NONLINEAR OPTICAL STUDIES OF POTENTIAL-SENSITIVE DYES

BY

HAOWEN LI

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Thesis Advisor: Professor Kenneth D. Singer

Department of Physics

CASE WESTERN RESERVE UNIVERSITY

MAY, 2007 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

______

candidate for the Ph.D. degree *.

(signed)______(chair of the committee)

______

______

______

______

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein.

To AMZ and my parents

Table of Contents Table of Contents...... 1 List of Tables ...... 3 List of Figures...... 4 List of Figures...... 4 Acknowledgements...... 7 Acknowledgements...... 7 Abstract...... 8 Chapter 1: Introduction...... 10 1.1 A brief introduction to membrane potential...... 10 1.2 Mechanisms of potential-sensitive dyes ...... 12 1.3 Styryl potential-sensitive dye applications ...... 15 1.4 Overview of this thesis...... 18 References...... 21 Chapter 2: Nonlinear optical theory for potential-sensitive dyes ...... 26 2.1 Macroscopic susceptibility...... 26 2.2 Microscopic susceptibility ...... 30 2.2.1 Early models ...... 31 2.2.2 Quantum mechanical calculation...... 33 2.2.3 Two-level model ...... 35 2.3 Relationship between microscopic and macroscopic optical nonlinearities...... 39 2.3.1 Oriented gas model ...... 40 2.3.2 In Langmuir-Blodgett films ...... 42 2.4 Measurement Techniques [38]...... 44 References...... 46 Chapter 3: Spectroscopic studies and the first hyperpolarizability of di-8-ANEPPS..50 3.1 A brief introduction to styryl dyes...... 51 3.2 Spectroscopic studies of di-8-ANEPPS...... 52 3.2.1 Materials and methods ...... 53 3.2.2 Results...... 55 3.2.3 Discussion...... 57 3.3 The first hyperpolarizability of di-8-ANEPPS ...... 63 3.3.1 Introduction...... 63 3.3.2 Experiment and Results ...... 65 3.3.3 Discussion...... 71 3.4 Conclusion ...... 73 Appendix...... 74 References...... 76 Chapter 4: Solvent effects on the non-linear optical response of a potentiometric biological imaging dye...... 79 4.1 Introduction...... 80 4.2 Methods and Theory ...... 82 4.3 Experimental...... 88 4.4 Results and discussion ...... 90

1 4.5 Conclusions...... 101 Appendix...... 102 References...... 104 Chapter 5: Applications of potential-sensitive dyes ...... 108 5.1 Introduction...... 108 5.2 Langmuir-Blodgett films ...... 110 5.2.1 Film Deposition ...... 110 5.2.2 Film characterization ...... 112 5.2.2.1 π – A Isotherm ...... 112 5.2.2.2 UV-VIS Absorption measurements ...... 114 5.2.2.3 Second harmonic generation (SHG) ...... 115 5.2.3 Conclusion ...... 117 5.3 Near-field Scanning optical microscopy...... 118 5.3.1 History of near-field scanning optical microcopy ...... 118 5.3.2 AlphaSNOM from WiTec...... 120 5.3.3 Samples...... 122 5.3.4 Latex projection pattern imaging...... 123 5.3.5 T-cell imaging...... 125 5.3.5.1 Comparison between confocal microsocopy and NSOM...... 125 5.3.5.2 Single T-cell NSOM imaging ...... 127 5.3.6 Conclusion ...... 127 5.4 Future studies of potential-sensitive dye applications ...... 129 References...... 131 Bibliography ...... 135

2 List of Tables Table 3-1 Solution properties. F is the polarity, ε the dielectric constant, and n the refractive index. Data for mixtures are calculated from algebraic sum of the cosolvent fractions using ε and n of chloroform as 4.81 and 1.4458, respectivelya...... 54

3 List of Figures Figure 1.1 Electrostatic potential profiles across a phospholipid bilayer...... 12 Figure 1.2 Generalized structure of styryl dyes. The structure shown is the primary contributor to the ground state, where the positive charge concentrates in the pyridine ring. One double bond could be replaced by two double bonds; the phenyl ring could be replaced by a naphthalene ring ...... 14 Figure 2.1 Typical organic molecules for nonlinear optical effects. The electron donor group (D) is connected to the electron acceptor group (A) through a π electron system. The π electron system can be those containing benzene analogs or stilbene analogs. R1 and R2 are usually carbon or nitrogen. (Adapted from reference [6] )....31 2 22 Figure 2.2 Dependence of μge (---), 1/ ωge (… ), and Δμ =μμee − gg () −−− and β (___) on the ground structure of one donor-acceptor polyene (From reference [27])…………...... 39 Figure 2.3 A Langmuir-Blodgett monolayer with an average tilt angle α relative to the surface normal...... 42 Figure 3.1 Structures of two styryl dyes, molecular structures are adapted and modified from reference [1, 2]...... 51 Figure 3.2 Absorption and fluorescence emission spectra of di-8-ANEPPS in DMSO solution. The concentration is 0.02mg/ml and the cell length is 10mm. Two dashed lines are the absorption and emission spectra, respectively, while the lines are the Log-normal fits...... 55

Figure 3.3 Wavenumbers of the maxima of absorption vabs (upper squares) and of emission vem (lower dots) for di-8-ANEPPS versus the polarity function F for mixtures of

DMSO and Chloroform. The average values of (vvabs+ em ) / 2 are marked by stars. They are all linearly fitted...... 56

Figure 3.4 Half of the stokes shift (vvabs− em ) / 2 of the maxima of absorption and emission for di-8-ANEPPS versus the polarity function F. Data are linear fitted with a slope of 36000± 1000 cm-1 and an intercept of - 4200± 200cm−1 ...... 57 Figure 3.5 Schematic pictures of the electronically adiabatic ground state and excited adia state free energy surface in the solvent coordinate. The energy gap Veq between the equilibrated ground state and excited state and the solvent reorganization energies in the g ex adiabatic ground state and excited state Λs and Λs are indicated. Adapted from reference [8] ...... 62 Figure 3.6 First hyperpolarizability β of di-8-ANEPPS versus dielectric constant. Five wavelengths have been measured for the molecule: Square: 800nm, Circle: 900nm, diamond: 976nm, triangle: 1020nm and star: 1064nm. The lines are not fits, they are only guides for the eye...... 63 Figure 3.7 Extended-cavity mode-locked Ti:Sapphire laser layout. The 6.66W pump is provided by an Argon laser at 488nm. M: mirror; L: lens; CM: intracavity mirror (transparent to pump); OC: output coupler...... 66

4 Figure 3.8 TCSPC-HRS setup. BS: beam splitter, A: analyzer, HW: Half-wave plate, L: Lens, PD: Avalanche photo diode, F: 10nm narrow band filter centered at 400nm, P: polarizer, PMA: single photon detector, TH200: multi-time-channel chip……………67 Figure 3.9 TCSPC histograms: (a) fluorescence signal of Rhodamine-6G; (b) SHG peak of MNA power; (c) and (d) histogram of di-8-ANEPPS. Circles denote broadband fluorescence signal obtained by removing the narrowband filter, while the Square data is obtained with the narrowband filter in place and depicts combined SHG and fluorescence signal...... 68 Figure 3.10 HRS experiment data: the crosses represent the photon counts of SHG plus fluorescence; the circles represent the photon of pure SHG...... 70 Figure 3.11 Electron orbitals of (a) HOMO and (b) LUMO. The full alkyl chains found on the dialkylamine donor and on the alkylpyridinium acceptor (left and right end of the figures, respectively) were not included in the calculation as they are not expected to significantly affect the electronic orbitals. The orbitals were generated using ZINDO semi-empirical quantum chemical calculations...... 72 Figure 4.1 Structure of di-8-ANEPPS...... 80 Figure 4.2 Experimental setup for HRS measurements ...... 88 Figure 4.3 Linear absorption spectra of di-8-ANEPPS in solutions A, B, C, D and E. Absorption spectrum shifts toward the red (lower wavenumber) as the polarity decreases from A (rightmost curve) to E (leftmost curve)...... 91 Figure 4.4 Solvent polarity (F) dependent measured quantities. a) Absorption frequency, b) transition moment, c) spectral half-width and d) first hyperpolarizability. (Solid square 800nm, open square 900nm, triangle 976nm, solid circle 1020nm, and open circle 1064nm in d.) ...... 92 Figure 4.5 Hyperpolarizability β of di-8-ANEPPS in DMSO solution (Dots), F=0.220 , and fit to Equations (4.5) Solid line and (4.6) Dashed line...... 94 Figure 4.6 Linear absorption spectra of di-8-ANEPPS in DMSO solution (Solution A, solid line). Dashed line is the Log-normal fit for the first electronic absorption band. Dot- Dash is the CDHO lineshape fit...... 96 Figure 4.7 Plot of Δμ versus F and the corresponding linear fit ...... 97 Figure 4.8 Relative second harmonic sensitivity as a function of frequency. Solid line is the model of Equation (4.13) (right-hand side), and data is the left-hand side of Equation (4.13)...... 98 Figure 4.9 Contributions of different terms of Equation (4.3) as a function of frequency. Labels indicate the variable of the partial derivative in that equation. ....100 Figure 5.1 Molecular structure of (a) di-8-ANEPPS (b) DPPC...... 111 Figure 5.2 Isotherms of (a)DPPC, (b) 1:1 molar ratio of di-8-ANEPPS/DPPC; (c) di- 8-ANEPS; (d) Twelve continuous di-8-ANEPPS isotherms, all on a pure water surface at room temperature...... 113 Figure 5.3 UV-VIS absorption of di-8-ANEPPS in Chloroform solution 0.01M (Right Y-axis scale, dashed line); di-8-ANEPPS monolayer (Left Y-axis scale, solid line)…………...... 115 Figure 5.4 SHG measurements for Di-8-ANEPPS monolayer deposited onto plain glass substrate (thickness 1mm) P-P polarization configuration (Line ) S-P polarization (Cross) ...... 117

5 Figure 5.5 Schematic diagram illustrating Synge’s proposal for achieving subdiffraction limit resolution...... 119 Figure 5.6 Left plot is the distance feedback using cantilever-NSOM sensors. The excitation as well as the feedback laser are focused onto the cantilever through the same lens system. The cantilever is loaded into the optical axis of the microscope with a high precision XY-stage for maximum throughout. Right side is cantilever with a hollow SiO2 tip. [Obtained from WiTec Manual] ...... 121 Figure 5.7 (a) Atomic force microscopy and (b) near-field scanning optical microscopy transmission image of the Latex projection pattern. (c) is the cross section of line on figure 5.7(a) . (d) is the cross section in the similar area of (figure 5.7(b))………...... 124 Figure 5.8 Fluorescence images of (a) T-cell stained with Alexa_fluor 568 from Zeiss LSM510 inverted confocal microscope; (b) T-cell stained with di-8-ANEPPS from Nikon microscope; (c) T-cell stained with Alexa_fluor 568 from WiTec confocal microscope; (d) T-cell with Alexa_fluor 568 from WiTec Fluorescence NSOM microscope...... 126 Figure 5.9 Single T-Cell imaging (a) NSOM fluorescence image (b) AFM image (c) shows the cross section of the short black line in (b) ...... 128

6 Acknowledgements

My past five and a half years at Case Western Reserve University has been a rewarding experience. I am especially grateful to my advisor, Professor Kenneth D. Singer, for his advice, support, and patience throughout my research and professional development.

I wish to thank all the members in the Organic Optoelectronics Laboratory group, both past and present. I am particularly thankful to Dr. Steven Kurti, who helped me getting familiar with our lab. I also wish to thank Jessica Merlin for many friendly talks, and for my thesis proof reading and editing. I enjoyed working together with Guilin Mao, Hefei Shi, Yeheng Wu, and Dr. Volodimyr Duzhko for various projects. Your friendships made my everyday working in the basement endurable and enjoyable.

I appreciate the discussion and collaboration with Professor Rolfe G. Petschek and Professor Jie Shan during our weekly seminars and on my research. Also I would like to thank their group members for broadening my scientific knowledge.

Most of my work had very close collaborations with several research groups at Case Campus. Here I would like to thank Dr. Anando Devadoss, Dechen Jiang and Prof. James D. Burgess from Chemistry Department for help with Langmuir-Blodgett experiments. I would like to thank Professor J. Adin Mann from Chemical Engineering for numerous guidances for thin-film depositions. I thank Zhilei Liu, Dr. Brian Todd, and Prof. Steven J. Eppell from Biomedical Engineering for involvement with AFM and NSOM experiments. I would like to thank Feng Xue and Professor Alan D. Levine from Medical School for biological sample preparation.

I thank Professor Walter Lambrecht for helpful discussion with numerical calculations and Professor Charles Rosenblatt for allowing me using equipments in his lab. Also, I would like to thank Mary MacGowan, Pat Bacevice, Lucy Rosenberg and Lori Rotar Morton for making my daily life easier at Physics Department.

I have also enjoyed the friendships with many first-year classmates. Just to name to few: Jeremy Heilman, Pavel Lukashev, Tim Peshek, Ishtiaque Syed, Mehdi Bagheri- Hamaneh, and Victor Taracila. I sincerely hope you all have a successful career.

Most of all, I am thankful for the love and support from my family. I won’t be able to go such far without their encouragement, patience, and inspiration.

7

Nonlinear Optical Studies of

Potential-sensitive Dyes

Abstract by Haowen Li

Di-8-ANEPPS and other similar potential-sensitive dyes have been extensively used to probe the electrical environment in biological membranes due to changes in its linear and nonlinear optical properties as a function of the local electric field. Ratiometric fluorescence, two-photon excited fluorescence (TPEF) and second harmonic generation

(SHG) have all been applied to study biological cell and membrane systems. Mechanism and sensitivity studies for these potential-sensitive dyes have become popular due to the increasing biological needs. In this thesis, fluorescence-free time-correlated single photon counting hyper Rayleigh scattering (TCSPC-HRS) has been used to determine the hyperpolarizability β. Experiment results are in agreement with quantum chemical calculations. Linear and nonlinear optical studies of di-8-ANEPPS in liquid solution have been carried out. This polarity-dependent solvent effects of linear absorption and hyper-

Rayleigh scattering in solution is combined together to analyze the spectral dependence of the sensitivity to the local environment. In this thesis, a two-level model including the spectral shift, changes in transition moment, excited and ground state dipole moment

8 difference, and spectral width is applied to study the sensitivity of di-8-ANEPPS in response to the local electric field (solvent polarity). Good agreement between the model and the measurements has been found. Studies of a di-8-ANEPPS monolayer have been tried and high resolution fluorescence near-field scanning optical microscopic (NSOM) imaging with T-Cells has been performed. Optical resolution beyond diffraction limit has been achieved.

9 Chapter 1: Introduction

1.1 A brief introduction to membrane potential

Over the last 50 years, since the discovery of the ion transporting membrane protein in 1957 [1] and the suggestion of the fluid mosaic model of the cell membrane structure in 1972 [2], it has become clear that the physiological process of ion transport across cell membranes is mediated via specialized transport proteins embedded in the lipid bilayer matrix. Ion transport is a fundamental mechanism underlying many important physiological processes, such as nerve signaling, muscle contraction, and energy transduction. Voltage-gated ion channels function by responding to changes of electric field inside the cell membrane. The intramembrane electric field has three distinct potential components: transmembrane potential, surface potential, and dipole potential. [3,

4] Figure 1.1 shows the potential profile across a phospholipid bilayer.

The transmembrane potential is the overall potential difference between the two aqueous solutions separated by the membrane. It is produced by selective ionic permeability of the cell membrane and the asymmetrical ionic concentration in the outer and inner bulk solutions across the membrane. Based on electroneutrality, the net charge on one side of the membrane would be equal and opposite to the net charge on the other side of the membrane. These equal and opposite charges separated by the membrane form a capacitative charge transmembrane potential which is probably the most important component of the membrane potential.

The surface potential is the electric potential difference between the membrane aqueous interface and the bulk aqueous solutions and depends on the density of

10 interfacial charged molecules. [5, 6] Most biological membranes contain zwitterionic lipids and acidic lipids on membrane-bound proteins. While the fixed charges are bound to the membrane, the counter ions are dissolved in the adjacent solution and diffused away from the membrane surface. This creates a charge separation and, consequently, an electric potential difference called the surface potential. It is on the order of a few tens of mVs and can be described using Gouy-Chapman-Stern theory [7-9] which gives the relationship between surface potential, the surface charge density and the ionic concentration in the aqueous phase.

The fact that hydrophobic anions bind several orders of magnitude stronger and translocate several orders of magnitude faster across a lipid bilayer than structurally similar cations [4, 10, 11] suggests a positive potential barrier inside the bilayer of several hundred mV, which is referred as dipole potential.[10] Unlike the surface potential, this barrier potential is independent of ionic strength and is presumed to originate from oriented dipoles in the membrane/water interface. The orientation of dipoles in the water molecules adjacent to the membrane, the polar head groups, and the ester linkages of the acyl chains to the glycerol backbone of the phospholipid could all account for such a potential difference between the interior of the bilayer and the aqueous phase.

Each of the above membrane potentials plays an important role of regulating the ion channels and influencing the conformation of membrane proteins. Thus it is important to accurately control and measure the intra-membrane potential. The patch clamp technique [12] has been widely used to control and measure the transmembrane potential. However, it cannot measure the variations in potential along the surface of the cell nor can it provide information on the potential profile across the membrane. Recently,

11 voltage sensitive dyes have become popular probes for the membrane potential. [13, 14]

These dyes bind to lipid membranes with their chromophore in the lipid headgroup region, where they are sensitive to the local electric field. Fluorescence excitation spectrum shifts of the membrane-bound dye have been utilized to probe the membrane potential.

Figure 1.1 Electrostatic potential profiles across a phospholipid bilayer. Ψs is the surface potential; ΨD is the dipole potential, ΔΨ is the trans- membrane potential.

1.2 Mechanisms of potential-sensitive dyes

L. B. Cohen and his co-workers led the pioneering effort to originate and promote the growth of the potential-sensitive probes for monitoring the changes in membrane potential. They screened thousands of available dyes for potential dependent responses to fast voltage clamp pulses on the squid axon and attempted to improve the synthetic

12 methods for promising dyes.[4, 15, 16] Waggoner demonstrated the extraordinarily high sensitivities of cyanine dyes for slow potential changes in cells in suspension.[17, 18]

Loew and his coworkers have synthesized a number of charge-shift probes after theoretical analyses of electrochromic mechanisms. [19-21]

There are at least three molecular mechanisms by which probes respond to a change in membrane potential. [14] A given dye may also respond via different mechanisms in different systems and may sometimes adopt several mechanisms simultaneously.

First, the fluorescence intensity can be altered by the membrane potential through a voltage-dependent redistribution mechanism, in which the fluorescent dye partitions into the membranes as a function of the electrical potential. A hydrophobic ionic dye will have a certain equilibrium distribution between the aqueous medium and the cell membrane. The membrane potential governs the distribution across and within the cell membrane. The more negative the potential, the greater the accumulation of positively charged dye, and vice versa. A change in membrane potential will result a different dye environment, thus resulting in a change in fluorescence, since the fluorescence of the probe dye is very sensitive to its molecular environment.

Second, the fluorescence intensity of a dye can be altered by the membrane potential through a re-orientational mechanism, in which the membrane-bounded dye tilts in response to an electric field. A dipolar probe dye may adopt a variety of orientations with respect to the membrane surface. The electrostatic interaction of the dipole with the membrane potential can affect these orientations. In this case, there are several ways for the orientation of the dye to change the fluorescence. Both changes in molecular

13 environment and reorientation-dependent dimerization could account for a different absorption and fluorescence emission spectra.

Third, the electric field could directly perturb the dye’s electronic transition through electrochromism. Electrochromism is the direct coupling of the membrane electric field with the electronic redistribution in a chromophore that accompanies excitation or emission.[22] The theory can be written in a simple form:

hqΔ=−⋅−Δ⋅−ΔναrE μ E E2 (0.1) where h is the Plank constant, Δν is the change in light frequency, q is the charge, r is the charge displacement, Δμ is the difference in dipole moment, and Δα is the difference in the polarity in the two electronic states involved in the transition. Upon application of an electric field E, the change in energy for the electronic transition depends on the three terms in the right side of Equation (1.1). The third term, which has a quadratic dependence on the electric field, is insignificant for fields in the physiological range

(hundreds of mV). However, the first two terms can be large enough and play an important role in the potential sensitive applications in biological membranes.

Figure 1.2 Generalized structure of styryl dyes. The structure shown is the primary contributor to the ground state, where the positive charge concentrates in the pyridine ring. One double bond could be replaced by two double bonds; the phenyl ring could be replaced by a naphthalene ring.

14 1.3 Styryl potential-sensitive dye applications

The styryl probes developed by Loew and coworkers [23] are currently popular fast-fluorescent probes. (See Figure 1.2) These chromophores bear a positive charge which is concentrated in the pyridine ring in the ground state and shifts to the aniline in the excited state. [20, 21] This charge shift can couple with the electric field across the membrane if the probe is appropriately oriented, leading to electrochromism. Most of the styryl dyes have absorption maxima within the range of 450nm to 550nm. The fluorescence emission spectra are shifted to longer wavelengths by 100 to 150nm which makes them very suitable to monitor the potential changes in membrane. In addition, the fluorescence quantum yields are 100 times stronger for membrane-bound dyes than the dyes in water, which reduces the contribution from any unbound dyes to the fluorescence.

[19]

Di-8-ANEPPS(Pyridinium, 4-[2-[6-(dioctylamino)-2-naphthalenyl]ethenyl]-1-(3- sulfopropyl)-, inner salt, C36H52N2O3S) and other similar styryl dyes have been extensively used as potential sensitive dyes to probe the electrical environment in biological membranes due to changes in their optical properties as a function of the local electric field. [13] Dual wavelength fluorescence ratios of the chromophore have been used to measure the local intramembrane electric field, including transmembrane potential,[24-26] surface potential, [27, 28]and dipole potential. [29, 30] For example, a ratiometric method is used by Gross et al. [29] to quantify the membrane dipole potential.

They found that the binding to the membrane of the dipolar compound phloretin, which is known to decrease the dipole potential,[31] caused a red shift of the fluorescence excitation spectrum, whereas the binding of 6-ketocholestanol, which is known to

15 increase the dipole potential, [31] causes a blue shift. In order to quantify the spectra shifts, Gross et al. determined the ratio R of the fluorescence intensities at two excitation wavelengths (440nm and 530nm) and a fixed emission wavelength (620nm). A decrease in R represents a decrease in the dipole potential, and correspondingly, an increase in R represents an increase in the dipole potential. They found out that their R values were approximately linearly related to the dipole potential, with a change in R of 0.8 corresponding to a change in the dipole potential of ~100mV.

Two-photon excited fluorescence (TPEF) and second harmonic generation (SHG) have now become popular tools in biological cell imaging. The popularity of TPEF has greatly expanded since the first implementation for biological microscopy in 1990 [32] and it has found useful applications in neuroscience, cell biology, and biophysics. SHG is also a nonlinear optical process, which has been developed by Shen [33] and Eisenthal

[34] to study surface and artificial monolayers, and shares many of the features of TPEF, such as greatly reduced out of plane photobleaching and phototoxicity, increased signal to noise ratio, and deeper penetration into the cell because of the use of the infrared light.

Huang [35] pioneered the application of SHG to study specific molecules in biological membranes, an approach that others continued by imaging physiological indicator dyes bound to cellular membranes. [36-39]

Because SHG is a coherent phenomenon that involves scattering and TPEF is an incoherent phenomenon that involves absorption and reemission, they have intrinsically different properties. SHG is an instantaneous response while TPEF will have a response time of a few picoseconds to nanoseconds depending on the life time of the involved excited states. One major constraint of SHG is the requirement of a non-centrosymmetric

16 environment. Because of the inherent asymmetry of lipid bilayers, both intracellular organelle and plasma membranes are suitable samples for probing with this methodology.

This is because the membrane structure forces a non-centrosymmetric environment, provided that only one leaflet is stained with the dye indicator. SHG and TPEF imaging can be performed simultaneously to provide complementary information. [37, 40, 41]

Moreaux et al. [42] demonstrated the presence of TPEF and the cancellation of SHG at the adhesion zone of two giant unilamellar vesicles (GUV) whose outside leaflets were stained with potential-sensitive dye. In the adhesion zone, the dye molecular distribution is centrosymmetrical because an approximately equal number of molecules are oriented in opposing direction and hence produce second harmonic with opposing phases. This leads to a total destructive interference of SHG, whereas it produces no change in phase- insensitive TPEF. In the same paper, they also demonstrated the decay of the SHG signal as a result of dye molecule flip-flop into both leaflets, which makes the net dye molecular distribution become progressively more centrosymmetric. The simultaneously acquired

TPEF does not change much, because it is relatively insensitive to the flip-flop rate and reflects only the small variations of the dye concentration.

Recently sensitivities of TPEF and SHG to the membrane potential have been studied extensively. [43-45] In comparison with the voltage sensitivity of 10%/100mV fluorescence intensity change, [46], Millard [44] found second harmonic sensitivities of up to 43%/100mV in the similar styryl dyes, which clearly shows SHG has great promise for allowing considerable improvements over existing fluorescence-based techniques to monitor the membrane potential change. Besides this, the mechanisms of the dyes contributing to the voltage response have been quantified and characterized by Mertz et

17 al. [47, 48] They found that the potential response may arise purely from electrically induced changes in the molecular electronic hyperpolarizability β, or by a combination of this and molecular reorientation by the external electric field. Several papers have theoretically and experimentally discussed the two-photon fluorescence or second harmonic sensitivity of di-8-ANEPPS and similar potential dyes. [49-51]

1.4 Overview of this thesis

Using potential-sensitive dyes to image the cell membrane and to probe the membrane potential has become widespread. One photon excited fluorescence, two photon excited fluorescence and second harmonic generation have all been used in these applications. A thorough nonlinear optical study of the potential-sensitive dyes will provide better understanding of their operational mechanisms and improve the sensitivities of the potential measurements. By physical characterization of the potential- sensitive dye, we could also find out the optimum wavelength for fluorescence or second harmonic imaging. Effects of using Langmuir-Blodgett nano-films to simulate the lipid membrane have been tried and near-field scanning optical microscope imaging of the cells stained with fluorescence dyes has also been performed.

Chapter 1 introduces the membrane physics and the electric properties of the cell membranes. Three important membrane potentials, i.e., surface potential, transmembrane potential, and dipole potential and their corresponding functions in ion channel regulation have been discussed. Three different mechanisms of potential-sensitive dyes such as redistribution, reorientation, and electrochromism have also been reviewed. Biological

18 applications of the styryl dyes such as di-8-ANEPPS have been discussed and finally the overview of this thesis is introduced.

Chapter 2 will start with the basic nonlinear optics theory and the interactions of light with matter. Several early empirical models for understanding the first hyperpolarizability of organic materials have been reviewed. A simple quantum- mechanical calculation is used to obtain expressions for hyperpolarizabilities and a two- level model approximation is introduced for charge-transfer dyes. Methods to obtain an optimized hyperpolarizability are discussed. Relationships between microscopic hyperpolarizability and macroscopic susceptibility have been discussed under the oriented gas model for organic materials. Characterization and measurement techniques for second-order nonlinear effects have also been reviewed.

Chapter 3 will introduce the styryl dyes and start with spectroscopic studies of di-

8-ANEPPS. The anomalous solvatochromism is presented and comparisons with theoretical models are performed. Time-correlated single photon counting hyper Rayleigh scattering (TCSPC-HRS) has been used to determine the hyperpolarizability component,

βzzz . Experiments are in agreement with quantum chemical calculations. The hyperpolarizability allows a determination of the value of the difference in dipole moment between the lowest excited and ground states. This quantity can be used to estimate the electrochromic response. Details of the TCSPC-HRS technique are also described.

