one of the alignment marks to be located at 0 0 This measurement step should be done with the Raith150 E beam and not the FEI FIB because the FEI FIB is not as accurate for absolute stage measurements In addition to measuring the x and y coordinates the Z focus change between the cantilever base and tip apex is measured The Z focus change is important to measure accurately because the E beam is sensitive to 1 um changes in focus and the distance between the cantilever base and apex is 3 um It is important to be able to precisely focus the beam in order to write bowtie nanoantennas with sharp features and small gaps 5 3 3 Chrome Etch Now the 4nm thick layer of chrome is etched by soaking in CR 14 chrome etchant for 5 seconds CR 14 is fairly specific to chrome and will not etch the Si3N4 in this short period of time 5 3 4 Float Coating of E beam Resist The next s
74. ering with Au bowtie nanoantennas J Chem Phys 124 061101 2006 15 Sundaramurthy A et al Toward Nanometer scale Optical Photolithography Utilizing the Near Field of Bowtie Optical Nanoantennas Nano Lett 6 355 360 2006 16 Schuck P J Fromm D P Sundaramurthy A Kino G S amp Moerner W E Improving the mismatch between light and nanoscale objects with gold bowtie nanoantennas Phys Rev Lett 94 017402 2005 17 Fromm D P Sundaramurthy A Schuck P J Kino G S amp Moerner W E Gap dependent optical coupling of single bowtie nanoantennas resonant in the visible Nano Lett 4 957 961 2004 18 Farahani J N et al Bow tie optical antenna probes for single emitter scanning near field optical microscopy Nanotech 18 125506 125510 2007 19 Farahani J N Pohl D W Eisler H amp Hecht B Single Quantum Dot Coupled to a Scanning Optical Antenna A Tunable Superemitter Phys Rev Lett 95 017402 1 017402 4 2005 20 Merlein J et al Nanomechanical Control of an Optical Antenna Nature Photonics 2 230 2008 21 Muhlschlegel P Eisler H Martin O J F Hecht B amp Pohl D W Resonant optical antennas Science 308 1607 1609 2005 22 Tang L et al Nanometre Scale Germanium Photodetector Enhanced by a Near Infrared Dipole Antenna Nature Photonics 2 226 2008 23 Oulton R F et al Plasmon Lasers at Deep Subwavelength Scale Nature 461 629 2009
75. f the light is reflected sin 0 2 2 n l where n and n are the indices of refraction for the two materials at the boundary and Oc is the angle beyond which all light will be reflected back into the first material The TIR method described below was used for measuring the plasmon resonance for gold bowtie nanoantennas using scattering 2 3 2 Optical Setup The optical setup for measuring scattering is relatively simple as seen in Figure 2 3 White light is taken from a tungsten lamp and sent through a polarizer in order to achieve S polarization of the excitation light perpendicular to the plane of incidence i e perpendicular to the plane of the figure The light is also sent through a lens L1 in Figure 2 3 which focuses the light down to a smaller spot on the sample This is 32 not a diffraction limited focal spot it just reduces the beam size to approximately Imm diameter Note that a large area of the sample is illuminated by the excitation light so the confocal emission pathway is used to ensure only 1 structure is measured at a time The key to this scattering setup is the use of a dove prism that is index matched to the glass substrate upon which the sample is fabricated When light enters the dove prism it is directed down onto the sample at an angle that ensures TIR off of the substrate glass air interface TIR ensures that the only light that is collected by the objective is light that has been scattere
76. f the CCD spectrometer system is clearly the most arduous task in this setup However once adjusted this path should only require re alignment upon changing the grating or substantially altering the laser wavelength The only 153 adjustment that should be done regularly is calibrating the X axis of the spectrometer readout using the pen lamp a task that takes just minutes C 4 Software C 4 1 Introduction There are 2 programs on 2 separate computers that are needed to acquire confocal images using the Topometrix microscope First a LabVIEW program is run on computer WEM 16 which integrates the number of counts from the APD into 10ms or longer bin times This computer then sends out an analog signal proportional to the number of counts measured to the AFM controller The AFM controller which controls stage movement is run by WEM12 This computer uses the Topometrix software to move the stage and build up images from the APD counts that WEM16 collects and processes in real time C 4 2 Using Bin APD counts LabVIEW Program A relatively simple LabVIEW program Bin APD Counts vi was written using LabVIEW version 8 6 on WEM16 LabVIEW is notorious for having poor compatibility between different versions of its software so avoid updates to LabVIEW on this computer This program is not very complicated however and can be fairly easily be rewritten in a new version especially by a user familiar with LabVIEW To run this progra
77. f you are grossly out of focus reset the focal length to something that makes sense near 5mm if you mounted the sample properly and work from there This is particularly helpful if you are zoomed out completely and the focus changes quickly 2 Astigmatism e Ifthe out of focus image is streaky and the streakiness changes direction 90 on either side of focus there is astigmatism Astigmatism also causes a loss of resolution at focus e Important The specimen should be at the center of focus before beginning This means that there is no streakiness in the image it is uniformly blurry This will make the stigmating the beam much easier e Adjust the stigmators one at a time to obtain the sharpest image e After adjusting the X and Y stimators once re focus and check for streakiness Iterate between focus and X Y stigmators as needed If necessary start at a lower mag and repeat the adjustment at the higher mag Check your final focus 3 Aperture alignment e Ifthe image moves as you change focus focus as best you can then choose the lens alignment icon e Focus wobble will be engaged causing the focus to wobble back and forth through the center of focus 133 e Click and drag on the crosshairs in the pop up window to minimize image movement e Deselect the lens alignment icon and check that the image no longer sweeps B 3 Milling with the lon Beam 1 Setup Ion beam for imaging e Choose the appropriate aperture f
78. fabricated bowtie nanoantenna during AFM imaging c SEM of SiN AFM tip before FIB milling Scale bar 1 um d SEM of same Si3N AFM tip after FIB milling Scale bar 1 um 5 3 E beam Lithography Approach 5 3 1 FIB milled Alignment Marks In order to use E beam lithography alignment marks on the cantilever near the tip are necessary When performing E beam lithography it is important to avoid 92 directly SEMing the intended write region when it coated in resist as SEMing develops the resist SEMing is however the only way to orient the sample so remote alignment marks are used to setup a local coordinate system in order to avoid direct exposure of the tip to the electron beam once it has been coated in resist Figure 5 2 is an SEM of the typical alignment crosses milled using a FIB where the actual tip is just visible at the top of the image Each line of the cross is 5 um long and lt 100 nm wide When making alignment marks the goal is to fabricate marks that are easy to locate hence the 5um length of the lines as well as highly precise in their center position lt 100nm widths of the lines Figure 5 2 SEM of calibration marks milled into an AFM cantilever Scale bar 5 um 5 3 2 Locating Alignment Marks Once the tip is flattened and has alignment marks the FIB is no longer needed The sample is loaded into the E beam without resist and with the chrome layer in order to measure the distance from the alignm
79. ffraction limited confocal spot of the objective 400nm in diameter for near IR light This alignment between the AFM tip and objective is the most important and often most difficult part of ANSOM Single molecule emission is collected through the same objective as the excitation light The majority of the fluorescence signal will originate from the concentrated spot of light Plasmonic tips have been shown to concentrate the light down to 10 10 10nm By raster scanning the sample a near field image is made of the surface and its resolution is only limited by the concentration of light by the tip 10nm not the diffraction limit 400nm 47 Objective Figure 2 10 Schematic of a typical ANSOM experiment A metal coated AFM tip is excited with light The light is concentrated down to 10nm due to the plasmon resonance of the structure which means the resolution of the imaging system is also 10nm Emission from the sample is collected back through an objective into a standard confocal emission pathway References 1 Davidovits P Egger M D Scanning Laser Microscope for Biological Investigations App Optics 10 1615 1971 2 Pawley J B in Handbook of Biological Confocal Microscopy ed Pawley J B 988 Plenum Press New York 1995 3 Corle T R amp Kino G S in Confocal Scanning Optical Microscopy and Related Imaging Systems Academic Press San Diego 1996 4 Moerner W E am
80. g spectrum from individual plasmonic nanostructures 2 4 Nanofabrication Techniques 2 4 1 Introduction Nanotechnology has become a popular field of study partially because of the new techniques available to fabricate and characterize nanostructures Electron Beam Lithography and Focused Jon Beam Milling will be discussed in this section as well as in Appendices A B 34 2 4 2 Electron Beam Lithography Electron Beam Lithography EBL has been used for over 50 years to fabricate micron and nanometer scale structures The basic process consists of defining the pattern using a focused electron beam to expose resist followed by a development step which removes the exposed resist The patterned resist can then be used as a mask for material deposition or etching Finally the resist is removed leaving behind only the patterned material This chapter specifically details the steps necessary to make bowtie nanoantennas using EBL on both conducting and insulating substrates in the Stanford Nanofabrication Facility SNF Figure 2 4 shows the process flow for fabricating gold bowtie nanoantennas onto a conductive substrate Indium Tin Oxide ITO In step 1 a square quartz coverslip Esco is cleaned by rinsing in acetone then plasma etching in Argon for 5 minutes Tom Carver in the Ginzton cleanroom facility then deposits a 50 nm thick layer of ITO usually in batches of 20 coverslips Next a 50 60nm thick layer of poly methyl methacrylate
81. g the chemical enhancement for surface enhanced Raman scattering with Au bowtie nanoantennas J Chem Phys 124 061101 2006 3 Eid J amp et al Real Time DNA Sequencing from Single Polymerase Molecules Science 323 133 138 2009 4 Aouani H et al Crucial Role of the Adhesion Layer on the Plasmonic Fluorescence Enhancement ACS Nano 3 7 2009 124 Appendix A EBL using Raith 150 A 1 1 Writing Bowtie nanoantennas with Raith 150 Load Samples Remove the stage from the vented system and load samples Replace stage and click Load Sample button Important be sure to squeeze door until the pump engages otherwise you will get a load lock error and have to call James Conway or a Raith champion to proceed Initial Setup Drive stage to home position gt click yes Reset coordinate system gt click yes Enter sample name gt Name your sample Set Column gt 10 KV Set Aperture gt 10 um Set Working Distance gt 5 mm Set stigmation amp aperture alignment to database values gt click yes Note that these are the settings for bowtie nanoantennas on glass If larger apertures column voltages are used the writes will be faster but the feature size will suffer Measure Beam Current Drive sample to correct Z position by setting Z position in Stage Control to 20mm in Absolute amp XYZ and then clicking GO Click go to Faraday cup on Holder Start
82. gure 1 9b are due to the number of different vibrational levels available for the molecule to relax down to the ground state at room temperature b 4 0 8 0 6 0 4 0 2 80700 750 600 850 900 950 Wavelength nm 13 Figure 1 9 a Simplified Jablonksi diagram for a typical fluorescence transition The emitter is pumped out of the ground state S and into vibrational sidebands of the electronic excited state S4 with rate Yas blue arrow Internal conversion fast non radiative transitions allows the molecule to relax into the lowest level of the excited state At this point the molecule relaxes back to the ground state either radiatively with rate y red arrow or non radiatively with rate y black wavy arrow Another internal conversion step black wavy arrow allows the molecule to relax to the lowest ground state level b Absorption blue and fluorescence emission red spectra from the molecule TPQDI 1 4 3 Single Molecule Fluorescence Over 20 years ago it was discovered that a single molecule could be optically detected first based on a measuring a single molecule s absorption of light and then demonstrated by measuring absorption by recording a single molecule s fluorescence emission at low temperature Room temperature detection of single molecule fluorescence quickly followed and the field of single molecule spectroscopy and imaging rapidly grew in particular in the biological community N
83. h extracts information from huge numbers of passages of molecules through the focal volume Examples of processes that affect fluorescence on these time scales are photon antibunching dark state bottlenecks photobleaching conformational dynamics FRET and diffusion FCS has been used to measure these processes in a number of free dye and biological systems see Ref for reviews of key work in this field 16 Figure 1 10 Experimental schematic for a typical FCS experiment A laser is focused tightly such that when fluorescent molecules yellow circles with trajectories in black in solution wander through the focus of the laser bright flashes of light are detected 1 5 2 Fundamentals When performing an FCS experiment either the autocorrelation is measured with a special purpose hardware autocorrelator or each photon s arrival time is measured precisely and then the autocorrelation G z of the time trace is computed in software using _ 61 t S1 t t EO aay 1 8 where I t is the intensity rate of photon emission at time t This autocorrelation function will reflect any dynamics in the dye that produce a change in the fluorescence emission before the molecule diffuses out of the focal volume typically a few 17 milliseconds For a confocal microscope an ellipsoidal Gaussian excitation volume can be assumed and G t for just simple diffusion is given by G T 1 9 ooo N 14T rp 1 a
84. hant ear d watch OM image while closing the chamber door to make sure the sample does not touch the lens e While holding the door firmly closed via the push bar click on pump command Keep pressure on the door for a few seconds tug to check the seal e Click cancel on the Confirm holder settings dialog box e Wait until Vac OK message appears at the bottom of the startup page about 3 5 mins Bring up electron and ion HV e Set electron and ion beams to 5 kV and 30 kV respectively e Electron spot size is normally set at 3 e Turn on HV for electron and ion beam by pressing HV buttons 130 6 7 Set initial height Click very top left blue Start freeze scan button in order to start SEMing IMPORTANT When the scanning begins the e beam confirm focus window pops up DO NOT CLICK OK ON THIS WINDOW UNTIL YOU OBTAIN A SEM IMAGE AND FOCUS AS INSTRUCTED see following This tells the computer how far your sample is from the lens With primary beam E icon highlighted choose either the SED or CDM E detector During your session feel free to test out both detectors for your sample and use the one that gives the best images e BEFORE ADJUSTING FOCUS adjust your contrast brightness knobs to give you an image This is a general rule if there is no image make sure the contrast brightness settings are at reasonable values e Focus it as you move up to
85. he desired position You need to calibrate these sensors and make sure they are linearized using software control The stage has been linearized by previous users and the stage files are saved in the following directory c programfiles Veeco SPMLab6 02 scanners Another important file is c programfiles Veeco SPMLab6 02 stages ini It must refer to the stages that you intend to use Because this version software is much newer than the Lumina scanner system that we have you need to make sure that it knows what the Lumina is and its two scanners the tripod and the tip scanner This file is backed up on floppy disk as well as saved on 2 computers WEMO01 Users Anika Backups Very Important Software WEM 12 VWEECO SPML602 WEM16 Users Backed Up Anika Very Important Software WEM 12 VEECO SPML602 The computer running the topometrix software has had its hard drive crash at least twice Check the hanging file for information about installing software onto a new hard drive should the current hard drive fail C 5 2 Calibration and linearization of stages Occasionally it is important to ensure that the stage is appropriately linearized You need to minimize the cross talk between the X and Y axes in the stage file and 159 this should be checked for each scan range that you intend to use I typically do the 50 um 20 um 10 um and 5 um files If you scan smaller the amount of non linearity is difficult to see anyway The Topomet
86. he measurement of the single molecule s excited state lifetime through Time Correlated Single Photon Counting TCSPC Since confocal microscopy is often implemented with a high speed detector like an APD each photon that is received on the detector can be time tagged with down to 5Ops precision for example by using a Micro Photon Devices PDM series APD in conjunction with a Picoharp 300 This time tagging ability is very useful in fluorescence microscopy because now quantities like the time spent by the molecule in the excited state excited state lifetime fluctuations and antibunching can be directly observed Figure 2 1 shows how this time tagging can be used to measure a single molecule s excited state lifetime using a pulsed excitation laser By using a pulsed laser the molecule is excited at regular intervals so the precise time the molecule is excited is known blue 30 lines in Figure 2 2 Signal photons are detected by the APD and the time of photon detection can be referenced to the last measured pulse from the laser this is called the time delay Each photon s time delay can be measured and this corresponds to the amount of the time the molecule spent in the excited state before emitting a photon In Chapter 3 this method is used to produce time delay histograms for SMs in order to tease out the excited state lifetime of a molecule and show how it is affected by a bowtie nanoantenna Pulsed Laser Fluorescence
