USF Libraries
USF Digital Collections

Integrated uv-vis multiangle-multiwavelength spectrometer for characterization of micron and sub-micron size particles

MISSING IMAGE

Material Information

Title:
Integrated uv-vis multiangle-multiwavelength spectrometer for characterization of micron and sub-micron size particles
Physical Description:
Book
Language:
English
Creator:
Kim, Yong-Rae
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Light scattering
Transmission
Fluorescence
Joint particle property distributions
Particle shape and composition
Dissertations, Academic -- Physics -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Characterization of micron and sub-micron size particles requires the simultaneous measurement of the joint particle property distribution (JPPD). The JPPD is comprised of particle size, shape, orientation, composition, optical properties, and surface properties. Measurement of each of the particle properties independently is a difficult task and it has been only partially successful. To determine as many particle properties as possible using optical methods it is necessary to simultaneously measure all aspects of the interaction of the incident light with the particles of interest. This approach leads to the concept of multidimensional spectroscopy suggested by Prof. Garcia-Rubio. Dr. Bacon proved the proposition by developing and testing a prototype multianglemultiwavelength (MAMW) spectrometer proposed by Prof. Garcia-Rubio.However, the prototype MAMW spectrometer has limitations in the amount of information it can obtain because of strong absorption of deep UV light and detector saturation due to the use of optical fibers and single integration time for the CCD detector. The Integrated UV-VIS MAMW spectrometer has been developed to overcome the limitations of the prototype MAMW spectrometer. Improvements have become possible through the use of UV lenses and integration time multiplexing (ITM). The Integrated UV-VIS MAMW spectrometer has the capabilities to perform low angle scattering measurements starting from 4o with simultaneous detection of multiwavelength light from 200 nm to 820 nm, UV-VIS transmission spectroscopy, and frequency domain fluorescence spectroscopy. Following the development, possible sources of errors were analyzed and data calibration procedures have been established toensure the validity and reproducibility of the measurement results.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Yong-Rae Kim.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 323 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001681168
usfldc doi - E14-SFE0001149
usfldc handle - e14.1149
System ID:
SFS0025470:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681168
003 fts
005 20060215104709.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 051012s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001149
035
SFE0001149
040
FHM
c FHM
090
2005
1 100
Kim, Yong-Rae.
0 245
Integrated uv-vis multiangle-multiwavelength spectrometer for characterization of micron and sub-micron size particles
h [electronic resource] /
by Yong-Rae Kim.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (Ph.D.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 323 pages.
520
ABSTRACT: Characterization of micron and sub-micron size particles requires the simultaneous measurement of the joint particle property distribution (JPPD). The JPPD is comprised of particle size, shape, orientation, composition, optical properties, and surface properties. Measurement of each of the particle properties independently is a difficult task and it has been only partially successful. To determine as many particle properties as possible using optical methods it is necessary to simultaneously measure all aspects of the interaction of the incident light with the particles of interest. This approach leads to the concept of multidimensional spectroscopy suggested by Prof. Garcia-Rubio. Dr. Bacon proved the proposition by developing and testing a prototype multianglemultiwavelength (MAMW) spectrometer proposed by Prof. Garcia-Rubio.However, the prototype MAMW spectrometer has limitations in the amount of information it can obtain because of strong absorption of deep UV light and detector saturation due to the use of optical fibers and single integration time for the CCD detector. The Integrated UV-VIS MAMW spectrometer has been developed to overcome the limitations of the prototype MAMW spectrometer. Improvements have become possible through the use of UV lenses and integration time multiplexing (ITM). The Integrated UV-VIS MAMW spectrometer has the capabilities to perform low angle scattering measurements starting from 4o with simultaneous detection of multiwavelength light from 200 nm to 820 nm, UV-VIS transmission spectroscopy, and frequency domain fluorescence spectroscopy. Following the development, possible sources of errors were analyzed and data calibration procedures have been established toensure the validity and reproducibility of the measurement results.
590
Adviser: Myung Keun Kim.
Co-adviser: Luis H. Garcia-Rubio
653
Light scattering.
Transmission.
Fluorescence.
Joint particle property distributions.
Particle shape and composition.
690
Dissertations, Academic
z USF
x Physics
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1149



PAGE 1

Integrated Uv-Vis Multiangle-Multiwavelength Spectrometer For Characterization Of Micron And Sub-Micron Size Particles by Yong-Rae Kim A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Physics College of Arts and Sciences University of South Florida Co-Major Professor: Luis Humbolt Garcia-Rubio Ph.D. Co-Major Professor: Myung Keun Kim, Ph.D. Nicholas Djeu, Ph.D. Pritish Mukherjee, Ph.D. Maria Kallergi, Ph.D. Date of Approval: December 9, 2004 Keywords: light scattering, transmission, fl uorescence, joint pa rticle property distributions, particle shape and composition Copyright 2005 Yong-Rae Kim

PAGE 2

Color illustrations are necessary for clear distinction of the intensities of UV-VIS Multiangle-Multiwavelength response surface plot s. Color illustrations can be found as hard copy at the University of South Flor ida library or electronically found from the Electronic library.

PAGE 3

DEDICATION I dedicate this dissertation to my late father-in-law Jeong-Hoon Lee who missed his beloved youngest daughter until his last moment. To my mother-in-law Ki-Soon Han. To my father Myung-Soo Kim. To my mo ther Jeong-Hee Seo. To my beloved wife Kyung-Hee Lee. To my son Sun-Joong Kim. W ithout their love, patience, and sacrifice, this work would be impossible.

PAGE 4

ACKNOWLEDGEMENTS I would like to express my deepest appreciation to my advisors Dr. Myung Keun Kim and Dr. Luis H. Garcia-Rubio. Without Dr. Kim and Dr. Garcia-Rubios valuable understanding, patience, and guidanc e, this research would not be successful. I would like to express my thanks to Dr. Nicholas Djeu, Dr. Maria Kallergi, and Dr. Pritish Mukherjee for their encouragement and serving on my di ssertation committee. I would like to thank Dr. Robert Potter for chairing my defense ex amination. I would like to thank Dr. Debra Huffman, Dr. Akihisa Nonoyama, Dr. Vineet Sh astry, Dr. Andres Cardenas, Dr. Michelle Cardenas, Dr. Christina Bacon, and Dr. J udy Fu, Tony Greco, Edward Zurek, Angela Brooke, Greta Klungness, Christie Stephans, an d Tracy Berg for their help throughout the research and writing dissertation. I woul d like to thank Sue wolf, Evelyn Keeton Williams, Phil Bergeron, and the staffs of the machine shop in the college of Marine Sciences for their support. I would like to th ank Berm-Jae Choi, Dr. David Austell, Mary Ellen Reilly, Dr. Hui Lee, Dr. Chul-Hwan Yoon for their friendship. I would like to express my thanks to my nephew Seong-Min Ho ng, sister-in-law SoonMi Lee, my wife Kyung-Hee Lee, and my son Sun-Joong Kim for their love and patience. Lastly, I would like to express my greatest th ank to God who always helps and guides me and my family with his amazing grace.

PAGE 5

i TABLE OF CONTENTS LIST OF TABLES v LIST OF FIGURES vi ABSTRACT xix CHAPTER 1. INTRODUCTION 1 CHAPTER 2. LITERATURE REVIEW 8 2.1. Particle Characterizatio n by Light Scattering 8 2.1.1. Particle Size or Size Dist ribution Measurement by Light Scattering 9 2.1.2. Particle Shape Measurement by Light Scattering 10 2.1.3. Particle Characterization by Polarized Light Scat tering 12 2.2. Particle Characterization by Multiwavelength Transmission Spectroscopy 13 2.3. Particle Characterization by Fluorescence Spectroscopy 15 CHAPTER 3. THEORY 16 3.1. Optical Constants 16 3.1.1. Optical Constants and Dispersion 17 3.1.2. Estimation of the Refractive Index 20 3.2. Polarization 21 3.2.1. Introduction to Polarization 22 3.2.2. Stokes Vector and Mueller Matrix 24 3.3. Light Scattering Theory 28 3.3.1. Introduction to Light Scattering Theory 29 3.3.2. Rayleigh Scattering Theory 35 3.3.3. Mie Theory 38 3.3.4. The Fraunhofer Diffraction 42 3.4. Transmission 43 3.4.1 Transmission and Absorbance 44 3.4.2 Turbidity 46 3.4.3 Beer-Lambert Law 48 3.5. Fluorescence Spectroscopy 49 3.5.1. Characteristics of Fluorescence Spectrum 50

PAGE 6

ii 3.5.2. Fluorescence Lifetimes and Quantum Yields 51 3.5.3. Fluorescence Quenching 53 3.5.4. Fluorescence Anisotropy 53 CHAPTER 4. DEVELOPMENT 55 4.1. Development Background 55 4.1.1. Description of the Prototype MAMW Spectrometer 56 4.1.2. Limitations of the Prototype MAMW Spectrometer 58 4.1.3. Development of the Integrated UV-VIS MAMW Spectrophotometer 61 4.2. Instruments 69 4.2.1. Light Source 69 4.2.2. Optics 73 4.2.3. Slits 75 4.2.4. Sample Cells 77 4.2.5. Goniometer 79 4.2.6. Spectrometer 79 4.3. Set up 84 4.3.1 Design Criteria for the Integrated UV-VIS MAMW Spectrophotometer 84 4.3.2 Description of the Inte grated UV-VIS MAMW Spectrophotometer Set up 83 4.4. Reproducibility 93 4.4.1. The Angular Adjustment of Goniometer 93 4.4.2. Optics Alignment 94 4.4.3. UV-VIS Beam SpectrumReproducibility Measur ements 93 4.5. Angle of Acceptance 98 4.5.1. Definition of the Angle of Acceptance 98 4.5.2. Effect of the Angle of Acceptance 100 CHAPTER 5. MEASUREMENT 109 5.1 Sample 109 5.1.1 Polystyrene Standards 110 5.1.2 Sample Preparation 113 5.2. Measurement 116 5.2.1. Signal Analysis 116 5.2.2. Measurement Procedures 118 5.3. Correction Factors 122 5.3.1.Reflection Correction 123 5.3.2. Refraction Correction 125 5.3.3. Scattering Path Length Correction 130 5.3.4. Scattering Volume Correction 135 5.3.5. Effect of Slit Width and Location 146 5.3.6. Data Normalization 161

PAGE 7

iii 5.4. Data Processing 171 5.4.1. Optical Density Data Processing 171 5.4.2. Scattering Data Processing 172 5.4.3. Fluorescence Data Processing 173 5.4.4. Remarks on Measured Results Plotting Using MATLAB 173 5.5 Calibration 176 5.5.1. Absolute Calibration 178 5.5.2. Relative Calibration 179 5.5.3. Calibration Constant 180 CHAPTER 6. RESULTS AND DISCUSSION 184 6.1. Measured Optical Density of Polystyrene Spheres 185 6.2. Results of UV-VIS MAMW Spectra Measurement 193 6.2.1. Characterization of Polystyrene Spheres by Size 194 6.2.2. M easured UV-VIS MAMW Spectr a of Non-spherical Polystyrene Standards 218 6.2.2.1 Non-spherical Particles 218 6.2.2.2 Light Scattering by Ensembles of Shaped Particles 219 6.2.2.3 Results of the UV-VIS MAMW Spectra Measurements of 1.9 m Size Spherical particles and Peanut-Shaped Particles 223 6.2.2.4. Proposed Set Up of an Integrated UV-VIS MAMW Spectrometer with Enhanced Backscattering Measurement Capacity 233 6.2.3. Characterization of Polystyrene Standards by Composition 236 6.3. Results of Frequency Domain Fluorescence Emission Spectrum Measurement 242 6.4. Results of the UVVIS MAMW Spectra Measurement of Whole Blood Samples 252 6.4.1.Sample Preparation and Measurement 252 6.4.2. Measured UV-VIS MAMW Spectra of Normal Whole Blood Sample 255 6.4.3. Measured UV-VIS MAMW Spectra of Sickle Cells 259 6.5. Discussion 265 6.5.1. The Upgraded Integrated UV-VIS MAMW Spectrometer 266 6.5.2. Development of the Multidimensional MAMW Spectrometer 268 CHAPTER 7. CONCLUSIONS 272 7.1. Summary of Research 272 7.2. Contributions 275 7.3. Recommendations for the Future Works 277 REFERENCES 281

PAGE 8

iv APPENDICES 287 Appendix A: MATLAB Program fo r UV-VIS MAMW Spectra Plot 288 Appendix B: Method of Correcti ng the Measured UV-VIS MAMW Spectra for the Continuously Varying Refractive Indices of Spectrosil Quartz 291 ABOUT THE AUTHOR End Page

PAGE 9

v LIST OF TABLES Table 4.1. Specification of lenses used in the Integrated UV-VIS MAMW Spectrometer 74 Table 5.1. List of polystyrene standards used for this dissertation research 111 Table 5.2. Differences in observation a ngle and refractive-indexcorrected actual scattering angle. 129 Table 5.3. Calculated scatteri ng volume in case of the 3 mm width slit installation for the Integrated UV-VIS MAMW spectrometer setup 141 Table 5.4.Calculated scattering volume in case of the 6 mm width slit installation for the Integrated UV-VIS MAMW spectrometer setup 141 Table 5.5. Slit width and distance used for the UV-VIS MAMW spectra Measurement 157 Table B.1 The refractive indices of fused silica 292 Table B.2 The refractive indices of Spectrosil quartz 292 Table B.3.Differences in observation a ngle and the refractive-index-corrected actual scattering angle. Th e manufacturerprovided refractive index of 1.551 at 200 nm was used for the correction. The results are compared with the data shown in Table 5.2. 297 Table B.3.Differences in observation a ngle and the refractive-index-corrected actual scattering angle. The calculated refracti ve index of 1.4532 at 800 nm was used for the correcti on. The results are compared with the data shown in Table 5.2. 297

PAGE 10

vi Table B.5. Calculated scattering volume fo r 3 mm width slit. The manufacturerprovided refractive index of 1.551 at 200 nm was used. The results are compared with the data shown in Table 5.3. 298 Table B.6. Calculated scattering volume for 6 mm width slit. The manufacturerprovided refractive index of 1.551 at 200 nm was used. The results are compared with the data shown in Table 5.4. 298 Table B.7. Calculated scattering volume for 3 mm width slit. The manufacturerprovided refrac tive index of 1.4532 at 800 nm was used. The results are compared with the data shown in Table 5.3. 299 Table B.8. Calculated scattering volume for 6 mm width slit. The manufacturerprovided refractive index of 1.4532 at 800 nm was used. The results are compared with the data shown in Table 5.4. 299

PAGE 11

vii LIST OF FIGURES Figure 1.1. Particles with different joint particle property di stribution (JPPD) 2 Figure 3.1. Optical property of a dielectric medium (a) th e dispersion relation and (b) the corresponding reflectance at normal incidence. 19 Figure 3.2 Wavelength depende nt variation of the complex refractive index of a dielectric material. 19 Figure 3.3 Electric field of a linearly polarized wave. 23 Figure 3.4 The polar plots of the angular functions n n and 40 Figure 3.5 Energy level diagram illustra ting absorption and emission of light by molecules. 51 Figure 3.6 Depopulation of the excited states by ra diative and nonradiative decay. 52 Figure.4.1. Schematic of the prototype MAMW spectrometer. 57 Figure 4.2. MAMW response surf aces of polystyrene 1 m standard in water. a) MAMW response surface of polystyrene 1 m standard in water measured by the prototype MAMW spectrometer. b) Simulated MAMW response surface of polystyrene 1 m standard in water. 59 Figure 4.3. Contour plots of MAMW re sponse surfaces of polystyrene 1 m standard in wate r. a) Contour plot of MAMW response surface of polystyrene 1 m standard in water measured by the prototype MAMW spectrometer. b) Contour plot of the simulated MAMW response surface of polystyrene 1 m standard in water. 60 Figure 4.4 Simulated Rayleigh ratios for particles with different sizes. 62 Figure 4.5. UV-VIS beam spectrum measur ed with the prototype MAMW spectrometer. 64

PAGE 12

viii Figure 4.6. UV-VIS beam spectrum meas ured with the UV-VIS MAMW spectrometer. 66 Figure 4.7. UV-VIS beam spectra recorded with the Integrated UV-VIS MAMW spectro meter with different integration times. 67 Figure 4.8. Photographs of Ocean Optics DT-1000 light source 70 Figure 4.9. Illustration of UV-VIS beam co mbination inside the DT-1000 light source 71 Figure. 4.10. Schematic of the slit with adjustable width. 76 Figure. 4.11. Schematic of the optical la yout of S-optical bench that is used for all the S2000 series spectrometer. 81 Figure 4.12. Grating efficiency curves for th e S-optical bench. 82 Figure 4.13. Photograph of the In tegrated UV-VIS MAMW spectrometer. 87 Figure 4.14. Schematic of the Integrated UV-VIS MAMW spectrometer. 88 Figure 4.15. Photograph of the shield ed Integrated UV-VIS MAMW spectrometer. 90 Figure 4.16. UV-VIS MAMW response surface of polystyrene 1 m standard in water. 92 Figure 4.17. Schematics of goniometer turntable a ngular position markers. 95 Figure 4.18. UV-VIS beam spectra varia tion during 10 hours continuous usage after the intensity stab ilization. 97 Figure 4.19. Schematic defining the a ngle of acceptance in case of the lens-pinhole system. 99 Figure 4.20. Optical density spectra of purified red blood cell measured by HP 8453 spectrometer and Perk in Elmer Lamda 900 spectrometer. 101 Figure 4.21. Schematic of the experimental set up measuring the effect of angle of acceptance changes on meas ured optical density spectra of particles with different sizes. 103

PAGE 13

ix Figure. 4.22. The optical density spectra of polystyrene 300 nm spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from sample cell using the experime ntal set up shown in the Fig.4.21. 104 Figure. 4.23. The optical density spectra of polystyrene 3 m spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from sample cell using the experi mental set up shown in the Fig.4.21. 105 Figure. 4.24. The optical density spectra of polystyrene 10 m spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from sample cell using the experimental set up shown in the Fig.4.21. 106 Figure 5.1. The optical properties of polystyre ne. a) Real part of refractive index. b) Imaginary part of refractive index. 112 Figure. 5.2. Measured optical dens ity variation of polystyrene 0.3 m spheres in water. a) Optical density variation depending on sample concentration. b) Optical de nsity variation depending on beam path length. 115 Figure 5.3. Reference solution signal intensities measured from 20o to 40o with 5o resolutions in observation angle. 119 Figure 5.4. Scattering geometry explaini ng the reflected light contributions in case of co nventional cylindrical cell. 126 Figure 5.5. Diagram explaining the geometry of scattered light refraction. 127 Figure 5.6. UV-VIS MAMW response surfaces of polystyrene 1.0 m standard in water. a) Plotted for the observation angle of 5o 40o. b) Plotted fo r the refraction corrected scattering angle of 4o 28o. 131 Figure 5.7. Contour plots of UV-VIS MAMW response surfaces of polystyrene 1.0 m standard in water, Fig.5.6. a) Plotted for the observation angle of 5o 40o. b) Plotted for the refracti on corrected scattering angle of 4o 28o. 132 Figure 5.8. Simulated plots of UV-VIS MAMW spectra of polystyrene 1.0 m standard in water. a) Si mulated plot of UV-VIS MAMW response surface for angular range of 4o 29o. b) Corresponding contour plot. 133 Figure 5.9. Diagram illustrating the necessity of scattered light path length correction in case of rectangular cuvette. 136

PAGE 14

x Figure 5.10. Schematic explaining scattering volume changes. 137 Figure 5.11. Schematic of geometry used for s cattering volume correction. 139 Figure 5.12. Scattering volume corrected UV-VIS MAMW response surface of polystyrene 1 m standard in water. a) UV-VIS MAMW response surface without volume correction. b) UV-VIS MAMW response surface when sine co rrection is done 142 Figure 5.12. (Continued). c) UV-VIS MAMW response surface when scatteri ng volume correction is done. d) Simulated UV-VIS MAMW response surface. 143 Figure 5.13. Scattering volume corrected UV-VIS MAMW response surface of polystyrene 8 m standard in water. a) UV-VIS MAMW response surface without volume correction. b) UV-VIS MAMW response surface when sine correction is done. 144 Figure 5.13. (Continued). c) UV-VIS MAMW response surface when scattering volume corr ection is done. d) Simulated UV-VIS MAMW response surface. 145 Figure 5.14. Effect of scattering vol ume correction on UV-VIS MAMW response surfaces as a function of scattering angle for a constant wave length at 633 nm. a) In case of polystyrene 1.0 m standard in water. Sine correction yields decreased scattered light intensity at low angles. 147 Figure 5.14. Continued. b) In case of polystyrene 8.0 m standard in water. Both calcu lated volume correction and Sine correction yields decreased scattered light intensity at low angles. 148 Figure 5.15. Diagram explaining the source of background UV light a) Without cuvette, no b ackground UV light enters the re ceiving optics. b) With cuvette, refracted background UV light enters the r eceiving optics. 151 Figure 5.16. UV-VIS MAMW scattering profil es of polystyrene 150 nm spheres at observation angles from 5o to 20o. 153 Figure 5.17. Measured UV-VIS MAMW spect ra of polystyrene 300 nm spheres in water. a) Measured with background UV light. b) Measured after eliminating background UV light. 154

PAGE 15

xi Figure 5.18. Modified slit structure for preventi on of background UV light. a) Before modification. B) After modification Double block and side block are added. 155 Figure 5.19. Effect of slit width variat ion on measured UV-VIS MAMW response surfaces of polystyrene 4.0 m standards in water. a) 3 mm width slit was used for measurement. b) 6 mm width slit was used for measurement. 159 Figure 5.20. Effect of distance variation be tween slit and sample cell on measured UV-VIS MAMW response surfaces of polystyrene 1.87 m spheres in water. a) Slit is1 cm away from cuvette. b) Slit is about 5 cm away from cuvette. 160 Figure 5.21. Incident UV-VIS beam profile used for normalization and resulting UV-VIS MAMW response surface. a) Incide nt UV-VIS beam profile measured at 0o. b) UV-VIS MAMW response surface of polystyrene 3 m standard in water obtained by normalizing measured results using the incident UV-VIS beam profile. 163 Figure 5.22. Reference UV-VIS beam profile used for normalization and resulting UV-VIS MA MW response surface. a) Reference UV-VIS beam Profile measured at 0o. b) UV-VIS MAMW response surface of Polystyrene 3 m standard in water obtained by normalizing measured results using the reference UVVIS beam profile. 164 Figure 5.23. Incident UV-VIS beam profile measured at 3o used for normalization and resultin g UV-VIS MAMW response surf ace. a) Incident UV-VIS beam profile measured at 3o. b) UV-VIS MAMW response surface of polystyrene 3 m standard in water obtaine d by normalizing measured results usin g the incident UV-VIS beam profile measured at 3o. 165 Figure 5.24. Reference UV-VIS beam profile measured at 5o used for normalization and resultin g UV-VIS MAMW response surf ace. a) Reference UV-VIS beam profile measured at 5o. b) UV-VIS MAMW response surface of polystyrene 3 m standard in water obtaine d by normalizing measured results usin g the reference UV-VIS beam profile measured at 5o. 166 Figure 5.25. Reference UV-VIS beam profile measured at 15o used for normalization and resulting UV-VIS MAMW response surfac e. a) Reference UV-VIS beam profile measured at 15o. b) UV-VIS MAMW response surface of polystyrene 3 m standard in water obtained by normalizing measured results using the reference UV-VIS beam profile measured at 15o. 167

PAGE 16

xii Figure 5.26. Incident UV-VIS beam profile r ecorded with different optics setup. Lens L4 was re moved and 3 mm width slits we re used. 169 Figure 5.27. Measured UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water obtained using UV-VIS beam profile in Fig. 5.26 for normalization. 170 Figure 5.28. Measured UV-VIS MAMW response surfaces of polystyrene 3 m spheres in wa ter obtained using UV-VIS beam profile in Fig. 5.26 for normalization. 170 Figure 5.29. Optical density spectra of polystyrene 1 m standard in water measured by the Integrated UV-VIS MAMW spectrometer (red) and HP8453 spect rometer (blue). Note the wavelength shift between two spectra. 175 Figure 5.30 Contour plot of whole bl ood UV-VIS MAMW response surface before wave length shift correction. MATLA B generated wavelength interval was used for plot. 177 Figure 5.31 Contour plot of whole blood UVVIS MAMW response surfaces after wavelength shift co rrection. OOIBase32 software generated wavelength interval was used for plot. 177 Figure 6.1 Normalized optical density sp ectra of polystyrene 20 nm spheres in water. 186 Figure 6.2 Normalized optical density sp ectra of polystyrene 500 nm spheres in water. 188 Figure 6.3 Normalized optical dens ity spectra of polystyrene 1 m spheres in water. 189 Figure 6.4 Normalized optical dens ity spectra of polystyrene 4 m spheres in water. 191 Figure 6.5 Normalized optical dens ity spectra of polystyrene 10 m spheres in water. 192 Figure 6.6. UV-VIS MAMW response surfaces of polystyrene 20 nm spheres in water. a) Measured response surface. b) Simulated response surface. 195

PAGE 17

xiii Figure 6.7. Contour plot of UV-VIS MAMW response surf aces of polystyrene 20 nm spheres in wa ter. a) Contour plot of measured response surface. b) Contour plot of simulated response surface. 196 Figure 6.8. wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 20 nm spheres in water. a). Wavelength view plot of measured response surface. b) Wavelength view plot of simulated response surface. 197 Figure 6.9 UV-VIS MAMW response surface s of polystyrene 500 nm spheres In water. a) Measured response surface. b) Simulated response surface. 200 Figure 6.10. Contour plot of UV-VIS MAMW response surf aces of polystyrene 500 nm spheres in water. a) Contour plot of measured response surface. b) Contour plot of simulated response surface. 201 Figure 6.11. Wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 500 nm spheres in water. a). Wavelength view plot of measured res ponse surface. b) Wavelength view plot of simulated response surface. 202 Figure 6.12 UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water. a) Measured response su rface. b) Simulated response surface. 203 Figure 6.13. Contour plot of UV-VIS MAMW response su rfaces of polystyrene 1 m spheres in wa ter. a) Contour plot of m easured response surface. b) Contour plot of simulated response surface. 204 Figure 6.14. Wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water. a). Wavelength view plot of measured res ponse surface. b) Wavelength view plot of simulated response surface. 205 Figure 6.15 UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water. a) Measured response surface. b) Simulated response surface. 207 Figure 6.16. Contour plot of UV-VIS MAMW response surf aces of polystyrene 4 m spheres in wa ter. a) Contour plot of m easured response surface. b) Contour plot of simulated response surface. 208

PAGE 18

xiv Figure 6.17. Wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water. a). Wavelength view plot of measured res ponse surface. b) Wavelength view plot of simulated response surface. 209 Figure 6.18. UV-VIS MAMW response surfaces of polystyrene 8 m spheres in water. a) Measured response surface. b) Simulated response surface. 211 Figure 6.19. Contour plot of UV-VIS MAMW response su rfaces of polystyrene 8 m spheres in water. a) Contour plot of measured response surface. b) Contour pl ot of simulated response surface. 212 Figure 6.20. Wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 8 m spheres in water. a). Wavelength view plot of measured res ponse surface. b) Wavelength view plot of simulated response surface. 213 Figure 6.21. UV-VIS MAMW response surfaces of polystyrene 10 m spheres in water. a) Measured response surface. b) Simulated response surface. 214 Figure 6.22. Contour plot of UV-VIS MAMW response surf aces of polystyrene 10 m spheres in wa ter. a) Contour plot of m easured response surface. b) Contour plot of simulated response surface. 215 Figure 6.23. Wavelength view plots of UV-VIS MAMW response surfaces of polystyrene 10 m spheres in water. a). Wavelength view plot of measured res ponse surface. b) Wavelength view plot of simulated response surface. 216 Figure. 6.24. Scanning electron microscope (SEM) images of polystyrene 1.9 m size standards. a) Spheri cal particles. b) Peanut shape particles. 220 Figure. 6.25. SEM images of two peanut shape particles placed in different orientations. a) Particle is oriented cl ose to perpendicular to the incident lig ht. b) Particle is oriented more like parallel to the incident light. 221 Figure 6.26. Microscope picture of concentrated peanut shape pa rticle solution. 226

PAGE 19

xv Figure 6.27. Measured UV-VIS MAMW response surfaces of polystyrene 1.9 m size standa rds in water. a) In case of polystyrene spheres. b) In case of peanut shape standards. Wavelengths extend from 230 nm to 800 nm. 224 Figure 6.28. Contour plots of measured UV-VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Contour plot in case of polystyrene spheres. b) Contour plot in case of p eanut shape standards. 225 Figure 6.29. Wavelength view plots of measured UV-VIS MAMW response surfaces of polys tyrene 1.9 m size standards in water. a) Wavelength view plot in case of polystyrene spheres. b) Wavelength view plot in case of peanut shape standards. 226 Figure 6.30. Measured UV MAMW response surfaces of polystyrene 1.9 m size standards in wate r. a) In case of polystyrene spheres. b) In case of peanut shape sta ndards. Wavelengths extend from 230 nm to 450 nm. 227 Figure 6.31. Contour plots of measured UV MAMW response surfaces of polystyrene 1.9 m si ze standards in water. a) C ontour plot in case of polystyrene spheres. b) Contour plot in case of pea nut shape standards. 228 Figure 6.32. Wavelength view plots of measured UV MAMW response surfaces of polystyrene 1.9 m size standards in wa ter. a) Wavelength view plot in case of pol ystyrene spheres. b) Wavelengt h view plot in case of peanut shape standards. 229 Figure 6.33. Measured VIS MAMW response surfaces of polystyrene 1.9 m size standards in wate r. a) In case of polystyrene spheres. b) In case of peanut shape sta ndards. Wavelengths extend from 450 nm to 800 nm. 230 Figure 6.34. Contour plots of measured VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Contour plot in case of polystyrene spheres. b) Contour plot in case of p eanut shape standards. 231 Figure 6.35. Wavelength view plots of measur ed VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Wavelength view plot in case of poly styrene spheres. b) Wavelengt h view plot in case of peanut shape standards. 232 Figure. 6.36. Schematic of the backscat tering measurement capacity enhanced Integrated UV-VIS MAMW spectrometer. 235

PAGE 20

xvi Figure. 6.37. Measured optical density spectra of polystyren 3.0 m, green fluorescen t polystyrene 3.0 m, and red dyed polystyrene 3.0 m spheres in water Red dyed polystyrene 3.0 m spheres has different profile be cause dye is surface coate d. HP 8452 spectrometer was used for this measurement. 237 Figure 6.38. Measured UV-VIS MAMW response surfaces of polystyrene 3.0 m size standards in water. a) Polystyrene 3.0 m spheres. b) Green fluorescent pol ystyrene 3.0 m spheres. 239 Figure 6.39. Contour plots of the measured UV-VIS MAMW response surfaces of polystyrene 3.0 m size standards in water. a) Contour plot of polystyrene 3.0 m spheres. b) Contour plot of green fluorescent polystyrene 3.0 m spheres. 240 Figure 6.40. Wavelength view plots of the measured UV-VIS MAMW response surfaces of polystyrene 3.0 m size standards in water. a) Wavelength view plot of pol ystyrene 3.0 m spheres. b) Wavelength view plot of green fluorescent polystyrene 3.0 m spheres. 241 Figure 6.41. Spectral information of D uke Scientific Corporation green fluorescent michrospheres. 244 Figure 6.42. Green fluorescent dye excita tion. a) UV excitation. b) Visible excitation. 245 Figure 6.43. Photograph of fluorescence m easurement set up. 246 Figure 6.44. Schematic of fluorescent measurement set up. 247 Figure 6.45. Fluorescence spectra measured by the Integrated UV-VIS MAMW spectrometer. 249 Figure 6.46. Fluorescence spectra (green) and scattered light from the particle of the same size (black) measured by the Integrated UV-VIS MAMW spectrometer. 249 Figure 6.47. Filtered absorption, fluorescence and scattered light. a) Measured at 90o. b) Measured at 20o. 250 Figure 6.48. Picture of red blood cells and si ckled blood cell. a) red blood cell has biconcave disc shape of ~ 8 m. b) The crescent shaped blood cell. 253

PAGE 21

xvii Figure 6.49. Normalized intensity profile of saline and SMBS dissolved saline solution. a) Pl otted for 190 nm to 820 nm range. b) Expanded plot from 500nm to 600 nm wher e hemoglobin doublets are located. 256 Figure 6.50. Measured MAMW response su rfaces of blood sample. a) MAMW response surf ace of pure red blood cells in saline measured by the prototype MAMW spectrometer1. b) UV-VIS MAMW response surface of normal whole blood in saline measured on the day of blood sample extraction. 257 Figure 6.51. Contour plots of measured MAMW response surfaces of blood sample. a) Contour plot of the pure re d blood cells response surface Fig 6.50.a. b) Contour plot of the normal whole blood in saline response surface Fig 6.50.b. 258 Figure. 6.52. UV-VIS MAMW response surface of normal and sickled blood samples. a) UV-VIS MAMW response surface of normal whole blood in saline measured on the second day of blood sample extraction. b) UV-VIS MAMW response surface of sickled blood in SMBS dissolved saline taken on the second day of blood sample extraction. 260 Figure 6.53. Contour plots of UV-VIS MAMW response surface of blood sample. a) Contour plot of normal whole bl ood sample response surface Fig 6.52.a. b) Contour plot of the sickled blood sample response surface Fig 6.52.b. 261 Figure 6.54. Wavelength view plots of UV-VIS MAMW response surface of blood sample a) Wavelength view plot of normal whole blood sample response surface Fig 6.52.a. b) Wavelength view plot of the sickled blood sample response surface Fig 6.52.b. 262 Figure.6.55. Measured optical dens ity spectra of oxyhemoglobin, deoxyhemoglobin and methemoglobin. 264 Figure 6.56. Schematic of a supposed MD MA MW spectrometer. 270 Figure B.1. Refractive indices of fused silica and Spectrosil quartz. 296

PAGE 22

xviii Integrated UV-VIS Multiangle-Multiwavelength Spectrometer for Characterization of Micron and Submicron Size Particles Yong-Rae Kim ABSTRACT Characterization of micron and submicron size particles requires the simultaneous measurement of the joint particle property distribution (JPPD). The JPPD is comprised of particle size, shape, orientat ion, composition, optical properties, and surface properties. Measurement of each of the par ticle properties independently is a difficult task and it has been only partially successful. To determine as many pa rticle properties as possible using optical methods it is necessary to simultaneously measure all aspects of the interaction of the incident light with the particles of interest. This approach leads to the concept of multidimensi onal spectroscopy suggested by Prof. Garcia-Rubio. Dr. Bacon proved the proposition by developing and testing a prototype multianglemultiwavelength (MAMW) spectrometer propos ed by Prof. Garcia-Rubio. However, the prototype MAMW spectrometer has limitations in the amount of information it can obtain because of strong absorption of deep UV light and detector saturation due to the use of optical fibers and single in tegration time for the CCD detector. The Integrated UV-VIS MAMW spectrometer has been developed to overcome the limitations of the prototype MAMW spectrometer. Improvements have become

PAGE 23

xix possible through the use of UV lenses and integration time multiplexing (ITM). The Integrated UV-VIS MAMW spectrometer ha s the capabilities to perform low angle scattering measurements starting from 4o with simultaneous detection of multiwavelength light from 200 nm to 820 nm UV-VIS transmission spectroscopy, and frequency domain fluorescence spectrosc opy. Following the development, possible sources of errors were analyzed and data ca libration procedures have been established to ensure the validity a nd reproducibility of the measurement results. The capabilities of the Integrated UVVIS MAMW spectrometer were tested by measuring UV-VIS MAMW spectra of polysty rene standards. The measured UV-VIS MAMW spectra clearly show differences due to particle size, shape, and compositional changes. Measurements of the UV-VIS MAMW spectra of sickled whole blood samples demonstrate that particle shape and compositional changes can be detected simultaneously. These results confirme d that the Integrated UV-VIS MAMW spectrometer could be a powerful tool for th e characterization of micron and sub-micron size particles. Alternate approaches to e nhance these capabilities further, i.e., the development of a new multidimensional MAMW spectrometer, are also described.