In Chapter 4 solvent effects of the hyperpolarizability β of di-8-ANEPPS has been discussed. We seek to examine the purely electronic mechanism for electric-field dependent second harmonic generation using solvent dependent optical properties. To do

19 so, we have carried out linear and nonlinear optical studies of the di-8-ANEPPS dye in liquid solution. We use polarity-dependent solvent studies of linear absorption and hyper-Rayleigh scattering in solution to probe the spectral dependence of the sensitivity to the local environment. We also model this sensitivity using a two-level model by including the spectral shift, changes in transition moment, excited and ground state dipole moment difference, and spectral width as the local electric field (solvent polarity) is varied. We model the relationship between first hyperpolarizability, β, and linear absorption spectrum as a function of solvent polarity and, by inference, local electric field.

We find good agreement between the model and the measurements.

In Chapter 5, using Langmuir-Blodgett nano-films as a model for biological membrane, we performed the measurements and reviewed the techniques for characterizing Langmuir-Blodgett films. Preliminary studies of high resolution fluorescence near-field scanning optical microscopic (NSOM) imaging with T-Cells have been performed. Optical resolution beyond the diffraction limit has been accomplished.

In the end, I will give a brief overview of future work related with the potential sensitive dyes.

20 References

1. Skou, J.C., The influence of some cations on an adenosine triphosphatase from peripheral nerves. Biochim. Biophys. Acta, 1957. 23: p. 394-401.

2. Singer, S.J. and G.L. Nicolson, The Fluid Mosaic Model of the Structure of Cell Membranes. Science, 1972. 175(18): p. 720-731.

3. Loew, L.M., The electrical properties of membranes. Biomembranes. Physical Aspects, ed. M. Shinitzky. 1993, Weinheim, Germany: VCH Publishers. 341-371.

4. Honig, B.H., W.L. Hubbell, and R.F. Flewelling, Electrostatic Interactions In Membranes And Proteins. Annual Review Of Biophysics And Biophysical Chemistry, 1986. 15: p. 163-193.

5. McLaughlin, S., The Electrostatic Properties Of Membranes. Annual Review Of Biophysics And Biophysical Chemistry, 1989. 18: p. 113-136.

6. McLaughlin, S., Electrostatic potentials at membrane-solution interfaces. Current Topics Membranes and Transport, ed. F. Bronner and J. Kleinzeller. 1977, New York: Academic Press. 71-144.

7. Barlow, C.A., The electrical double layer. Physical Chemistry, an Advanced Treatise, ed. H. Eyring. 1970, New York: Academic Press Inc. 167-246.

8. Davies, J.T. and E.K. Rideal, Interfacial Phenomena. 1963, New York: Academic Press Inc.

9. Grahame, D.C., The electrical double layer and the theory of electrocapillarity. Chemical Reviews, 1947. 41: p. 441-501.

10. Hladky, S.B. and D.A. Haydon, Membrane conductance and surface potential. Biochim. Biophys. Acta, 1973. 318: p. 464-468.

11. Liberman, E.A. and V.P. Topaly, Permeability of biomolecular phospholipid membranes for fat-soluble ions. Biophysics, 1969. 14: p. 477-487.

12. Neher, E. and B. Sakmann, Single channel currents recorded from membrane of denervated frog muscle fibres. Nature, 1976. 260: p. 799-802.

13. Loew, L.M., Potentiometric dyes: Imaging electrical activity of cell membranes. Pure And Applied Chemistry, 1996. 68(7): p. 1405-1409.

14. Loew, L.M., How to choose a potentiometric membrane probe. Spectroscopic Membrane Probes, ed. L.M. Loew. Vol. II. 1988: CRC Press. 139-151.

21 15. Gupta, R.K., B.M. Salzberg, A. Grinvald, L.B. Cohen, K. Kamino, S. Lesher, M.B. Boyle, A.S. Waggoner, and C.H. Wang, Improvements In Optical Methods For Measuring Rapid Changes In Membrane-Potential. Journal Of Membrane Biology, 1981. 58(2): p. 123-137.

16. Cohen, L.B., B.M. Salzberg, H.V. Davilla, W.N. Ross, D. Landowne, A.S. Waggoner, and C.H. Wang, Changes in axon fluorescence during activity: molecular probes of membrane potential. Journal Of Membrane Biology, 1974. 19: p. 1.

17. Waggoner, A.S., Dye indicators of membrane potential. Annual Review Of Biophysics And Bioengineering, 1979. 8: p. 847.

18. Sims, P.J., A.S. Waggoner, C.H. Wang, and J.F. Hoffmann, Studies on the mechanism by which cyanine dyes measure membrane potential in red blood cells and phosphatidyl chroline vesicles. Biochemistry, 1974. 13: p. 3115.

19. Fluhler, E., V.G. Burnham, and L.M. Loew, Spectra, Membrane-Binding, And Potentiometric Responses Of New Charge Shift Probes. Biochemistry, 1985. 24(21): p. 5749-5755.

20. Loew, L.M. and L.L. Simpson, Charge-Shift Probes Of Membrane-Potential - A Probable Electrochromic Mechanism For Para-Aminostyrylpyridinium Probes On A Hemispherical Lipid Bilayer. Biophysical Journal, 1981. 34(3): p. 353-365.

21. Loew, L.M., G.W. Bonneville, and J. Surow, Charge shift optical probes of membrane potential. Theory. Biochemistry, 1978. 17: p. 4065.

22. Platt, J.R., Electrochromism, a possiblechange of color producible in dyes by an electric filed. Journal Of Chemical Physics, 1956. 25: p. 80.

23. Loew, L.M., Design And Characterization Of Electrochromic Membrane Probes. Journal Of Biochemical And Biophysical Methods, 1982. 6(3): p. 243-260.

24. Zhang, J., R.M. Davidson, M.D. Wei, and L.M. Loew, Membrane electric properties by combined patch clamp and fluorescence ratio imaging in single neurons. Biophysical Journal, 1998. 74(1): p. 48-53.

25. Bedlack, R.S., M.D. Wei, and L.M. Loew, Localized Membrane Depolarizations And Localized Calcium Influx During Electric Field-Guided Neurite Growth. Neuron, 1992. 9(3): p. 393-403.

26. Montana, V., D.L. Farkas, and L.M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurements Of Membrane-Potential. Biochemistry, 1989. 28(11): p. 4536-4539.

22 27. Xu, C., B.M. Slepchenko, and L.M. Loew, Surface potential measurements on biological membranes with voltage-sensitive dyes. Biophysical Journal, 2002. 82(1): p. 546A-546A.

28. Zhang, J., M.D. Wei, and L.M. Loew, Change of surface potential induced by intracellular digestion of membrane proteins. Biophysical Journal, 1998. 74(2): p. A402-A402.

29. Gross, E., R.S. Bedlack, and L.M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurement Of The Membrane Dipole Potential. Biophysical Journal, 1994. 66(2): p. A387-A387.

30. Bedlack, R.S., M.D. Wei, S.H. Fox, E. Gross, and L.M. Loew, Distinct Electric Potentials In Soma And Neurite Membranes. Neuron, 1994. 13(5): p. 1187-1193.

31. Franklin, J.C. and D.S. Cafiso, Internal Electrostatic Potentials In Bilayers - Measuring And Controlling Dipole Potentials In Lipid Vesicles. Biophysical Journal, 1993. 65(1): p. 289-299.

32. Denk, W., J.H. Strickler, and W.W. Webb, 2-Photon Laser Scanning Fluorescence Microscopy. Science, 1990. 248(4951): p. 73-76.

33. Shen, Y.R., Surface-Properties Probed By 2nd-Harmonic And Sum-Frequency Generation. Nature, 1989. 337(6207): p. 519-525.

34. Eisenthal, K.B., Liquid interfaces probed by second-harmonic and sum-frequency spectroscopy. Chemical Reviews, 1996. 96(4): p. 1343-1360.

35. Huang, J.Y., A. Lewis, and L. Loew, Nonlinear optical properties of potential sensitive styryl dyes. Biophys J, 1988. 53(5): p. 665-70.

36. Moreaux, L., O. Sandre, and J. Mertz, Membrane imaging by second-harmonic generation microscopy. Journal Of The Optical Society Of America B-Optical Physics, 2000. 17(10): p. 1685-1694.

37. Campagnola, P.J., M.D. Wei, A. Lewis, and L.M. Loew, High-resolution nonlinear optical imaging of live cells by second harmonic generation. Biophysical Journal, 1999. 77(6): p. 3341-3349.

38. BenOren, I., G. Peleg, A. Lewis, B. Minke, and L. Loew, Infrared nonlinear optical measurements of membrane potential in photoreceptor cells. Biophysical Journal, 1996. 71(3): p. 1616-1620.

39. Bouevitch, O., A. Lewis, I. Pinevsky, J.P. Wuskell, and L.M. Loew, Probing Membrane-Potential With Nonlinear Optics. Biophysical Journal, 1993. 65(2): p. 672-679.

23 40. Campagnola, P.J., H.A. Clark, W.A. Mohler, A. Lewis, and L.M. Loew, Second- harmonic imaging microscopy of living cells. Journal Of Biomedical Optics, 2001. 6(3): p. 277-286.

41. Moreaux, L., O. Sandre, M. Blanchard-Desce, and J. Mertz, Membrane imaging by simultaneous second-harmonic generation and two-photon microscopy. OPTICS LETTERS, 2000. 25(5): p. 320-322.

42. Moreaux, L., O. Sandre, S. Charpak, M. Blanchard-Desce, and J. Mertz, Coherent scattering in multi-harmonic light microscopy. Biophysical Journal, 2001. 80(3): p. 1568-1574.

43. Ries, R.S., H. Choi, R. Blunck, F. Bezanilla, and J.R. Heath, Black lipid membranes: visualizing the structure, dynamics, and substrate dependence of membranes. Journal Of Physical Chemistry B, 2004. 108(41): p. 16040-16049.

44. Millard, A.C., L. Jin, M.D. Wei, J.P. Wuskell, A. Lewis, and L.M. Loew, Sensitivity of second harmonic generation from styryl dyes to transmembrane potential. Biophysical Journal, 2004. 86(2): p. 1169-1176.

45. Millard, A.C., L. Jin, A. Lewis, and L.M. Loew, Direct measurement of the voltage sensitivity of second-harmonic generation from a membrane dye in patch- clamped cells. Optics Letters, 2003. 28(14): p. 1221-1223.

46. Loew, L.M., L.B. Cohen, J. Dix, E.N. Fluhler, V. Montana, G. Salama, and J.Y. Wu, A Naphthyl Analog Of The Aminostyryl Pyridinium Class Of Potentiometric Membrane Dyes Shows Consistent Sensitivity In A Variety Of Tissue, Cell, And Model Membrane Preparations. Journal Of Membrane Biology, 1992. 130(1): p. 1-10.

47. Pons, T., L. Moreaux, O. Mongin, M. Blanchard-Desce, and J. Mertz, Mechanisms of membrane potential sensing with second-harmonic generation microscopy. JOURNAL OF BIOMEDICAL OPTICS, 2003. 8(3): p. 428-431.

48. Moreaux, L., T. Pons, V. Dambrin, M. Blanchard-Desce, and J. Mertz, Electro- optic response of second-harmonic generation membrane potential sensors. OPTICS LETTERS, 2003. 28(8): p. 625-627.

49. Pons, T. and J. Mertz, Membrane potential detection with second-harmonic generation and two-photon excited fluorescence: A theoretical comparison. Optics Communications, 2006. 258(2): p. 203-209.

50. Millard, A.C., L. Jin, J.P. Wuskell, D.M. Boudreau, A. Lewis, and L.M. Loew, Wavelength- and time-dependence of potentiometric non-linear optical signals from styryl dyes. Journal Of Membrane Biology, 2005. 208(2): p. 103-111.

24 51. Fisher, J.A.N., B.M. Salzberg, and A.G. Yodh, Near infrared two-photon excitation cross-sections of voltage-sensitive dyes. Journal Of Neuroscience Methods, 2005. 148(1): p. 94-102.

25 Chapter 2: Nonlinear optical theory for potential-sensitive dyes

In this chapter, general considerations of organic materials for quadratic nonlinear applications are discussed. First a simple description of the interaction of light with matter is presented and used as a basis for understanding nonlinear optics. In organic bulk systems, because the molecular units interact very weakly due to the absence of net charge or intermolecular charge transfer, an oriented gas model is often used to relate the molecular properties to the corresponding bulk properties. In this oriented gas model, the organic bulk optical nonlinearity is determined primarily by the nonlinear optical properties of the individual molecular unit. Thus, it is important to understand the electronic origins of the microscopic nonlinearity and its dependence on molecular structures. Several theoretical approaches to the molecular origin of the microscopic nonlinearity are described, relationships between microscopic hyperpolarizability and macroscopic susceptibility are reviewed, and, finally, measurement techniques for second-order nonlinear optical effects are presented.

2.1 Macroscopic susceptibility

Let us consider the wave equation for the propagation of light through a nonlinear optical medium. [1-3] Maxwell’s equations

∇⋅=D 4πρ (2.1)

∇⋅=B 0 (2.2)

1 ∂B ∇×E =− (2.3) ct∂

26 14∂D π ∇×HJ = + (2.4) ct∂ c involve macroscopic electromagnetic variables E, H, D, B, J and ρ. Equations (2.1-2.4) must be completed by further relations among the vectors E, D, B, H. In order to account for the dielectric and magnetic response of matter, the electric dipole density P and magnetic dipole density M need to be introduced:

DE= + 4π P (2.5)

BH= + 4π M (2.6)

We are primarily interested in the region that contains no free charge and no free current, so that

ρ = J = 0 (2.7) and if we assume the material is nonmagnetic, so that

BH= (2.8)

It is possible to eliminate the magnetic variables and obtain the following wave equation:

1 ∂ ∇×∇×ED =− (2.9) ct22∂

Now using equation (2.5) to eliminate D, we obtain:

14∂22EPπ ∂ ∇×∇×E + =− (2.10) ct22∂ c 2∂ t 2

This is the general form of the wave equation in nonlinear optics. By using the vector identity,

∇×∇×()()EEE =∇∇⋅ −∇2 (2.11)

27 The first term on the right side of the equation (2.11) vanishes in the case of linear optics of isotropic source-free media, and remains small and can be dropped under the slowly- varying amplitude approximation in nonlinear optics. Equation (2.10) can be rewritten as:

14∂22EPπ ∂ −∇2E + = − (2.12) ct22∂ c 2∂ t 2

The polarization density P can be split into its linear and nonlinear parts as:

PP()ωωωωωω=+(1) () P NL () =+++ P (1) () P (2) () P (3) () (2.13)

and

(n) ()n PEEE()ω =−χωωωω ( ;12 , ,nn ) ( ωω 1 )( 2 ) ( ω ) (2.14) where χ(1) is the linear susceptibility, χ(2) and χ(3) are the second- and third-order nonlinear optical susceptibilities, respectively. The displacement, D, can be decomposed into its linear and nonlinear parts as:

DE=+(4ππ P(1) )4 + P NL = D (1) + 4 π P NL (2.15)

Then, equation (2.12) becomes:

14∂∂22DP(1)π NL −∇2E + = − (2.16) ct22∂∂ ct 2

In the case of dispersive medium, we need to consider each frequency component of the field separately. The electric field, linear displacement, and polarization can be represented as the sums of their various frequency components:

Er(,)tt= ∑ 'En (,) r (2.17) n

The summation is performed over positive frequencies only. The same treatment can be applied to the displacement D(1) and polarization P density. In addition, each frequency component can be represented in terms of its complex amplitude:

28 −itωn Ernn(,)tecc=+ Er () .. (2.18)

The relationship between the linear displacement, D(1) , and electric field can be expressed in terms of a frequency dependent dielectric tensor according to:

(1) (1) DrErnn()=+ () 4πεω P = (n ) E(r) n (2.19)

Finally, the equation of propagation in a nonlinear medium can be written as:

ωπω224 −∇2Er() −nnεω ( ) ⋅ Er () = PNL () r (2.20) nnncc22n

Because PNL contains two or more electric field factors, a set of (n+ 1) coupled equations is needed for the nth order nonlinear optical process. For second harmonic generation, the details are neglected and only the results are presented here. Under the condition that

E(2ω) is much smaller than the input field E(ω) and considering a wave propagating in the z direction with the input face of the medium fixed at z = 0,

32 2 128πω deff 22 sin(Δkz / 2) 2 IzI2ωω= 32 (0)[ ] (2.21) cnnωω2 Δ kz/2 where Δk is the phase mismatch between the fundamental and the second harmonic

electric field. The quantities I2ω and Iω , nω and n2ω are the intensities and the index of refractions at the fundamental and second harmonic frequencies, respectively. The intensity I in a material of index of refraction n is proportional to the squared magnitude

2 1 of the electric field, I = cn E . The quantity d = χ (2) is the effective second-order 8π eff 2 coefficient for second harmonic generation.

29 2.2 Microscopic susceptibility

Organic molecules are primarily composed of carbon atoms and other atoms from the second row of the periodic table, which may give rise to two types of orbitals with contrasting properties: σ orbitals and π orbitals. The strong delocalization of the π electrons makes the π-electron distribution highly deformable in conjugated electronic systems, giving rise to large optical nonlinearities. Historically, from empirical observation, organic molecules that have extended conjugated π networks have excellent second order characteristics. Further, chromophores with a push-pull structure, i.e., possessing intra-molecular electron donor and acceptor pairs, give rise to nonlinear optical enhancements. Aromatic heteroatom substitutions in the conjugated system like nitrogen or sulfur can further increase the donor-acceptor interaction.[4, 5] Figure 2.1, adapted from reference [6], shows the basic structure of organic molecules for nonlinear optics. While empirical studies have laid the foundations for nonlinear optical response, the molecular structure and the NLO response are better explained by contemporary quantum mechanical calculations. Before we proceed with the quantum chemical analysis, it is helpful to review early models for interpreting first hyperpolarizability β.

30

Figure 2.1 Typical organic molecules for nonlinear optical effects. The electron donor group (D) is connected to the electron acceptor group (A) through a π electron system. The π electron system can be those containing benzene analogs or stilbene analogs. R1 and R2 are usually carbon or nitrogen. (Adapted from reference [6] )

2.2.1 Early models

The equivalent internal field (EIF) developed by Oudar and Chemla [7] is one of the first attempts to interpret second harmonic response in organic substances. For conjugated molecules, the strong delocalization of the π electrons could induce cooperative effects and produce very large nonlinear optical responses. However, in non-substituted molecules such as benzene, the contributions of the delocalized electrons to the first hyperpolarizability, β, vanish due to their centrosymmetric structures. The EIF model proposed that the second order response in these organic chromophores could be predicted from the ground state deformation of the π electron distribution due to electropositive or electronegative substituents. The relationship between the first hyperpolarizability β and the distortion of the π electron distribution can be expressed by:

3γΔμ β = (2.22) α

31 where α is the molecular linear polarizability, Δμ is the induced mesomeric moment and γ is the second hyperpolarizability. The EIF model correlation of β and Δμ has been verified for both mono-substituted benzenes and stilbenes.[5, 8]

Another early model for understanding β is to use additivity model to account for different contributions coming from structural elements of a molecule. In the case of the first hyperpolarizability, bond additivity is normally used instead of atomic additivity. A natural step towards the optimization of molecular nonlinearities would be to utilize two substituents whose β values are of opposite signs, connected on the same π electron bridge in order to add their effects. For instance, the first hyperpolarizability β of p- nitroaniline would be approximately equal to the sum of the β’s of aniline and nitrobenzene.

The EIF and additivity model work well for weakly coupled systems, however, they fail to explain the highly asymmetric π-organic ones. This is because in the EIF and additivity models, β is governed by ground state electron distribution and that the substituents act independently with each other. In the highly asymmetric π-organic systems such as the push-pull organic chromophores, the excited states play a major role for the nonlinear optical response.[9] . The large β in these molecules are mainly due to an intramolecular charge-transfer interaction between the acceptor and the donor. Thus the total β would be the sum of these two contributions:

β = ββadd+ CT (2.23)

where βadd is the additivity part for the different substituents, and βCT is the intramolecular charge-transfer part from the interaction of donor and acceptor . The charge-transfer term can be given by:

32 2 3e ωgef geΔμ ge βCT = 22 22 22 22 (2.24) 2(m ωge−−ωω )( ge 4 ω )

where ω is the energy of the incident photon, ωge the transition frequency between the

ground state g and excited state e. The quantity fge is the oscillator strength for ground

state to excited state transition, and Δμge is the difference between ground and excited state dipole moment. With this expression, it is now possible to relate the hyperpolarizability β to several readily available physical data for the molecules. It was

shown that βCT is responsible for the large portion of the experimentally determined β values for the charge-transfer chromophores like disubstituted polyenes [9] and isomers of nitroaniline. [10]

2.2.2 Quantum mechanical calculation

In order to calculate the nonlinear optical response of organic molecules, the

Hamiltonian and computational methods need to be specified. For the selection of the

Hamiltonian, one can take an ab initio approach to start with a complete Hamiltonian or use semi-empirical parameters. Semi-empirical calculations usually assume some sort of simplication to the full Hamiltonian and use adjustable parameters to fit the results of the real experiments. Thus the semi-empirical methods are always much faster than ab initio calculations, although probably less accurate.

For the computation methods, one can use derivative or sum-over-state methods.

The derivative method relates different derivatives of the energy and dipole moment to various terms in the power series expansions. For example, the first hyperpolarizability β

33 would be simply given by the third derivative of the energy or the second derivative of the dipole moment with respect to the applied electric field. The sum-over-state method is based on a perturbation theory developed by Ward [11] to account for the effects of the externally applied electric field on the target molecule. It expresses the polarizability and hyperpolarizabilities as a sum of dipole integrals over ground states and various excited states. Popular semi-empirical derivative quantum chemical computation methods include PPP [12, 13], Extended Huckel [14], CNDO [15], and ZINDO [16].

Based on reference [17], we can write the Sum-Over-State quantum mechanical expression for the hyperpolarizability β from time dependent perturbation theory. For a molecule interacting with an external electric field, the time evolution of the density matrix at thermal equilibrium ρ = emm−Em/ kT can be described by Liouville equation: ∑m

∂ρ 1 = [,]H ρ (2.25) ∂ti where m and Em are the eigenstates and energies of the unperturbed system. The

Hamiltonian HH=+⋅0 μ E includes a dipolar interaction μ ⋅ EqrE=⋅ which couples the electronic position r with charge q to the perturbation. Because a solution to equation

(2.25) can be obtained by successive approximation of the density matrix:

ρρ=+++(0)(1)(2)(3) ρ ρ ρ (2.26)

The expectation value of the molecular polarizability and hyperpolarizability are given by:

=[]ρρμαβγ =∑∑ < p()nn >= Trp [ () ] =< >+< >+< >+< >+ (2.27) nn==00 where μ is the permanent molecular dipole moment, α , β and γ are the linear polarizability, quadratic and cubic hyperpolarizability tensors, respectively. Substituting equation (2.26) in equation (2.25) and collecting terms of the same order:

34 ∂ρ ()n 1 mm' =+⋅[(HE ,ρμρ()nn ) ( , (− 1) ) ] (2.28) ∂ti 0'mm mm ' or

(1)n− ()n (,μρ⋅ E )mm' ρmm' = (2.29) ()ωω− mm'

(0) At normal temperature, all molecules are assumed to be in the ground state, ρgg =1 and

combining equations (2.29) and (2.27) yields the quantum mechanical expression for βijk :

−1 gllmmgμμij μ k gllmmg μμ ij μ k βωωωijk (;,312 )=+2 ∑∑ [ 2()()()() lm ωωωωgl++32 gm ωωωω gl −− 32 gm

gμμ llmmgg μ μμ llmmg μ ++kj i kj i ()()ωωωωgl++23 gm ()() ωωωω gl −− 23 gm

gllmmgμμ μ gllmmg μμ μ ++ik j ik j ()()ωωωωgl++31 gm ()() ωωωω gl −− 31 gm

gllmmggllmmgμμ μ μμ μ ++jk i jk i ()()ωωωωgl++13 gm ()() ωωωω gl −− 13 gm

gllmmggllmmgμμ μ μμ μ ++ki j ki j ()()ωωωωgl−+21 gm ()() ωωωω gl +− 21 gm

gllmmggllmmgμμ μ μμ μ ++ji k ji k](2.30) ()()ωωωωgl−+12 gm ()() ωωωω gl +− 12 gm

2.2.3 Two-level model

From the above hyperpolarizability β expression, in order to determine its nonlinear optical response, the full electronic structure of the molecule needs to be known.

35 However, as a result of the energy denominators, transitions which have large transitional moments and low energy gaps are expected to have a stronger contribution. Depending on the input optical frequency, the resonances that exist in the frequency dependence of the hyperpolarizability can be exploited to single out the contribution of individual states.

In the simplest approximation, one can consider the contribution from a single excited state [9], essentially making a two-level system. Let lme= = , and limiting the sums to just the ground (g) and excited state (e), we can get:

22 2 ωωωωωge(3 ge +−12 3 ) βωωωzzz (;,312 )= 22 22 22 2 ()()()ωge−−−ωω123 ge ωω ge ω

×−[(geeeggegμμzz μ z )] μ z

22 22 ωge(3ωωωωμμ ge+−12 3 ) ge Δ = 22 22 22 2 (2.31) ()()()ωge−−−ωω123 ge ωω ge ω

where μge and Δμ denote the transitional moment and the change in dipole moment

between the ground and the excited states. ( Δμ =−μμee gg ) The two-level approximation well describes one-dimensional push-pull conjugated chromophores near resonance, in which charge can move only along the main direction of the molecule. In the special case

of second harmonic generation, i.e. ω312= 222ωωω==, the two-level approximation for β is [9]:

22 3ωμge geΔ μ βωωωzzz (2 ; , ) = 22 2 2 2 (2.32) ()(4)ωge−−ωω ge ω

The expression (2.32) is identical with equation (2.24) and has been commonly used to

define the dispersion-free hyperpolarizability β0 and to relate measurement at different

36 wavelength [18, 19]. The frequency dependence can be separated into a product of

intrinsic hyperpolarizability β0 and a dispersion factor F(ω):

β zzz ()ωωβ= F ()0 (2.33)

2 3μgeΔμ β0 = 22 (2.34) ωge

4 ωge F()ω = 222 2 (2.35) ()(4)ωge−−ωω ge ω

With this two-level model expression, it is possible to establish trends in the relationship between β and the molecular structure. From the two-level model, β

2 increases with increasing Δμ and μge . This means as the difference in polarity between ground state and excited state increases, β is expected to increase. As the oscillator strength increases, an increase of β is also expected. A red shift of the optical absorption spectra (decreasing ωge) will also mostly result an increase in β. The model also shows that when ω or 2ω is close to the transition frequency ωge, β will be greatly enhanced. .

Investigations of chemical properties leading to large second order nonlinear optical properties have been an active area of research.[20-24]

Marder et al. [23, 25-28] used a two-state, four-orbital, independent electron analysis to compute the optimum condition for the hyperpolarizability β. They find bond length alternation (BLA: the average difference in length between single and double bonds in the molecule) is a key parameter that can be correlated with hyperpolarizability.

They used this model to examine how the three parameters in the two-level β expression

(change in dipole moment between the two states Δμ , the square of the transition

2 22 moment μge , and the inverse square of the transition energy 1/ ωge ) change with the

37 various donor and acceptor strengths in the conjugated organic molecules. An optimal combination of donor and acceptor strength for a given bridge can be obtained for a maximized hyperpolarizability β.

They also found out that an internal (with solvents of various polarity) or external electric field will change the molecular chemical structure, BLA, and electronic structures of organic polymethine and donor-acceptor polyene dyes. [25, 27] An increased solvent polarity or external electric field will drive the dye from a neutral, bond-alternated, polyene-like structure, through a symmetric cyanine-like structure, and finally to a zwitterionic (charge-separated) bond-alternated polyene like structure.