87. hematic of bowtie Figure 3 2 nanoantenna gold coated with TPQDI molecules black arrows in PMMA light blue on a transparent substrate b TPQDI molecular structure c SEM of Au bowtie nanoantenna bar 100 nm d FDTD calculation of local intensity enhancement bar 100 nm e Red blue absorption emission spectra of TPQDI in toluene Green Scattering spectrum from bowtie shown in c measured as in Ref Black line laser excitation wavelength After Ref 53 Imaging unenhanced single molecule fluorescence a Confocal fluorescence scan of a low concentration lt 1 molecule diffraction limited spot sample of TPQDI in PMMA without bowtie nanoantennas scale bar 4 um b Fluorescence time trace of a single unenhanced TPQDI molecule aligned along the excitation polarization axis Data collected with 79 kW cm then scaled for direct comparison with Figure 3 3b c Histogram of unenhanced single molecule TPQDI brightness values from same low concentration TPQDI doped PMMA sample Data collected with 79 kW cm After Ref 56 Figure 3 3 Measuring enhanced fluorescence from single molecules on bowtie nanoantennas a Confocal scan of 16 bowties coated with high concentration 1 000 molecules diffraction limited spot TPQDI in PMMA collected with 2 4 kW cm scale bar 4 um b Fluorescence XIX time trace of TPQDI PMMA coated bowtie nanoantenna shown in Fig Ic Blinking dynamics and eventual photobleachi
88. hing steps Figure 3 4b is a plot of the fr values measured for 129 single molecules as a function of bowtie gap size The smallest gap bowties yielded the highest fr s up to a factor of 1340 consistent with smaller gap bowties having higher local field strengths than larger gap bowties and single triangles Of course broad distribution of fr values occurs because not all molecules are optimally located 57 3 5 Finite Difference Time Domain Simulations As shown above the bowtie nanoantennas enhance single molecule fluorescence an order of magnitude more than any other reported plasmonic structure to my knowledge Fluorescence can be enhanced both in both by increases in absorption and in emission so in order to understand this system finite difference time domain simulations were used to simulate these enhancements First the absorption of light by a molecule is proportional to IE as also shown in Figure 3 5a Thus the enhancement of the absorption of light fg is simply the change in the squared field strength or optical intensity due to the plasmonic antenna 2 Eagal E fe 3 2 2 inc The bowtie nanoantenna has been shown to locally enhance IE up to a factor of 1 000 which corresponds to the maximum expected value for fz The change in IE will be calculated below for a molecule located in the precise center of the bowtie s gap which we believe to be the position of highest enhancement and for molecu
89. ie on an AFM tip fabricated by Arvind Sundaramurthy using E beam lithography Scale bar 1 um 5 4 Focused lon Beam Process Flow 5 4 1 Introduction An alternative to using E beam lithography to fabricate the bowtie nanoantenna is to instead use the FIB In this scheme the entire tip is covered in gold and the FIB removes all the gold in a 4um area except for a bowtie shaped region This fabrication is much easier but there is one severe drawback When the FIB mills it deposits Ga ions into the substrate which alters the optical properties of the bowtie and renders the tips essentially unusable for fluorescence enhancement 100 experiments This section will detail the fabrication involved as well as optical experiments showing the lack of fluorescence enhancement when using FIB bowtie nanoantennas on AFM tips 5 4 2 Chrome Etch and Gold Deposition The first step in this process is to etch the conductive chrome layer off of the tip in CR 14 chrome etchant Calibration marks will not be used so the chrome is no longer necessary Next a uniform layer of 4 nm titanium and 20 nm gold is deposited onto the AFM cantilever Figure 5 8a Note that this gold layer serves to make the sample conductive for the FIB milling steps ahead a Gold Deposition b Focused lon Beam Milling Figure 5 8 Schematic of FIB Process Flow a A 4 nm titanium sticking layer followed by a 20 nm gold layer are deposited by Tom Carver in the Gin
90. ightly bent or occasionally not bent at all The bending is not due to the weight of the resist but rather due to the tension of the PMMA film This means that sometimes with great care and luck the tip can be gently tapped with tweezers from underneath to break the film and relieve the tension within the PMMA film While this is a problem with enough persistence some tips 95 will survive this step with minimal to no bending Note that bending alters the change in the focal distance as well as the change in lateral distance between the alignment marks and apex tip so these measurements will be incorrect if significant bending has occurred The second problem with float coating is the irreproducibility in the thickness of the resist While float coating forms locally uniform resist films the thickness over the entire film can change from 20 200 nm The dose used in exposing resist is critically dependent upon the thickness so it is difficult to correctly expose a resist film with unknown thickness A moderately high line dose 360pA s is used so that the resist will likely be fully exposed Figure 5 4 SEM showing cantilever bending after float coating of E beam resist 96 5 3 5 Chrome Deposition Since the AFM tip is non conductive a 4nm thick layer of chrome must once again be deposited onto the AFM tip Figure 5 5a Note that this is the second chrome deposition required and that the two chrome depositions cannot be co
91. ing but hurts signal transmission Typically a 75 100 um pinhole is best particularly in the infrared where spherical aberration and the long wavelength slightly enlarge the image spot Alignment of the confocal pinhole is the most critical aspect of confocal alignment Since the microscope has a fixed tube length the beam is converging just after the exit port of the microscope If working in the infrared it is best to align the pinhole to this beam using a visible laser for instance the 633nm HeNe on the setup first If the infrared is aligned to the same position as viewed on the Genwac camera then the alignment will be close after the initial alignment with the visible laser In order to align the beam initially use an index card or power meter to measure the beam after the pinhole Walk the X and Y axes of the pinhole until the power is maximized The output should be very symmetrical and extremely sensitive to the focus of the objective lens Make sure to check that you have found the real maximum not a local maximum by going well past the maximum signal position in each direction The Z position of the pinhole is not nearly as sensitive as the X and Y positions but should be adjusted to ensure that the light is maximized through the pinhole when the microscope beam is focused through the eyepiece or on the 146 Genwac This places the pinhole properly at the microscope image plane which is formed by the microscope tube le
92. ing it a useful A level precision tool Laser Photo diode Ce Cantilever Piezo electric y B scanner i Figure 2 9 Schematic for typical AFM experiment A cantilever with a sharp AFM tip is scanned over a sample surface Nanometer scale tip deflections from the sample surface are measured by reflecting a laser off of the back of the AFM tip and onto a quadrant photodiode which senses different intensities based on the tip deflection Figure from Ref 2 5 3 Apertureless Scanning Near field Optical Microscope Setup A typical ANSOM setup has two main components an AFM head and a confocal optical pathway I have used a commercial setup called the Topometrix Lumina that combines an inverted optical microscope with an AFM head where the 46 sample stage can be precisely scanned in X and Y with closed loop ositioning accuracy This device is no longer sold as Topometrix was purchased by Thermomicroscopes and then by Veeco A simplified schematic of the setup is shown in Figure 2 10 For full detail on how to use this setup see the detailed users manual For ANSOM the main idea is that an AFM tip that has been modified to have a plasmon resonance either by a metal coating or by lithographically fabricating a plasmonic antenna on top is put in contact with a transparent substrate containing emitters to be studied either fluorophores or Raman active molecules The tip is then aligned such that it is centered on the di
93. ingle C shaped nanoapertures Appl Phys Lett 85 648 650 2004 14 Taminiau T H Stefani F D Segerink F B amp van Hulst N Optical Antennas Direct Single Molecule Emission Nat Phot 2 234 237 2008 15 Binnig G Quate C F amp Gerber C Phys Rev Lett 56 930 1986 49 Chapter 3 Large Single Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna The research reported in this chapter has been previously published in A Kinkhabwala Z Yu S Fan Y Avlasevich K Millen and W E Moerner Large Single Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna Nature Photonics 3 654 657 2009 published online October 18 2009 3 1 Introduction Due to the size mismatch between light and nanoscale objects like single molecules it is important to be able to control light molecule interactions Plasmonic nanoantennas create highly enhanced local fields when pumped resonantly leading to increased Raman scattering but whether fluorescence enhancement occurs depends upon a variety of factors While sharp metal tips and colloids can enhance fluorescence the highly enhanced optical fields of lithographically fabricated bowtie nanoantennas provide a structure that is more controllable and can potentially be integrated Using gold bowties we have observed enhancements of a single molecule s fluorescence up to 1340x an order of magnitude higher than repor
94. ion blue and fluorescence emission red spectra from the molecule TPQDI 16 Figure 1 10 Experimental schematic for a typical FCS experiment A laser is focused tightly such that when fluorescent molecules yellow circles with trajectories in black in solution wander through the focus of the laser bright flashes of light are detected 19 Figure 1 11 Zero mode waveguide geometry for high concentration FCS Yellow circles are molecules that occasionally enter the hole in the aluminum and emit fluorescence into the collection optics 22 Figure 2 1 a Schematic of typical excitation pathway for single molecule confocal microscopy b Schematic of emission pathway for confocal microscope showing the placement of a pinhole at the image plane which provides Z sectioning 28 Figure 2 2 Time tagging of photons is accomplished by measuring the time delay between a signal photon and the sync signal of a pulsed laser 33 Figure 2 3 Schematic of TIR optical setup used to measure scattering from plasmonic structures 35 xvi Figure 2 4 Process Flow for E beam Lithography of Bowtie Nanoantennas onto conductive substrate 1 Deposit 50nm thick layer of the transparent conductive oxide Indium Tin Oxide ITO onto a quartz coverslip Spin 50nm of PMMA using Laurel spincoater 2 Expose bowtie pattern into resist using Raith 150 E beam writer 3 Develop exposed resist in 1 3 MIBK IPA solution for 35s and rinse in IPA for 40s 4 Deposit 4nm Titani
95. ion in a two level system without a plasmonic antenna The blue arrow shows absorption of light rate of absorption of light Yaps is proportional to the incident electric field squared lEn For emission the radiative and non radiative pathways from the excited state must be considered b Jablonski diagram for fluorescence transition of a two level system coupled to a plasmonic antenna Absorption of light is still proportion to IEl but now the electric field is modified by the antenna to become Emeta The emission pathways have also been modified There are now 3 classes of pathways one radiative and two non radiative to consider 59 Since the radiative and non radiative pathways are changing the quantum efficiency QE of the molecule must be calculated For a molecule in free space the intrinsic QE is K j 3 3 Sr ae er while the QE for a molecule coupled to an antenna is yY r 7 ae KVI ai ee 3 4 After some simple algebra the enhancement or quenching of the QE from a single molecule coupled to an antenna becomes A Y 1 n n ar f5 3 5 It is not immediately obvious whether a molecule s QE will be enhanced or quenched by analyzing this equation alone so finite difference time domain simulations were used to calculate the changes in the radiative and non radiative rates of a molecule located in the center of a bowtie s gap in order to use in Equation 3 5 Simulations of a single
96. jective and after reflecting off of each successive mirror letting the beam shoot across the table to make sure you are perfectly level When you cannot use a fiber e g using the pulsed laser since fibers broaden the pulse width a 500 800um pinhole also works as a spatial filter Shoot a 141 collimated laser beam into this pinhole and let the beam diffract out from there You can typically get 25 of the laser through this pinhole depends on pinhole size and laser beam used This output beam will need to be allowed to slowly diffract out to a large enough diameter to overfill the back aperture of the microscope While only pseudo collimated this technique produces an excellent confocal spot C 1 2 Beam Size Once the beam is Gaussian and collimated it should also have a beam diameter that is greater than or equal to the back aperture of the microscope objective used If this condition must be satisfied in order to obtain diffraction limited images By choosing the collimating objective wisely NA 0 18 is currently used the beam will have the correct size already but a telescope can be built to magnify the beam if necessary C 1 3 Excitation Filter Diode lasers do not emit at a single wavelength but rather have a spectrum sharply peaked at the desired wavelength The tails of this spectrum are weak in comparison to the peak but for single molecule spectroscopy every photon that leaks into the emission pathway co
97. l easily damage it The program is now open and setup for AFM scanning but different settings are needed for confocal scanning Go to Setup gt Acquire Make sure that the Fwd and Rev checkboxes are only clicked for IN1 this is where the analog signal from the LabVIEW program is sent Since this is an AFM program it will scan every line in the forward and reverse direction This is a useful feature for AFM scanning because you need that information to judge if a feature is real but it is not necessary for confocal scanning Unfortunately there is not a way around it so the user will just have two images of the same object at the end of the scan Now the program is set to collect data from WEM16 Press the button with 3 yellow arrows to start a scan Of course the LabVIEW program Bin APD photons vi needs to be running and collecting data otherwise the Topometrix program will just plot 0 for the entire image The scan range scan rate and resolution 157 should all be set to acquire data at the rate the LabVIEW program is binning it That means if the LabVIEW program has 10ms bins the Topometrix program settings should yield 10 ms pixel One set of conditions for a 10 ms bin time is Scan Range 20 um Scan Rate 20 um Resolution 100 which yields 10 ms bin times and 200 nm x 200 nm pixels These settings could be altered to yield a 100nm pixel size by doubling the resolution and halving the Scan Rate Image
98. le emitting at 820 nm wavelength After Ref a 62 bowtie modified QE versus the intrinsic QE Notice that if a molecule with 7 100 is used then the quantum efficiency is actually quenched to y 25 but if a molecule with 4 2 5 to 25 in used then the bowtie modified quantum efficiency is enhanced to the same 7 25 The QE enhancement factor is plotted directly in Figure 3 7b Notice that for molecules with 7 gt 25 the QE is quenched while for molecules with lower QE there is significant enhancement If the intrinsic QE is too low however the modified QE may have a large QE enhancement f but it will have a low final QE 7 The optimal QE to balance these two effects is 4 2 This means that the emission from high QE molecules cannot be enhanced so by choosing a relatively low QE molecule higher overall fluorescence enhancements can be achieved Based on the simulations above one can estimate the total fluorescence enhancement factor ff fr Sefa 3 9 For a molecule located in the gap of the bowtie the absorption enhancement is calculated to be fg 180 As shown above the emission or quantum efficiency enhancement depends upon intrinsic quantum efficiency and for TPQDI q 2 5 is calculated from Equation 3 5 to be f 9 3 Taking these two factors into account in Equation 3 9 yields a predicted total fluorescence enhancement ratio fr 1690 63 o Of 02 03 04 05 06 07 08 09 1
99. lecules even at 100 nM concentration To see this one finds that fluorescence enhancement from bowties immersed in a 100uM solution of IR800cw in ethanol is only barely detectable Under the assumption that the surface was already saturated with sticky molecules at 100nM concentration by increasing the concentration to 100uM only the background would increase drowning out the enhanced fluorescence from enhanced molecules stuck to 76 the surface near the bowtie Figure 4 2g enforces this point by plotting the amplitude of a Gaussian fit to 25 bowties immersed in various concentrations of IR800cw in ethanol divided by the background signal level the signal to background ratio or S B and shows that S B decreases steadily with increasing IR800cw concentration from 100nM to 100uM One might wonder if the highly anisotropic and concentrated optical field of the bowtie is producing trapping by gradient forces but this is not the case since when the optical intensity increases the lingering time of molecules near the enhanced region drops Finally Figure 4 2e is an image of bowties immersed in a luM concentration of ICG in ethanol While bowties submerged in ICG in water easily showed enhanced fluorescence Figure 4 2d this image shows that the effect disappears in ethanol By using separate observations of the photobleaching behavior of molecules in the presence of an ITO coated surface without bowties ICG was found to stick to the ITO surface in
