Single Molecule Detection

Transcription

1 March 22, 2005

2 Contents 1 Introduction 4 2 Phenomenological Absorption Fluorescence Theory Electronic States Absorption Luminescence Vibrational Relaxation Fluorescence Internal Conversion Triplet State and Intersystem Crossing The Three Level System Experimental Microscope Setup Sample Preparation Basic Alignment of the Microscope Analysis Analysis Software Further reading 21 6 Appendix Laser

3 6.2 Wide field illumination lens Dichroic mirror Microscope objective Raman cut off filter CCD Camera DiI

4 1 Introduction Considerable progress in experimental instrumentation has moved the detection and investigation of single molecules from a scientists dream [1, 2, 3] to reality. In 1982, scanning tunnelling microscopy (STM) [4] and, a little later, atomic force microscopy (AFM) [5], opened the way for measurements on single atoms or molecules. An ultrafine tip, is used to explore the surface of the object, and through very short interactions, images with an atomic resolution and surface positioning of 0.1 nm can be created [6, 7]. Besides such sophisticated techniques optical detection techniques have been established and enabled studies of physical, chemical and even biological processes [8, 9] on a single molecular level. Despite a variety of techniques to push spatial resolution beyond the diffraction limit exist [10], typical single molecule experiments are performed within the diffraction limit provided by microscopy instruments. One of the preconditions to detect single molecules at room temperature - to make them spatially distinguishable - is achieved by microscopy, which provides the ability to look at extremely small (diffraction limited) spots. In confocal microscopy a small volume is probed and fluorescence images are constructed from raster scanning of the sample. This method has become one of the working horses [11] in single molecule detection. Alternatively, wide field instruments based on extremely sensitive array detectors such as Charge Coupled Devices (CCD) and Electron Multiplying Charge Coupled Devices (EMCCD) are especially suited to investigate a large number of molecules in parallel and to bridge single molecule data and ensemble measurements. The other precondition, to make the molecule detectable can be achieved by extremely sensitive detection techniques such as laser induced fluorescence. Most important is the choice of a chromophore with appropriate photo physical properties. The signal in a single molecule experiment stems from a repetitive excitation-emission cycle. For common fluorophores, the optical excitation-emission cycle rates are as large as 10 8 per second with a few µw of laser power focused to a diffraction limited spot. With the collection efficiencies of high aperture microscope objective lenses (up to 80%) the signals of single molecules can be detected with excellent signal to background ratio. 4

5 2 Phenomenological 2.1 Absorption The attenuation of a probe beam is described by the Lambert Beer law ( ) I(λ) log = ε(λ)cd (1) I 0 and depends on the concentration of the absorbing species (C), the molar extinction coefficient (ε) and the absorption length (d). ( ) I(λ) log = σ(λ)n d (2) I 0 in terms of a molecular absorption cross section with [ ] [ ] mol molecules N = l cm 3 (3) Example: Rhodamine 6G, ε max, σ max are values at the peak absorption wavelength [ ] ε max = l, σ max = cm 2 (4) molcm if we look at one molecule in a volume of 1µm in size, this gives a concentration of Molecules/cm 3 or mol/l and we have to detect an absorbance of 10 7! This can also be regarded as the ratio between the spot area and the absorption cross section. σ max [cm 2 ] ε max [mol 1 cm 1 ] Atoms Molecules Infrared Raman Table 1: Some typical absorption cross sections 5