The potential-sensitive dye used in our study, di-8-ANEPPS, which is a hemicyanine dye (charged push-pull polyene), lies in the BLA D or E region, as we can see from figure 2.2 (From reference [27]) Depending on the polarity of the used solvents, the changed BLA parameter will cause the molecule have a change of hyperpolarizability

β in different solvents. Based on the analysis here, the β for di-8-ANEPPS is a negative value, so is the difference in ground and excited states dipole moment Δμ .

38

2 22 Figure 2.2 Dependence of μge (---), 1/ ωge (…), and Δμ =μμee − gg ( −−− ) and β (___) on the ground structure of one donor-acceptor polyene (From reference [27])

2.3 Relationship between microscopic and macroscopic optical nonlinearities

Similar to the expression for macroscopic polarization, the interaction of the radiation field with a molecular unit is described

pEp()ω =+μαω () ⋅ () ω +NL () ω (2.36) and

NL pEEEEE()ω= β (−+−+ ωω ;12 , ω ):( ω 1 )( ω 2 ) γ ( ωω ; 123 , ω , ω ) ( ω 1 )( ω 2 )( ω 3 ) (2.37)

39 where μ is the permanent molecular dipole moment, and α , β and γ are the linear polarizability, quadratic and cubic hyperpolarizability tensors, respectively. In organic substances the molecules interact through van der Waals forces or hydrogen bonds, which are one or two orders of magnitude weaker than the chemical bonds holding the atoms together within the molecules. Therefore, the molecules conserve their individuality in the condensed medium and they can be considered as the microscopic building units of the bulk medium. Then it is possible to obtain the nonlinear optical properties of the bulk medium from those of the constituent molecules by additivity, although great care needs to be taken for second order nonlinear process such as symmetry, orientation, and local field factors. A centrosymmetric, isotropic orientation in the bulk medium will lead to a zero χ(2) no matter how large the β is for the constituent molecule. Also, the local field factor can strongly affect the response of the individual molecule to the external electric field.

2.3.1 Oriented gas model

In 1982, Oudar and Zyss [29] used an oriented gas model to analyze the relationship between the hyperpolarizabilities of the molecular constituents and the macroscopic nonlinear coefficients for organic crystals. Singer et al. [30] extended the analysis to less ordered materials such as liquid crystals and polymers. In general, the

(2) macroscopic second-order coefficient χ IJK can be expressed in terms of the molecular microscopic βijk by:

(2)ω ωω12 * χIJK(;,)−= ωωω12Nf I f J f K <−> β IJK (;,) ωωω 12 ijk (2.38)

40 * where N is the number density and < β IJK > is an orientational average of the molecular first hyperpolarizability. The indices i, j, and k define the coordinate systems of the macroscopic materials, while I, J, and K define those of the molecules. The two coordinate systems are related by Euler angles. Equation (2.38) can be rewritten as:

(2) ω ωω12 χ IJK(;,)−=ωω12 ωNf I f J f K <>− R Ii R Jj R Kk β ijk (;,) ωω 12 ω

=ΩΩ−Nfω fωω12 f(())(;,) d G R R R β ωω ω (2.39) IJ K∫ IiJjKkijk 12

Here RIi are the elements of transformation matrix that rotates β from molecular to laboratory frame. G (Ω ) is a statistical orientation distribution function. β and G (Ω ) are the important parameters that define the performance of the macroscopic systems.

The f’s are the local field factors, which relate the externally applied electric field to the local electric field actually acting on the individual molecules in the bulk phase.

Under different circumstances, the frequency dependent local field factors are given by the Lorenz-Lorentz or Onsager models. [31]

Under an external electric field, considering the effects of an induced dipole, the electric field inside a dielectic sphere is a Lorentz local field. The Lorenz-Lorentz correction factor is commonly used in determining the molecular hyperpolarizability β of solutions and liquids. Its expression for the local field correction is:

ε + 2 f ()ω = ω (2.40) 3

2 where εω is the dielectric constant which is equals to n , n is the refractive index.

The Lorentz local field model applies to nonpolar molecules, while the Onsager local field model applies to polarizable dipole molecules. Taking into account the dipolar

41 reaction field in mixtures of polar liquids, the local field factors for this process are known as the Onsager local field corrections and are given by:

22 nn()[()ωωi + 2] fi ()ω = 22 (2.41) 2()nnω + i ()ω where n(ω) is the refractive index of the mixture and ni(ω) is the refractive index of species i.

2.3.2 In Langmuir-Blodgett films

Now let us consider a special case for the relationship between microscopic and macroscopic optical nonlinearities. Langmuir-Blodgett (LB) films are controlled organized monolayers of amphiphilic molecules which are transferred from the air-water interface onto a solid substrate. LB films have potential applications such as sensors, detectors,[32] biological membranes, [33] and nonlinear optical materials.[34-36]

Chapter 5 will discuss the details of preparation and characterization of these layers. Here the relationships between the nonlinear optical properties of the constituent molecules and films will be reviewed.

Figure 2.3 A Langmuir-Blodgett monolayer with an average tilt angle α relative to the surface normal

42 For a LB monolayer with an average tilt angle α to the surface normal (See figure

2.3), an azimuthal distribution φ, the nonlinear susceptibility for second harmonic generation is given by:

(2) χωωωs,IJK= Nf s I() f J () f K (2) b IJK (2.42) where the subscript s refers to a surface layer. The quantity Ns is the surface concentration and bIJK is the hyperpolarizability in the macroscopic frame. If we assume the amphiphilic molecule has a rod-shape and is highly polarizable, then βzzz is the dominant component of the molecular hyperpolarizability tensor. If we also assume that the distribution of angles, α, is relatively sharp, and φ is assumed to be a uniform distribution, then the two values of bIJK can characterize the film [37]:

3 bzzz= β zzz (2.43) and

<>cosαα sin2 b = β (2.44) p() zxx2 zzz where p indicates identical expressions for the permutations of the indices. Thus, the two

(2) independent components of χs are given by:

(2) 3 χ s,zzz= zzz (2.45) and

(2)1 2 χ s,(pzxx )=<2 Nf s'cossinααβ > zzz (2.46) where f and f’ are the local field factors. In order to determine the value of α, the ratio of the two independent tensor components need to be measured, assuming we know f and f’ or make assumption that they are close to unity. By defining a ratio A as:

43 (2) 2χ spzxx,( ) A = (2) (2) (2.47) χχs,,()zzz+ 2 s p zxx

which is related to the molecular orientation by:

< cosα sin 2 α > A = (2.48) <>cosα

If the distribution of the tilt angle α is sharply peaked, then

α sin−1 (A ) (2.49)

Therefore, second harmonic generation is a powerful tool for gaining insight into the structure and order in LB films.

2.4 Measurement Techniques [38]

The second-order susceptibility can be measured both absolutely and relatively by several techniques. The absolute methods include the phase matched method [39] and parametric fluorescence method. [40] Relative methods for characterizing materials with respect to a standard material include the Maker fringe method [39, 41, 42] and the power method. [43] The Maker fringe method is one of the most useful methods for determining the value of nonlinear coefficients. In chapter 5, Langmuir-Blodgett films consisting of di-8-ANEPPS will be characterized using this method.

It is more convenient and appropriate to characterize the individual molecules of the material rather than the bulk material. This is because the bulk optical properties of organics are largely determined by the individual molecular units. Various techniques have also been employed to measure the molecular hyperpolarizability. The techniques that are most widely used are electric field induced second harmonic generation (EFISH)

44 [10, 44, 45] and hyper Rayleigh scattering (HRS).[46, 47] In chapter 3 and 4, the principles of HRS and experimental details will be explained and measurements of molecular hyperpolarizability of di-8-ANEPPS will be performed using HRS.

Solvatochromic effects in solution have also been used as another measurement technique for molecular hyperpolarizability. [48, 49] It has been utilized in chapter 4 to analysis the solvent effects in the potential sensitive dyes.

45 References

1. Boyd, R.W., Nonlinear Optics. Second Edition ed. 2003: Academic Press.

2. Shen, Y.R., The Principles of Nonlinear Optics. 1984, New York: John Wiley & Sons, Inc.

3. Bloembergen, N., Nonlinear Optics. 1965, New York: W.A. Benjamin.

4. Kanis, D.R., M.A. Ratner, and T.J. Marks, Design And Construction Of Molecular Assemblies With Large 2nd-Order Optical Nonlinearities - Quantum- Chemical Aspects. Chemical Reviews, 1994. 94(1): p. 195-242.

5. Prasad, P.N. and D.J. Williams, Introduction to Nonlinear Optical Effects in Molecules and Polymers. 1991: John Wiley & Sons, Inc.

6. Bosshard, C., M. Bosch, I. Liakatas, M. Jager, and P. Gunter, Second-Order Nonlinear Optical Organic Materials: Recent Developments. Nonlinear Optical Effects and Materials, ed. P. Gunter. 2000: Springer.

7. Oudar, J.L. and D.S. Chemla, Theory of Optical Second-Order Susceptibilities of benzene substitutes. Opt. Comm., 1975. 13: p. 164-168.

8. Zyss, J., Hyperpolarizabilities of substituted conjugated molecules. II. Substituent effects and respective – contributions. Journal Of Chemical Physics, 1979. 70(7): p. 3341.

9. Oudar, J.L. and D.S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state dipole moment. Journal Of Chemical Physics, 1977. 66(6): p. 2664.

10. Oudar, J.L., Optical nonlinearities of conjugated molecules. Stilbene derivatives and highly polar aromatic compounds. Journal Of Chemical Physics, 1977. 67: p. 446-457.

11. Ward, J.F., Calculation of Nonlinear Optical Susceptibilities Using Diagrammatic Perturbation Theory. Reviews of Modern Physics, 1965. 37: p. 1.

12. Pople, J.A., Electron interaction in unsaturated hydrocarbons. Transactions of the Faraday Society, 1953. 49: p. 1375.

13. Pariser, R. and R.G. Parr, A Semi-Empirical Theory of the Electronic Spectra and Electronic Structure of Complex Unsaturated Molecules. I. Journal Of Chemical Physics, 1953. 21(3): p. 466.

46 14. Hoffmann, R., An Extended Hückel Theory. I. Hydrocarbons. Journal Of Chemical Physics, 1963. 39(6): p. 1397.

15. Pople, J.A., D.P. Santry, and G.A. Segal, Approximate Self-Consistent Molecular Orbital Theory. I. Invariant Procedures. Journal Of Chemical Physics, 1965. 43: p. S129.

16. Zerner, M., Semiempirical Molecular Orbital Methods. Reviews in Computational Chemistry, ed. K.B. Lipkowitz and D.B. Boyd. Vol. 2. 1991.

17. Cheng, L.T., Second-Order Nonlinear Optical Properties of Organic Materials. Handbook of Photonics, ed. M.C. Gupta. 1996: CRC Press.

18. Kaatz, P. and D.P. Shelton, Polarized hyper-Rayleigh light scattering measurements of nonlinear optical chromophores. Journal Of Chemical Physics, 1996. 105(10): p. 3918-3929.

19. Zyss, J., T.C. Van, C. Dhenaut, and I. Ledoux, Harmonic Rayleigh-Scattering From Nonlinear Octupolar Molecular Media - The Case Of Crystal Violet. Chemical Physics, 1993. 177(1): p. 281-296.

20. Kenis, P.J.A., O.F.J. Noordman, S. Houbrechts, G.J. van Hummel, S. Harkema, F. van Veggel, K. Clays, J.F.J. Engbersen, A. Persoons, N.F. van Hulst, and D.N. Reinhoudt, Second-order nonlinear optical properties of the four tetranitrotetrapropoxycalix[4]arene conformers. Journal Of The American Chemical Society, 1998. 120(31): p. 7875-7883.

21. Whitaker, C.M., E.V. Patterson, K.L. Kott, and R.J. McMahon, Nitrogen and oxygen donors in nonlinear optical materials: Effects of alkyl vs phenyl substitution on the molecular hyperpolarizability. Journal Of The American Chemical Society, 1996. 118(41): p. 9966-9973.

22. Moylan, C.R., R.J. Twieg, V.Y. Lee, S.A. Swanson, K.M. Betterton, and R.D. Miller, Nonlinear-Optical Chromophores With Large Hyperpolarizabilities And Enhanced Thermal Stabilities. Journal Of The American Chemical Society, 1993. 115(26): p. 12599-12600.

23. Marder, S.R., D.N. Beratan, and L.T. Cheng, Approaches For Optimizing The 1st Electronic Hyperpolarizability Of Conjugated Organic-Molecules. Science, 1991. 252(5002): p. 103-106.

24. Marder, S.R., J.W. Perry, and W.P. Schaefer, Synthesis Of Organic Salts With Large 2nd-Order Optical Nonlinearities. Science, 1989. 245(4918): p. 626-628.

25. Marder, S.R., C.B. Gorman, F. Meyers, J.W. Perry, G. Bourhill, J.L. Bredas, and B.M. Pierce, A Unified Description Of Linear And Nonlinear Polarization In Organic Polymethine Dyes. Science, 1994. 265(5172): p. 632-635.

47 26. Marder, S.R., L.T. Cheng, B.G. Tiemann, A.C. Friedli, M. Blancharddesce, J.W. Perry, and J. Skindhoj, Large 1st Hyperpolarizabilities In Push-Pull Polyenes By Tuning Of The Bond-Length Alternation And Aromaticity. Science, 1994. 263(5146): p. 511-514.

27. Bourhill, G., J.L. Bredas, L.T. Cheng, S.R. Marder, F. Meyers, J.W. Perry, and B.G. Tiemann, Experimental Demonstration Of The Dependence Of The 1st Hyperpolarizability Of Donor-Acceptor-Substituted Polyenes On The Ground- State Polarization And Bond-Length Alternation. Journal Of The American Chemical Society, 1994. 116(6): p. 2619-2620.

28. Gorman, C.B. and S.R. Marder, An Investigation Of The Interrelationships Between Linear And Nonlinear Polarizabilities And Bond-Length Alternation In Conjugated Organic-Molecules. Proceedings Of The National Academy Of Sciences Of The United States Of America, 1993. 90(23): p. 11297-11301.

29. Zyss, J. and J.L. Oudar, Relations Between Microscopic And Macroscopic Lowest-Order Optical Nonlinearities Of Molecular-Crystals With One- Dimensional Or Two-Dimensional Units. Physical Review A, 1982. 26(4): p. 2028-2048.

30. Singer, K.D., M.G. Kuzyk, and J.E. Sohn, 2nd-Order Nonlinear-Optical Processes In Orientationally Ordered Materials - Relationship Between Molecular And Macroscopic Properties. Journal Of The Optical Society Of America B-Optical Physics, 1987. 4(6): p. 968-976.

31. Bottcher, C.J.F., Theory of Electric Polarization. Vol. 1. 1973: Elsevier.

32. Roberts, G., Langmuir-Blodgett Films, ed. G. Roberts. 1990, New York: Plenum Press.

33. Sackmann, E., Supported membranes: Scientific and practical applications. Science, 1996. 271(5245): p. 43-48.

34. Marowsky, G., A. Gierulski, R. Steinhoff, D. Dorsch, R. Eidenschnik, and B. Rieger, Efficiency Studies Of 2nd-Harmonic-Active Organic-Dye Coverages. Journal Of The Optical Society Of America B-Optical Physics, 1987. 4(6): p. 956-961.

35. Girling, I.R., N.A. Cade, P.V. Kolinsky, J.D. Earls, G.H. Cross, and I.R. Peterson, Observation Of 2nd-Harmonic Generation From Langmuir-Blodgett Multilayers Of A Hemicyanine Dye. Thin Solid Films, 1985. 132(1-4): p. 101-112.

36. Girling, I.R., P.V. Kolinsky, N.A. Cade, J.D. Earls, and I.R. Peterson, 2nd Harmonic-Generation From Alternating Langmuir-Blodgett Films. Optics Communications, 1985. 55(4): p. 289-292.

48 37. Heinz, T.F., H.W.K. Tom, and Y.R. Shen, Determination Of Molecular- Orientation Of Monolayer Adsorbates By Optical 2nd-Harmonic Generation. Physical Review A, 1983. 28(3): p. 1883-1885.

38. Sutherland, R.L., Handbook of Nonlinear Optics. 2003: Marcel Dekker, Inc.

39. Kurtz, S.K., Measurement of nonlinear optical susceptibilities. Quantum Elecronics, ed. H. Rabin and C.L. Tang. Vol. I. 1975, New York: Academic Press. 209-281.

40. Byer, R.L. and S.E. Harris, Power and bandwidth of spontaneous parametric emission. Physical Review, 1968. 168: p. 1064.

41. Jerphagnon, J. and S.K. Kurtz, Maker Fringes: A Detailed Comparison of Theory and Experiment for Isotropic and Uniaxial Crystals. Journal Of Applied Physics, 1970. 41: p. 1667-1681.

42. Maker, P.D., R.W. Terhune, M. Nisenoff, and C.M. Savage, Effects of dispersion and focusing on the production of optical harmonics. Physical Review Letters, 1962. 8: p. 21.

43. Kurtz, S.K. and T.T. Perry, A power technique for the evaluation of nonlinear optical materials. Journal Of Applied Physics, 1968. 39: p. 3798.

44. Singer, K.D. and A.F. Garito, Measurements Of Molecular 2nd Order Optical Susceptibilities Using Dc Induced 2nd Harmonic-Generation. Journal Of Chemical Physics, 1981. 75(7): p. 3572-3580.

45. Levine, B.F. and C.G. Bethea, Second and third order hyperpolarizabilities of organic molecules. Journal Of Chemical Physics, 1975. 63: p. 2666-2682.

46. Clays, K. and A. Persoons, Hyper-Rayleigh Scattering In Solution. Review Of Scientific Instruments, 1992. 63(6): p. 3285-3289.

47. Clays, K. and A. Persoons, Hyper-Rayleigh Scattering In Solution. Physical Review Letters, 1991. 66(23): p. 2980-2983.

48. Bosshard, C., G. Knopfle, P. Pretre, and P. Gunter, 2nd-Order Polarizabilities Of Nitropyridine Derivatives Determined With Electric-Field-Induced 2nd-Harmonic Generation And A Solvatochromic Method - A Comparative-Study. Journal Of Applied Physics, 1992. 71(4): p. 1594-1605.

49. Paley, M.S., J.M. Harris, H. Looser, J.C. Baumert, G.C. Bjorklund, D. Jundt, and R.J. Twieg, A Solvatochromic Method For Determining 2nd-Order Polarizabilities Of Organic-Molecules. Journal Of Organic Chemistry, 1989. 54(16): p. 3774-3778.

49 Chapter 3: Spectroscopic studies and the first hyperpolarizability of di-8-ANEPPS

After a brief introduction to styryl dyes, the linear and nonlinear optical properties of di-8-ANEPPS are discussed. Its absorption spectrum shifts to the blue and its fluorescence emission spectrum shifts to the red with solvents of high polarity. This anomalous solvatochromism has been discussed within a Monopole-Dipole model.

Another two-valence bond state model has also been applied to explain this anomalous solvatochromism and the dependence between solvent polarity and molecular hyperpolarizability. Time-correlated single photon counting hyper-Rayleigh scattering

(TCSPC-HRS) has been used to determine the hyperpolarizability component βzzz at 800 nm for di-8-ANEPPS. Experiments are in agreement with quantum chemical calculations.

The hyperpolarizability allows a determination of the value of the difference in dipole moment between the lowest excited and ground states. This quantity can be used to estimate the electrochromic response. Details of the TCSPC-HRS technique are also described.

50 3.1 A brief introduction to styryl dyes

Loew and his coworkers pioneered the design and characterization of potentiometric dyes based on electrochromism. The dyes developed in his lab are in the structure class called styryl dyes. [1] These are amphiphilic membrane staining dyes which usually have a pair of hydrophobic hydrocarbon tails acting as membrane anchors and a hydrophilic group which aligns the chromophore perpendicular to the membrane and aqueous interface. The chromophore is believed to undergo a large electronic charge shift as a result of excitation from the ground state to the excited state, which underlies the electrochromic mechanism relating the sensitivity of these dyes to membrane potential. The ASP ((aminostyryl)pyridinium) dyes developed by Loew [2] demonstrated the successful design of electrochromic membrane probes. Later ANEP

(aminonaphthylethenylpyridinium) dyes [3] such as di-4-ANEPPS and di-8-ANEPPS become popular in various applications because of their large, consistent response in various membrane systems.

di-5-ASP di-8-ANEPPS

Figure 3.1 Structures of two styryl dyes, molecular structures are adapted and modified from reference [1, 2]

Figure 3.1 shows the structure difference of a typical ASP dye di-5-ASP and an

ANEP dye di-8-ANEPPS. The chromophore of ANEP probe is lengthened compared to

51 the ASP series by the substitution of a naphthalene ring system in the place of a phenyl ring. This has the effect of lengthening the path for the excitation induced charge shift without introducing any additional flexibility to the chromophore. From the two-level model analysis [4] for the first hyperpolarizability β of charge-transfer dyes, we can find that dyes that have extended conjugated π networks, aromatic heteroatom (nitrogen or sulfur) substitution, and electron donor/acceptor pairs and large change in dipole moment between the ground and excited states will tend to have large β values. The ANEP chromophore with a strongly donating dialkyamino group and an electron accepting pyridinium nucleus certainly satisfies all of these criteria. Research work [5] does show that di-8-ANEPPS has a very large second-order response, about an order of magnitude larger than that of Rhodamine 6G [6], making it an excellent probe for second harmonic studies.

3.2 Spectroscopic studies of di-8-ANEPPS

Before we describe the studies of the nonlinear optical properties of di-8-

ANEPPS, it is worthwhile noting that the linear optical properties of this molecule are unusual. When di-8-ANEPPS is dissolved in a solution, with increasing solvent polarity, the absorption spectrum is blue-shifted, while the fluorescence emission spectrum is red- shifted. Such behavior is VERY unusual. First, the absorption and emission spectra shift in opposite direction; It is unusual as such opposite shifts of absorption and emission spectra are unexpected for absorption and emission processes involving the same excited electronic states. Second, the absorption spectra shifts to the blue with increasing solvent polarity, which is rather uncommon for hemicyanines. (Charged push-pull polyenes

52 which are unsaturated organic compounds that contain alternating single and double carbon-carbon bonds) In order to rationalize this anomalous solvatochromism, Fromherz developed a monopole-dipole model, which is based on the Born-Marcus type theory of reversible charging for positive chromophores alone, disregarding the negative sulfonate.

[7] Recently Laage et al. [8] also developed a theoretical description for hemicyanine molecules in solvents of different polarity. They used a two valence bond state model to explain the unusual spectroscopic behavior and provide a treatment of the nonlinear optical properties of hemicyanine dyes. They pointed out that actually these two are intimately connected: A corrected treatment of the absorption and its solvatochromism is required to account for the nonlinear optical properties and their environmental sensitivity. Here we are going to use these models to analyze the spectroscopic properties of di-8-ANEPPS.

3.2.1 Materials and methods The di-8-ANEPPS used here was purchased from Invitrogen and also synthesized by a procedure described by Loew and coworkers. [9] In order to study the optical properties of di-8-ANEPPS as a function of solvent polarity, two solvents of different polarity, dimethyl sulfoxide (DMSO) and chloroform, were mixed in various ratios. The di-8-

ANEPPS was first dissolved in a stock solution of DMSO at a concentration of 0.1mg/ml and then diluted to 0.02mg/ml into five solutions: A: pure DMSO solvent; B: 80%

DMSO and 20% Chloroform; C: 60% DMSO and 40% Chloroform; D: 40% DMSO and 60% Chloroform; E: 20% DMSO and 80% Chloroform.

53

Table 3-1 Solution properties. F is the polarity, ε the dielectric constant, and n the refractive index. Data for mixtures are calculated from algebraic sum of the cosolvent fractions using ε and n of chloroform as 4.81 and 1.4458, respectivelya.

F, a polarity indictor in solvents, which is a function of solvent dielectric constant and refractive index, can be written as: [10]

2 1 εii−−n εε Fn(,)ε =− (2 ) (3.1) εiiεεε++22n i

εi , an empirical parameter, is taken as the intermediate dielectric constant. (εi = 2) Based on the refractive index and dielectric constant data of DMSO and chloroform, a series of solutions of controlled polarity can be formed, as shown in Table 3.1. The dielectric constants and refractive indices of mixtures in the table were calculated by assuming linear scaling by fraction.

UV-VIS absorption spectra were obtained by using a Cary 5000 UV-Vis-NIR spectrophotometer. The fluorescence emission spectra were obtained using a PTI 841 spectrofluorometer. For fluorescence emission, the excitation wavelength is 450nm and the fluorescence is collected from 460nm to 950nm. The absorption spectra and

54 fluorescence emission spectra were all fitted with a Log-normal function [11] in order to find their peak position. Figure 3.2 shows the absorption spectrum, fluorescence spectrum and their corresponding Log-normal fits of di-8-ANEPPS in DMSO solution.

200000 1.4

1.2 160000 1.0

120000 0.8

0.6 80000

0.4 40000

0.2 Absorption (O.D.) Fluorescence Emission

0 0.0 12000 14000 16000 18000 20000 22000 24000 26000 Wavenumber (cm-1)

Figure 3.2 Absorption and fluorescence emission spectra of di-8-ANEPPS in DMSO solution. The concentration is 0.02mg/ml and the cell length is 10mm. Two dashed lines are the absorption and emission spectra, respectively, while the lines are the Log-normal fits.

3.2.2 Results

The wavenumber vabs and vem of the maxima of absorption and emission in the solvent mixtures are shown in Figure 3.3 for di-8-ANEPPS. We see that the absorption is shifted to the blue and the emission is shifted to the red with increasing polarity of the solvent. Based on the linear fit as a function of F, the slope for the absorption maxima

55 is (36±× 2) 1041cm− , the slope for emission maxima is (351)10−±×41cm− , we can conclude that the solvatochromism is almost symmetrical. The average of the

wavenumber of absorption and emission (vvabs+ em ) / 2 for di-8-ANEPPS in different solvent polarities is 16800± 20cm−1 , which is exactly the same value for di-4-ANEPPS.

[7]

20000

) 18000 -1 cm ( 16000

14000 Abs/Emi

12000 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 F

Figure 3.3 Wavenumbers of the maxima of absorption vabs (upper squares) and of emission vem (lower dots) for di-8-ANEPPS versus the polarity function F for mixtures of

DMSO and Chloroform. The average values of (vvabs+ em ) / 2 are marked by stars. They are all linearly fitted.

The values of (vvabs− em ) / 2 are also plotted in Figure 3.4. The Stokes shift also increases linearly with the polarity function F,.

56 3600

) 3400 -1

cm 3200 (

3000

2800

2600 Wavenumber

2400 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 F

Figure 3.4 Half of the stokes shift (vvabs− em ) / 2 of the maxima of absorption and emission for di-8-ANEPPS versus the polarity function F. Data are linear fitted with a slope of 36000± 1000 cm-1 and an intercept of - 4200± 200cm−1

3.2.3 Discussion

As we can see, the symmetry of solvatochromism, i.e., (vvabs+ em ) / 2 remains constant over a range of solvent polarity. Based on the framework of Born-Marcus theory

(Monopole-Dipole Model),[7] this is possible only if the absolute value of charge and the

dipole moment is identical in the ground state and excited state; i.e. qqEG= and

μEG= μ , where qE and qG, μE and μG are the charges and dipole moments in the ground state and excited state, respectively.