PAGE 24

1 CHAPTER 1. INTRODUCTION Micron and sub-micron size particle char acterization is important for industry, environmental monitoring, biology, and medi cine because of improved product quality, higher throughput and increased efficiency, reduced environmental pollution, compliance with regulatory requirements (FDA, EPA, OSHA), and as a new methodology for early detection of, especially, spreading diseases by non-invasive detection of microorganisms, etc.46, 56 Methods of particle characterization in clude light-based methods, sedimentation, chromatography, sieves, acoustic, and gas adsorption, etc.67 These methods exploit the joint particle property distri bution (JPPD) to characterize particles. The JPPD is comprised of particle size or size distribut ion, shape, orientation, composition, structure, optical properties, and surface propert ies (area, charge, porosity, etc.).2,10 The JPPD can serve as the spectroscopic finge rprint of particle. Fig. 1.1 shows some particles with different JPPD.56 However, most of the particle characterization methods measure only one property and lead to incons istencies in the results and errors in the estimation because micron and sub-micron size particle properties ar e interactive, interrelated and thus, they are dependent on each other.18 Moreover, as particle hom ogeneity changes not only in space but also in time, serial JPPD measuremen t using different procedures can lead to

PAGE 25

2 Figure 1.1. Particles with different joint particle property di stribution (JPPD). Adapted from Reference 56.

PAGE 26

3 the additional inaccuracy in the obtained re sults. To overcome th e limitation of single particle property measurement, Prof. Ga rcia-Rubio has suggest ed the concept of multiangle-multiwavelength (MAMW) spectrome ter that incorporated multiwavelength static light scattering an d transmission spectroscopy.2,3 By developing and testing a prototype MAMW spectrometer that employs a tungsten-halogen lamp as light source, optical fibers as light delive ry tool, and a miniature CCD diode array spectrometer for detector, Dr. Bacon had prove d that MAMW spectrometer would be a possible modality of detecting different JPPD (mainly size, shape, and composition) simultaneously.2 Measured JPPD of different po lystyrene and silica standards we re consistent with results obtained by theoretical simulation. On the other hand, in spite of these ea rly accomplishments, there has been a necessity to develop a new MAMW spectromete r. The use of optical fibers as light delivery tool and single integration time for the CCD detector limited the capabilities of the prototype MAMW spectrometer. It util ized multiwavelength light scattering only within the spectral range of 400 nm ~ 800 nm and angular range of 30o ~ 90o, respectively.2 In addition, incorporation of spectro scopic techniques other than scattering, like transmission, fluorescence, and polarizat ion measurements were not possible. Consequently, information available from meas urements or range of particles that could be characterized by the prototype MAMW spectrometer had limitations. Thus, the necessity of developing a new MAMW spectro meter that can overcome the limitations of the prototype and enhance the JPPD detecti on capability has draw n significant research interest. This gives motivation for this di ssertation research, Development of the

PAGE 27

4 Integrated UV-VIS MAMW spectrometer for characteri zation of micron and sub-micron size particles. The limitations of the prototype MAMW sp ectrometer are overcome by the use of fused silica UV lenses as light delivery to ol and integration time multiplexing (ITM), optimized use of the 2 x 10 8 dynamic range of Ocean Op tics S2000 CCD diode array spectrometer.43 The adoption of UV lenses enables the use of broadband UV light source, increases the intensity of the incident light to sample cell, and toge ther with the use of narrow width slits and ITM, enhances the low angle scatte ring capability. The incorporation of ITM avoids detector saturation due to CCD linearity. By overcoming the limitations of the prototype MAMW spect rometer, detection of fluorescence, transmission, and low angle scattered light b ecomes possible. Because it can perform not only scattering but transmission and fluores cence spectroscopy, it is an integrated spectrometer. The Integrated UV-VIS MAMW spectrometer has the following capabilities to perform: 1. low angle scattering starting from 4o with simultaneous detection of multiwavelength light from 200 nm to 820 nm, 2. UV-VIS transmission spectroscopy with wave lengths for analysis ranging from 200 nm to 820 nm and 3. frequency domain UV-VIS fluorescence spectroscopy. The capability 1 is the unique characte ristics of the Integrated UV-VIS MAMW spectrometer. As of Nov. 15, 2004, the INSPEC (physics reference search engine) search

PAGE 28

5 results show no particle characteriza tion reports using both multiangle and multiwavelength UV-VIS light scattering. In ad dition, it has room to adopt polarization optics to measure the Mueller matrix, di chroism in transmission spectroscopy and fluorescence anisotropy. Consequently, the capacity of the MAMW spectrometer has been enhanced considerably. The enhanced lo w angle scattering capab ility and the use of UV-VIS light source allow the characterization of polystyre ne spheres ranging from 20 nm to 10 m with further extension possibilities. Moreover, measured UV-VIS MAMW spectra clearly show differences due to partic le shape or compositional changes. Besides, UV-VIS MAMW spectra of sickled whole bloo d samples confirmed that particle shape and compositional changes can be measured simultaneously with the Integrated UV-VIS MAMW spectrometer. All the results proved th e possibility that the Integrated UV-VIS MAMW spectrometer could be a prospective mo dality that can characterize micron and sub-micron size particles by the simultaneous de tection of the JPPD, specifically, size, shape, and composition. The Integrated UVVIS MAMW spectrometer has potential to be applied to viral particle detecti on and nano-technology in addition to the characterization of microorganisms or micron and sub micron size industrial particle characterization. The capabilities of the Integrated UV-VI S MAMW spectrometer can be enhanced further. This can be accomplished by either upgrading the current prototype Integrated UV-VIS MAMW spectrometer or developing a new multidimensional MAMW spectrometer. The multidimensional MAMW spectrometer will perform multidimensional spectroscopy simultaneously and allow obtaining optimum information

PAGE 29

6 necessary to characterize pa rticles. Methods to implemen t upgrading of the Integrated UV-VIS MAMW spectrometer and devel oping a new multidimensional MAMW spectrometer are described in the la ter part of this dissertation. The main contributions of this di ssertation research are as follows. 1. Development of the Integrated UVVIS MAMW spectrometer that can perform multiwavelength st atic light scattering, transmission, and frequency domain fluorescence spectroscopy us ing single light source and single detector. 2. Development of data correction proc edures for the Integrated UV-VIS MAMW spectrometer. 3. Simultaneous JPPD (size, shape and composition) measurements of polystyrene and biological standards. 4. Suggestions of how to upgrade the In tegrated UV-VIS MAMW spectrometer and to develop a new multidimensional MAMW spectrometer. Dissertation Layout This dissertation has seven chapters. Th e introductory chapter is followed by: CHAPTER 2. LITERATURE REVIEW brief su mmary of references pertinent to particle characterization by th e detection of the JPPD using light based spectroscopy, CHAPTER 3. THEORY descri ption of static light sc attering, transmission, and frequency domain fluorescence spectroscopy theory that are indispensable to the

PAGE 30

7 development of the Integrated UV-VIS MAMW spectrometer and the interpretation of the experimental results, CHAPTER 4. DEVELOPMENT de scription of background information, specifications of the instrument ations, design criteria, and error analysis including reproducibility and the effect of the angle of acceptance, CHAPTER 5. MEASUREMENT description of experimental methods and resulting data analysis including correction factors and calibration proce dures, CHAPTER 6. RESULTS AND DISCUSSION description of measurement re sults, the prospects of the Integrated UVVIS MAMW spectrometer, and ways to upgrade the Integrated UV-VIS MAMW spectrometer and develop a new multidimensional MAMW spectrometer, and CHAPTER 7. SUMMARY summary of this dissertation re search activity, the results, and the lists of recommended work.

PAGE 31

8 CHAPTER 2. LITERATURE REVIEW The objective of this disser tation research is the development of the Integrated UV-VIS MAMW spectrometer for micron and s ub-micron size particle characterization by the simultaneous measurement of the JPPD. It has been achieved by the incorporation of multiwavelength UV-VIS static light sca ttering, UV-VIS transmission, and frequency domain fluorescence spectroscopy in one inst rument. Each of the above spectroscopic methodologies has its own strength in the de tection of the indivi dual particle property.7, 19, 32 Considering the depth and breadth of spectroscopy and the amount of relevant publications, whole literature review is virtually impossible and unpractical. Thus, very brief review of literature and previous research activities directly related to the topic are described in this section. 2.1. Particle Characterization by Light Scattering Light scattering has been widely used in particle characte rization because it provides nondestructive and in stantaneous way of measur ing particle properties.10 Particle size, size distribution, and simple geometrical shape information are primarily obtained by light scattering measurement.7,10 Moreover, nowadays general technical

PAGE 32

9 development and the improvement of computa tional capability lead to the investigation of other characteristics such as irregularity, structure, composition and refractive index.30 2.1.1. Particle Size or Size Distributio n Measurement by Light Scattering Angular light scattering pr ovides a way of obtaining particle size information because scattering pattern relies on both th e size parameter and the relative refractive index.31 As particle size increases, the patter n becomes more and more complicated and sufficiently unique so that it can be used as a measure of particle size. In many instances, if refractive index of particle is known, the determination of particle size only requires the angular location of the maxima an d minima in the sc attering pattern.31 If particles are no longer of nearly the sa me size, it is necessary to obtain the particle size distribution (PSD ) instead of extracting size in formation for characterization. In practice the PSD is widely used because ideal monodisperse systems of particles are very rare. The acquisition of the PSD can be done by determining a normalized distribution function p(a) that gives the fraction of particles with radii between a and a + a.31 Forward scattering is also an attracti ve method of particle sizing. Near the forward direction, scattering by large particles is dominated by diffraction that depends on particle area, regardless of its shape, a nd is independent of re fractive index. Because of these advantages of refractive index independence and insensitivity to shape, forward scattering has been actively used to size non-spherical pa rticles with unknown optical properties.7

PAGE 33

10 Light scattering has been an important t ool to investigate pa rticle size or size distribution for various samples of importance in many different app lication areas due to its rapid and noninvasive measurement. The cap acity of this technique will be enhanced further with the advances in av ailable theory and instrument.30 2.1.2. Particle Shape Measurement by Light Scattering Interests in the characterization of non spherical particles by light scattering have increased significantly in recent years due to a wide variety of application areas. Unlike the homogeneous spherical particle characteri zation problem, the characterization of non spherical particles becomes more complicated because it needs the particle orientation information in addition to the parameters required to describe shape.12 To avoid the complexity, a tremendous amount of theoretical calculations have been made using Mie theory even if typical scattering particles ar e not always perfect spheres. This has been justified by two obvious and important reason s; I) many of the qualitative and even quantitative characteristics of particle s cattering are satisfactor ily described by areaequivalent spheres and II) Mi e theory calculations have been easy to carry out even before the advent of current high speed computers.21 However, uncritical use of Mie theory leads to imperfect and in some cases to spurious and misdirecting results in a number of applications of practical interests. Mie theory provides exact solutions only to the case s of homogeneous sphere, layered sphere, homogeneous circular cylinder, layered circular cyli nder, and ellipsoidal cylinder.12 For the other cases, approximate methods, like Rayleigh, Rayleigh-Debye-

PAGE 34

11 Gans (RDG), anomalous diffraction theory (ADT), Fraunhofer diffraction, coupleddipole model and T-matrix, are available.7,30,60 For particles with sizes larger than the Rayleigh particles but close to the wavelength of the incident light, Rayl eigh-Debye-Gans (RDG) approximation is available. Scattering by arbitrarily shap ed particles can be treated using RDG approximation because this approxima tion incorporates the form factor.31 Fairly accurate estimates of the light scattering can be acquired from RDG theory for suspension of randomly oriented, asymmetric particle s of slightly irregular shape. Van de Hulst formulated the anomalous diffraction theory. It is based on the assumptions that the refractive index of particle is different from that of the environment very slightly and the particle size parameter is large.60 One of distinct advantages of the anomalous diffraction theory is that it is sufficiently genera l and applicable to bodies of different shapes directly.55 As a result, scatteri ng of large particles of arbitrary shape and orientation are investigat ed using this theory.30, 31 T-matrix is a method based on integral fo rm of the scattering problem caused by particles of arbitrary shape and large volume. It is actively being used at present because numerical results can be obtained by this method.7,38, 49 Consequently, T-matrix method or Watermans extended boundary condition method (EBCM) becomes one of the most robust and efficient methods to investig ate scattering by non-spherical particles.49 Barber and Latimer applied RDG, ADT and EBCM to the scattering of the ellipsoids of revolution and found that EBCM substantially yi elded reasonably accurate results in most cases studied.33

PAGE 35

12 Coupled dipole model regards an arbitrary particle as an array of N polarizable subunits to calculat e total scattering.14 In spite of its relative simplicity, coupled dipole model was not popular because of the vast co mputer storage and central processing unit (CPU) time requirement necessary for modeli ng particles with large size parameters. However, an increasing number of peopl e have begun to employ the coupled-dipole method recently due to the development of faster computers and more efficient programming techniques.14 Considerable amount of efforts have b een made to extend the realm of light scattering theory beyond Mie theory. As a re sult, a diversity of theories has been developed to explain scatteri ng by non-spherical particles. However, the acquisition of particle shape information by light scatte ring still remains as a big challenge. 2.1.3. Particle Characterization by Polarized Light Scattering Polarized light scattering seems to be suitable for the characterization of nonspherical particles because particle shape and/or orientation significantly changes the polarization status of light.36, 52 Therefore, polarized light scattering has been actively used in the characterization of particles with different orientation and shape because of its sensitivity. Smith et al. measured the cross correlation between the pol arization status of scattered laser speckle as a function of sca ttering angle for a range of spherical and nonspherical particle suspensions and concluded that this measur ement has the possibility of non-spherical particle dete ction and characterization.52

PAGE 36

13 Biological particles provide a wide variet y of geometrical shapes and internal structures that can be uniformly altered and regulated during light scattering experiments. Consequently, polarized light scattered from well-prepared biological particles such as bacteria, pollen, spores, and red and white blood cells in solution provides information particular to the biologica l condition of the particle.6 W.S. Bickel et al. investigated biological scatterers by measuring the Muelle r matrix and concluded that polarization effects in light scattering can yield a very sensitive tool to study the structural changes of cells during their life cycles.5 2.2. Particle Characterization by Multiw avelength Transmission Spectroscopy Multiwavelength transmission spectra could be used to extract the particle size of nearly spherical particles by matching the meas ured features with t hose calculated from Mie theory. It is valid for monodisperse a nd sufficiently dilute system of particles.15, 16, 19 The optical density of particle disper sion (combined absorption and scattering characteristics) provides information that, in principle, can be employed to estimate the particle size distribution (PSD) and to identify the chemical composition of the suspended particles. The particle com position affects its optical pr operties, i.e., the complex refractive index of the particles that will re sult in absorption and scattering of light.2 Absorption depends on the particle size in addition to the optical properties of the particles. Consequently, unique absorption spectral features of solutions for a very dilute system of particles can be used to id entify particles or to determine the PSD.2

PAGE 37

14 Multiwavelength transmission spectra can be obtained from the measured optical density spectra by eliminati ng scattering spectra using the mathematical deconvolution method.35 The recovery of PSD from the measur ed multiwavelength transmission spectra requires to solve the Fredholm integral of th e first kind that incl udes the particle size distribution, volume fraction, a nd scattering efficiencies Qsca.29 The scattering efficiencies can be calculated using classical scattering theories if wavelength, relative refractive index, and the particle diameter are given. Therefore, the acquisition of particle size distribution from the measured multiwavele ngth transmission spect ra relies on (1) the proper inverse algorithm for a solution to Fre dholm integral equation of the first kind and (2) the development of proper multiwavele ngth transmission spectra measurement methods.29 The inverse solution of the measured multiwavelength transmission spectra can be obtained by either analytical inversion64 or numerical inversion method.15,16 The analytical inversion method has the advantage of yielding a fast solu tion to the inverse problem considered. On the other hand, this method needs the prior knowledge of the turbidity over the whole spectrum of wave nu mbers that is not practical. Therefore, a procedure of stabilizing or smoothing is required.29 Using an analytical inversion algorithm, Wang and Hallett succee ded to recover the latex sphere size distribution.64 Numerical inversion method that employs nonlinear least-squares algorithms and regularization techniques does not require the prior information of turbidity and thus, provides accurate recovery of the PSD w ith a reasonably fast computational time.29 Elicabe and Garcia-Rubio successf ully recovered the PSD of polystyrene lattices with range of 50 ~ 3950 nm using the numerical inversion technique in their simulated

PAGE 38

15 studies.15,16 Multiwavelength transmi ssion spectroscopy using the numerical inversion technique has been applied to the quantita tive interpretation of the measured human blood platelet spectra.35 2.3. Particle Characterization by Fluorescence Spectroscopy Fluorescence has been well known for over 300 years and has proven to be a viable method in biology and chemistry due to the superb detection limits on material composition.9 However, fluorescence spectroscopy has not been actively used for particle characterization other than or ganic or biological sample identification. This can be attributed to the fact that most intrinsic fluorephores, ex cept rare earths, are organic aromatic molecules and fluorescence measur ement requires complicated instruments.32 On the other hand, application of fluorescen ce spectroscopy to medicine like biological particle identificati on and medical diagnostics has attr acted keen rese arch interests recently.20, 44 For instance, Gray et al. showed that UV fluorescent spectra could be used to identify bacteria like E. coli, S. aures, and S. typhimurium.20 Characterization of particles contained in human body fluids using fluorescence spectroscopy seems to be very realistic.

PAGE 39

16 CHAPTER 3. THEORY This chapter describes the basic aspects of static light scattering, transmission, and frequency domain fluorescence spectroscopy theory. Static light scattering theory treated in this chapter includes Rayleigh s cattering theory, Mie theory and Fraunhoffer Diffraction. These are indispensable for th e development of the Integrated UV-VIS MAMW spectrometer and for the interpretation of the experimental results. In describing transmission theory, priority is given to clarifying ambiguities existing in terms and definitions. Fundamental aspects of fre quency domain fluorescence spectroscopy are briefly described as an in troduction to fluorescence spectroscopy. In addition, optical constants that explain inherent optical properties of material as well as the fundamentals of polarization theory including th e Mueller matrix are introduced. 3.1. Optical Constants The optical properties of a particle ar e completely specified by frequencydependent optical constants namely, the complex refractive index and the complex dielectric constant.7 They are interdependent and either of them can be chosen depending on the application. The complex refractive inde x that is comprised of the real and the imaginary parts is a critical parameter in th e light scattering theory. The difference in the

PAGE 40

17 real part of the refractive index of the me dium and that of the particle causes the scattering of light. The imaginary part of the particles refractive i ndex, i.e. absorption coefficient, decides the amount of light at tenuation. Precise calculation of the complex refractive index is essential for the estimation of the absorption and scattering of light by the particle. The deriva tion of optical constants using Lo rentzs simple classical harmonic oscillator model, brief investigation of the optical properties of di electric material, and the description of the ways to evaluate th e complex refractive index constitute this section. 3.1.1. Optical Constants and Dispersion Optical constants describe the intrinsic op tical properties of materials. There are two sets of commonly used optical constant s, namely the complex refractive index, ik n N and the complex dielectric constant, i. They are not independent and related to each other by expressions7 nk k n 2 ,2 2 (3.1) 2 22 2 2 2 k n (3.2) The complex refractive index is preferred to describe wave propagation while the complex dielectric constant is favored to e xplain absorption and scattering by particles,

PAGE 41

18 especially, small compared with the wavele ngth. These optical cons tants can be derived by Lorentzs simple harmonic oscillator model that describes the classical aspects of optical properties of material.7 In this model, the electrons behave as though they are held by springs and are subjected to damping force and restoring force. The applied electromagnetic fields provide driving force to the electrons. The result is known as a dispersion equation45 j j j j e ei f m e N n 2 2 0 2 21 (3.3) where eN is the number of electrons, e is charge on an electron, me is the mass of an electron, 0 is the permittivity of free space, j is the resonance frequency of material, is the angular frequency of the radiation, jf is the oscillator st rength for the resonance frequency j and j is the damping constant of the electron (responsible for absorption). Fig. 3.1.a shows the frequency dependence of the real and imaginary parts of the refractive index of a dielectric medium and Fi g. 3.1.b exhibits that of the reflectance at normal incidence.7 The refractive index n increases as frequency increases (normal dispersion) except the region where is close or equal to the resonance frequency j (decreasing n anomalous dispersion). The regi on of high absorption around the resonance frequency causes an associated regi on of high reflectance. Thus, a material

PAGE 42

19 Figure 3.1. Optical property of a dielectric medium (a) the di spersion relation and (b) the corresponding reflectance at normal incidence. Adapted from Reference 7. Figure 3.2 Wavelength dependent va riation of the complex refract ive index of a dielectric material. Adapted from Reference 50. a b

PAGE 43

20 with 1 k is highly reflecting not highly absorbi ng and little light that gets into the material is rapidly attenuated.7 Fig.3.2 shows typical wavelength dependent profile of the complex refractive index of a dielectric material.50 There is a strong absorpti on in the ultraviolet region because the photon energy corresponds to ener gy differences between filled and empty electron energy levels. The continuity of both empty and filled levels results in a broad region of high absorption. As wave length increases, absorption index k becomes small and almost constant. Therefore, the material is transparent in the visible region. All visibly transparent materials have strong absorption in the UV region. The resonance absorption in the infrared region, can be attr ibuted to the absorption caused by lattice vibration.7 3.1.2. Estimation of the Refractive Index Eq. (3.3) is inconvenient to calculate th e refractive index because it requires the information of atomic properties. Hence, simple approximation methods have been developed. The Cauchy dispersion equation, de duced from Eq. (3.3), is an empirical approximation formula relating the real part of the refractive inde x to the wavelength It is given by ...4 2 2 1 0 d d d n (3.4) The number of parametersnd is determined by the requi red accuracy to fit the experimentally obtained data.45

PAGE 44

21 The complex refractive index can be de scribed as a harmonic complex number and thus, fulfills the Kramers-Kronig transform. As a result, the real part of the refractive index can be computed if the imaginary part of the refractive i ndex is known or vice versa, using the transforms expr essed by Eq. (3.5) and Eq. (3.6). d k P n0 2 2) ( 2 1 ) ( (3.5) d n P k0 2 2) ( 2 ) ( (3.6) where is the variable representing frequency in the integral, and P is the principal value of the integral.7 In addition to these approximation methods the optical constants often can be obtained from experiment. However, both the real and imaginary parts of the complex refractive index cannot be measured directl y. Therefore, they must be derived from measurable quantities using a proper theory.7 3.2. Polarization In analytical spectroscopy, the polarization stat us of light, either in cident or detected, provides valuable information that is indisp ensable for the study of the objectives under investigation. The Stokes vector and the Mueller-matrix approach for light scattering,

PAGE 45

22 fluorescence anisotropy, and dichroisms (linear and circular) in ab sorption spectroscopy are examples of experimental methods that pr oduce a wealth of info rmation by taking the advantages of polarization. Fundamentals of polarizati on theory are briefly revi ewed in this section. Derivation of the polarization status of light, the Stokes v ector approach to express polarization status as observable quantity, a nd the Mueller matrix concept required for complete characterization of the optical prop erties of material are explained in this section. This information is necessary for the description of light scattering and fluorescence theory that will be th e subject of following sections. 3.2.1. Introduction to Polarization The Maxwell equations in an infinite medium when there are no electric and magnetic sources are:27 0 1 0 t B c E E (3.7) 0 0 t E c B B In Eq. (3.7) cis the speed of light in vacuum, is permittivity and is permeability. The simplest solution of these equations is general homogeneous plane wave propagating in the direction of n k k

PAGE 46

23t i x ie ) E E ( ) t x ( E k 2 2 1 1 (3.8) where n and ,2 1 are a set of real mutually ort hogonal unit vectors, k is propagation number ( ) and E1 and E2 are amplitudes with complex numbers.27 The polarization status of ) ( t x E is determined by the magnitude and the phase of E1 and E2. In case of linea rly polarized wave, E1 and E2 have the same phase and the polarization vector making an angle p = tan-1( 1 2E E) with 1 and a magnitude E = 2 2 2 1E E as shown in Fig 3.3 Figure 3.3. Electric field of a linearly polar ized wave. Adapted from Reference 27.

PAGE 47

24 The Eq. (3.8) expresses ellip tically polarized wave if E1 and E2 have different phases. If E1 and E2 have the same magnitude, but differ in phase by 90o, the Eq. (3.8) represents circularly polari zed wave and becomes: t i x ie i E t x E k 2 1 0) ( ) ( (3.9) with the common real amplitude E0.27 Suppose that axes are select ed so that the wave is propagating in the positive z direction, while 1 and 2 are in the x and y directions, respectively. Then the wave in case of) (2 1 i is called left circularly polarized while the wave for) (2 1 i is described as righ t circularly polarized.27 3.2.2. Stokes Vector and the Mueller Matrix If a plane electromagnetic wave is expre ssed in the form of Eq. (3.8) with known coefficients (), ,2 1E Ethe information of the polarizati on status of the wave can be acquired without much difficulty. However, in pr actice, it is indispensable to decide the polarization status by observing the beam and thus, Eq. (3.8) is not a convenient form to use for this purpose. To avoid this complicat ion, each of the scalar coefficients in Eq. (3.8) can be defined as a ma gnitude times a phase factor: 2 12 2 1 1, i ie a E e a E (3.10)

PAGE 48

25 where 2 1 a a are the amplitudes and 2 1, are the phases.27 Let the subscripts 1 and 2 designate parallel and perpendi cular direction respectively. Th en electric fields can be rewritten as parallel and perpe ndicular to the plane of interest60 ,k k // ////t i z i i t i z i ie e a E e e a E (3.11) The Stokes parametersV U Q I , ,, which completely describe the intensity and the polarization status of a light wa ve using directly observable quantities, are related to the electric field vector as follows:60 sin 2 ) ( cos 2 ,// // // // // // 2 2 // * // // 2 2 // * // // a a E E E E i V a a E E E E U a a E E E E Q a a E E E E I (3.12) where is the phase difference // and represents the complex conjugate. The Stokes parameters have an obvious interpretation. For example, I is the total intensity, Q is parallel or perpendicular linear polarization to the scattering plane, U is linear polarization at o to the scattering plane, and V is left or right circular polarization.5

PAGE 49

26 The Stokes parameters for a quasi-monochromatic light are written as4 , ,* // // // // * // // * // // E E E E i V E E E E U E E E E Q E E E E I (3.13) where the angular brackets indicate time av erages over an interval longer than the periodf / 1 T where f is the optical frequency (~1015/s). There is a relationship among the Stokes parameters:4 2 2 2 2V U Q I (3.14) The equal sign holds for the totally polarized light while inequality represents the cases of partially polarized light and 0 V U Q defines the natural or unpolarized light. The Stokes vector, a column vector, the four elem ents of which are the Stokes parameters can represent a beam of arbitrary polarizati on including partially polarized light. When light interacts with optical compone nts such as lenses, filters, polarizers, surfaces, scattering media, etc ., the polarization status of light is changed in general.25 This interaction of light with any optical elements or mate rial can be described as a multiplication of the four-component Stokes ve ctor with a 4 x 4 matrix. This sixteen-

PAGE 50

27 element matrix is called the Mueller matrix or the scattering matrix if scattering is involved. The Mueller matrix S, S 44 43 42 41 34 33 32 31 24 23 22 21 14 13 12 11S S S S S S S S S S S S S S S S (3.15) completely characterizes the optical prope rties of any components or material.4 Once an incident four-component Stokes vector iV and a Mueller matrix ] [ S are given, a final Stokes vector fV is determined from the relation i fS V V ] [ or i i i i f f f fV U Q I S S S S S S S S S S S S S S S S V U Q I 44 43 42 41 34 33 32 31 24 23 22 21 14 13 12 11 (3.16) Eq. (3.16) can be applied to a series of optical elements placed in a beam.5 The final polarization status of the beam can be obt ained by multiplying the corresponding Mueller matrix of each element.4 If incident light is perfec tly polarized, Jones vector, another representation of polarized light that has the advantages of be ing applicable to cohe rent beams, provides

PAGE 51

28 extremely concise representation. The details of Jones Vector and corresponding Jones Matrix approach can be found in Reference 50. 3.3. Light Scattering Theory If an electromagnetic wave irradiates an object, the electric fi eld of the incident wave forces electric charges in the object to begin oscillatory moti on which results in the acceleration of electric charges. Scattering is the re-radiation of electromagnetic energy in all directions by these accelerated electric charges in the object.7 If part of the incident electromagnetic energy is transformed into other forms, like thermal energy, by the excited elementary charges, it is called ab sorption. Scattering a nd absorption cannot be considered separately in gene ral because the sum of the atte nuation of incident radiation by both scattering and absorption accounts for th e total extinction of the incident light.7 In scattering, the incident radiation frequenc y is different from the natural frequency of molecules in the object. The static light scattering theory for dilu ted particle system can be classified into rigorous theory and approximation.30 For rigorous theory, parameters can be calculated to any desired accuracy by boundary condition me thod and integration method. The former, solution is figured out by a pplying the electromagnetic bounda ry conditions and the latter, solution is found from an integral over the volume of the scatterer.30 Due to the complexity of the scattering pr oblem, rigorous solutions requi re considerable amount of computer time and storage even for the simplest shapes and smallest particles. Therefore, a wide range of approximation theories has be en developed. Each of them may be used

PAGE 52

29 for certain ranges of size and refractive inde x only. However, they are easy to calculate and often provide much insight into the scattering process.30 In this research, Mie theory is adopted as the primary interpretation model because the exact solution explaining the sca ttering of light by a sphere that has size comparable to the wavelength of inci dent light can be obtained using it.7 In addition, Mie theory can provide approximations for s pheres with a wide range of sizes or nonspherical particles with simple geometri cal shape. Besides Mie theory, Rayleigh approximation and diffraction theo ries are also described. 3.3.1. Introduction to Light Scattering Theory If an infinite plane wave illuminates a single particle, the total scattered power by the particle scaP is given i sca scaI C P (3.17) where iI is the incident light intensity and scaC is the scattering cross section.30 The total absorbed power absP is defined in the same way i abs absI C P (3.18) where absC is the absorption cross section.30 The cross section, a quantity with

PAGE 53

30 dimensions of area, is the energy taken away from the incident radiation. The scattering cross section scaC is the energy removed by scattering and the absorption cross section absC corresponds to the energy pulled off by absorption. The combined effect of absorption and scattering is extinction and accounts for the total energy eliminated from the incident radiation.7 The total power extinguished is i ext extI C P (3.19) The extinction cross section extC can be written as the sum of the absorption cross section absC and the scattering cross section scaC because of the conservation of energy:7 sca abs extC C C (3.20) The Efficiencies for extinction, scattering, a nd absorption are defined by normalizing the corresponding cross section with the particle size: G C Q G C Q G C Qabs abs sca sca ext ext (3.21) where G is the particle cross-sectional area pr ojected onto a plane perpendicular to the incident beam.7 The efficiencies for extinction, abso rption, and scattering are related to each other by Eq. (3.22)

PAGE 54

31 sca abs extQ Q Q (3.22) The efficiencies are merely dimensionless cross sections. The scattered electric field at sufficiently large distances from the particle has the form of an outgoing spherical wave7 A r i e Er i sca kk (3.23) where A is the real vector field independent from time but depend on position. scaE is transverse to the radial direction of propagation r Eq. (3.23) may be rewritten as ) ( k) ( k F E r i e Ez r i sca (3.24) ) ( F is the dimensionless vector scattering amplitude, is the angle between the incident light and the scattered light, is the azimuthal angle, and z is the direction of the incident radiation.7,30 The corresponding magnetic field is7 ) ( ksca scaE r H (3.25)

PAGE 55

32 The Poynting vector that specifies the electromagnetic energy flux is7 ) Re( 2 1*H E S (3.26) By substituting Eq. (3.24) and Eq. (3.25) into Eq. (3.26), we can fi nd the Poynting vector for the scattered field30 r F r I Si sca ) ( k2 2 2 (3.27) The integration of scaS over the surface of the spherical volume of the medium surrounding the scattering particle yi elds the scattering cross section30 4 2 2 2 00 2 2k ) ( sin k ) ( d F d d F I P Ci sca sca (3.28) The term 2 2 k ) ( F is named as the differential scat tering cross section or Rayleigh ratio and often designated as /d dCsca. Physically, the differential scattering cross section describes the angular distribution of the scattere d light: the amount of light scattered into a unit solid angle ( ) about a given direction.7 The phase function p

PAGE 56

33 which expresses probability that during a scat tering event a photon with initial direction is scattered in the direction is defined by normalizing (3.28)30 scaC F p2 2k ) ( 4 ) ( (3.29) and 2 001 sin ) ( 4 1 d d p (3.30) The extinction cross section is obt ained from the optical theorem7 } ) Re{( k 40 2 F r I P Ci ext ext (3.31) It is convenient to relate in cident and scattered electric fields through the matrix form7 i i // z) (r i sca sca //E E S S S S r i e E E1 4 3 2 kk (3.32) where//E is the electric field compone nt polarized parallel to and E is that polarized perpendicular to the scattering plane. Eq. (3 .32) defines the amplitude scattering matrix.

PAGE 57

34 Its elements ) , j ( Sj4 3 2 1 depend in general on the scattering angle and the azimuthal angle .7 The Muller matrix in Eq. (3.15) is called the scattering matrix if scattering is involved. Its 16 elements are given in terms of the amplitude scattering matrix elements and ,4 3 2 1S S S S For example, 11S, which defines the angular distribution of the scattere d light in case of unpolariz ed incident light, becomes7 2 4 2 3 2 2 2 1 112 1 S S S S S (3.33) Expressions for the other elements can be found in Reference 7. 11S is directly related to the ratio of scattered to incident intensity ) (i scaI I 2 2 11k r S I Ii sca (3.34) where scaI is the scattered light intensity.7 If particle shape is isotropic then 4 3and S S of the amplitude scattering matrix in Eq. (3.32) become nullified because no crosspolarization is introduced. For an infinite plane incident wave, the azimuthal angle at o0 can be ignored and it follows that ) 0 ( ) 0 ( ) 0 (2 1S S S Substituting this into Eq. (3.31) leads to

PAGE 58

35 the expression for the extinction cross section7 )} (0 { Re k 4o 2S Cext (3.35) 3.3.2. Rayleigh Scattering Theory Rayleigh scattering occurs when the size of the scattering particle is much smaller than the wavelength of the light being scattere d. In case of the Rayl eigh scattering, the expressions for the amplitude scattering matrix elements S1 and S2 are7 cos a 2 3 a 2 31 2 1 1 S S (3.36) where 1a is a scattering coefficient 2 1 3 2 a2 2 3 1 m m x i (3.37) The size parameter x and the relative refractive index m are defined as N N m a N a x1 1k k 2 k (3.38)

PAGE 59

36 where 1N and N are the complex refractive i ndices of particle and medium, respectively.7 The resulting scattering matrix becomes7 cos 0 0 0 0 cos 0 0 0 0 ) cos (1 2 1 1) (cos 2 1 0 0 1) (cos 2 1 ) cos (1 2 1 k 4 a 92 2 2 2 2 2 2 1r (3.39) The scattered light intensity scaI for unpolarized incident light with intensity iI is given7 i scaI m m r a N I ) cos 1 ( 2 1 82 2 2 2 2 4 6 4 (3.40) For a sphere, the scatteri ng cross section and the ab sorption cross section are approximately7 2 2 2 4 4 6 52 1 3 128 m m N a Csca, (3.41)

PAGE 60

37 ) 2 ( ) 1 ( Im 42 2 3m m a x Cabs Dividing Eq. (3.41) by ,2a we can find the scattering efficiency and the absorption efficiency7 2 2 2 42 1 3 8 m m x Qsca (3.42) 2 1 Im 42 2m m x Qabs For sufficiently small particles the absorpti on and the scattering efficiencies can be approximated7 41 1 Q Qsca abs (3.43) The extinction spectrum varies as / 1 if it is dominated by absorption and as 41/ if it is dominated by scattering. In general shorte r wavelengths are extinguished more than longer wavelengths and this accounts for blue sky or red sunset.