6 2.2 Fluorescence Fluorescence is the phenomenon in which absorption of light of a given wavelength by a fluorescent molecule is followed by the emission of light at longer wavelengths. The distribution of wavelength-dependent excitation intensity that causes fluorescence is known as the fluorescence excitation spectrum, and the distribution of wavelength-dependent intensity of emitted energy is known as the fluorescence emission spectrum. Fluorescence detection has three major advantages over other light-based investigation methods: high sensitivity, high speed, and safety. Sensitivity is an important issue because the fluorescence signal is proportional to the absorption and thus the concentration of the substance being investigated. Whereas by means of absorbance measurements single molecules so far only have been detected at cryogenic temperatures, fluorescence techniques have opened up access to a variety of fluorophores at low temperatures as well as at room temperature. Because it is a non-invasive technique, fluorescence does not interfere (mechanically or electrically) with a sample. The excitation light levels required to generate a fluorescence signal are relatively low, reducing the effects of photo-bleaching. 3 Theory 3.1 Electronic States 1 When discussing the nature of electronic states, it is important to distinguish between the terms electronic state and electronic orbital. An orbital is defined as the volume element in which there is a high probability (99.9%) of finding an electron. It is calculated from a one-electron wave function and is assumed to be independent of all other electrons in the molecule. Electronic states, on the other hand, are concerned with the properties of all the electrons in all the orbitals. In other words, the wave function of an electronic state is a combination of the wave functions of each of the electrons in each of the orbitals of the molecule. Another important distinction is that between exited electronic states and 1 A good, sound introduction to fluorescence spectroscopy can be found in ref. [12]. 6

7 the transition state. Generally a transition state corresponds to a vibrationally exited ground state (i.e. ground state in a strained configuration), where as exited electronic states may contain no excess vibrational energy, but are still much higher in energy than the ground state. In fact a molecule in an exited state is best regarded as a completely new entity, only remotely related to the same molecule in the ground state. An exited state will have a completely different electron distribution from the ground state, a different geometry, and more than likely will undergo chemical reaction quite different from those of the ground state. Electronic states of organic molecules can be grouped into two broad categories, singlet states and triplet states. A singlet state is one in which all of the electrons in the molecule have their spins paired. Triplet states are those in which one set of electron spin have become unpaired. As will be seen later, triplet states and singlet states differ significantly in there properties as well as in there energies. A triplet state will always lie lower in energy than its corresponding singlet state. 3.2 Absorption Figure 1 show a partial energy diagram for a photoluminescence system. One should note that each of the electronic states (ground or exited) has a number of vibrational levels superimposed on it. The vibrational levels arise because a molecule in a given electronic state may absorb small increments of energy corresponding to changes in vibrational modes, although retaining the same electronic configuration. Another significant fact to be noted is the degree of overlap between the vibrational levels of exited states such as S 2 and S 1. By convention the singlet states should be stacked in a column while the triplet states are stacked in another vertical column displaced to the right of the singlet column. One should also notice the overlap between the vibrational levels of the triplet state T 1 and its corresponding singlet state S 1. The energy of a photon (E = hc/λ) required to produce a particular exited state is the difference in energy between that state and the ground state as shown in figure 1. Thus there is a range of wavelength that can lead to a transition between any two electronic states, which accounts for the fact that electronic absorption spectra generally occur as broad bands, rather than as single lines. 7

8 Radiationlees transitions Processes involving photons Figure 1: Partial energy diagram for a photoluminescent system (Jablonski diagram) As one might notice there is no arrow describing a transition from the ground state to the triplet state T 1. These transitions are forbidden and thus highly improbable. A good role of thumb is that singlet-triplet processes have a probability of about 10 6 that of a corresponding singlet-singlet or triplettriplet processes. 3.3 Luminescence Luminescence processes can be interpreted only in terms of the exited state from which luminescence emission occurs and its relationship to the ground state of the molecule. Although the simple picture of photon absorption 8

9 by a molecule subsequent by a reemission of a photon to give luminescence seems to be quite straightforward, there are nonradiative processes which precede and/or compete with photon emission. In the following section mean lifetimes of all processes will be stressed as they are important for determining the luminescence behavior of a molecule Vibrational Relaxation One my assume that all molecules are in the lowest vibrational level of the ground state at room temperature. The actual time required for a photon absorption, i.e. the time required for a molecule to go from one electronic state to another, is sec. This time is short relative to the time required for all other electronic processes and nuclear motion. This means that immediately after excitation a molecule has the same geometry and is in the same environment as it was in the ground state. In this situation it can do one of two things: emit a photon from the same vibrational level to which it was excited initially. undergo changes in vibrational levels prior to emission of radiation. Which of these two processes is dominant depends on the environment of the molecule. For an isolated molecule in the gas phase, the only way to lose vibrational energy is to emit an infrared photon, which is less probable than undergoing an electronic transition to return to the ground state. Therefore one tends to see photon emission from higher vibrational levels of exited states in gas phase spectra at low pressures. In a solution, however, thermal relaxation of a vibrationally excited molecule is quite rapid through transfer of excess vibrational energy from the solute molecule to the solvent. In fact, this process is so efficient that all the excess vibrational energy of the excited state is lost, this process occurs in to sec. This means that before an excited molecule in a solution can emit a photon, it will undergo vibrational relaxation, and therefore photon emission will always occur from the lowest vibrational level of an excited state. 9