Although di-8-ANEPPS is a zwitterionic molecule which is electrically neutral, in order to use the Monopole-Dipole model, the counterion sulfonate moiety is ignored in

57 the solvatochromism. This implies that the solvation shell of the sulfonate remains unchanged upon excitation and that all re-solvation refers to the positively charged pyridinium in di-8-ANEPPS. In this case, the Monopole-Dipole model for di-8-ANEPPS looks like this: A sphere of radius a is thought to be located around the center of the chromophore. In the ground state, an elementary charge e0 is located eccentrically at a distance +δ/2 toward the pyridinium. In the excited state, the charge is located eccentrically at a distance -δ/2 toward the naphthylamine. These two charge distributions may be represented as superpositions of a constant quadrupole with two charges e0/2 at position +/- δ/2 and of a dipole with charge +e0/2 at position +δ/2 and -e0/2 at position -

δ/2 in the ground state and of a reversed dipole in the excited state with exchanged positions of the charges. The quadrupole can be replaced by an invariant monopole qE = qG = e0 in the center of the sphere and the dipoles are approximated by point dipole μGx=⋅eeδ 0 /2and μEx=−eeδ ⋅ 0 /2in the center if the off-center displacement is small comparing with the radius of the sphere,. ( ex is a unit vector in the x direction, pointing to pyridinium)

The Stokes shift (vvabs− em ) / 2 reflects the change of the reorganization energy:

2 vvabs−− em()μμ E G =+3 F(,) nε const . (3.2) 24πε0hca where h is the Plank constant, c is the speed of light, and a is defined previously the radius of the molecule sphere. μE and μG can be expressed by the charge shift δ. With the

2 fine structure constant ehc00/ 2εα== 1/137 , the Stokes shift can be rewritten:

vv− αδ2 abs em =+F(,) nε const . (3.3) 22π a3

58 The Stokes shift depends on the extent of intramolecular charge shift. Here we can estimate the radius of the spherical model for di-8-ANEPPS by using the quantum

chemical computations of charge shift δ. Based on the slope for the Stokes shift Δν sol equals to 36000± 1000 cm-1 for di-8-ANEPPS,δ 23/ a is calculated to be 3.1. If we use

0.33nm for δ [7], (This is for di-4-ANEPPS, di-8-ANEPPS should have a similar value) the radius of the sphere is 0.33nm, which is smaller compared to the di-4-ANEPPS value of 0.49nm. Our different fitting mechanism might explain the difference, as the intercept is fixed for di-4-ANEPPS by Fromherz [7] while for our experiment, the slope and the intercept are both variables. Another reason might be the limited range of our solvent polarities, since it only varies from 0.18 to 0.22.

It is possible to relate the electrochromic and solvatochromic effects of the potential-sensitive chromophores. [12] When a dye with charge shift δ is placed in an external electric field, its absorption and emission bands will be shifted by a linear Stark shift according to electrochromism. The amphiphilic dye can be bound to the outer surface of cell membranes so that their charge shift δ has a component δcosθ along the membrane normal. (θ is the dye tilt angle) The electrochromic spectral shift can be expressed as

e Δ=ΔΨν 0 δθcos (3.4) electro hc where ΔΨ is the transmembrane potential, e0 is the monopole charge, h is the Plank constant, c is the speed of light. By using the slope from the solvatochromic effect,

22 e0 δ Δ=ν sol 3 , a general relationship between electrochromic and solvatochromic 4πε0hca effect can be written as:

59 4πε Δ=ΔΨΔν 0 νθa3/2 cos (3.5) electrohc sol

However, there are still two uncertainties with a quantitative prediction of electrochromism from solvatochromism. One is the determination of a, the radius of the sphere, and the other is the tilt angle θ.

Laage et. al [8] employed a two-valence bond state model for the anomalous solvatochromism. By including a nonequilibrium solvation (Frank-Condon transition) and the geometrical coordinate associated with a change in single and double bond alternation for the ground and excited states, they managed to explain the blue shift for the absorption spectra and red shift for the fluorescence emission spectra for hemicyanine dyes with increasing solvent polarity. In their study, a hemicyanine dye with similar donor and acceptor size is assumed and the counterion is also ignored. Before the excitation, the solvent configuration is in equilibrium with the charge distribution of the ground state molecule. Upon excitation from the electronic ground state to the excited state, the charge transfer occurs from one endgroup to the other. Thus after the excitation, the solvent configuration is out of equilibrium with the new charge distribution of the

Frank-Condon excited state. Hence, the absorption transition energy is a sum of two contributions: the energy gap between the equilibrated adiabatic ground and excited state

adia ex Veq and the adiabatic solvent reorganization energy in the excited state Λs . (Illustrated in Figure 3.5, adapted from reference [8]) Similarly, for fluorescence, the emission

adia transition energy is the difference between the adiabatic equilibrium energy Veq and the

g solvent reorganization energy, but this time in the adiabatic ground state Λs , so that:

adia ex EVabs=+Λ eq s (3.6)

60 and

adia g EVem=−Λ eq s (3.7) where Eabs and Eem are absorption and emission transition energies, respectively.

Since the two endgroups have similar sizes, the electrostatic interaction energy

adia with the solvent is approximated equal, so that the adiabatic equilibrium energy gap Veq will not change with solvent polarity. However, the solvent reorganization energy will increase for solvents of higher polarity, both for the ground state and excited states. Thus, when the solvent polarity increases, we see that the absorption energy increases while the emission energy decreases, i.e., the absorption spectrum exhibits a blue shift and the

fluorescence spectrum exhibits a red shift. It also indicates that (vvabs+ em ) / 2 remains

constant over a range of solvent polarity and Stokes shift (vvabs− em ) / 2 increases by solvents of high polarity. For normal uncharged push-pull polyene, it is the energy gap

adia between the equilibrated adiabatic ground and excited states Veq that changes dominantly. So for the normal uncharged push-pull polyene, a more polar solvent shifts both the absorption and emission spectra in the same direction, and keep the Stokes shift as a constant.

61

Figure 3.5 Schematic pictures of the electronically adiabatic ground state and excited adia state free energy surface in the solvent coordinate. The energy gap Veq between the equilibrated ground state and excited state and the solvent reorganization energies in the g ex adiabatic ground state and excited state Λs and Λs are indicated. Adapted from reference [8]

The first hyperpolarizability β in solvents of different polarity (Data measurement in Chapter 4) versus dielectric constant is plotted in Figure 3.6. The hyperpolarizability β was measured at five wavelengths selected with minimum two photon excited fluorescence. The first hyperpolarizability β decreases with dielectric constant, which also agrees with the prediction by the two-valence bond model. [8]

Laage et. al’s two-valence bond state model is not as limited as Fromherz’s monopole-dipole model assuming a single spherical cavity for the molecule. For di-8-

ANEPPS, the single spherical cavity is not very appropriate as the molecule has a long bridge between the donor and the acceptor moieties, and while it sits in the membrane,

62 the two endgroups will be surrounded by media with different dielectric constants. Thus the two-valence bond state model opens a way for modeling di-8-ANEPPS or other similar styryl dyes in biological membranes. Details of the two-valence bond model can be found in references. [8, 13, 14]

800nm 900nm 1020nm 3500 1064nm 976nm 3000 ) 2500 esu

2000 -30 10 (

1500 β

1000

10 15 20 25 30 35 40 45 50 Dielectric constant

Figure 3.6 First hyperpolarizability β of di-8-ANEPPS versus dielectric constant. Five wavelengths have been measured for the molecule: Square: 800nm, Circle: 900nm, diamond: 976nm, triangle: 1020nm and star: 1064nm. The lines are not fits, they are only guides for the eye.

3.3 The first hyperpolarizability of di-8-ANEPPS

3.3.1 Introduction Di-8-ANEPPS has been extensively used as a potential sensitive dye due to changes in its optical properties as a function of the local electric field. [1, 15]

Ratiometric fluorescence measurements have been used to detect membrane potential

63 changes with the help of di-8-ANEPPS. [16] More recently, it has been found that probing membrane potential using second harmonic generation with di-8-ANEPPS provides better sensitivity as well as spatial and temporal resolution, while reducing photo bleaching and photo toxicity.[17-19] Considerable work has been carried out to study the mechanism of electric field sensitivity in probe dyes [2, 18-24] and the hyperpolarizability β values of similar potential sensitive dyes have been measured in a monolayer on the surface of water. [5] It is well-known that the electronic electrochromic response as well as solvatochromism is related to the change in dipole moment between the ground and excited states. In the two-level model, [4] the hyperpolarizability is also dependent on the same quantity. Therefore, measurements of the molecular hyperpolarizability in solution are helpful in not only describing second harmonic generation, but can also be used to gain insight into the electric field response as well. To this end, we report on such measurements in this chapter.

Hyper-Rayleigh Scattering (HRS) has been a powerful technique for extracting the first-order hyperpolarizability β. [25, 26] Its capability of measuring both polar and non-polar materials as well as charged and uncharged materials gives it a distinct advantage over techniques such as electric field induced second harmonic generation

(EFISH). However, multiphoton induced fluorescence is always a competing process in the measurement, and this is especially an issue in di-8-ANEPPS as it is highly fluorescent. The fluorescent emission spectrum could overlap with second harmonic generation (SHG) making it difficult to separate the two spectrally.

Two techniques have been introduced to take advantage of the fact that the fluorescent signal and SHG are not emitted simultaneously. Clays et al have developed a

64 technique in the Fourier domain using synchronous detection. [27] In addition, time- domain separation was introduced by Noordman and Van Hulst. [28] Using time- correlated single photon counting (TCSPC) technology, we have improved on the technique of Noordman and Van Hulst to measure the hyperpolarizability using hyper-

Rayleigh scattering (TCSPC-HRS). TCSPC measurements can not only separate second harmonic from fluorescence, but as a highly sensitive photon counting technique it yields a high signal to noise ratio with a Ti:Sapphire pump.

3.3.2 Experiment and Results A mode-locked Ti:Sapphire laser was used as the source. The subpicosecond pulses are more than adequate to allow subnanosecond time resolution in our TCSPC apparatus.

Since the fluorescence lifetime can be as long as 20ns, a low repetition rate is required so that the temporal separation between pulses exceeds the fluorescent lifetime. [29] To this end, we employed a “home-built” extended cavity, mode-locked Ti: sapphire laser system with a 6.6 meter cavity as shown in Figure 3.7. The output arm is about 4 times longer than the back mirror arm, making it highly asymmetric, which facilitates mode- locking. With a 6.7W pump (Argon laser), 0.8W of output can easily be generated. This system is mode-locked at 800nm with a pulse duration of about 120fs. The repetition rate is 45MHz, which gives 22ns duty time, long enough to allow for the decay of the fluorescence signal between laser pulses.

65 Extended arm

OC Pump LCM

TiSa crystal

M Prisms

Figure 3.7 Extended-cavity mode-locked Ti:Sapphire laser layout. The 6.66W pump is provided by an Argon laser at 488nm. M: mirror; L: lens; CM: intracavity mirror

(transparent to pump); OC: output coupler.

In the TCSPC-HRS setup shown in Figure 3.8, the TCSPC module consists of a fast multi-time-channel chip, TimeHarp200 (from PicoQuant, Inc.), a photon counting photomultiplier tube (PMA185 from PicoQuant, Inc.), and an avalanche photodetector.

In principle, the TCSPC operates when the laser output triggers the START input of the

TCSPC module, and the arrival of a single photon at the photomultiplier tube stops the timer at the STOP input. A histogram of single photon arrival times is produced. This requires that the STOP pulse is always correlated with a single photon event. In order to ensure that only a single photon event is detected, the photon count rate must be well below one photon per laser pulse. Typically, single photon events below the 10% level are required. In practice, then, the single photon event is used to START the timer, and a laser pulse under a fixed delay is used to STOP in order to minimize dead time. The delay is adjusted with lengths of cable in order to capture the signal events in an appropriate time window.

66 BS A HP L Cell Ti:Sa

F PD P

PMA TH200 STOP START

Figure 3.8 TCSPC-HRS setup. BS: beam splitter, A: analyzer, HW: Half-wave plate, L: Lens, PD: Avalanche photo diode, F: 10nm narrow band filter centered at 400nm, P: polarizer, PMA: single photon detector, TH200: multi-time-channel chip.

Figure 3.9 shows histograms for the different types of photons detected. Figure 3.9(a) shows the histogram of a typical fluorescent photon, from the fluorescent dye

Rhodamine-6G, which shows a sharp rise and slow decay line. The lifetime of

Rhodamine-6G is measured to be about 6ns. Figure 3.9(b) depicts a histogram of a typical SHG signal, generated by 2-methyl-4-nitroaniline (MNA powder) which shows a sharp peak and no apparent fluorescence. By fitting to a Gaussian function, the SHG histogram pulse is found to have a FWHM of 280ps yielding the time resolution of the system limited mostly by the photomultiplier tube jitter. Figure 3.9(c) and (d) ((d) zooms in (c)) show the histogram of the sample, di-8-ANEPPS, which exhibits both strong SHG and multiphoton induced fluorescence. As shown in the figure 3.9 (c) and (d), when the

400nm narrow band filter is removed, the broadband fluorescence completely buries the

SHG signal and only fluorescence is obvious (squares and circles). The fluorescence peak

(squares) is later than that of the SHG (squares), determining the time gate used in separating the signals.

67

Figure 3.9 TCSPC histograms: (a) fluorescence signal of Rhodamine-6G; (b) SHG peak of MNA power; (c) and (d) histogram of di-8-ANEPPS. Circles denote broadband fluorescence signal obtained by removing the narrowband filter, while the Square data is obtained with the narrowband filter in place and depicts combined SHG and fluorescence signal.

The TCSPC-HRS setup shown in Figure 3.8 is very similar to the standard HRS setup [25, 26] except for the signal detection where TCSPC-HRS measures the arrival time distribution of single photons rather than the intensity. As shown in Fig. 3.9(d), the

SHG intensity is calculated by integrating the SHG peak over the indicated time window

([TL, TR], where TR is set at the position of the fluorescence peak) and subtracting the linear part of the fluorescent count in the window yielding the total counts of pure SHG,

I(2ω). Both the pure SHG signal and the uncorrected signal were quadratic in the incident power verifying that the signal is a composite of SHG and two-photon excited fluorescence.

68 After resolving the SHG from the fluorescence using the histogram, the hyperpolarizability can be measured in the same way as a standard HRS experiment.

Figure 3.10 depicts the concentration dependence of the raw data and the extracted SHG signal. If the SHG is not absorbed, a quadratic dependence gives SHG signals

2 2 as I(2ω) = G(N solvent β solvent + N solute β solute ) .[25, 26] Since the SHG contribution from the solvent is almost constant and the absorption of the SH signal (for relatively weak absorption) can be described as an exponential function, we can further write the

2 expression as: I(2ω) = (A+ BNsolute)exp(−cNsolute) . Here, B ∝ βsolute , A is the solvent contribution, and c is the absorption factor of the solute. Disperse Red 1(DR1) has been used as an external reference. [30] Its hyperpolarizability value at 1360nm was

-30 determined from EFISH measurements to be βEFISH = 125×10 esu. [31] We convert this

-30 to hyperpolarizability due to HRS yielding β HRS =52×10 esu (where, for one-

dimensional molecules like DR1 β HRS = 6 / 35β EFISH ).[32, 33] Since the present experiment was carried out at 800nm, the two-level model [4] was used to extrapolate the

DR1 hyperpolarizability, giving a value of 72×10-30esu at 800 nm. Using this value, the

-30 hyperpolarizability of di-8-ANEPPS is determined to be βHRS=(272±27)×10 esu at

800nm.

69

21 18 ) 4 15 10 ( 12

9 6 Photons 3 0 0.0 0.3 0.6 0.9 16 -3 Number density (10 cm )

Figure 3.10 HRS experiment data: the crosses represent the photon counts of SHG plus fluorescence; the circles represent the photon of pure SHG.

The depolarization ratio is a useful check of our results since it can reveal the dimensionality of the molecule. In HRS experiments the depolarization ratio is defined

ZX as DZZ = I ⊥ / I|| , where I ⊥ and I|| represent the SHG intensity when the incident and outgoing polarization states are either orthogonal or parallel to each other, respectively.

We obtain a value of 0.23 ± 0.07 which agrees with the expected value for a one- dimensional molecule of 0.20. [33]

Based on the planar, nearly rod-like structure of this potential-sensitive dye, it is reasonable to assume that the hyperpolarizability tensor only has two in-plane

components: β ZZZ and a small β ZXX . We can calculate them explicitly by using equation

(3.8) as derived in Appendix:

70 3 − 2ρ D ZX = ZZ 15 +18ρ (3.8) 6 16 β = β + ρ ZZZ 35 105

where ρ = ββZXX/ ZZZ .

XZ For a one-dimensional system, ρ = 0 andDZZ = 0. 2 , which is predicted by Kaatz et

al [34] β = 6 / 35β ZZZ , which is the relationship between the conventional measurement of β for a one-dimensional molecule in EFISH to HRS. [32, 33] For the potential

−30 sensitive dye, βZZZ =±×(643 64) 10 esu , and − 0.20 < β ZXX / β ZZZ < 0.12 , confirming the approximation of a one-dimensional molecule.

3.3.3 Discussion Huang et al. measured a similar dye, di-4-ANEPPS, using a monolayer on water

−30 surface at 532nm to give [5] β ZZZ = 600×10 esu , which is in reasonable agreement with our experimental value considering the different wavelength. Also they estimated

from semiempirical calculations that β ZXX / β ZZZ lies between -0.3 and 0.3. Our di-8-

ANEPPS measurement are consistent with their calculations.

71

(a)

(b)

Figure 3.11 Electron orbitals of (a) HOMO and (b) LUMO. The full alkyl chains found on the dialkylamine donor and on the alkylpyridinium acceptor (left and right end of the figures, respectively) were not included in the calculation as they are not expected to significantly affect the electronic orbitals. The orbitals were generated using ZINDO semi-empirical quantum chemical calculations.

The well-separated charge transfer absorption band of di-8-ANEPPS shown in

Figure 3.2 allows theoretical estimation of the hyperpolarizability tensor using only the first term in the sum-over-states expression (the two-level model). It is well known in the two-level model that the hyperpolarizability is proportional to the change in dipole moment (charge transfer) between the ground and first excited states, as well as to the square of the transition dipole moment between these two states. Both of these quantities have been calculated using the semi-empirical ZINDO[35] program within Gaussian03.

[36] First, geometry optimization was performed using the semi-empirical AM1

Hamiltonian (also in Gaussian03). The di-8-ANEPPS molecule was found to be a planar, pi-conjugated molecule, as expected. This pi-conjugation along the length of the

72 molecule makes for very efficient charge transfer from the dialkylamine donor to the pyridinium acceptor. The results of the ZINDO Configuration Interaction Singles (CIS) calculation resulted in a first excited state that is mainly characterized by the transfer of one electron from the HOMO to the LUMO as seen in Figure 3.11. The transition moment calculated with ZINDO was 11.75 D. Integration of the lowest lying absorption peak in Figure 3.2 yields an experimental value for the transition moment of 9.7 D in

DMSO. The discrepancy is, in part, due to (has the correct sign for) solvent effects not

accounted for in the ZINDO calculation. The change in dipole moment (Δμg1) calculated by ZINDO was -14.6 D. Using the two-level model and the experimental value for the

transition moment gives Δ=−μg1 14.0 D in good agreement with the ZINDO calculation, although at least a solvent correction would be expected here as well. The computed

-30 vector hyperpolarizability from the pure ZINDO calculations is βzzz=876×10 esu, which is an overestimation resulting mainly from the moderate difference between the measured and computed transition moment. The determination of the quantity Δμg1 is related to the electrochromic and solvatochromic effects. This quantity allows estimation of the electrochromic response of two-photon fluorescence and second harmonic generation.

These results will be discussed in chapter 4.

3.4 Conclusion

In a summary, spectroscopic studies of di-8-ANEPPS have been performed. Its absorption spectrum shifts to the blue while the fluorescence emission spectrum shifts to the red by solvents of high polarity. This symmetrical solvatochromism, the linear and

73 nonlinear optical properties of di-8-ANEPPS have been reviewed with two current models for the anomalous solvatochromism.

The hyperpolarizability tensor components βzzz and βzxx of di-8-ANEPPS have been measured at 800 nm using the TCSPC HRS method, which minimizes contributions to the second harmonic signal from two-photon fluorescence. These measurements confirmed the nearly one-dimensional nature of the chromophore and provide a good experimental determination of the difference in dipole moment between the ground and the first excited states, which will be helpful in elucidating the origin of the potential sensitive response in biological applications.

Appendix

In this appendix we calculate the two in-plane Cartesian components β ZZZ and

small β ZXX , based on our previous work [37].

β The component ratio is defined as: ρ = ZXX (A1) β ZZZ

For a one-dimensional system, Kleinman symmetry only leaves two rotational invariants,

β1SS and β3SS so that we can express the relationship between the two different forms as:

2 9 2 β1SS = (1+ 2ρ)β ZZZ 15 3 (A2) 6 β 2 = (1− 2ρ)β 2 3SS 15 7 ZZZ

74 2 XZ I ⊥ β ⊥ The depolarization ratio is defined as: DZZ = ∝ 2 (A3) I|| β||

2 2 The quantities β ⊥ and β|| can further be written as a function of β ZZZ and ρ , as follows:

2 1 2 4 2 1 2 2 β ⊥ = β1SS + β3SS = ( − ρ)β ZZZ 15 3 15 7 35 105 (A4) 2 3 2 2 2 1 6 2 β|| = β1SS + β3SS = ( + ρ)β ZZZ 5 5 7 7 35

2 2 The HRS average hyperpolarizabilty value can be written as: β = β ⊥ + β|| (A5)

From equations (A1) to (A5), we can obtain the following relationship:

3 − 2ρ D ZX = ZZ 15 +18ρ (A6) 6 16 β = β + ρ ZZZ 35 105

75 References

1. Loew, L.M., Potentiometric dyes: Imaging electrical activity of cell membranes. Pure And Applied Chemistry, 1996. 68(7): p. 1405-1409.

2. Loew, L.M., G.W. Bonneville, and J. Surow, Charge shift optical probes of membrane potential. Theory. Biochemistry, 1978. 17: p. 4065.

3. Fluhler, E., V.G. Burnham, and L.M. Loew, Spectra, Membrane-Binding, And Potentiometric Responses Of New Charge Shift Probes. Biochemistry, 1985. 24(21): p. 5749-5755.

4. Oudar, J.L. and D.S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state dipole moment. Journal Of Chemical Physics, 1977. 66(6): p. 2664.

5. Huang, J.Y., A. Lewis, and L. Loew, Nonlinear optical properties of potential sensitive styryl dyes. Biophys J, 1988. 53(5): p. 665-70.

6. Shen, Y.R., Surface-Properties Probed By 2nd-Harmonic And Sum-Frequency Generation. Nature, 1989. 337(6207): p. 519-525.

7. Fromherz, P., Monopole Dipole Model For Symmetrical Solvatochromism Of Hemicyanine Dyes. Journal Of Physical Chemistry, 1995. 99(18): p. 7188-7192.

8. Laage, D., W.H. Thompson, M. Blanchard-Desce, and J.T. Hynes, Charged push- pull polyenes in solution: Anomalous solvatochromism and nonlinear optical properties. Journal Of Physical Chemistry A, 2003. 107(31): p. 6032-6046.

9. Hassner, A., D. Birnbaum, and L.M. Loew, Charge-Shift Probes Of Membrane- Potential - Synthesis. Journal Of Organic Chemistry, 1984. 49(14): p. 2546-2551.

10. Brunschwig, B.S., S. Ehrenson, and N. Sutin, Solvent Reorganization In Optical And Thermal Electron-Transfer Processes - Solvatochromism And Intramolecular Electron-Transfer Barriers In Spheroidal Molecules. Journal Of Physical Chemistry, 1987. 91(18): p. 4714-4723.

11. Siano, D.B. and D.E. Metzler, Band shapes of the electronic spectra of complex molecules. Journal Of Chemical Physics, 1969. 51(5): p. 1856.

12. Hubener, G., A. Lambacher, and P. Fromherz, Anellated hemicyanine dyes with large symmetrical solvatochromism of absorption and fluorescence. Journal Of Physical Chemistry B, 2003. 107(31): p. 7896-7902.

13. Thompson, W.H., M. Blanchard-Desce, V. Alain, J. Muller, A. Fort, M. Barzoukas, and J.T. Hynes, Two valence bond state model for molecular

76 nonlinear optical properties. Comparison with push-pull polyene solution measurements. Journal Of Physical Chemistry A, 1999. 103(19): p. 3766-3771.

14. Thompson, W.H., M. Blanchard-Desce, and J.T. Hynes, Two valence bond state model for molecular nonlinear optical properties. Nonequilibrium solvation formulation. Journal Of Physical Chemistry A, 1998. 102(39): p. 7712-7722.

15. Bedlack, R.S., M.D. Wei, and L.M. Loew, Localized Membrane Depolarizations And Localized Calcium Influx During Electric Field-Guided Neurite Growth. Neuron, 1992. 9(3): p. 393-403.

16. Gross, E., R.S. Bedlack, and L.M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurement Of The Membrane Dipole Potential. Biophysical Journal, 1994. 67(1): p. 208-216.

17. Campagnola, P.J., M.D. Wei, A. Lewis, and L.M. Loew, High-resolution nonlinear optical imaging of live cells by second harmonic generation. Biophysical Journal, 1999. 77(6): p. 3341-3349.

18. Ries, R.S., H. Choi, R. Blunck, F. Bezanilla, and J.R. Heath, Black lipid membranes: visualizing the structure, dynamics, and substrate dependence of membranes. Journal Of Physical Chemistry B, 2004. 108(41): p. 16040-16049.

19. Bouevitch, O., A. Lewis, I. Pinevsky, J.P. Wuskell, and L.M. Loew, Probing Membrane-Potential With Nonlinear Optics. Biophysical Journal, 1993. 65(2): p. 672-679.

20. Loew, L.M., L.B. Cohen, B.M. Salzberg, A.L. Obaid, and F. Bezanilla, Charge- Shift Probes Of Membrane-Potential - Characterization Of Aminostyrylpyridinium Dyes On The Squid Giant-Axon. Biophysical Journal, 1985. 47(1): p. 71-77.

21. Loew, L.M. and L.L. Simpson, Charge-Shift Probes Of Membrane-Potential - A Probable Electrochromic Mechanism For Para-Aminostyrylpyridinium Probes On A Hemispherical Lipid Bilayer. Biophysical Journal, 1981. 34(3): p. 353-365.

22. Moreaux, L., T. Pons, V. Dambrin, M. Blanchard-Desce, and J. Mertz, Electro- optic response of second-harmonic generation membrane potential sensors. OPTICS LETTERS, 2003. 28(8): p. 625-627.

23. Pons, T., L. Moreaux, O. Mongin, M. Blanchard-Desce, and J. Mertz, Mechanisms of membrane potential sensing with second-harmonic generation microscopy. JOURNAL OF BIOMEDICAL OPTICS, 2003. 8(3): p. 428-431.

24. Moreaux, L., O. Sandre, and J. Mertz, Membrane imaging by second-harmonic generation microscopy. JOURNAL OF THE OPTICAL SOCIETY OF AMERICA B-OPTICAL PHYSICS, 2000. 17(10): p. 1685-1694.

77 25. Clays, K. and A. Persoons, Hyper-Rayleigh Scattering In Solution. Review Of Scientific Instruments, 1992. 63(6): p. 3285-3289.

26. Clays, K. and A. Persoons, Hyper-Rayleigh Scattering In Solution. Physical Review Letters, 1991. 66(23): p. 2980-2983.

27. Olbrechts, G., K. Wostyn, K. Clays, and A. Persoons, High-frequency demodulation of multiphoton fluorescence in long-wavelength hyper-Rayleigh scattering. Optics Letters, 1999. 24(6): p. 403-405.

28. Noordman, O.F.J. and N.F. vanHulst, Time-resolved hyper-Rayleigh scattering: Measuring first hyperpolarizabilities beta of fluorescent molecules. Chemical Physics Letters, 1996. 253(1-2): p. 145-150.

29. Spence, D.E., P.N. Kean, and W. Sibbett, 60-Fsec Pulse Generation From A Self- Mode-Locked Ti-Sapphire Laser. Optics Letters, 1991. 16(1): p. 42-44.

30. Hubbard, S.F., R.G. Petschek, and K.D. Singer, Spectral content and dispersion of hyper-Rayleigh scattering. Optics Letters, 1996. 21(21): p. 1774-1776.