100. les located at different z positions The change in emission of a molecule coupled to a nanoantenna is more complicated than the case for the change in excitation Figure 3 5a shows that once the molecule is excited from the ground So to the excited S1 state there are fast non radiative vibrational relaxation pathways black wavy arrow to the lowest excited state At this point the molecule effectively waits to relax to the ground state It can either decay radiatively and emit a photon red emission pathway y or it can decay non radiatively by creating high order vibrations or phonons black emission pathway 58 Ynr When a plasmonic antenna is placed near the emitter the picture changes Figure 3 5b Now the radiative pathway is not simply emission into free space but rather the emitter is coupling to the plasmonic antenna and then the antenna is emitting a photon into free space y Plasmonic antennas however are lossy Ohmic losses at optical wavelengths so once the emitter couples to the antenna the antenna could just lose energy by Joule heating from resistive currents yar Finally the non radiative pathways intrinsic to the molecule even without the plasmonic antenna are still present so they must also be included nr Yr Var nr y abs Ep i abs Ei Figure 3 5 Jablonski diagrams for fluorescence transition near and away from a plasmonic antenna a Jablonski diagram for a fluorescence transit
101. licon Nitride Photonic Crystal Nanocavities Opt Exp 15 17231 2007 119 12 Makarova M Vuckovic J Sanda H amp Nishi Y Silicon based Photonic Crystal Nanocavity Light Emitters Appl Phys Lett 89 221101 2006 13 McCuthceon M amp Loncar M Design of a Silicon Nitride Photonic Crystal Nanocavity with a Quality Factor of One Million for Coupling to a Diamond Nanocrystal Opt Exp 16 19136 2008 14 Akahane Y Asano T Song B amp Noda S High Q Photonic Nanocavity in a Two Dimensional Photonic Crystal Nature 425 944 2003 15 Avlasevich Y Muller S Erk P amp Mullen K Novel Core Explanded Rylenebis Dicarboximide Dyes Bearing Penatcene Units Facile Synthesis and Photophysical Properties Chem Eur J 13 6555 2007 16 Zhou H et al Lithographically Defined Nano and Micro Sensors using Float Coating of Resist and Electron Beam Lithography J Vac Sci Technol B 18 3594 2000 17 Rivoire K et al Lithographic Position of Fluorescent Molecules on High Q Photonic Crystal Cavities Appl Phys Lett 95 123113 2009 18 Martiradonna L Stomeo T De Giorgi M Cingolani R amp De Vittorio M Nanopatterning of Colloidal Nanocrystal Emitters Dispersed in a PMMA Matrix by E beam Lithography Microelecton Eng 83 1478 2006 120 Chapter 7 Conclusions 7 1 Conclusions Fluorescence is an important technique used throughout biology and so understanding how molec
102. lignment marks this second write can be aligned to the initial photonic crystal write to within 20 nm In this way all of the unwanted resist can be selectively exposed leaving the cavity s resist untouched A simple development step removes the E beam exposed resist and leaves behind a photonic crystal with the cavity region coated in fluorescent molecules Figure 6 3a b are confocal fluorescence images of a cavity before and after electron beam exposure and development measured with the same excitation power There is still strong emission from molecules coupled to the cavity after exposure though diminished by the exposure process but there is no emission from the nearby areas The contrast in Figure 6 3b is higher showing that by removing the molecules the background signal from molecules not located over the cavity is much lower Figure 6 3c shows a PL spectrum Q 4500 measured on the same cavity after localization of the resist to the cavity demonstrating that the remaining molecules are still spectrally coupled to the photonic crystal cavity An atomic force microscope image Figure 6 3d confirms that DNQDI doped PMMA from Figure 6 3b is localized to the cavity and is 12 nm in height and 700 nm by 400 nm laterally The AFM image shows a misalignment of approximately 300 nm between the cavity region and the lithography defined DNQDI PMMA region With optimization of the overlay process it should be possible to reduce this error to less
103. location and contrast brightness so check pattern positioning after inserting 3 Choose appropriate ion beam aperture e Use the following rule to choose the ion beam aperture a Areapattem Hm lt Aperture size pA lt 6 Areapattern m 4 Retract the needle e Make sure to retract the needle when you are finished depositing Do not translate or change tilt with the needle inserted B 5 Shutdown 1 Turn off GIS sources e Make sure any GIS needles are retracted and their heaters are turned off 2 Return to home settings e Return to a 10pA ion beam aperture e Return sample tilt to 0 e If using UHR mode return to SRH search mode e Set both x and y to 0 3 Turn off CDM E detector e Make sure that contrast and brightness for CDM E detector are zero This detector needs to be turned off in this way to reduce wear Check that it is off 135 10 11 even if you did not use this detector sometimes it is used by accident or a previous user forgot to turn it off e Select SED detector This detector can be left with its contrast brightness settings on Turn off beams e Turn off electron and ion high voltage by deselecting HV for both e Leave the ion source on if the next user will be on within 4 6 hrs else turn off the ion source Vent the chamber Select OM Remove your sample Pump the chamber e Make sure you obtain the Vac OK message If the chamber is not pumping
104. lt 20nm so is used for all bowtie nanoantenna fabrication Once development has finished the metallization step begins Step 4 In this step Tom Carver deposits a 4nm titanium layer followed by a 20nm gold layer The titanium layer is necessary as a sticking layer between the gold and ITO In Figure 2 4 it can be seen that the bowties are now adhered to the ITO but there is still unwanted gold remaining The final step liftoff Step 5 removes this unwanted gold by sonicating the sample in acetone to dissolve the remaining underlying PMMA removing the leftover Ti Au layer with it The sample is sonicated until all the excess gold is removed which can be seen by eye Usually this requires only a few seconds but it can sometimes take a minute to complete If it is not possible to fabricate bowtie nanoantennas on a conductive substrate then a slightly different technique is necessary to perform EBL As seen in Figure 2 5 the first step now consists of first spinning a 50 60nm thick layer of PMMA baking then depositing a 4nm layer of chrome chromium metal This thin layer of chrome prevents the sample from charging during the electron beam exposure in step 2 After exposure the chrome layer is then removed in step 3 using Chrome etchant Cyantek 38 CR 14 exposing the PMMA surface The sample is then developed Step 4 metal is deposited Step 5 and finally liftoff step 6 is performed as described previously Most importantl
105. m e Red blue absorption emission spectra of TPQDI in toluene Green Scattering spectrum from bowtie shown in c measured as in Ref Black line laser excitation wavelength After Ref Bowtie nanoantennas were fabricated in gold using electron beam lithography Raith 150 onto 50 nm thick indium tin oxide ITO coated quartz coverslips Experimental measurements of fr for a SM were performed by coating electron beam fabricated gold bowtie nanoantennas with the relatively low fluorescence quantum efficiency 7 2 5 but solubilized near IR dye N N Bis 2 6 diisopropylpheny l 1 6 11 16 tetra 4 1 1 3 3 tetramethylbutyl phenoxy quaterrylene 3 4 13 14 bis dicarboximide TPQDIJ doped in a thin poly methyl methacrylate PMMA layer Figure 3 1a TPQDI was doped into 1 wt vol of 75k MW PMMA Polysciences Inc in distilled toluene and spun onto the bowtie sample at 2500x RPM to achieve a 52 final thickness of 30 nm In addition to its low quantum efficiency TPQDI Figure 3 1b was chosen for the overlap of its absorption and emission spectra with the measured bowtie plasmon scattering resonance Figure 3 le 3 3 Confocal Imaging of Unenhanced Single Molecules In order to measure the enhancement of fluorescence from a molecule coupled to a bowtie nanoantenna the fluorescence expected from an unenhanced molecule must first be determined A 780 nm diode laser was used to excite fluorescence from TPQDI in a PMMA film in a
106. m on WEM16 simply click on the Bin APD Counts vi file located on the desktop or here C Users Backed Up Anika Labview programs Bin APD photons vi This program is commented in the Labview file so this section is just devoted to use of the code A screen capture of the front panel of the program is seen 154 in Figure C 7 Note that when using LabVIEW the program can be altered so avoid adding or deleting any of the options on the front panel unless you save a backup copy and have experience with LabVIEW programming Bin APD photons vi Front Panel File Edit View Project Operate Tools Window Help E IN 13pt Application Font JERES 2 Pl Timing Parameters Sample Clock Source Dev1 Ctr1InternalOutput j Timing Counter Generating Counter Be devijetri Frequency He A 100 00 J Counting Counter Counter s z A Dev1 ctr0 x Initial Count Jo y Count Direction Scot Up Edge S FRising y Waveform Chart Poto Za Counts from last 10ms period 0 Available samples if gt 1 then program is not reading fast enough jo stop lt gt Figure C 7 Screen capture of Bin APD photons program front panel In order to run the program simply click on the white arrow located at the top of the program just as any LabVIEW program starts The program will begin drawing out a time trace that corresponds to
107. mbined since the bowtie should not be fabricated on top of a chrome layer This means that the first chrome layer must be etched off then resist float coated and finally a new chrome layer deposited for E beam exposure 5 3 6 Standard E beam Lithography Steps The tip is now loaded into Raith150 for exposure of bowtie shaped features into the resist Figure 5 5b The alignment marks are located to setup a local coordinate system and then the bowtie shape is exposed on the apex of the tip Directions for using the Raith150 for a bowtie exposure can be found in Appendix A After the resist has been exposed the chrome must once again be etched off Figure 5 5c in Cr 14 chrome etchant for 5 seconds The exposed resist on the tip is then removed by development in 1 4 Methyl Isobutyl Ketone MIBK Isopropanol for 35 s followed by soaking for 40 s in pure Isopropanol Figure 5 5d Finally the tip is given to Tom Carver for 4 nm titanium sticking layer and 20 nm gold deposition Figure 5 5e 97 a Deposit Chrome b Expose Resist c Etch Chrome aoe d Develop Resist e Deposit Ti Au f Liftoff Aa A Aa Figure 5 5 E beam lithography process flow for nonconductive substrate a Deposit chrome onto float coated resist layer b Expose resist using Raith 150 E beam Lithography Tool c Etch chrome layer in CR14 chrome etchant to expose resist layer d Develop resist in 1 4 Methyl Isobutyl Ketone Isopropanol for 35 s and Isopropanol fo
108. mited to 5Onm For bowtie nanoantennas particularly in gold which mills easily the 1pA aperture was used to achieve the highest resolution Check that your ion beam and electron beam mags are coupled At 10 20KX center a recognizable feature while still in e beam mode Choose primary beam I icon and adjust contrast brightness and focus DO NOT MOVE THE STAGE Instead re center the feature using the beam shift knobs shifts the beam NOT the stage This will align the ion beam and electron beam images 9 Proceed with your sample B 2 Focusing and Stigmating the Electron and lon beams To obtain a good image and ion etching deposition results you must stigmate and focus both the electron and ion beams properly This takes practice As you learn how to operate the instrument make sure you are learning these procedures well as they will determine the smallest feature size possible with the FIB Focusing and astigmatism corrections should be made at least one magnification step above where you want to take a final image These procedures are 132 identical for both beams Be aware of possible beam damage as you align so you may want to adjust these away from your area of interest 1 Focusing e Use the hardware knobs or right click and drag the mouse e Be aware of the sensitivity of the coarse and fine focus knobs as one or the other will make more sense depending on your magnification e I
109. molecules 103 In this section measurements of the enhancement or quenching of fluorescence by a FIB bowtie on an AFM tip will be shown In order to test for fluorescence enhancement using a FIB bowtie on an AFM tip a thin film 30 nm of TPQDI doped PMMA was spun onto a clean coverslip and loaded into a confocal microscope The FIB bowtie AFM tip was then raster scanned above the sample and fluorescence was recorded as a function of bowtie position Figure 5 11 Since the sample remains fixed the recorded fluorescence image is the fluorescence as a function of AFM tip position so if the bowtie is enhancing fluorescence there will be more fluorescence photons detected when it is scanned through the focal volume of the objective Since the sample is stationary the same molecules are excited for the entire tip scan which means this experiment relies on the fact that a bulk sample of TPQDI excited with low power shows very little photobleaching over a 5 minute time scale Scan tip Objective 104 Figure 5 11 Schematic of setup used to test for enhancement of bulk TPQDI fluorescence using a FIB bowtie on an AFM tip Blue circles are bulk high concentration TPQDI molecules The above experiment was performed for both a FIB bowtie AFM tip and a sharpened gold coated AFM tip Figure 5 12a b For the FIB bowtie AFM tip there actually appears to be a slight quenching of fluorescence with the bowtie is scanned over the focu
110. mon polariton excited at a metal dielectric interface 1 2 4 Localized Surface Plasmon Resonance Now consider the case of exciting a plasmon in a nanoscale object such as a small metal sphere or colloidial particle In such a confined system the plasmon cannot propagate and is instead referred to as a localized surface plasmon resonance LSPR Figure 1 3 shows schematically how electrons in a small metal colloid respond to an electromagnetic field at the LSPR Notice that the electric field pushes the electrons to one side of the particle leaving behind positively charged holes on the other side By separating and localizing the electrons and holes on opposite ends of the nanoparticle there is a local buildup of electric field which produces an optical antenna For a metal sphere of diameter lt lt A the LSPR resonance is located at ENTER EQUATION and differently shaped small objects have different resonance conditions E Figure 1 3 Response of free electrons in a metal colloid to an AC electromagnetic field tuned to the particle s plasmon resonance 1 2 5 Gold Bowtie Nanoantenna Plasmon Resonance Metallic spheres are often used experimentally because making small metallic particles by colloidial chemistry is are relatively simple and spherical systems are analytically solvable using Mie theory but they do not give the highest electric field enhancements or tightest field confinement Many other geometries have bee
111. mp Webb W W Biological and chemical applications of fluorescence correlation spectroscopy a review Biochemistry 41 697 705 2002 20 Dittrich P amp Schwille P Photobleaching and stabilization of fluorophores used for single molecule analysis with one and two photon excitation App Phys B 73 829 2001 21 Burland D M Miller R D amp Walsh C M Second order nonlinearity of poled polymer systems Chem Rev 94 31 75 1994 22 Edman L Mets U amp Riger R Conformational transitions monitored for single molecules in solution Proc Natl Acad Sci U S A 93 6710 6715 1996 89 Chapter 5 Toward Bowtie Nanoantennas as Apertureless Scanning Near field Probes 5 1 Introduction Gold bowtie nanoantennas have been shown to greatly enhance the fluorescence from low QE molecules Chapter 3 but so far only bowtie nanoantennas fabricated onto glass coverslips have been discussed These structures could be very useful if instead fabricated onto a scanning tip and used for ANSOM Chapter 2 forming a bowtie on a tip BOAT Then the bowtie could be positioned directly above a molecule using standard AFM technology instead of relying on randomly distributing molecules around the bowtie Other scanning plasmonic tips exist such as metal coated fibers with sub diffraction limited apertures NSOM sharpened metal AFM tips ANSOM and other novel ANSOM probes such as aluminum bowtie nanoantenn
112. n in the Jablonksi diagram in Figure 1 9a On the excitation side a laser is chosen whose energy is equivalent to or greater than the direct ground state So to excited state S allowed dipole transition blue arrow The singlet state S refers to 12 the fact that all of the electrons in this state are paired with another electron of anti parallel spin A transition between two singlet states is an optically allowed transition because it does not require a spin flip to occur Figure 1 9b shows schematically the absorption blue and emission red spectra from the molecule TPQDI see Chapter 3 for more details on this dye The excitation laser can be chosen at any wavelength where there is absorption by the molecule Once the molecule is in the excited state internal conversion or fast non radiative vibrational relaxations black wavy arrow occur in a few ps and the molecule relaxes to the lowest level of the electronic excited state At this point the molecule typically remains in the excited state for few ns before relaxing to the ground state This relaxation can either be radiative red arrow where the molecule emits a lower energy photon or non radiative black arrow where the molecule does not emit a photon but simply gives off the energy as heat and moves through other levels to relax to the ground state internal conversion or intersystem crossing for instance The width of the emission and the peaks in the spectrum in Fi
113. n studied 6 10 11 12 1 14 21 such as metal wires 2 strips us cones 5 and bowties for a diverse set of applications including but not limited to photodetectors plasmonic lasers solar cells apertureless near field microscopy and photolithography as Summarized in the excellent review of Ref The bowtie shape was first studied at mid infrared wavelengths in the laboratory of Gordon Kino and was shown to produce efficient antenna effects This work was then scaled to fabricate nanoscale antennas in the near IR consisting of two 80 nm triangles separated by as small a distance as possible typically 10 nm SEM in Figure 1 4 When this subwavelength object is excited resonantly electrons in both triangles of this antenna move in the direction of the electric field causing a concentration of negative charge in the tip of one triangle and a concentration of positive charge in the tip of the other triangle Figure 1 5a a field configuration which switches for every half cycle of the applied electromagnetic wave This configuration leads to a very strong buildup of field in the gap as seen in the finite difference time domain simulation in Figure 1 5b Figure 1 4 SEM scanning electron microscopy image of a gold bowtie nanoantenna fabricated with electron beam E beam lithography Scale bar 40 nm Figure 1 5 a Schematic of electron and hole concentration due to excitation of the bo
