USF Libraries
USF Digital Collections

Identification of the nucleation locus in emulsion polymerization processes

MISSING IMAGE

Material Information

Title:
Identification of the nucleation locus in emulsion polymerization processes
Physical Description:
Book
Language:
English
Creator:
Shastry, Vineet, 1973-
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
droplet characterization
spectroscopy
liquid-liquid systems
surfactants
dispersions
Dissertations, Academic -- Chemical Engineering -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Particle Nucleation is the forcing function in the Emulsion Polymerization processes and it plays an important role in dictating the final properties of the latex produced. Identification of the main nucleation sites and characterizing them in terms of their size and composition is important for elucidating the mechanism of particle nucleation. This research focuses on identifying the most likely nucleation locus in emulsion polymerization processes by characterizing the initial conditions of the reaction mixture. In order to achieve this objective, a methodology was devised, which used a non-reacting model emulsion system instead of the original emulsion. The model emulsion system selected has the same dispersion properties as that of the monomer emulsion system, but different optical properties. The model emulsion system enabled the study of the distribution of the emulsifier using Uv vis spectroscopy. This approach also eliminated the time constraint associated with sampling during a polymerization reaction. A quantitative deconvolution using the turbidity equation, was done on the transmission Uv vis spectra of the emulsions. This enabled the characterization of the emulsions in terms of their particle size distribution, particle number and the composition of the droplet populations comprising them. The studies conducted provide the experimental evidence for a previously unidentified nano-droplet population of size range 30 to 100nm in diameter. To further support this experimental evidence, calculations were performed to obtain the emulsifier distribution over the nano-droplet population. The calculations suggest the probability of existence of the nano-droplet population to be much higher than the probability of the existence of the swollen micelles. The results, depending upon the emulsification conditions, indicate the presence of about 15 % to 80% of the dispersed phase in the nano-droplet population. The large interfacial area offered by the nano-droplet population due to their high particle numbers and high percentage of the dispersed oil phase in them, make them the most probable particle nucleation loci in emulsion polymerization processes. Designed experiments were performed to experimentally observe the changes in the nano-droplet populations. The effects of the process variables, namely pH, surfactant concentration and temperature, on the size and compositional characteristics of the nano-droplet population were investigated. The results suggested that the surfactant to oil ratio was the dominating factor governing the size and the weight percent of the dispersed phase in the nano-droplet population.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2004.
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 Vineet Shastry.
General Note:
Includes vita.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 224 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 - 001455191
oclc - 54702708
notis - AJP4865
usfldc doi - E14-SFE0000221
usfldc handle - e14.221
System ID:
SFS0024917:00001


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 001455191
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 040305s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000221
035
(OCoLC)54702708
9
AJP4865
b SE
SFE0000221
040
FHM
c FHM
090
TP145
1 100
Shastry, Vineet,
d 1973-
0 245
Identification of the nucleation locus in emulsion polymerization processes
h [electronic resource] /
by Vineet Shastry.
260
[Tampa, Fla.] :
University of South Florida,
2004.
502
Thesis (Ph.D.)--University of South Florida, 2004.
504
Includes bibliographical references.
500
Includes vita.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 224 pages.
520
ABSTRACT: Particle Nucleation is the forcing function in the Emulsion Polymerization processes and it plays an important role in dictating the final properties of the latex produced. Identification of the main nucleation sites and characterizing them in terms of their size and composition is important for elucidating the mechanism of particle nucleation. This research focuses on identifying the most likely nucleation locus in emulsion polymerization processes by characterizing the initial conditions of the reaction mixture. In order to achieve this objective, a methodology was devised, which used a non-reacting model emulsion system instead of the original emulsion. The model emulsion system selected has the same dispersion properties as that of the monomer emulsion system, but different optical properties. The model emulsion system enabled the study of the distribution of the emulsifier using Uv vis spectroscopy. This approach also eliminated the time constraint associated with sampling during a polymerization reaction. A quantitative deconvolution using the turbidity equation, was done on the transmission Uv vis spectra of the emulsions. This enabled the characterization of the emulsions in terms of their particle size distribution, particle number and the composition of the droplet populations comprising them. The studies conducted provide the experimental evidence for a previously unidentified nano-droplet population of size range 30 to 100nm in diameter. To further support this experimental evidence, calculations were performed to obtain the emulsifier distribution over the nano-droplet population. The calculations suggest the probability of existence of the nano-droplet population to be much higher than the probability of the existence of the swollen micelles. The results, depending upon the emulsification conditions, indicate the presence of about 15 % to 80% of the dispersed phase in the nano-droplet population. The large interfacial area offered by the nano-droplet population due to their high particle numbers and high percentage of the dispersed oil phase in them, make them the most probable particle nucleation loci in emulsion polymerization processes. Designed experiments were performed to experimentally observe the changes in the nano-droplet populations. The effects of the process variables, namely pH, surfactant concentration and temperature, on the size and compositional characteristics of the nano-droplet population were investigated. The results suggested that the surfactant to oil ratio was the dominating factor governing the size and the weight percent of the dispersed phase in the nano-droplet population.
590
Adviser: Garcia-Rubio, Luis
653
droplet characterization.
spectroscopy.
liquid-liquid systems.
surfactants.
dispersions.
690
Dissertations, Academic
z USF
x Chemical Engineering
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.221