31. Kuzyk, M.G. and C.W. Dirk, Characterization techniques and tabulations for organic nonlinear optical materials. 1998: Marcel Dekker Inc. 489.

32. Wang, C.H., J.N. Woodford, C. Zhang, and L.R. Dalton, Resonant and nonresonant hyper-Rayleigh scattering of charge-transfer chromophores. Journal Of Applied Physics, 2001. 89(8): p. 4209-4217.

33. Bersohn, R., Y.-H. Pao, and H.L. Frisch, Double-Quantum Light Scattering by Molecules. Journal Of Chemical Physics, 1966. 45: p. 3184-3198.

34. Kaatz, P. and D.P. Shelton, Polarized hyper-Rayleigh light scattering measurements of nonlinear optical chromophores. Journal Of Chemical Physics, 1996. 105(10): p. 3918-3929.

35. Ridley, J. and M. Zerner, An Intermediate Neglect of Differential Overlap Technique for Spectroscopy: Pyrrole and the Azines. Theoretica Chemica Acta (Berl.), 1973. 32: p. 111.

36. Frisch, M.J., G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, and e. al., Gaussian 03 Revision C.02. 2004, Gaussian Inc: Wallingford CT.

37. Ostroverkhov, V., R.G. Petschek, K.D. Singer, L. Sukhomlinova, R.J. Twieg, S.X. Wang, and L.C. Chien, Measurements of the hyperpolarizability tensor by means of hyper-Rayleigh scattering. Journal Of The Optical Society Of America B- Optical Physics, 2000. 17(9): p. 1531-1542.

78 Chapter 4: Solvent effects on the non- linear optical response of a potentiometric biological imaging dye

In order to understand the electronic mechanisms for the electric-field sensitive second harmonic response of di-8-ANEPPS in biological membranes, we have measured its nonlinear optical hyperpolarizability using a combination of Hyper-Rayleigh scattering (HRS) and linear optical properties in solvent environments of various polarity.

Changes in the first hyperpolarizabilites are discussed by modeling experimental inputs from the solvatochromic effect in linear absorption and appropriate dispersion relations near the two-photon resonance. We find that both frequency dependent and frequency independent effects make significant contributions to the response, and the strongest response is on the low energy side of the two-photon resonance, while the weakest response is on the high energy side of the two-photon resonance.

79 4.1 Introduction

First developed by Loew and colleagues,[1, 2] the fluorescent dye di-8-ANEPPS

(4-(2-[6-(dioctylamino)-2-naphthalenyl]ethenyl)-1-(3-sulfopropyl) pyridinium inner salt) as shown in Figure (4.1), consists of a quaternized pyridinium electron acceptor moiety linked by a π-bond to an amino substituted naphthalene electron donor. [3] It has been extensively used as a potential sensitive dye to probe, via linear and nonlinear optics, the electrical environment in biological membranes due to changes in its optical properties as a function of the local electric field. Considerable work has been done in order to determine the mechanism of the nonlinear optical sensitivity and the mode of optimum use of ANEPPS and other probe dyes. [4-11] Studies in Loew’s laboratory have shown that electrochromism (Stark shift) is an important component of the dye response to potential sensitivity. [9-11] The chromophore’s electron density is concentrated in the amino-naphthalene ring in the ground state and shifts to the pyridinium ring in the excited state; this charge-shift couples with an electric field within a cell membrane, resulting in electrochromism.

- + O3S(H2C)3 N C8H17 N C8H17

Figure 4.1 Structure of di-8-ANEPPS

It has been found that probing membrane potential using second harmonic generation is superior to fluorescence modalities because of the requirement for noncentrosymmetric media leading to high membrane specificity and contrast, low

80 background, and high spatial resolution in deep scattering specimens. [12-15] Recently, the molecular orientation of di-8-ANEPPS and similar dyes in black lipid membranes

(BLM)[4] and giant unilamellar vesicles (GUV)[5, 6] and its effects on potential sensitivity have also been discussed using second harmonic generation. The potential response may arise purely from electrically induced changes in the molecular electronic hyperpolarizability β, or by a combination of this and molecular reorientation by the external electric field. Several papers have theoretically and experimentally discussed the two-photon fluorescence or second harmonic sensitivity of ANEPPS and similar potential dyes.[16-20] In particular, Pons and Mertz[16] recently examined theoretically the Stark shift mechanism near the two-photon resonance.

In this chapter, we seek to examine the purely electronic mechanism for electric- field dependent second harmonic generation using solvent dependent optical properties.

To do so, we have carried out linear and nonlinear optical studies of the di-8-ANEPPS dye in liquid solution. It is well-known that solvent polarity is closely related to the local electric field surrounding a dissolved molecule [21]. We use polarity-dependent solvent studies of linear absorption and hyper-Rayleigh scattering in solution to probe the spectral dependence of the sensitivity to the local environment. We also model this sensitivity using a two-level model by including the spectral shift, changes in transition moment, excited and ground state dipole moment difference, and spectral width as the local electric field (solvent polarity) is varied. We model the relationship between first hyperpolarizability, β, and linear absorption spectrum as a function of solvent polarity and, by inference, local electric field. We find good agreement between the model and

81 the measurements. Our work suggests a cuvette-based linear optical technique for estimating the sensitivity response of potentiometric dye candidates.

We note that while di-8-ANNEPS is not the only second-harmonic potentiometric dye currently in use, we have chosen to study it here because of the large body of literature on it, and because of its good solubility in a variety of organic solvents. We believe that the results and conclusions found here are more generally applicable and plan further studies on other dyes to verify this.

4.2 Methods and Theory

Hyper-Rayleigh scattering (HRS) has been established as a powerful tool for measuring the first-order electronic hyperpolarizability β.[22, 23] It involves illumination of a sample in isotropic solution at frequency ω and detection of the scattered incoherent second-harmonic photons at 2ω. Since HRS does not involve the use of an external static electric field, it can measure both polar and non-polar as well as charged and uncharged materials,[24] making it particularly appropriate for measurement of di-8-ANEPPS.

In order to obtain the hyperpolarizability β for di-8-ANEPPS, an external referencing method is used. A well-studied molecule, para-nitroaniline (pNA) is used as the reference compound. The HRS intensity is measured as a function of concentration for the di-8-ANEPPS and pNA under identical experimental conditions. Due to the incoherent nature of the HRS process, for a two component solvent-solute system, the

HRS intensity can be written as a sum of the intensities from the solvent and the dye

222−CNsolute molecules [25]: IGN(2ωβ )=<>+<> (solvent solvent N solute β solute ) Ie0 where G is a

82 geometrical, local field and instrumental factor. The quantities Nsolvent and Nsolute are the

2 2 number densities of the solvent and solute, respectively, and < βsolvent > and <>βsolute are the molecular hyperpolarizabilities squared and averaged over orientation fluctuations.

−CNsolute The quantity I0 is the intensity of the fundamental laser, and e is the Beer–Lambert

Law to account for the absorption of the HRS if the scattered light wavelength lies in the absorption band of the dye solution. If we use A and B to substitute the variables in the

2222 above equation we obtain: AGIN= 0 solvent<>β solvent , BGI= 0 <>βsolute , Thus:

−CNsolute IABNe2ω =+()solute . By varying the concentration of solute, in the limit of low

concentration, the number density of the solvent Nsolvent does not vary appreciably so that we can obtain the concentration dependence by a fit of the experimental data with intercept A, linear factor B, and extinction factor C. By comparing the data for pNA and

2 BpNA<>β pNA the dye, we obtain: = 2 . With the knowledge of βpNA, the βdye can be easily Bdye<>β dye calculated.

Since the di-8-ANEPPS chromophore is an asymmetric conjugated π-electron system possessing a large hyperpolarizability in the charge-transfer direction, it might be described using a one-dimensional, two-level model (TLM), originally developed by

Oudar and Chemla: [26]

2 3Δ⋅μμ 4 ge ω0 β = 22 2 2 2 2 (4.1) ω00()(4)ωωωω−− 0

83 Here is Planck’s constant, ω0 the resonance frequency, Δμ = μeg− μ the

difference in the dipole moment between the excited (e) and ground (g) states, and μge the transition dipole moment. As we show later, depolarization measurements do indeed suggest that the chromophore is quasi-one-dimensional and, therefore, makes the two- level model reasonably descriptive. Since we are carrying out measurements near the two-photon resonance, an appropriate model for damping must be included to properly describe the dispersion. In general, this implies that

β =Δβωωμμγ(2;0 , ,ge ,) (4.2) where the functional form is that which appropriately describes the broadening mechanism depending on the broadening factor, γ.

The second harmonic sensitivity to local field is actually described by the EFISH, or electric-field-induced second harmonic coefficient, a third order nonlinear optical response function. However, direct EFISH measurements in solution or even in membranes will yield both electronic (the quantity wanted here) and orientational responses. In solution, the EFISH orientational response is much larger.

The electrochromic and solvatochromic response, that is the change in intensity due to the change in the local field, for optical second harmonic generation can be

2 dI d ()2 2 estimated from = ()χ E , where Eloc is the local electric field that is, in the dEloc dE loc case considered here, related to solvent polarity The nonlinear optical susceptibility

χ (2) is proportional to the hyperpolarizability β so we can calculate the relative changes of the amplitude of β as a function of local electric field, or the sensitivity as:

84 2 12ddβ β 2 = . Eloc can be expressed using F, a polarity indictor in solvents, β dElocβ dE loc which is a function of solvent dielectric constant and refractive index: [27]

1 ε −−n2 εε ii and ε = 2, an empirical parameter, is taken as the Fn(,ε )=− (2 ) i εiiεεε++22n i intermediate dielectric constant.

The sensitivity, or relative change in the amplitude of β as a function of polarity

(local electric field) can be written as:

11d ββ∂Δμ ∂⎤ 1∂μge ∂ β⎤ =+⎥ ⎥ ββμdF∂∂Δ F βμ ∂∂ F ⎦μωγ,, ge ⎦⎥Δμω,, γ ge 0 0 (4.3)

11∂ω0 ∂⎤ββ∂γ ∂⎤ ++⎥⎥ βω∂∂FF βγ ∂∂ 0 ⎦⎦Δμμ,, γ ge Δμμ,,ge ω0 where barred denote mean values used to numerically evaluate the partial derivatives of the analytical expression for β . Equation (4.3) provides a model for understanding the mechanism for the electric field dependent second harmonic response (electronic contribution to the electric field induced second harmonic generation, EFISH). We will measure the left-hand side of Equation (4.3) directly using hyper-Rayleigh scattering in solutions of various F. We will compare these HRS measurements to the right-hand side of the equation using solvatochromic measurements of linear absorption to determine the

∂μge ∂ω0 ∂γ ∂Δμ F dependence of μge, ω0, and γ, i.e. ,,. The F dependence of Δμ, , will ∂FFF∂∂ ∂F be determined from hyper-Rayleigh measurements and an appropriate model for β. The

85 1 ∂⎤β factors on the right of the form ⎥ require an analytical model for dispersion of β ∂u ⎦vwx,,

β near the two-photon resonance as well as average values for parameters held fixed.

Models for the dispersion of β near the two-photon resonance have been widely discussed in the literature. This discussion basically centers on the appropriate description of broadening near the resonance. The original model is based on a simple Classical

Damped Harmonic Oscillator (CDHO) introduced as ω00=−ωγi , where β can be written as[28]:

2 Δ⋅μμge 1 β = { 2 (2)()ωγωωγω++ii ++ 00 (4.4) 11 ++} (ω00−−iiγωωγωωγωωγω 2 )( −− ) ( 00 ++ ii )( −− )

Ref. [28] has shown that Equation (4.4) can be written analogously to the Oudar-Chemla

TLM:

2 3Δ⋅μμ 2 ge ()ωγ0 − i β = 22222 (4.5) [(ω00−−iiγωωγ ) ][( −− ) 4 ω ]

This model is generally taken to describe homogeneously broadened media, which is clearly not descriptive of a complex molecule in liquid solution. Nonetheless, we will compare this with other models below to show that under certain circumstances, it yields a reasonable approximation.

Wang [29] has pointed out the inadequacy of Equation (4.5) in describing the dispersion of β near the two-photon resonance. The vibronic structure as the broadening mechanism was introduced to better describe the dispersion. In order to describe the dispersion in this way, the vibrational modes must be known in detail by, for instance,

86 using computational modeling of the experimental absorption spectrum and resonance

Raman excitation profiles. [30-32] In order to avoid the resonance Raman experiments and computational modeling, Kelly [33] has developed a semi-empirical approach using a direct mathematical transformation of the linear absorption band shape to obtain the hyper-Rayleigh dispersion, essentially a Kramers-Kronig transform relating the absorption coefficient to the index of refraction. In this case, Kelly finds:

2 Δμ μωωge(2 e + ) β (2−=ωωω , , )2 × [ + χ (2)] ω (4.6) 2(πωωπωωωωcc eee−++ )2()(2)

Here Kelly has used ωe as the average electronic transition frequency (averaged over the vibronic states, which we take as identical to ω0). The term χ ()2ω is the linear susceptibility near the two-photon resonance. It is determined empirically as the imaginary part is obtained from the linear absorption spectra, and the real part from the

Kramers-Kronig transform of the imaginary part. It should be noted that terms in the energy denominator containing positive frequencies have been omitted since they do not resonate at the frequencies in this experiment.

These two dispersion approaches will be used in the analysis of our data. Kelly’s model will allow us to obtain accurate values for Δμ. The CDHO model will provide a reasonably accurate analytical description for evaluation of Equation (4.3) since Kelly’s model cannot be used to evaluate Equation (4.3) as it is semi-empirical, and, therefore, not analytical.

87 4.3 Experimental

The samples of di-8-ANEPPS used in our study were both purchased from Invitrogen, and synthesized by a procedure described by Loew and coworkers. The synthesis started with the Bucherer reaction of 6-bromo-2-naphthol to give 2-amino-6-bromonaphthalene.

This aminonaphthalene was next reacted with 1-bromooctane to afford the dialkylated 2-

(N,N-dioctyl)-6-bromonaphthalene which was coupled to 4-vinylpyridine using the palladium mediated Heck reaction. The resulting pyridine functionalized product was alkylated on nitrogen with 1,3-propane sultone to give the di-8-ANEPPS.

Figure 4.2 Experimental setup for HRS measurements

Figure (4.2) depicts the HRS experiment setup. A Q-switched Nd:YAG laser

(Surelite II, Continuum) with a repetition rate of 10Hz and 7-ns pulse was used. The

88 1064nm laser output was followed by a frequency doubler and tripler to produce 355nm light, which was used to pump an OPO to obtain a variable wavelength output. The laser light passes through an analyzer and a Berek compensator set to λ/2 retardation for the fundamental wavelength. The compensator is mounted on a computer-controlled rotation stage in order to adjust the polarization and intensity of the fundamental light. A small fraction of the incident fundamental light is focused onto a quartz plate for monitoring the fluctuations of the fundamental laser light so as to eliminate the shot-to-shot noise from the signal. (Reference channel) The fundamental light is focused into the sample solution in a square quartz cuvette (10x10 mm). The focal length is about 5cm, with the HRS signal collected at 90 degrees. In order to efficiently collect the second-order scattered light, a large aperture aspherical condenser is used (focal length is 2.5cm). The collimated light passes through a polarizer followed by a plano convex lens to focus the light into a photomultiplier tube. (Signal channel) Depending on which second harmonic wavelength is measured, a suitable color filter and an interference notch filter (10nm bandwidth) is used to block the fundamental light. The polarizer in front of the detector is added to measure the depolarization ratio of the Hyper-Rayleigh scattering in order to check the dimensionality of the molecule. The depolarization ratios we measured are 0.28± 0.5 for pNA and 0.24± 0.03 for di-8-ANEPPS which agree with the expected value of a one- dimensional molecule[34, 35]. Hyper-Rayleigh scattering measurements were carried out at five wavelengths: 800nm, 900nm, 976nm, 1020nm, and 1064nm. It was difficult to use wavelengths beyond this range because of the competing signal from background fluorescence. Appropriate wavelengths for measurement were determined from the collected emission spectra that showed little competing fluorescent emission.

89 In order to determine the hyperpolarizability β as a function of solvent polarity, two solvents of different polarity, dimethyl sulfoxide (DMSO) and chloroform, were mixed in various ratios. We found the two solvents to be completely miscible. The di-8-

ANEPPS was first dissolved in a stock solution of DMSO at a concentration of 0.1mg/ml and then diluted to 0.02mg/ml into five solutions: A: pure DMSO solvent; B: 80%

DMSO and 20% Chloroform; C: 60% DMSO and 40% Chloroform; D: 40% DMSO and 60% Chloroform; E: 20% DMSO and 80% Chloroform. Based on the refractive index and dielectric constant data of DMSO and chloroform, these solutions form a series of solutions of controlled polarity, as shown in Table (3.1).

The emission spectra from the hyper-Rayleigh scattering experiments were initially collected using a spectrometer outfitted with an intensified CCD camera detector to verify in all cases (wavelengths and solvents) that the second harmonic signal was well-separated from any fluorescence emission. Following this, hyper-Rayleigh scattering was measured at the appropriate polarization using photomultiplier tubes.

Para-nitroaniline (pNA) in acetone has been used as an external reference. Its hyperpolarizability value at 1064nm is 24.5× 10−30 esu [36] and the Oudar-Chemla TLM was used to extrapolate the β values at other wavelengths. This extrapolation is reasonable because all of the wavelengths were far from the two-photon resonance where dispersion must be more carefully considered.

4.4 Results and discussion

Figure (4.3) shows the linear absorption spectra of the Di-8-ANEPPS in DMSO and mixtures of DMSO and chloroform solvents. We can see the absorption spectrum

90 shifts towards the blue at higher polarity (higher DMSO content), which agrees with the spectroscopic observation of similar dyes by Fromherz [37].

1.4

1.2

1.0

0.8

0.6

0.4 Optical density (OD)

0.2

0.0 15000 20000 25000 30000

Wave number (cm-1)

Figure 4.3 Linear absorption spectra of di-8-ANEPPS in solutions A, B, C, D and E. Absorption spectrum shifts toward the red (lower wavenumber) as the polarity decreases from A (rightmost curve) to E (leftmost curve).

From the linear absorption spectra, ω0 , μge , and γ can be calculated for each solvent and are plotted as a function of solvent polarity F in Figure (4.4). The electronic

transition frequencyω0 can be obtained from absorption peak. The transition moment μge can be calculated from the integrated molar absorptivity as [38]:

91 2 ε[]ϖ μ =×9.185 10−3 ndϖ (4.7) ge ∫ ϖ

(a) (b)

11.2 20500 11.0 ) 20000 10.8 -1

(D) cm (

eg 10.6 e 19500 μ ω 10.4

0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 F F

(c) (d)

2250 4000 2200 3000

2150 )

) esu

-1 2100

2000 −30 cm 2050 (

10 ( γ β 2000 1000 1950 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220

F F

Figure 4.4 Solvent polarity (F) dependent measured quantities. a) Absorption frequency, b) transition moment, c) spectral half-width and d) first hyperpolarizability. (Solid square 800nm, open square 900nm, triangle 976nm, solid circle 1020nm, and open circle 1064nm in d.)

92 where n is the index of refraction, ε is the molar absorptivity in liters per mole per centimeter, ϖ is the wavenumber in inverse centimeters and the transition moment is in

Debye. The γ is taken as the half-width at half maximum of the spectral transition, which can be obtained from fitting of the absorption spectra, as explained later in this chapter.

dω dμ dγ d β A linear fit is used to obtain 0 , ge and . We also plot from dF dF dF dF hyper-Rayleigh scattering data for all wavelengths. Figure (4.4) shows the four plots and

dω dμ dγ linear fits. The fits yield 0 =±36500 7700, eg =−16.9 ± 2.1 and =±5850 810 . dF dF dF

These results and a model for the dispersion of β allow determination of the last three terms on the right-hand side of Equation (4.3). We also determine the relative

1 d β sensitivity values for the wavelengths 800nm, 900nm, 976nm, 1020nm, and β dF

1064nm (respectively: −±5.9 3.1 , −4.5± 4.0 , −9.0± 3.8 , −13.8± 4.3 and −±16.5 5.0 ).

These values will provide the data for the left-hand side of Equation (4.3).

In order to determine the first term on the right-hand side of Equation (4.3) involving Δμ, we need to determine Δμ from the hyper-Rayleigh scattering data. To do this, we plot the values of β at various wavelengths as shown in Figure (4.5). The quantity Δμ is determined by fitting the data in Figure (4.5) to an appropriate model for the dispersion of β. We will fit to β as obtained by calculating the complex magnitude of both Equations (4.5) and (4.6) using Δμ as an adjustable parameter. The other factors in Equations (4.5) and (4.6) are fixed as determined from the linear absorption data.

93

3.00E-027

2.50E-027

2.00E-027

1.50E-027

(esu)

β 1.00E-027

5.00E-028

0.00E+000 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Wave number (cm-1)

Figure 4.5 Hyperpolarizability β of di-8-ANEPPS in DMSO solution (Dots), F=0.220 , and fit to Equations (4.5) Solid line and (4.6) Dashed line.

We show the fit to the absorption spectrum in pure DMSO as an example in

Figure (4.6) to both a standard Lorentzian shape and also to a log-normal function[39].

We see that the log-normal function appears to be a better fit. However, we have determined Δμ from both models. In order to obtain values of ω0, γ and μge for calculating Δμ from Equation (4.5), we need to fit the absorption ε(ω) to a standard

Lorentzian shape, [40]

2γ εω()= A 22 (4.8) ()ω −ωγ0 +

94 where γ is the HWHM of the spectral transition and ω0 is the transition frequency. The

transition moment μge is calculated from Equation (4.7). Based on this Lorentzian fit to

-1 the linear absorption spectrum, we use mean values of ω0, μge and γ as 20048cm , 10.7D,

2114cm-1 as the fixed parameters of the fit to β in Figure (4.5), using the CDHO model of

Equation (4.5) , we obtain Δμ of 20.2 D.

In order to calculate Δμ from Equation (4.6), we employ the fit of the absorption data to the log-normal function [39],

ε b 1 ω − a εω( )=−−0 exp(c22 )exp{ [ln( )] } for ω > a (4.9) ω − acb2 2 and

ε()ω = 0 for ω ≤ a (4.10) where the parameters a , b, c are related to resonant frequency ω0 and HWHM of the spectral transition γ by

2 ω0 = ab+−exp( c ) (4.11) and

{exp[2c (2ln 2)1/2 ]− 1} γ =−bcexp(2 ) (4.12) 2exp[c (2ln 2)1/2 ]

We calculate the imaginary part of the linear susceptibility χ ()2ω from

4πω σ ()ωω= Im[()]x . The real part of χ (2ω ) needed to evaluate equation (4.6) can be a 3cn

2'Im[(')]∞ ωχω obtained using the Kramers-Kronig relationship:[41] Re[χ (ωω )]= Pd '. ∫ 22 πωω0 ' −

95

1.4

1.2

1.0

0.8

0.6

0.4 Optical density (OD) density Optical 0.2

0.0 15000 20000 25000 30000

Wave number (cm-1)

Figure 4.6 Linear absorption spectra of di-8-ANEPPS in DMSO solution (Solution A, solid line). Dashed line is the Log-normal fit for the first electronic absorption band. Dot- Dash is the CDHO lineshape fit.

Using all of this in Equation (4.6), we find Δμ of 13.0D. We believe that this value for Δμ is more accurate than that obtained from the CDHO model both heuristically, and because of its agreement with previous results,[11] which listed an estimated Δμ of

14D. Note that it is likely that Δμ and β this dye are negative which cannot be determined by hyper-Rayleigh scattering. However, Equation (4.3) is insensitive to whether or not

Δμ is negative. We carried out this procedure for all of the values of F from various solvent mixtures, with the results plotted in Figure 4.7. We note that we found similar slopes in both models, but the more accurate value of Δμ from Equation (4.6).

96

28 27 26 25 24 ) 23 D

22

Δμ ( 21 20 19 18 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.220 F

Figure 4.7 Plot of Δμ versus F and the corresponding linear fit

Having all of the factors in the right-hand side of Equation (4.3) of the form

∂ of , we now must use an analytical expression for β to obtain the other factors. ∂F

Kelly’s semiempirical Equation (4.6) is not appropriate since there is no convenient and accurate analytical expression for the real part of χ (2ω) . Therefore, we must use the

CDHO model. We note that the fits of Figure (4.5) do not clearly differentiate between the shapes of the model (although the Δμ for the best fit is better for Kelly’s model). We will apply the CDHO model, since it is analytic, and find that it is reasonable to do so.

97

0 -2 -4 -6 -8 -10

-12 -14 -16 -18 Sensitivity (a.u.) -20 -22 0 2000 4000 6000 8000 1000012000140001600018000 ω (cm-1)

Figure 4.8 Relative second harmonic sensitivity as a function of frequency. Solid line is the model of Equation (4.13) (right-hand side), and data is the left-hand side of Equation (4.13).

Rewriting Equation (4.3) for the CDHO model of Equation (4.5), we obtain,

112d β ∂Δμ ∂μ ∂ω ∂γ =++ge AB0 + (4.13) ωγ00,, ωγ βμμdFΔ∂ Fge ∂ F ∂ F ∂ F where A and B are given in the Appendix. The bars denote that the average values of the quantities (averaged over the various values corresponding to different F) are used in the calculation. In Figure (4.8) the data points indicate the sensitivity of the second harmonic intensity to a field as given by the left-hand side of Equation (4.13) and the results depicted in Figure (4.4d). The curve in Figure 4.8 is obtained by evaluating the right-

98 hand side of Equation (4.13) and the results depicted in Figures (4.4a-c) and (4.7). We see that the data and model agree without adjustable parameters. We conclude that

CDHO model given in Equation (4.13) is a reasonable description of the sensitivity of the second harmonic signal to local field. The imperfect fit is likely due to inadequacies of

CDHO description and the use of a two-level model, in general.

It is interesting to ask how each term in Equation (4.13) contributes to the overall sensitivity. Figure (4.9) depicts the contribution of each term to the sensitivity. First, it is clear that the contribution of the broadening γ is small, except perhaps near the resonance. This might explain why the CDHO model yields reasonable results despite its obvious shortcomings. Pons and Mertz [16] concluded that the highest sensitivity is obtained at the low energy side of resonance by only considering the Stark shift (third term on the right-hand side of Equation (4.13). We concur with this conclusion, but find that the other contributions make this all the more dramatic. When considering the other contributions that are constant in frequency, the sensitivity on the high energy side is minimal. It is clear that changes in the dipole moments with local field make significant contributions to the sensitivity as also seen in Figure (4.9). We also note that our analysis agrees well with trends of the sensitivities of similar dyes in biological imaging experiments [17, 19], given the difference in the absorption spectra between our solutions and the biological environments. From the curve trend, at the fundamental wavenumber

~9000 cm-1 (or at wavelength 1.1μm), the relative sensitivity is maximal. We note that to the extent that the two-level model of the imaginary part of the second hyperpolarizability describes the sensitivity of the two-photon excited fluorescence

(TPEF) response, [16] our results and conclusions can be applied to that imaging

99

8 6 4 ω0 2 γ 0 -2 Δμ

-4 μeg -6 -8 Sensitivity (a.u.) -10 -12 0 2000 4000 6000 8000 1000012000140001600018000 ω (cm-1)

Figure 4.9 Contributions of different terms of Equation (4.3) as a function of frequency. Labels indicate the variable of the partial derivative in that equation.

modality. We note, however, that, in the mode of two-color fluorescence ratio potentiometric imaging, [2] our model will not necessarily apply, as this modality implies that the frequency shift will dominate and frequency independent factors will not contribute.

Finally, we can make some remarks concerning solvatochromism. In the introduction, we indicated that the local field is closely related to the local electric field as the relationship between solvatochromism and electrochromism. This relationship is model dependent, and will involve the Onsager reaction field, for example and phenomenological cavity radius parameter. For this reason, we indicate only relative units in Figure (4.8).