114. n the spectrometer beam path is a broadband reflective mirror M1 which is good for 400 1000 nm It is mounted on a translation stage that enables translation of the beam across the input slit of the monochromator C 3 3 Focusing lens The next component in the beam path is a f 16 mm camera lens FL in Figure C 6 which is used because it is highly color corrected This focuses the beam through the entrance slit of the monochromator The focal length of this lens is matched to perfectly fill the grating of the monochromator This can be observed by sending a visible laser through the system and placing an index card on the grating mount be careful to avoid contact with the grating which is extremely fragile The beam should be round have a Gaussian profile and centered on the grating filling the entire active area 149 Figure C 6 Optics in the CCD Spectrometer assembly Note the input mirror M mounted on a translation stage the focusing camera lens FL the entrance slit S the collimating concave mirror CM the grating G focusing concave mirror FM and the CCD camera located at the exit focal plane of the monochromator C 3 4 Entrance slit A 150 um slit S in Figure C 6 is used aligned vertically on the input to the spectrometer For initial alignment remove this slit and have the beam properly fill the grating see above Then add the slit and re align to achieve the same condition Measure the
115. na and thus will have much lower fr cf blue vs grey dashed in Figure 3 6b 3 6 Excited State Lifetime Measurements The discussions above suggest that the enhancement of quantum efficiency should also produce a change in the SM total decay lifetime tp To probe the enhanced molecules excited state lifetimes a mode locked Ti Sapphire laser pulse length 200 fs was tuned to 780 nm and used in conjunction with a fast time resolution APD MPD PDM 100 series as well as a time correlated single photon counting analyzer Picoharp 300 to measure total decay lifetime In order to measure tr for a SM on a bowtie despite the presence of background fluorescence from other molecules a special procedure was implemented Key to the scheme is the fact that all fluorescence photons from the molecule coated bowtie were time tagged A typical binned time trace of these time tagged photons is seen in Figure 3 8a The black line in Figure 3 8a shows the time interval where the enhanced molecule is emitting while the red line indicates the time interval when the molecule has blinked off In both time intervals all other molecules are also emitting but they are assumed to not change and thus represent a constant background Time delay histograms are formed based on the fluorescence photons before black in Figure 3 8b and after red in Figure 3 8b a single molecule photobleaching step The algebraic difference in shape of these two time delay histograms is
116. ne diimide DNQDJ which was chosen for its broadband emission over the desired wavelength range 700 nm 850 nm good photostability and high fluorescence quantum efficiency QE 40 The structure of the molecule and its emission spectrum are shown in Figure 6 1 To couple DNQDI to photonic crystal cavities Figure 6 1c the molecule was dissolved into a solution of 1 poly methyl methacrylate PMMA in distilled toluene In standard lithographic processing this solution is then spun onto a surface leaving behind a smooth thin film of dye doped polymer resist However spinning onto an uneven surface such as a photonic crystal membrane causes unwanted aggregation of the dye doped PMMA Instead the solution was float coated whereby the photonic crystal sample is submerged into a water bath and a single drop of the dye doped PMMA in toluene solution is dropped onto the surface of the water bath The drop quickly disperses across the surface leaving a locally uniform layer of hydrophobic dye doped resist floating on top of the water bath The water is then pipetted away allowing the PMMA layer to gently rest on top of the photonic crystal sample The sample is baked at 90 C for 30 minutes to ensure that all the water is fully evaporated The concentration of DNQDI in the PMMA layer is approximately 5 molecules 100 nm 6 3 Optical Characterization of High Q Cavity Modes We first characterize cavities passively prior to depositing mol
117. ng are due to 1 molecule that has been enhanced by a factor of 1340 After Ref 58 Figure 3 4 Measurement of fr for SMs as a function of bowtie gap size a Figure 3 5 Figure 3 6 Histogram of gap sizes of all bowties measured b Scatter plot of 129 SM fluorescence brightness enhancements fr as a function of bowtie gap size for all bowties measured in a After Ref 59 Jablonski diagrams for fluorescence transition near and away from a plasmonic antenna a Jablonski diagram for a fluorescence transition in a two level system without a plasmonic antenna The blue arrow shows absorption of light rate of absorption of light Yabs is proportional to the incident electric field squared IE PY For emission the radiative and non radiative pathways from the excited state must be considered b Jablonski diagram for fluorescence transition of a two level system coupled to a plasmonic antenna Absorption of light is still proportion to IE but now the electric field is modified by the antenna to become Ejnetai The emission pathways have also been modified There are now 3 classes of pathways one radiative and two non radiative to consider 61 Electromagnetic simulations of SM fluorescence near a gold bowtie nanoantenna a Spectrum of calculated electric field intensity enhancement versus wavelength in the center of a bowtie with 14 nm gap Inset the simulated structure side view consists of a SiOz refractive index n 1
118. ng objective NA 0 18 the rear mirror Mp and the final mirror Mp 140 Schematic of typical optics used for a confocal microscope A single mode fiber SMF is used to produce a Gaussian beam profile for the excitation path followed by additional optics that control the beam s polarization power and spectrum The beam path is a confocal setup because the emission pathway is focused through a pinhole allowing for z sectioning An emission filter ensures only fluorescence reaches the detector 141 Alignment of the confocal beam using the Genwac CCD camera a The beam is centered in intensity but off of the ideal optical axis need to walk beam using both mirrors b Mirror M4 adjusted to move beam closer to XXX Figure C 5 Figure C 6 Figure C 7 ideal position but now is going in at an angle c Adjust Mirror M5 angle to fix angle and achieve properly aligned beam 144 Confocal microscope output optics Note the pinhole PH placed at the microscope image plane the collimating lens CL the 90 reflector flips in and out the focusing lens FL for the APD detector 146 Optics in the CCD Spectrometer assembly Note the input mirror M mounted on a translation stage the focusing camera lens FL the entrance slit S the collimating concave mirror CM the grating G focusing concave mirror FM and the CCD camera located at the exit focal plane of the monochromator 150 Screen capture of Bin APD pho
119. nna with 2 9MW cm laser intensity is plotted in solid black c e Fit parameters used for fit curves shown in b using equation 4 3 85 Turning now to the other fluorophore IR800cw even though this molecule is not optimal for bowtie FCS the FCS curves can still be recorded at low powers as shown in Figure 4 9a b The lower S B ratio makes the FCS curves have lower contrast and thus more challenging to measure Notice that the absolute G 100ns for these curves in Figure 4 9a for IR800cw is much lower than in Figure 4 9a for ICG a consequence of a lower S B ratio for IR800cw As was measured for ICG bowtie FCS the photobleaching time for IR800cw bowtie FCS is found to decrease as the excitation intensity increases Normalized G6 m 0 ee 10 10 10 10 10 10 10 10 10 10 10 10 z ms z ms 1 N o gt o N A O 1 Tokoto ms 05 1 10 10 10 Laser Intensity kWicm Laser Intensity kWicm Laser Intensity kWicm 0 Figure 4 9 a FCS curves for a bowtie immersed in 100nM IR800cw in ethanol when illuminated with 0 14 kW cm blue 0 47 kW cm red 1 3 kW cm green 4 6 kW cm pink and 13 8 kW cm cyan The grey curve indicates the FCS curve for the same 100nM IR800 in ethanol solution but without a bowtie nanoantenna at 1 3 kW cm laser intensity b FCS curves from a are normalized to their value at t 100 ns and clearly show that the photobleaching time decreases as the lase
120. ns inside microscope chassis C 2 2 Collimating the Emission Signal Following the pinhole there is a lens f 50 8 mm achromat to re collimate the beam This lens is fairly simple to align and shouldn t need to be adjusted often C 2 3 Emission filters Depending on the emission filters usually bought from Omega Optical or Chroma Engineering multiple long pass emission filters may be necessary to reject laser light at the proper frequency The further the filter s turn on is red shifted from the laser the fewer filters perhaps only 1 are needed C 2 4 Aligning the Avalanche Photodiode APD There is a lens f 50 8 mm achromat that focuses the fluorescence signal onto the APD chip The APD is mounted on an X Y Z stage to center the beam onto the chip approximate size 200 um X 200 um Align the beam by using back reflected light from a glass coverslip First roughly align the APD stage with the APD off such that the beam is close to the chip cover up the chip with an index card to be safe Again if working in the IR this is best to do with a visible laser that is closely aligned to the IR beam path so that the IR does not have to be visualized Once the alignment is close switch to the IR source attenuate the laser turn on the APD and finish aligning ATTENUATE THE LASER LIGHT OD gt 8 DO NOT ALLOW COUNTS ON APD TO EXCEED 10 s 147 Both Perkin Elmer PE and MPD current produce APD s The PE AP
121. o reposition the sample if necessary Step 5 Place in 90 C over for 30min to bake out remaining water Sample is now covered in thin layer of PMMA and can be removed from silicon piece 43 Figure 2 8 Schematic of FIB milling A beam of ions is focused onto the surface and material is ablated away Notice that Gallium ions red circles become implanted deep within the sample Alternatively if a gas is introduced into the system such as a platinum precursor gas the ions can act to deposit platinum instead of ablate the surface This allows for controlled deposition of a metal or dielectric but there will still be significant gallium implantation Figure from R 45 Figure 2 9 Schematic for typical AFM experiment A cantilever with a sharp AFM tip is scanned over a sample surface Nanometer scale tip deflections from the sample surface are measured by reflecting a laser off of the back of the AFM tip and onto a quadrant photodiode which senses different intensities based on the tip deflection Figure from Ref 47 Figure 2 10 Schematic of a typical ANSOM experiment A metal coated AFM tip is excited with light The light is concentrated down to 10nm due to the plasmon resonance of the structure which means the resolution of the xviii imaging system is also 10nm Emission from the sample is collected back through an objective into a standard confocal emission pathway 49 Figure 3 1 Enhanced fluorescence experimental outline a Sc
122. o enhance fluorescence so the FIB approach should be avoided 5 5 Conclusions In summary two methods were developed to fabricate bowtie nanoantennas onto AFM tips An E beam lithography approach developed by Arvind Sundaramurthy was initially attempted because E beam bowties have shown remarkable ability to enhance fluorescence of single molecules Ultimately this technique could not be reproduced and thus failed to produce any tips for testing A simpler strategy involving FIB milling was then attempted and bowtie nanoantennas were successfully fabricated onto AFM tips Unfortunately these bowties did not enhance fluorescence of TPQDI and so they were not usable as ANSOM tips A highly resonant bowtie nanoantenna would still be very useful ANSOM tip but as this chapter demonstrates it is a difficult fabrication problem to solve 106 References 1 Synge E H A suggested method for extending the microscopic resolution into the ultramicroscopic range Philosopical Magazine 6 356 1928 2 Pohl D W Denk W amp Lanz M Optical Stethoscopy Image Recording with resolution Lambda 20 App Phys Lett 44 651 1984 3 Lewis A Isaacson M Harootunian A amp Muray A Development of a 500A Spatial Resolution Microscope I Light is Efficiently Transmitted Through Lambda 16 Diameter Apertures Ultramicroscopy 13 227 1983 4 Zenhausern F Martin Y amp Wickramasinghe H K Scanning interferometric apertur
123. olecule Fluorescence Enhancements Produced by a Gold Bowtie Nanoantenna Nat Photonics 3 654 2009 87 2 Eid J amp et al Real Time DNA Sequencing from Single Polymerase Molecules Science 323 133 138 2009 3 Levene M J et al Zero Mode Waveguides for Single Molecule Analysis at High Concentrations Science 299 682 686 2003 4 Uemura S et al Real time tRNA Transit on Single Translating Ribosomes at Codon Resolution Nature 464 1012 1017 2010 5 Fromm D P Sundaramurthy A Schuck P J Kino G S amp Moerner W E Gap dependent optical coupling of single bowtie nanoantennas resonant in the visible Nano Lett 4 957 961 2004 6 Fromm D P et al Exploring the chemical enhancement for surface enhanced Raman scattering with Au bowtie nanoantennas J Chem Phys 124 061101 2006 7 Fromm D Improving the Size Mismatch Between Light and Single Molecules using Metallic Nanostructures Stanford Ph D Thesis 2005 8 Tam F Goodrich G P Johnson B R amp Halas N J Plasmonic enhancement of molecular fluorescence Nano Lett 7 496 501 2007 9 Bakker R M et al Enhanced localized fluorescence in plasmonic nanoantennae Appl Phys Lett 92 043101 2008 10 Ringler M et al Shaping emission spectra of fluorescent molecules with single plasmonic nanoresonators Phys Rev Lett 100 203002 2008 11 Biteen J S Lewis N S Atwater H A Mertens H amp Polman
124. olecule tends to spend in a diffraction limited volume of water and Taarkis 4 us For IR800cw Tp is 274 us and Tyark is 1 us 81 Normalized GG 3 E 1 10 10 10 10 10 10 t ms Figure 4 6 FCS of 10pM ICG in water blue and 10pM IR800cw in ethanol red without bowtie nanoantenna Fits to Eqn 4 2 are shown as dashed lines 4 6 Bowtie Enhanced FCS Time traces of the fluorescence emission intensity for single bowties immersed ina 1 uM solution of IR800cw in ethanol and ICG in water are shown in Figure 4 7 In both cases flashes of fluorescence can be seen whenever a molecule enters the enhanced field region of the bowtie nanoantenna and until the molecule eventually photobleaches No single molecule fluorescence flashing events are measured in the absence of the bowtie nanoantennas at 1uM concentrations of either dye as is expected since with large N the bursts cannot be observed and the contrast in the autocorrelation disappears see below Notice that the contrast between single enhanced molecules and background is much higher for ICG than for IR800cw This difference supports the conclusion that ICG is a better molecule for bowtie FCS than IR800cw since it has a lower intrinsic QE and hence a higher bowtie induced fluorescence enhancement 82 a 300 200 100 Counts ms b 1000 500 Counts ms 1 2 3 4 5 Time s Figure 4 7 a Fluorescence time trace binned to 1ms for a bowtie immersed in 1u
125. omic Force Microscopy 46 2 5 3 Apertureless Scanning Near field Optical Microscope Setup 47 Chapter 3 Large Single Molecule Fluorescence Enhancements Produced by a Bowtie Nanoantenna 50 3 1 Introduction 52 3 2 Experimental Schematic 53 3 3 3 4 3 5 3 6 3 7 3 8 Confocal Imaging of Unenhanced Single Molecules Single Molecule Fluorescence Measurements on Bowtie Nanoantennas Finite Difference Time Domain Simulations Excited State Lifetime Measurements Excitation Polarization Dependence Conclusions Chapter 4 Fluorescence Correlation Spectroscopy at High Concentrations using Gold Bowtie Nanoantennas 4 1 4 2 4 3 4 4 4 5 4 6 4 7 Introduction Experimental Schematic Bulk Bowtie Enhanced Fluorescence of Molecules in Solution Emission Spectra of Bowtie enhanced Fluorescence FCS of Low Concentration Dye Solutions Bowtie Enhanced FCS Conclusions 55 57 60 67 70 71 73 73 75 78 80 82 84 89 Chapter 5 Toward Bowtie Nanoantennas as Apertureless Scanning Near field Probes 5 1 Introduction 5 2 Initial Preparation of AFM Tip 5 3 E beam Lithography Approach 5 3 1 FIB milled Alignment Marks 5 3 2 Locating Alignment Marks 5 3 3 Chrome Etch 5 3 4 Float Coating of E beam Resist 5 3 5 Chrome Deposition Xi 90 91 92 93 93 94 95 95 98 5 3 6 Standard E beam Lithography Steps 5 3 7 Liftoff 5 3 8 E beam Fabrication Con
126. onance and electric field enhancement were unknown Therefore in addition to the BOAT fabrication bowtie nanoantennas were FIB milled on flat quartz substrates Figure 5 10a for scattering studies Due to the optics used for scattering studies in this thesis see Chapter 2 for optical setup it is only possible to measure the resonance for bowties on flat substrates The scattering spectra for two bowties with 25nm gap sizes fabricated with E beam lithography or FIB milling are compared in Figure 5 10b Notice that the peak of the resonance is approximately the same in both cases but the width of the resonance is much broader for the FIB bowties This broadening may indicate that while the resonance is located in the same position the enhanced fields are lower for the FIB fabricated bowtie 0 6 0 4 0 2 Normalized Scattering 550 600 650 700 750 800 850 900 950 Wavelength Figure 5 10 Scattering study of FIB milled bowties a SEM of FIB bowtie nanoantenna on a flat quartz substrate with 20nm gap b Comparison between scattering spectra for E beam and FIB fabricated bowties on quartz substrates with similar gap sizes 5 4 5 Optical Results from FIB Bowties on AFM tips Ultimately the best test for whether the FIB bowtie nanoantenna is useful is to see if it enhances the fluorescence of the fluorophore TPQDI since the application for these tips is apertureless near field imaging of Raman active or fluorescent