PAGE 61

383.3.3. Mie Theory Mie theory that is a thorough description of the interaction of an infinite plane wave with a dielectric sphere de scribes scattering if the particle size is similar to or larger than the wavelength of the incident light. To obt ain an analytical solution, the fields need to be expanded in a proper coordinate syst em part of which is made on the particle surface. By adopting spherical coordinate sy stem, Mie could find the exact solution for the scattering problem for a sphere.7 For a spherical particle, the two amplitude scattering matrix elements 2 1and S S are7 } ) (cos b ) cos ( a { 1) (n n 1 2n ) (n n n 1 n 1 Sn and (3.44) } ) (cos a ) cos ( b { 1) (n n 1 2n ) (n n n 1 n 2 Sn where the angular functions n n and which decides the dependence of the fields, are7 ) (cos P sin 1 ) (cos 1 n n ) (cos P d d ) (cos 1 n n (3.45)

PAGE 62

39 Here, 1 nP are the associated Lege ndre polynomials. Fig. 3.4 shows the polar plots of n n and for n = 1-5.7 As n grows, the forward-direction lobe becomes narrower. This behavior of n n and explains the enhancement of forw ard scattering in case of large particle because the increased sphere size incorporates the higher order n n and in the scattering diagram.7 The scattering coefficients na for the electric multipoles and nb for the magnetic multipoles are given by7 ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( a' n n n n n n n n nmx x x mx m mx x x mx m and (3.46) ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( b' n n n n n n n n nmx x m x mx mx x m x mx where ) ( J ) 2 1 ( ) ( 1/2 n 1/2 nx x x, ) ( H ) 2 1 ( ) ( 1/2 n 1/2 nx x x(3.47) are the Riccati-Bessel functions, and ) ( J1/2 nx, ) ( H1/2 nx are spherical Bessel functions7

PAGE 63

40 Figure 3.4. The polar plots of the angular functions n n and Cited from Reference 7.

PAGE 64

41 The scattering coefficients n nb and a, which are functions of the size and optical properties of the particle a nd the nature of the surrounding medium in addition to the wavelength and the polarization status of th e incident light, allow determining all the observable quantities associated with scattering and absorption like cross sections and the scattering matrix elements. The cross sections for Mie sc attering are given as follows60 ) b a ( ) 1 n 2 ( k 22 n 2 n 1 n 2 scaC, and (3.48) )} b a Re( { ) 1 n 2 ( k 2n n 1 n 2 extC The efficiencies are ) b a ( ) 1 n 2 ( k 22 n 2 n 1 n 2 a Qsca and (3.49) )} b a Re( { ) 1 n 2 ( k 2n n 1 n 2 a Qext

PAGE 65

423.3.4. The Fraunhofer Diffraction Fraunhofer diffraction is valid if the in cident light is unpolarized, particles are large, and scattering is close to the forward direction. Fo r a large sphere of radius a, the scattering amplitude which is inde pendent of the azimuthal angle is7 sin ) sin ( 2 ) cos 1 ( ) (1 2x x J x S (3.50) where J1 is the Bessel function of the first kind. The scattered light intensity is given30 2 2 1 2 4 2cos 1 2 1 sin ) sin ( 2 16 x x J a I Ii sca (3.51) where the term inside the bracket defines th e Airy function. The angular position of the first minimum in intensity can be found from the expression30 a 22 1 sin (3.52) If the refractive index of the particle is very low ( 1 1 m), the particle transmits light almost without deflection. This transmitted light interferes with diffracted

PAGE 66

43 light and results in the anomalous diffraction. The scattering efficiency for the anomalous diffraction is60 ) cos 1 )( 4 ( sin ) 4 ( 22 extQ (3.53) where ) 1 ( 2 m x (3.54) 3.4. Transmission If a plane homogeneous wave traver ses a medium, its intensity becomes exponentially attenuated by scattering and ab sorption as the wave crosses the medium. Transmission is the ratio of the intensities measured with and without the sample interposed between light source and detector.7 Transmission measurement is indispensable for the estimation of the JPPD b ecause the acquired spectra can be used to derive the chemical composition information in addition to the particle size distribution. Transmission has been expressed in di fferent forms depending on the knowledge that is sought from the material system under study. These include the Bouger-Lambert law, the turbidity, and the Beer-Lambert law.19 The Bouger-Lambert law is the most general expression relating transmission to th e attenuation of the incident radiation by sample material. It doesnt impose any restri ctions on either the attenuation mechanism

PAGE 67

44 or the material under investigation. The tu rbidity provides the basis for understanding physical phenomena of incident radiation extinction. On the other hand, the BeerLambert law allows the estimation of chemical composition or chromophore concentration. Because these are merely di fferent expressions of the same phenomena, i.e., they have the same functional form, th ere is enough room for confusion about their derivation, interpretation and usage. Therefore, the description of transmission theory in this section is focused on the clarification of similarities and differences among the Bouger-Lambert law, the turbidity and the Beer-Lambert law. 3.4.1 Transmission and Absorbance If a plane wave traverses an ab sorbing medium of path length l, the intensity of the wave after crossing the medium is described by:7 l i trse I I (3.55) where iI is the intensity of the wave before entering the medium and is the absorption (also called attenuation or extinction) coefficient. The absorption coefficient is a distinctive property of material. can be derived from the Poynting vector of a plane wave and expressed as7 k4 (3.56)

PAGE 68

45 Here, the imaginary part of the complex refractive index k decides the rate at which electromagnetic energy is removed from the wave as it propagates through the medium. The transmission T is defined as the ra tio of the intensities iI and trsI 7 l i trse I I T (3.57) The logarithmic form of Eq. (3.57) is known as the Bouger & Lambert law19 l I Ii trs ) ( ln (3.58) The Bouger-Lambert law specifies the absorp tion or the extincti on of the incident radiation due to material. Ther e is no restriction in the attenuation mechanism or material to apply the Bouger-Lambert law. The absorbance A is a different way to relate the ratio of the intensities of the incident light and the transmitted light. It is defined as the log of the inverse of the transmission59 T I I I I Ai trs trs ilog log log (3.59)

PAGE 69

46 It is evident that the measurement of iI and trsI leads to the information of the absorption coefficient and the imaginary part of the complex refractive index k. However, this kind of measurement should be done carefully to avoid or minimize unwanted effects like reflections from cuvette walls.7 3.4.2 Turbidity When light passes through a particulat e medium, it loses intensity due to absorption and scattering by pa rticles in the medium. This loss of electromagnetic wave intensity defines extinction. If multiple sc attering is negligible, the transmission Eq. (3.57) can be rewritten7 ) exp( l I I Text i trs (3.60) where the extinction coefficient ext is given by7 ) (sca abs p ext p extC C N C N (3.61) In Eq. (3.61), pN is the total number of particles per unit volume, sca abs extC C Cand ,are the extinction, absorption and scatte ring cross sections, respectively.

PAGE 70

47 Turbidity is defined as the logarithm of the ratio of transmitted to incident radiation if absorption and scat tering are present. For diluted solutions, it has been shown that the turbidity is given by19 l C N l I Iext p ext trs i ln (3.62) The turbidity ) ( measured at a given wavelength can be related to the number, size, and optical constants of suspended isotropi c spherical particles through the Mie theory19 dD D f D x m Q l Next p) ( ) ( ) 4 ( ) (2 0 (3.63) Where D is the particle diameter, and f (D) is the normalized particle size distribution for spherical particles. The Mie extinction efficiency) (x m Qextis a function of the particle size parameter x and the complex refractive index rati o of the particle to the surrounding medium m. Though it is not easy, turbidity Eq. (3.63) can be solved by inversion techniques, either analytical inversio n or numerical inversion.29 The resulting solution yields information regarding particle number, size dist ribution, or optical properties of material. Thus, it can be used in different ways de pending on the information sought and the data available. For example, if the optical prope rties are known as functions of wavelength, Eq. (3.63) can be used to estimate the particle size distribution or vice versa.35

PAGE 71

48 Measurement of extinction requires the ex clusion of the forward-scattered light entering the detector. However, it is difficult to practice experimentally, especially for large particles because forward scattering is enhanced with incr easing particle sizes.7 Therefore, the term optical dens ity (OD) that is defined as T1 log, is frequently used to express the magnitude of the measured transmission7,31 Taking this into account, optical density is preferably used to show the magn itude of the measured transmission in this dissertation following Bohren and Huffman.7 3.4.3 Beer-Lambert Law If the attenuation of the incident wa ve is caused by the chromophores of the absorbing medium, transmission can be expres sed using the molar absorption coefficient ) ( the concentration of chromophores c, and the sample path-length l l c i trse I I T) ( (3.64) where ) ( 4 ) (k (3.65)

PAGE 72

49 The molar absorption coefficient ) ( is specific to the chromophores of substance and may be regarded as a subset of the absorption coefficient defined in Eq. (3.56) that can be applicable to any absorbing material. Eq. (3.64) can be rewritten to obtain the BeerLambert law19 l I Ii trsc ) ( ln (3.66) The Beer-Lambert law expresses that the decreased incident light intensity is proportional to the molar absorption coefficient ) ( the concentration of chromophores c, and the sample path-length l. It is well known that the Beer-Lambert law is a special case of Mie theory. The turbidity expression becomes the Beer-Lambert law if there is no scattering due to small particle sizes, less than 10 nm for polystyrene, or the sample to medium refractive index ratio close to1.19 If these requirements cannot be fulfilled, then scattering effects can be accounted for and removed from the measured extinction spectrum by deconvolution.35 3.5. Fluorescence Spectroscopy If a molecule absorbs electromagnetic ra diation, it can be stimulated to an electronically excited state. The molecules inherent structure and the physiochemical characteristics of its local environment d ecide the molecules pathway to the ground state. In some situations, it returns to the ground state by emitting electromagnetic

PAGE 73

50 radiation. Fluorescence is the emission of light from the singlet excited states in which the electron has the oppos ite spin orientation as the ground-state electron.32 Therefore, return to the ground state is spin-allowed and thus, end in the fast emission rates (108 s-1). Typical fluorescence lifetime is about 10 ns.32 Fluorescence has been known for more than 300 years and has proven to be an indisp ensable tool in biology, biochemistry, and chemistry because of its superb detection limits.9 In this part of theory, the basic aspects of fluorescence spectroscopy and introduction to fluorescence anisot ropy are described. 3.5.1. Characteristics of Fluorescence Spectrum The energy level diagram illustrates absorption and emission of light by molecules as shown in Fig 3.5.32 When the sample under inves tigation is i rradiated, the molecules with the lowest vibrational ener gy primarily absorb th e excitation energy. Electrons in the substance are generally excited to some higher vibrational levels of either 2 1S or S and then, with a few uncommon exceptions, promptly relax to the lowest vibrational level of 1S. This process is calle d internal conversion.32 Because internal conversion usually occurs in 10-12 s, far shorter than fluorescence lifetime, it results in fluorescence emission from a thermally equilibrated excited state, i.e., the lowest-energy vibrational state of 1S. If molecules in the 1Sstate experience a spin conversi on to the first triplet state 1T, this process is called intersystem crossing.32 Phosphorescence is the emission of light from triplet states. It is forbidden transiti on and thus, results in the slow emission rates (103-100 s-1).32 The phosphorescence lifetimes are normally milliseconds to seconds.

PAGE 74

51 The characteristics of fluorescence emissi on are I) Stokes Shift the wavelength of the emitted light is longer than that of th e excited light, II) Kashas Rule the emission spectra are typically independent of the ex citation wavelength, III) mirror image in general, the emission is the mirror image of the 1 0S S absorption alone, not of the total absorption spectrum.32 Figure 3.5 Energy level diagram illustrati ng absorption and emission of light by molecules. Adapted from Reference 32. 3.5.2. Fluorescence Lifetimes and Quantum Yields The fluorescence quantum yield is defined by the ratio of the number of emitted photons to the number of absorbed photons and is expressed by32 nrk (3.67)

PAGE 75

52 where is the radiative rate constant and nrk is the nonradiative rate constant. Both radiative and nonradiati ve processes depopulate the excite d state as shown in the Fig. 3.6.32 Figure. 3.6. Depopulation of the excited st ates by radiative and nonradiative decay. Adapted from Reference 32. The fluorescence lifetime is the average time that the molecule spends in the excited state before its retu rn to the ground state. Fluor escence lifetime is given by32 nr flk 1 (3.68) Because fluorescence emission is a random pro cess, few molecules emit their photons at exactly flt In case of a single exponential decay, 63% of the molecules have decayed S1 S0 Relaxation (10-12s) S1 kn r Absorption

PAGE 76

53 before flt while 37 % decay after flt Due to complexity, the determination of the lifetime from multiple exponential decays requires numerical analysis.9,32 3.5.3. Fluorescence Quenching The decreases of fluorescence intensity by a wide variety of processes are called quenching.32 For example, collisional quenching ha ppens if the excited-state fluorophore is deactivated upon contact with some other mo lecule, the quencher, in solution. In case of static quenching, fluorophores form nonfl uorescent complexes with quenchers and this process takes place in the ground state and does not depend on diffu sion or molecular collisions. Quenching can also arise by numerous nonmolecular mechanisms like the attenuation of the incident light by the fl uorophore itself or othe r absorbing species. Fluorescence quenching yields information necessary to understand the role of the excited-state lifetime to detect dynamic pr ocesses in solution or in macromolecules.32 3.5.4. Fluorescence Anisotropy Anisotropy measurements are essential for the biochemical applications of fluorescence and yield information on the size an d shape of proteins or the rigidity of various molecular environments.32 Photoselective excitation of fluorophores by polarized light provides the grounds of anisotropy measurement because fluorophores favorably absorb photons of which electric vectors are aligned parallel to the transition moment of the fluorophore. The selectiv e excitation follows a partia lly oriented population of

PAGE 77

54 fluorophores (photoselection) and partially polarized fluorescence emission.32 The fluorescence anisotropy flr and polarization flP are given by9,32 fl fl fl fl flI I I I r 2// // fl fl fl fl flI I I I P// // (3.69) where fl flI I and// represent the fluorescence intensities of the vertically (//) and horizontally () polarized emission if the sample is excited with the vertically polarized light. Anisotropy can be converted into pol arization and vice versa using Eq. (3.70)32 fl fl flr r P 2 3 fl fl flP P r 3 2 (3.70)

PAGE 78

55 CHAPTER 4. DEVELOPMENT This chapter provides information pertinen t to the development of the Integrated UV-VIS MAMW spectrometer. For development background, the analysis of the research activities related to the protot ype MAMW spectrometer that motivated the development of the Integrated UV-VIS MAMW spectrometer is described. The development became possible by observing and fu lly enabling the characteristics of each instrument that constitutes the Integrated UV-VIS MAMW spectrometer. Specifications of each instrument, the design criteria, a nd overall lay out of the Integrated UV-VIS MAMW spectrometer are described in this chap ter. The credibility of the measured UVVIS MAMW spectra depends on the reproduci bility of the data and the angle of acceptance. Therefore, the last two sections of the chapter discuss procedures necessary to secure the reproducibility of the measured data and the analysis of the effect of the angle of acceptance. 4.1. Development Background This section provides the development background of the Integrated UV-VIS MAMW spectrometer. It includes the descri ption of the prototype MAMW spectrometer and the analysis of its limitation. Methods that overcome the shortfalls of the prototype

PAGE 79

56 MAMW spectrometer and lead to the deve lopment of the Integrated UV-VIS MAMW spectrometer are described in detail. 4.1.1. Description of the Prototype MAMW Spectrometer Dr. Bacon developed a MAMW spectrometer that has the capacity of measuring the scattered light from 30o to 90o within the wavelength range of 400 ~ 800nm.2 Throughout this dissertation, we will designate it as the prototype MAMW spectrometer. It utilized Oriel Instruments model 6333 100-W tungsten-halogen lamp that was installed in Oriel 60020 convective lamp housing as light source, a 200 m UV transparent optical fiber as incident light delivery tool, and an Ocean Optics S2000 production miniature CCD diode array spectrometer (S2000 production spectrometer) as detector. In addition, a fused silica lens with 13 mm diameter a nd 25 mm focal length was attached in front of the detector to collect scattered light The schematic of the prototype MAMW spectrometer is shown in Fig.4.1.2 The prototype MAMW spectrometer had b een tested by measuring the JPPD of the selected particle standa rds simultaneously and validated by comparing the measured results to the ones obtained by simulations. Th ese standards include polystyrene spheres (500 nm, 1 m, 10 m, hollow 1 m, and red dyed 3 m), silica (470 nm, 1 m), and silver chloride cube (500 nm) in water and pure red blood cells in saline. Fig.4.2.a and Fig. 4.3.a show the MAMW response surface of polystyrene 1 m standard in water measured by the prototype MAMW spectrome ter and the contour plot, respectively.2

PAGE 80

57 Figure.4.1. Schematic of the pr ototype MAMW spectrometer. Cited from Reference 2.

PAGE 81

58 Fig.4.2.b and Fig. 4.3.b represent the simulated MAMW response surface of polystyrene 1 m standard in water and its contour plot, respectively.2 The fact that the measured MAMW responsive surfaces of polystyrene and silica standards are consistent with those obtained by simulation had proved the proposi tion that MAMW spectroscopy could be a modality of micron and sub-micron size particle characterization by the simultaneous detection of the JPPD. 4.1.2. Limitations of the Prot otype MAMW Spectrometer The prototype MAMW spectrometer has some limitations in both measurement ranges and applicable methodologies despit e its success in proof of the particle characterization by the simultaneous JPPD m easurement concept. These limitations are as listed: 1. Lack of the low-angle scattering meas urement capability: the prototype MAMW spectrometer was unable to perform scattere d light measurements at angles lower than 30o, 2. Inability to perform measurements in th e UV range: measurements were done only in the visible to short NIR range of 400 ~ 800 nm and 3. Inability to adopt spectros copic techniques other than li ght scattering: incorporation of transmission and fluorescence spectroscopy or polarization optics was not possible. Measurement of the low-angle scattered li ght is crucial for the characterization of large size particles including bi ological particles such as ba cteria or blood cells because

PAGE 82

59 Figure 4.2. MAMW response surf aces of polystyrene 1 m standard in water. a) MAMW response surface of polystyrene 1 m standard in water measured by the prototype MAMW spectrometer. b) Simulated MAMW response surface of polystyrene 1 m standard in water. Ci ted from Reference 2. a b

PAGE 83

60 Figure. 4.3. Contour plots of MAMW response surfaces of polystyrene 1 m standard in water. a) Contour plot of MAMW response surface of polystyrene 1 m standard in water measured by the prototype MAMW spectro meter. b) Contour plot of the simulated MAMW responsive surface of polystyrene 1 m standard in water. Cited from Reference 2. a b

PAGE 84

61 increased particle size enhances forward sc attering or diffraction as shown in Fig. 4.4, simulated Rayleigh ratios for pa rticles with different sizes.17 The UV spectroscopy is important for biol ogical particle characterization due to the absorption of UV light by DNA (260 nm) and protein (280 nm). Moreover, the use of UV light can reinforce the capability of sma ll size particle charac terization because the increased ratio of particle size to the wave length of the incident light enhances the interference effects. Therefor e, the adoption of UV light is desirable for the MAMW spectrometer. Besides, transmission and fl uorescence spectroscopy will add additional dimensions to the particle characterizati on capabilities of the MAMW spectrometer because they are superb in the detection of particle composition. Taking all these into consideration, the necessity of developing a MAMW spectrometer that encompasses all the merits stated above has drawn significan t research interests, especially for the characterization of bi ological particles. 4.1.3. Development of the Integrat ed UV-VIS MAMW Spectrometer The limitations of the prototype MAMW spectrometer are attributed to: 1. the use of single integration time fo r the S2000 production spectrometer and 2. the adoption of optical fiber as the incident light delivery tool. Firstly, considering the fact that the scat tered light intensity va ries several orders of magnitudes depending on particle sizes, wa ys to avoid the dete ctor saturation should be devised. Typically, this can be done by th e insertion of neutral density filter (NDF).7

PAGE 85

62 Figure 4.4 Simulated Rayleigh ratios for part icles with different sizes. Cited from Reference 17.

PAGE 86

63 However, there was no room for NDF in case of the prototype MAMW spectrometer because of the use of fiber. Another way to a void the detector saturation is adjusting the signal intensity by manipulating integrat ion time. In case of Ocean Optics S2000 spectrometer, this can be done by the OOIBase 32 spectrometer operating software that is installed in the computer. The reason is uncl ear but single detector integration time of 2000 ms had been used for all the measur ements done with the prototype MAMW spectrometer. Because this integration time wa s set for the measurement of the scattered light from 30o to 90o, which has several orders of magnitude weaker intensity compared to that of the low-angle scattered light, it is apparent that the detector saturation made the low-angle scattering and transmissi on measurements not possible. Secondly, optical fiber strongly abso rbs UV light and thus, resulted in the disparity in intensities between UV li ght and VIS light as shown in Fig.4.5.2 As a result, UV light was not used in the prototype MA MW spectrometer. Moreover, optical fiber allows only small portion of light to be delivered to the sample cell due to its small cross section. Consequently, weak signal intensity made fluorescence measurement unfeasible. Methods to overcome the limitations of the prototype MAMW spectrometer have been sought. These include: 1) use of UV-VIS fused silica lens es as light delivery tool, 2) adoption of a method of integration ti me multiplexing (ITM), optimized use of the dynamic range of Ocean Optics S2000 production spectrometer, to avoid detector saturation and

PAGE 87

64 Figure 4.5. UV-VIS beam spectrum measured w ith the prototype MAMW spectrometer. Note the disparity in intensities between UV and visible beam due to the use of optical fiber as incident light delivery tool. Cited from Reference 2.

PAGE 88

65 3) installation of a c ouple of narrow width slits befo re and after the sample cell (cuvette) to enhance the capability of th e low-angle scattered light detection. Typical uncoated fused silica UV lens is transparent from 190 nm to 2100 nm.39 Therefore, very little UV light is absorbed by lens and thus, the disparity between the intensities of UV light and visible light become s negligible with the adoption of lenses as shown in Fig. 4.6. Besides, lens has large cross section compared to the fiber. As a result, the intensity of light delivered via lens is far stronger than that delivered via fiber. Consequently, the measurement of th e fluorescence spectra with the MAMW spectrometer becomes feasible with the adoption of lenses. Ocean Optics S2000 spectrometer has 2 x 10 8 dynamic range that is determined by multiplying the ratio of the maximum to the minimum integration time (2 x 104 for 3 msec sec to 65 sec) by the ratio of the maximum to the minimum signal counts for single scan (2 x 103 for 2 to 4000 signal counts).43 Therefore, if we choose the proper integration time, detector saturation can be avoided and the detection of transmitted or scattered light for a wide range of angles without the use of NDF becomes possible. Fig. 4.7 illustrates the UV-VIS beam spectra record ed with different integration times. It shows that the intensity of UV-VIS beam is proportional to the used integration time. For the Integrated UV-VIS MAMW spectrometer, integration time of 5,000 msec (1 x 103) has been used. Together with a single scan dynamic range, th is yields dynamic range of 105 ~ 106. The intensities of signals caught with different integration times must be corrected for the standard integration time and this can be accomplished by multiplying the proper integration time ratios.

PAGE 89

66 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 3000 3500 4000 Wavelength, nmCounts5-13-04 uv mamw beam profile Figure 4.6. UV-VIS beam spectrum measured with the Integrated UV-VIS MAMW spectrometer. Disparity in intensities between UV beam and visible beam is negligible due to the use of lenses. Wavelength (nm) Counts

PAGE 90

67 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 3000 3500 4000 4500 CountsIntegration Time Multiplexing S2000 Integration Time blue: 6 msec red: 48 msec Figure 4.7. UV-VIS beam spectra measured with the Integrated UV-VIS MAMW spectrometer with different integration times The UV-VIS beam intensity is proportional to the integration time. No neutral density filter was used for this measurement. Wavelength (nm) Counts

PAGE 91

68 The detection of the low-a ngle scattered light requires the use of well-collimated narrow width incident beam. This becomes possible with the use of a collimation lens and a couple of narrow widt h slits installed before and after the sample cell. The narrow width slits adjust the width of the incident beam and the scattered light that enters the receiving optics. Unlike fiber, the collimati on lens ensures large collimation length of the incident beam near the sample cell and thus, enables the optimum separation of the sample cell and the receiving optics. Together with the adjustable integration time, the use of narrow width slits and collimation le ns enables the detection of the low-angle scattered light. The development of the Integrated UV-VIS MAMW spectrometer becomes feasible by the adoption of UV lenses, in tegration time multiplexing, and narrow width slits. Because it can perform not only scatte ring but also transmission and fluorescence spectroscopy, it is an integrated spec trometer. The Integrated UV-VIS MAMW spectrometer has the capabil ities of performing: 1. low angle scattering measurement starting from 4o with simultaneous detection of multi-wavelength light from 200 nm to 820 nm, 2. UV-VIS transmission measurement with wa velengths for analys is ranging from 200 nm to 800 nm, and 3. UV-VIS fluorescence spectra measurement. As a result, the Integrated UV-VIS MA MW spectrometer can yield abundant information necessary to characterize micr on and sub-micron size pa rticles. Moreover,

PAGE 92

69 the use of lenses allows room to install polarization optics to measure the 16-element Mueller matrix for light scattering, linear or circular dichrois m in transmission, and fluorescence anisotropy. 4.2. Instruments Specifications of the instruments used to build the Integrated UV-VIS MAMW spectrometer are described in this section. Particularly, the char acteristics of each instrument are provided in detail because the development of the Integrated UV-VIS MAMW spectrometer was possible by fully enabling them. 4.2.1. Light Source Ocean optics DT-1000 deuterium tungsten-hal ogen light source is selected as the light source of the Integrated UV-VIS MAMW spectrometer. It prov ides the continuous UV to short NIR light in th e spectral range of 200 ~ 1100 nm. Its dimension is 310 mm (L) x 172 mm (W) x 175 mm (H). The output po wer of deuterium lamp is 30 W and that of tungsten-halogen lamp is 6.5 W. The peak -to-peak stability is 0.05 % (maximum) and drift is 0.5% per hour.43 Fig. 4.8.a is the photograph of DT-1000 light source and Fig. 4.8.b shows its inside structure. Fig. 4.9 is th e schematic of bulbs and a UV-VIS beam focusing lens. Visible light emitted from the tungsten-halogen (T-H) bulb diffuses everywhere. As a result, the shadow of f ilament is removed and the diffused beam

PAGE 93

70 Figure 4.8. Photographs of Ocean Optics DT-1 000 light source. a) Exterior view. b) Inside structure. a b T-H bulb Deuterium bulb UV lens

PAGE 94

71 Figure 4.9. Illustration of UV-VIS beam combination inside the DT-1000 light source. T-H (VIS) bulb Deuterium (UV) bulb Fiber Connector UV-VIS lens

PAGE 95

72 becomes spatially uniform. Part of the diffused beam passes through the hole made in aluminum plate that separa tes tungsten-halogen and deut erium bulbs and enters the deuterium bulb, the source of UV light. The me tal reflectance plate inside the deuterium bulb also has a hole and allows the penetration of the visible light. Th erefore, without the use of any beam-combiner, the visible light is combined with the UV light and the combined beam can be placed in a single optical path. The combined UV-VIS beam enters the fused silica lens with 5 mm diameter and 10 mm focal length. The lens originally has the role of focusing the UV-VIS beam to the tip of optical fiber that is connected to the DT-1000 light source. The output UV-VIS beam from the DT-1000 light source is collected and collimated by fused silica lens with 50 mm diameter and 76 mm focal length and sent to the samp le cell. In case of the Oriel 6333 visible light source, which was used for the prototype MAMW sp ectrometer, the lamp stability was not good because it uses air convection system to cool dow n the bulb. The intensity fluctuation of visible output became noticeable after 3 hours operation.2 On the other hand, for the DT1000 light source, a cooling fan is adopted to cool down the bulbs and thus, very good stability of the output beam intensity is achie ved. Once the lamp output is stabilized, less than 1 percent of the maximum intensity fluc tuation was observed even after the 10 hours of continuous usage. Spatially uniform beam quality and long-term output power stability make the DT-1000 light source suitable as th e light source for the Integrated UV-VIS MAMW spectrometer.

PAGE 96

73 4.2.2. Optics Multiangle-multiwavelength spectroscopy cove rs spectral range from UV to short NIR region. Therefore uncoated UV grade fuse d silica lenses, which are transparent from 190nm to 2100 nm, are selected to deliver inci dent and scattered light. Total of five lenses are used in the construction of th e Integrated UV-VIS MAMW spectrometer. For convenience, each lens will be designated by L1, L2, L3, L4 and L5, respectively. L1 represents the lens for the in cident UV-VIS beam collimati on. L2 is the lens for focusing the incident light to the sample cell in cas e of fluorescence measurement. L3 designates the objective lens that collects light from th e sample cell. L4 represents the lens for steering the collected light to the focusing lens and L5 is the focusing lens that focuses the received beam to the spectrometer. Table 4.1 shows the specifica tion of these lenses. All the lenses except L5 are plano-convex lenses because this type of lens is proper to obtain good collimated beam. Ideally the use of achromatic lenses is desirable because they minimize chromatic aberration, the variat ion of lens focal length with different wavelength light. The effect of chromatic ab erration is significan t in case of UV light. However, there are no commercially available achromatic lenses that cover spectral range from UV to visible and thus, the idea of using the achromatic lenses is abandoned. In case of L5, bi-convex lens is chosen in order to secure more power and thus, improve the signal sensitivity at the detector end.

PAGE 97

74 Table 4.1. Specification of lenses used for the Integrated UV-VIS MAMW spectrometer Lens Manufactur er Model Focal length Diameter f-number* (f/#) L1 Thorlabs LA4078 75mm 51mm 1.5. L2 Newport SPX019 76mm 25mm 3.0 L3 Newport SPX016 50mm 25mm 2.0 L4 Newport SPX016 50mm 25mm 2.0 L5 Newport SBX019 25mm 25mm 1.0 f/# = focal length of lens / lens diameter

PAGE 98

75 4.2.3. Slits Slits are necessary to obtain a narrowwidth uniformly collimated beam to irradiate the sample cell and to allow only th e scattered or transmitted light within the angle of acceptance enters the detector. Initially a couple of slits with adjustable width were made using aluminum plates and razo r blades. Fig. 4.10 shows the schematic of these slits. Both slits have the same di mension of 51 mm x 51 mm x 3 mm. The size of razor blade is 17 mm (W) x 42 mm (H). A rect angular aperture of 9 mm (W) x 25 mm (H) was made in the middle of each plate. The maximum achievable slit width is determined by this aperture width. Desired slit width can be obt ained by sliding two razor blades that are placed in the grooves on the top and bottom of the plate. It is pointed out that slits, especially the one between th e cuvette and the objective lens, should be made using sharp edges such as razor blad es to prevent any stray light entering the receiving optics through the reflection at the edges.58 However, it was observed that the edge of the razor blade also deflects light due to its dimens ion. Thus, the idea of using the sharp edges was discarded. To minimize the refl ection from the slit or the slit mount, the razor blades are either wrapped with black ta pes or stained using a black marker pen and the aluminum plates are c overed with black felt. During the last phase of resear ch, the structure of the slits was slightly modified to minimize the amount of background UV light, wh ich originates from stray reflection, entering the receiving optic s. This is described in detail in Section 5.3.5.

PAGE 99

76 Figure. 4.10. Schematic of the slit with ad justable width. The si ze of the rectangular aperture is 9mm (W) x 25 mm (H). razor blade The edge of razor blade razor blade Groove Aperture Groove

PAGE 100

77 4.2.4. Sample Cell The sample cell plays an important role in the UV-VIS MAMW spectroscopy because the shape of the sample cell also imposes limitation on measurable scattering angles. Moreover, the required corre ctions of the measured results vary depending on the shape of the sample cell. Different shapes of sample cells are available for the light scattering measurements. These include rectangular, cylindrical, semi-octagonal, octagonal, and other cells of various shape.2 Dr. Bacon summarized the relation between the shape of the sample cell and the limitation it imposes on the scattering measurements.2 In case of cylindrical cells, the measurement of light scattering for a wide range of angles is possible. In addition, re fraction correction is not necessary. However, they were not used for the MAMW spectromete r due to the fact that low angles are not accessible because of multiple reflections from the walls.53 Instead, two different quartz sample cells, a rectangular fluorescent cuvett e and an octagonal ce ll were used for the measurement of the UV-VIS MAMW spectra. The octagonal cell that permits meas urements on all sides of the cuvette was adapted from the semi-octagonal cell developed by Brice et al.2,8 It was used as the sample cell for the prototype MAMW spectro meter. It is a cust om product made by Precision Glass, Inc. The cuvette is made of Spectrosil quartz plates and transparent from170 nm to 2700 nm. In order to minimize refraction effects, only 1 mm thick plates were used to make it. Its width is 25 mm a nd height is 50 mm. The octagonal cell yields eight perpendicular faces at observa tion angles of 45, 90, 135, 180, 225, 270, 315, and 360 degrees. Consequently, it allows the simultaneous measurement of scattering,

PAGE 101

78 fluorescence and transmission spectra. On the other hand, the corners of the octagonal cell obstruct scatte ring measurements at 22.5, 67.5, 112.5, 157.5, 202.5, 247.5, 292.5, and 337.5 degrees. The measurement results acquired using the octagonal cell may require corrections for reflection and refracti on. The octagonal cuvette is not much used for the Integrated UV-VIS MAMW spectromete r because it is very fragile while its replacement is not easy due to expens ive cost and substantial waiting time for production. Therefore, its usage was limited for the scattering measurements that require measurements at observation angles greater than 35o. For main experiments, a commercially available quart z rectangular fluorescent cell was adopted (Starna Cells, Inc. Model # 23-Q-10). All four windows of the cell are polished and transparent. Its dimension is 12.5 mm (W) x 12.5 mm (L) x 48 mm (H) and path length is 10 mm. Because it is made of Spectrosil Far UV quartz, it has no background fluorescence This cuvette was used for the scattering measurements of up to 35o in observation angle due to the refraction from the second face. The scattering measurement results acquired using the rect angular cuvette particularly needs the correction of: 1) refraction because the sca ttered light does not incide nt perpendicularly on the face of the cuvette except at th e angles of 0, 90, 180, and 270 degrees and 2) the variation of the scattered-light path length because the scattered-light path length is not identical at all the angles.