100 Our method suggests an assay for estimating the electronic sensitivity using only linear spectra. In this work, we utilized hyper-Rayleigh scattering to determine Δμ.

However, Δμ can be estimated using solvent shifts for both linear absorption and emission where model dependent parameters, such as the cavity radius, are made to cancel. The excited state dipole moment of a quasi-one-dimensional dye can be determined through, [21]

ΔEemission μμeg= (4.14) ΔEabsorption where the ground state dipole moment must be independently known. With these measurements, the right hand side of Equation (4.3) can be determined using linear optical techniques.

4.5 Conclusions

We have studied the electric field dependence for second harmonic generation in a potentiometric dye used in biological imaging using solvatochromic studies of hyper-

Rayleigh scattering and linear absorption. Through a simple model relating the sensitivity of second harmonic generation to solvatochromic data of linear absorption, we can understand the contributions to the second harmonic sensitivity. We find that, in addition to the Stark shifts, that changes in dipole moments with local field, make significant contributions. We find that the dispersion of the sensitivity indicates that the most sensitive wavelength for measurements is one half width lower in energy than the absorption peak, and the least sensitive is one half width higher in energy than the

101 absorption peak. Our model and the use of solvatochromic linear absorption studies provides a reasonable assay for assessing the potential sensitivity of new chromophores with the proviso that additional measurements are needed to estimate the magnitude of the Δμ term, which is important. We believe that our study has more general applicability, but further work is necessary to establish this.

Appendix

Here we calculate the partial derivatives of β with respect to ω0 and γ. From

2 3Δ⋅μμ 2 ge ()ωγ0 − i Equation (4.5) β = 22222, we obtain: [(ω00−−iiγωωγ ) ][( −− ) 4 ω ]

2 3()Δ⋅μμ γ22 + ω β =×ge 0 2

422426422462 ((γ +++++− 5γω 4 ω ) 2(2 γ 5 γω γω 20 ω ) ω0

4224422680.5− +−(6γγωωωγωωω 10 + 33 )000 + 2(2 − 5 ) + )

1 ∂ β 1 ∂ β We also calculate A = and B = explicitly: β ∂ω0 β ∂γ

862442686422462 A =−[2ω0 (γγωγωγωωγγωγωωω − 5 − 32 − 60 − 16 + (4 − 15 + 32 + 20 ) 0

42242 268 2242242 +−3(2γγωωγωωωγωγγωω 5 )0000 +− (4 5 ) + )] /[( + )(( + 5 + 4 )

64224624 2244 2 268 ++2(2γ 5γω +− γω 20 ω ) ω0000 +− (6 γ 10 γω + 33 ω ) ω + 2(2 γ − 5 ω ) ω + ω )]

862268 6 422462 B =−[2γ (γγωγωωγγωγωωω + 5 − 20 − 16 + (4 + 15 + 32 + 60 ) 0

4224422682242242 ++(6γγωωωγωωωγωγγωω 15 − 32 )0000 + (4 + 5 ) + )]/[( + )(( + 5 + 4 )

64224624 2244 2 268 ++2(2γ 5γω +− γω 20 ω ) ω0000 +− (6 γ 10 γω + 33 ω ) ω + 2(2 γ − 5 ω ) ω + ω )]

102 The A , B values from Equation (4.13) are obtained by substituting the average ω0 ,γ ω0 ,γ values of ω0, γ for the different polarity solvent mixtures.

103 References

1. Bedlack, R.S., M.D. Wei, and L.M. Loew, Localized Membrane Depolarizations And Localized Calcium Influx During Electric Field-Guided Neurite Growth. Neuron, 1992. 9(3): p. 393-403.

2. Loew, L.M., Potentiometric dyes: Imaging electrical activity of cell membranes. Pure And Applied Chemistry, 1996. 68(7): p. 1405-1409.

3. Fluhler, E., V.G. Burnham, and L.M. Loew, Spectra, Membrane-Binding, And Potentiometric Responses Of New Charge Shift Probes. Biochemistry, 1985. 24(21): p. 5749-5755.

4. Ries, R.S., H. Choi, R. Blunck, F. Bezanilla, and J.R. Heath, Black lipid membranes: visualizing the structure, dynamics, and substrate dependence of membranes. Journal Of Physical Chemistry B, 2004. 108(41): p. 16040-16049.

5. Pons, T., L. Moreaux, O. Mongin, M. Blanchard-Desce, and J. Mertz, Mechanisms of membrane potential sensing with second-harmonic generation microscopy. JOURNAL OF BIOMEDICAL OPTICS, 2003. 8(3): p. 428-431.

6. Moreaux, L., T. Pons, V. Dambrin, M. Blanchard-Desce, and J. Mertz, Electro- optic response of second-harmonic generation membrane potential sensors. OPTICS LETTERS, 2003. 28(8): p. 625-627.

7. Moreaux, L., O. Sandre, and J. Mertz, Membrane imaging by second-harmonic generation microscopy. JOURNAL OF THE OPTICAL SOCIETY OF AMERICA B-OPTICAL PHYSICS, 2000. 17(10): p. 1685-1694.

8. Bouevitch, O., A. Lewis, I. Pinevsky, J.P. Wuskell, and L.M. Loew, Probing Membrane-Potential With Nonlinear Optics. Biophysical Journal, 1993. 65(2): p. 672-679.

9. Loew, L.M., L.B. Cohen, B.M. Salzberg, A.L. Obaid, and F. Bezanilla, Charge- Shift Probes Of Membrane-Potential - Characterization Of Aminostyrylpyridinium Dyes On The Squid Giant-Axon. Biophysical Journal, 1985. 47(1): p. 71-77.

10. Loew, L.M. and L.L. Simpson, Charge-Shift Probes Of Membrane-Potential - A Probable Electrochromic Mechanism For Para-Aminostyrylpyridinium Probes On A Hemispherical Lipid Bilayer. Biophysical Journal, 1981. 34(3): p. 353-365.

11. Loew, L.M., G.W. Bonneville, and J. Surow, Charge shift optical probes of membrane potential. Theory. Biochemistry, 1978. 17: p. 4065.

12. Sacconi, L., D.A. Dombeck, and W.W. Webb, Overcoming photodamage in second-harmonic generation microscopy: Real-time optical recording of neuronal

104 action potentials. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2006. 103(9): p. 3124-3129.

13. Nuriya, M., J. Jiang, B. Nemet, K.B. Eisenthal, and R. Yuste, Imaging membrane potential in dendritic spines. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2006. 103(3): p. 786-790.

14. Dombeck, D.A., L. Sacconi, M. Blanchard-Desce, and W.W. Webb, Optical recording of fast neuronal membrane potential transients in acute mammalian brain slices by second-harmonic generation microscopy. Journal Of Neurophysiology, 2005. 94(5): p. 3628-3636.

15. Campagnola, P.J., M.D. Wei, A. Lewis, and L.M. Loew, High-resolution nonlinear optical imaging of live cells by second harmonic generation. Biophysical Journal, 1999. 77(6): p. 3341-3349.

16. Pons, T. and J. Mertz, Membrane potential detection with second-harmonic generation and two-photon excited fluorescence: A theoretical comparison. Optics Communications, 2006. 258(2): p. 203-209.

17. Millard, A.C., L. Jin, J.P. Wuskell, D.M. Boudreau, A. Lewis, and L.M. Loew, Wavelength- and time-dependence of potentiometric non-linear optical signals from styryl dyes. Journal Of Membrane Biology, 2005. 208(2): p. 103-111.

18. Fisher, J.A.N., B.M. Salzberg, and A.G. Yodh, Near infrared two-photon excitation cross-sections of voltage-sensitive dyes. Journal Of Neuroscience Methods, 2005. 148(1): p. 94-102.

19. Millard, A.C., L. Jin, M.D. Wei, J.P. Wuskell, A. Lewis, and L.M. Loew, Sensitivity of second harmonic generation from styryl dyes to transmembrane potential. Biophysical Journal, 2004. 86(2): p. 1169-1176.

20. Kuhn, B., P. Fromherz, and W. Denk, High sensitivity of stark-shift voltage- sensing dyes by one- or two-photon excitation near the red spectral edge. Biophysical Journal, 2004. 87(1): p. 631-639.

21. Suppan, P. and N. Ghoneim, Solvatochromism. 1997: Cambridge, Royal Society of Chemistry.

22. CLAYS, K. and A. PERSOONS, HYPER-RAYLEIGH SCATTERING IN SOLUTION. PHYSICAL REVIEW LETTERS, 1991. 66(23): p. 2980-2983.

23. Clays, K. and A. Persoons, Hyper-Rayleigh Scattering In Solution. Review Of Scientific Instruments, 1992. 63(6): p. 3285-3289.

24. Zyss, J., T.C. Van, C. Dhenaut, and I. Ledoux, Harmonic Rayleigh-Scattering From Nonlinear Octupolar Molecular Media - The Case Of Crystal Violet. Chemical Physics, 1993. 177(1): p. 281-296.

105 25. Chui, T.W. and K.Y. Wong, Study of hyper-Rayleigh scattering and two-photon absorption induced fluorescence from crystal violet. Journal Of Chemical Physics, 1998. 109(4): p. 1391-1396.

26. Oudar, J.L. and D.S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state dipole moment. Journal Of Chemical Physics, 1977. 66(6): p. 2664.

27. Brunschwig, B.S., S. Ehrenson, and N. Sutin, Solvent Reorganization In Optical And Thermal Electron-Transfer Processes - Solvatochromism And Intramolecular Electron-Transfer Barriers In Spheroidal Molecules. Journal Of Physical Chemistry, 1987. 91(18): p. 4714-4723.

28. Berkovic, G., G. Meshulam, and Z. Kotler, Measurement and analysis of molecular hyperpolarizability in the two-photon resonance regime. Journal Of Chemical Physics, 2000. 112(9): p. 3997-4003.

29. Wang, C.H., Effects of dephasing and vibronic structure on the first hyperpolarizability of strongly charge-transfer molecules. Journal Of Chemical Physics, 2000. 112(4): p. 1917-1924.

30. Moran, A.M., D.S. Egolf, M. Blanchard-Desce, and A.M. Kelley, Vibronic effects on solvent dependent linear and nonlinear optical properties of push-pull chromophores: Julolidinemalononitrile. Journal Of Chemical Physics, 2002. 116(6): p. 2542-2555.

31. Moran, A.M., C. Delbecque, and A.M. Kelley, Solvent effects on ground and excited electronic state structures of the push-pull chromophore julolidinyl-n-N,N '-diethylthiobarbituric acid. Journal Of Physical Chemistry A, 2001. 105(45): p. 10208-10219.

32. Moran, A.M. and A.M. Kelley, Solvent effects on ground and excited electronic state structures of p-nitroaniline. Journal Of Chemical Physics, 2001. 115(2): p. 912-924.

33. Kelley, A.M., Frequency-dependent first hyperpolarizabilities from linear absorption spectra. Journal Of The Optical Society Of America B-Optical Physics, 2002. 19(8): p. 1890-1900.

34. Heesink, G.J.T., A.G.T. Ruiter, N.F. Vanhulst, and B. Bolger, Determination Of Hyperpolarizability Tensor Components By Depolarized Hyper Rayleigh- Scattering. Physical Review Letters, 1993. 71(7): p. 999-1002.

35. Bredas, J.L., F. Meyers, B.M. Pierce, and J. Zyss, On The 2nd-Order Polarizability Of Conjugated Pi-Electron Molecules With Octupolar Symmetry - The Case Of Triaminotrinitrobenzene. Journal Of The American Chemical Society, 1992. 114(12): p. 4928-4929.

106 36. Kaatz, P. and D.P. Shelton, Polarized hyper-Rayleigh light scattering measurements of nonlinear optical chromophores. Journal Of Chemical Physics, 1996. 105(10): p. 3918-3929.

37. Fromherz, P., Monopole Dipole Model For Symmetrical Solvatochromism Of Hemicyanine Dyes. Journal Of Physical Chemistry, 1995. 99(18): p. 7188-7192.

38. Klessinger, M. and J. Michl, Excited States and Photochemistry of Organic Molecules. 1995.

39. Siano, D.B. and D.E. Metzler, Band shapes of the electronic spectra of complex molecules. Journal Of Chemical Physics, 1969. 51(5): p. 1856.

40. Milonni, P.W. and J.H. Eberly, Lasers. 1998: J. Wiley and Sons.

41. Yariv, A., Quantum Electronics, 2nd Edition. 1975.

107 Chapter 5: Applications of potential- sensitive dyes

5.1 Introduction

Di-8-ANEPPS (Pyridinium, 4-[2-[6-(dioctylamino)-2-naphthalenyl]ethenyl]-1-(3- sulfopropyl)-, inner salt) and other similar styryl dyes have been used extensively in cell membrane potential imaging due to changes in their optical properties as a function of the local electric field. [1] Ratiometric fluorescence measurements have been used to measure the local intramembrane electric field. [2-8] Nonlinear optical techniques including two-photon excited fluorescence and second harmonic generation have also become increasing popular tools in cell imaging. [9-17] A computer search from ISI web of science for di-4-ANEPPS and di-8-ANEPPS returns close to 200 results. Considering the limited nature of the database and that some authors didn’t include the names of the dyes used in their abstracts, this is easily a 50% underestimate of the actual number of papers published during 1980 and 2006. If we consider other similar potentiometric spectroscopic dyes, the numbers for papers published will be even much higher: For example, a similar search for “potentiometric AND membrane” returns more than 2200 results.

108 A key factor controlling the SHG efficiency of these styryl dyes in a biological membrane is their orientation and organization in the complex system. [18, 19]

Phospholipid monolayers and bilayers formed using Langmuir-Blodgett technique have been used as models for biological membrane. [20-22] It is advantageous to use this technique to study the molecular organization and orientation as it reduces the complexity of real biological membranes. Di-8-ANEPPS has a pair of hydrocarbon chains acting as membrane anchors and a hydrophilic group which aligns it perpendicular to the membrane/aqueous interface. [1] As a result, di-8-ANEPPS is an amphiphilic membrane staining dye which is suitable for fabricating Langmuir-Blodgett films. However, using di-8-ANEPPS and some phospholipids such as DPPC (L-α- dipalmitoylphosphatidylcholine) to build bilayers or multilayers proves to be problematic.

In this chapter, the difficulties of the sample fabrication will be reported, and properties of di-8-ANEPPS monolayers, DPPC monolayer, and mixture of di-8-ANEPPS and DPPC monolayer will be investigated and their significance will be discussed.

Since the plasma membrane regulates many essential cellular processes, it is important to determine its structure and function. The “fluid mosaic” model developed by

Singer and Nicolson [23] suggests that the plasma membrane can be viewed as a two dimensional fluid with proteins and lipids free to diffuse around the surface of the cell.

Although this model forms the basis of the modern concept of the plasma membrane, decades of research have revealed significant deviation from its predictions. Now the emerging picture of the plasma membrane is a considerably inhomogeneous landscape of barriers, patches, and domains. Near-field scanning optical microscopy (NSOM), whose resolution exceeds the optical diffraction limit, has great potential for plasma membrane

109 imaging. It has the potential to address questions regarding scales of tens of nanometers in the plasma membrane. Actually, NSOM has already been carried out with model lipid bilayers as well as lipid microstructures isolated from cells [24, 25]. In this chapter, the principles of near-field scanning optical microscopy will be reviewed and some preliminary studies of near-field fluorescence imaging on dried cells will also be reported.

5.2 Langmuir-Blodgett films

The Langmuir-Blodgett technique, first introduced by Irving Langmuir and applied extensively by Katharine Blodgett, involves the transfer of organic monolayers of amphiphilic molecules from the air-water interface to a substrate. [26-28] The surface pressure and temperature of the monolayer on the water surface can be controlled so that the organic film is in a condensed and stable state. Phospholipid bilayers which are supported on flat solid substrates have become popular model membrane systems to study the physical properties and the physiology of biological membranes. One of the advantages of these supported planar bilayers over many other model membrane systems is that they are unilamellar and geometrically well defined. Here the phospholipid bilayer and multilayers using di-8-ANEPPS and DPPC have been attempted and the films have also been characterized.

5.2.1 Film Deposition

DPPC in chloroform (20mg/ml) was purchased from Avanti polar Lipids and used without purification. Di-8-ANEPPS was purchased from Invitrogen and was dissolved in

110 Chloroform (0.5mg/ml). The mixture of DPPC and di-8-ANEPPS was prepared with a

1:1 molar ratio. Figure 5.1 shows the structures of DPPC and di-8-ANEPPS.

(a)

(b)

Figure 5.1 Molecular structure of (a) di-8-ANEPPS (b) DPPC

The preparation of the Langmuir-Blodgett films was performed using a KSV

5000 Langmuir-trough. The subphase was Milli-Q water at pH = 6.8. The interfacial surface pressure was monitored with a Wilhelmy balance using a platinum plate. The amphiphilic solutions were spread from chloroform solution onto the pure water subphase. After a relaxation time of about 30 minutes for solvent evaporation, the monolayers were compressed continuously with a compression barrier speed of

5mN/m/min until it reached to the desired surface pressure. The monolayers were left at this pressure for one hour for equilibration. A previously immersed plain glass slide

(hydrophilic treated) or mica was withdrawn from the subphase at a speed of 1~3 mm/min, depending on the displayed transfer ratio. The transfer ratio is defined as the ratio of the area of monolayer removed from the water surface to the area of substrate coated by the monolayer. For di-8-ANEPPS and DPPC, the transfer ratios for substrate withdrawn are very close to 1, which indicates a good deposition. However, after about

111 ten minute’s exposure in the air, when the substrate is immersed in the subphase again, the transfer ratio is close to -1. This means the monolayer is peeling off the slide and respreading on the water surface.

5.2.2 Film characterization

There are a number of methods to investigate monolayer Langmuir-Blodgett films.

Here we only list methods which have been applied to our samples. Those are: π-A isotherm, UV-VIS absorption spectra and surface second harmonic generation.

5.2.2.1 π – A Isotherm

One of the most important indicators of the monolayer properties of a material is given by the π(surface pressure)-A(molecular area) isotherm: It is a plot of surface pressure as a function of the area of water surface available to each molecule. This is carried out at constant temperature and is known as surface pressure/area isotherm. π - A isotherms for three different monolayers at the air-water interface are shown in Figure

5.2(a)-(d). They are di-8-ANEPPS, di-8-ANEPPS/DPPC 1:1 ratio mixture, and DPPC monolayers. Figure 5.2 (a) depicts the isotherm of DPPC, which shows several phase transitions that agree with the literature very well. [22] However, for mixtures of di-8-

ANEPPS/DPPC Figure 5.2(b) and di-8-ANEPPS monolayers Figure 5.2(c), no phase transition can be found, and the monolayers were not very stable. This is in contradiction with the statement from reference [29]. In Figure 5.2 (d), hysteresis appears for twelve

112

(a) (b)

(c) (d)

Figure 5.2 Isotherms of (a)DPPC, (b) 1:1 molar ratio of di-8-ANEPPS/DPPC; (c) di- 8-ANEPS; (d) Twelve continuous di-8-ANEPPS isotherms, all on a pure water surface at room temperature.

continuous isotherms of di-8-ANEPPS, with the molecular area getting smaller and smaller. Basically, when the surface is pressed to a certain surface pressure and then very slowly relaxed to zero pressure, the isotherm is not reversible. It appears that di-8-

113 ANEPPS partially dissolves in the water at a high pressure. After equilibrating with the solution, it is possible to make a di-8-ANEPPS monolayer.

5.2.2.2 UV-VIS Absorption measurements

UV-VIS absorption spectra were obtained by using a Cary 5000 UV-Vis-NIR spectrophotometer both for monolayers and for solutions. A clean glass slide or a cuvette with chloroform inside has been used as a reference for monolayers and solutions, respectively. Figure 5.3 indicates that di-8-ANEPPS deposited as a LB film has a blue shift of the absorption band compared with that of the corresponding chloroform solutions. Due to solvatochromic effects, it is highly possible that variations in the monolayer environment of the dye can significantly modify the absorption spectra, and more generally its linear and nonlinear optical properties. [30]. The DPPC monolayer and the mixture of DPPC/di-8-ANEPPS have negligible absorption at wavelength longer than

300nm, and their absorption spectra are not included in the plot.

114

0.05

0.04 0.2

0.03

0.02 0.1

0.01 Absorption(O.D)

0.00 0.0 300 400 500 600 700 800 Wavelength(nm)

Figure 5.3 UV-VIS absorption of di-8-ANEPPS in Chloroform solution 0.01M (Right Y-axis scale, dashed line); di-8-ANEPPS monolayer (Left Y-axis scale, solid line)

5.2.2.3 Second harmonic generation (SHG)

Surface second harmonic generation has been a powerful tool to characterize [31,

32] the organization, orientation and second-order susceptibility of chromophores at the surface/interface. The nonlinear optical properties of the monolayer films were

2ω investigated by SHG. In this experiment, the transmitted SHG Iij are measured, where i and j specify the polarization of the incident beam and transmitted beam, respectively. A light beam polarized in the plane of incidence is referred to P polarization while light beam polarized perpendicular to the plane of incidence is referred to S polarization. The

SHG measurements were made using a Nd:YAG laser at 1064nm. The laser has an output of 100μJ reaching the sample and a pulse duration of about 8 ns. The incident laser beam was focused onto the monolayer surface (Spot size is about 1mm in diameter).

115 The transmitted SHG signal was separated from the fundamental laser using a BG18 color filter and a 532nm interference notch filter. A gated boxcar was used to collect the transmitted SHG signal as the substrate was rotated from -70 degrees to 70 degrees relative to normal incidence.

The transmitted SHG intensity for a di-8-ANEPPS monolayer is shown in Figure

5.4. The interference fringe of the SHG intensity arises from the interaction of the second harmonic waves from the dye monolayers on both sides of the glass substrate. The complete destructive interference is evident: That the minimum goes to zero indicates that the thin films on both sides of the substrate are essentially identical. They possess the same molecule orientation or polar angle, number density, chemical structure and

22ωω thickness. From Figure 5.4, IIPP/5.4 SP = so that dd33/3.4 31 = . Based on the analysis given in Chapter 2, Assuming the distribution angle α is sharply peaked, then

α sin−10 (A )≈ 37 Using quartz as a reference, we also calculated the effective

−8 susceptibility to be desueff =×1.6 10 . This value is smaller than expected, since di-8-

−30 ANEPPS have a very large β (From Chapter 3, β zzz ∼ 650× 10 esu ) and the proximity of the laser harmonic wavelength (532nm) to its absorption band (~500nm).

The DPPC monolayer does not produce any measurable SHG signal; however, the

1:1 mixture of di-8-ANEPPS and DPPC does produce some weak SHG. (about one order smaller than the pure di-8-ANEPPS monolayer) Weak interference fringes also arise, but the minima don’t totally destructively go to zero. This indicates inhomogeneity in the film structure.

116 2.5

2.0

1.5

1.0

0.5 SHG intensity(a.u.)

0.0 -60 -40 -20 0 20 40 60 Degree

Figure 5.4 SHG measurements for Di-8-ANEPPS monolayer deposited onto plain glass substrate (thickness 1mm) P-P polarization configuration (Line ) S-P polarization (Cross)

5.2.3 Conclusion

Di-8-ANEPPS monolayers were fabricated and characterized along with DPPC and di-8-ANEPPS/DPPC mixtures. We have found it difficult to reproduce good quality monolayers and multilayers. We tried several methods in order to improve the films, such as changing the pH values of the subphase, calcium ions, and mixtures with DPPA(1,2-

Dipalmitoyl-sn-Glycero-3-Phosphate), Arachidic acid (C20H40O2). However, none of them produced films that could yield reproducible second harmonic generation. Actually several authors [29, 33, 34] have reported the similar difficulties. While many of our attempts were unsuccessful, this initial work will provide insight into further studies that may prove more fruitful.

117 5.3 Near-field Scanning optical microscopy

In order to investigate materials on the nanoscale, a microscope with resolution below the optical diffraction limit is necessary. There are several high resolution imaging techniques available, such as the electron microscopes and scanning probe microscopes.

However, optical imaging can provide information unobtainable by any other means. Of course, until recently, one major draw back of the conventional optical microscopy was the inability to obtain resolution on length scale smaller than 250nm. Fortunately, with progress made in the last decade, several techniques such as 4π microscopy [35, 36], I5M microscopy [37], STED (Stimulated emission depletion) [38, 39], and NSOM (Near-field scanning optical microscopy) [40, 41] have all extended the optical resolution beyond the diffraction limit. Here in this chapter, after a brief introduction to NSOM, we report the results of NSOM imaging for biological cells.

5.3.1 History of near-field scanning optical microcopy

The resolution in classical optical microscopy is limited by the wavelength of the excitation light. The diffraction limit is defined as 0.61λ /(NA . .) , [42] where λ is the wavelength of the light and NA..= n sinθ , the numerical aperture of the lens system.

Here n is the index of refraction of the surrounding medium and θ is the half-angle of the maximum cone of light that can enter or exit the lens. Since the numerical aperture of any lens system can only be extended slightly beyond unity, the maximum resolution can be simplified to λ /2.

In 1928, Synge proposed that by using a metal plate with a tiny hole much smaller than the wavelength of light, optical imaging resolution beyond the diffraction limit

118 might be achieved. [43] He suggested illuminating this tiny hole and scanning the plate in close proximity across a sample. If the plate-sample distance is much smaller than the diameter of the hole, then the resolution is limited by the diameter of the light source (the hole) instead of by the wavelength of the light. Figure 5.5 shows the configuration.

Figure 5.5 Schematic diagram illustrating Synge’s proposal for achieving subdiffraction limit resolution.

It was not until 1972 that Ash and Nicolls were able to verify the theory with electromagnetic waves in the microwave range. [44] With their setup, they achieved a resolution of 0.5mm at a wavelength of 3cm, yielding an impressive image resolution of nearly λ/60. In 1984, Pohl and his co-workers [45] were able to demonstrate, for the first time, resolution below the diffraction limit in the optical wavelength range. For their experiments they used etched and coated quartz tips. A breakthrough toward routine application of the technique was carried out at Bell Labs by Betzig et al. [40, 41] , who

119 used pulled and metal coated optical single mode fibers to demonstrate optical resolution approaching 50nm or approximately λ/10 in the visible spectrum.

5.3.2 AlphaSNOM from WiTec [46]

Although many good results have been obtained with fiber tips, they have notable disadvantages, including but not limited to: time-consuming tip production, low light transmission in the tips and the maintenance of a critical tip-sample distance. The near- field scanning optical microscope used in our lab (AlphaSNOM from WiTec, Germany) goes a step further and uses micro-structured cantilever sensors and beam deflection feedback for distance control. [46] These cantilever sensors have the size and shape of standard AFM cantilevers and are fabricated in a batch process. The cantilever consists of silicon with a hollow silicon dioxide pyramid at the end. This pyramid has a tiny hole of below 100nm at its tip and the whole system is aluminum coated. The NSOM excitation laser is focused from the backside into the pyramid. A small fraction of the light can pass through the aperture to excite the sample in the near-field. A second feedback laser is focused onto the cantilever and the reflected light is used for standard beam-deflection feedback, as is done in most AFMs. Figure 5.6 shows the configuration and the cantilever.

In order to take advantage of the high resolution obtained by near-field scanning optical microscopy, it is crucial to maintain a short distance between the tip and the sample. The near-field is the region where the electric field is confined to the dimensions of the aperture and its optical resolution is determined only by the size of this aperture.

As the distance between aperture and sample increases, the intensity of the evanescent field drops off exponentially while at the same time, diffraction effects becomes stronger,

120 causing a transition to that of a diffraction limited conventional optical microscope.

Figure 5.5 clearly describes the short range of the near-field region.