127. onfocal microscope approximately diffraction limited collection and spectrometer from the same photonic crystal cavity in a and b after molecules are deposited on cavity X polarized emission is shown in blue Y polarized emission is shown in red Inset Fluorescence measurements of fundamental cavity mode black box Line indicates Lorentzian fit with Q 10 000 d Quality factors measured for high Q cavity mode from reflectivity open circles before molecule deposition and fluorescence after molecule deposition for structures with lattice constant a and hole radius r a tuned so that the fundamental cavity resonance shifts across the fluorescence spectrum of the molecule Blue open 114 circles indicate reflectivity measurements for the cavities that were also measured in fluorescence blue closed circles After Ref 1 6 4 Fluorophore Cavity Coupled Fluorescence Emission Spectra After characterizing the structure initially a molecule doped polymer film is float coated on top of the entire sample We measure fluorescence from the molecule Figure 6 2c using a 633 nm helium neon excitation laser in a confocal microscope setup When measuring fluorescence from the cavity region collected from an approximately diffraction limited area of a photonic crystal cavity we observe sharp polarized resonances identical to those in our reflectivity measurements demonstrating the molecules are coupled to the cavity modes From confocal images
128. ongersma the final reader on my committee is an expert in the area of plasmonics and taught an excellent class early in my career as a graduate student This class fostered my early love of plasmonics My graduate career began in the Moerner lab under the guidance of Dr Dave Fromm who taught me the basics of plasmonics and optical microscopy knowledge that significantly aided my early development as a scientist Dr Jim Schuck a postdoc when I joined the lab also guided me and has even been a helpful resource after finishing his work in the Moerner lab and moving on to LBNL The last member of the early bowtie team was Arvind Sundaramurthy who taught me a great vi deal about nanofabrication of bowtie nanoantennas Frank Jackel later joined the bowtie team and I very much appreciated his help and guidance through the middle portion of my graduate career I have had a number of collaborators throughout my time at Stanford The most important collaborators for the work contained in this thesis are Dr Zongfu Yu of Prof Shanhui Fan s lab Kelley Rivoire of Prof Jelena Vuckovic s lab and Dr Yuri Avlasevich of Prof Klaus Miillen s lab Zongfu is an amazing theory collaborator for the bowtie work and has helped immensely in understanding the effects I measured experimentally Dr Avlasevich was kind enough to share the DNQDI and TPQDI molecules which made much of this work possible Finally Kelley Rivoire is an expert in photonic
129. ope for single molecule imaging and spectroscopy The main consideration for imaging single molecules at room temperatures is that they tend to photobleach after emitting 27 10 10 photons 6 It is therefore important to maximize the signal to noise ration SNR ee Nie B D SNR 2 1 where Naet is the number of fluorescence photons collected on the detector B is the number of background photons from the sample and D is the number of dark counts produced by the detector all in a fixed integration time To maximize the SNR Naet should be as high as possible which is accomplished by collecting every fluorescence photon possible and minimizing loss in the emission pathway as well as minimizing B and D which involves using non fluorescent substrates selecting the detector from stray light and shielding the detector from stray light Interfering signals can be produced by many parts of the setup First of all the room lights add to this signal so the first consideration in building any single molecule microscope is to ensure the emission pathway is well shielded for room lights Next dark counts from the detector typically a silicon avalanche photodiode APD are always present The dark counts from an APD can be as low as 50 counts s but only if the detector is initially very good and if the detector is treated well The detector should never be turned on unless the user is sure the number of photons it
130. or beamsplitter 2 Center beam position monitor on CCD again Assume beam is only too far left perfectly centered in vertical axis Using mirror M4 Figure C 3 first move beam away from the optimal position so that the beam intensity becomes uneven You want to move it a maximum of one half the defocused spot size Then center the beam up with mirror M5 Figure C 3 The beam will have moved closer to the center 144 of the optical axis Repeat as necessary until the axis is completely centered Then do the orthogonal optical axis 3 Check the centering of the beam by focusing through the focus center The beam position should remain perfectly centered and concentric If not the beam is tilted and step 2 should be repeated C 2 Output Optics Figure C 3 and Figure C 5 show the important output optics for a confocal setup This section will review the optics necessary for a confocal setup in particular alignment of the confocal pinhole choosing the right emission filters and aligning an avalanche photodiode APD 145 Figure C 5 Confocal microscope output optics Note the pinhole PH placed at the microscope image plane the collimating lens CL the 90 reflector flips in and out the focusing lens FL for the APD detector C 2 1 Confocal Pinhole The diameter of the confocal pinhole Pinhole in Figure C 3 determines the amount of z confinement a smaller pinhole rejects more background i e scatter
131. or your feature e Adjust contrast brightness focus and coincidence you may have to work very quickly if you use a large aperture e Freeze the image or grab a 1 Ion beam frame 2 Setup Pattern e Select the pattern shape s from the icons along the menu bar and draw what you want to mill If complicated patterns are required contact experienced users for ways to draw patterns offline in Matlab e On the work page select serial sequential or parallel milling e Make sure ion beam is selected as the beam you will use and is the primary beam e Choose the appropriate material resource file The si mtr file can generally be used regardless the actual material The actual depth will need to be calibrated for any material e Adjust the milling dimensions as required the computer will determine the approximate milling time e Grab a 1 I frame again to check the setup e Click start stop patterning icon to begin the mill You can take 1 E and 1 I frames as you mill B 4 Pt deposition with the lon Beam 1 Setup Ion Beam for imaging as described for milling 134 2 Turn on Pt heater and insert GIS needle e Turn on the Pt heater and allow it to warm up indicator becomes red before use e Make sure you are at eucentric height and insert the Gas Injection System GIS Pt needle If you are above eucentric height then the needle will hit the sample Inserting the needle may cause a small shift in the image both
132. ormance Currently a liquid nitrogen cooled slow scan camera is used This camera is cooled to 120 C by filling with 152 liquid nitrogen in order to minimize dark counts During cool down of this camera ensure that the software controlling the camera Winspec is running the shutter to the camera is closed and the camera is actively collecting data This ensures that the camera is not damaged during cooling The read out time of this camera is slow compared to the cameras for wide field imagine in the lab requiring 0 37 sec frame While other cameras are faster and can be used they tend to have higher dark counts Be careful when using CCDs with high gains making sure that the gain is linear across the spectral range used this is especially problematic in scattering experiments covering a large spectral range C 3 10 Final alignment With the CCD camera attached use the back reflected laser beam attenuated to safe levels for the detector to maximize signal on the camera You may have to remove the slit to see maximum signal levels and then put the slit back in and realign adjusting i the input mirror angle and position ii the focusing camera lens iii camera angle and focal position iv grating angle to image the desired spectral range v grating mounting bracket vertical adjustment should be adjusted to ensure that the beam is focused on the center of the CCD chip C 3 11 Final comments The alignment o
133. ow instead of performing bulk experiments and only measuring average behavior it is possible to measure the behavior of every target molecule and understand heterogeneity in behavior In order to detect a single molecule s fluorescence a very good fluorophore is necessary A good fluorophore has several properties First it absorbs light well which corresponds to a high absorption cross section Second a good fluorophore tends to emit radiatively instead of non radiatively which is reflected in the fluorescence quantum efficiency QE The QE is the probability that a molecule will emit a fluorescence photon per photon absorbed and is given by 14 n 1 7 Fy 4 fe where 7 is the QE while y and y are the radiative and non radiative decay rates respectively If the QE is 100 then the molecule always emits a photon and never decays non radiatively making it an ideal fluorophore for most applications Finally a good fluorophore must be stable and capable of emitting many photons before photobleaching Photobeaching refers to any change in the molecule that occurs during illumination typically involving a chemical reaction that alters the molecule s identity causing it to no longer fluoresce One common pathway for a molecule to photobleach can occur when the molecule enters a triplet state a forbidden transition from the singlet state that occurs with low probability The triplet state represents an intermediate from which
134. ows for sub diffraction limited imaging This section will describe the basics of AFM and ANSOM for use with bowtie nanoantennas fabricated onto AFM tips 2 5 2 Atomic Force Microscopy Atomic Force Microscopy AFM was invented by Binning Quate and Gerber in 1986 as a way to study the topography of any surface non destructively For a relatively recent technique AFM has become central to characterizing many nanoscale structures and many modalities have beed discussed such as working in fluids sensing magnetic fields sensing charges and deposition by dip pen lithography to name a few This section only briefly discusses the main aspects of AFM necessary to understand ANSOM In the simplest AFM experiment running in contact mode a sharp tip is brought into contact with the sample surface Figure 2 9 The tip is then dragged across the sample surface either by scanning the tip or the sample in order to track the topography of the surface Tip deflections are measured by bouncing a laser off of 45 the back of the cantilever and onto a quadrant photodiode By subtracting the signal from the bottom half of the photodiode from the top half the tip deflection can routinely be measured to nm accuracy in the Z direction X and Y resolution is dependent on the sharpness of the AFM tip and is often limited to 10nm Recently however it was shown that AFM is sensitive enough to resolve the atoms of a single pentacene molecule mak
135. p Fromm D P Methods of Single Molecule Fluorescence Spectroscopy and Microscopy Rev Sci Instrum 74 3597 3619 2003 5 Lounis B L Deich J Rosell F I Boxer S G amp Moerner W E Photophysics of DsRed a red fluorescent protein from the ensemble to the single molecule level J Phys Chem B 105 5048 5054 2001 6 Soper S A Nutter H L Keller R A Davis L M amp Shera E B The Photophysical Constants of Several Fluorescent Dyes Pertaining to Ultrasensitive Fluorescence Spectroscopy Photochem Photobiol 57 972 977 1993 48 7 Lakowicz J R in Principles of fluorescence spectroscopy 954 Springer Science New York 2006 8 http www fibics com fib tutorials introduction focused ion beam systems 4 9 Zhou H et al Lithographically Defined Nano and Micro Sensors using Float Coating of Resist and Electron Beam Lithography J Vac Sci Technol B 18 3594 2000 10 http sahussain wordpress com 2007 1 1 03 can we see the atomic dimension 11 Farahani J N et al Bow tie optical antenna probes for single emitter scanning near field optical microscopy Nanotech 18 125506 125510 2007 12 Farahani J N Pohl D W Eisler H amp Hecht B Single Quantum Dot Coupled to a Scanning Optical Antenna A Tunable Superemitter Phys Rev Lett 95 017402 1 017402 4 2005 13 Matteo J A et al Spectral analysis of strongly enhanced visible light transmission through s
136. power through the slit ideally gt 95 transmission is possible C 3 5 Concave mirror This is a collimating mirror CM in Figure C 6 and directs a collimated beam onto the grating 150 C 3 6 Grating The grating G in Figure C 6 is the dispersive element of the monochromator The grating is mounted with 5 minute epoxy onto a mounting bracket This is a kinematic mount that can be adjusted so that the dispersed beam hits the center of the CCD camera vertical adjustment Further two specs are important for the grating the blaze the peak operating wavelength and the groove spacing more grooves disperse light more quickly This is a ruled grating and replacements may still be vailable from Genesis Labs 1 970 241 0889 To ensure optimal throughput it is important to operate near the blaze wavelength the general rule is to work between 2 3 and 3 2 of the blaze wavelength The number of grooved rulings describes a tradeoff between resolution and the spectral range provided on the detector Using the LN cooled Princeton Instruments camera 512 pixels pixel 18 um the following is observed Pitch Blaze Typical Approx Used for Spectral range Resolution 150 grv mm 500 nm 400 1000 nm 4nm scattering 600 nm 300 400 500 800nm 2 fluorescence 300 nm 600 1000 850 1000 nm 1 Raman 150 nm Table 1 These are the gratings currently available for the Monospec18 spectrometer Th
137. r 40 s e Deposit 4 nm titanium and 20 nm gold f Liftoff resist by various methods described below 5 3 7 Liftoff The final step in this process is liftoff Figure 5 5f which never fully worked satisfactorily Normally liftoff is performed by placing the substrate in acetone and sonicating briefly but sonication cannot be performed on AFM tips since the cantilevers break Sometimes liftoff can be performed by simply soaking the substrate in acetone but this never worked on the AFM tips I also tried soaking in heated acetone heated PG remover as well as oxygen plasma etching but was never able to completely remove the non bowtie shaped metal Figure 5 6a shows one tip after development and metal deposition but before lift off For this tip I wrote an array of bowties with one targeted to the apex of the tip red lines show this targeting was successful Figure 5 6b shows this tip after liftoff by soaking in acetone Most 98 of the resist is removed except for the resist near the tip itself the most important area A bowtie that was written at the base of the tip is shown in Figure 5 6c This SEM shows that the array of bowties in the lifted off region were written and developed successfully This particular bowtie looks jagged and uneven because when writing the bowtie the focus was set for the apex of the tip so this bowtie was written 3 um out of focus Since this bowtie was written out of focus and yet lifted off corr
138. r intensity increases Fits to each curve using equation 4 1 are plotted with dashed black lines The FCS curve for 86 a 10pM solution of IR800cw in the absence of a bowtie nanoantenna with 1 9MW cm laser intensity is plotted in solid black c e Fit parameters used for fit curves shown in b using equation 4 1 4 7 Conclusions Bowtie FCS has been shown as a viable alternative to zero mode waveguides when studying molecules immobilized on the surface of a substrate at high uM concentrations As a proof of principle bowtie FCS successfully measured the photobleaching turn off times of high 1uM concentration of ICG in water as a function of laser intensity While this method is currently limited to molecules that linger in the enhanced region many experiments of this type are possible For instance an enzyme could be attached to the surface near the bowtie and whenever it acts on a fluorescently labeled substrate molecule at uM concentrations then the molecule will be held near the bowtie for an extended period of time allowing for easy measurement In a similar fashion a biomolecule with a ligand binding site can be attached to the surface and then fluorescently labeled ligands which bind to the biomolecule can be easily detected and the unbinding times directly measured Acknowledgements I would like to thank Dr Zongfu Yu for his help in analyzing the acquired data References 1 Kinkhabwala A et al Large Single M
139. r z sectioning An emission filter ensures only fluorescence reaches the detector SMF s are delicate optical components so care should be taken when using them Do not kink the fibers at all because they will break They are not cables and must be treated with care the maximum diameter of curvature that one should impart ona SMF is 8 For excess fiber bundle it up and tie wrap it together tying it to the racks out of the way Also be sure the keep the ends capped when not in use in order to prevent dust from accumulating on the end faces After the beam exits the SMF it needs to be re collimated using a lens An objective is currently used L2 in Figure C 3 It is important for NA of the objective lens should match that of the SMF which is about 0 1 0 2 for most fibers The output beam should be perfectly collimated which should be checked by shooting the beam across several meters of free space mirrors help here and imaging the spot on a card or beam profiler Make sure there are no foci in the beam that the intensity profile is perfectly symmetrical if not the output of the fiber is tilted with respect to the objective and that it is Gaussian in shape It is helpful to take a beam stop and put an index card on it with a mark signifying the height of the fiber output which should be equal to the microscope input height for ease of alignment Use this beam stop to center the height of the output from the collimating ob
140. re reduced because the feature sizes are larger Small differences in cavity Q measured with reflectivity versus fluorescence Figure 6 2d are primarily due to fit error 6 5 Lithographically Defining Molecule Position over Photonic Crystal Cavity Since the molecules are doped into PMMA an E beam lithography resist it is straightforward to selectivity expose and develop the polymer film using E beam lithography so that molecule doped PMMA remains only at the location of the photonic crystal cavity Figure 6 1c While float coating deposits resist uniformly over a small region PMMA thickness variations were observed from one coating to the next so electron beam doses were varied for different cavities on one sample Figure 6 3a shows a scanning confocal image of photoluminescence from a photonic crystal cavity coated with DNQDI doped PMMA before patterning The fluorescence is flat to within 3 5 with slightly more emission from the cavity region likely a result of enhanced outcoupling from molecules coupled to the cavity mode 116 Using the Raith150 E beam lithograpy tool it is possible to expose an area of resist based upon alignment marks located on the sample During the initial electron beam lithography step that defines the photonic crystals alignment marks were also written After photonic crystal fabrication and float coating of dye doped resist the sample is then reloaded into the Raith150 Depending upon the size of the a