10 3.3.2 Fluorescence Once a molecule arrives at the lowest vibrational level of an exited singlet state, it can do a number of things, one of which is to return to the ground state by photon emission. This process is called fluorescence. The lifetime of an excited singlet state is approximately 10 9 to 10 7 sec and therefore the decay time of fluorescence is of the same order of magnitude. If fluorescence is unperturbed by competing processes, the lifetime of fluorescence is the intrinsic lifetime of the excited singlet state. The quantum efficiency of fluorescence is defined as the fraction of molecules that will fluoresce. It should be noted that even though a quantum of radiation is emitted in fluorescence this quantum will be lower in energy on the average than the quantum absorbed by the molecule, due to vibrational relaxation (both after absorption and after emission). The change in photon energy causes a shift of the fluorescence spectrum to longer wavelength, relative to the absorption spectrum, this is referred to as the Stokes Shift. To summarize, the process of fluorescence consist of photon absorption by a molecule to go to an excited singlet state, relaxation from higher vibrational levels of that state to its lowest vibrational level, photon emission to a vibrationally exited level of the ground state, and again relaxation of the molecule to the lowest vibrational level of the ground state Internal Conversion In addition to fluorescence, one also encounters radiationless processes where molecules in an excited singlet state may return to the ground state without the emission of a photon, converting all the excitation energy into heat. The process called internal conversion, is not well understood and its efficiency is very difficult to measure. Generally internal conversion is an inefficient process and is probably only a small fraction of the total excitation energy in most molecules. This is particularly true in aromatic hydrocarbons, which we use in this experiment. So far the discussion was limited to the S 0 S 1 transitions, and the question might arise as to the nature of exited state processes if a molecule is excited to a higher singlet state such as S 2. In such cases the molecule will undergo vibrational relaxation as discussed before. The fate of the molecule as it reaches the zeroth vibrational level of S 2 depend of the energy separation between the exited singlet states. Generally the separation between excited 10

11 singlet states in an aromatic molecule is smaller than the energy separation between the lowest singlet state S 1 and the ground state S 0. This means that the lowest vibrational level of S 2 will overlap with higher vibrational levels of the S 1 state, that do not involve extremely large differences in configuration from the zeroth vibrational level of the S 1 state. This situation gives rise to a high degree of coupling between the vibrational levels of the S 2 and S 1 states, which provides an extremely efficient path for crossing from the S 2 state to the S 1 state. In fact this process is so efficient that the molecule undergoes internal conversion from the S 2 state to the lowest vibrational level of the S 1 state in about the same time that it requires to convert from an exited vibrational level of the S 1 state to its zeroth vibrational level (i.e to sec). Because of this situation one may formulate the following rule: a molecule may be considered to undergo internal conversion to the lowest vibrational level of its lowest excited singlet state in a time that is short, relative to photon emission, regardless of the singlet state to which it was excited initially Triplet State and Intersystem Crossing Although population of triplet states by direct absorption from the ground state is insignificant, a more efficient process exists for population of triplet states from the lowest excited singlet state in many molecules. This process is referred to as intersystem crossing, and is a spin-dependent internal conversion process. As singlet-triplet processes are generally less probable than singlet-singlet processes, one may be startled that a singlet-triplet process such as intersystem crossing can occur within the lifetime of an excited singlet state (10 8 sec). The mechanism for intersystem crossing involve vibrational coupling between the excited singlet state and a triplet state. Remembering that singlet-triplet processes are less probable than singlet-singlet processes by a factor of 10 5 to 10 6, and that radiationless vibrational processes (such as internal conversion) occur in approximately sec, the time required for a spin-forbidden vibrational process would be approximately 10 8 to 10 7 sec, which is the same order of magnitude as the lifetime of an excited singlet state. Therefore intersystem crossing can compete with fluorescence emission from the zeroth vibrational level of an excited singlet state but cannot compete with vibrational deactivation from higher vibrational level of a singlet state. Once intersystem crossing has occurred the molecule undergoes the usual 11