PAGE 102

79 The methods of implementing th e required corrections that ar e originated from the shape of the sample cell are describe d in detail in section 5.3. 4.2.5. Goniometer A goniometer from the Brice-Phoenix t ype spectrophotometer system was used to build the UV-VIS MAMW spectrometer. The Brice-Phoenix type spectrophotometer is so reliable that it has been actively us ed for light scattering experiments requiring absolute calibration.8 The diameter of the goniometer turntable is ~ 40 cm. The minimum angular resolution that is achie vable by this turntable is 0.2o. All the angular adjustments were carried out by manually rota ting the goniometer turntable. 4.2.6. Spectrometer The nature of MAMW spectroscopy requires the use of portable spectrometer that can be installed on the goniometer arm. Ocean Optics S2000 production CCD diode array spectrometer was chosen as the spectromete r for both the prototype and the current Integrated UV-VIS MAMW spectrometer becau se of its good spectral resolution, high light throughput, and low levels of stray light detection.2 S2000 production spectrometer is a kind of S2000 spectrometer but different from typical S2000 spectrometer in appearance. Unlike typical S2000 spectrometer, optical bench of S2000 production spectrometer, S-optica l bench that is used for the all S2000 series spectrometer, is mounted on an elec tronics board with 100 mm x 130 mm size. S2000 production spectrometer is different fr om PC 2000 spectrometer that also has

PAGE 103

80 s-optical bench mounted on the pc board fo r installation into the motherboard of a desktop PC.43 S2000 production or PC 2000 spectrometer has the same features of typical S2000 spectrometer becaus e all of them use S-optical bench. Therefore, all these are virtually identical spectrometers. The dimension of S-optical bench is 89 mm x 63 mm x 18 mm. The S-optical bench is designed to accept light from a single optical fiber. Fig. 4.11 shows the schematic of S-optical bench.43 The focal length of the S2000 spectrometer is 42 mm and the f-number (f/#), which can be obtained by di viding the focal length of the collimating mirror by its diameter, is 3.3.62 A Czerny-Turner type grating that has groove density of 600 lines/mm and spectral range of 200 ~ 850 nm with blaze wavelength at 400 nm is fixed in place at the time of manufacture. Th e efficiency curve for this grating is shown in Fig.4.12.43 Light diffracted by grating is caugh t on the Sony ILX511 shallow-well linear CCD array silicon detector. It is com posed of 2048 pixels and each pixel has size of 14 m x 200 m. The well depth of each pixel is 62,500 electrons. It has the sensitivity of 90 photons per count or 2.9 x 10-17 joules per count at 400 nm. The corrected linearity is greater than 99.8%. The ma ximum signal to noise ratio of detector is 250:1 at full signal and its spectral range is 200 ~ 1100 nm. For our S2000 production spectrometer, the standard window of the CCD dete ctor is replaced by a quartz window coated with a phosphor material in order to enhance the performance of the spectrometer for UV applications. The entrance slit is 25 m wide and 1 mm high. The corresponding pixel resolution for 25 m width is ~ 4.2 pixels.

PAGE 104

81 Figure. 4.11. Schematic of the op tical layout of S-optical benc h that is used for all the S2000 series spectrometer.43 (Courtesy of Ocean Optics, Inc.)

PAGE 105

82 Figure 4.12. Grating efficiency cu rves for the S-optical bench.43 # 2 grating that covers 200 nm to 800 nm is currently used for the S2000 production spectrome ter. (Courtesy of Ocean Optics, Inc.)

PAGE 106

83 The optical resolution, measured as Fu ll Width Half Maximum (FWHM), of the spectrometer is given by43 Optical resolution (nm) = (pixels) elements detector Total (pixels) resolution pixel x (nm) range spectral Gratings (4.1) Thus, the optical resolution of the curr ent S2000 production spectrometer is 1.33 nm. The rate of data acquisition is determin ed by the data acquisition card. An Ocean Optics ADC 500 analog-to-digital conversion (ADC) board has been used for the data acquisition. It has12 bit, 4 channel conversion and maximu m 500 kHz A/D frequency. The spectrometer is controlled by OOIBase32 Spectrometer Operating Software, a 32-bit, user-customizable, advanced acquisition and display program. It can collect data from up to 8 spectrometer channels simultane ously and display the results in a single spectral window. With OOIBase32, different spectroscopic measurements are possible using scope mode, absorbance mode, transmissi on mode, and relative irradiation mode. The scope mode can be used to measure sa mple signal, reference, and background noise directly. Time domain experiment is also pos sible via external hard ware trigger function, though not used in this research.43 In addition, OOIBase32 soft ware has a function to control spectrometer integration time. Th e nominal dynamic range of S2000 spectrometer is 2 x 108. However, the practical dynamic range is limited by signal to noise ratio and it can be enhanced by using signa l averaging functions that in creases the signal to noise ratio. In OOIBase32 software, time base and spat ial base signal averaging are available.

PAGE 107

84 The time based analysis is done by a recursive filter of ns average samples where the newest sample replaces the oldest. If obser ved spectral structures are broad, spatial averaging or boxcar averaging can be used to improve signal to noise ratio. This can be done by averaging np pixel values side by side and re places the center value with an averaged value. Because spatial and ti me based algorithms are uncorrelated, the improvement in signal to noise ratio is the product of the two processes.62 Ocean Optics S2000 series spectrometers are optimized for minimum stray light. It is manufactured to remove possible sources of stray light except scattered light from the grating. Because the exposure to stray light from the grating is rather uniform across the CCD array, this can be measured and mathematically removed. OOI base32 software has algorithms to do this. In addition, this software automatically corre cts electrical-dark si gnal and prevents the drift of dark current.62 4.3. Set Up This section describes the de sign criteria and the overall lay out of the Integrated UV-VIS MAMW spectrometer. Especially, optic s lay out and alignment procedures are described in detail. 4.3.1 Design Criteria for the Integr ated UV-VIS MAMW Spectrometer Different spectrometer requires differe nt design criteria depending on its applications. In case of the Integrated UV-VIS MAMW spectrometer following design criteria should be satisfied:2,56

PAGE 108

85 1. Maximum sensitivity it should have se nsitivity over orders of magnitude to detect scattered light origin ates from a wide range of scattering angles, 2. Quick measurement multiangle-multiwav elength spectra should be measured within the temporal equilibrium of samples, 3. Information rich Measured sp ectra should produce ample information necessary to characterize complicated particles using the JPPD, 4. Reproducibility stability of the instruments should be guaranteed and 5. Statistically meaningful data measur ed data should be st atistically reliable. Among them requirements 1 to 3 are directly related to the available instruments, especially, the output power of light source and detector se nsitivity. Therefore, much attention was paid to secure the stability of the instruments and the acquisition of statistically meaningful data during the design stage. 4.3.2 Description of the Integrated UV-VIS MAMW Spectrometer Set Up The prototype MAMW spectrometer had been built inside the Brice-Phoenix type spectrometer. For this type of spectrometer, sample cell and detector are set in a metal box to prevent ambient light entering the de tector. However, this box was removed during the construction of the Integrated UV-VIS MAMW spectrometer for large room and ease of alignment. Background light can be prevented from entering the detection optics by turning off room light during the measurement.

PAGE 109

86 Fig. 4.13 is the photograph of the Integr ated UV-VIS MAMW spectrometer and Fig.4.14 is its schematic. The light source, Ocean optics DT-1000 UV-VIS light source, is secured on the optical bench. A fused silica neutral density filter is placed in front of the lamp and used to adjust the beam in tensity during the alig nment procedure if necessary. The UVVIS light proceeds to the collimation lens L1. Because of the large diameter of lens L1, 50 mm, there is no lo ss of incident light and good collimated beam with comparable UV and visible light cross sections is obtained. The diameter of UV beam is 22 mm and that of visible beam is 18 mm at the location of sample cell. In addition, large diameter reduces aberration effects. A focusi ng lens L2 may be installed after the lens L1 in case of fl uorescence measurement. This lens focuses the incident light for fluorescence excitation. L2 is remove d during the scattering or transmission measurement. The theory of light scatteri ng requires parallel in cident light, not converging or focused beam. If the incident light is converging or focused, its intensity varies across the beam path and thus, de fining the scattering volume becomes very difficult. Therefore, the use of narrow width slits that permits parallel incident beam is desirable. A narrow width slit with adjustable s lit width is installed in front of the sample cell. Initially, the location of this slit was 1 cm away from the sample cell. However, during the last phase of resear ch, it was relocated a few cm away from the sample cell to minimize the amount of background UV light en tering the receiving optics. The sample holder is located in the middle of the goniomet er turntable. Another slit is installed in front of the object lens L3 that is focused at the center of the sample cell.

PAGE 110

87 Figure 4.13. Photograph of the Integr ated UV-VIS MAMW spectrometer.

PAGE 111

88 Figure 4.14. Schematic of the Integr ated UV-VIS MAMW spectrometer.

PAGE 112

89 L3 collects scattered, transmitted or fluorescence light from the sample cell. The focal length of L3 is 76 mm. With this focal length, optimum angular resolution can be achieved without much loss of the scattered light from the sample cell. The collected beam is focused to the lens L4 that prevents the loss of collected beam and thus, increases the intensity of light that is sent to th e focusing lens L5. The location of L4 was determined empirically by observing the sh ape and intensity of the incident UV-VIS beam spectrum on the monitor screen. The focu sing lens L5 focuses the received light to the detector. The narrow slit width of the S2000 spectrometer, 25 m, prevents the majority of stray light entering the spectr ometer. The f/# of S2000 spectrometer is 3.3. Therefore, a lens with the same f/#, 25m m diameter and 76 mm focal length, should be used to focus the detected beam to the sp ectrometer because it optimizes the amount of detected light incident on the spectromet er grating. However, due to the physical dimension of the goniometer arm, it was unable to install 76 mm focal length lens. Instead, two shorter foca l length lenses (L4, and L5) were used to increase the amount of light entering the spectrometer as much as possible. The receiving optics and the spectrometer are installed on th e goniometer arm that can be rotated for scattering or fluorescence measurement. The alignment of the goniometer turntable was accomplished by obtaining the maximum signal intensity at 0o. The receiving optics and the spectrometer are shielded to prevent unwanted background ligh t entering the detector as shown in Fig.4.15. In addition, all the shiny part s are covered with black felt to minimize background light. The construction of the Integrated UV-VIS MAMW spectrometer

PAGE 113

90 Figure 4.15. Photograph of the shielded Integr ated UV-VIS MAMW spectrometer. The receiving optics and the spectrometer were covered to prevent unwanted backgr ound light entering the detector

PAGE 114

91 requires the accurate optics ali gnment due to the use of lens es. Detected signal is very sensitive to the alignment. For precise alignmen t, the position of optics or detector should be adjustable in all directions. This can be done by installing all the lenses and the detector on the stages (Newport, Model MT-X linear stage) that allow 9.5 mm movement in the perpendicular di rection to the incident light. By placing these stages on the rails, one in front of the sample cell and another af ter the sample cell, the adjustment in the direction parallel to the incident light also b ecomes possible. The rail after the cuvette has the role of the goniometer arm. All the op tics and spectrometer ar e locked after the alignment. The goniometer arm can rotate from 0o to 150o in the clockwis e direction and 0o to 140o in the counterclockwise direction. However, the act ual scattering measurement was done up to 60o in the clockwise direction due to th e weak scattered light intensity at large angles or the difficulty of alignm ent. Fig.4.16 is the UV-VIS MAMW response surface of polystyrene 1 m standard in water measured up to 60o in observation angle using octagonal cuvette. If mo re intense light source is av ailable, measurable angular range can be extended further. Because the majo rity of the scattered light is limited in the low-angle direction, measurement of up to 35o in observation angle yielded information necessary to characterize pa rticle standards used in th is dissertation research.

PAGE 115

92 Figure 4.16. UV-VIS MAMW response surface of polystyrene 1 m standard in water. Measurement was performed up to 60o in observation angle usi ng the octagonal cuvette. Observation angle (deg) Wavelength (nm) Log (Isca/Ii)

PAGE 116

93 4.4. Reproducibility The reliability of the particle characte rization results obtaine d by the Integrated UV-VIS MAMW spectrometer depends on th e accuracy and repr oducibility of the measured data. Errors that affect the accura cy and reproducibility of the experimental results can be classified into instrumental errors and experimental or human errors. Instrumental errors, which are inherent in th e instruments and thus inevitable, include the fluctuation of the incident light intensity and the detector stability. However, as stated in Section 4.2, Ocean Optics DT-1000 light source is very stable light source and S2000 production spectrometer is optimized for the mi nimum stray light. Besi des, its electrical dark signal is automatically corrected to prev ent the drift of dark current by the operating software OOIBase32. On the other hand, experiment al errors mechanical stability of the system, manual rotation of the goniometer arm, optics alignment, and sample preparationcan be reduced during the develo pment or measurement stage by exercise of care, implementation of improvement ideas, and practicing relevant protocols. Much effort was made to reduce the experimental e rrors. This resulted in the achievement of substantial UV-VIS beam spectrum reproducibil ity, less than 1 % fl uctuation after 10 hours continuous usage. 4.4.1. The Angular Adjustment of Goniometer. The goniometer of the UV-VIS spectrometer is mechanically stable and reliable. However, the manual rotation of the goniometer arm was the greatest source of error in case of the prototype MAMW spectrometer.2 This may be attributed to the single angle

PAGE 117

94 marker used to decide the rotated angl e of the goniometer arm (Fig.4.17.a) due to parallax by eyes. To solve this situation, the principle of micrometer was applied. Several lines with 1o resolution were drawn on the couple of white paper tape pieces. Then these pieces are attached to both sides of the angle marker as shown in Fig.4.17.b. By using these line-drawn paper tapes, it has been possible to reduce the angular resolution of the goniometer to approximately 0.05o from previous 0.2o in the prototype MAMW spectrometer. As a result, errors caused by th e manual rotation of the goniometer arm are reduced considerably. 4.4.2. Optics Alignment Although the use of the lenses as light deli very tool leads to the development of the Integrated UV-VIS MAMW spectrometer, it also resulted in the difficulty and delicacy of the optics alignmen t. Slight misalignment of optics easily resulted in the change of whole UV-VIS beam spectrum. Ther efore, securing the beam reproducibility was a significant challenge for the devel opment of the UV-VIS MAMW spectrometer. This issue was resolved by the use of the overlay function of the OOIBase32 software that allows simultaneous display of up to 8 spectra. The UV-VIS beam is aligned to acquire the best possible spectrum. The scope mode of the OOIBase32 software is used for this alignment. The shape of the UV-VI S beam spectrum is empirically determined because the calibration of UV-VIS beam sp ectrum for the whole wavelength region of 200 nm to 800 nm is virtually impossible.63 Once the optimum UV-VIS beam spectrum is acquired, it is recorded as the standard UV-VIS beam spectrum and stored in the

PAGE 118

95 Figure 4.17. Schematics of goniometer turntable angular position markers. a) Single line marker used in the prototype MAMW spect rometer. b) Micrometer-principle-applied multiple markers used for the Integrated UV-VIS MAMW spectrometer. 90o 90o a b

PAGE 119

96 computer. Thereafter, if there is any change of optics or the necessity of re-alignment, the standard UV-VIS beam spectrum is disp layed using the overlay function of the OOIBase32 software. By overlapping the in cident UV-VIS beam spectrum to this standard UV-VIS beam spectrum and adju sting the location of optics, precision alignment becomes possible. Optics holders like posts, bases or x-pos itioners are locked at their locations once the best UV-VIS beam spectrum is acquired. In addition, the micrometer principle is applied to the adjust ment of the optics holders too and the linedrawn paper tapes were attached to the optics posts to help the rec overy of alignment. 4.4.3. UV-VIS Beam Spectrum Reproducibility Measurement The implementation of the measures necessary to minimize the experimental errors results in the substantial improvement in the reproducibility of the incident beam spectrum compared to that of the protot ype MAMW spectrometer. For the prototype MAMW spectrometer, 6 % error in the reprodu cibility of the incident visible beam spectrum due to the readjustment of the goniometer arm was reported.2 On the other hand, the UV-VIS beam spectra measured during the 10 hours continuous usage of the Integrated UV-VIS MAMW spectrometer yielded less than 1 % error in the reproducibility as shown in the beam profile Fig.4.18. This ensures th e reliability of the data measured with the Integrated UV-VIS MAMW spectrometer.

PAGE 120

97 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 3000 Wavelength, nmCounts5-17-04 uv-vis mamw beam profiles for 10 hours Blue: 1hr Green: 2hr Red: 10 hr Figure 4.18. UV-VIS beam spectra variations during 10 hours continuous usage after the intensity stabilization. The maximum difference is less than 1 %. Note that the spectra measured after an hour (Blue) and two hours usage (Green) are not easy to differentiate because of the excellent reproducibility of the incident UV-VIS beam spectra. Wavelength (nm) Counts

PAGE 121

98 4.5. Angle of Acceptance Angle of acceptance plays an important role for transmission or light scattering experiment. The use of improper angle of acceptance can easily lead to spurious experimental results and thus, the reproducibility of the expe rimental results may not be guaranteed. These are primarily due to the angul ar averaging effects that depends on the angle of acceptance variation. A method to defi ne angle of acceptance for the Integrated UV-VIS MAMW spectrometer and the effect of the angle of acceptance on the measured results are investigated in this section. 4.5.1. Definition of the Angle of Acceptance The angle of acceptance is described as th e sum of half the angle of divergence or convergence of the incident beam and half the angle subtended by detector.61 However, if the incident beam is well collimated, the a ngle of acceptance can be defined as the half the angle received by detector. If lens is us ed for receiving optics, the exact angle of acceptance is defined by the lens-pinhole opti cal system. Fig. 4.19 shows a typical lenspinhole optical system.22 For this set up, the angle of acceptance is ) ( tan1 o p aaf r (4.2) where aa is the angle of acceptance, pris the radius of the pinhole and ofis the focal length of the objective lens.7 In case of the Integrated UV-VIS MAMW spectrometer, the incident UV-VIS beam is well collimated and th e receiving optics is comprised of lenses.

PAGE 122

99 Figure 4.19. Schematic defining the angle of acceptance in case of the lens-pinhole system. Adapted from Reference 22. Detector Lo aa rp fo Lo: Objective lens aa: Angle of acceptance fo: Focal length of pinhole rp: Radius of pinhole

PAGE 123

100 Therefore, the definition of the angle of acceptance for the lens-pinhole optical system should be adopted as the angle of accep tance for the Integrated UV-VIS MAMW spectrometer. However, without the use of achromatic lens es, estimation of the exact angle of acceptance is not possible due to chromatic aberration of UV-VIS beam. With current set up, which employs plano-c onvex lenses, focusing both UV and visible beam into a single spot is not possible. In addition, UV light is invisible and the estimation of focused beam spot size for UV beam is not an easy task. As a substitution, the definition of the angle of acceptance us ed for the prototype MAMW spectrometer, half the angle received by the detector, is adopted as the approximate angle of acceptance for the Integrated UV-VIS MAMW spectrometer. 4.5.2. Effect of the Angle of Acceptance The use of proper angle of acceptance is im portant in the extinction measurements, especially for large particles due to the enhanced forward scattering in the narrow forward direction. The use of inadequate angle of acceptance may lead to spurious results and lack of reproducibility. For example, Fig.4.20 shows two transmission spectra of purified red blood cell.41 The top represents transmission spectrum measured by HP 8453 spectrometer and the bottom is diffusive transmission spectrum measured by Perkin Elmer Lamda 900 spectrometer. Although the same sample is used for measurements, the resulting spectra have totally different profiles due to the different angle of acceptances of the two spectrometers. This is an exampl e that the extinction or forward scattering measurement results can be affected by the angle of acceptance of the spectrometer.

PAGE 124

101 OD spectrum of erythrocytes (HP vs Perkin-Elmer)0 0.2 0.4 0.6 0.8 1 1.2 1.4 1902903904905906907908909901090 Wavelength (nm)OD HP 8453 Perkin-Elmer Figure 4.20. Optical density spectra of pur ified red blood cell measured by HP 8453 spectrometer and Perkin Elmer Lamda 900 sp ectrometer. Inconsistency between two spectra is due to different angle of accep tance of the two spectrometers. Cited from Reference 41.

PAGE 125

102 The effect of the angle of acceptance is not absolute but relative and it depends on particle sizes. For small particles, forward scattering is negligible and thus, measured extinction will not be much affected by the change of the angle of acceptance. On the other hand, large particles e nhance forward scattering as particle size increases. Consequently, large particles will be more affected by the change of the angle of acceptance. Experiment was performed to c onfirm this proposition using polystyrene spheres of three different sizes-300 nm, 3 m, and 10 m. For this measurement, the Integrated UV-VIS MAMW spectrometer was temporarily disintegrated. The schematic for this experimental setup is shown in the Fig. 4.21. All the lenses except the collimation lens L1 are removed because lens es virtually increase beam path length and hence, reduce the angle of acceptance. Extinct ion of each sample was measured at three different locations within the incident UV-VI S beam collimation length. This is virtually identical to measuring the extinction of each sample with different angle of acceptance. The measured results are plotted as optical density because the forward-scattered light cannot be excluded. Fig. 4.22 is the optical density spectra of polystyrene 300 nm spheres in water measured at approximately 2 cm, 7 cm and 17 cm away from the sample cell, respectively. Fig. 4.23 a nd Fig. 4.24 correspond to polystyrene 3 m spheres and polystyrene 10 m spheres, respectively. In cases of polystyrene 300 nm spheres and polystyrene 10 m spheres, the results are not much affected by the location of the spectrometer. On the other hand, the optic al density spectra of polystyrene 3 m spheres vary depending on the distance from the samp le cell. For polystyrene300 nm spheres, particle size is small and thus, forward scattering is negligible Hence, the change of the

PAGE 126

103 Figure 4.21. Schematic of the experimental se t up measuring the effect of angle of acceptance changes on the measur ed optical density spectra of pa rticles with different sizes.

PAGE 127

104 200 300 400 500 600 700 800 0 0.2 0.4 0.6 0.8 1 1.2 1.4 OD PS 300 nm Optical dnsity Cuvette to S2000 distance blue: 2 cm red: 7 cm black:17 cm Figure. 4.22. The optical density spectra of polystyrene 300 nm spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from the sample cell using the experimental set up shown in the Fig.4.21. Wavelength (nm) Nomalized Optical Density

PAGE 128

105 200 300 400 500 600 700 800 0.3 0.4 0.5 0.6 0.7 0.8 0.9 OD PS 3 micron Optical dnsity Cuvette to S2000 distance blue: 2 cm red: 7 cm black:17 cm Figure. 4.23. The optical density spectra of polystyrene 3 m spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from the sample cell using the experimental set up shown in the Fig.4.21. Op tical density measured at 17 cm shows correct profile. Wavelength (nm) Normalized Optical Density

PAGE 129

106 200 300 400 500 600 700 800 0 0.2 0.4 0.6 0.8 1 1.2 OD PS 10 micron Optical dnsity Cuvette to S2000 distance blue: 2 cm red: 7 cm black:17 cm Figure. 4.24. The optical density spectra of polystyrene 10 m spheres in water measured at approximately 2 cm, 7 cm, and 17 cm away from the sample cell using the experimental set up shown in the Fig.4.21. Wavelength (nm) Normalized Optical Density

PAGE 130

107 angle of acceptance did not affect th e results. In case of polystyrene 10 m spheres, forward-scattered light is limited in the na rrow angle and hence, the observation of the effect of the angle of accepta nce change requires longer inci dent beam collimation length that is not obtainable with current setup. For polystyrene 3 m spheres, the measured optical density spectra subject to the angle of acceptan ce changes. This proves the proposition that the effect of the angle of acceptance change is not absolute but depends on particle sizes. Therefore, the effect of the angle of acceptance should be considered during the set up of the spectrometer. For extinction measurement, the use of sm all angle of acceptance is desirable. In case of the Integrated UV-VIS MAMW spect rometer, the measur ed optical density spectra of polystyrene spheres with different sizes are consistent with those obtained using HP8453 diode array spectrometer. The us e of objective lens virtually increases collimation length and thus, results in narro wing the angle of acceptance. However, for scattering experiment, less stri ct standard can be applied to determine the angle of acceptance because forward-scattered light is limited to the narrow angles close to the forward direction. In addition, the reduction of the angle of accep tance sacrifices the signal sensitivity and thus, limits the measurab le angular range for scattering experiment. On the other hand, the angle of acceptance canno t be increased without limit because the increment of the angle of acceptance results in averaging out the maxima and the minima of the measured spectra. Therefore, acquiri ng the exact scattering fe ature is not possible with large angle of acceptance. Dr. Bacon investigated the relation between the angle of

PAGE 131

108 acceptance and the averaging effect by theore tical simulations and concluded that the angle of acceptance of 4o is allowable for the measurem ent of the MAMW spectra of polystyrene 10 m spheres without the loss of the maxima and the minima.2 Therefore, 4o was chosen as the upper limit of the approxima te angle of acceptance for the Integrated UV-VIS MAMW spectrometer. For actual measurement, slit width of up to 6 mm that corresponds to approximately 2o of angle of acceptance was used. However, the determination of slit width requires the consideration of additional correction factors like refraction, reflection, or pa rticle sizes. These are discussed in Section 5.3.

PAGE 132

109 CHAPTER 5. MEASUREMENT This chapter describes topics related to measurement and data analysis including sample preparation, correction factors, and calibration. The reliabili ty of spectroscopic measurement results depends not only on the in strument used but also on the status of sample. Therefore, careful sample prepar ation and handling are indispensable. The measured results require correction before calibration. The necessary correction factors for the Integrated UV-VIS MAMW spectrome ter include: refraction, scattering volume and data normalization. Procedur es to implement these correc tion factors are provided. In addition, the effect of the background UV light and data representation strategies are discussed. Quantitative validation of m easurement results requires calibration. Considering the experimental set up and meas urement purpose, a relative calibration is considered desirable for the Integrated UV-VIS MAMW spectrometer. Procedures necessary to accomplish the relative calibrati on are described in the calibration section. 5.1 Sample The reliability and reproducibility of any spectroscopic measurement results depend not only on the instrumentation but also on th e reproducibility of th e sample preparation protocol. In this research, the feasibility of the Integrated UV-VIS MAMW spectrometer

PAGE 133

110 to characterize micron and sub-micron size particles was tested using commercially available polystyrene standards. Therefore, this section provides information related to the polystyrene standards and the sample prep aration protocol. In addition to polystyrene, whole blood sample was chosen as an example of application of UV-VIS MAMW spectroscopy to biol ogical systems. 5.1.1 Polystyrene Standards Polystyrene standards with different sizes or compositions were selected for this research because their optical propertie s are well known, and are actively used as calibration standards for instruments measuring light scattering.28 Table 5.1. lists the data reported by the manufactures for the polystyrene standards used in this dissertation. The optical property of polystyrene is shown in Fig. 5.1.26 Both real and imaginary parts of refractive index have a resonant band belo w 230 nm. Therefore, polystyrene strongly absorbs, reflects and refracts deep UV light. Th ereafter, the real part of refractive index, Fig.5.1(a), decays exponentially until 500 nm and becomes stable at wavelengths longer than 500 nm. The imaginary part has slight values only in the wa velengths between 230 nm and 280 nm as exhibited in Fig.5.1(b). Due to the resonant absorption of polystyrene, the measured UV-VIS MAMW spectra have we ak intensities at wavelengths below 230 nm. As a result, the measured UV-VIS MAMW spectra were plotted for the wavelengths longer than 230 nm.

PAGE 134

111 Table 5.1. List of polystyrene standards used for this dissertation research No Standard Manufacturer Catalog No. Mean Diameter 1 Polystyrene 20 nm sphere Duke Scientific Corporation 3020A 20nm 2.0 nm 2 Polystyrene 150 nm sphereDuke Scientific Corporation 3150A 150nm .0 nm 3 Polystyrene 300 nm sphereDuke Scientific Corporation 3300A 300nm 5.0 nm 4 Polystyrene 500 nm sphereDuke Scientific Corporation 3500A 499nm 5.0 nm 5 Polystyrene 1.0 m sphere Duke Scientific Corporation 4010A 1.020 m 0.022 m 6 Polystyrene 3.0 m sphere Duke Scientific Corporation 4203A 3.063 m 0.027 m 7 Polystyrene 4.0 m sphere Duke Scientific Corporation 4204A 4.000 m 0.033 m 8 Polystyrene 8.0 m sphere Duke Scientific Corporation 4208A 7.979 m 0.055 m 9 Polystyrene 10.0 m sphere Duke Scientific Corporation 4210A 10.15 m 0.06 m 10 Green Fluorescent-Dyed Polystyrene 3.0 m sphere Duke Scientific Corporation G0300 N/A 11 Red-Dyed Polystyrene 3.0 m bead Polysciences, Inc 17137 2.923 m 0.105 m 12 Polystyrene 1.87 m sphere Magsphere, Inc.PS1587A* 1.87 m 0.12 m 13 Peanut-shaped Polystyrene Latex 1.85 m standard Magsphere, Inc.PS1305B* 1.85 m 0.53 m Lot number

PAGE 135

112 Figure 5.1. The optical properties of polystyre ne. a) Real part of refractive index. b) Imaginary part of refractive index. Adapted from Reference 26. a b

PAGE 136

113 5.1.2 Sample Preparation Static light scattering theory is based on dilute sample con centration, i.e., enough distance between particle and particle to a void interactions among pa rticles. Therefore, sample preparation is crucially impor tant for the UV-VIS MAMW spectroscopy. Improper sample preparation can easily lead to spurious results due to multiple scattering, aggregation, settling, or particle deposition. For instance, sample aggregation changes the particle size or shape and sett ling affects the concentration. Th erefore, much attention has to be paid for sample preparation. The details of the sample preparation procedures are provided in the protocol for sample prepar ation. Note that sample sonication is implemented before the dilution of polystyrene standards with solvent, after the dilution, and just before the measurement because ultrasonic vibrations can break aggregated particles apart.23 Protocol for Sample Preparation 1. Prepare 0.005% Sodium Dodecyl Sulfat e (SDS) solution as solvent for polymer standards by diluting SDS with deionized (DI) water. 2. Sonicate polymer standards for 30 sec before dilution. 3. Dilute polymer standards with solv ent. The proper concentrations of polymer suspensions vary depending on the sizes of polystyrene standards. They should be determined using the measured UV-VIS transmission spectra. 4. Sonicate the diluted polymer suspensi ons for 3 min. If prepared samples are not used immediately, sonicate them for 3 ~ 5 min just before the measurements.

PAGE 137

114 For the UV-VIS MAMW spectroscopy, an op tical density (OD) = 1 was used as guideline in order to stay w ithin the linear res ponse of the detector Fig. (5.2.a) is an example of variations of the measured opt ical density spectra of polystyrene 0.3 m suspensions depending on sample concentr ation. The measurements were performed using an identical cuvette. As sample con centration was increased, less light entered the detector. Consequently, the measured optical density of the sample was linearly increased within the range of the linearity of the detect or (red, green). Further increment of sample concentration led to the block of short wavelength UV light. This resulted in the observation of noisy signal cause d by detector noise that can be easily appreciated on the measured optical density spectrum (blue). Not only sample concentration but also the sample cell path length affects the measured optical density. For the measurem ent of optical densit y spectra shown in Fig.5.2.b, sample cells with different path length s were used while sample concentrations were kept constant. In case of the octagona l cuvette, its path length is about 2.5 times longer than the fluorescent cuvette. This results in the enhanced optical density proportional to the beam path length ratio (r ed) compared to that measured using the fluorescent cuvette (blue). The use of uncontaminated reference solu tion and very clean cuvette is also important for the UV-VIS MAMW spectroscopy.66 This is especially critical for the measurement of scattering by extremely small or very large polystyr ene particles because of their weak scattering intensitie s in the UV region at low angles.