141. respect to the long axis of the bowtie Due to differences in dichroic reflectivity the parallel orientation data were taken at 1 2 kW cm while the perpendicular data were taken at 5 9 kW cm but the parallel data is scaled here to 5 9 kW cm for easy comparison b Red Blue SM TPQDI excited with light polarized parallel perpendicular to the long axis of the bowtie Black dashed lines connect measurements from the same molecule After Ref 3 8 Conclusions In this work single molecules of TPQDI were used as probes of fr near gold bowtie nanoantennas Using the dominant emission that arises from the most highly 69 enhanced molecule fluorescence brightness enhancements of up to 1340 were observed in agreement with electromagnetic calculations of radiative nonradiative and electromagnetic intensity enhancements SM lifetimes show additional information about the decay processes for each molecule independent of the local optical intensity enhancement The bowtie nanoantenna provides a useful balance between enhancement and loss for SM emission applications In particular emission decay times as short as 10 ps were observed which means that a high emission rate room temperature single photon source can be fabricated using a SM in a bowtie gap References 1 Chance R R Prock A amp Silbey R J Molecular Fluorescence and Energy Transfer Near Interfaces Adv Chem Phys 37 1 65 1978 2 Muhlschlegel P
142. rix instruction book gives a great walk through on this process in p 7 2 through 7 18 I will not repeat them here except to say that it s important to set up the grating that you scan in AFM mode quite square with respect to X and Y else there will be a lot of cross talk Finally make sure to note which stage files are changed discussed in the instructions Before doing any recalibration it is important to save a copy of the stage calibration file in a safe spot so that if you need to go back to it you can C 5 3 Hardware signals in out of ECU controller The AFM and NSOM signals go from the head unit through a thick multi pin cable and plug into a PC board that is mounted on the optical table This board then goes out to the respective inputs in the ECU controller The two APD channels are brought in through IN 1 and IN2 respectively These inputs are whatever you want provided they are voltages 10 to 10 V range To get the APD signals TTL or NIM pulses you must bring these pulses into a NI board that serves as an integrator i e how many pulses do you see within a 10 ms time The architecture of the controller unit is quite open and fairly intuitive it s simple to get signals in out of this box References 1 Fromm D Improving the Size Mismatch Between Light and Single Molecules using Metallic Nanostructures Stanford Ph D Thesis 2005 160 161
143. s Figure 5 12c while there is a definite enhancement seen for the apex of a sharpened gold AFM tip Figure 5 12d as expected from similar experiments in the literature SwoT sjuno SWOT sjuNO Figure 5 12 Fluorescence enhancement attempt with FIB bowtie on an AFM tip and sharpened gold AFM tip a schematic of the FIB bowtie AFM tip b Schematic of sharp gold coated AFM tip c FIB bowtie AFM tip was scanned over a bulk TPQDI in PMMA sample The sample remained fixed while the tip was scanned thus imaging the enhancement of fluorescence as a function of tip position When the bowtie is positioned over the objective focus the fluorescence is quenched Scale bar 1um d When a sharpened gold coated AFM tip is scanned over the sample an enhancement of fluorescence is measured Scale bar lum 105 Since the FIB bowtie actually quenched instead of enhanced the fluorescence of TPQDI this approach was abandoned Aluminum FIB bowtie antennas have been shown to enhance the fluorescence from colloidal quantum dots by a factor of s3 ee This enhancement is very low compared to the enhancements measured in this thesis for low quantum efficiency molecules coupled to E beam fabricated bowties 1300 and this difference is likely due to the gallium implantation into the substrate and metal that lowers the antenna efficiency in addition to the high QE of quantum dots Overall the gold FIB bowtie nanoantenna on an AFM tip does not appear t
144. s ICG in ethanol 4 4 Emission Spectra of Bowtie enhanced Fluorescence Fluorescence spectra were taken of both bulk and bowtie enhanced fluorescent molecules with the same imaging power and integration time Figure 4 4a This required use of the confocal fluorescence microscope and imaging of the emission from the sample on the entrance slit of a grating spectrometer with a CCD array detector at the exit slit The optical arrangement has been described in Ref 7 The measured spectra are typical for room temperature fluorescence measurements and do 78 not show sharp features typically associated with Raman transitions which rules out SERS effects As expected the fluorescence spectra taken in the presence of the bowtie nanoantenna have much higher signal In Figure 4 4b the bowtie and no bowtie spectra for IR800cw are normalized in order to show that the shape of the fluorescence spectrum does not markedly change with the presence of the bowtie nanoantenna In principle plasmonic antennas can change the fluorescence emission of molecules coupled to them but the bowtie s resonance is relatively broad and well matched to the molecules emission spectra so this does not occur a Bowtie 100nM IR800cw R No bowtie 100nM IR800cw Bowtie 1uM ICG 7 No bowtie 1uM ICG S o 5 800 830 860 890 920 Wavelength nm b g 1 Bowtie 100nM IR800cw S 08 Nobowtie 100nM IR800cw a0 o
145. s are saved into a format specific to the Topometrix program but this program should not be used for data analysis because it tends to crash Export the files to txt format and import them into Matlab for analysis C 5 Scanning stages C 5 1 Piezoelectric Scanner There are two stages for the Topometrix AFM system the tripod sample scanner and the tip scanner only used for AFM Typically the sample is scanned though the tip can be scanned as well for AFM only operation You CANNOT scan both at the same time This would require a second controller box an option if careful tip positioning is needed the ECU controller The tripod scanner is known as LX149707 tripod scanner 50 um range and the tip scanner is referenced in the software as X089704 Accurex scanner 100 um range There are dozens of other scanner files in the Veeco SPM Lab 6 02 software but only these are used However you can control other stages with this controller if you want i e the PI closed loop stage has a driver file this is just a note for future experiments 158 Both of these scanners are closed loop piezo stages meaning that the piezos are completely linear in their movement and that you can return to a known position clicking around on the screen This is done by a capacitive sensor that reads out the extension of the piezo stack A servo loop in the controller monitors this and adjusts the voltage to the piezo between 0 100 V to achieve t
146. s is not due to diffusion as in Fig 4 6 but instead the long time decay reports on photobleaching times at different molecule positions and orientations on the surface Since an enhanced molecule can be in a number of different positions and orientations and still be measured then a continuum of different photobleaching times underlies the FCS curve Photobleaching is often a Poisson process with exponential waiting time but here a distribution of characteristic times must be present This type of multi exponential behavior is commonly modeled with a stretched exponential E Therefore the bowtie FCS curves were fit with the following equation l TIT bhor G r e Clee 4 3 where N relates to the concentration Tphoto is the photobleaching time parameter and is the usual stretching parameter As usual when 1 the FCS curve is a single exponential but as f approaches zero the exponential is stretched and is representative of the sum of more and more exponentials The fits agree well with the data and are plotted as dashed lines in Figure 4 8b The extracted fit parameters are plotted in Figure 4 8c e as a function of pumping intensity with 95 confidence interval bootstrapped errors In particular notice that in Figure 4 8d as the excitation power increases the inverse of Tproro increases consistent with photobleaching behavior For the bowtie FCS curves p values between 0 15 to 0 32 are observed indicating that the FCS curves
147. single mode fiber SMF in Figure C 3 Laser couplers are available in the lab that have a built in lens L1 in Figure C 3 The core of a SMF 139 is small typically 1 5 3 um in diameter and the output should be perfectly Gaussian You can check this by imaging the output onto an index card or beam profiler If the beam is distorted then either the fiber is dirty or damaged The end facets of the SMF can be cleaned dragging a MeOH coated lens tissue over the surface of the fiber If one end of the fiber is slightly damaged you can still use it just make sure the perfect end is the fiber output to microscope Coupling efficiency will suffer but generally plenty of power is available for confocal experiments Finally the mode output of the fiber labeled on the fiber end is important to consider the IR fiber currently installed is good for 700 900 nm but has low transmission in the visible 785nm diode M1 laser M3 a w a Polarizer M4 785nm excitation filter OD wheel L4 800LP L3 Microscope APD Figure C 3 Schematic of typical optics used for a confocal microscope A single mode fiber SMF is used to produce a Gaussian beam profile for the excitation path followed by additional optics that control the beam s polarization power and spectrum The beam path is a confocal setup because the 140 emission pathway is focused through a pinhole allowing fo
148. ssword 2 Check Vacuum status e Check that the Vacuum and High Tension HT hardware buttons are lit This ensures that the vacuum s turbopump is operational 3 Turn on Ion Source e Check that the ion source is on button is colored yellow If it is off button colored gray turn on the ion source by clicking the ion source button to warm up Emission will fluctuate for a few minutes but should stabilize at 2 1 2 3 amps 129 e The extractor is always set to 12 00 kV but the suppressor can be changed by the user from 2150 to 2150V If the emission is not stable between 2 1 2 3 uamps slowly adjust the suppressor to compensate If the suppressor is at its maximum value and the current is still too low the source needs to be heated contact a trainer or qualified user to heat the source The FIB can still be used if the current is only slightly before 2 1 uamps but make sure to let someone know if needs to be heated 4 Load sample e Choose OM in detectors menu In RH start up window choose Vent then OK Venting takes 3 5 minutes until the front door of the microscope can be opened e When vented insert your sample a wear gloves b make sure set screw is not engaged when you place your sample in the mount Tighten the set screw gently barely finger tight DO NOT OVERTIGHTEN c adjust the top of the sample to 5 mm from the lens use eucentric height adjuster aka elep
149. t al Bow tie optical antenna probes for single emitter scanning near field optical microscopy Nanotech 18 125506 125510 2007 12 Tam F Goodrich G P Johnson B R amp Halas N J Plasmonic enhancement of molecular fluorescence Nano Lett 7 496 501 2007 13 Taminiau T H Stefani F D Segerink F B amp van Hulst N Optical Antennas Direct Single Molecule Emission Nat Phot 2 234 237 2008 14 Zhang J Fu Y Chowdhury M H amp Lakowicz J R Metal enhanced single molecule fluorescence on silver particle monomer and dimer Coupling effect between metal particles Nano Lett 7 2101 2107 2007 15 Bakker R M et al Enhanced localized fluorescence in plasmonic nanoantennae Appl Phys Lett 92 043101 2008 16 Muskens O L Giannini V Sanchez Gil J A amp Gomez Rivas J Strong Enhancement of the Radiative Decay Rate of Emitters by Single Plasmonic Nanoantennas Nanolett 7 2871 2875 2007 17 Brolo A G et al Surface plasmon quantum dot coupling from arrays of nanoholes J Phys Chem B 110 8307 8313 2006 18 Gerard D et al Nanoaperture enhanced fluorescence Towards higher detection rates with plasmonic metals Phys Rev B 77 045413 2008 19 Ringler M et al Shaping emission spectra of fluorescent molecules with single plasmonic nanoresonators Phys Rev Lett 100 203002 2008 71 20 Bek A et al Fluorescence Enhancement in Hot Spots of AFM Designed Gold
150. t by OD 6 So after the pinhole and collimating lens typically extra long pass or bandpass emission filters at least one but sometimes two are placed to ensure that only fluorescence photons reach the APD In addition the laser source should be filtered to ensure that it is only a single frequency any broadband emission to long wavelengths of the laser emission line represent serious interfering signals because they can mimic fluorescence Lasers in particular diodes often emit multiple frequencies so a filter should be placed in the excitation path to prevent leakage of any long wavelength light from the excitation laser directly into the emission pathway 29 All the above sources of noise B and D can be eliminated or minimized but shot noise on the detected molecule photons will always be present due to the particle nature of light Shot noise shows up as Tig scaling in Equation 2 1 if B and D are negligible Since shot noise can never be eliminated the SNR has a fundamental value dependent upon the number of photons collected Despite the limited number of total photons from single molecules and the limitations arising from the various contributions to the SNR single molecule imaging and spectroscopy is feasible in a confocal microscope if care is taken when building the optical setup 2 2 4 Time Correlated Single Photon Counting When the laser spot is placed on a single molecule confocal microscopy allows for t
151. ted 7 8 10 11 22 previously Electromagnetic simulations reveal that this is due to greatly 50 enhanced absorption and an increased radiative emission rate resulting in enhancement of intrinsic quantum efficiency calculated to be a factor of 9 times despite additional nonradiative Ohmic effects from currents induced in the metal Bowtie nanoantennas thus show great potential for high contrast selection of single nanoemitters 3 2 Experimental Schematic A single fluorescent molecule SM with transition dipole 7 acts as a nanoscale optical sensor of the local field E near a bowtie nanoantenna because its transition rate is proportional to Z E P while its emission can either couple to the far prop u p 23 24 field via the nanoantenna or quench via Ohmic losses Low quantum efficiency emitters have been noted to have much higher potential fluorescence brightness enhancements fr than high quantum efficiency emitters because their intrinsic quantum efficiency has greater potential to be improved by the antenna s presence 26 51 700 800 900 Wavelength nm Figure 3 1 Enhanced fluorescence experimental outline a Schematic of bowtie nanoantenna gold coated with TPQDI molecules black arrows in PMMA light blue on a transparent substrate b TPQDI molecular structure c SEM of Au bowtie nanoantenna bar 100 nm d FDTD calculation of local intensity enhancement bar 100 n
152. ted to a single molecule s dynamics and reveal that half of the fluorescence from this particular bowtie is due to a single molecule In other words the digital step like sudden drop near 293 s is an unambiguous signature that a single molecule photobleached and the step size shows its contribution to the total signal Spowrie While the exact position and orientation of this molecule is not known it is highly likely that the molecule is located fairly near the position of maximum field enhancement i e between the two triangle tips discussed further below The fluorescence enhancement factor fr for this SM was determined with the following formula S owtie Ea fe a 3 1 un max L bowtie where Spowrie and Phowtie are the SM fluorescence signal and laser excitation power used for Figure 3 3a while Sun max and P apply to Figure 3 2a At later times a different single molecule could often be observed to photobleach enabling measurement of its 55 fr factor and so on on the same bowtie In effect the single molecules randomly sample the possible enhancements that can occur for various positions and orientations near the bowtie and the distribution of values will be discussed below Figure 3 3 Measuring enhanced fluorescence from single molecules on bowtie nanoantennas a Confocal scan of 16 bowties coated with high concentration 1 000 molecules diffraction limited spot TPQDI in PMMA collected with 2 4 kW cm sc
153. tep in the E beam lithography process is to coat the tip in resist PMMA Unfortunately spin coating is not an option because it leaves an uneven layer of resist on an AFM tip see Chapter 2 for details so another method must be 94 used namely float coating Briefly this procedure outlined in Figure 5 3 consists of placing the AFM tip into a water bath and then putting one drop of 1 PMMA in toluene onto the surface of the water The toluene evaporates leaving behind a smooth layer of PMMA behind on the water surface since PMMA does not dissolve in water When the water is pipetted out the PMMA layer evenly coats the tip and is then baked at 90 C for 30 minutes to ensure all the water has been baked out a Place in Water Bath b Float Coat 1 PMMA c Remove Water and Bake in Toluene N ZAR Figure 5 3 Float coating of resist onto an AFM tip a Tip is placed in a water bath b 1 drop of a 1 PMMA in toluene solution is dropped onto the water s surface A thin layer of PMMA forms as the toluene evaporates c Water is pipetted out letting the resist gently rest upon the AFM tip The tip is baked at 90 C for 30 minutes to remove any remaining water There are two major problems with float coating the first of which is demonstrated in Figure 5 4 most of the time the AFM tip bends during float coating The bending can be as extreme as depicted in Figure 5 4 where the tip is no longer even pointing vertically just sl