Figure 5.6 Left plot is the distance feedback using cantilever-NSOM sensors. The excitation as well as the feedback laser are focused onto the cantilever through the same lens system. The cantilever is loaded into the optical axis of the microscope with a high precision XY-stage for maximum throughout. Right side is cantilever with a hollow SiO2 tip. [Obtained from WiTec Manual]

The AlphaSNOM consists of a combination of scanning near-field optical microscopy, confocal microscopy, and atomic force microscopy. The NSOM cantilever sensors are integrated into a NSOM objective. By simply turning the turret of the microscope, it can be easily switched from confocal to near-field microscopy.

Several imaging methods can be applied for near-field scanning. It can be used in near-field transmission mode, or reflection mode. By using additional optical filters, it can be used in near-field fluorescence transmission mode or reflection mode. In the fluorescence imaging mode, after the laser excites the sample, the microscope objective captures the excitation and fluorescence light at the same time. Optical filters are used to block the fundamental wavelength while a photomultiplier detects the fluorescence. For

121 our experiments, NSOM, fluorescence NSOM and fluorescence confocal microscopy are used, all in a transmission setting. A 532nm laser Coherent Verdi-V8 is used to excite the dye and the fluorescence is collected using a 610nm long pass filter. A microscope objective of 20X focuses the laser beam onto the sample through the cantilever while a

60X (N.A.= 0.8) objective is used to collect the transmitted light.

5.3.3 Samples

There are three samples available for the NSOM experiments. The first one is a standard test sample: Latex projection pattern, purchased from Kentax UHV equipment in Germany. It consists of a glass substrate with small aluminum islands. The sample is prepared as follows: In the first step, a monolayer of latex spheres with 450nm diameter is deposited on a glass substrate. A very thin aluminum layer about 10nm thickness is then evaporated onto this structure. After this, the latex spheres are removed and the aluminum forms a hexagonal structure of small islands. This sample is an ideal test structure to determine the near-field optical resolution.

The second sample is a fixed T lymphocyte cell which was stained with a primary antibody and then labeled with a secondary antibody stained with Alexa_fluor 568. It sits on a cover slip and is dehydrated before the imaging. Alexa_fluor 568 is purchased from

Molecular Probes, Oregon, which has an absorption spectrum peaked at 568nm, and emission spectrum peaked at ~603nm.

The third sample is also a fixed T lymphocyte cell, this time it is stained with di-

8-ANEPPS. When di-8-ANEPPS stained with lipids, it has an absorption peak around

470nm and an emission peak around 630nm. Because of the nature of the wide

122 absorption band, it is possible to use the same set up as the Alexa_fluor 568, although the excitation efficiency will be much lower.

5.3.4 Latex projection pattern imaging

It is useful to first use the latex projection pattern to test the WiTec system. As the dimensions of the pattern is known, one would be able to check the lateral size from the atomic force microscopy image, and also be able to check the obtained optical resolution from the simultaneously acquired NSOM image. A transmission NSOM setup is used in this experiment. The laser intensity emanating from the fiber is close to 0.50mW. Only

0.1 of this intensity is required for the imaging experiment. The AFM set point is equal to

0.4 and the scan time per line is 2 seconds. The proportional gain and the intensity gain are set to their default values, 4.0 and 2.0, respectively. Figure 5.7 (a)-(d) show the acquired images. Figure 5.7(a) is the AFM image and (b) is the corresponding NSOM image. The hexagonal structure can be easily identified from Figure 5.7 (a). In order to check the lateral distance between two aluminum islands, (or to check the diameter of the latex sphere), a cross section of the line drawn from (a) is presented. By checking the distance between two peaks, dx = 0.477μm is obtained. Considering the latex sphere is

0.45μm in diameter, the uncertainty of the lateral distance measurement is less than 10% .

Similar cross sections can be obtained for Figure 5.7 (b), which as shown in (d). Between the peak and the valley, we we find that dx = 0.116μm. Taking our 532nm laser wavelength into consideration, we calculate ~λ/5 resolution; certainly better than the λ/2 diffraction limit. Actually, the resolution of the NSOM imaging, in this case is determined by the feature size, instead of the imaging optics. If we use an extremely small sphere latex projection pattern, an even higher resolution might be observed.

123

(a) (b)

(c)

(d)

Figure 5.7 (a) Atomic force microscopy and (b) near-field scanning optical microscopy transmission image of the Latex projection pattern. (c) is the cross section of line on figure 5.7(a) . (d) is the cross section in the similar area of (figure 5.7(b))

124 5.3.5 T-cell imaging

5.3.5.1 Comparison between confocal microsocopy and NSOM

In these series of experiments, fluorescence NSOM experiments are performed.

The fluorescence light is collected instead of 532nm excitation light. First the confocal fluorescent image was obtained from a commercial Zeiss LSM510 inverted confocal microscope. The excitation wavelength is chosen to be 534nm, and the emission is collected using a 650nm bandpass (20nm width) filter. This image can be found in Figure

5.8 (a). Figure 5.8 (b) is the fluorescence T-cell imaging stained with di-8-ANEPPS using a Nikon fluorescence microscope with 530nm excitation and 630nm long pass emission detection. (Artificial color is not used in this image) The WiTec instrument is also capable of collecting fluorescence confocal imaging by itself, and by rotating the turret, it can switch between the confocal and near-field modes. Figure 5.8 (c) is the fluorescence confocal imaging obtained by WiTec using 532nm as the excitation wavelength and fluorescence is collected at 610nm. A 20X objective lens focuses 2 mW of 532nm onto the sample, while a 60X (N.A. 0.8) lens collects the fluorescence to the detector. The scan area is about 75μm by 75μm. Figure 5.8 (d) is the NSOM imaging for T-cell stained with Alexa_fluor 568. With the same configuration, the AFM access point is 0.5 while the proportional gain and intensity gain are still at their default: 4.0 and 2.0 respectively.

The NSOM fluorescent imaging by WiTec is as good as or superior than the commercial confocal fluorescent imaging by Zeiss LSM510. Furthermore, NSOM imaging has the ability to focus on a tiny cell and scan it over a very small scale, which cannot be done by the confocal microscope.

125 (a) (b)

(c) (d)

Figure 5.8 Fluorescence images of (a) T-cell stained with Alexa_fluor 568 from Zeiss LSM510 inverted confocal microscope; (b) T-cell stained with di-8-ANEPPS from Nikon microscope; (c) T-cell stained with Alexa_fluor 568 from WiTec confocal microscope; (d) T-cell with Alexa_fluor 568 from WiTec Fluorescence NSOM microscope.

126 5.3.5.2 Single T-cell NSOM imaging

Under the same imaging condition as above, a single cell is scanned and the near- field fluorescent imaging together with the topography of the cell is shown in Figure 5.9.

The scan area is 10μm by 10μm.Complex structures and variable intensities in the cell are evident. Based on the intensity cross section (Figure 5.9 (c)), the optical resolution for this image is close to 200nm. However, a careful comparison of the optical image with its simultaneously obtained topographical image needs to be done to identify artifacts. It is worth to note that the optical aperture would be slightly shifted from the point of contact

(50~100nm) as the mechanical contact is somewhere at the rim of the optical aperture.

NSOM fluorescence imaging has also been performed for T-cells stained with di-

8-ANEPPS. (Results are not shown here) However, as the excitation laser wavelength

532nm is not as close to the absorption peak of di-8-ANEPPS (470nm), therefore, the overall contrast and quality is not as good as those stained with Alexa_fluor 568.

5.3.6 Conclusion

In summary, we have successfully performed NSOM imaging with dried and fixed T-cells. The image resolution is beyond the conventional diffraction limit. Near- field scanning optical microscopy creates new opportunities for imaging biological plasma membranes.

127 (a) (b)

(c)

Figure 5.9 Single T-Cell imaging (a) NSOM fluorescence image (b) AFM image (c) shows the cross section of the short black line in (b)

128 5.4 Future studies of potential-sensitive dye applications

Currently our laboratory has two projects related to applications of the chromophore di-8-ANEPPS. One is potential mapping in a simulated biological membrane; the other is live cell imaging using our current high-resolution NSOM facility.

For potential mapping, films comprised of n layers, where n will vary from 2 to 6, will be fabricated so that one of the layers will consist of a mixed dye/amphiphilic layer.

Several samples for each n will be fabricated locating the dye at various depths within the multilayer assembly. This film will be immersed in an electrolyte solution and a potential applied. When the electric double-layer is fully formed and the current stops flowing, it can be assumed that the potential has dropped completely across the multilayer film. In this case, second harmonic generation will be measured in these films as a function of applied potential and depth location of the dye layer, thus mapping the electric field as a function of depth. The organization and orientation of the di-8-ANEPPS in the embedded monolayer as a function of the applied electric field could also be determined. Since Langmuir-Blodgett multilayer techniques have proven to be problematic in our studies, some alternative methods such as Black lipid membrane [10] and giant unilamellar vesicles [12] might have to be used as a model membrane.

The other project in our laboratory involves high-resolution imaging for biological samples, especially live cell imaging. A near-field scanning optical microscope will be utilized in this research.

Since we have been able to image dried and fixed T-cells stained with di-8-ANEPPS, naturally the next step would be obtaining high resolution live cell imaging. NSOM imaging of living cells under in vitro conditions has been problematic. The main issue

129 involves the softness of living cells resulting in unacceptably high deformation due to tip- sample interactions while imaging. In order to successfully image live cells, the feedback mechanism must be compatible with immersion, and the mechanical forces between the imaging tip and the cell shouldn’t cause deformation of the cell. And in order to image live biological cells, the rate of the raster scan must be significantly higher than the typical rate of changes of cell thickness and motility.

Recently several papers have been published showing the possibilities of imaging live cells using NSOM [47-49] and one of them [49] involves the use of a low spring constant AFM tip with a milled hole on it. This suggests our WiTec system, which is similar with their setup, is capable of live cell imaging.

130 References

1. Loew, L.M., Potentiometric dyes: Imaging electrical activity of cell membranes. Pure And Applied Chemistry, 1996. 68(7): p. 1405-1409.

2. Xu, C., B.M. Slepchenko, and L.M. Loew, Surface potential measurements on biological membranes with voltage-sensitive dyes. Biophysical Journal, 2002. 82(1): p. 546A-546A.

3. Zhang, J., R.M. Davidson, M.D. Wei, and L.M. Loew, Membrane electric properties by combined patch clamp and fluorescence ratio imaging in single neurons. Biophysical Journal, 1998. 74(1): p. 48-53.

4. Zhang, J., M.D. Wei, and L.M. Loew, Change of surface potential induced by intracellular digestion of membrane proteins. Biophysical Journal, 1998. 74(2): p. A402-A402.

5. Gross, E., R.S. Bedlack, and L.M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurement Of The Membrane Dipole Potential. Biophysical Journal, 1994. 67(1): p. 208-216.

6. Bedlack, R.S., M.D. Wei, S.H. Fox, E. Gross, and L.M. Loew, Distinct Electric Potentials In Soma And Neurite Membranes. Neuron, 1994. 13(5): p. 1187-1193.

7. Bedlack, R.S., M.D. Wei, and L.M. Loew, Localized Membrane Depolarizations And Localized Calcium Influx During Electric Field-Guided Neurite Growth. Neuron, 1992. 9(3): p. 393-403.

8. Montana, V., D.L. Farkas, and L.M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurements Of Membrane-Potential. Biochemistry, 1989. 28(11): p. 4536-4539.

9. Millard, A.C., L. Jin, M.D. Wei, J.P. Wuskell, A. Lewis, and L.M. Loew, Sensitivity of second harmonic generation from styryl dyes to transmembrane potential. Biophysical Journal, 2004. 86(2): p. 1169-1176.

10. Ries, R.S., H. Choi, R. Blunck, F. Bezanilla, and J.R. Heath, Black lipid membranes: visualizing the structure, dynamics, and substrate dependence of membranes. Journal Of Physical Chemistry B, 2004. 108(41): p. 16040-16049.

11. Campagnola, P.J., H.A. Clark, W.A. Mohler, A. Lewis, and L.M. Loew, Second- harmonic imaging microscopy of living cells. Journal Of Biomedical Optics, 2001. 6(3): p. 277-286.

131 12. Moreaux, L., O. Sandre, S. Charpak, M. Blanchard-Desce, and J. Mertz, Coherent scattering in multi-harmonic light microscopy. Biophysical Journal, 2001. 80(3): p. 1568-1574.

13. Moreaux, L., O. Sandre, M. Blanchard-Desce, and J. Mertz, Membrane imaging by simultaneous second-harmonic generation and two-photon microscopy. OPTICS LETTERS, 2000. 25(5): p. 320-322.

14. Moreaux, L., O. Sandre, and J. Mertz, Membrane imaging by second-harmonic generation microscopy. Journal Of The Optical Society Of America B-Optical Physics, 2000. 17(10): p. 1685-1694.

15. Campagnola, P.J., M.D. Wei, A. Lewis, and L.M. Loew, High-resolution nonlinear optical imaging of live cells by second harmonic generation. Biophysical Journal, 1999. 77(6): p. 3341-3349.

16. BenOren, I., G. Peleg, A. Lewis, B. Minke, and L. Loew, Infrared nonlinear optical measurements of membrane potential in photoreceptor cells. Biophysical Journal, 1996. 71(3): p. 1616-1620.

17. Bouevitch, O., A. Lewis, I. Pinevsky, J.P. Wuskell, and L.M. Loew, Probing Membrane-Potential With Nonlinear Optics. Biophysical Journal, 1993. 65(2): p. 672-679.

18. Pons, T., L. Moreaux, O. Mongin, M. Blanchard-Desce, and J. Mertz, Mechanisms of membrane potential sensing with second-harmonic generation microscopy. JOURNAL OF BIOMEDICAL OPTICS, 2003. 8(3): p. 428-431.

19. Moreaux, L., T. Pons, V. Dambrin, M. Blanchard-Desce, and J. Mertz, Electro- optic response of second-harmonic generation membrane potential sensors. OPTICS LETTERS, 2003. 28(8): p. 625-627.

20. Sackmann, E., Supported membranes: Scientific and practical applications. Science, 1996. 271(5245): p. 43-48.

21. Kalb, E., S. Frey, and L.K. Tamm, Formation Of Supported Planar Bilayers By Fusion Of Vesicles To Supported Phospholipid Monolayers. Biochimica Et Biophysica Acta, 1992. 1103(2): p. 307-316.

22. Tamm, L.K. and H.M. McConnell, Supported Phospholipid-Bilayers. Biophysical Journal, 1985. 47(1): p. 105-113.

23. Singer, S.J. and G.L. Nicolson, The Fluid Mosaic Model of the Structure of Cell Membranes. Science, 1972. 175(18): p. 720-731.

24. Tokumasu, F., J. Hwang, and J.A. Dvorak, Heterogeneous molecular distribution in supported multicomponent lipid bilayers. Langmuir, 2004. 20(3): p. 614-618.

132 25. Hwang, J., L.A. Gheber, L. Margolis, and M. Edidin, Domains in cell plasma membranes investigated by near-field scanning optical microscopy. Biophysical Journal, 1998. 74(5): p. 2184-2190.

26. Blodgett, K.B., Films built by depositing successive monomolecular layers on a solid surface. Journal Of the American Chemcial Society, 1935. 57: p. 1007.

27. Blodgett, K.B., Monomolecular films of fatty acids on glass. Journal Of the American Chemcial Society, 1934. 56: p. 495.

28. Langmuir, I., The mechanism of the surface phenomenon of flotation. Trans. Faraday Soc., 1920. 15(62-74).

29. Huesmann, H., G. Gabrielli, and G. Caminati, Monolayers and Langmuir- Blodgett films of the electrochromic dye Di-8-ANEPPS. Thin Solid Films, 1998. 329: p. 804-807.

30. Laage, D., W.H. Thompson, M. Blanchard-Desce, and J.T. Hynes, Charged push- pull polyenes in solution: Anomalous solvatochromism and nonlinear optical properties. Journal Of Physical Chemistry A, 2003. 107(31): p. 6032-6046.

31. Shen, Y.R., Surface-Properties Probed By 2nd-Harmonic And Sum-Frequency Generation. Nature, 1989. 337(6207): p. 519-525.

32. Heinz, T.F., H.W.K. Tom, and Y.R. Shen, Determination Of Molecular- Orientation Of Monolayer Adsorbates By Optical 2nd-Harmonic Generation. Physical Review A, 1983. 28(3): p. 1883-1885.

33. Gambinossi, F., P. Baglioni, and G. Caminati, Self-aggregation and phase separation of a styryl dye in monolayer at the liquid-air interface and in Langmuir-Blodgett films. Physical Chemistry Chemical Physics, 2004. 6(7): p. 1587-1591.

34. Gambinossi, F., M. Mannini, P. Baglioni, and G. Caminati, Spectroscopic properties of Langmuir-Blodgett films containing a potential-sensitive dye. Materials Science & Engineering C-Biomimetic And Supramolecular Systems, 2003. 23(6-8): p. 897-902.

35. Egner, A., S. Jakobs, and S.W. Hell, Fast 100-nm resolution three-dimensional microscope reveals structural plasticity of mitochondria in live yeast. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2002. 99(6): p. 3370-3375.

36. Hell, S. and E.H.K. Stelzer, Properties Of A 4pi Confocal Fluorescence Microscope. Journal Of The Optical Society Of America A-Optics Image Science And Vision, 1992. 9(12): p. 2159-2166.

133 37. Gustafsson, M.G.L., D.A. Agard, and J.W. Sedat, (IM)-M-5: 3D widefield light microscopy with better than 100 nm axial resolution. Journal Of Microscopy- Oxford, 1999. 195: p. 10-16.

38. Hell, S.W., Toward fluorescence nanoscopy. Nature Biotechnology, 2003. 21(11): p. 1347-1355.

39. Klar, T.A., E. Engel, and S.W. Hell, Breaking Abbe's diffraction resolution limit in fluorescence microscopy with stimulated emission depletion beams of various shapes. Physical Review E, 2001. 64(6).

40. Betzig, E. and J.K. Trautman, Near-Field Optics - Microscopy, Spectroscopy, And Surface Modification Beyond The Diffraction Limit. Science, 1992. 257(5067): p. 189-195.

41. Betzig, E., J.K. Trautman, T.D. Harris, J.S. Weiner, and R.L. Kostelak, Breaking The Diffraction Barrier - Optical Microscopy On A Nanometric Scale. Science, 1991. 251(5000): p. 1468-1470.

42. Rayleigh, L., On the theory of optical images with special reference to the microscope. Philosophical Magazine, 1896. 5: p. 42.

43. Synge, E.A., A Suggested Method for Extending Microscopic Resolution into the Ultra-Microscopic Region. Philosophical Magazine, 1928. 6: p. 356-362.

44. Ash, E.A. and G. Nicholls, Super-resolution Aperture Scanning Microscope. Nature, 1972. 237: p. 510-512.

45. Pohl, D.W., W. Denk, and M. Lanz, Optical Stethoscopy - Image Recording With Resolution Lambda/20. Applied Physics Letters, 1984. 44(7): p. 651-653.

46. Manual, O., Scanning Near-field Optical Microscope. 2002, Ulm, Germany: WiTec GmBH.

47. Hoppener, C., J.P. Siebrasse, R. Peters, U. Kubitscheck, and A. Naber, High- resolution near-field optical imaging of single nuclear pore complexes under physiological conditions. Biophysical Journal, 2005. 88(5): p. 3681-3688.

48. Koopman, M., A. Cambi, B.I. de Bakker, B. Joosten, C.G. Figdor, N.F. van Hulst, and M.F. Garcia-Parajo, Near-field scanning optical microscopy in liquid for high resolution single molecule detection on dendritic cells. Febs Letters, 2004. 573(1- 3): p. 6-10.

49. Kapkiai, L.K., D. Moore-Nichols, J. Carnell, J.R. Krogmeier, and R.C. Dunn, Hybrid near-field scanning optical microscopy tips for live cell measurements. Applied Physics Letters, 2004. 84(19): p. 3750-3752.

134 Bibliography

Ash, E. A., and G. Nicholls, Super-resolution Aperture Scanning Microscope: Nature, v. 237, p. 510-512. 1972

Barlow, C. A., The electrical double layer: Physical Chemistry, an Advanced Treatise: New York, Academic Press Inc, 167-246 p., 1970

Bedlack, R. S., M. D. Wei, S. Fox, E. Gross, and L. M. Loew, Endogenous Electric- Fields Within The Neuronal Plasmalemma - Nonuniformities Suggesting Differential Voltage-Dependence: Biophysical Journal, v. 66, p. A387-A387. 1994

Bedlack, R. S., M. D. Wei, S. H. Fox, E. Gross, and L. M. Loew, Distinct Electric Potentials In Soma And Neurite Membranes: Neuron, v. 13, p. 1187-1193. 1994

Bedlack, R. S., M. D. Wei, and L. M. Loew, Localized Membrane Depolarizations And Localized Calcium Influx During Electric Field-Guided Neurite Growth: Neuron, v. 9, p. 393-403. 1992

BenOren, I., G. Peleg, A. Lewis, B. Minke, and L. Loew, Infrared nonlinear optical measurements of membrane potential in photoreceptor cells: Biophysical Journal, v. 71, p. 1616-1620. 1996

Berkovic, G., G. Meshulam, and Z. Kotler, Measurement and analysis of molecular hyperpolarizability in the two-photon resonance regime: Journal Of Chemical Physics, v. 112, p. 3997-4003. 2000

Bersohn, R., Y.-H. Pao, and H. L. Frisch, Double-Quantum Light Scattering by Molecules: Journal Of Chemical Physics, v. 45, p. 3184-3198. 1966

Betzig, E., and J. K. Trautman, Near-Field Optics - Microscopy, Spectroscopy, And Surface Modification Beyond The Diffraction Limit: Science, v. 257, p. 189-195. 1992

Betzig, E., J. K. Trautman, T. D. Harris, J. S. Weiner, and R. L. Kostelak, Breaking The Diffraction Barrier - Optical Microscopy On A Nanometric Scale: Science, v. 251, p. 1468-1470. 1991

Blodgett, K. B., Monomolecular films of fatty acids on glass: Journal Of the American Chemcial Society, v. 56, p. 495. 1934

Blodgett, K. B., Films built by depositing successive monomolecular layers on a solid surface: Journal Of the American Chemcial Society, v. 57, p. 1007. 1935

135 Bloembergen, N., Nonlinear Optics: New York, W.A. Benjamin. 1965

Bosshard, C., M. Bosch, I. Liakatas, M. Jager, and P. Gunter, Second-Order Nonlinear Optical Organic Materials: Recent Developments: Nonlinear Optical Effects and Materials, Springer. 2000

Bosshard, C., G. Knopfle, P. Pretre, and P. Gunter, 2nd-Order Polarizabilities Of Nitropyridine Derivatives Determined With Electric-Field-Induced 2nd-Harmonic Generation And A Solvatochromic Method - A Comparative-Study: Journal Of Applied Physics, v. 71, p. 1594-1605. 1992

Bottcher, C. J. F., Theory of Electric Polarization, v. 1, Elsevier. 1973

Bouevitch, O., A. Lewis, I. Pinevsky, J. P. Wuskell, and L. M. Loew, Probing Membrane-Potential With Nonlinear Optics: Biophysical Journal, v. 65, p. 672- 679. 1993

Bourhill, G., J. L. Bredas, L. T. Cheng, S. R. Marder, F. Meyers, J. W. Perry, and B. G. Tiemann, Experimental Demonstration Of The Dependence Of The 1st Hyperpolarizability Of Donor-Acceptor-Substituted Polyenes On The Ground- State Polarization And Bond-Length Alternation: Journal Of The American Chemical Society, v. 116, p. 2619-2620. 1994

Boyd, R. W., Nonlinear Optics, Academic Press. 2003

Bredas, J. L., F. Meyers, B. M. Pierce, and J. Zyss, On The 2nd-Order Polarizability Of Conjugated Pi-Electron Molecules With Octupolar Symmetry - The Case Of Triaminotrinitrobenzene: Journal Of The American Chemical Society, v. 114, p. 4928-4929. 1992

Brunschwig, B. S., S. Ehrenson, and N. Sutin, Solvent Reorganization In Optical And Thermal Electron-Transfer Processes - Solvatochromism And Intramolecular Electron-Transfer Barriers In Spheroidal Molecules: Journal Of Physical Chemistry, v. 91, p. 4714-4723. 1987

Byer, R. L., and S. E. Harris, Power and bandwidth of spontaneous parametric emission: Physical Review, v. 168, p. 1064. 1968

Campagnola, P. J., H. A. Clark, W. A. Mohler, A. Lewis, and L. M. Loew, Second- harmonic imaging microscopy of living cells: Journal Of Biomedical Optics, v. 6, p. 277-286. 2001

Campagnola, P. J., M. D. Wei, A. Lewis, and L. M. Loew, High-resolution nonlinear optical imaging of live cells by second harmonic generation: Biophysical Journal, v. 77, p. 3341-3349. 1999

Chemla, D. S., and J. Zyss, Nonlinear Optical Properties of Organic Molecules and Crystals, v. 1 and 2: New York, Academic Press. 1987

136 Cheng, L. T., Second-Order Nonlinear Optical Properties of Organic Materials: Handbook of Photonics, CRC Press. 1996

Chui, T. W., and K. Y. Wong, Study of hyper-Rayleigh scattering and two-photon absorption induced fluorescence from crystal violet: Journal Of Chemical Physics, v. 109, p. 1391-1396. 1998

Clarke, R. J., The dipole potential of phospholipid membranes and methods for its detection: Advances In Colloid And Interface Science, v. 89, p. 263-281. 2001

Clays, K., and A. Persoons, Hyper-Rayleigh Scattering In Solution: Physical Review Letters, v. 66, p. 2980-2983. 1991

Clays, K., and A. Persoons, Hyper-Rayleigh Scattering In Solution: Review Of Scientific Instruments, v. 63, p. 3285-3289. 1992

Cohen, L. B., B. M. Salzberg, H. V. Davilla, W. N. Ross, D. Landowne, A. S. Waggoner, and C. H. Wang, Changes in axon fluorescence during activity: molecular probes of membrane potential: Journal Of Membrane Biology, v. 19, p. 1. 1974

Davies, J. T., and E. K. Rideal, Interfacial Phenomena: New York, Academic Press Inc. 1963

Denk, W., J. H. Strickler, and W. W. Webb, 2-Photon Laser Scanning Fluorescence Microscopy: Science, v. 248, p. 73-76. 1990

DiFranco, M., J. Capote, and J. L. Vergara, Optical imaging and functional characterization of the transverse tubular system of mammalian muscle fibers using the potentiometric indicator di-8-ANEPPS: Journal Of Membrane Biology, v. 208, p. 141-153. 2005

Dombeck, D. A., L. Sacconi, M. Blanchard-Desce, and W. W. Webb, Optical recording of fast neuronal membrane potential transients in acute mammalian brain slices by second-harmonic generation microscopy: Journal Of Neurophysiology, v. 94, p. 3628-3636. 2005

Egner, A., S. Jakobs, and S. W. Hell, Fast 100-nm resolution three-dimensional microscope reveals structural plasticity of mitochondria in live yeast: Proceedings Of The National Academy Of Sciences Of The United States Of America, v. 99, p. 3370-3375. 2002

Eisenthal, K. B., Liquid interfaces probed by second-harmonic and sum-frequency spectroscopy: Chemical Reviews, v. 96, p. 1343-1360. 1996

Fisher, J. A. N., B. M. Salzberg, and A. G. Yodh, Near infrared two-photon excitation cross-sections of voltage-sensitive dyes: Journal Of Neuroscience Methods, v. 148, p. 94-102. 2005

137 Fluhler, E., V. G. Burnham, and L. M. Loew, Spectra, Membrane-Binding, And Potentiometric Responses Of New Charge Shift Probes: Biochemistry, v. 24, p. 5749-5755. 1985

Franklin, J. C., and D. S. Cafiso, Internal Electrostatic Potentials In Bilayers - Measuring And Controlling Dipole Potentials In Lipid Vesicles: Biophysical Journal, v. 65, p. 289-299. 1993