12 internal conversion process (10 13 to sec) and falls to the zeroth vibrational level of the triplet state. Since the difference in energy between the zeroth vibrational level of the triplet state and the zeroth vibrational level of the lowest exited singlet state is large compared to thermal energy, repopulation of a singlet state from a triplet state is highly improbable. There are two factors which tend to enhance a radiationless transition between the lowest triplet state and the ground state. First the energy difference between the triplet state and the ground state is smaller than the difference between the lowest singlet state and the ground state. This tends to enhance vibrational coupling between these two states, and therefore to enhance internal conversion. Second, and more important, the life time of a triplet state is much longer than that of an excited singlet state (about 10 4 to 10 sec) and therefore loss of excitation energy by collisional transfer is generally enhanced. In fact, this second process is so important that in solution at room temperature it is often the dominant pathway for the loss of triplet state excitation energy. If a molecule is placed in a rigid medium where collisional processes are minimized, a radiative transition between the lowest triplet state and the ground state is observed. This emission is called phosphorescence. As phosphorescence originates from the lowest triplet state, it will have a decay time approximately equal to the lifetime of the triplet state (ca to 10 sec). Therefore phosphorescence is often characterized by an afterglow which is not observed for fluorescence. 3.4 The Three Level System To investigate the kinetics of a single fluorophore it is sufficient to regard the molecule as a three level system where we take the following three states into account 0, 1, and 3 for groundstate S 0, singlet excited state S 1 and triplet excited state T 1. We use the derivation of the rates and populations for an ensemble of fluorophores. By appropriate normalization we can regard the results as occupation probabilities of a single fluorophore. Fig. 2 shows the three level system with the possible transitions and transition rates k ij. The excitation is described by k01 = σ I hν with the absorption crossection σ, the excitation intensity I, and the photon energy hν. Stimulated emission is described by k10 = σ I hν, spontaneous relaxation as k 10 (which is the sum of k nr +k rad, for nonradiative and radiative relaxation processes) and intersystem crossing by 12

13 1 0 k 01 * k 10 * k 10 k 13 k 30 3 Figure 2: Schematic representation of a closed three level system with the possible transitions and the corresponding transition rates k ij. k 13 = k ISC k 10 with the intersystem crossing rate k ISC. (k 10 + k 13 ) = λ 1 where τ 1 = 1 λ 1 is the lifetime of the excited state S 1 and k 30 = λ 3, where τ 3 = 1 λ 3 is the lifetime of the triplet state T 1. We define N 0, N 1 and N 3 as occupation numbers of the three levels and N as the total number of fluorophores. The following three equations have to be fulfilled: dn 1 dt = k01n 0 (k10 + k 10 + k 13 )N 1 (5) dn 0 dt = k01n 0 + (k10 + k 10 )N 1 + k 30 N 3 (6) dn 3 dt = k 13 N 1 k 30 N 3 (7) N = N 0 + N 1 + N 3 (8) In steady state all the changes have to vanish. Equs. 5, 8 and 7 therefore lead to: N 1 = k01 k10 + k N k 13 (9) N 0 = N (N 1 + N 3 ) (10) N 3 = k 13 k 30 N 1 (11) 13