154. than 50 nm 117 Counts Wavelength nm Figure 6 3 Aligning molecules to a photonic crystal cavity a Scanning confocal image of fluorescence from DNQDI doped PMMA float coated onto a photonic crystal membrane Pixel size is 200nm and scale bar indicates 2 um b Scanning confocal image of DNQDI fluorescence after E beam lithography is used to remove all molecules except for the ones coating the cavity region at the center The same imaging laser power as in a was used Pixel size is 80nm and scale bar indicates 2 um c Fluorescence spectrum from the fundamental mode of photonic crystal cavity after selective removal of molecules by E beam lithography d Atomic force microscopy image showing localization of DNQDI doped PMMA to the cavity region PMMA thickness is 12nm Scale bar indicates 500nm After Ref 6 6 Conclusions In conclusion we have demonstrated the coupling of fluorescent molecules to photonic crystal cavities with resonances in the far red and near infrared wavelengths and quality factors up to 12 000 By exposing and developing the molecule s polymer host using E beam lithography we have also been able to localize the molecules to the cavity region exclusively Our results show that molecules can be coupled to high 118 quality factor photonic crystal cavities as well as localized to the nanoscale cavity using standard lithographic techniques References 1 Noda S Fujita M amp
155. the 5000 8000X magnification range Focus well and the free working distance FWD bottom of the screen will now read the focal length of the objective lens which equals the true distance from the sample to the bottom of the lens Click OK on the pop up window to calibrate Z to FWD e At this point the SEM capabilities are set up for basic use Translate to your sample using the joystick mouse and or stage table reconfirm Z if necessary and you can perform any SEM imaging you need Set eucentric height e Move to desired location on sample and click Z FWD button e Eucentric height is about 5mm for this machine On the workpage raise the sample to 5mm and refocus if necessary e At 10 15KX place a recognizable feature on the center crosshair e Tilt the sample a few degrees and re center the feature vertically with respect to the screen using the Z knob on the stage door not using the joystick or mouse 131 Increment up a few more degrees and repeat Continue you can increase the size of the increments until you get to 52 perpendicular to the ion beam Check that the feature is vertically centered at 0 and 52 8 Obtain ion beam image adjust electron ion beam coincidence After setting the eucentric height and while still at 52 select a relatively non destructive ion beam aperture The 1pA aperture can make features as small as 30nm while the 10pA aperture is li
156. the beam on the left computer 125 e Start imaging on the right computer e Open control panel on right computer e Make sure that in detectors tab Signal A InLens e Move until centered on the cup black hole e Click Measure Current gt should be 0 01 nA for 10um aperture at 10kV Setup sample coordinate system e On left computer s microscope control box set mag to 1000X 100um make sure NOT to select 100x 1000um e Check database values click Set e Go to your sample by going to correct clip clip 1 6 e Start imaging move to lower left corner of coverslip e Set origin correction click Adjust e Set angle correction using 2 points on the bottom of the sample e Note that setting the origin and angle correction will help if you need to image your sample after measuring it optically otherwise it will be difficult to locate the bowties by SEM blindly Set up the electron beam Gun Align e Move to a spot near where we want to write but still 1mm away e Click Apertures Tab e Click emission mode near gun align button e Click gun align button e Go to imaging mode 4 e Open crosshairs center the crosshairs in the circle e Click normal mode Setup the electron beam Aperture Align e Focus sample and burn calibration spot e First attempt to focus on dirt on the sample If there is no dirt in the nearby area try burning a really long calibration dot by pressing the crosshair button with
157. the cavity These cavities have been used to demonstrate nanoscale on chip devices and to probe fundamental quantum interactions between light and matter Experiments in this regime however are limited by the precision with which cavity and emitters can be spatially aligned and by the spectral range of the emitters that can be coupled to the narrow cavity resonance Emitters are most often distributed randomly in the photonic crystal slab and spatial alignment to the photonic crystal cavity occurs by chance Recently several techniques have been developed to position emitters with respect to cavities these techniques rely primarily on either a mechanical transfer process to bring an emitter to the surface of the cavity or the fabrication of a cavity at the location of a previously detected emitter 8 Neither method is easily scalable to arrays of cavities and emitters or achievable with conventional semiconductor fabrication processes Here we demonstrate coupling of near IR fluorescent molecules to cavities with quality factors above 10 000 and show that we can selectively position these molecules on top of a nanocavity using conventional lithographic techniques 6 2 Sample Fabrication and Preparation Typically photonic crystal cavity resonances are fabricated too far in the infrared to overlap with a fluorophore s absorption and emission spectra but advances in the growth and lithography of gallium phosphide have allowed the fa
158. the molecule can return to the singlet ground state eventually However the triplet state is long lived and so there is time for the molecule to react with triplet oxygen and cause the molecule to no longer fluoresce photooxidation This pathway is just an example the exact photobleaching mechanism for any particular molecule would have to be studied in detail and is not the subject of this work A good fluorophore will emit 10 photons before eventual degradation 1 5 Fluorescence Correlation Spectroscopy 1 5 1 Motivation Using fluorescence correlation spectroscopy FCS to measure dynamics from a fluctuating fluorescence signal was first described in 1972 before the invention of single molecule spectroscopy It was eventually realized that the ability to measure 15 FCS allowed for very dilute solutions to be studied and fluorescence dynamics from a dL In this collection of many single molecule intensity bursts to be analyze technique a laser is tightly focused into a dilute solution of molecules Figure 1 10 A confocal fluorescence microscope then measures the bright flashes of fluorescence from small numbers or single molecules passing through the focused laser spot When these flashes of fluorescence are analyzed any process that changes the fluorescence on time scales from nanoseconds to the transit time of a few milliseconds can be measured The typical analysis involves autocorrelation of the emission signal whic
159. the single molecule s time delay histogram blue in Figure 3 8c Notice that this particular single molecule s time 65 delay histogram completely overlaps with the instrument response function IRF green in Figure 3 8c This overlap means that the molecule was emitting from the excited state too fast for the APD to resolve its lifetime If the lifetime is longer than 10 ps then deconvolution of the measured instrument response function IRF allowed extraction of the lifetime from the data a 150 100 Counts Counts 0 01 02 03 04 c Time Delay ns Normalized Counts 0 01 02 03 04 Time Delay ns Figure 3 8 Measuring excited state lifetime from a single molecule coupled to bowtie nanoantenna a Time trace of fluorescence from a single bowtie nanoantenna Black and red lines indicate times before and after one molecule photobleaches b Time delay histograms from time trace in a corresponding the before black and after red photobleaching step c Blue Normalized single molecule time delay histogram formed by subtracting the red from the black curves in b Green is the instrument response function The deconvolved lifetime for this curve was less than 10 ps the minimum value we were able to determine experimentally After Ref Compared to measurements of fr changes in tp only monitor changes in nonradiative and radiative processes but not changes in the absorption rate The 66
160. thout bowties Blue 1uM ICG in ethanol Red 1 uM ICG in water Black 1uM IR800cw in ethanol Green 1 uM IR800cw in water If photobleaching drop in signal is measured beyond the first 10ms bin then molecules must be sticking to the surface and cannot be replaced since molecules only remain in the focal volume for no more xxiii Figure 4 4 Figure 4 5 than Ims unless they are stuck to the surface Therefore the only solution that did not show sticking is ICG in ethanol 79 a Spectra integrated over 10s from a 100nM concentration solution of TR800cw in ethanol with blue and without red a bowtie present as well as spectra from a 1uM concentration solution of ICG in water with green and without black a bowtie present Notice that none of the spectra contain Raman peaks b Normalized spectra from 100nM IR800cw with blue and without red a bowtie present Notice that the shape of the spectrum does not change depending on the bowtie s presence or absence For both figures the laser filter cuts off emission 800nm and shorter causing aberrations in this spectral region particularly at 810 nm 81 In order to measure autocorrelations at short time scales the fluorescence emission is split onto two detectors using a cube 50 50 beam splitter 82 Figure 4 6 FCS of 10pM ICG in water blue and 10pM IR800cw in ethanol red without bowtie nanoantenna Fits to Eqn 4 2 are shown as dashed lines 83 Figure 4 7 a Fluoresc
161. tical waveguides and high Q cavities by creating defects in the periodic photonic crystal lattice 1 3 2 Planar Photonic Crystal Cavities Planar photonic crystal cavities are fabricated out of thin membranes of semiconductor where the semiconductor is used as a high index dielectric in the wavelength regime of the semiconducting gap A hexagonal closed packed array of air holes is etched into this film as in Figure 1 8a Three holes in the middle of the photonic crystal are not etched this defect produces the cavity region When a photon enters the cavity region it is trapped in three dimensions In the Z dimension 10 light is confined due to total internal reflection off of the top and bottom surfaces of the thin membrane In the X and Y dimensions photonic crystal plane light is trapped due to Distributed Bragg Reflection DBR because the light energy falls within the energy band gap of the photonic crystal ____ b Figure 1 8 a Scanning electron microscope SEM image of a photonic crystal cavity b Electric field profile of photonic crystal cavity excited at resonance for the fundamental cavity mode After Ref The quality or Q factor is proportional to how long light is trapped inside the cavity before escaping and is simply f o 1 6 where fo and Af are the frequency center position and width of the resonance The Q for GaP photonic crystal cavities discussed in this thesis can be as high as 10
162. tips were chosen because Si3Nz is transparent to optical wavelengths and should not interfere with the bowtie s plasmon resonance unlike the more common Si AFM tips The AFM tip is coated with a 4nm thick layer of chrome by Tom Carver in the Ginzton cleanroom Figure 5 1la c This layer is necessary because the tip is insulating and cannot be imaged by an electron beam or milled by an ion beam without it The tip is then loaded into a FEI Strata 235DB dual beam FIB SEM in order to flatten it Figure 5 1b d The tip is sculpted to have a 500 nm x 500 nm flat area for lithography as well as a post bar of 30nm height This post is useful because ANSOM experiments typically require the AFM tip be brought into contact 91 with the surface a process that might damage the delicate bowtie nanoantenna Since the post is the tallest object on the tip it will be the first part to come into contact with the sample allowing the bowtie to be close to the sample and yet remain undamaged From here the E beam and FIB fabrication processes diverge a Deposit Chrome b Flatten Tip al c d 4A Figure 5 1 Initial flattening of an AFM tip using FIB a Schematic of AFM tip before FIB processing A thin 4 nm layer of chrome is deposited uniformly on the tip to prevent charging during FIB milling and SEM imaging b After FIB milling the tip is flattened except for a short 30 nm post which will be used to protect the eventually
163. to scattering of the Ga ions For this reason it is not recommended to use excessive FIB lithography for plasmonic antennas It has been shown that moderate enhancements of fluorescence can be achieved with FIB lithography antennas and that using FIB lithography to mill apertures is also reasonable since such a small dose of Ga is necessary but overall EBL should be used whenever possible Ga Gas Assisted Etching or Selective Deposition optional es AQ r ee o o gt amp o n e e ENNS Figure 2 8 Schematic of FIB milling A beam of ions is focused onto the surface and material is ablated away Notice that Gallium ions red circles become implanted deep within the sample Alternatively if a gas is introduced into the system such as a platinum precursor gas the ions can act to deposit platinum instead of ablate the surface This allows for controlled deposition of a metal or dielectric but there will still be significant gallium implantation Figure from 44 2 5 Apertureless Near Field Optical Microscopy 2 5 1 Introduction As discussed in Chapter 1 in apertureless near field optical microscopy ANSOM a plasmonic structure often a metal coated Atomic Force Microscopy AFM tip is optically excited while raster scanning in direct contact with a sample surface The plasmonic structure acts to concentrate light beyond the diffraction limit which ultimately all
164. tons program front panel 155 XXX1 Chapter 1 Introduction 1 1 Overview Richard Feynmann stated in 1959 that There s Plenty of Room at the Bottom predicting the recent explosion in nanotechnology research Nanotechnology is the study of materials systems at nanometer scale dimensions At first glance just making an object small may not seem interesting but a material often behaves differently on nanoscale dimensions than in bulk which has led to many interesting problems as well as new opportunities in miniaturization While nanotechnology has touched many areas of research this thesis concerns nano optics the study of light on the nanoscale by using two devices to control local electromagnetic fields the bowtie nanoantenna and the photonic crystal cavity In particular these two structures will be used to modify and control optical emission from nanoscale emitters 1 2 Optical Plasmonic Nanoantennas 1 2 1 Motivation Antennas are inescapable today They are used to receive and transmit radio and microwave range electromagnetic waves in devices such as cell phones televisions laptops and radios These antennas are capable of capturing and concentrating these fields efficiently to subwavelength dimensions and usually converting them to currents in an external circuit but in this thesis the concern is not with the external circuit but only with the local concentration of the electromagnetic field Notably scaling
165. ul ways to modify one of the most important diagnostic tools available to biologists today namely fluorescence Chapter 2 describes the optical and nanofabrication techniques used throughout the thesis In chapter 3 bowtie nanoantennas are used to enhance a single molecule fluorescence by a factor of 1 300 and electromagnetic simulation are used to understand this enhancement Chapter 4 extends the work of chapter 3 to include molecules in solution so that dynamics in the fluorescence signal can be measured Chapter 5 describes attempts to fabricate a bowtie nanoantenna onto an AFM tip so that the bowtie could eventually be positioned relative to a molecule In Chapter 6 fluorescent molecules are lithographically patterned onto photonic crystal cavities and the coupling between the molecules and cavity is measured Finally Appendices A B and C provide specific details on operating the Raith150 E beam lithography tool the FEI Strata Focused Ion Beam tool and aligning the home built confocal microscope used in this thesis Where collaborators are involved they are mentioned at the start of each chapter 20 References 1 Drude P Zur Elektronentheorie der Metalle Annalen der Physik 306 566 1900 2 Ashcroft N W amp Mermin N D in Solid State Physics ed Crane D G 826 Harcourt College Publishers New York NY 1976 3 Raether H in Surface Plasmons on Smooth and Rough Surfaces and on Grating Springer Verlag Berlin 19
166. ularly to protect the EBL machine from outgasing samples so the longer lower temperature bake is necessary if EBL is to be performed but may be skipped for other experiments Step 1 Place Step 2 Pipette sample into 1 drop of water bath PMMA solution Step 3 Allow Step 4 Pipette solvent to out water evaporate Step 5 Heat at 90 C for 30min Figure 2 7 Float Coating resist onto uneven substrate AFM tip Step 1 Place sample AFM tip pictured on top of a silicon piece in a water bath Step 2 Drop 1 drop of 1 PMMA in toluene onto the water bath using a 100uL pipette tip Step 3 Allow drop to disperse on top of water bath s surface for 5 minutes so that thin PMMA film forms and toluene evaporates completely Step 4 Pipette out water using 1000uL pipette tip Pipette out water far away from the sample and push the Silicon piece to reposition the sample if necessary Step 5 Place in 90 C over for 30min to bake out remaining water Sample is now covered in thin layer of PMMA and can be removed from silicon piece 42 2 4 4 Focused lon Beam Lithography The Focused Ion Beam FIB milling method was first developed in 19751 1 The FIB machine operates in much the same way as the E beam lithography systems but instead of shooting electrons at the surface it shoots Ga ions In each technique a focused beam of either ions or electrons is scanned over a sample to form a pattern In E beam lithography the electrons act to
167. ule light interactions can be altered using nanophotonic structures is an important field of study This thesis has shown that molecular fluorescence can be controlled using the bowtie nanoantenna as well as the photonic crystal cavity Chapter 3 discussed the exceptionally large enhancement of a single molecule s fluorescence caused by coupling the molecule to a lithographically fabricated gold bowtie nanoantenna When an initially low QE fluorescent molecule is positioned in the gap of a bowtie nanoantenna with its transition dipole moment oriented correctly its fluorescence is increased roughly 1 300 fold This enhancement was found to be due to an increase in the absorption of light by the molecule as well as a to shortening of the molecule s lifetime leading to an overall increase in the QE This result is useful in fields which require single molecule sensitivity in highly concentrated samples The above work is extended in Chapter 4 to show that enhancements of fluorescent molecules in liquid environments are possible It was found that single molecule FCS experiments could be performed on relatively high concentrations of molecules The only molecules that had measurable enhancement were ones that 121 adhered to the surface near the bowtie nanoantenna FCS experiments showed that the photobleaching time scaled inversely with the excitation intensity These results suggest that experiments on enzymes bound to a surface that act on fl