PAGE 138

115 200 300 400 500 600 700 800 0 0.5 1 1.5 2 2.5 3 3.5 4 Wavelength, nmAbsorbance05-09-04-a S2000 Ps 0.3 Micron Absorbances 200 300 400 500 600 700 800 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Wavelength, nmAbsorbance5-20-04 uv-vis mamw PS 1 Absorbances Blue: FL cuvette Red: Octagonal Cuvette Figure. 5.2. Measured optical density variati on of polystyrene standards in water. a) Optical density variation of polystyrene 0.3 m spheres depending on sample concentration. b) Optical density variation of polystyrene 1.0 m spheres depending on the beam path length. Wavelen g th ( nm ) OpticalDensity Wavelen g th ( nm ) OpticalDensit y a b Sample concentrations(au): Red< Green< Blue Blue:Rectangular cuvette Red : Octa g onalcuvette

PAGE 139

116 5.2. Measurement One of the merits of the Integrated UV-VIS MAMW spectrometer is that one single instrument can meas ure scattering, optical dens ity or transmission, and fluorescence for particle charac terization. Although these are to tally different branches of spectroscopy, their measurement requires comm on procedures measurement of sample and reference signal intensity for scattering and fluorescence and additional dark current measurement for optical density. However, da ta processing requires different procedures. In this section, signal analysis and m easurement procedures are described. 5.2.1. Signal Analysis When incident light irradiates a sample in the sample cell, light absorbed or scattered by the sample or fluorescence light from the sample can be detected by the spectrometer and yields signals to be displayed on a monitoring system. If ) ( osI designates observed sample signal intensity due to light from the sample detected by the spectrometer located at angle then ) ( osI can be written as65 ) ( ) ( ) ( ) ( s s ts osB D I I (5.1) where ) ( sB and ) ( sD represent background light intensit y and dark current recorded during the sample signal measurement and ) ( tsI is the true sample signal intensity. Likewise, the reference signal intensity due to reference solution at observation angle

PAGE 140

117 becomes65 ) ( ) ( ) ( ) ( r r tr orB D I I (5.2) where ) ( and ), ( ), ( ), ( tr r r orI D B I correspond to terms in Eq. (5.1). The sample scattering signal intensity ) ( scaI can be found by subtracting the observed reference signal intensity from the observed sample signal intensity:65 ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( r s r s tr ts or os scaB B D D I I I I I (5.3) The terms in the second and the third parent heses can be neglected; the dark current terms depend only on the integration time and the electronic noise of S2000 production spectrometer that is very stable and the bac kground light terms are very small and nearly constant because the measurements are pe rformed in a dark room. In addition, any background light sources like the lamp indica tor light are blocked and the detection optics are covered to minimize background lig ht entering the detector. Fig. 5.3. shows reference solution signal inte nsities measured from 20o to 40o in observation angle with 5o resolution. These intensities are scaled by multiplying the measured intensities by the appropriate integration time ra tio. At angles larger than 20o, the scattering intensities resulting from the reference solution are neglig ible and the detected signal intensities are

PAGE 141

118 primarily due to the dark current. Alt hough the measurements were performed at different observation angles with different integration time, th e resulting reference signal intensities which are primarily due to the da rk current and the bac kground light are highly reproducible and very stable. Consequently, the difference between the observed sample signal intensity and the observed referen ce signal intensity yields an excellent approximation to the true sample signal intensity ) ( sI ) ( ) ( tr tsI I (5.4) The terms ) ( and ) ( tr tsI I include correction factors due to the spectrometer set up such as reflection, refraction, etc. These terms can be corrected after the measurement. In the case of transmission, a separate measurem ent of the dark current intensity at 0o is required because the dark current is the primary source of noise. 5.2.2. Measurement Procedures For the measurement of the joint particle property distribution, sample quality and a clean cuvette are as important as the instrument and the experimental set up. Therefore, the cuvette needs to be washed thoroughly to minimize reflection from the cuvette wall due to contamination and the quality of the standard needs to be confirmed by measuring the optical density with the reference spectrometers HP8452 or HP 8453 spectrometer before any UV-VIS MAMW spectra measurements. Once the sample has been prepared,

PAGE 142

119 250 300 350 400 450 500 550 600 650 700 750 800 0 10 20 30 40 50 60 70 80 90 100 s5m references 2 Blue:20 deg, Green:25 deg, Red: 30 deg Black: 35 deg, Magenta: 40 deg Figure 5.3. Reference solution signal intensities measured from 20o to 40o with 5o resolutions in observation angle. Signal intensities were adjusted for the integration time used for the measurement. Wavelength (nm) Counts

PAGE 143

120 turn on the DT-1000 UV-VIS li ght source and wait for about 40 minutes to warm up the lamps. Then the incident UV-VIS beam spect rum needs to be measured using the scope mode of OOIBase32, S2000 spectrometer opera ting software, and compared to the UVVIS beam spectrum that was recorded during the optics alignment for the Integrated UVVIS MAMW spectrometer. The optics alignmen t procedure is described in Section 4.4.2. If both UV-VIS beam spectra are not consiste nt, then optics alignment should be checked. After lamp warm up and the confirmation of the incident UV-VIS beam spectrum, the optical density needs to be measured and recorded first. As stated in the chapter 4, two quartz cuve ttes, rectangular fluor escent cuvette and octagonal cuvette, were used for the UVVIS MAMW spectra measurements. Their shapes and relevant correction factors are described in Section 4.2.4 and the procedures of implementing the required corrections are discussed in Section 5.3. If the rectangular fluorescent cuvette is used, scattering measurements are performed up to 35o in observation angle due to the refractio n caused by the shape of the cuvette On the other hand, if the octagonal cuvette is used for the measurement, the upper limit of the measurable scattering angle is not restricted by sample cell shape. For scattering measurement, integration times of up to 5000 msec have been used as needed depending on the sample and scat tering angle. Changing integration time requires time for the signal to re-stabili ze prior to take sample spectra. For each measurement, the spectrometer settings for si gnal averaging include an average sample number of 5 and an average boxcar number of 10. These numbers yield clean signals

PAGE 144

121 while minimizing measurement time. The detail s of the average sample number and the average boxcar number are explained in Section 4.2.6 The total measurement time depends on the integration time. For 5000 msec integration time, the signal is recorded every 25 seconds. Due to signal stabilizati on and data recording the overall scattering measurements take 30-40 minutes, 15-20 minut es for sample scattering measurement and additional 15-20 minutes for reference sca ttering measurement. Typically, reference scattering measurement was performed after sample scattering measurement because it requires the information of integration time used for sample scattering measurement. Reference scattering measurement requires thorough washing of cuvette to minimize scattering due to particles remaining in the cuvette wall. In the case of fluorescence, measurements at 90o are repeated after changing the optics set up. The details of the measurement procedures are summarized in the protocol for measurement. Protocol for Measurement 1. Measure the optical density of samp le using HP 8452 or HP 8453 reference spectrometer to check the sample quality. 2. Turn on the DT-1000 UV-VIS light source and wait for lamp warm up. It takes 20 min for tungsten-halogen lamp and 40 min for deuterium lamp. 3. Measure the incident UV-VIS beam spectru m at 0 deg using the scope mode of OOIBase32 software. Use the overlay f unction of OOIBase32 software and ensure that the measured UV-VIS beam spectrum is consistent with the UV-VIS

PAGE 145

122 beam spectrum recorded during the optics a lignment of the UV-VIS MAMW spectrometer. If both spectr a are not consistent, check the optics alignment. 4. Block the incident UV-VIS beam a nd measure the dark current at 0o using the scope mode. Remove the block after the measurement. 5. Fill the cuvette with solvent and measure the reference spectrum at 0o using the scope mode. 6. Fill the cuvette with sample and measure the sample spectrum at 0o using the scope mode. Record the optical density as experiment file. 7. Measure the sample scattering from 5o to 35 o in 5o increment while varying integration time as needed. Use the scope mode of OOIBase32 software for this and all the remaining measurements. 8. Fill the cuvette with solvent and meas ure the reference scattering from 5o to 35 o in 5o increment while varying integrati on time as needed. For both sample scattering and reference scattering measurem ents, the same integration time must be used at the corresponding angles. 9. Measure the fluorescence intensity and the reference intensity at 90o in case of fluorescence measurement. 10. Turn off the DT-1000 UV-VIS light s ource and record th e operation hours. 5.3. Correction Factors This section describes the procedures fo r implementing the necessary corrections for the measured UV-VIS MAMW spectra. Th e correction factors include reflection,

PAGE 146

123 refraction, scattering path length and scat tering volume corrections. Among them reflection, refraction and scatte ring path length corrections are closely related to the shape of the sample cell used for the scattering measurement as discussed in the section 4.2.4. In case of scattering volume correction, it is inevitable rega rdless the type of sample cell used for scattering measuremen t because the scattering volume continuously changes as the goniometer arm rotates through th e angles of interest. In addition to the above correction factors, the effect of s lit structure and slit-to-cuvette distance on the measured UV-VIS MAMW spectra is discussed. Also the correlation between the shapes of the incident UV-VIS beam spectra used fo r the data normalizati on and the shapes of the resulting UV-VIS MAMW spectra is examined. 5.3.1 Reflection Correction Differences in refractive indices across the interfaces between cuvette wall and liquid and cuvette wall and air result in the reflection of scattered light. Therefore, a reflection correction is required to compensate for scatte ring due to the reflected light. Reflection at the interface of cuvette wall and ai r is much larger than that of liquid and cuvette wall because the refract ive index of air is 1. Accord ing to Tommimatsu & Palmer, 57 the required reflection correction depends on the type of sample cell and it differs depending on the shape, material, thickness, et c. of the sample cell. Fig.5.4 illustrates the scattering geometry for the explanation of the reflected light contributions in a conventional cylindrical cell.51 In this case, ) (detscaI is the intensity of scattered light arriving at the detector, ) ( scaI is the intensity of the directly scattered light and

PAGE 147

124 ) 180 (oscaI is the intensity of the scattered light due to the reflection of the incident beam51 ) 180 ( ) 2 exp( ) ( ) (o det sca sc p r sca scaI r R f I I (5.5) where the coefficient rf compensates all deviations from the ideal reflection model like the contribution of the light that is reflected more than once, pR is Fresnels reflectivity coefficient for normal incidence, and ) 2 exp(scr is the attenuation of the scattered light due to reflection. Here, represents the absorption coe fficient of sample solution and scr is the radius of the sample cell.51 The Fresnels reflectivity coefficient for normal incidence pR is expressed as51 2 2 1 2 1 i i i i pn n n n R (5.6) where 2 1andi in nare the refractive indices of the material on both sides of the interface for the wavelength of the incident light. For rectangular cuvette, pRneeds to be replaced by the Fresnel formulas for reflection of light obliquely incident on a plane boundary.7 For the Integrated UV-VIS MAMW spectro meter, reflection corrections require measurement of scattered light at observation angles greater than 140o that is not feasible with the current setup. Nevertheless, reflect ion correction can be neglected for forward

PAGE 148

125 scattering because reflection does not cause appr eciable differences in the intensity ratio until after 90o.53 Reflection correction is indispen sable for large angle scattering (backward direction) where the reflected light intensity is comparable to the scattered light intensity coming from the scattering volume.51 Given these consid erations reflection corrections were not implemented in this research. 5.3.2 Refraction Correction The geometry of the rectangular cuvette requires refraction correction because the angle between the cuvette wall and the scattered light is not perpendicular. Fig.5.5 illustrates the refraction of the scattered light due to the refractive index differences at interfaces between sample suspension and th e cuvette wall and between the cuvette wall and air.2 The scattering angle can be obtained by applying the Snells law consecutively to the above interfaces q q sn n sin sin (5.7) a a q qn n sin sin (5.8) where a q sn n nand ,are the refractive indi ces of sample suspensi on, cuvette material, and air, respectively, at the wavelength of interest and a q and represent

PAGE 149

126 ) ( Is det sca Ii: incident light ) (detscaI: scattered light arri ving at the detector : scattering angle rf: (180o ) Figure 5.4. Scattering geometry explaining the reflected light contributions in case of conventional cylindrical cell. Adapted from Reference 51.

PAGE 150

127 ns: refractive index of sample suspen sion at the wavelength of interest nq: refractive index of cuvette material at the wavelength of interest na: refractive index of air : scattering angle q: refracted angle at the interface of sample suspension and cuvette wall a: refracted angle at the inte rface of cuvette wall and air. Figure 5.5. Diagram explaining the geometry of scattered-light refraction. For illustration purpose, the horizontal side of the cuvette is magnified. Adapted from Reference 2.

PAGE 151

128 angle, refracted angle at the interface of sample suspension and cuvette wall, and refracted angle at the interface of cuvette wall and air, respectively. Actually, a is the observation angle and an= 1. The incident UV-VIS beam is multiwavele ngth. Therefore, the exact refraction and scattering volume corrections require the use of the con tinuous refractive indices of cuvette material, Spectrosil Quartz, at all the wavelengths from 230 nm to 800 nm. However, the examination of the refraction as well as the scattering volume correction results acquired using the refr active indices of Spectrosil Quartz at 200nm, 254nm, and 800 nm show that the complicated corrections using the continuous refractive indices of Spectrosil Quartz can be avoided if a proper refractive index is selected for the corrections. The analysis of the above correction result s as well as the procedure obtaining the unknown refractiv e indices of Spectrosil Quartz are provided in Appendix B. For the refraction correction, qn= 1.506 at 254 nm that is the median value of the manufacturer-provided si x different refractive indices of the Spectrosil quartz was chosen.54 As for the refractive index of polystyre ne standards, the refractive index of water at 254 nm, 1.3716, was selected because polystyrene standards diluted in water were used for measurements and the concentrat ions of the standards were thin. Table 5.2 shows the results of refr action corrections obtained for the angles from 5o to 40o with 5o resolution. The corrected scattering angles are approximated as 4o, 7o, 11o, 14o, 18o, 21o, 25o and 28o, respectively.

PAGE 152

129 Table 5.2. Differences in observation a ngle and refractive-index-corrected actual scattering angle. Observation angle a Angle refracted by the cuvette wall q Corrected scattering angle 5o 3.3o 3.6o 10o 6.6o 7.2o 15o 9.9o 10.9o 20o 13.1o 14.4o 25o 16.3o 17.9o 30o 19.4o 21.4o 35o 22.4o 24.7o 40o 25.3o 28.0o

PAGE 153

130 Fig.5.6 is the UV-VIS MAMW response surface of polystyrene 1.0 m standard in water. Fig.5.6.a is plotted for the observ ation angle and Fig.5.6.b is plotted for the refraction corrected scattering angle. Both response surfaces are identical in shape, intensity and wavelengths but refraction corrected Fig.5.6.b is angularly contracted as expected. Fig.5.7 is the contour plot of the UV-VIS MAMW response surface of polystyrene 1.0 m standard in water. Fig.5.7.a is th e contour plot before the refraction correction and Fig.5.7.b is the contour plot after the refrac tion correction. These contour plots again confirm the angular contraction by refraction correction. Fig.5.8.a is the simulated plot of the UV-VIS MAMW response surface of polystyrene 1.0 m standard in water and Fig.5.8.b is the corresponding contour plot. The comparison of the UV-VIS MAMW response surface before the refraction correction with that after the refraction correction shows that the re fraction-corrected UV-VIS MAMW response surface fits better to the theore tically-simulated UV-VIS MAMW response surface. Therefore, experimental results acquire d using a rectangular cuvett e have been corrected for refraction. 5.3.3 Scattering Path Length Correction For rectangular cuvette, the scattering path length increases as the scattering angle increases until the scattered light passes the co rners of the cuvette where the faces of the cuvette parallel to the incident light meet the 1st face of the cuvette where the forward scattered light passes. Consequently, more scat tered light can be absorbed by the sample

PAGE 154

131 Figure 5.6. UV-VIS MAMW respons e surfaces of polystyrene 1.0 m standard in water. a) Plotted for the observation angle of 5o 40o. b) Plotted for the refraction-corrected scattering angle of 4o 28o. Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) b Wavelength (nm) Observation Angle (deg) Log (Isca/Ii) a

PAGE 155

132 Figure 5.7. Contour plots of the UV-VIS MAMW response surfaces of polystyrene 1.0 m standard in water, Fig.5.6. a) Plotted for the obser vation angle of 5o 40o. b) Plotted for the refraction corrected scattering angle of 4o 28o. Observation Angle (deg) Wavelength (nm) a Scattering Angle (deg) Wavelength (nm) b

PAGE 156

133 Figure 5.8. Simulated plots of UV-VI S MAMW spectra of polystyrene 1.0 m standard in water. a) Simulated pl ot of the UV-VIS MAMW response surface for the angular range of 4o 29o. b) Corresponding contour plot. Wavelength (nm) Scattering Angle (deg) a Scattering Angle (deg) Wavelength (nm) b Log (Isca/Ii)

PAGE 157

134 suspension at large scattering angle than at low scattering angle until the scattered light reaches the edge of the 1st face. Therefore, scattering path length correction is inevitable. Fig.5.9 shows the geometry of the scat tering path length variation as a function of scattering angle in case of the rectangular cu vette. The reduced scattered light intensity rscaI at scattering angle is sl sca rscae I I ) ( ) (0 (5.9) where ) (0 scaI is the scattered light intensity at o0, is the absorption coefficient and sl is the scattering path length increment. sl can be rewritten as a function of scattering angle h sd l ) 1 cos 1 ( (5.10) where hd is half the inside path length of the rectangular cuvette. If Eq. (5.10) and Eq. (2.56) are substituted into Eq. (5.9), then hd k sca rscae I I) 1 cos 1 ( 4) ( ) ( (5.11)

PAGE 158

135 For polystyrene, the imaginary part of the refractive index k is 0.04 at 230 nm, 0.006 at 250 nm, and 0 at wavelengths longe r than 283 nm. Therefore, scattering path length correction may be ignored in case of polystyrene standa rds. For other standards, scattering path length correct ion requires the information of optical properties of the sample. If this information is not availabl e, the octagonal cuvette, which has virtually identical scattering path length at every angle, can be used to avoid the sc attering path length correction. 5.3.4 Scattering Volume Correction The scattering volume is defined as the inte rsection of the incident beam with the detector field of view that continuously ch anges as the goniometer arm rotates as shown in Fig. 5.10.7 Therefore, it is inevitable to correct the measured scattered light intensity using a proper correction factor. Typically the scattering volu me correction is done by multiplying sin by the measured scattered light intensity at the corresponding scattering angle.7 However, as Wyatt pointed out, the simple multiplication of sin may become a new source of error, espe cially, at low scattering angles.66 This seems to originate from the fact that the value of sin at every scattering angle is always constant for that angle regardless the spectrometer setup while the actual scattering volume at every scattering angle varies depending on the sa mple cell path length, the incident beam height, and the width of the slits. For instance, in case of the current UV-VIS MAMW spectrometer setup that uses the slits with 6 mm width and the rect angular fluorescent

PAGE 159

136 hd : half the inside path length of rectangular cuvette sl : scattering path length increment Figure 5.9. Diagram illustrating the necessity of the scattering path length correction in case of a rectangular cuvette.

PAGE 160

137 F1, F2: Possible polarization filters : scattering angle Figure 5.10. Schematic explaining scattering vo lume changes. Cited from Reference 7.

PAGE 161

138 cuvette with 10 mm path length, the scat tering volume has the maximum 6 x 10 x 20 mm3 (1200 mm3) at 0o and the minimum 6 x 6 x 20 mm3 (720 mm3) at 90o. Here 20 mm represents the approximate height of the incident UV-VIS beam. The resulting maximum to minimum scattering volume ratio is 1.67. On the other hand, if the slit widths are changed into 3mm, the maximum and the minimum scattering volume become 600 mm3 and 180 mm3, respectively. The corresponding maxi mum to minimum s cattering volume ratio increases to 3.33. Furthermor e, in addition to the constant sin regardless the spectrometer setup, the value of sin approaches 0 as scattering angle is nearing 0o and thus, results in the reduced scattered light in tensity at low angles. Therefore, it can be concluded that the scattering volume changes may not be represented as sin, especially at low angles if the incident beam and the de tector field of view have some widths and the sample cell path length is not large. Consequently, the scattering volume correction should be done not by the univers al application of the simple sin multiplication but by multiplying the ratio of the scattering volume change at each scattering angle that is calculated from the actual experimental setup. Hereafter, two different terms, sine correction and calculatedvolume correction, will be used to differentiate the scattering volume correction done by multiplying sin and that performed using the calculated scattering volume. Fig. 5.11 shows a diagram used to calc ulate the approximate scattering volume changes for the Integrated UV-VIS MAMW spectrometer. For the calculation of the scattering volume, the refraction-corrected sc attering angles in Table 5.2 are used.

PAGE 162

139 Figure 5.11. Schematic of geometry us ed for scattering volume correction.

PAGE 163

140 The results are summarized in Table 5.3 in case of the 3 mm width slit usage for the UVVIS MAMW spectrometer setup and Table 5.4 fo r the 6 mm width slit installation case. The multiplication factors are obtained by dividing the scattering volume at each scattering angle into the scattering volume at 4o. These results are used to correct the scattering volume when plotting the UV-VI S MAMW response surfaces. Fig. 5.12 shows the UV-VIS MAMW response surfaces of polystyrene 1.0 m standard in water. Fig. 5.12.a is the plot wit hout scattering volume correction and Fig. 5.12.b is the case when sinis used for correction. Fig. 5.12.c shows the results of using the multiplicative factors in Table 5.4 and Fig. 5.12.d is the si mulated response surface. In case of the sinecorrected Fig. 5.12.b, the scattering intens ity at low angles is decreased while the scattering intensity at large a ngles is enhanced compared to Fig. 5.12.a, the case without correction. On the other hand, the calculated -volume-corrected Fig. 5.12.c shows similar profile to Fig. 5.12.a. Compared with the simulated plot Fig. 5.12.d, the calculatedvolume correction matches well at low angles while the sine correction fits better at large angles in case of the polystyrene 1.0 m standard in water. Fig. 5.13 shows the UV-VIS MAMW response surfaces of polystyrene 8.0 m standard in water. Fi g. 5.13.a is the plot without scattering volume correction and Fig. 5.13.b is the plot obtained by multiplying sin. Fig. 5.13.c is the calculated-volume -corrected plot obtained using the multiplicative factors in Table 5.4 and Fig. 5.13.d shows the simulated response surface. Note that the sine-corrected Fig. 5.13.b ma tches better to the simu lated plot Fig. 5.13.d than the calculated-volume-corrected Fi g.13.c in case of the polystyrene 8.0 m standard.

PAGE 164

141 Table 5.3. Calculated scatteri ng volume in case of the 3 mm wi dth slit installation for the Integrated UV-VIS MAMW spectrometer setup. Observation Angle a (deg) Scattering Angle (deg) Scattering Volume (mm3) Multiplicative Factor 5 4 560 1.0 10 7 520 1.08 15 11 500 1.12 20 14 480 1.17 25 18 450 1.24 30 21 400 1.40 35 25 370 1.51 Table 5.4. Calculated scatteri ng volume in case of the 6 mm wi dth slit installation for the Integrated UV-VIS MAMW spectrometer setup. Observation Angle a (deg) Scattering Angle (deg) Scattering Volume (mm3) Multiplicative Factor 5 4 1150 1.0 10 7 1120 1.03 15 11 1070 1.07 20 14 1030 1.12 25 18 1020 1.13 30 21 1010 1.14 35 25 970 1.19

PAGE 165

142 Figure 5.12. UV-VIS MAMW respons e surfaces of polystyrene 1.0 m standard in water plotted for the demonstration of the scat tering volume correction. a) UV-VIS MAMW response surface without scattering volume correction. b) UV-VIS MAMW response surface with the sine correction. Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) a Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) b

PAGE 166

143 Figure 5.12. (Continued). c) UV-VIS MAMW response surf ace with the scattering volume correction. d) Simulated UV-VIS MAMW response surface. Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) a Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) b

PAGE 167

144 Figure 5.13. Scattering volume correct ed UV-VIS MAMW response surface of polystyrene 8 m standard in water plotted for the de monstration of the scattering volume correction. a) UV-VIS MAMW response surface without scattering volume correction. b) UV-VIS MAMW response surface w ith the sine correction. Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) a Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) b

PAGE 168

145 Figure 5.13. (Continued). c) UV-VIS MAMW response surface when scattering volume correction is done. d) Simulate d UV-VIS MAMW response surface. Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) a Wavelength (nm) Scattering Angle (deg) Log (Isca/Ii) b

PAGE 169

146 Fig. 5.14 shows the effect of the scatteri ng volume correction. Fig. 5.14.a shows the effect of the scattering volume correction for the polystyrene 1.0 m standard in water as a function of the observation a ngle at a wavelength of 633 nm, and Fig. 5.14.b shows the effect of scattering volume correction for the polystyrene 8.0 m standard in water as a function of the observation angle at a wave length of 633 nm. For the polystyrene 1.0 m standard, the sine correction does not fit at low angles. For the polystyrene 8.0 m standard, both the calculated-volume correcti on and the sine correc tion show a decreased scattering intensity at low angles. The reas ons why the calculated-volume correction does not fit well at large angles or for large particles compared with the sine correction are not clear. Nevertheless, for UV-VIS MAMW sca ttering, the calculated-volume correction is adopted as the best available approximati on because the scattering volume change depends on individual spectrometer set up and scattering cell geomet ry and a correction will have to be applied every time. The unconditional use of the sine correction, especially at low angle scat tering, should be avoided. Fu rther study is recommended to elucidate the best scattering volume co rrection, especially, at low angles. 5.3.5 Effect of Slit Width and Location Refraction, reflection from the cuvette wa ll, and scattering volume are the main correction factors for the light scattering experiment. However, for the Integrated UVVIS MAMW spectrometer, the stray reflection of UV light from the slit edge and the slit mount in front of the cuvette should also be considered.

PAGE 170

147 4 6 8 10 12 14 16 18 20 22 24 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 Angle, deg Ps1 micron MAMW crossection at 633 nm Blue: Exp, No scattering volume correction Red: Exp, Sine correction Green: Exp, Volume correction Black: Simulation Figure 5.14. Effect of scattering volum e correction on the UV-VIS MAMW response surfaces as a function of scattering angl e and a constant wavelength of 633 nm. a) polystyrene 1.0 m standard in water. The sine co rrection yields d ecreased scattered light intensity at low angles. a

PAGE 171

148 4 6 8 10 12 14 16 18 20 22 24 0 0.05 0.1 0.15 0.2 0.25 Angle, deg Ps8 micron MAMW crossection at 633 nm Blue: Exp, No scattering volume correction Red: Exp, Sine correction Green: Exp, Volume correction, 6 mm slit width Black: Simulation Figure 5.14. Continued. b) polystyrene 8.0 m standard in water: Both the calculatedvolume correction and the sine correction yield decreased scattered li ght intensity at low angles. Scattering Angle (deg) b

PAGE 172

149 At first, the location of the slit in front of the cuvette was made as close as possible to the cuvette in order to minimize any diffraction of UV light from the slit. The possibility of the UV-VIS light reflection from slit edge and slit mount due to the close distance between the slit and the cuvette wa s ignored because the use of narrow width slits that can minimize the errors caused by stray reflection.53 However, the change in the slit-to-cuvette distance became inevitable during the last stage of research due to the following observations: 1) reference scattering measurement at 5o with a few thousand milliseconds integration time that was far higher than the typical integration time used at this angle showed strong scattered light intensity in the UV region and 2) no strong UV light intensity was recorded if this measurement was repeated without the cuvette. The source of this background UV light was inve stigated. Among the possibilities, part of the incident UV-VIS beam became the sour ce of the background UV light was excluded because: 1) during the spectrometer set up the incident UV-VIS beam path was traced using a white index card to ensure that no part of the incident beam or background light enters the slit in front of the objective lens if the observation angles are equal to or greater than 5o and

PAGE 173

150 2) if part of the incident UV-VIS beam is refracted and enters the detector, there should be signal saturation due to its st rong intensity and the large integration time. However, no such signal saturation was observed. The fact that no backgr ound UV light was recorded without the cuvette attributed the source of this background UV light to the stra y reflection from the slit edge and the slit mount. The reason can be easily explained usi ng Fig.5.15, the diagrams illustrating the source of the background UV light. If there is no cuvette, there is no refraction of any UV light as shown in Fig.5.15.a and thus, result s in no strong background UV light detection. On the other hand, if there is a cuvette, part of the incident UV beam is reflected by the cuvette wall and then reflected agai n by the slit edge or the slit mount. The stray UV light reflected from the slit edge or the slit mount incidents on th e cuvette obliquely. Then it is refracted by the cuvette wall as shown in Fig.5.15b. The short wavelength UV light is refracted more than the long wavelength UV light because of the large refractive index differences between spectrosil quartz and air at deep UV light wavelength. This reflectedthen-refracted UV light becomes th e source of the background UV light. Methods to prevent the reflected UV light from entering the receiving optics were considered. These include: 1) reduction of the width of bot h slits from 6 mm to 3 mm, 2) changing the location of the slit in front of the cuvette. It was relocated at the position 5 cm away from the sample cell. Howe ver, it should be noted that the slit

PAGE 174

151 Figure 5.15. Diagram explaining the source of background UV light. a) Without cuvette, no background UV light ente rs the receiving optics. b) With cuvette, reflected-thenrefracted background UV light enters the receiving optics.

PAGE 175

152 in front of the cuvette cannot be lo cated too far away from the cuvette because of the UV beam divergence after passing the slit and 3) As shown in Fig.5.16, changing the slit stru cture from single block (Fig.5.16.a) to double block and erecting a side wall adjacen t to the slit entrance on the incident beam path side (Fig.5.16.b) In addition, thorough cleaning of the cuvette after every sample change reduced further the amount of background UV light. The cleanne ss of the cuvette can be assessed using the transmission spectrum of deionized wate r. Implementation of these methods ensured the systematic reduction of the backgr ound UV light and the reproducibility of the measurement results. As a result, the measurement of the UV-VIS MAMW spectra of polystyrene 20 nm spheres that is the mi nimum size standard available from Duke Scientific became possible. Previously, th e measurement of the UV-VIS MAMW spectra of polystyrene spheres less than 150 nm wa s impossible due to the strong background UV light at low angles as show n in Fig.5.17. Besides, Fig.5.18, the measured UV-VIS MAMW spectra of polystyrene 300 nm spheres, showed peculiar features in the short wavelength UV region at low observation angles. It should be noted that slit structures and slit-to-cuvette distances used for th e measurement of the UV-VIS MAMW spectra presented in this dissertation were varied depending on the samples. Slit widths and slitto-cuvette distances used for the measur ement of each polystyrene standard are summarized in the Table 5.5. The changes in the slit structure and the slit-to-cuvette distance depending on the samples resulted in the lack of the consistency of the setup

PAGE 176

153 Figure 5.16. Modified slit structure for the prevention of the background UV light. a) Before the modification. b) After the m odification Double bloc k and side block are added. Slit Razor Blade Ii Slit Razor Blade Ii Side Block Double Block a b

PAGE 177

154 250 300 350 400 450 500 550 600 650 700 750 800 -200 -150 -100 -50 0 50 100 Ps 0.15 MAMW Scattering Aspects 1 B: 5 deg G: 10 deg Red: 15 deg Black: 20 Deg Q:Large integration time? Figure 5.17. UV-VIS MAMW scattering profil es of polystyrene 150 nm spheres at observation angles from 5o to 20o. Note the negative scattered light intensity in case of 5o. Wavelength (nm) Counts

PAGE 178

155 Figure 5.18. Measured UV-VIS MAMW spectra of polystyrene 300 nm spheres in water. a) Measured with the background UV light. b) Measured after the elimination of the background UV light. a b

PAGE 179

156 used for the measurement. Therefore, it is inev itable to investigate to what extent the loss of the consistency of the setup is acceptable for the UV-VIS MAMW spectra measurement. First of all, it should be noted that re gardless the slit structure and the slit-tocuvette distance only the scattered light that are paraxial or pa rallel to the scattering angle of measurement enters the spectrometer due to the use of lenses. In other words, even if the slit structure and the slit-t o-cuvette distance is changed th e detection of the scattered light that is originated from angles other than the scatteri ng angle of measurement is not probable. On the other hand the amount of the background UV light that enters the detector may vary depending on slit structure and slit-to-cuvette distance. Consequently, if scattered light intensity of the standard is far stronger than the background UV light intensity, the measurement result is not much affected by the setup changes and vice versa. For micron size polystyrene standards (1 m 8 m), their scattered light intensity in the UV region is strong and thus, the background UV light does not much affect the measured results. Fig 5.19 s hows the UV-VIS MAMW response surfaces of polystyrene 4 m standards in water. 3 mm slit width was used for the measurement reported in Fig. 5.19.a and 6 mm slit width wa s used for the measurement reported in Fig. 5.19.b. The results are consistent although slit s with different widths were used for the measurements. The variation in the slit-to-cuvette di stance also has a negligible effect on the measurement results. Fig.5.20 shows the UV-VIS MAMW response surfaces of polystyrene 1.87 m spheres in water. For this measurement, slits with the same width

PAGE 180

157 Table. 5.5. Slit width and distance used for the UV-VIS MAMW spectra measurement Standard Slit width (mm) Distance between slit and cuvette (cm) Polystyrene 20 nm sphere 3 5 Polystyrene 500 nm sphere 3 5 Polystyrene 1.0 m sphere 6 1 Polystyrene 4.0 m sphere 6 1 Polystyrene 8.0 m sphere 6 1 Polystyrene 10.0 m sphere 3 5 Polystyrene 1.87 m sphere 3 5 Peanut-shaped Polystyrene 1.85 m standard 3 5 Polystyrene 3.0 m sphere 6 1 Green fluorescent-dyed Polystyrene 3.0 m sphere 6 1 Whole blood samples (Normal, Sickled) 6 1

PAGE 181

158 reported in Fig. 5.19.a and 6 mm slit width wa s used for the measurement reported in Fig. 5.19.b. The results are consistent although slit s with different widths were used for the measurements. The variation in the slit-to-cuvette di stance also has a negligible effect on the measurement results. Fig.5.20 shows the UV-VIS MAMW response surfaces of polystyrene 1.87 m spheres in water. For this measurement, slits with the same width were used while the slit-to-cuvette distance wa s changed from 1 cm (Fig.5.20 a.) to 5 cm (Fig.5.20 b.). Both results are consistent even if the slit-to-cuvette distance was varied. Therefore, it can be concluded that the change s in the slit structure or the slit-to-cuvette distance do not much affect the measured UV-VIS MAMW spectra of micron-size polystyrene standards th at are smaller than 10 m. On the other hand, sub-micron size polysty rene standards have weak scattered light intensity in the short UV wavelengt h region. Thus, the background UV light can significantly distort the measured results. For polystyrene 10 m standard, it has weak scattered light intensity in the UV region due to the absorption of UV light by free surfactants detached from the polystyrene standards and the diffraction caused by the increased particle size. Therefore, the measurement of the UV-VIS MAMW spectra of polystyrene 10 m standard in the deep UV region is not capable if there is strong background UV light. In conclusion the elimination of the background UV light by changing slit structure and slit-to-cuvette distance is indispensable for the UV-VIS MAMW spectra measurement of sub-micron size polyst yrene standards or polystyrene 10 m standard.

PAGE 182

159 Figure 5.19. Effect of the slit width vari ation on the measured UV-VIS MAMW response surfaces of polystyrene 4.0 m standard in water. For bot h measurements the slit-to cuvette distance was kept constant. a) 3 mm s lit width was used for the measurement. b) 6 mm slit width was used for the measurement. a b

PAGE 183

160 Figure 5.20. Effect of the slit-to cuvett e distance variation on the measured UV-VIS MAMW response surfaces of polystyrene 1.87 m spheres in water. For both measurement, the slit width was kept constant. a) Slit is1 cm away from cuvette. b) Slit is about 5 cm away from cuvette. a b

PAGE 184

161 Further systematic investigation is reco mmended to elucidate the best slit width and the slit-to-cuvette distance for the In tegrated UV-VIS MAMW spectrometer. Micro positioning equipment and modular optics will be required for this study to ensure the reproducibility of th e optics alignment. 5.3.6 Data Normalization Comparison of the results obtained by the theoretical simulation with the experimental measurement necessitates data no rmalization to remove the effects of the sample concentration and/or particle number variation.2 In the case of optical density, normalization can be accomplished by dividi ng the measured spectra by the area under the corresponding optical density curve.1 For UV-VIS MAMW scattering, the criteria or standard for data normalization can be found fr om Eq. (2.34) that requires the use of the incident light intensity to normalize the scattere d light intensity. Ideal ly, the incident light intensity iI should be measured at 0o without the sample cell. However, the use of the sample cell and reference solution decreases the incident light intensity due to reflection, refraction, etc. Therefore, the referen ce UV-VIS beam intensity measured at 0o with the cuvette filled with the reference solution may be chosen as the incident light intensity for the normalization of the scattered light intensity. For the normalization of the UV-VIS MAMW scattering measurement results, not only the incident UV-VIS beam intensity, but the shape of the incident UV-VIS beam spectrum is also important because the measur ed scattered light intensity is normalized not by a single data point but by a mu ltiwavelength light spectrum (i.e., the 570

PAGE 185

162 wavelengths used to plot the UV-VIS MAMW spectra). This requi res rigorous optics alignment and ensuring the reproducibility of the incident UV-VIS beam spectrum. If the shape of the reference beam spectrum is signifi cantly different from that of the incident beam spectrum, the resulting UV-VIS MAMW response surfaces can also be different. Consequently, the reference beam that has the different spectral shape compared with that of the incident beam is not acceptable for the normalization standard. For instance, Fig.5.21.a, the incident UV-VIS beam spectrum measured at 0o, Fig.5.22.a, the reference UV-VIS beam spectrum measured at 0o, and Fig.5.24.a, the reference UV-VIS beam spectrum measured at 5o show similar spectral shapes. Hence, the resulting UV-VIS MAMW response surfaces obtained by normalizi ng the measured scattered light intensity using any one of the above UV-VIS beam spectra are also consistent. This becomes evident from Fig.5.21.b, Fig.5.22.b, and Fig.5.24.b, the UV-VIS MAMW response surfaces of the polystyrene 3 m standard in water acquired using Fig.5.21.a, Fig.5.22.a, and Fig.5.24.a as the normalization standard, respectively. Therefore, any one of these three UV-VIS beam spectra Fig.5.21.a, Fig.5.22.a and Fig.5.24.a seems to be acceptable as the normalization standard. On the other hand, Fig.5.23. a, the incident UV-VIS beam spectrum measured at 3o and Fig.5.25.a, the reference UV-VIS beam spectrum measured at 15o show different spectral sh ape compared with Fig.5.21. a. As a result, the UV-VIS MAMW response surfaces of the polystyrene 3 m standard in water Fig.5.23.b and Fig.5.25.b, obtained using Fig.5.23.a and Fi g.5.25.a as the normalization standard respectively, show different and noisy respons e surfaces. Therefore, it can be concluded

PAGE 186

163 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 3000 Wavelength (nm)Counts Figure 5.21. Incident UV-VIS beam spectrum used for normalization and the resulting UV-VIS MAMW response surface. a) Incident UV-VIS beam spectrum measured at 0o. b) UV-VIS MAMW response surface of the polystyrene 3 m standard in water obtained by normalizing the measured results usi ng the incident UV-VIS beam spectrum. a b

PAGE 187

164 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 Wavelength (nm)Counts Figure 5.22. Reference UV-VIS beam spectrum used for normalizati on and the resulting UV-VIS MAMW response surface. a) Referenc e UV-VIS beam spectrum measured at 0o. b) UV-VIS MAMW response surface of the polystyrene 3 m standard in water obtained by normalizing the measured results usi ng the reference UV-VIS beam spectrum. a b

PAGE 188

165 200 300 400 500 600 700 800 0 500 1000 1500 2000 2500 3000 3500 Wavelength (nm)Counts Figure 5.23. Incident UV-VIS beam spectrum measured at 3o used for normalization and the resulting UV-VIS MAMW response surf ace. a) Incident UV-VIS beam spectrum measured at 3o. b) UV-VIS MAMW response surface of the polystyrene 3 m standard in water obtained by normalizing the measured results using the incident UV-VIS beam spectrum measured at 3o. a b

PAGE 189

166 200 300 400 500 600 700 800 0 20 40 60 80 100 120 Wavelength (nm)Counts Figure 5.24. Reference UV-VIS b eam spectrum measured at 5o used for normalization and the resulting UV-VIS MAMW response surface. a) Reference UV-VIS beam spectrum measured at 5o. b) UV-VIS MAMW response surface of the polystyrene 3 m standard in water obtained by normalizing the measured results using the reference UVVIS beam spectrum measured at 5o. a b

PAGE 190

167 200 300 400 500 600 700 800 10 15 20 25 30 35 40 45 50 55 60 Wavelength (nm)Counts Figure 5.25. Reference UV-VIS b eam spectrum measured at 15o used for normalization and the resulting UV-VIS MAMW response surface. a) Reference UV-VIS beam spectrum measured at 15o. b) UV-VIS MAMW response surface of the polystyrene 3 m standard in water obtained by normalizing the measured results using the reference UVVIS beam spectrum measured at 15o. a b

PAGE 191

168 that the UV-VIS beam spectra shown in Fig. 5.23.a and 5.25.a are not adequate for the normalization standard. Clearly, the features of the measured UV-VIS MAMW response surfaces depend on the shape of the normalization beam spectrum. Therefore, it is necessary to investigate which UV-VIS beam may be used as the nor malization standard. Fig. 5.26 shows an incident UV-VIS beam spectrum recorded with different optics setup; the lens L4 was removed and 3 mm width slits were used. Fig. 5.27 and Fi g.5.28 are the measured UVVIS MAMW response surfaces of polystyrene 1 m spheres and polystyrene 4 m spheres in water, respectively, obtained us ing this set up. Alth ough the shape of this incident UV-VIS beam spectrum is different from that shown in Fig.5.21.a., the corresponding UV-VIS MAMW re sponse surfaces are distinctive for each particle standard and consistent with the results shown in Fig.5.6.a or Fig.5.19.a except the wavelength range at longer than 750nm wher e the incident UV-VIS beam spectrum Fig. 5.26 shows weak intensity. Therefore, it can be concluded that if the incident UV-VIS beam spectrum has enough intensity for the w hole wavelength range, slight changes in the shape of the UV-VIS beam spectrum do not significantly affect the measured UVVIS MAMW response surfaces. Together with the flexibility to choose the normalization standard, this proves that the particle char acterization by the simu ltaneous detection of the JPPD using MAMW spectromete r is a robust technique. At present, calibration of the UV-VIS beam spectrum at wavelengths shorter than 250 nm is impossible.63 For that reason, systematic study to establish the acceptable UVVIS beam spectrum for the Integrated UV-VIS MAMW spectrometer and the

PAGE 192

169 200 300 400 500 600 700 800 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm)Counts Figure 5.26. Incident UV-VIS beam spectrum re corded with different optics setup. Lens L4 was removed and slits with 3 mm width were used.