14 The following expression for N 1 can now be calculated from Equ. 9: N 1 = k 01 k 10 + k 10 + k 13 1 ( ) N (12) k k10 +k 10+k k 13 k 30 To use the above derived results for a single three level system we set N = 1 and interpret the occupation numbers N i as occupation probabilities. The fluorescence rate n fl can then be written as n fl = N 1 k rad, (13) and the transition rate from the metastable triplet state into the ground state and with it the rate of triplet processes n tr = N 3 k 30. (14) For very strong excitation k10 = k 01 Equ. 12 simplifies to ( N 1,sat = 2 + k ) 1 13, (15) k 30 and the maximum photon emission rate from the singlet state is R = N 1,sat k rad = and therefore the saturation intensity is k rad 2 + k 13 /k 30, (16) I sat = hνk rad 2σ [ 1 + k13 /k k 13 /k 30 ]. (17) 14

15 4 Experimental 4.1 Microscope Setup Figure 3: Schematics of the wide field microscopy setup. The setup consists of a home built inverted microscope and is illustrated in Fig. 3. The light of a frequency doubled Nd:YAG laser is directed via a dichroic mirror into a microscope objective to illuminate the sample. A wide field lens creates a spot of 30µm in diameter in the focal plane of the objective. Assuming that all light emitted by the laser (25 mw) is transferred to the sample this gives a maximum intensity of about 6kW/cm 2. A microscope cover slide mounted on a XYZ translation stage is precisely adjustable with piezoelectrically driven micrometer adjustment screws. A drop of immersion oil with low autofluorescence is placed between the microscope slide and the objective to achieve index matching of the high N.A. (1.3) objective. The fluorescence light is collected by the same objective and passes the dichroic mirror. Stray light is blocked by a Raman cut off filter. After focusing by a mm zoom objective, the expanded (40-100x) fluorescence images are recorded by a CCD camera connected to a PC. The high resolution chip (1376x1040 pixel, size µm) has a quantum efficiency of > 60 % and low dark noise is provided by peltier cooling of the detector. With an A/D conversion factor of 2 electrons per count approximately 10 photons emitted from a molecule generate a count in one of the 15

16 pixels. At the CCD chip the fluorescence spot of a single molecule appears distributed over a range of 30µm (5 pixels, FWHM). Taking into account an expansion factor of 100, this corresponds to a spot size of 300 nm at the sample, which is close to the diffraction limit. 4.2 Sample Preparation A typical sample consists of a microscope cover slide coated with a thin polymer film (PMMA) containing an extremely small amount of dye (DiI) 2. All preparations can be made at room temperature and ambient pressure in the presence of air. In order to remove fluorescent impurities the cover glasses were baked for several hours at 500 C. calculate the dye concentration if the PMMA film is assumed to be about 50 nm thick and the molecules should be well resolvable by the microscope (diffraction limited spot 300nm). prepare a 5%(wgt.) solution of PMMA in toluene. mix a part of it with the appropriate amount of dye (start from stock solution). clean the spincoater plate and tape a coverslip to the plate. put one drop of polymer solution to the rotating ( 3000rpm) sample. To immerse the sample in N 2, a flexible tube of 8 mm inner diameter is fixed above the sample and connected to a nitrogen gas bottle. A gentle, laminar flow of N 2 removes the air from the sample surface and, via fast diffusion the oxygen molecules from the sample. 4.3 Basic Alignment of the Microscope make sure that the laserbeam propagates at constant height. check vertical alignment of mirror 1 and dichroic mirror by back reflection. 2 for details see section