168. um as a sticking layer and 20nm Gold 5 Liftoff remaining PMMA by sonicating sample in acetone for a few seconds leaving behind bowtie nanoantennas 38 Figure 2 5 Process Flow for E beam Lithography of Bowtie Nanoantennas onto Figure 2 6 insulating substrate 1 Spin 50nm of PMMA using Laurel spincoater Deposit thin layer 4nm of Chrome to make sample temporarily conductive 2 Expose bowtie pattern into resist using Raith 150 E beam writer 3 Remove Chome in Chrome etch Cyantek CR 14 4 Develop exposed resist in 1 3 MIBK IPA solution for 35s and rinse in IPA for 40s 5 Deposit 4nm titanium as a sticking layer and 20nm gold 6 Liftoff remaining PMMA by sonicating sample in acetone for a few seconds leaving behind bowtie nanoantennas 40 A Spin coating resist onto a flat substrate yields a smooth even layer B Spin coating onto an uneven substrate leads to uneven coverage and buildup of resist at the base of features 41 Figure 2 7 Float Coating resist onto uneven substrate AFM tip Step 1 Place sample AFM tip pictured on top of a silicon piece in a water bath Step xvii 2 Drop 1 drop of 1 PMMA in toluene onto the water bath using a 100uL pipette tip Step 3 Allow drop to disperse on top of water bath s surface for 5 minutes so that thin PMMA film forms and toluene evaporates completely Step 4 Pipette out water using 1000uL pipette tip Pipette out water far away from the sample and push the Silicon piece t
169. unts and these background photons will overwhelm the fluorescence signal You will need to filter the laser to a narrow excitation band by buying the appropriate laser line pass filter such as 785nm excitation filter shown in Figure C 3 from Chroma Engineering or Omega Optical In addition if a fiber is used in the excitation beam path a laser line pass filter will reject Raman scattering 142 from the fiber which is particularly important when working far from the optimal fiber wavelength C 1 4 Polarization For any imaging system it is important to know and control the polarization of the excitation beam path The easiest polarization to work with is linear and can be achieved with a linear polarizer labeled polarizer in Figure C 3 Linearly polarized light that is polarized horizontally or vertically with respect to the table will bounce off of dichroic mirrors and preserve its polarization Half wave plates can be used to rotate the polarization and quarter wave plates can be used to create circularly polarized light but be careful dichroic mirrors will not preserve the polarization of circularly polarized light or light polarized along other axes Great care is needed to excite a sample with perfectly circularly polarized excitation C 1 5 Alignment into Microscope Two mirrors are used to couple light into the back of the microscope because the beam needs to enter in a perfectly straight line These two mirrors allow one to
170. uorescence photons per 10 ms for 79 kW cm excitation 10 4 2 5 0 0 2 4 Time s Counts Coun c 30 20 10 0 100 150 200 250 Single Molecule QDI Brightness Figure 3 2 Imaging unenhanced single molecule fluorescence a Confocal fluorescence scan of a low concentration lt 1 molecule diffraction limited spot sample of TPQDI in PMMA without bowtie nanoantennas scale bar 4 um b Fluorescence time trace of a single unenhanced TPQDI molecule aligned along the excitation polarization axis Data collected with 79 kW cm then scaled for direct comparison with Figure 3 3b c Histogram of unenhanced single molecule TPQDI brightness values from same low concentration TPQDI doped PMMA sample Data collected with 79 kW cm After Ref 54 3 4 Single Molecule Fluorescence Measurements on Bowtie Nanoantennas After fluorescence from unenhanced molecules was characterized we proceeded to measure fluorescence from molecules enhanced by gold bowtie nanoantennas Figure 3 3a shows a confocal scan from an array of 16 bowties coated with a high concentration of TPQDI in PMMA 1 000 molecules diffraction limited spot or 3 molecules 10 nm This image required a far lower pumping intensity than Fig 3 2a In order to see a SM out of the many covering the bowtie the fluorescence as a function of time is shown in Figure 3 3b Discrete blinking and eventual photobleaching of 50 of the total signal can be attribu
171. uorescently tagged substrate molecules at high concentrations are possible with bowtie nanoantennas Chapter 5 takes a different approach from the previous chapters Instead of randomly positioning molecules around the bowtie nanoantenna the goal of this work was to fabricate a bowtie nanoantenna on the end of a scannable AFM tip so that the bowtie could be positioned precisely above a molecule The fabrication of a resonant bowtie nanoantenna onto an AFM tip proved difficult to achieve with E beam lithography due to problems with liftoff Bowtie nanoantennas were eventually fabricated onto AFM tips using FIB milling but these antennas were not resonant and did not enhance the fluorescence of the molecule TPQDI likely due to gallium implantation from the FIB milling Chapters 3 and 4 show that if fabricated successfully a bowtie nanoantenna AFM tip could be a useful device for studying nanoscale emitters and enhancing their fluorescence thus extending the ANSOM and related near field imaging technique Finally Chapter 6 considers a very different nanophotonic structure the photonic crystal cavity A lithographic approach to positioning molecules on the cavity was developed whereby a dye doped polymer film is float coated onto the entire photonic crystal and E beam lithography development removes all dye doped resist everywhere except the photonic crystal cavity region These molecules were shown to be spectrally coupled to the photonic crystal
172. uture experiments where the sticky surface of our experiments is replaced with a surface of enzyme molecules or proteins with an affinity for fluorescent dyes In this way when biomolecules attached to the surface in the enhanced region of the bowtie nanoantenna incorporate 13 or bind fluorescently labeled substrate or ligand molecules flashes of fluorescence will be measured until the fluorophore either photobleaches is released by the enzyme or unbinds from the biomolecule This experiment is analogous to experiments described in Chapter 1 using zero mode waveguides to measure individual biological events such as DNA replication and RNA translation 4 2 Experimental Schematic The two fluorescent molecules used in this study are indocyanine green Sigma Aldrich ICG shown in Figure 4 1c and IR800cw carboxylate Li cor IR800cw shown in Figure 4 1d The absorption and emission spectra of the two dyes shown in Figure 4 1b overlap well with the plasmon resonance from a 10 nm gap Au bowtie nanoantenna so the bowtie could potentially enhance absorption and emission from both molecules From Table 1 the QE of ICG in water is 2 4 but increases to 14 in ethanol Considering the work in Chapter 3 this suggests that ICG will have higher fluorescence enhancement in water than in ethanol but moderate fluorescence enhancement should still be expected in ethanol Similarly since the QE of IR800cw in ethanol is 28 it should have
173. ve enhancement factor is 2 ae oe f epee 3 6 pe eC Y k oe as seen in Figure 3 6d e These two ratios can be used in equation 3 5 in order to estimate the change in QE f due to the bowtie s presence Figure 3 7a plots the 61 a 300 b 5 5 e 200 c v v E E v v g 100 2 lt lt C Lu Lu 0 700 800 900 1 000 Wavelength nm c d e Figure 3 6 Electromagnetic simulations of SM fluorescence near a gold bowtie nanoantenna a Spectrum of calculated electric field intensity enhancement versus wavelength in the center of a bowtie with 14 nm gap Inset the simulated structure side view consists of a SiO refractive index n 1 47 substrate a 50 nm layer of ITO n 2 and a 30 nm layer of PMMA n 1 49 The gold bowtie structure is 20 nm thick on a 4 nm layer of titanium b Radiative red and non radiative green enhancement factors along the center of the gap for wavelength 820 nm z measures the distance above the ITO PMMA interface Black dashed line shows the enhancement factor for electric field intensity at 780 nm Blue curve shows the fluorescence enhancement factor for quantum efficiency 2 5 molecules and grey dash line for quantum efficiency 100 molecules c e Illustration of the simulated structure side view section through the two triangle tips showing regions of fluorescence Blue radiative Red and non radiative Green enhancement factors for a molecu
174. will measure is lt 10 s These are the two sources of D dark counts Additional interfering sources are laser power dependent and contribute to background B whose Poisson variance represents a noise contribution The biggest source of background is often due to improper filtering of the laser Fluorescence 28 works by exciting a molecule at one wavelength and collecting light of a longer wavelength or lower energy which has been emitted by the molecule This effect called the molecular Stokes shift is a result of excited state relaxation after optical excitation of the molecule The absorption cross section of a single molecule at room temperature is very low 1 in every 10 photons will be absorbed by a molecule which means that there is a great deal of excitation laser signal that needs to be removed from the emission pathway By using a transparent sample such as molecules spun onto a glass coverslip only 4 of the excitation laser will reflect off of each glass surface into the emission pathway This 4 signal is still MUCH greater than the fluorescence signal however so appropriate choice of emission filters is important The first filter in the emission pathway also the last optic in the excitation pathway is a dichroic mirror beamsplitter Figure 2 1 A dichroic mirror reflects the excitation laser wavelength while transmitting the longer wavelength fluorescence signal but it does not attenuate the scattered pump ligh
175. wtie at its plasmon resonance b Map of IEI for gold bowtie nanoantenna pumped at 856nm from Ref By measuring the scattering of light experimental details discussed in Chapter 2 by these structures the plasmon resonance frequency can be experimentally determined It was found that small gap bowties 10nm have resonances around 820nm Figure 1 6a As the gap size increases the resonance first blue shifts due to larger overall antenna size before eventually red shifting toward the single triangle resonance These results illustrating the coupling between the two plasmons of the two triangle to produce a plasmon resonance for the overall bowtie agree well with finite difference time domain FDTD simulations of the plasmon resonance g 860 sj g 360 84 A G r p 0 B 2 820 820 S o qg gt E 800 4 e 800 7 gee a 780 4 780 a 760 E 2 o 9 g 760 740 a Pol E 740 Pol 720 ost 720 eve iva x 700 gt lt q x 700 gt lt a 680 a T r T T j 1 amp 680 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Separation Distance Triangle Height Separation Distance Triangle Height Figure 1 6 a Peak scattering wavelength versus bowtie gap size measured as gap size triangle height for long axis excitation polarization direction b Peak scattering wavelength versus bowtie gap size for short axis excitation polarization direction Figure from Ref 1
176. xposed resist in methyl isobutyl ketone MIBK which important for the development step 36 1 Spin Resist 2 Expose Resist Top View PMMA Onn PMMA PAZA nm ITO Quartz Quartz 3 Develop Resist 4 Deposit Metal 20nm Gold 4nm Titanium 5 Liftoff Figure 2 4 Process Flow for E beam Lithography of Bowtie Nanoantennas onto conductive substrate 1 Deposit 50nm thick layer of the transparent conductive oxide Indium Tin Oxide ITO onto a quartz coverslip Spin 50nm of PMMA using Laurel spincoater 2 Expose bowtie pattern into resist using Raith 150 E beam writer 3 Develop exposed resist in 1 3 MIBK IPA solution for 35s and rinse in IPA for 40s 4 Deposit 4nm Titanium as a sticking layer and 20nm Gold 5 Liftoff remaining PMMA by sonicating sample in acetone for a few seconds leaving behind bowtie nanoantennas 37 Step 3 is developing the exposed resist The sample is removed from the Raith 150 and developed by soaking it in a 1 3 solution of MIBK IPA for 35s and subsequently soaking in a pure IPA solution for 40s to prevent further development Since PMMA is a positive resist this step removes the exposed resist and leaves behind bowtie shaped holes in the resist Negative e beam resists do exist and they become less soluble during the development step and would leave behind bowtie shaped pillars of resist Positive resists in particular PMMA performs better for smaller feature sizes
177. y Scale bar 40 nm 9 a Schematic of electron and hole concentration due to excitation of the bowtie at its plasmon resonance b Map of IEI for gold bowtie nanoantenna pumped at 856nm from Ref 10 a Peak scattering wavelength versus bowtie gap size measured as gap size triangle height for long axis excitation polarization direction b Peak scattering wavelength versus bowtie gap size for short axis excitation polarization direction Figure from Ref 11 Measurement of enhanced IEI fields near a gold bowtie nanoantenna as a function of bowtie gap size using TPPL Figure from Ref s 12 Figure 1 8 a Scanning electron microscope SEM image of a photonic crystal cavity Figure 1 9 b Electric field profile of photonic crystal cavity excited at resonance for the fundamental cavity mode After Ref ay 14 a Simplified Jablonksi diagram for a typical fluorescence transition The emitter is pumped out of the ground state So and into vibrational XV sidebands of the electronic excited state S1 with rate yap blue arrow Internal conversion fast non radiative transitions allows the molecule to relax into the lowest level of the excited state At this point the molecule relaxes back to the ground state either radiatively with rate y red arrow or non radiatively with rate yn black wavy arrow Another internal conversion step black wavy arrow allows the molecule to relax to the lowest ground state level b Absorpt
178. y this process does not require ITO deposition so it can be performed on top of any flat substrate Note that after fabrication is complete the sample is insulating so scanning electron microscopy SEM imaging is not possible without depositing another thin metal layer 1 Spin Resist 2 Expose Resist Top View Deposit Chrome e 4nm Chrome PMMA 50 60nm PMMA PAVA 3 Etch Chrome 4 Develop Resist PMMA PAVA PMMA I E 5 Deposit Metal 6 Liftoff 20nm Gold PMMA 4nm Titanium Quartz Quartz Figure 2 5 Process Flow for E beam Lithography of Bowtie Nanoantennas onto insulating substrate 1 Spin 50nm of PMMA using Laurel spincoater Deposit thin layer 4nm of Chrome to make sample temporarily conductive 2 Expose bowtie pattern into resist using Raith 150 E beam writer 3 Remove Chome in Chrome etch Cyantek CR 14 4 Develop exposed resist in 1 3 MIBK IPA solution for 35s and rinse in IPA for 40s 5 Deposit 4nm titanium as a sticking layer and 20nm gold 39 6 Liftoff remaining PMMA by sonicating sample in acetone for a few seconds leaving behind bowtie nanoantennas E beam lithography is a very useful clean technique for fabricating nanostructures but it does take time to master fully Once a process has been developed such as the bowtie process described above bowtie nanoantennas can be routinely and precisely fabricated Focused Ion Beam FIB milling will be discussed later and should be considered
179. y for single molecule confocal microscopy b Schematic of emission pathway for confocal microscope showing the placement of a pinhole at the image plane which provides Z sectioning Once the sample is excited fluorescent molecules located anywhere inside the laser s beam path inside the sample will emit light and roughly half of this emission is collected back through the same objective Figure 2 1B Note that above and below the focal plane the optical intensity is lower but the number of emitting molecules is larger The fluorescence is collected and eventually focused through a confocal pinhole The pinhole is the most important part of any confocal setup because it provides the Z sectioning capability of the microscope By placing the pinhole at an image plane in the emission pathway only emission from molecules located at the Z focus of the microscope are focused through the pinhole and eventually collected on the detector In this way emission from a roughly 1um Z slice depending on the wavelength and size of the pinhole of the sample is imaged onto the point detector which is usually a photon counting avalanche photodiode In order to build up a full image the sample is simply raster scanned over the laser and the signal as a function of position is recorded thereby building the image pixel by pixel 2 2 3 Technical Issues for Single Molecule Imaging Special considerations are necessary when building a confocal microsc
180. zton cleanroom b The FIB is used to mill away gold in the pattern of a bowtie nanoantenna 5 4 3 Focused lon Beam Milling Finally the tip is loaded into the FIB in order to mill the bowtie shape into tip Figure 5 8b and Figure 5 9 Appendix B has detailed instructions for using the FIB 101 with particular information for writing patterns with small feature sizes such as the bowtie nanoantenna The key points to remember are to use a small beam aperture 1 pA current and to focus and stigmate the beam as well as possible At Stanford the FEI Strata 235DB tool is capable of making 30 nm feature sizes too large for the small gap bowties needed so the Raith ionLiNE capable of lt 10nm features was used instead in Dortmund Germany in collaboration with Dr Sven Bauerdick and Dr Jason Sanabia to mill the small gap bowtie in Figure 5 9 The vertical lines in Figure 5 9 are due to the beam pattern used in the pattern to mill the bowtie nanoantenna and do not correspond to gold metal remaining on the tip Figure 5 9 SEM of a FIB BOAT fabricated on Raith s ionLiNE FIB tool Scale bar is 200 nm 5 4 4 Scattering measurements on flat substrate FIB bowties Thus far in this thesis only E beam bowties have been characterized and shown to be useful for enhancing single molecule fluorescence For FIB fabricated bowties the effects of the ion beam particularly gallium implantation into the 102 substrate on the plasmon res

Coupling Fluorescent molecules to nanophotonic structures

Contents

Download Pdf Manuals

Related Search

Related Contents