PAGE 193

170 Figure 5.27. Measured UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water obtained using the UV-VIS beam spectrum in Fig. 5.26 for normalization. Figure 5.28. Measured UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water obtained using the UV-VIS beam spectrum in Fig. 5.26 for normalization.

PAGE 194

171 investigation of the effect of the spectral sh ape changes of the incident UV-VIS beam to the measured UV-VIS MAMW spectra are recommended. 5.4. Data Processing This section describes the data proce ssing procedures for each of scattering, optical density and fluorescence measurements. Once measured data is processed, it requires the adjustment using co rrection factors before plotti ng and further interpretation. All the data processing and pl otting were performed using MATLAB; the code used for data processing and plotting is li sted in Appendix A. To plot the measured results, it is recommended that the spectrometer generated wa velength interval be used directly to avoid possible wavelength shifts and ar tificial smoothing ar ising from MATLAB algorithms. 5.4.1. Optical Density Data Processing Optical density (OD) is measured using the absorbance mode of the Ocean Optics OOIBase32 software that controls the S 2000 production spectrometer. The absorbance mode automatically calculates optical de nsity by utilizing the following equation )) 0 ( ) 0 ( ( )) 0 ( ) 0 ( ( log 1 log D I D I T ODs r (5.12)

PAGE 195

172 where T represents transmission and ) 0 (rI, ) 0 (sI and ) 0 ( D designate the reference signal intensity, the sample signal intensity and the dark current at 0o, respectively. The optical density measured in the absorban ce mode is ready for plotting or further processing. However, for comparison with opt ical density spectra measured using HP 8453 spectrometer the optical densities are normalized by the area under the curve. 5.4.2 Scattering Data Processing Once the scattered light intensity at observation angle is found by subtracting the reference intensity from the scattered light intensity at that angle, the correction factors are multiplied. For the current pr ototype UV-VIS MAMW spectrometer only refraction and scattering volume corrections ar e taken into consider ation while reflection and scattering path length corrections are i gnored as discussed in Section 5.3. Then the integration time ratio, calculated at every scattering angle for each sample, is multiplied. The integration time used for the measurement of UV-VIS MAMW scattering varies depending on sample and scattering angle as described in Section 5.2.2. After data correction and multiplication for the integrati on time ratio, the resulting scattered light intensity is normalized by the reference intensity at 0o. For the UV-VIS MAMW response surface plot, the log value of the normalized s cattered light intensity is used. All these procedures are summarized in Eq. (5.13) ) 0 ( ) ( ) ( ) ( log ) (r r s plot scaI I I C I (5.13)

PAGE 196

173 where ) (plot scaI is the plotted scattered light intensity at scattering angle ) ( C is the correction term taking into acc ount the relevant correction factors and integration time ratio, ) ( scaI is the scattered light intensity, and ) ( rI is the reference light intensity at scattering angle The resulting UV-VIS MAMW response surfaces are plotted for the refraction corrected scattering angle, the observation angle. 5.4.3 Fluorescence Data Processing Data processing for fluorescence simply re quires the subtraction of the measured reference signal intensity from the measured fluorescence signal intensity: ) 90 ( ) 90 (r fl flI I I (5.14) where flI is fluorescence intensity, ) 90 (flI and ) 90 (rI represents fluorescence signal intensity and reference signa l intensity measured at 90o, respectively. In this dissertation research, additional data corrections are not performed in case of fluorescence because the primary purpose of the fluorescence measurement is to demonstrate that the Integrated UV-VIS MAMW spectromete r can perform fluorescence spectroscopy. 5.4.4. Remarks on Measured Result Plotting Using MATLAB For validation purposes, the measured optic al density spectra recorded with the Integrated UV-VIS MAMW spectrometer were compared to the measured optical density spectra obtained using an HP 8453 spectrome ter. Although the spectra recorded from

PAGE 197

174 both spectrometers showed the same features there were unexplaina ble wavelength shifts as shown in Fig. 5.29 where th e optical density spectra of 1 m polystyrene standards in water are shown. If the optical density spect ra measured with another S2000 spectrometer were used for comparison, no appreciable wavele ngth shifts were observe d. At first, these wavelength shifts were attributed to the diffe rent angle of acceptance of the Integrated UV-VIS MAMW spectrometer. As described in Section 4.1.3, the Integrated UV-VIS MAMW spectrometer employs lenses as light delivery tool while the Ocean Optics S2000 spectrometer and accompanying light source are designed to use optical fibers to deliver light. Consequently, the incident beam divergence of the integrated UV-VIS MAMW spectrometer can be different from that of the S2000 spectrometer. This leads to the different angle of acceptance of th e Integrated UV-VIS MAMW spectrometer because the angle of acceptance can be aff ected by the incident beam divergence as described in Section 4.5. The relation between the angle of acceptance of the Integrated UV-VIS MAMW spectrometer and the wavelength shifts of the measured spectra was systematically investigated. However, all th e efforts including the use of a homemade spatial filter that minimizes the incident UV-VIS beam divergence were not successful. Thereafter, MATLAB program that processe s the measured data was re-examined because of the possibility that the wavelength interval generated by mathematical software like MATLAB may not match th at generated by S2000 spectrometer.68 For instance, there are 1674 data points (an array of 1674 CCDs) from 230.04 nm to 800.13 nm in case of S2000 spectrometer and to match these data points,

PAGE 198

175 200 300 400 500 600 700 800 1 1.2 1.4 1.6 1.8 2 2.2 x 10-3 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 5.29. Optical density spectra of polystyrene 1 m standard in water measured with the Integrated UV-VIS MAMW spectrometer (red) and the HP8453 spectrometer (blue). Note the wavelength shifts between two spectra.

PAGE 199

176 the MATLAB code for the wavelength axis of the UV-VIS MAMW response surface was written as x = [230.04: (800.13 230.04)/1 674: 800.13]. However, it was turned out that the resulting wavelengt h resolution does not match that generated by the S2000 spectrometer operating software (OOIBase32) After noticing this, the MATLAB code was rewritten to use the wavelength interval defined by the OOIBase32 software directly; the plots made by new code showed no wavele ngth shifts between the results obtained by the Integrated UV-VIS MAMW spectromete r and HP8453 spectrometer. Fig. 5.30 and Fig. 5.31 show contour plots of the w hole blood UVVIS MAMW response surfaces before and after the wavelength shift correc tion, respectively. The same observation angles were used for both plots. Before th e wavelength shift corre ction (Fig. 5.30), the peak of hemoglobin absorption band was fo rmed at around 400 nm. This does not match with the result obtained using the prot otype MAMW spectrometer (Fig.6.53.a.).2 After the wavelength shift correction, both plots (Fig. 5.31 and Fig.6.53.a.) show consistent hemoglobin absorption peak position at around 415 nm. Care should be taken not to use the wavelength intervals generated by mathemat ical software but to use the spectrometer generated wavelength interval wh en plotting the measured data. 5.5 Calibration Calibration is necessary for the quantitative validation of the joint particle property distribution estimated from UV-VIS MAMW spectra measurement.2 Calibration methods can be divided into absolute calib ration and relative calibration. Due to its

PAGE 200

177 Figure 5.30 Contour plot of the whole blood UV-VIS MAMW response surface before wavelength shift correction. The MATLAB gene rated wavelength interval was used for plotting. Fig. 5.31 Contour plot of the whole bl ood UV-VIS MAMW response surface after the wavelength shift correction. The OOIBase32 soft ware generated wavelength interval was used for plotting.

PAGE 201

178 simplicity, the relative calibration was selected for the Integrated UV-VIS MAMW spectrometer. This type of calibration requires the comparison of the measurement results with the theoretically predicted values for the sample system. The implementation of whole calibration procedures requires preci se simulation program. Our current UV-VIS MAMW simulation program is being upgraded Therefore, only the outline of the calibration procedures for the UV-VIS MAMW s cattering is provided in this section. For this dissertation research, the measured optical density or fluorescence can be validated by direct comparison with the results obtaine d using our reference spectrometer HP8453 spectrometer or the data provided by the manufacturer. 5.5.1. Absolute Calibration Absolute calibration is a method of acquiring calibration constants from the geometry of the apparatus.66 As such, it requires the calculation of all the geometrical factors involved as well as the measurement of the intensity of the incident light on the sample cell. Consideration of the geometrical factors includes accurate determinations of the detector field of view, corrections for refractive indices of the sample cell and the suspending medium, reflection, and estim ation of the scattering volume, etc.66 Absolute calibration is very difficult to perform and time consuming due to the number and the complexity of the factors considered. More over, still some errors can exist and the method is not flexible since it has to be revise d every time the system is modified. Unless absolute scattering values are essential, fo r example, the case of fundamental constant evaluation from light scattering measuremen ts, absolute calibration seems to be

PAGE 202

179 unnecessary. As a consequence, this is not adopted as a calibration method for the integrated UV-VIS MAMW spectrometer. 5.5.2. Relative Calibration Relative calibration compares the resu lts from independently characterized standards with theoretical calculation to establish the calibration constants.28, 66 It is relatively easy to perform and flexible. Ther efore, it is commonly used as a calibration method for light scattering as well as fluorescence spectroscopy.32 A number of relative calibration methods for light s cattering instrument have been developed. These include:28, 66 1. comparison of the scattering from pure liqui ds such as benzene and toluene with literature values, 2. comparison of the angular sca ttering intensity wi th the turbidity (i ntegration of the light scattered overall angl es) for a solution of the non-absorbing Rayleigh point scatterers and 3. comparison of the measured scattering inte nsities of Mie scatterers or suspension of latex particles with the corresponding intensitie s predicted from Mie theory. Latex particles can be ideal Mie scatterers because of their spherical shapes. Methods1 or 2 are not the proper choices be cause the current prot otype integrated UVVIS MAMW spectrometer does not yet have su fficient sensitivity to measure scattering from pure solvents and the measurable angular range is limited. Taking into consideration

PAGE 203

180 the sensitivity and configur ation of the current protot ype Integrated UV-VIS MAMW spectrometer, it is consider ed that Method 3 is the mo st appropriate to use. 5.5.3. Calibration Constants Obtaining the proper calibration constants at relevant scattering angles and multiplying them by the measured scattering in tensities for the sample at every angle completes the whole calibration procedure. In case of the Integrated UV-VIS MAMW spectrometer, the calibration constant at ever y angle is not a single number but an array of numbers corresponding to each of the wa velengths used. Detailed procedures to estimate calibration constant are given as follows 1. Selection of ca libration standards: Choice of proper size standards is impor tant to obtain the correct calibration constants. Monodisperse polysty rene standards are suitable, not only because they are readily available in a large number of sizes, but also because the optical properties of polystyrene are well establishe d as functions of wavelength and the necessary theoretical calculations can be readily pe rformed. For the calibrati on of the integrated UV-VIS MAMW spectrometer, it is recommended to us e the multiple standards in the size range of 20 nm ~ 10 m rather than a single standard.

PAGE 204

181 2. Minimization of undesired effects: Before evaluating the calibration consta nt, it is necessary to minimize any undesired effects by correcting measured results Details of correction factors and data correction procedures for the Integrated UV-VIS MAMW spectrometer are described in the section 5.3. 3. Scattering intensity adjustment: The calculation of calibration constants requ ires the use of the least squares fit for all angles at all wavelengths. However, for simplicity, the approximation methods of estimating calibration constants rather than the rigorous approach es using the least squares fit are provided in this dissertation. The adjusted scattering intensity of standard at the scattering angle and the wavelength can be estimated by subtracting the re ference scattering intensity from the standard scattering intensity at th e corresponding angl e and wavelength: ) ( I ) ( I ) ( Iref dard S meas dard S meas dard S meas Adj , ,tan tan tan (5.15) where ) and , ,tan tan tan ( I ) ( I ), ( Iref dard S meas dard S meas dard S meas Adj represents the adjustedmeasured scattering intensity of the standard the measured scattering intensity of the standard, and the scattering intensity of th e reference recorded during the standard scattering measurement at the scattering angle and the wavelength respectively.

PAGE 205

182 4. Approximate Calibration Constant ) K( The approximate calibration constant ) K( at scattering angle and wavelength can be obtained by calculating the ra tio of the theoretically-calculated scattering intensity of the sta ndard to the adjusted-measured scattering intensity of the standard as shown in Eq. (5.16)24 ) ( I ) ( I Kdard S meas Adj dard S simul Theor , ) (tan tan (5.16) where ) ( Idard S simul Theor,tan represents the theoretically-calculated scattering intensity of the standard at the scattering angle and the wavelength 5. Data calibration Multiplying the adjusted-measured scattering intensity of the sample by the approximate calibration constant completes the calibration of measured results:24 ) ( I ) K( ) ( ISample meas Adj Cal sample , (5.17) ) ( I ) ( I ) ( Iref Sample meas Sample meas Sample meas Adj , (5.18)

PAGE 206

183 where ) ( ICal sample, is the calibrated scattering intensity of the sample and ) ( ISample meas Adj,, ) ( I ) ( Iref Sample meas Sample meas and represents the adjusted-meas ured scattering intensity of the sample, the measured scattering intensity of the sample, the scattering intensity of the reference recorded during the sample scat tering measurement at the scattering angle and the wavelength respectively. The implementation of the complete ca libration procedures requires a simulation program that can predict the JPPD of particles precisely. At present, theoretically calculated results obtained using our cu rrent UV-VIS MAMW simulation program and experimental results do not match well fo r small particles but show UV-VIS MAMW scattering features for large particles. Furthe r work is necessary to elucidate the reason for the discrepancy, especi ally for small particles.

PAGE 207

184 CHAPTER 6: RESULTS AND DISCUSSION The capabil ity of the Integrated UV-VIS MAMW spectrometer to characterize micron and sub-micron size particles by the si multaneous measurement of the JPPD was tested. Optical density, UV-VIS MAMW spect ra, and fluorescence spectra of polystyrene standards including polyst yrene spheres with si zes from 20 nm to 10 m, peanut-shaped non-spherical particles, and green fluorescen t polystyrene spheres were measured. The results demonstrate that the Integrated UVVIS MAMW spectrometer can detect not only particle size, but also shap e and composition information. As a demonstration of potential applicat ions to biological systems, the UV-VIS MAMW spectra of normal whole blood sample and sickled whole blood sample were measured. The results clearly demonstrate th at particle shape and compositional changes can be detected simultaneously by the meas urement of UV-VIS MAMW spectra, and that applications, such as medical diagnosis, where such knowledge is important can be developed with this instrumentation. The analysis of the particle size standa rds measured with the current prototype UV-VIS MAMW spectrometer, together with th e analysis of its optical configuration indicate that the particle characterization capabilities of the Integrated UV-VIS MAMW spectrometer can be further e nhanced. Ways to upgrade the Integrated UV-VIS MAMW

PAGE 208

185 spectrometer are discussed. In addition, methods to add new dimensions to the Integrated UV-VIS MAMW spectrometer are described in this section. The Multidimensional (MD) MAMW spectrometer will pe rform multidimensional spectroscopy measurements simultaneously and thus, maximize the information necessary to completely characterize particle suspensions. 6.1. Measured Optical Density of Polystyrene Spheres The optical density of polystyrene s pheres with sizes of 20 nm, 500 nm, 1 m, 4 m, and 10 m were measured with the Integrat ed UV-VIS MAMW spectrometer. The measured spectra show features of inte rference, reddening, ripple structures, and diffraction as well as absorption due to el ectronic transitions. To validate the MAMW observations, measurements were repeated using the HP 8453 diode array spectrometer and the results obtained from both spectrometers were compared after normalization with the area under curve. Fig. 6.1 shows the normalized optical densit y spectra of polystyrene 20 nm spheres in water. Note that the measured optical density spectra were plotted for the wavelength range of 220 nm to 500 nm. The size of particles is small enough to be described by Rayleigh-Debye-Gans theory. The extinction features below 280 nm are dominated by the absorption of polystyrene. Th e observed spectral features are consistent with the absorption profile of polystyrene provided in Fig.5.1.b. The extinction features below 240 nm are due to absorption by el ectronic transition. Besides absorption, scattering also contributes to the extinction of incident light. As the wavelength of the

PAGE 209

186 250 300 350 400 450 500 0 0.005 0.01 0.015 0.02 0.025 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 6.1 Normalized optical density spectra of polystyrene 20 nm spheres in water. Note that wavelength extends from 220 nm to 500 nm. Normalized Optical Density

PAGE 210

187 incident light increases, appr oximate absorption efficiency a and scattering efficiency 4a decrease. As a result, no more extincti on of incident light is observed at wavelengths longer than 450nm. Fig. 6.2 shows the normalized optical de nsity spectra of polystyrene 500 nm spheres in water. The measured optical dens ity spectra were plotted for the wavelength range of 200 nm to 820 nm. The broadband peak is an interference peak. Interference caused by phase difference between the forward-s cattered light and the incident light that traverses the same physical path outside a sphe re gives rise to inte rference structure: a series of broad, regularly spaced extinction maxima and minima.7 The requirement for interference between the forwar d-scattered and the incident light can be obtained by analyzing the numerators of scattering coefficients na and nb .7 The sudden suppression of the first interference peak near = 230 nm is caused by a resonant absorption band of polystyrene. Another characte ristics of Fig. 6.2 is reddeni ng: a monotonic decrease of extinction with the increasing wavelength of th e incident light. Reddening is caused by enhanced absorption and scattering at the shor ter-wavelength blue li ght than the longerwavelength red light.7 Fig. 6.3 shows the normalized optical density spectra of polystyrene 1.0 m spheres. Compared to Fig. 6.2, it features more interferen ce structures due to the increased particle size. Extinction is frequent ly dominated by scatteri ng if the particle is about the same size as or larger than the wavelength of the incident light. Although it is not shown in the spectra measured by th e Integrated UV-VIS MAMW spectrometer due to insufficient resolution, ripple structures can be seen in the optical density spectrum

PAGE 211

188 200 300 400 500 600 700 800 0 0.5 1 1.5 2 2.5 3 3.5 4 x 10-3 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 6.2 Normalized optical density spectra of polystyrene 500 nm spheres in water. Normalized Optical Density

PAGE 212

189 200 300 400 500 600 700 800 1 1.2 1.4 1.6 1.8 2 2.2 x 10-3 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 6.3 Normalized optical dens ity spectra of polystyrene 1 m spheres in water.

PAGE 213

190 measured by the HP8453 spectrometer at wave lengths between 270 nm and 300 nm. Like interference structures, ripple structures are also strongly damped if absorption becomes large. Fig. 6.4 shows the normalized optical density spectra of polystyrene 4.0 m spheres in water. It featur es interference peaks, rippl e structure, and diffraction. Compared to Fig.6.3, the number of interference peak s are increased a nd shifted toward the longer wavelengths because of the larger particle size. Fig.6.4 clearly shows small ripple structure: sharp and highly irregular fi ne structure appears at wavelengths longer than 400 nm. Ripple structure arises if the denominators of the scattering coefficients vanish.7 The strong extinction peak in the deep UV region is due to diffraction that becomes apparent if the particle size is more than 10 times larger than the wavelength of the incident light.18 Fig.6.5 shows the normalized optical de nsity spectra of polystyrene 10 m spheres in water. Due to the particle si ze, diffraction becomes the dominant cause of extinction. However, the surfactan t used to stabilize the partic le surface also contributes to the observed extinction in the deep UV re gion. Therefore, the strong extinction peaks observed at wavelengths shorter than 250 nm are not only caused by diffraction but they are partially due to the surfact ant. Clearly, the interpretati on of the extinction peaks of large polystyrene particles at wavelengt hs below 250 nm requires extra caution.18 In summary, the optical density spectra of polystyrene sphe res with sizes ranging from 20 nm to 10 m were measured using the Integrated UV-VIS MAMW

PAGE 214

191 200 300 400 500 600 700 800 1 1.5 2 2.5 3 3.5 x 10-3 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 6.4 Normalized optical dens ity spectra of polystyrene 4 m spheres in water. Normalized Optical Density

PAGE 215

192 200 300 400 500 600 700 800 0 0.005 0.01 0.015 Wavelength(nm)Normalized Optical Density Red: UV-VIS MAMW Blue: HP8453 Figure 6.5 Normalized optical dens ity spectra of polystyrene 10 m spheres in water. Normalized Optical Density

PAGE 216

193 spectrometer. The measured spectra show the characteristics of re ddening, interference, ripple structures, diffraction, and absorption du e to electronic transi tions. The results are in good agreement with the optical density sp ectra of corresponding polystyrene spheres in water recorded with the HP 8453 spectrometer. This demonstrated the capability of the Integrated UV-VIS MAMW spectrometer to perform transmission spectroscopy. For the optical density spectra of polystyrene 1.0 m spheres, ripple struct ures are not explicit compared to the spectra measured by the HP 8453 spectrometer. This is caused by low resolution of the Integrated UV-VIS MAMW spectrometer. The resolution can be enhanced by using narrower width sl its for S2000 production spectrometer. 6.2. Results of UV-VIS MAMW Spectra Measurement The capabilities of the Integrated UV-VI S MAMW spectrometer to characterize particles by the simultaneous measurement of th e JPPD, mainly size, shape, and chemical composition, were tested by measuring th e UV-VIS MAMW spectra of polystyrene standards with different sizes, shape, and composition. The measured UV-VIS MAMW spectra were displayed as UV-VIS MAMW re sponse surface: 3-dimensional (scattering angle, wavelength and the log intensity ra tio) plot of the measured UV-VIS MAMW spectra, 2-dimensional (scattering angle and wavelength) contour plot of the response surface, and 2-dimensional (wavelength and the log intensity ratio) wavelength-view plot of the response surface. The measuremen ts demonstrate that the UV-VIS MAMW response surfaces determined by the JPPD of particles can serve as the spectroscopic fingerprint of particles. Compared to the prototyp e MAMW spectrometer, the

PAGE 217

194 incorporation of broadband UV light source and low angle scattering capacity widens the range of particles that can be charac terized with the MAMW spectrometer. 6.2.1. Characterization of Polystyrene Spheres by Size UV-VIS MAMW spectra of polystyrene spheres with sizes of 20 nm, 500 nm, 1 m, 4 m, 8 m, and 10 m were measured to test the capabilities of the Integrated UVVIS MAMW spectrometer for characterizi ng particles by sizes. The resulting UV-VIS MAMW response surfaces of standards are clearly distinguishabl e. The measurement results demonstrate that particles as small as 20 nm and as large as 10 m can be readily characterized using the UV-VIS MA MW scattering measurements. Fig 6.6 through Fig. 6.8 are the UV-VIS MAMW spectra of polystyrene 20 nm spheres in water. Fig 6.6.a shows the m easured UV-VIS MAMW response surface and Fig. 6.6.b represents the simulated UV-VIS MAMW response surface plot. The height of response surface is the log intensity ratio of scattered light to incident light. The surface is a function of the scattering a ngle and the wavelength of the incident light. The scattering angle is limited from 4o to 25o due to refraction by the rectan gular cuvette wall. Note that the plotted wavelength range extends from 230 nm to 450 nm. The 3-dimensional view of the UV-VIS MAMW response surface makes it possible to see the absorption and scattering components simultaneously. Fig 6.7.a shows the contour plot of the measured response surface and Fig 6.7.b represents the contour plot of the simulated response surface. The colors designate the height of the surface and the color bar to the

PAGE 218

195 Figure 6.6. UV-VIS MAMW response surfaces of pol ystyrene 20 nm spheres in water. a) Measured response surface. b) Simulated res ponse surface. Note that wavelength extends from 230 nm to 450 nm. a b

PAGE 219

196 Figure 6.7. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 20 nm spheres in water. a) Contour plot of the meas ured response surface. b) Contour plot of the simulated response surface. a b

PAGE 220

197 Figure 6.8. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 20 nm spheres in water. a) Wave length-view plot of th e measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 221

198 right of the plot clarifies the height. Fig 6.8.a shows the wavelength-view plot of the measured response surface and Fig 6.8.b represents the wavelengt h-view plot of the simulated response surface. The wavelength-view plots are functions of log intensity ratio and the wavelength of the incident light. The distinct aspect of the measured UV-VIS MAMW spectra of polystyrene 20 nm spheres is the bulk absorption band of polystyrene shown at wavelengths below 280 nm. The resonant absorption at wavelengths below 240 nm is caused by electronic transition. Together with absorption, scat tering provides additional size information because scattering intensity is enhanced with increasing particle si zes. Scattering extends toward longer wavelengths if particle size increases. The measurement of polystyrene 20 nm spheres MAMW spectra becomes possible due to the inclusion of multiwavele ngth UV light source. The minimum particle size that might be characterized with the Integrated UV-VIS MAMW spectrometer is probably 5 nm in case of polystyrene.19 This indicates that the realm of the application area of the Integrated UV-VIS MAMW spectro meter can be extended to virus detection and nano-particle characterization. The fact that small particle scattering is not only determined by the size but its shape and com position makes this prospect even brighter because of the simultaneous JPPD detecti on capability of the Integrated UV-VIS MAMW spectrometer. Fig 6.9.a, Fig 6.10.a, and Fig 6.11a represent the measured UV-VIS MAMW response surface of polystyrene 500 nm spheres in water, the contour plot of Fig 6.9.a, and the wavelength-view plot of Fig 6.9.a, respectively. Fig 6.9.b, Fig 6.10.b, and Fig

PAGE 222

199 6.11.b show the simulated UV-VIS MAMW re sponse surface of polystyrene 500 nm spheres in water, the contour plot of Fi g 6.9.b, and the wavelength-view plot of Fig 6.9.b, respectively. Note that th e plotted wavelength range extends from 230 nm to 700 nm. The UV-VIS MAMW spectra of polystyrene 500 nm spheres mainly feature interference and reddening. Comparison of Fi g.6.9.a with Fig.6.6.a shows the enhanced scattering structures at angles larger than 15o in Fig.6.9.a because of the increased particle size. Both the measured and the simulated UV-VIS MAMW spectra show that absorption and scattering are limited to the shorter wa velength region. This implies that UV light scattering is a valuable tool to characterize small size particles. The noise in the visible region, which can be clearly appreciated in the wavelength-view plot, a ppear to be caused by the tungsten-halogen beam spectrum that has considerable structure unlike the UV beam spectrum. Fig 6.12.a, Fig 6.13.a, and Fig 6.14.a repr esent the measured UV-VIS MAMW response surface of polystyrene 1 m spheres in water, the c ontour plot of Fig 6.12.a, and the wavelength-view plot of Fig 6.12.a respectively. Fig 6.12.b, Fig 6.13.b, and Fig 6.14.b show the simulated UV-VIS MAMW response surface of polystyrene 1 m spheres in water, the contour plot of Fi g 6.12.b, and the wavelength-view plot of Fig 6.12.b, respectively. Note that the plotted wa velength range extends from 230 nm to 800 nm. The measured UV-VIS MAMW spectra of polystyrene 1.0 m spheres in water show further enhancement of low angle scatte ring due to the larger particle size.

PAGE 223

200 Figure 6.9 UV-VIS MAMW response surfaces of polystyrene 500 nm spheres in water. a) Measured response surface. b) Simulate d response surface. Note that wavelength extends range from 230 nm to 700 nm. a b

PAGE 224

201 Figure 6.10. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 500 nm spheres in water. a) Contour plot of the m easured response surface. b) Contour plot of the simulated response surface. a b

PAGE 225

202 Figure 6.11. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 500 nm spheres in water. a) Wave length-view plot of the measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 226

203 Figure 6.12 UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water. a) Measured response surface. b) Simulated response surface. a b

PAGE 227

204 Figure 6.13. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water. a) Contour plot of the meas ured response surface. b) Contour plot of the simulated response surface. a b

PAGE 228

205 Figure 6.14. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 1 m spheres in water. a) Wave length-view plot of the measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 229

206 Particle size increases result in the enhan ced scattering at low angle and at longer wavelengths as well as the increased fine structures on both measured and simulated UVVIS MAMW spectra. Fig 6.14. a, the wavelengthview plot of the measured response surface, suggests that interference is th e main characteristic of polystyrene 1.0 m spheres. Noise probably due to the structur ed tungsten-halogen beam spectrum also appears in the visible region. Experimental a nd simulation results ar e in better agreement compared to those of smaller size particles. Fig 6.15.a, Fig 6.16.a, and Fig 6.17.a repr esent the measured UV-VIS MAMW response surface of polystyrene 4 m spheres in water, the c ontour plot of Fig 6.15.a, and the wavelength-view plot of Fig 6.15.a respectively. Fig 6.15.b, Fig 6.16.b, and Fig 6.17.b show the simulated UV-VIS MAMW response surface of polystyrene 4 m spheres in water, the contour plot of Fi g 6.15.b, and the wavelength-view plot of Fig 6.15.b, respectively. Note that the plotted wa velength range extends from 230 nm to 800 nm. The measured UV-VIS MAMW spectra of polystyrene 4.0 m spheres in water show more fine structure spread on the whole response surfac e or contour and the wavelength-view plots. Increased interferen ce peaks also can be seen at low angles. Although it is unclear from the measured spectr a, the ripple structur e can be observed in the simulated plots. Deep and straight stru ctures shown at wavelengths below 300 nm in both experimental and simulation plots seem to be caused by diffraction. Fig 6.18.a, Fig 6.19.a, and Fig 6.20.a repr esent the measured UV-VIS MAMW response surface of polystyrene 8 m spheres in water, the c ontour plot of Fig 6.18.a,

PAGE 230

207 Figure 6.15 UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water. a) Measured response surface. b) Simulated response surface. a b

PAGE 231

208 Figure 6.16. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water. a) Contour plot of the meas ured response surface. b) Contour plot of the simulated response surface. a b

PAGE 232

209 Figure 6.17. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 4 m spheres in water. a) Wave length-view plot of the measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 233

210 and the wavelength-view plot of Fig 6.18.a, respectively. Fig 6.18.b, Fig 6.19.b, and Fig 6.20.b show the simulated UV-VIS MAMW response surface of polystyrene 8 m spheres in water, the contour plot of Fi g 6.18.b, and the wavelength-view plot of Fig 6.18.b, respectively. Note that the plotted wa velength range extends from 230 nm to 800 nm. Due to the increased particle size, the measured and the simulated UV-VIS MAMW spectra of polystyrene 8.0 m spheres in water show a large peak shift toward the short near infrared (NIR) region. Increased particle size also enhances diffraction and results in the straight struct ures shown on the majority of response surfaces. Simulated plots are in good agreement with experimental results. Fig 6.21.a, Fig 6.22.a, and Fig 6.23.a repr esent the measured UV-VIS MAMW response surface of polystyrene 10.0 m spheres in water, the contour plot of Fig 6.21.a, and the wavelength-view plot of Fig 6.23.a, respectively. Fig 6.21.b, Fig 6.22.b, and Fig 6.23.b show the simulated UV-VIS MAMW response surface of polystyrene 10.0 m spheres in water, the contour plot of Fi g 6.21.b, and the wavelength-view plot of Fig 6.23.b, respectively. Note that the plotted wa velength range extends from 230 nm to 800 nm. Straight structures that seem to be caused by diffraction spread on the whole response surfaces of polystyrene 10.0 m spheres in water. UV-VIS MAMW spectra of polystyrene 8.0 m spheres and that of polystyrene 10.0 m spheres are clearly distinguishable at low angles whil e looking similar at large angles.

PAGE 234

211 Figure 6.18. UV-VIS MAMW response surfaces of polystyrene 8 m spheres in water. a) Measured response surface. b) Simulated response surface. a b

PAGE 235

212 Figure 6.19. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 8 m spheres in water. a) Contour plot of the meas ured response surface. b) Contour plot of the simulated response surface. a b

PAGE 236

213 Figure 6.20. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 8 m spheres in water. a) Wave length-view plot of the measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 237

214 Figure 6.21. UV-VIS MAMW response surfaces of polystyrene 10 m spheres in water. a) Measured response surface. b) Simulated response surface. a b

PAGE 238

215 Figure 6.22. Contour plot of the UV-VIS MAMW response surfaces of polystyrene 10 m spheres in water. a) Contour plot of the measured response surface. b) Contour plot of the simulated response surface. a b

PAGE 239

216 Figure 6.23. Wavelength-view plots of th e UV-VIS MAMW response surfaces of polystyrene 10 m spheres in water. a) Wave length-view plot of th e measured response surface. b) Wavelength-view plot of the simulated response surface. a b

PAGE 240

217 This confirms that the incorporation of low angle scattering capability into the Integrated UV-VIS MAMW spectrometer results in the la rge size particle characterization capability. Further shift of th e large peak toward the short NIR region can be inferred from the comparison of both measured and simulated UV-VIS MAMW spectra of polystyrene 10.0 m spheres in water with those of polystyrene 8.0 m spheres. This suggests that the use of a spectrometer grat ing which can resolve short NIR light and narrower width slit in the detection op tics may enhance the large size particle characterization capability of the Integrat ed UV-VIS MAMW spectrometer further. The capabilities of the Integrated UV-VI S MAMW spectrometer for particle size analysis and characterization was tested by measuring the UV-VIS MAMW spectra of polystyrene spheres with sizes of 20 nm, 500 nm, 1 m, 4 m, 8 m, and 10 m. The measurement results showed that particle s as small as 20 nm and as large as 10 m can be characterized with the Integrated UV-VIS MAMW spectrometer due to the incorporation of broadband UV light source and low angle scattering capacity. In addition, the measured results showed the po ssibility of widening the particle size range that can be characterized by the MAMW spectrometer. Methods of upgrading the Integrated UV-VIS MAMW spectrometer for fu rther enhancement of large size particle characterization capability have also been discussed.