17 remove microscope objective and adjust vertical beam by M2, use back reflection from a mirror on objective base plate. move beam in center of objective hole by use of translation stages (dichroic mirror and M2). insert objective and adjust beam to pass the objective on axis by translation stages. mount a coverglass (incl. one drop of immersion oil) and adjust distance of the coverglass to give a focus in front of the camera. adjust the camera position until you get the spot centered on the CCD array. If height and tilt are correctly adjusted the position of the spot should be independent of focusing, use strongly attenuated laser beam and computer control for camera. insert filter(s) mount sample with fluorescing species, adjust distance of the coverglass to give a focus in front of the filters. Gently move the sample mount by means of the MRA controller (set controller to motor C), and push knob left, two reflective spots appear, continue until the image becomes focused. Single spots should show up. proceed with your measurement. 4.4 Analysis The illuminated area is detected in a software selected region of the CCD array (recommended 512x512 pixels). Series of images can be recorded with appropriate exposure time. The images are stored on a PC and evaluated by home written routines based on a data visualization software package (PV-Wave, Visual Numerics). First, a virtual data cube as indicated in Fig. 4 is constructed from a series of sequentially recorded images. The front view shown in Fig. 4(a) shows a translucent image of the x-y plane of the data cube, i.e. the maximum of all intensities collected throughout the image series along the time axis. Subtracting the background intensity helps to analyze the data. Since we have to deal with an inhomogeneous background it can be calculated using a 3D median filter. Also indicated is the basic evaluation procedure of an experiment: in the x-y plane of 17

18 N y t x (a ) (b ) (c ) Figure 4: Schematic representation of a 3D data cube: (a) generated from a series of sequentially recorded images. The first frame shown here gives a translucent image of the data cube displaying the maximum intensities collected throughout the picture series. Different pixels or areas, as indicated in (a) can be selected and figures (b,c) illustrate the result of the analysis. For a single pixel or small area a time trace of a fluorescing single molecule (b) is extracted or by summation over a large area an ensemble of molecules (c) is obtained. 18

19 the data cube, points or areas of interest can be selected. The integrated intensities over these areas are plotted along the t-axis. Fig. 4 (b-c) show the typical intensity-time traces, a single pixel or small integration region (b) that contains a single spot in the x-y plane - a single molecule - and a large area of integration (c) representing an ensemble of molecules. Once experimental conditions favoring single-molecule detection are satisfied, a number of criteria have to be met to ascertain that the observed signal actually comes from a single emitter. These criteria, summarized in the following list, are direct consequences of common photophysical properties of fluorophores: 1. The observed density of emitters is compatible with the known concentration of individual molecules and scales with the original bulk concentration. 2. The observed fluorescence intensity level is consistent with that of a single emitting molecule. 3. Each immobilized emitter has a well-defined absorption or emission dipole (for organic dyes usually linear, but undefined for other emitters such as fluorescent semiconductor nanocrystal) 4. Fluorescence emission exhibits only two levels (on/off behavior due to blinking or photobleaching) over time scales where no changes in the environment are expected. 5. If there are two or more emission levels, photophysical property changes are correlated with blinking. 6. The emitted light exhibits antibunching, which is characteristic of a single quantum emitter. 7. The Spectrum of the molecule corresponds to the bulk Analysis Software The provided software is based on PV-WAVE and SXM (University of Basel). The programs menu and mouse button system is slightly different from other ones, therefore a short introduction: Clicking the mouse buttons causes the following actions: 19

20 Left mouse button executes the function. Shift + left mouse button calls the help text for the function. Right mouse button lets you set relevant options for the function. The main menu is at the bottom right of the screen in the lower part of the menu window. By clicking one of the main menu entries, the corresponding submenu appears in the upper part. For data analysis for this experiment a special user menu has been compiled, offering access to the most important functions for data analysis: Read Starts a file dialog that lets you select a file. Select the first file of a series that you want to analyze. In the next screen you will be asked how many frames of the specified data set you want to read. Write data Write a data channel as an ASCII *.txt file for further analysis, e.g. to fit a decay constant. Save image Save an image to disk using a common image file format, e.g. *.jpeg. Cut plane3d Allows to cut a selectable image out of the stack of images. Is also well suited to just browse through the images for inspection. Topview Shows a topview of a data channel that can be selected in the next window. (It handles only single images) Lineview Shows a lineview of a 1D data channel, e.g. a time trace. Show Max 3D Show a translucent image, which is buildt by displaying at each pixel position the maximum intensity measured in the series. Median 3D Subtracts a 3D median filtered image from the data. Is used to subtract background. Cut line 3D Lets interactively select a pixel at which the time series is extracted from the image series into a new channel. Shows the actual time series while moving the mouse over the image. Cut subarea 3D Lets select an area over which the intensity is integrated for each image in the series and stores it as a time series in a new channel. 20