Frisch, M. J., G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, and e. al., Gaussian 03 Revision C.02, Wallingford CT., Gaussian Inc. 2004

Fromherz, P., Monopole Dipole Model For Symmetrical Solvatochromism Of Hemicyanine Dyes: Journal Of Physical Chemistry, v. 99, p. 7188-7192. 1995

Gambinossi, F., P. Baglioni, and G. Caminati, Self-aggregation and phase separation of a styryl dye in monolayer at the liquid-air interface and in Langmuir-Blodgett films: Physical Chemistry Chemical Physics, v. 6, p. 1587-1591. 2004

Gambinossi, F., M. Mannini, P. Baglioni, and G. Caminati, Spectroscopic properties of Langmuir-Blodgett films containing a potential-sensitive dye: Materials Science & Engineering C-Biomimetic And Supramolecular Systems, v. 23, p. 897-902. 2003

Girling, I. R., N. A. Cade, P. V. Kolinsky, J. D. Earls, G. H. Cross, and I. R. Peterson, Observation Of 2nd-Harmonic Generation From Langmuir-Blodgett Multilayers Of A Hemicyanine Dye: Thin Solid Films, v. 132, p. 101-112. 1985

Girling, I. R., P. V. Kolinsky, N. A. Cade, J. D. Earls, and I. R. Peterson, 2nd Harmonic- Generation From Alternating Langmuir-Blodgett Films: Optics Communications, v. 55, p. 289-292. 1985

Gorman, C. B., and S. R. Marder, An Investigation Of The Interrelationships Between Linear And Nonlinear Polarizabilities And Bond-Length Alternation In Conjugated Organic-Molecules: Proceedings Of The National Academy Of Sciences Of The United States Of America, v. 90, p. 11297-11301. 1993

Grahame, D. C., The electrical double layer and the theory of electrocapillarity: Chemical Reviews, v. 41, p. 441-501. 1947

Gross, E., R. S. Bedlack, and L. M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurement Of The Membrane Dipole Potential: Biophysical Journal, v. 66, p. A387-A387. 1994

Gupta, R. K., B. M. Salzberg, A. Grinvald, L. B. Cohen, K. Kamino, S. Lesher, M. B. Boyle, A. S. Waggoner, and C. H. Wang, Improvements In Optical Methods For Measuring Rapid Changes In Membrane-Potential: Journal Of Membrane Biology, v. 58, p. 123-137. 1981

138 Gustafsson, M. G. L., D. A. Agard, and J. W. Sedat, (IM)-M-5: 3D widefield light microscopy with better than 100 nm axial resolution: Journal Of Microscopy- Oxford, v. 195, p. 10-16. 1999

Hassner, A., D. Birnbaum, and L. M. Loew, Charge-Shift Probes Of Membrane-Potential - Synthesis: Journal Of Organic Chemistry, v. 49, p. 2546-2551. 1984

Heesink, G. J. T., A. G. T. Ruiter, N. F. Vanhulst, and B. Bolger, Determination Of Hyperpolarizability Tensor Components By Depolarized Hyper Rayleigh- Scattering: Physical Review Letters, v. 71, p. 999-1002. 1993

Heinz, T. F., H. W. K. Tom, and Y. R. Shen, Determination Of Molecular-Orientation Of Monolayer Adsorbates By Optical 2nd-Harmonic Generation: Physical Review A, v. 28, p. 1883-1885. 1983

Hell, S., and E. H. K. Stelzer, Properties Of A 4pi Confocal Fluorescence Microscope: Journal Of The Optical Society Of America A-Optics Image Science And Vision, v. 9, p. 2159-2166. 1992

Hell, S. W., Toward fluorescence nanoscopy: Nature Biotechnology, v. 21, p. 1347-1355. 2003

Hladky, S. B., and D. A. Haydon, Membrane conductance and surface potential: Biochim. Biophys. Acta, v. 318, p. 464-468. 1973

Hoffmann, R., An Extended Hückel Theory. I. Hydrocarbons: Journal Of Chemical Physics, v. 39, p. 1397. 1963

Honig, B. H., W. L. Hubbell, and R. F. Flewelling, Electrostatic Interactions In Membranes And Proteins: Annual Review Of Biophysics And Biophysical Chemistry, v. 15, p. 163-193. 1986

Hoppener, C., J. P. Siebrasse, R. Peters, U. Kubitscheck, and A. Naber, High-resolution near-field optical imaging of single nuclear pore complexes under physiological conditions: Biophysical Journal, v. 88, p. 3681-3688. 2005

Huang, J. Y., A. Lewis, and L. Loew, Nonlinear optical properties of potential sensitive styryl dyes: Biophys J, v. 53, p. 665-70. 1988

Hubbard, S. F., R. G. Petschek, and K. D. Singer, Spectral content and dispersion of hyper-Rayleigh scattering: Optics Letters, v. 21, p. 1774-1776. 1996

Hubener, G., A. Lambacher, and P. Fromherz, Anellated hemicyanine dyes with large symmetrical solvatochromism of absorption and fluorescence: Journal Of Physical Chemistry B, v. 107, p. 7896-7902. 2003

139 Huesmann, H., G. Gabrielli, and G. Caminati, Monolayers and Langmuir-Blodgett films of the electrochromic dye Di-8-ANEPPS: Thin Solid Films, v. 329, p. 804-807. 1998

Hwang, J., L. A. Gheber, L. Margolis, and M. Edidin, Domains in cell plasma membranes investigated by near-field scanning optical microscopy: Biophysical Journal, v. 74, p. 2184-2190. 1998

Jerphagnon, J., and S. K. Kurtz, Maker Fringes: A Detailed Comparison of Theory and Experiment for Isotropic and Uniaxial Crystals: Journal Of Applied Physics, v. 41, p. 1667-1681. 1970

Kaatz, P., and D. P. Shelton, Polarized hyper-Rayleigh light scattering measurements of nonlinear optical chromophores: Journal Of Chemical Physics, v. 105, p. 3918- 3929. 1996

Kalb, E., S. Frey, and L. K. Tamm, Formation Of Supported Planar Bilayers By Fusion Of Vesicles To Supported Phospholipid Monolayers: Biochimica Et Biophysica Acta, v. 1103, p. 307-316. 1992

Kanis, D. R., M. A. Ratner, and T. J. Marks, Design And Construction Of Molecular Assemblies With Large 2nd-Order Optical Nonlinearities - Quantum-Chemical Aspects: Chemical Reviews, v. 94, p. 195-242. 1994

Kapkiai, L. K., D. Moore-Nichols, J. Carnell, J. R. Krogmeier, and R. C. Dunn, Hybrid near-field scanning optical microscopy tips for live cell measurements: Applied Physics Letters, v. 84, p. 3750-3752. 2004

Kelley, A. M., Frequency-dependent first hyperpolarizabilities from linear absorption spectra: Journal Of The Optical Society Of America B-Optical Physics, v. 19, p. 1890-1900. 2002

Kenis, P. J. A., O. F. J. Noordman, S. Houbrechts, G. J. van Hummel, S. Harkema, F. van Veggel, K. Clays, J. F. J. Engbersen, A. Persoons, N. F. van Hulst, and D. N. Reinhoudt, Second-order nonlinear optical properties of the four tetranitrotetrapropoxycalix[4]arene conformers: Journal Of The American Chemical Society, v. 120, p. 7875-7883. 1998

Klar, T. A., E. Engel, and S. W. Hell, Breaking Abbe's diffraction resolution limit in fluorescence microscopy with stimulated emission depletion beams of various shapes: Physical Review E, v. 64. 2001

Klessinger, M., and J. Michl, Excited States and Photochemistry of Organic Molecules: New York, VCH. 1995

Koopman, M., A. Cambi, B. I. de Bakker, B. Joosten, C. G. Figdor, N. F. van Hulst, and M. F. Garcia-Parajo, Near-field scanning optical microscopy in liquid for high

140 resolution single molecule detection on dendritic cells: Febs Letters, v. 573, p. 6- 10. 2004

Kuhn, B., P. Fromherz, and W. Denk, High sensitivity of stark-shift voltage-sensing dyes by one- or two-photon excitation near the red spectral edge: Biophysical Journal, v. 87, p. 631-639. 2004

Kurtz, S. K., Measurement of nonlinear optical susceptibilities: Quantum Elecronics, v. I: New York, Academic Press, 209-281 p., 1975

Kurtz, S. K., and T. T. Perry, A power technique for the evaluation of nonlinear optical materials: Journal Of Applied Physics, v. 39, p. 3798. 1968

Kuzyk, M. G., and C. W. Dirk, Characterization techniques and tabulations for organic nonlinear optical materials, Marcel Dekker Inc, 489 p., 1998

Laage, D., W. H. Thompson, M. Blanchard-Desce, and J. T. Hynes, Charged push-pull polyenes in solution: Anomalous solvatochromism and nonlinear optical properties: Journal Of Physical Chemistry A, v. 107, p. 6032-6046. 2003

Langmuir, I., The mechanism of the surface phenomenon of flotation: Trans. Faraday Soc., v. 15. 1920

Leray, A., L. Leroy, Y. Le Grand, C. Odin, A. Renault, V. Vie, D. Rouede, T. Mallegol, O. Mongin, M. H. V. Werts, and M. Blanchard-Desce, Organization and orientation of amphiphilic push-pull chromophores deposited in Langmuir- Blodgett monolayers studied by second harmonic generation and atomic force microscopy: Langmuir, v. 20, p. 8165-8171. 2004

Levine, B. F., and C. G. Bethea, Second and third order hyperpolarizabilities of organic molecules: Journal Of Chemical Physics, v. 63, p. 2666-2682. 1975

Liberman, E. A., and V. P. Topaly, Permeability of biomolecular phospholipid membranes for fat-soluble ions: Biophysics, v. 14, p. 477-487. 1969

Loew, L. M., Design And Characterization Of Electrochromic Membrane Probes: Journal Of Biochemical And Biophysical Methods, v. 6, p. 243-260. 1982

Loew, L. M., How to choose a potentiometric membrane probe: Spectroscopic Membrane Probes, v. II, CRC Press, 139-151 p., 1988

Loew, L. M., The electrical properties of membranes: Biomembranes. Physical Aspects: Weinheim, Germany, VCH Publishers, 341-371 p., 1993

Loew, L. M., Potentiometric dyes: Imaging electrical activity of cell membranes: Pure And Applied Chemistry, v. 68, p. 1405-1409. 1996

141 Loew, L. M., G. W. Bonneville, and J. Surow, Charge shift optical probes of membrane potential. Theory: Biochemistry, v. 17, p. 4065. 1978

Loew, L. M., L. B. Cohen, J. Dix, E. N. Fluhler, V. Montana, G. Salama, and J. Y. Wu, A Naphthyl Analog Of The Aminostyryl Pyridinium Class Of Potentiometric Membrane Dyes Shows Consistent Sensitivity In A Variety Of Tissue, Cell, And Model Membrane Preparations: Journal Of Membrane Biology, v. 130, p. 1-10. 1992

Loew, L. M., L. B. Cohen, B. M. Salzberg, A. L. Obaid, and F. Bezanilla, Charge-Shift Probes Of Membrane-Potential - Characterization Of Aminostyrylpyridinium Dyes On The Squid Giant-Axon: Biophysical Journal, v. 47, p. 71-77. 1985

Loew, L. M., and L. L. Simpson, Charge-Shift Probes Of Membrane-Potential - A Probable Electrochromic Mechanism For Para-Aminostyrylpyridinium Probes On A Hemispherical Lipid Bilayer: Biophysical Journal, v. 34, p. 353-365. 1981

Maker, P. D., R. W. Terhune, M. Nisenoff, and C. M. Savage, Effects of dispersion and focusing on the production of optical harmonics: Physical Review Letters, v. 8, p. 21. 1962

Mannini, M., F. Gambinossi, P. Baglioni, and G. Caminati, Immobilization of a fluorescent dye in Langmuir-Blodgett films: Bioelectrochemistry, v. 63, p. 9-12. 2004

Manual, O., Scanning Near-field Optical Microscope: Ulm, Germany, WiTec GmBH. 2002

Marder, S. R., D. N. Beratan, and L. T. Cheng, Approaches For Optimizing The 1st Electronic Hyperpolarizability Of Conjugated Organic-Molecules: Science, v. 252, p. 103-106. 1991

Marder, S. R., L. T. Cheng, B. G. Tiemann, A. C. Friedli, M. Blancharddesce, J. W. Perry, and J. Skindhoj, Large 1st Hyperpolarizabilities In Push-Pull Polyenes By Tuning Of The Bond-Length Alternation And Aromaticity: Science, v. 263, p. 511-514. 1994

Marder, S. R., C. B. Gorman, F. Meyers, J. W. Perry, G. Bourhill, J. L. Bredas, and B. M. Pierce, A Unified Description Of Linear And Nonlinear Polarization In Organic Polymethine Dyes: Science, v. 265, p. 632-635. 1994

Marder, S. R., J. W. Perry, and W. P. Schaefer, Synthesis Of Organic Salts With Large 2nd-Order Optical Nonlinearities: Science, v. 245, p. 626-628. 1989

Marowsky, G., A. Gierulski, R. Steinhoff, D. Dorsch, R. Eidenschnik, and B. Rieger, Efficiency Studies Of 2nd-Harmonic-Active Organic-Dye Coverages: Journal Of The Optical Society Of America B-Optical Physics, v. 4, p. 956-961. 1987

142 McLaughlin, S., Electrostatic potentials at membrane-solution interfaces: Current Topics Membranes and Transport: New York, Academic Press, 71-144 p., 1977

McLaughlin, S., The Electrostatic Properties Of Membranes: Annual Review Of Biophysics And Biophysical Chemistry, v. 18, p. 113-136. 1989

Millard, A. C., L. Jin, A. Lewis, and L. M. Loew, Direct measurement of the voltage sensitivity of second-harmonic generation from a membrane dye in patch- clamped cells: Optics Letters, v. 28, p. 1221-1223. 2003

Millard, A. C., L. Jin, M. D. Wei, J. P. Wuskell, A. Lewis, and L. M. Loew, Sensitivity of second harmonic generation from styryl dyes to transmembrane potential: Biophysical Journal, v. 86, p. 1169-1176. 2004

Millard, A. C., L. Jin, J. P. Wuskell, D. M. Boudreau, A. Lewis, and L. M. Loew, Wavelength- and time-dependence of potentiometric non-linear optical signals from styryl dyes: Journal Of Membrane Biology, v. 208, p. 103-111. 2005

Milonni, P. W., and J. H. Eberly, Lasers, J. Wiley and Sons. 1998

Montana, V., D. L. Farkas, and L. M. Loew, Dual-Wavelength Ratiometric Fluorescence Measurements Of Membrane-Potential: Biochemistry, v. 28, p. 4536-4539. 1989

Moran, A. M., C. Delbecque, and A. M. Kelley, Solvent effects on ground and excited electronic state structures of the push-pull chromophore julolidinyl-n-N,N '- diethylthiobarbituric acid: Journal Of Physical Chemistry A, v. 105, p. 10208- 10219. 2001

Moran, A. M., D. S. Egolf, M. Blanchard-Desce, and A. M. Kelley, Vibronic effects on solvent dependent linear and nonlinear optical properties of push-pull chromophores: Julolidinemalononitrile: Journal Of Chemical Physics, v. 116, p. 2542-2555. 2002

Moran, A. M., and A. M. Kelley, Solvent effects on ground and excited electronic state structures of p-nitroaniline: Journal Of Chemical Physics, v. 115, p. 912-924. 2001

Moreaux, L., T. Pons, V. Dambrin, M. Blanchard-Desce, and J. Mertz, Electro-optic response of second-harmonic generation membrane potential sensors: OPTICS LETTERS, v. 28, p. 625-627. 2003

Moreaux, L., O. Sandre, M. Blanchard-Desce, and J. Mertz, Membrane imaging by simultaneous second-harmonic generation and two-photon microscopy: OPTICS LETTERS, v. 25, p. 320-322. 2000

Moreaux, L., O. Sandre, S. Charpak, M. Blanchard-Desce, and J. Mertz, Coherent scattering in multi-harmonic light microscopy: Biophysical Journal, v. 80, p. 1568-1574. 2001

143 Moreaux, L., O. Sandre, and J. Mertz, Membrane imaging by second-harmonic generation microscopy: Journal Of The Optical Society Of America B-Optical Physics, v. 17, p. 1685-1694. 2000

Moylan, C. R., R. J. Twieg, V. Y. Lee, S. A. Swanson, K. M. Betterton, and R. D. Miller, Nonlinear-Optical Chromophores With Large Hyperpolarizabilities And Enhanced Thermal Stabilities: Journal Of The American Chemical Society, v. 115, p. 12599-12600. 1993

Neher, E., and B. Sakmann, Single channel currents recorded from membrane of denervated frog muscle fibres: Nature, v. 260, p. 799-802. 1976

Noordman, O. F. J., and N. F. vanHulst, Time-resolved hyper-Rayleigh scattering: Measuring first hyperpolarizabilities beta of fluorescent molecules: Chemical Physics Letters, v. 253, p. 145-150. 1996

Nuriya, M., J. Jiang, B. Nemet, K. B. Eisenthal, and R. Yuste, Imaging membrane potential in dendritic spines: Proceedings Of The National Academy Of Sciences Of The United States Of America, v. 103, p. 786-790. 2006

Olbrechts, G., K. Wostyn, K. Clays, and A. Persoons, High-frequency demodulation of multiphoton fluorescence in long-wavelength hyper-Rayleigh scattering: Optics Letters, v. 24, p. 403-405. 1999

Ostroverkhov, V., R. G. Petschek, K. D. Singer, L. Sukhomlinova, R. J. Twieg, S. X. Wang, and L. C. Chien, Measurements of the hyperpolarizability tensor by means of hyper-Rayleigh scattering: Journal Of The Optical Society Of America B- Optical Physics, v. 17, p. 1531-1542. 2000

Oudar, J. L., Optical nonlinearities of conjugated molecules. Stilbene derivatives and highly polar aromatic compounds: Journal Of Chemical Physics, v. 67, p. 446-457. 1977

Oudar, J. L., and D. S. Chemla, Theory of Optical Second-Order Susceptibilities of benzene substitutes: Opt. Comm., v. 13, p. 164-168. 1975

Oudar, J. L., and D. S. Chemla, Hyperpolarizabilities of the nitroanilines and their relations to the excited state dipole moment: Journal Of Chemical Physics, v. 66, p. 2664. 1977

Paley, M. S., J. M. Harris, H. Looser, J. C. Baumert, G. C. Bjorklund, D. Jundt, and R. J. Twieg, A Solvatochromic Method For Determining 2nd-Order Polarizabilities Of Organic-Molecules: Journal Of Organic Chemistry, v. 54, p. 3774-3778. 1989

Pariser, R., and R. G. Parr, A Semi-Empirical Theory of the Electronic Spectra and Electronic Structure of Complex Unsaturated Molecules. I.: Journal Of Chemical Physics, v. 21, p. 466. 1953

144 Platt, J. R., Electrochromism, a possiblechange of color producible in dyes by an electric filed: Journal Of Chemical Physics, v. 25, p. 80. 1956

Pohl, D. W., W. Denk, and M. Lanz, Optical Stethoscopy - Image Recording With Resolution Lambda/20: Applied Physics Letters, v. 44, p. 651-653. 1984

Pons, T., and J. Mertz, Membrane potential detection with second-harmonic generation and two-photon excited fluorescence: A theoretical comparison: Optics Communications, v. 258, p. 203-209. 2006

Pons, T., L. Moreaux, O. Mongin, M. Blanchard-Desce, and J. Mertz, Mechanisms of membrane potential sensing with second-harmonic generation microscopy: JOURNAL OF BIOMEDICAL OPTICS, v. 8, p. 428-431. 2003

Pople, J. A., Electron interaction in unsaturated hydrocarbons: Transactions of the Faraday Society, v. 49, p. 1375. 1953

Pople, J. A., D. P. Santry, and G. A. Segal, Approximate Self-Consistent Molecular Orbital Theory. I. Invariant Procedures: Journal Of Chemical Physics, v. 43, p. S129. 1965

Prasad, P. N., and D. J. Williams, Introduction to Nonlinear Optical Effects in Molecules and Polymers, John Wiley & Sons, Inc. 1991

Rayleigh, L., On the theory of optical images with special reference to the microscope: Philosophical Magazine, v. 5, p. 42. 1896

Ridley, J., and M. Zerner, An Intermediate Neglect of Differential Overlap Technique for Spectroscopy: Pyrrole and the Azines: Theoretica Chemica Acta (Berl.), v. 32, p. 111. 1973

Ries, R. S., H. Choi, R. Blunck, F. Bezanilla, and J. R. Heath, Black lipid membranes: visualizing the structure, dynamics, and substrate dependence of membranes: Journal Of Physical Chemistry B, v. 108, p. 16040-16049. 2004

Roberts, G., Langmuir-Blodgett Films: New York, Plenum Press. 1990

Ross, W. N., B. M. Salzberg, L. B. Cohen, A. Grinvald, H. V. Davilla, A. S. Waggoner, and C. H. Wang, Changes in absorption, fluorescence, dichroism, and birefringence in stained axons: optical measurement of membrane potential: Journal Of Membrane Biology, v. 33, p. 141. 1977

Sacconi, L., D. A. Dombeck, and W. W. Webb, Overcoming photodamage in second- harmonic generation microscopy: Real-time optical recording of neuronal action potentials: Proceedings Of The National Academy Of Sciences Of The United States Of America, v. 103, p. 3124-3129. 2006

145 Sackmann, E., Supported membranes: Scientific and practical applications: Science, v. 271, p. 43-48. 1996

Shen, Y. R., The Principles of Nonlinear Optics: New York, John Wiley & Sons, Inc. 1984

Shen, Y. R., Surface-Properties Probed By 2nd-Harmonic And Sum-Frequency Generation: Nature, v. 337, p. 519-525. 1989

Siano, D. B., and D. E. Metzler, Band shapes of the electronic spectra of complex molecules: Journal Of Chemical Physics, v. 51, p. 1856. 1969

Sims, P. J., A. S. Waggoner, C. H. Wang, and J. F. Hoffmann, Studies on the mechanism by which cyanine dyes measure membrane potential in red blood cells and phosphatidyl chroline vesicles: Biochemistry, v. 13, p. 3115. 1974

Singer, K. D., and A. F. Garito, Measurements Of Molecular 2nd Order Optical Susceptibilities Using Dc Induced 2nd Harmonic-Generation: Journal Of Chemical Physics, v. 75, p. 3572-3580. 1981

Singer, K. D., M. G. Kuzyk, and J. E. Sohn, 2nd-Order Nonlinear-Optical Processes In Orientationally Ordered Materials - Relationship Between Molecular And Macroscopic Properties: Journal Of The Optical Society Of America B-Optical Physics, v. 4, p. 968-976. 1987

Singer, S. J., and G. L. Nicolson, The Fluid Mosaic Model of the Structure of Cell Membranes: Science, v. 175, p. 720-731. 1972

Skou, J. C., The influence of some cations on an adenosine triphosphatase from peripheral nerves: Biochim. Biophys. Acta, v. 23, p. 394-401. 1957

Solletti, J. M., M. Botreau, F. Sommer, W. L. Brunat, S. Kasas, T. M. Duc, and M. R. Celio, Elaboration and characterization of phospholipid Langmuir-Blodgett films: Langmuir, v. 12, p. 5379-5386. 1996

Solletti, J. M., M. Botreau, F. Sommer, T. M. Duc, and M. R. Celio, Characterization of mixed miscible and nonmiscible phospholipid Langmuir-Blodgett films by atomic force microscopy: Journal Of Vacuum Science & Technology B, v. 14, p. 1492- 1497. 1996

Spence, D. E., P. N. Kean, and W. Sibbett, 60-Fsec Pulse Generation From A Self-Mode- Locked Ti-Sapphire Laser: Optics Letters, v. 16, p. 42-44. 1991

Suppan, P., and N. Ghoneim, Solvatochromism, Cambridge, Royal Society of Chemistry. 1997

Sutherland, R. L., Handbook of Nonlinear Optics, Marcel Dekker, Inc. 2003

146 Synge, E. A., A Suggested Method for Extending Microscopic Resolution into the Ultra- Microscopic Region: Philosophical Magazine, v. 6, p. 356-362. 1928

Tamm, L. K., and H. M. McConnell, Supported Phospholipid-Bilayers: Biophysical Journal, v. 47, p. 105-113. 1985

Thompson, W. H., M. Blanchard-Desce, V. Alain, J. Muller, A. Fort, M. Barzoukas, and J. T. Hynes, Two valence bond state model for molecular nonlinear optical properties. Comparison with push-pull polyene solution measurements: Journal Of Physical Chemistry A, v. 103, p. 3766-3771. 1999

Thompson, W. H., M. Blanchard-Desce, and J. T. Hynes, Two valence bond state model for molecular nonlinear optical properties. Nonequilibrium solvation formulation: Journal Of Physical Chemistry A, v. 102, p. 7712-7722. 1998

Tokumasu, F., J. Hwang, and J. A. Dvorak, Heterogeneous molecular distribution in supported multicomponent lipid bilayers: Langmuir, v. 20, p. 614-618. 2004

Waggoner, A. S., Dye indicators of membrane potential: Annual Review Of Biophysics And Bioengineering, v. 8, p. 847. 1979

Wang, C. H., Effects of dephasing and vibronic structure on the first hyperpolarizability of strongly charge-transfer molecules: Journal Of Chemical Physics, v. 112, p. 1917-1924. 2000

Wang, C. H., J. N. Woodford, C. Zhang, and L. R. Dalton, Resonant and nonresonant hyper-Rayleigh scattering of charge-transfer chromophores: Journal Of Applied Physics, v. 89, p. 4209-4217. 2001

Ward, J. F., Calculation of Nonlinear Optical Susceptibilities Using Diagrammatic Perturbation Theory: Reviews of Modern Physics, v. 37, p. 1. 1965

Whitaker, C. M., E. V. Patterson, K. L. Kott, and R. J. McMahon, Nitrogen and oxygen donors in nonlinear optical materials: Effects of alkyl vs phenyl substitution on the molecular hyperpolarizability: Journal Of The American Chemical Society, v. 118, p. 9966-9973. 1996

Xu, C., B. M. Slepchenko, and L. M. Loew, Surface potential measurements on biological membranes with voltage-sensitive dyes: Biophysical Journal, v. 82, p. 546A-546A. 2002

Yariv, A., Quantum Electronics, 2nd Edition: New York, Wiley. 1975

Yoshimura, T., Enhancing 2nd-Order Nonlinear Optical-Properties By Controlling The Wave-Function In One-Dimensional Conjugated Molecules: Physical Review B, v. 40, p. 6292-6298. 1989

147 Zerner, M., Semiempirical Molecular Orbital Methods: Reviews in Computational Chemistry, v. 2. 1991

Zhang, J., R. M. Davidson, M. D. Wei, and L. M. Loew, Membrane electric properties by combined patch clamp and fluorescence ratio imaging in single neurons: Biophysical Journal, v. 74, p. 48-53. 1998

Zhang, J., M. D. Wei, and L. M. Loew, Change of surface potential induced by intracellular digestion of membrane proteins: Biophysical Journal, v. 74, p. A402- A402. 1998

Zyss, J., Hyperpolarizabilities of substituted conjugated molecules. II. Substituent effects and respective – contributions: Journal Of Chemical Physics, v. 70, p. 3341. 1979

Zyss, J., and J. L. Oudar, Relations Between Microscopic And Macroscopic Lowest- Order Optical Nonlinearities Of Molecular-Crystals With One-Dimensional Or Two-Dimensional Units: Physical Review A, v. 26, p. 2028-2048. 1982

Zyss, J., T. C. Van, C. Dhenaut, and I. Ledoux, Harmonic Rayleigh-Scattering From Nonlinear Octupolar Molecular Media - The Case Of Crystal Violet: Chemical Physics, v. 177, p. 281-296. 1993

148