PAGE 241

218 6.2.2. Measured UV-VIS MAMW Spectra of Non-spherical Polystyrene Standards Current methodologies for the characteriza tion of non-spherical particles by light scattering include:7 1. Forward scattering measurement the resu lts are compared with those obtained from area-equivalent spheres, 2. Polarization measurement esp ecially linear polarization and 3. Backscattering measurements. At present, the first method is actively used to size non-spherical particles. For the second and the third methods, emphasis is on the acq uisition of spectra that can be used to differentiate spherical and non-s pherical particles consistently. Motivated by the fact that new light scattering techniques which can be used to charact erize non-spherical particles are needed, experiments using a 1.85 m peanut-shaped polystyrene latex and a 1.87 m spherical polystyrene latex were designed a nd performed to test the capabilities of the Integrated UV-VIS MAMW spectrometer fo r the differentiation between spherical particles and non-spherical particles. The re sults demonstrate the sensitivity of the Integrated UV-VIS MAMW spectrome ter to particle shape effects. 6.2.2.1 Non-spherical Particles Peanut-shaped polystyrene latex particles of average size 1.85 m (Lot # PS1305B) and spherical polystyrene latex particles of average size 1.87 m (Lot # PS1587A) were purchased from Magsphere Inc. Peanut-shaped particles are chemically fused and cannot

PAGE 242

219 be separated or broken into pieces by physic al methods like ultr asonic shaking. Fig. 6.24 shows scanning electron microscopy (SEM) images of the spherical particles and peanutshaped particles respectively. For SEM imaging, dried sample solutions at higher concentrations than those utilized for th e measurement of the UV-VIS MAMW spectra were used for the ease of imaging. 6.2.2.2 Light Scattering by Ensembles of Non-spherical Particles Scattering properties of non-spherical pa rticles are different from those of spherical particles due to not only particle shape but also continuous ly changing particle orientation. Therefore, resulti ng spectra can be different even if the same shape particles are taken for scattering measurements. Fig. 6.25.a and Fig. 6.25.b, SEM images of two peanut-shaped particles, provide an example of the same shape particles placed in different orientations (perpendicular and para llel). Consequently, the particle orientation relative to the incident light must also be eval uated in order to assess the particle shape. Fig. 6.26 shows a microscopy picture of the concentrated peanut -shaped particles. For ease of imaging, a concentrated stock so lution was used. Although the distribution of particles and their orientati on shown under the microscope are continuously changing, the measured UV-VIS MAMW spectra of peanut-s haped particles are reproducible because the Integrated UV-VIS MAMW spectrometer measures the scattered light from the ensemble of particles, not fr om single or a few particles.

PAGE 243

220 Figure. 6.24. Scanning electron microscopy (S EM) images of polystyrene 1.9 m size standards. a) Spherica l particles. b) Peanut-shaped particles. a b

PAGE 244

221 Figure. 6.25. SEM images of two peanut -shaped particles placed in different orientations. a) Particle is oriented perpe ndicular to the incident light. b) Particle is oriented parallel to the incident light. a b Ii Ii Ii: Incident light

PAGE 245

222 Figure 6.26. Microscope picture of con centrated peanut-shaped particle suspension.

PAGE 246

223 6.2.2.3 Results of the UV-VIS MAMW Spectra Measurements of 1.9 m Size Spherical Particles and Peanut-Shaped Particles Fig 6.27.a, Fig 6.28.a, and Fig 6.29.a repr esent the measured UV-VIS MAMW response surface of polystyrene 1.87 m spheres in water, the contour plot of Fig 6.27.a, and the wavelength-view plot of Fig 6.27.a, respectively. Fig 6.27.b, Fig 6.28.b, and Fig 6.29.b show the measured UV-VIS MAMW response surface of peanut-shaped polystyrene 1.85 m standards in water, the c ontour plot of Fig 6.27.b, and the wavelength-view plot of Fig 6.27.b, respectively. The results are different across the complete wavelength range from 230 nm to 800 nm. The more dramatic differences bei ng apparent in the UV region. To see the differences clearly, the measured spectra ar e plotted again for the UV region (Fig. 6.30 ~ Fig. 6.32) and for the visible region (Fig. 6.33 ~ Fig. 6.35). The observed differences seem to result from: 1) absorbance differences between sphe rical and peanut-shaped non-spherical particles due to beam path length variations caused by peanut-shaped particle orientation changes and 2) destructive interference in the vici nity of peanut-shaped particles, especially by UV light due to the increased particle size parameter resulting from shorter wavelength light.

PAGE 247

224 Figure 6.27. Measured UV-VIS MAMW response surfaces of polystyrene 1.9 m size standard in water. a) Spheres. b) Peanut -shaped particles. Note that the plotted wavelength range extends from 230 nm to 800 nm. b a

PAGE 248

225 Figure 6.28. Contour plots of the measur ed UV-VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut-shaped particles. Note that the plotted wavelength ra nge extends from 230 nm to 800 nm. a b

PAGE 249

226 Figure 6.29. Wavelength-view pl ots of the measured UV-VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut -shaped particles. Note that the plotted wavelength ra nge extends from 230 nm to 800 nm. b a

PAGE 250

227 Figure 6.30. Measured UV MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut -shaped particles. Note that the plotted wavelength range extends from 230 nm to 450 nm. a b

PAGE 251

228 Figure 6.31. Contour plots of the measured UV MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut-shaped particles. Note that the plotted wavelength range extends from 230 nm to 450 nm. a b

PAGE 252

229 Figure 6.32. Wavelength-view pl ots of the measured UV MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut-shaped particles. Note that the plotted wavelength range extends from 230 nm to 450 nm. a b

PAGE 253

230 Figure 6.33. Measured VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut -shaped particles. Note that the plotted wavelength range extends from 450 nm to 800 nm. a b

PAGE 254

231 Figure 6.34. Contour plots of the measur ed VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut-shaped particles. Note that the plotted wavelength range extends from 450 nm to 800 nm. a b

PAGE 255

232 Figure 6.35. Wavelength-view pl ots of measured VIS MAMW response surfaces of polystyrene 1.9 m size standards in water. a) Spheres. b) Peanut-shaped particles. Note that the plotted wavelength range extends from 450 nm to 800 nm. a b

PAGE 256

233 Therefore, it maybe concluded that the we ll-known supposition that scattering properties of non-spherical particles can be approximated to those of spheres w ith equivalent sizes may hold depending on particle size parameter. In conclusion, the measurement of th e UV-VIS MAMW spectra of both peanutshaped particles and spherical particles of the same sizes shows clear differences, especially in the UV regi on. This demonstrates: 1) the capability of UV-VIS MAMW spect roscopy to distinguish particle shape and 2) the potential of the Integrated UV-VIS MAMW spectrometer to be another modality of non-spherical particle characterizati on by light scattering. 6.2.2.4. Proposed Set Up of an Integrat ed UV-VIS MAMW Spectrometer with Enhanced Backscattering Measurement Capacity As described in the beginning of this s ection, polarized light scattering and back scattering provide both theoretical and e xperimental methodologies for the study and characterization of non-spherica l particles because of their sensitivity to the particle shape and/or orientation.7 Therefore, the implementation of these methodologies will further enhance the capabilitie s of the Integrated UV-VIS MAMW spectrometer to infer or characterize non-spherica l particles. Currently, th e Integrated UV-VIS MAMW spectrometer employs unpolarized light only. In addition, the geometry of our current prototype UV-VIS MAMW spectrometer hinders the measurement of backscattering.

PAGE 257

234 Methods of implementing backscattering a nd polarized light scattering measurements are discussed in this section. Fig. 6.36 shows a schematic of an en hanced Integrated UV-VIS MAMW spectrometer design that will have the capabil ity of backscattering and polarized light scattering measurements. It incorporates polarization optics for the Mueller-matrix measurement and a new double slit for the backscattering measurement. A pair of linear polarizer a nd half-wave plate will be in stalled in both the incident beam path and the scattered beam path. Theo retically, this will enable the measurement of the 16-element Mueller matrix (the comb ination of all the possible polarized light scattering measurements). As stated in sect ion 3.2.2, measurement of the Mueller matrix allows of the acquisition of all the information obtaina ble from light scattering. Measurement of backscattered light can be implemented by the use of a double slit in the path of the incident beam. One slit will be used as entrance for the incident beam to the sample cell and the other slit wh ich will be located at a calculated position that will allow the optimum detection of back scattered light. Both slits will be located closely and thus, the incident beam is required to be shaped before arriving at the double slit. This can be accomplished by installing addi tional single rectangular slit just after the collimation lens L1.

PAGE 258

235 Figure. 6.36. Schematic of the Integr ated UV-VIS MAMW spectrometer with enhanced backscattering measurement capacity.

PAGE 259

236 6.2.3. Characterization of Polystyrene Standards by Composition The capability of the Integrated UV-VIS MAMW spectrometer to characterize particle composition was tested usin g Duke Scientific polystyrene 3.0 m spheres (Duke Scientific catalog No. 4203A) and green fluorescent polystyrene 3.0 m spheres (Duke Scientific catalog No. G0300). Both standards have the same specifications except for the compositional differences. The green fluorescen t dye has an excitation maxima at 468 nm and an emission maxima at 508 nm. It also has a weak excitation band at around 330 nm. Before the measurement of the UV-VIS MA MW spectra, the effect of the dye on the incident UV-VIS beam spectrum was inve stigated. For this purpose, the optical density spectra of polystyrene 3.0 m spheres, green fluorescent polystyrene 3.0 m spheres, and red dye coated polystyrene 3.0 m spheres in water were measured. The red dye coated polystyrene 3.0 m spheres are from Polysciences, Inc. (Polyscience catalog No. 17137). For convenience, each st andard will be designated as ps3.0 (polystyrene 3.0 m spheres), pf 3.0 (green fluorescent polystyrene 3.0 m spheres), and pr 3.0 (red dye coated polystyrene 3.0 m spheres) respectively. Fig. 6.37 shows the measured optical density spectra of ps3.0, pf3.0, and pr3.0 polystyrene spheres in water For comparison purposes, the measured spectra are normalized with the area under the curve. The optical density spect ra of ps3.0 and pf3.0 standards show no appreciable differences in the visible region. On the other hand, that of pr3.0 standard shows reduced optical density in the visible region between 430 nm and 580 nm. This is attributed to different manufact uring procedures of the pf3.0 and pr 3.0 standards. In case of pf3.0, green fluorescent dye is inco rporated into the polymer matrix

PAGE 260

237 200 300 400 500 600 700 800 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 Wavelength, nmAbsorbance05-09-04 Ps 3, Pr 3, Pf 3 Micron Absorbance B: Ps 3 R: Pr 3 G: Pf 3 Figure. 6.37. Measured optical density spec tra of 3.0 m size polystyrene: spheres, green fluorescent, and red dyed. The spectru m of the red dyed polystyrene spheres has a different profile because the dye is surface co ated. The HP 8452 spectrometer was used for this measurement. Blue: Polystyrene 3 m Green: Green fluorescent Polystyrene 3 m Red: Red dyed Polystyrene 3 m Normalized Optical Density Wavelength (nm)

PAGE 261

238 during polymerization. This method preven ts dye leaching in aqueous media and minimizes photobleaching.13 Therefore, green fluorescent dye behaves like intrinsic particle chromophore. For the pr3.0 standard, the dye is coat ed on the surface of sphere.48 As a result, majority of the dye stays at the su rface of particle and t hus, the analysis of the measurement results clearly requires the inhomogeneous particle model for the acquisition of correct particle size information. Consequent ly, pr 3.0 standard was not chosen for the feasibility test in this dissertation research. The measurements were performed using th e octagonal cuvette to see the effect of the fluorescent dye clearly. B ecause refraction is negligible, the measured angular range could easily be extended up to 60o. The results are expressed in terms of the observation angle not the scattering angle. Fig 6. 38.a, Fig 6.39.a, and Fig 6.40.a represent the measured UV-VIS MAMW response surface of polystyrene 3.0 m spheres in water, the contour plot of Fig 6.38.a, and the wavelengt h-view plot of Fig 6.38.a, respectively. Fig 6.38.b, Fig 6.39.b, and Fig 6.40.b show the measured UV-VIS MAMW response surface of green fluorescent polystyrene 3.0 m spheres in water, the contour plot of Fig 6.38.b, and the wavelength-view pl ot of Fig 6.38.b, respectively. Both response surfaces show almost the same features except the locations at around 300 nm and 450 ~ 550 nm and approximate angular range of 30o ~ 60o where the intensity of the response surface of pf 3.0 standard is enhan ced due to the fluorescent dye. This can be seen clearly from the contour pl ots and the wavelength-vi ew plots. Typically, light scattering or optical density measurem ents do not yield fluorescence information.

PAGE 262

239 Figure 6.38. Measured UV-VIS MAMW response surfaces of polystyrene 3.0 m standards in water. a) Polystyrene spheres. b) Green fluorescent pol ystyrene spheres. In case of b), both particle size and composition (green fl uorescence) information is available from the response surface. The measurements were conducted with an octagonal cuvette and the results are plotte d as function of the observation angle. Wavelength (nm) Observation Angle (deg) Log10 (Isca/Ii) a Wavelength (nm) Observation Angle (deg) Log 10 (Isca/Ii) b

PAGE 263

240 Figure 6.39. Contour plots of the measur ed UV-VIS MAMW response surfaces of polystyrene 3.0 m standards in water. a) polystyrene spheres. b) green fluorescent polystyrene In case of b), fluorescence information can be seen at around 500 nm. a b Observation Angle (deg) Wavelength (nm) Observation Angle (deg) Wavelength (nm)

PAGE 264

241 Figure 6.40. Wavelength-view pl ots of the measured UV-VIS MAMW response surfaces of polystyrene 3.0 m standards in water. a) polystyrene spheres. b) green fluorescent polystyrene. The size and fluorescence co mposition information can be readily appreciated. a Wavelen g th ( nm ) Log10 (Isca/Ii) Wavelength (nm) Log10 (Isca/Ii) b

PAGE 265

242 However, the measurement of the UV-VIS MAMW spectra can yield particle size and fluorescence information simultaneously. This is especially useful for biological particle characterization. The capability of the Integrated UV-VIS MA MW spectrometer to measure particle composition was tested using polystyrene 3.0 m size spheres and green fluorescent polystyrene 3.0 m size spheres. The resulting UV-VIS MAMW spectra show differences due to particle composition gr een fluorescent dye. This demonstrates the capability of the Integrated UV-VIS MAMW spectrometer to measure particle composition. This capability is also demonstrat ed in other parts of this dissertation: observation of polystyrene bulk absorption band from polystyrene 20 nm spheres UVVIS MAMW spectra and the disappearance of hemoglobin doublet in sickled red blood cells. All these results demonstrate the pot ential of the Integrated UV-VIS MAMW spectrometer for the characteri zation of micron and sub-micron size particles in terms of size and chemical composition. 6.3. Results of Frequency Domain Fluores cence Emission Spectrum Measurement The capability of the Integrated UV-VI S MAMW spectrometer to perform the frequency domain fluorescence spectroscopy was tested by measuring the fluorescence emission spectrum from green fluorescent polystyrene 3.0 m spheres that was acquired from Duke Scientific Cor poration. The green fluorescent dye has the excitation maxima at 468 nm and the emission maxima at 508 nm as shown in Fig. 6.41.13 Its Stokes shift is 40 nm. In addition, it has weak excita tion band at around 330 nm. Fig.6.42.a shows UV

PAGE 266

243 excitation and Fig.6.42.b exhibits visible excitation of green fluorescent dye.13 Both excitations lead to the emissi on of fluorescent light with peak intensity at around 515 nm. Fig.6.43 shows the photograph of fluorescence measurement set up and Fig. 6.44 illustrates the schematic of the experimental set up. A fused silica convex lens with f/# 2.0 (25 mm diameter and 50 mm fo cal length, L2) was installed in front of the cuvette to focus the incident UV-VIS beam into the samp le cell. Lens L4 described in Section 4.2.2 was removed to improve the signal sensitiv ity at the detector end. If necessary, a bandpass filter with the center wavelengt h at 470 nm and 10 nm FWHM (THORLABS, Inc., Model No FB470-10) or a bandpass filter with the cente r wavelength at 510 nm and 10 nm FWHM (THORLABS, Inc., Model No FB 510-10) was alternately installed at the location of lens L4 to meas ure a portion of the excitation (470 nm) or the emission (510 nm) spectrum. For typical fluorescence measurement setup, the bandpass filter that transmits only the excitation li ght and blocks the other light is installed in front of the sample cell. However, with the current prototype UV-VIS MAMW spectrometer setup, it was inevitable to install the bandpass filter fo r the excitation light after the sample cell because of the weak visible light intensity from DT-1000 light source. Deviations from the typical fluorescence measurement setup, i. e. the use of single detector and the alternate installation of bandpass filters, b ecame possible due to the use of goniometer, a fluorescence sample with known excitation a nd emission spectra, and the CCD detector (S2000 production spectrometer) that allows th e simultaneous detection of the broadband UV-VIS-NIR beam.

PAGE 267

244 Figure 6.41. Spectral informati on of Duke Scientific Corporation green fluorescent michrospheres. 13

PAGE 268

245 Figure 6.42. Green fluorescent dye excitation. a) UV excitation. b) Visible excitation. Both excitation lead to the emission of fluor escent light with peak intensity at around 515 nm.13 b a

PAGE 269

246 Figure 6.43 Photograph of fluore scence measurement set up. Note that the locations of lens L5 and S2000 production spectrometer were moved closer to the sample cell to maximize the amount of light de tected by the spectrometer.

PAGE 270

247 Figure 6.44. Schemati c of the fluorescence measurement set up.

PAGE 271

248 Note that lens L5 and S2000 production spectr ometer were moved closer to the sample cell to maximize the amount of light de tected by the spectrometer. During the measurement, the integration time was held at 5000 msec and the average sample number for the signal averaging was held at 5. Ther efore, the detected fluorescence emission spectrum was displayed on the monitor screen and if necessary recorded with the OOIBase32 software at every 25 sec. Fig.6.45 shows the measured excitati on and fluorescence emission spectra. It displays excitation (absorption) spectrum w ith peak intensity at 490 nm and fluorescence emission spectrum with peak intensity at 515 nm. The absorption and the fluorescence emission spectra are overlap due to broad excitation and emission band. Consequently, the locations of peak intensities are slightly shifted from those shown in the manufacturers data sheet. Th e broad but weak peak at around 330 nm is due to a UV absorption band. A series of small peaks after 550 nm are caused by scattered light. This was confirmed by repeating the measurement using polystyrene 3.0 m spheres. Both results, the measured spectra of polystyrene 3.0 m spheres and those of green fluorescent polystyrene 3.0 m spheres, are plotted in Fig. 6.46 for comparison purposes. Fig. 6.47. a shows the bandpass-filtered absorption (red 470 nm) and fluorescence emission (green 510 nm) spectra of the green fluorescent polystyrene spheres and the corresponding bandpass-filter ed scattered-light spectra(blue 470 nm, black 510 nm) from polystyrene spheres measured at 90o.

PAGE 272

249 250 300 350 400 450 500 550 600 650 700 0 200 400 600 800 1000 1200 1400 1600 Wavelength, nmIntensity-countsfluorescence spectra Figure 6.45. The excitation and fluorescence emission spectra measured with the Integrated UV-VIS MAMW spectrometer. It displays the exc itation (absorption) spectrum with peak intensity at 490 nm and the emission (fluorescence) spectrum with peak intensity at 515 nm. 250 300 350 400 450 500 550 600 650 700 0 200 400 600 800 1000 1200 1400 1600 Green : gf3 Black: ps3 Figure 6.46. The excitation and fluorescence emi ssion spectra (green) and the scatteredlight spectrum of the pa rticle of the same size (black) m easured with the Integrated UVVIS MAMW spectrometer. Wavelen g th ( nm ) Counts Wavelen g th ( nm ) Counts

PAGE 273

250 460 470 480 490 500 510 520 530 -100 0 100 200 300 400 500 600 700 Wavelength, nmIntensity-aufiltered fluorescence and scattered light intensity at 90 deg red: fl excitation at 470 nm green: fl emission at 510 nm blue: ps3.0 at 470 nm black: ps3.0 at 510 nm 460 470 480 490 500 510 520 530 540 0 200 400 600 800 1000 1200 1400 Wavelength, nmIntensity-aufilter-20 deg red: fl excitation at 470 nm green: fl emission at 510 nm blue: ps3.0 at 470 nm black: ps3.0 at 510 nm Figure 6.47. Band pass-filtered absorption, fluorescence and scattered light spectra. a) Measured at 90o. b) Measured at 20o. Wavelen g th ( nm ) Counts Scattered li g h t Excitation Fluorescence a Scattered light Wavelength (nm) Counts b Excitation Fluorescence

PAGE 274

251 Fig. 6.47. b represents the results of th e same measurements repeated at 20o. In case of Fig. 6.47. a, the intensities of both absorpti on and fluorescence light are stronger than those of the scattered-light from polystyre ne spheres. However, in Fig. 6.47. b, the scattered light has str onger intensities compared with those of the absorption or the fluorescence light. This further confirms the measurement of fluorescence spectra by the Integrated UV-VIS MAMW spectrometer. The possibility of performing the fre quency domain fluorescence spectroscopy with the Integrated UV-VIS MAMW spectrome ter was tested and confirmed. The result is consistent with the data provided by th e manufacturer of the fluorescence standard. However, practical application of fluores cence spectroscopy with th e Integrated UV-VIS MAMW spectrometer requires the use of stronger light source, especially in the visible region. The power of the tungsten-halogen bulb used in this measurement was only 6.5 W. Weak visible light intensity was partiall y compensated by using large CCD detector (S2000 production spectrometer) integration time s (5000 msec). One of the merits of the CCD detector usage in the frequency domain fluorescence spectroscopy is that no scan of absorption or fluorescence light is require d. Unlike photomultiplier tubes (PMT), the CCD detector can immediately record whol e absorption and fluorescence spectra. This allows room to increase the detector integra tion time substantially and thus, permits the use of tungsten-halogen (visible) or Deuteriu m (UV) lamp that has weak power compared to several hundred W Xenon lamp as the ex citation light source Therefore, single Deuterium tungsten-halogen light source can be used for UV-VIS MAMW scattering, transmission, and fluorescence measurement.

PAGE 275

252 Additional fluorescence measurements us ing various samples are recommended to elucidate further requirements inevitable to the practical application of the frequency domain fluorescence spectroscopy with the Integrated UV-VIS MAMW spectrometer. 6.4. Results of the UV-VIS MAMW Spectra Measurement of Whole Blood Samples UV-VIS MAMW spectra of normal whole blood samples and sickled whole blood samples were measured. The resulting sp ectra showed differences due to changes in red blood cell (RBC) morphology and compositional changes in sickled blood samples. These experiments confirmed the cap ability of the Integrated UV-VIS MAMW spectrometer to detect particle shape and compositional change simultaneously. The implication of the measured results and the potential of the Integrated UV-VIS MAMW spectrometer for biological part icle characterization and for th e application to the medical area are discussed. 6.4.1.Sample Preparation and Measurement Whole blood samples from a sickle cell anemia patient were obtained from Florida Blood Services (FBS). Part of the bl ood sample was kept in its normal status and part of it was used to induce the sickling of cells by deoxygenating the sample. Sickle cell anemia induces red blood cell (RBC) morphol ogy changes from ellipsoid into holly-leaf or crest shape when oxygen is deprived.11 Fig 6.48.a shows a picture of normal red blood cells which have the shape of biconcave discs and Fig 6.48.b shows a picture of sickle cells.37

PAGE 276

253 Figure 6.48. Picture of red blood cells and sickled blood cells. a) Red blood cells have a biconcave disc shape of ~ 8 m. b) The crescent-non-spherical blood cell (bottom, left) shows sickled red blood cells. Cited from Reference 37. a b

PAGE 277

254 Because Red blood cells and sickle cells have different morphology and composition, the observation of the UV-VIS MAMW spectra of both samples and the comparison of the results provide an opportunity to confirm the simultaneous detection of particle shape and compositional changes. In the laboratory, sickling can be in duced by adding proper amount of Sodium meta Bisulfate (SMBS) to the blood samp le. SMBS deoxygenates red blood cells. The addition of 1 to 2 drops of 2 % SMBS solution to one drop of blood enables the observation of sickle cells under the microscope.11 On the first day of drawing sample, 100 l of 2% SMBS solution was added to 100 l blood to induce sickling. However, the microscope observation showed no sickle cel ls. Probably, a leak of air, which was remained on the top empty portion of sample tube, into the blood sample seemed to prevent deoxygenation. On the second day, more concentrated SMBS solution was added to blood samples and each sample tube was filled and sealed tightly to prevent air leakage. Dr. Debra Huffman at the College of Marine Sciences, University of South Florida confirmed that more than 80 ~ 90 % of sickling was induced in one sample. UV-VIS MAMW spectra of the normal blood sample were measured on the first day and those of the normal and the sickled blood samples were taken on the second day. For UV-VIS MAMW measurements, both the no rmal and the sickled blood samples are diluted in the saline at the c oncentration of approximately 10-4. Fig. 6.49.a shows the normalized intensity profile of saline and SMBS saline solution, respectively. The scope mode of OOIBase32 software was used fo r these measurements. SMBS is a strong absorber of UV light and results in the re duction of the incident light intensity at

PAGE 278

255 wavelengths below 250 nm. However, above 250 nm, both reference solutions show the same features. Therefore, measured results are not affected by the optical property differences of reference solutions. This can al so be confirmed in the expanded plot of Fig. 6.49.a, shown in Fig. 6.49.b that is plotted fo r the wavelength range from 500 nm to 600 nm where hemoglobin doublets are located. The measured results are plotted from 250 nm to avoid any artifacts due to differences in the optical properties of the suspending media. 6.4.2. Measured UV-VIS MAMW Spectra of Normal Whole Blood Sample Fig. 6.50.a is MAMW response surface of pur e red blood cells in saline measured with the prototype MAMW spectrometer2 and Fig 6.51.a is its c ontour plot. Fig. 6.50.b is UV-VIS MAMW response surface of norma l whole blood in saline taken by the Integrated UV-VIS MAMW spectrometer on th e day of blood sample extraction and Fig 6.51.b is its contour plot. Because red blood cells are the most nume rous element in blood, their spectral features are similar to those of whole blood.35 For Fig. 6.50.a and Fig. 6.51.a, only partial aspects of RBC spectra are shown. Especiall y, the major hemoglobin absorption band at 415 nm was truncated. On the other ha nd, Fig. 6.50.b and Fig. 6.51.b provide complete blood spectra from 250 nm to 800 nm. The ab sorption band at 260 nm ~280nm originates from DNA, RNA, and protein. Absorption bands at around 330 nm, 415 nm and doublet at around 570 nm are due to hemoglobin.47

PAGE 279

256 200 300 400 500 600 700 800 0 0.5 1 1.5 2 2.5 3 3.5 x 10-3 Saline SMBS 500 510 520 530 540 550 560 570 580 590 600 0.008 0.0085 0.009 0.0095 0.01 0.0105 0.011 0.0115 0.012 Saline SMBS Figure 6.49. Normalized intensity profile of sa line and SMBS dissolved saline solution. a) Plotted for the range of 190 nm to 820 nm b) Expanded plot from 500nm to 600 nm where hemoglobin doublets are located. Note that the scope mode of OOIBase32 software was used for these measurements. a b Wave length (nm) Wave length (nm) Normalized Intensity (au) Normalized Intensity (au)

PAGE 280

257 Figure 6.50. Measured MAMW response surfaces of blood samples. a) MAMW response surface of pure red blood cells in saline measured with the prototype MAMW spectrometer.2 b) UV-VIS MAMW response surface of normal whole blood in saline measured on the day of blood sample extraction. a b Scattering Angle (deg) Wavelength (nm) Log (Isca/Ii)

PAGE 281

258 Figure 6.51. Contour plots of the measured MAMW response surfaces of blood sample. a) normal red blood (Fig 6.50.a).2 b) normal whole blood in saline (Fig 6.50.b). a b Wavelength (nm) Scattering Angle (deg)

PAGE 282

259 Therefore, it can be concluded that th e incorporation of UV light into the MAMW spectrometer enhanced the capability of biol ogical particle characterization by providing biologically important UV absorption or fluorescence spectra originating from biologically relevant chromophores (DNA, RNA, and proteins, etc). 6.4.3. Measured UV-VIS MAMW Spectra of Sickle Cells Fig 6.52.a, Fig 6.53.a, and Fig 6.54.a repr esent the UV-VIS MAMW response surface of normal whole blood in saline meas ured with the Integrated UV-VIS MAMW spectrometer on the second day of blood sample extraction, the contour plot of Fig 6.52.a, and the wavelength-view plot of Fig 6.52.a, respectively. Fig 6.52.b, Fig 6.53.b, and Fig 6.54.b show the UV-VIS MAMW response surf ace of sickled blood in SMBS saline solution taken with the Integrated UV-VIS MAMW spectrometer on the second day of blood sample extraction, the c ontour plot of Fig 6.52.b, and th e wavelength-view plot of Fig 6.52.b, respectively. Compared to the normal blood response surface Fig. 6.52.a, the sickle cell response surface Fig. 6.52.b shows a round profile at low angles, widened log intensity ratio, reduced protein (280nm) and hemoglobin (415 nm) absorption band intensities, and the disappearance of hemoglobi n doublets at around 570nm. This can also be confirmed from the corresponding contour plots of Fig. 6.53.a and Fig. 6.53.b or from the wavelength-view plots of Fig. 6.54.a and Fi g. 6.54.b. Particularly, th e wavelength-view plots provide a clear distinction of the spectral differences.

PAGE 283

260 Figure. 6.52. UV-VIS MAMW response surface of normal and sickled blood samples. a) UV-VIS MAMW response surface of normal whole blood in saline measured on the second day of blood sample extraction. b) UV-VIS MAMW response surface of sickled blood in SMBS dissolved saline taken on th e second day of blood sample extraction. a b

PAGE 284

261 Figure 6.53. Contour plots of the UV-VIS MA MW response surface of blood sample. a) normal whole blood sample response surface (Fig 6.52.a ). b) sickled blood sample response surface (Fig 6.52.b). a b

PAGE 285

262 Figure 6.54. Wavelength-view plots of th e UV-VIS MAMW response surface of blood sample. a) normal whole blood sample response surface (Fig 6.52.a). b) sickled blood sample response surface (Fig 6.52.b). a b

PAGE 286

263 The round profile at low angle and the widene d log intensity ratio of the sickled blood samples seem to be due to particle morphology change s from ellipsoid into holly-leaf or crest shape. The disappear ance of the hemoglobin doublet s at around 570 nm in the sickled blood sample can be attribut ed to the deoxygenation of hemoglobin.47 Fig.6.55 shows the measured optical density spect ra of oxy-hemoglobin, de-oxyhemoglobin and met-hemoglobin, respectively.42 De-oxyhemoglobin has reduced oxygen and methemoglobin has no bound oxygen.42 Sickled blood cell hemo globin changes show the features of met-hemoglobin rather than de -oxyhemoglobin. In the wavelength-view plot of sickle cells, Fig. 6.54.b, hemoglobin doubl ets at around 570 nm totally disappeared and the hemoglobin absorption band at 415 nm is shifted toward shorter wavelength region compared to Fig. 6.54.a. However, in the case of me t-hemoglobin, the absorption peak at 415 nm is enhanced while that of sickled blood samples is decreased. This implies that the mechanism of de-oxygenati on in the hemoglobin of the sickled blood samples may not be the same as that of de-oxyhemoglobin or met-hemoglobin. Additional study to clarify the decrease of hemoglobin absorption band at 415 nm in sickled blood samples is recommended. Comparison of the UV-VIS MAMW response surfaces of normal blood sample Fig. 6.50.b and Fig. 6.52.a shows the enhanced hemoglobin absorption band at 415 nm and hemoglobin doublets at around 570 nm in Fig. 6.52.a. This suggests that the sample was oxygenated further through storage.

PAGE 287

264 -1000 1000 3000 5000 7000 9000 11000 13000 15000 1902903904905906907908909901090 Wavelength (nm)Extinction Coefficient oxy deoxy met Figure.6.55. Measured optical density spectra of oxy-hemoglobin (oxy), deoxyhemoglobin (deoxy) and met-hemoglobin (met).42

PAGE 288

265 The UV-VIS MAMW spectra of normal blood in saline and sickled blood in SMBS saline solution were measured. The results show spectral differences due to RBC shape and compositional changes. This confir ms the capability of the Integrated UV-VIS MAMW spectrometer to detect particle sh ape and compositional changes simultaneously. Potential applications in the area of biology and medici ne are clearly important. 6.5. Discussion The capability of the Integrated UV-VI S MAMW spectrometer to characterize micron and sub-micron size particles by th e simultaneous detect ion of the JPPD, especially particle size, shape, and compos ition, has been proved by the measurement of UV-VIS MAMW spectra of different standards including polystyrene spheres with sizes from 20 nm to 10 m, peanut-shaped particles, green fluorescent polystyrene spheres, normal whole blood sample, and sickled whole blood sample. As of Nov.14, 2004, INSPEC (physics reference search engine) sear ch results show no report of particle size, shape, and composition information detec tion by single angular light scattering measurement. All the UV-VIS MAMW spect ra measurement re sults show strong application possibilities in micr on and sub-micron size industri al particle characterization, biological particle characteri zation including bacteria and vi ruses, medical diagnostics (reduce the number of unnece ssary biopsies), nano-technolo gy, etc. In addition, it can provide a tool to test the light scattering theories that predict particle shape and composition, a map to direct research direction (for example, UV laser light scattering for nano-particle study and NIR laser light scattering for large si ze particle characterization),

PAGE 289

266 and a method of extracting information nece ssary for particle characterization (for instance, the derivation of the refractive indi ces of particles by comparing measured and simulated UV-VIS MAMW response surfaces). Moreover, transmission and frequency domain fluorescence spectra were able to be measured with the Integrated UV-VIS MAMW spectrometer. This opens the possibi lity of 3-dimensional MAMW spectroscopy (angular scattering, transmission, and fluorescen ce) for particle characterization with single instrument. In add ition, the Integrated UV-VI S MAMW spectrometer is inexpensive to build and the governing technology is very simple. It can also be portable and thus, possible to use in the fields. Data measured in remote places can be transferred and interpreted in real time online. No significant problems hinder the commercial production of the Integrat ed UV-VIS MAMW spectrometer. However, the research is just in th e beginning stage considering the immense possibilities of the MAMW sp ectrometer. There is room to innovate the current Integrated UV-VIS MAMW spectrometer prototype and to develop new multidimensional MAMW spectrometers. Meth ods of upgrading the Integrated UV-VIS MAMW spectrometer as well as developing a new multi dimensional MAMW spectrometer are discussed in the following section. 6.5.1. The Upgraded Integrated UV-VIS MAMW Spectrometer The capabilities of the Integrated UV-VI S MAMW spectrometer can be upgraded by incorporating: 1. Strong light source,

PAGE 290

267 2. UV-VIS achromatic lenses, 3. CCD detector with better sensi tivity and grating efficiency, 4. Motorized goniometer rotation, 5. Polarization optics and 6. Modular optics. The use of strong light sour ce, UV-VIS achromatic lenses and upgraded CCD detector will increase the overall sensitivity of the Integrated UV-VIS MAMW spectrometer drastically and widen the an alyzable wavelength ranges. Computer-controlled motorized goniometer rotation will reduce measurement ti me and the adoption of polarization optics will amplify the dimension of the UV-VIS MAMW scattering from current 1dimension to 16-dimension via the Mueller-matrix meas urement. Modular optics will secure the optimum use of the Integrated UV-VIS MA MW spectrometer capabilities because it allows prompt change of optics while retain ing the optical alignmen t. Currently, the UVVIS MAMW spectra can display up to 6840 data points per sample, 570 wavelengths from 230 nm to 800 nm range with 1 nm resolution and 12 angles from 5o to 60o range with 5o resolution. If this number is compared to the data points that can be typically obtained from a single wavelength scattering experiment for the measured angular range from 5o to 170o with 2o resolution, the UV-VIS MAMW s cattering yields about 80 times or more data points. However, if the listed upgrading is implemented, the UV-VIS MAMW spectra can display approxim ately 74,000 data points per sample (5o to 170o with 2o resolution and 200 nm to 1100 nm with 1 nm resolution) for unpolarized incident

PAGE 291

268 light and more than 1 million data points per sample with the use of polarization optics. Besides, modulation optics will allow the pr ompt measurement of transmission spectra including dichroism and fluorescence spect ra including anisotr opy. Consequently, the available information that will be obtai ned from the upgraded Integrated UV-VIS MAMW spectrometer will make this technology an ideal candidate for empirical particle characterization. Even nowadays, exact calcul ation of light scatte ring by particles with complicated shape or compos ition is virtually impossible. Therefore, for practical applications, especially for complicated biol ogical particle charac terization or medical diagnostics, methods of comparing the measured spectra to those stored in the spectral data base are actively sought. The upgraded UV-VIS MAMW integrated spectrometer is ideal methodology for this purpose because of its capacity to produce ample available information. Technologies that will be used for the construction of the upgraded UV-VIS MAMW integrated spectrometer are not far aw ay. They already exist or will be available in the near future primarily due to drive for the commercial applications of optics products. Therefore, the c onstruction of the upgraded Integrated UV-VIS MAMW spectrometer is not unrealistic. 6.5.2. Development of the Multidim ensional MAMW Spectrometer The Integrated UV-VIS MAMW spectrometer has been developed for the simultaneous detection of the JPPD of micr on and sub-micron size particles. Single UVVIS MAMW scattering measurement yields particle size, shap e, and composition

PAGE 292

269 information. However, it uses goniometer for the rotation of CCD detector and thus, true meaning of simultaneous detection of the JPPD is not accomplished. In addition, the capability of fluorescence and transmission spectra measurement was proven but these measurements were performed in serial due to the use of single CCD detector and the need to change the optics. Because of the te mporal changes of part icle characteristics, like settling or biological cell multiplication, simultaneous measurement of the JPPD is indispensable for particle characterization. Ways to accomplish the simultaneous dete ction of the JPPD and to optimize the use of existing spectroscopic techniques are described in the concept of the multidimensional (MD) MAMW spectrometer. Unlike the upgraded Integrated UV-VIS MAMW spectrometer, the MD MAMW spectro meter will employ technologies of the future. Consequently, the description is more conceptual than specific. Fig. 6.56 shows a schematic of the proposed MD MAMW spectrometer. It consists of a light source, a sample cell, a transmission and di ffraction detector, a fluorescence detector, and multiple numbers of multiangle (MA) scattering detectors. The details of the instrumentation will be determined based on the available state of the art technology at the time of construction. To be a multidimensional spectrometer, the MD MAMW spectrometer will be able to meas ure the following features simultaneously: 1. Clockwise and counter-clockwise multiangle scattering-multiple number of detectors will be installed. Each detection optics will have the capability of performing the measurement of the 16-element Mueller matrix.