21 Darktimes Calculates a histogram of darktimes of a time series (blinking analysis). 5 Further reading Beside of the already mentioned references: [13, 14, 15, 16] References [1] James Clerk-Maxwell. Nature, 8:437, An atom is a body which cannot be cut in two. A molecule is the smallest possible portion of a particular substance. No one has ever seen or handled a single molecule. Molecular science, therefore, is one of those branches of study which deal with things invisible and imperceptible by our senses, and which cannot be subjected to direct experiment. [2] Erwin Schrödinger. Brit. J. Phil. Sci., 3: , that we never experiment with just one electron, atom or (small) molecule. [3] Richard P. Feynman. There s plenty of room at the bottom, an invitation to enter a new field of physics. [4] C.J. Chen. Introduction to Scanning Tunneling Microscopy. Oxford University Press, New York, [5] R. Wiesendanger, editor. Scanning Probe Microscopy and Spectroscopy. Cambridge University Press, [6] Basché T.; Moerner W. E.; Orrit M. and Wild U. P., editors. Singlemolecule optical detection, imaging and spectroscopy. Verlag Chemie, Weinheim, [7] Drexler K. E. Nanosystems: molecular machinery, manufacturing, and computation. Wiley Interscience, [8] Frontiers in chemistry: Single molecules. Science, 283, special issue. 21

22 [9] T. Basché R. Rigler, M. Orrit, editor. Single molecule spectroscopy: Nobel conference lectures, volume 67 of Springer series in Chemical Physics. Springer Verlag Berlin, [10] V. Sandoghdar. Trends and developments in near field optical microscopy. In N. Garcia M. Alleghrini and O. Marti, editors, Proc. International School of Physics Enrico Fermi Course CXLIV, [11] R. N. Zare S. Nie. Annu. Rev. Biophys. Biomol. Struct., 26:567, [12] Joseph R. Lakowicz. Principles of fluorescence spectroscopy. Kluwer Academic/Plenum, New York, [13] I. Renge U.P. Wild C.G.H. Hübner, A. Renn. Direct observation of the triplet lifetime quenching of single dye molecules by molecular oxygen. J. Chem. Phys., 115:9619, [14] B. Hecht Y. Lill. Single dye molecules in an oxygen-depleted environment as photostable organic triggered single-photon sources. Appl. Phys. Lett., 84(10): , [15] Microscopy Resorce: [16] All you need to know about dyes and fluorescent labels: 22

23 6 Appendix 6.1 Laser Crystalaser ( Model number GCL-025-L; Linewidth < 0.2nm; Coherence length 30 mm, Output noise(rms) < 0.5%. 6.2 Wide field illumination lens The illuminated area diameter w s depends on the initial beam diameter w l, the focal length of the wide field lens, f w, and the focal length of the microscope objective, f o. Example: f o = 2mm; f w = 80mm; w l = 1mm; w s = 25µm. 6.3 Dichroic mirror Omega XF2017 (560DRLP) ( Figure 5: Transmission function of the dichroic mirror XF Microscope objective Nikon Planfluo 100x, 1.3, oil, infinity corrected. 23

24 6.5 Raman cut off filter Omega XR3002 ( Figure 6: Transmission function of the raman cutoff filter XF CCD Camera PCO ( Sensicam QE; Super Quantum Efficiency ; Low Noise 4e- ; High Resolution (1376 x 1040) ; 12 Bit Dynamic Range ; Exposure from 1ms s ; Peltier Cooled; Binning Horizontal and Vertical. Figure 7: Spectral sensitivity of the CCD camera. 24

25 6.7 DiI The dye used for the experiment is DiI (1,1 -dioctadecyl-3,3,3,3 - tetramethylindocarbocyanine perchlorate), also called DiIC18(3) (For further information see Figure 8: Absorption and emission spectra of DiI. H 3 C CH 3 H 3 C CH 3 N + CH CH CH N (CH 2 ) 17 CH 3 ClO 4 - (CH 2 ) 17 CH 3 Figure 9: Structure of DiI. 25