PAGE 293

270 Figure 6.56. Schematic of a proposed MD MAMW spectrometer.

PAGE 294

271 2. Forward scattering and/or diffraction, 3. Backscattering, 4. Transmission including dichroism and 5. Fluorescence including anisotropy. The MD MAMW spectrometer will employ both multiwavelength light source and laser depending on applications. In addition, it w ill also perform the time-domain spectroscopy. Lastly, proper fields (electric, magnetic) will be applied to the sample cell to enhance the characterization capabilities. Once the MD MAMW spectrometer is developed, the limitation of its application will be determined by our imagination alone.

PAGE 295

272 CHAPTER 7. CONCLUSIONS This chapter describes the summary of the research, list of contributions and recommended work. The description of the instruments, which are required for the upgrading of the Integrated UV-VIS MAMW spectrometer and described in Section 6.5.1, is recapitulated in the section re lated to the recommended work. 7.1. Summary of Research The Integrated UV-VIS MAMWspectrometer has been developed to overcome the drawbacks of the prototype MAMW spect rometer and enhance the JPPD detection capabilities. The development of the Integrated UV-VIS MAMW spectrometer becomes possible through the use of fused silica UV lenses and integration time multiplexing (ITM). The adoption of UV lenses allows th e use of a broadband UV light source, intensifies the incident light to the sample ce ll, and together with the use of narrow width slits and ITM, enhances the low-angle scatte ring capabilities. The incorporation of ITM avoids detector saturation within the lineari ty of the CCD detector. By overcoming the limitations of the prototype MAMW spectro meter, measurements of fluorescence, transmission, and low-angle scattered light become possible. The Integrated UV-VIS

PAGE 296

273 MAMW spectrometer can perform angular scattering measurements starting at 4o with the simultaneous detection of multiwavelength light from 200 nm to 820 nm, UV-VIS transmission spectroscopy with wavelengths for analysis ranging from 200 nm to 820 nm, and frequency domain UV-VI S fluorescence spectroscopy. After the development of the current prototype Integrated UV-VIS MAMW spectrometer, possible sources of error were analyzed to ensure the reproducibility of the results. In addition, data corre ction and calibration procedures have been established to ensure the credibility of the experimental re sults. Refraction correction, scattering volume correction, and data normalization are the neces sary correction factors for the quantitative analysis of the Integrated UVVIS MAMW spectrometer results. The capabilities of the Integrated UV-VIS MAMW spectrometer were tested by measuring: the optical density, UV-VIS MAMW spectra, and the fluorescence spectra of polystyrene standards. The measured optic al density spectra show features of interference, reddening, ripple structures, and diffraction as well as absorption due to electronic transitions. The results were va lidated by comparing to measured spectra recorded with a reference spectromete r (HP 8453 diode array spectrometer). The measured UV-VIS MAMW spectra proved th at the JPPD can provide spectroscopic fingerprints of particles. The use of a broadband UV light source enabled the measurement of polystyrene standards as small as 20 nm and the incorporation of lowangle scattering measurements made it possible to distinguish polystyrene 8 m spheres and polystyrene 10 m spheres clearly. The possibility of the Integrated UV-VIS MAMW spectrometer to infer spherical par ticles and non-spherical particles from the

PAGE 297

274 measured UV-VIS MAMW spectra was tested using 1.85 m peanut-shaped polystyrene latex and 1.87 m polystyrene sphere particles. The resulting response surfaces showed a clear distinction in the UV region. Differences in the absorption of UV light and destructive interference due to differences in the beam path length re sulting from particle orientation appear to manifest in difference s in the measured spectra. This demonstrated the feasibility of the Integrated UV-VIS MAMW spectrometer to infer particle shape. Methods to measure the Mueller matrix a nd backscattering have been described to improve the identification of particle shape and orienta tion. The capabilities of the Integrated UV-VIS MAMW spectrometer to measure particle composition were tested by measuring the UV-VIS MAMW spectra of polystyrene 3.0 m spheres and green fluorescent polystyrene 3.0 m spheres. The resulting UV-VIS MAMW spectra show differences due to particle compos ition. A single UV-VIS MAMW scattering measurement provides information on the pa rticle size and composition simultaneously. Fluorescence spectra of green fluorescent polystyrene 3 m spheres were measured. The measured spectra have the features show n in the manufacturer-provided data. As an application to biological particle systems, the UV-VIS MAMW spectra of normal whole blood sample and sickled whole blood sample were measured. The resulting spectra showed sp ectral changes in the low angle region, probably due to changes in the morphology of the RBCs, the bl ueshift of hemoglobin absorption band at 415 nm, the reduction of protein and hemoglobin absorption bands and the disappearance of the hemoglobin doublet at around 570 nm in the sickled blood sample.

PAGE 298

275 These clearly demonstrated that particle shape and compositional changes can be detected simultaneously with the Inte grated UV-VIS MAMW spectrometer. Ways of upgrading the Integrated UV-VIS MAMW spectrometer and developing a new multidimensional MAMW spectrometer concept are provided to enhance the particle characterization capabilities. The upgrade of the Integrated UV-VIS MAMW spectrometer is based on the currently availa ble technologies while the development of a new multidimensional MAMW spectrometer implies the incorporation of future technologies. The upgraded Integrated UV-VIS MAMW spectrometer will generate more than one million data points per sample per measurement. In conclusion, the simultaneous measurement of the JPPD, particularly, size, shape, and composition using the Integrated UV-VIS MAMW spectrometer has a bright prospect for the characterization of micron a nd sub-micron size particle s. It has potential applications in the areas of industrial part icle characterization, biological particle characterization, medical diagnostics, nano-te chnology, etc. The possibility of readily upgrading the current In tegrated UV-VIS MAMW spectrometer and the new MD MAMW spectrometer make this prospect even brighter. 7.2. Contributions The contributions of this di ssertation research include 1. Development of the Integrated UV-VIS MAMW spectrometer that can perform static light scattering, transmissi on, and frequency domain fluorescence spectroscopy using single light s ource and single detector only.

PAGE 299

276 2. Development of data correction proce dures for the Integrated UV-VIS MAMW spectrometer. 3. Improved performance of the UV-VIS MAMW scattering experiments with starting wavelength from 200 nm and starting angle from 4o. 4. Improved performance of the UV-VIS transmission spectroscopy. 5. Improved performance of the frequency domain fluorescence spectroscopy. 6. The JPPD measurements of polystyrene standards. 7. Inference of particle shape by light scattering. 8. Observation of particle co mposition by light scattering. 9. Measurement of the UV-VIS MAMW spectra of normal whole blood sample and sickled whole blood sample. 10. Demonstrated that changes in particle shape and composition can be detected simultaneously. 11. Suggestions: how to measure the Mu eller matrix and backscattering. 12. Suggestions: how to upgrade the Inte grated UV-VIS MAMW spectrometer. 13. Suggestions: concept of a new multi dimensional MAMW spectrometer. 14. Recommendations: how to build the Integrated UV-VIS MAMW spectrophotometer. 15. Recommendations: use multiwavelength UV-VIS light spectroscopy for micron and sub-micron size particle characterization.

PAGE 300

277 7.3. Recommendations for Future Work To improve the capabilities of and/or to extend the applicat ion areas of the Integrated UV-VIS MAMW spectrometer, the following theoretica l, instrumental, and experimental studies are recommended. 1. Recommendations for theoretical work: a. Develop a simulation program that can provide results consistent with measured data, especially for sub-micr on size particles, are necessary for the establishment of quantitative ca libration and validation procedures. b. Use the T-Matrix approach to deve lop a generalized particle shape characterization software. c. Statistical validation of the measured multidimensional spectra using the above programs is recommended. By comparing the measured and the simulated results at all the angles and the wavelengths, it will improve the sensitivity and specificity of the characterization results. d. Investigation of light sc attering theory that can yield better interpretation of experimental results, especia lly related to particle shape and compositional changes, is recommended.

PAGE 301

278 2. Recommendations for hardware implementations: a. Computer controlled rotation of the goniometer arm is recommended. b. Ensuring the mechanical stability of the optical bench is strongly recommended. Currently, the optical co mponents are installed on the mini optical rails and reproducible precisi on alignment requires considerable amount of time and effort. c. Modular design of optics is r ecommended, so that, depending on experiment, proper optics can be substituted and/or added without the loss of alignment and reproducibility. d. Use of achromatic lenses that can cover a broad UV-VIS wavelength range is recommended. The use of ac hromatic lens may enhance the intensity of the detected light more than 10 times. In addition, the quality of the optics alignment can be improved by using achromatic lenses. e. Use of stronger light source with lo ng lifetime is recommended. Current tungsten-halogen bulb power is only 6.5 W. Hence, it does not provide adequate light intensities for measurem ent of the scattered light at large angles. The lifetime of the bulb is less than 80 hrs. Therefore, the bulb has to be changed at 80 hours and this may be a potential to lose optics alignment. f. Upgrading of the CCD detector grat ing is recommended. Current grating covers UV-VIS region only. By employing a grating that can be used from

PAGE 302

279 UV to NIR, large particle charac terization capability of the Integrated UVVIS MAMW spectrometer can be enhanced further. g. Improvement of the CCD detector sensitivity is recommended. The current detector has sensitivity of 12 bits only. The use of 16-bit detector will enlarge measurable scattering angle and reduce overall measurement time. h. Installation of polarizat ion optics is recommended. This will enable the measurement of the Mueller matr ix, dichroism, and fluorescence anisotropy simultaneously. 3. Recommended experiments: Measurement of the UV-VIS MAMW sp ectra of several synthetics and biological particles that can further demonstrate the pote ntial of this technology for particle characterization is recommended. For the systems discussed below the sample preparation and measurement protocols have to be developed. a. Measurement of the JPPD of microorganisms is recommended. Some microorganisms have similar sizes and compositions; as a result their characterization through extinction or fluorescence measurements alone may not be sufficient. Comparativ e studies using the UV-VIS MAMW spectrometer are desirable. b. Measurements of the UV-VIS MAMW spectra of polystyrene spheres larger than 10 m are recommended.

PAGE 303

280 c. Measurement of the UV-VIS MAMW spectra of TiO2 samples are recommended to obtain the information pertinent to particle shape changes. d. Fluorescence measurements using various samples are recommended.

PAGE 304

281 REFERENCES 1. C. E. Alupoaei. Modeling of the Transmi ssion Spectra of Microorganisms. M.S. Thesis, University of South Florida, Tampa, FL, December 2001. 2. C. P. Bacon. Simultaneous Characterizati on of Particle Properties (Size, Shape, and Composition) from the Development of the Multiangle-Multiwavelength Spectrometer System. Ph.D. Thesis, Univer sity of South Florida, Tampa, FL, August 1999. 3. C. Bacon and L. H. Garcia-Rubio, Multiangle-Multiwavelength detection for particle characterization, Particle Size Distribution III: Assessment and Characterization, T. Provder, Ed., ACS Symposium Series 693 30-38 (American Chemical Society, Washington DC, 1998). 4. W. S. Bickel and W. M. Bailey, Stokes vectors, Mueller matrices, and polarized light scattering, Am. J. Phys. 53 468-478 (1985). 5. W. S. Bickel, J. F. Davidson, D. R. Huffman, and R. Kilkson, Application of polarization effects in light sca ttering: A new biophysical tool, Proc. Nat. Acad. Sci. USA 73 486-490 (1976). 6. W. S. Bickel and M. E. Stafford, Bio logical particles as irregularly shaped scatterers, Light Scattering by Irre gularly Shaped Particles, D. W. Schuerman, Ed., 299-305 (Plenum Press, New York, NY, 1980). 7. C. F. Bohren and D. R. Huffman, Absorption and Scattering of Light by Small Particles (John Wiley & Sons, New York, NY, 1983). 8. B. A. Brice, M. Halwer, and R. Sp eiser, Photoelectric light-scattering photometer for determining high molecular weights, Journal of the Optical Society of America 40 768-778 (1950). 9. F. V. Bright, Modern molecular fluorescence spectroscopy, Focus on Analytical Spectrometry: A Compendium of App lied Spectroscopy Focal Point Articles (1994-1997) J. A. Holcombe, G. M. Hieftje, and V. Majidi Ed., 104-109 (Society for Applied Spectroscopy, Frederick, MD, 1998). 10. Coulter Corporation, Particle Character ization LS Short Course Note (1997). 11. G. A. Daland and W. B. Castle, A si mple and rapid method for demonstrating sickling of the red blood cel ls: the use of reducing agents, J. Lab. Clin. Med. 33 1082-1088 (1948).

PAGE 305

282 12. N. Damaschke, G. Gouesbet, G. Grehan and C Tropea, Optical techniques for the characterization of non-s pherical and non-homogeneous particles, Meas. Sci. Technol, 9 137-140 (1998). 13. http://www.dukescientific.com. 14. C.E. Dungey and C.F. Bohren, Light scattering by nonspherica l particles: a refinement to the coupled-dipole method, Journal of the Optical Society of America A 8, 81-87 (1991). 15. G. E. Elicabe and L. H. Garcia-Rubio, Latex particle size distribution from turbidimetry using i nversion techniques, J. Coll. Int. Sci. 129 192-200 (1989). 16. G. E. Elicabe and L. H. Garcia-Rubio, Latex particle size distribution from turbidimetric measurements, Adv. Chem. Ser. 227 83-104 (1990). 17. X. Fu, Detection and Identification of Microorganisms using a combined Flow Field-Flow Fractionation Spectroscopy. Ph.D. Thesis University of South Florida, Tampa, FL, October 2003. 18. L. H. Garcia-Rubio, The effect of mol ecular size on the absorption spectra of macromolecules, Macromolecules 20 3070-3075 (1987). 19. L. H. Garcia-Rubio at University of South Florida, Pr ivate Communication. 20. P. C. Gray, I. R. Shokair, S. E. Rosenthal, G. C. Tisone, J. S. Wagner, L. D. Rigdon, G. R. Siragusa, and R. J. Hein en, Distinguishabil ity of biological material by use of ultraviolet multispectral fluorescence, Appl. Opt. 37, 60376041 (1998). 21. J. M. Greenberg, Focusi ng in on particle shape, Light Scattering by Irregularly Shaped Particles, D. W. Schuerman, Ed., 7-24 (P lenum Press, New York, NY, 1980). 22. R. O. Gumprecht and C. M. Sliepcevic h, Scattering of light by large spherical particles, J. Phys. Chem. 57 90-95 (1953). 23. F. S. Harris, G. C. Sherman, and F. L. Morse, Experimental comparison of scattering of coherent and incoherent light, IEEE Transactions on Antennas and Propagation AP-15 141-147 (1967). 24. L. Hespel, A. Delfour, and B. Guillame, Mie light-scattering gr anulometer with an adaptive numerical filtering method. II. Experiments, Appl. Opt. 40 974-985 (2001).

PAGE 306

283 25. A. H. Hielscher, A. A. Eick, J. R. Mouran t, D. Shen, J. P. Freyer, and I. J. Bigio, Diffuse backscattering Mueller matri ces of highly sc attering media, Opt. Express. 1 441-453 (1997). 26. T. Inagaki, E. T. Arakawa, R. N. Hamm, and M. W. Williams, Optical properties of polystyrene from the near-infrared to the x-ray region and convergence of optical sum rules, Phy. Rev. B 15 3243-3253 (1977). 27. J. D. Jackson, Classical Electrodynamics 2nd ed., Chapter 7 (John Wiley & Sons, New York, NY, 1975). 28. B. R. Jennings and H. Plummer, L ight-scattering photometer calibration, Brit. J. Appl. Phys. (J. Phys. D) 1 1201-1209 (1968). 29. H. Jiang, J. Pierce, J. Kao, and E. Sevi ck-muraca, Measurement of particle-size distribution and volume fraction in c oncentrated suspensions with photon migration techniques, Appl. Opt. 36 3310-3318 (1997). 30. A. R. Jones, Light scattering for particle characterization, Progress in Energy and Combustion Science 25 1-53 (1999). 31. M. Kerker, The Scattering of Light and Other Electromagnetic Radiation (Academic Press, New York, NY, 1969). 32. J. R. Lakowicz, Principles of Fluorescence Spectroscopy 2nd ed. (Kluwer Academic / Plenum Publishers, New York, NY, 1999). 33. P. Latimer and P. Barber, Scatteri ng by ellipsoids of revolution. A comparison of theoretical methods, J. Coll. Int. Sci. 63 310-316 (1978). 34. The Mathworks Inc., MATLAB 5.3. (1999). 35. Y. D. Mattley. An Investigation of the Spectroscopic Properties of Platelets During Activation and Storage: Implementa tion of a New Interpretation Model. Ph.D. Thesis, University of South Florida, Tampa, FL, December 2000. 36. M. P. Menguc and S. Manickavasagam, C haracterization of size and structure of agglomerates and inhomoge neous particles vi a polarized light, International J. of Engineering Science. 36, 1569-1593 (1998). 37. Microsoft Corporation, Microsoft Encarta Encyclopedia 2002 (2002).

PAGE 307

284 38. M. I. Mishchenko, Light scattering by si ze-shape distributions of randomly oriented axially symmetric particles of a size comparable to a wavelength, Appl. Opt. 32, 4652-4666 (1993). 39. Newport Corporation, The Newport Resource 2003 (2003). 40. http://www.newrise-llc.com/fused-silica.html. 41. A. Nonoyama. Using multiwavele ngth UV-visible spectroscopy for the characterization of red blood cells: An investigation of hypochromism. Ph.D. Thesis, University of South Florida, Tampa, FL, November 2004. 42. A. Nonoyama at University of South Florida, Private Communication. 43. Ocean Optics, Inc., 2 Product Catalog (2002). 44. Biomedical Topical Meetings OSA Technical Digest (Opt ical Society of America, Washington DC, 2000). 45. F. L. Pedrotti and L. S. Pedrotti, Introduction to Optics Chapter 27 (Prentice-Hall International, Englewood Cliffs, NJ, 1993). 46. http://www. Photocor.com. 47. UV Atlas of Organic Compounds, Volume II. (Plenum Press, New York, NY, 1966). 48. Polysciences Inc., Private Communication. 49. A. Quirantes, Light scattering prope rties of spheroidal coated random orientation, J. Quant. Spectrosc. Radiat. Transfer 63, 263-275 (1999). 50. B.E.A. Salek and M.C. Teich, Fundamentals of Photonics (John Wiley & Sons, New York, NY, 1991). 51. H. Schnablegger and O. Gl atter, Sizing of colloidal particles with light scattering: corrections for beginning multiple scattering, Applied Optics 34 3489-3501 (1995). 52. P. R. Smith, O. Kusmartseva, and R. Naimimohasses, Evidence for particleshape sensitivity in the correlation between polarization states of light scattering, Opt. Lett. 26 1289-1291 (2001).

PAGE 308

285 53. K. A. Stacey, Light Scattering in Physical Chemistry Chapter 3 (Butterworths Scien tific Publications, London, 1956). 54. Starna Cells Inc., Private Communication. 55. G. L. Stephens, Scattering of plan e waves by soft obstacles: anomalous diffraction theory for circular cylinders, Appl. Opt. 23 954-959 (1984). 56. http://www.thermo-oriel.com. 57. Y. Tomimatsu and K. J. Palmer, Refl ection corrections fo r light-scattering measurements for various ce lls with the Brice-type photometer, J. Phys. Chem. 67 1720-1722 (1963). 58. H. Utiyama, Light scattering instruments, Light Scattering From Polymer Solutions M. B. Huglin, Ed., 41-60 (Academic Press, London, 1972). 59. H. Utiyama, Calibration and correction factors, Light Scattering From Polymer Solutions M. B. Huglin, Ed., 61-88 (Academic Press, London, 1972). 60. H. C. van de Hulst, Light Scattering by Small Particles (Dover, New York, NY, 1981). 61. P. Walstra, Discussion of errors in turbidimetry, Brit. J. Appl. Phys. 16 11871192 (1965). 62. R. Walters, Series OEM GUIDE, Ocean Optics, Inc. (2000). 63. R. Walters at Ocean Optics, Inc., Private Communication. 64. J. Wang and F. R. Hallett, Spherical part icle size determination by analytical inversion of the UV-visible-NIR extinction spectrum, Appl. Opt. 35 193-197 (1996). 65. L. S. Wright, A. Chowdhury, and P. Russ o, Static light scattering instrument for rapid and time-resolved particle sizi ng in polymer and colloid solutions, Rev. Sci. Instrum. 67 3645-3648 (1996). 66. P. J. Wyatt, Review: Light scattering and the absolute characterization of macromolecules, Analytica Chimica Acta 272 1-40 (1993).

PAGE 309

286 67. R. Xu, Particle size distribution analysis using light scattering, Liquid and Surface-Borne Particle Measurement Handbook, J. Z. Knapp, T. A. Barber, and A. Lieberman, Ed., 745-777 (Marcel Dekker, New York, NY, 1996). 68. E. Zurek at University of Sout h Florida, Private Communication.

PAGE 310

287 APPENDICES

PAGE 311

288 APPENDIX A: MATLAB Program for UV-VIS MAMW Spectra Plot of the Measured Results % UV-VIS MAMW % Yong-Rae Kim % The following program performs UV-VIS MAMW spectra plots clear; clc; % Load data files for sample % Sample v case % up incident beam spectrum, ur reference beam spectrum % vssample spectrum measured at 0o, vaOD spectrum % vr reference scatteri ng, vssample scattering load up.txt;load ur.txt; load va.txt;load vs.txt; load vr05.txt;load vr10.txt;load vr15.txt;load vr20.txt; load vr25.txt;load vr30.txt;load vr35.txt; load vs05.txt;load vs10.txt;load vs15.txt;load vs20.txt; load vs25.txt;load vs30.txt;load vs35.txt; % Use the wavelength interval that is designated by S2000 production spectrometer. wavelength = up(:,1); % Define the wavelength interval for plot. ind1 = (wavelength>=230 ) & (wavelength<=800); x = wavelength(ind1); % Use refraction corrected angles as scattering angles. y = [4;7;11;14;18;21;25];

PAGE 312

289 APPENDIX A: (Continued) % Define the normalization standard file ur = ur(ind1,2); rm = ur; % Define reference scattering files vr05 = vr05(ind1,2); vr10 = vr 10(ind1,2); vr15 = vr15(ind1,2); vr20 = vr20(ind1,2); vr25 = vr 25(ind1,2); vr30 = vr30(ind1,2); vr35 = vr35(ind1,2); kr05 = vr05; kr10 = vr10; kr15 = vr15; kr20 = vr20; kr25 = vr25; kr30 = vr30; kr35 = vr35; % Define sample scattering files vs05 = vs05(ind1,2); vs10 = vs 10(ind1,2); vs15 = vs15(ind1,2); vs20 = vs20(ind1,2); vs25 = vs 25(ind1,2); vs30 = vs30(ind1,2); vs35 = vs35(ind1,2); ks05 = vs05; ks10 = vs10; ks15 = vs15; ks20 = vs20; ks25 = vs25; ks30 = vs30; ks35 = vs35; % Define the UV-VIS MAMW scattering intensities s5 = ks05 kr05; s10 = ks 10 kr10; s15 = ks15 kr15; s20 = ks20 kr20; s25 = ks25 kr25; s30 = ks30 kr30; s35 = ks35 kr35; % Multiply volume correction factors % 3mm width slit case s5 = (1.0).*s5; s10 = (1.08).*s10; s15 = (1.12).*s15; s20 = (1.17).*s20; s25 = ( 1.24).*s25; s30 = (1.40).*s30; s35 = (1.51).*s35;

PAGE 313

290 APPENDIX A: (Continued) % Multiply integration time ratio % The ratio shown here is example of sample v case. E ach sample has its own ratio. rm = (5000/8).*rm ; s5 = 2.*s5 ; s10 =2.*s10; s15 = 2.*s15 ;s20 = 2.*s20 ; s25 = 1.*s25 ; s30 = 1.*s30 ; s35 = 1.*s35 ; % Normalize the scattering intensity w ith the reference beam spectrum rm % Check matrix dimension sm = [s5'./rm';s10'./rm';s15'./rm';s20'./rm'; s25'./rm';s30'./rm'; s35'./rm']; % UV-VIS MAMW response surface plot [x_grid, y_grid] = meshgrid(x,y); z=log10(sm); surf(x_grid, y_grid, z); xlabel('Wavelength (nm)'), ylabel('Observation Angle (deg)'); zlabel('Intensity log10 (I/Ii)'); % To make smooth plot, use interpolation shading interp; % Contour plot of the UV-VIS MAMW response surface contour(y_grid, x_grid, z); pcolor(y_grid, x_grid, z); xlabel('Wavelength (nm)'), ylabel('Observation Angle (deg)'); shading interp; colorbar;

PAGE 314

291 APPENDIX B: Method of Correcting the Measured UV-VIS MAMW Spectra for the Continuously Varying Refractive Indices of Spectrosil Quartz The refractive indices of Spectrosil quartz for the whole wa velength range of the incident light can be derived using a polynom ial that can fit the known refractive indices of Spectrosil quartz at selected wavelengths. MATLAB or other mathematical software can be used to find this polynomial. On ce the sought-for polynomial is acquired, the refractive indices of Spectrosilquartz at any wavelengths can be calculated using this polynomial. Spectrosil quartz is a kind of synthetic fused silica. As a result, the refractive indices of fused silica can be used to figure out the sought-f or polynomial. Table B.1 lists the known refractive indices of fused silica.40 DATDEMO nonlinear data fitting code of the MATLAB version 5.3 was used to process this information to acquire the sought-for polynomial.34 The resulting polynomial sq rip is x x sq rie c e c p2 12 1 (B.1) where 2 1and c c are linear parameters and 2 1and are nonlinear parameters. There are several methods to calc ulate the parameters 2 1, c c, 2 1and Among them, BroydenFletcher-Golfarb-Shanno method was chosen to perform the data fitting and Mixed Polynomial Interpolation method was selected for line search.34 The results are

PAGE 315

292 APPENDIX B: (Continued) Table B.1. The refractive indices of fused silica.40 Wavelength (nm) Refractive index 200 1.55051 220 1.52845 250 1.50745 300 1.48779 320 1.48274 360 1.47529 400 1.47012 450 1.46557 488 1.46302 500 1.46233 550 1.46008 588 1.45860 600 1.45804 633 1.45702 650 1.45653 700 1.45529 750 1.45424 800 1.45332 850 1.45250 900 1.45175

PAGE 316

293 APPENDIX B: (Continued) 4608 4736868193 1 2085 1280668348 12 1 c c (B.2) and 432 0000174875 0 5957 0131969832 02 1 (B.3) The accuracy of the acquired refractive indices using the above polynomial was examined by comparing them with the manuf acturer-provided data. The results that are summarized in Table B.2 show excellent ag reement between the calculated and the known refractive indices of Spectrosil quartz. The maximum error is only 0.14 % at 1000 nm. Fig.B.1 shows that the known refractive indices of Spectrosil quartz and fused silica accurately fit for the curve obtain ed by plotting this polynomial Eq. (B.1). This further confirms the validity of the met hod of acquiring the unknown refractive indices of Spectrosil quartz using a proper polynomial. Theref ore, the calculated refractive indices can be used for the correction of the meas ured UV-VIS MAMW spectra for the whole wavelength range of the incident light. After finding the unknown refractive indices of Spectrosil quartz, the refraction correction and the sc attering volume correction were performed again to elucidate whether the corrections using the continuously changing refrac tive indices are necessary

PAGE 317

294 APPENDIX B: (Continued) or not. The manufacturer-provi ded refractive index of 1.551 at 200 nm and the calculated refractive index of 1.4532 at 800 nm were used. The results are summarized in Table B.3 and Table B.4 for the refraction (scattering angle) correction and Table B.5 through Table B.8 for the scattering volume correcti on. These results are compared with the results shown in Table 5.2, Table 5.3, and Ta ble 5.4 that are acquired using the median value of the manufacturer-provide d refractive index of Spectrosil quartz, 1.504 at 254 nm. For refraction correction, the maximum erro r is 2.8% if the refractive index for 200 nm is used and 4.2 % if the refractive index for 800 nm is used. For the scattering volume correction the results are consistent if the refractive index for 200 nm is used and the maximum error is 4.0 % if the refractive index for 800 nm is used. In summary, if the median value of the continuously varying re fractive indices of Spectrosil quartz is chosen for the correction of the measured UV-VIS MAMW spectra, the errors due to differences in the refractive indices of Spectrosil quartz for the wavelength range from 200 nm to 800 nm is smaller than 5 %. Therefore, it can be concluded that the correction of the m easured UV-VIS MAMW spectra for the continuously varying refrac tive indices of Spectrosil quartz is unnecessary.

PAGE 318

295 APPENDIX B: (Continued) Table B.2. The refractive indices of Spectrosil quartz.54 Wavelength (nm) Manufacturerprovided Refractive index Calculated Refractive Index Error (%) 200 1.551 1.549 0.13 254 1.506 1.507 0.07 300 1.488 1.487 0.07 400 1.470 1.469 0.07 600 1.458 1.459 0.07 1000 1.450 1.448 0.14

PAGE 319

296 APPENDIX B: (Continued) 200 300 400 500 600 700 800 900 1000 1.44 1.46 1.48 1.5 1.52 1.54 1.56 1.58 Figure B.1. Refractive indices of fused silica ( __ ) and Spectrosil quartz (O). The data in Reference 54, Reference 40, and the calculated re sult of Eq. (B.1) were used for plotting. Wavelength (nm) Refractive Index O: Spectrosil quartz, manufacturer provided. : Fused silica, from Reference 40. __ : Fused silica, calculated.

PAGE 320

297 APPENDIX B: (Continued) Table B.3. Differences in observation angl e and the refractive-inde x-corrected actual scattering angle. The manufacturerprovided refractiv e index of 1.551 at 200 nm was used for the correction. The results are compared with the data shown in Table 5.2. Observation angle a (deg) Refracted angle by cuvette wall q (deg) Corrected scattering angle (deg) Error (%) 5 3.2 3.5 2.8 10 6.4 7.1 1.4 15 9.6 10.6 2.8 20 12.7 14.1 2.1 25 15.8 17.6 1.7 30 18.8 20.9 2.3 35 21.7 24.2 2.0 Table B.4. Differences in observation angl e and the refractive-inde x-corrected actual scattering angle. The calculated refract ive index of 1.4532 at 800 nm was used for the correction. The results are compar ed with the data shown in Table 5.2. Observation angle a (deg) Refracted angle by cuvette wall q (deg) Corrected scattering angle (deg) Error (%) 5 3.4 3.7 2.8 10 6.9 7.5 4.2 15 10.3 11.3 3.7 20 13.6 14.9 3.5 25 16.9 18.5 3.4 30 20.1 22.1 3.3 35 23.2 25.5 3.2

PAGE 321

298 APPENDIX B: (Continued) Table B.5. Calculated scattering volume for 3 mm width slit. The manufacturerprovided refractive index of 1.551 at 200 nm was used. The results are compared with the data shown in Table 5.3. Observation Angle a (deg) Scattering Angle (deg) Scattering volume (mm3) Error (%) 5 4 560 0 10 7 520 0 15 11 500 0 20 14 480 0 25 18 450 0 30 21 400 0 35 24 370 0 Table B.6. Calculated scattering volume for 6 mm width slit. The manufacturerprovided refractive index of 1.551 at 200 nm was used. The results are compared with the data shown in Table 5.4. Observation Angle a (deg) Scattering Angle (deg) Scattering volume (mm3) Error (%) 5 4 1150 0 10 7 1120 0 15 11 1070 0 20 14 1030 0 25 18 1020 0 30 21 1010 0 35 24 970 0

PAGE 322

299 APPENDIX B: (Continued) Table B.7. Calculated scatteri ng volume for 3 mm width slit. The manufacturerprovided refractive index of 1.4532 at 800 nm was used. The results are compared with the data shown in Table 5.3. Observation Angle a (deg) Scattering Angle (deg) Scattering volume (mm3) Error (%) 5 4 560 0 10 8 520 0 15 11 500 0 20 15 470 2.1 25 19 460 2.2 30 22 410 2.5 35 26 385 4.0 Table B.8. Calculated scattering volume for 6 mm width slit. The manufacturerprovided refractive index of 1.4532 at 800 nm was used. The results are compared with the data shown in Table 5.4. Observation Angle a (deg) Scattering Angle (deg) Scattering volume (mm3) Error (%) 5 4 1150 0 10 8 1130 0.9 15 11 1070 0 20 15 1050 1.9 25 19 1030 1.0 30 22 1000 1.0 35 26 970 0

PAGE 323

ABOUT THE AUTHOR Yong-Rae Kim received a Bachelor of Science Degree in Physics from Kyung Hee University at Seoul, South KOREA in 1984. Then he continued his research in the area of Nuclear Physics and got a Master of Science Degree in Physics from Kyung Hee University in 1986. After working a few year s as a Graduate A ssistant at Kyung Hee University, he entered the Master of Science program at the University of South Florida in 1994. He majored Laser Spectroscopy and rece ived a Master of Science Degree in Physics from the University of South Flor ida in 1996. He continued his study for a Ph.D. Degree in Applied Physics and engaged in the research of Biom edical Spectroscopy.