PAGE 1

Identification Of The Nucleation Locus In Emulsion Polymerization Processes by Vineet Shastry A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemical Engineering College of Engineering University of South Florida Major Professor: Luis Garcia-Rubio, Ph.D. Aydin Sunol, Ph.D. Oscar D. Crisalle, Ph.D. Julie Harmon,Ph.D. Wilfrido Moreno, Ph.D Date of Approval: January 15, 2004 Keywords: droplet characteriza tion, spectroscopy, liquid-liquid systems, surfactants, dispersions Copyright 2004 Vineet Shastry

PAGE 2

i TABLE OF CONTENTS LIST OF TABLES.vi LIST OF FIGURES..vii LIST OF SYMBOLS.....xv ABSTRACT...xviii CHAPTER 1: INTRODUCTION....1 1.1 Motivation.........1 1.2 Significance of the Emulsion Polymerization Process.....2 1.3 A Brief Description of Emulsi on Polymerization Process...........4 1.3.1 Important Definitions.............4 1.3.2 Emulsion Polymerization Process......5 1.4 Significance of the Research....9 1.5 Outline of the Dissertation......10 1.6 Contributions..11 1.6.1 Identification of the Main Locus for Nucleation in Emulsion Polymerization.................................................................11 1.6.2 Characteri zing the Emulsifier Distribution as a Function of Initial Emulsification Conditions in an Emulsion Polymerization Reaction..............................................................11 1.6.3 Implementation of Spectroscopic Techniques to Identify the Nucleation Mechanism....12 1.6.4 Emulsion Characterization...12

PAGE 3

ii CHAPTER 2: LITERATURE REVIEW.........13 2.1 Introduction...................13 2.2 Proposed Nucleation Mechanisms and their Limitations 2.3 Direction of Ongoing Research at USF.....26 CHAPTER 3: PROB LEM DEFINITION ....... 3.1 Introduction...........31 3.2 Approach...............31 3.2.1 Preliminary Studies....33 3.2.2 Algorithms Implemented.......33 3.2.3 Simulations and Experiment al Analysis Performed..........36 3.3 Inferences Drawn from the Simulation Studies....................38 3.4 Hypothesis Proposition.....38 3.4.1 Hypothesis No 1.........39 3.4.2 Hypothesis No 2.........39 3.4.3 Hypothesis No 3.........39 3.5 Relevance of the Proposed Hypotheses for Identifying Nucleation Mechanism........39 3.5.1 Hypothesis No 1.....39 3.5.2 Hypothesis No 2.........40 3.5.3 Hypothesis No 3.....42 3.6 Conclusion.........45 Chapter 4: EXPERIMENTAL WORK UNDERTAKEN............58 4.1 Introduction.......58

PAGE 4

iii 4.2 Rationale for Using the Model System..........59 4.3 Experiments with Model System...61 4.3.1 Materials and Methods.......61 4.3.1.1 Materials.. 4.3.1.2 Methods.......62 4.3.2 Equipment and Experimental Setup.......64 4.3.3 Experimental Design Strategy....65 4.3.4 Relevance of the Experimental Variables.......66 4.3.4.1 Effect of Surfactan t Concentration on Initial Distribution of the Particle Populations.......66 4.3.4.2 Effect of Temperature on Initial Distribution of the Particulate Distribution of the Populations......67 4.3.4.3 Effect of pH on Initial Di stribution of the Particulate Populations..67 4.3.5 Experimental Design Strategy....68 4.3.5.1 Surfactant Concentration. 4.3.5.2 Levels of pH .......70 4.3.5.3 Temperature.71 4.3.6 Experimental Procedure..........71 4.4 Data Analysis.............72 CHAPTER 5: RESULTS AND DISCUSSIONS....82 5.1 Introduction.... 5.2 Effects of Manipulated Variables 5.2.1 Effect of Surfactant Concentration.

PAGE 5

iv 5.2.2 Effect of Temperature...................86 5.2.3 Effect of pH......87 5.3 Surfactant Stabilized Area Analysis........93 5.4 Conclusion...98 CHAPTER 6: CONCLUSION AND RECOMMENDED FUTURE WORK.......110 6.1 Conclusions................111 6.2 Contributions. 6.3 Recommendations for Future Work..113 REFERENCES....114 APPENDICES.....119 Appendix A Size Dependent Scattering and Absorption Particle Characteristics........................120 Appendix B: Spectroscopic Calculations...........124 B.1 Introduction...........124 B.2 Sample Calculations..........126 B.3 Inferences Drawn from the Results Obtained from the Spectral Manipulations Appendix C: Optical Properties......134 C.1 Introduction.......134 C.2 Characteri zation of the Optical Properties....136 C.2.1 Optical Properties for Decane (Dispersed Phase)......136 C .2.2 Optical Properties for Water (Continuous Phase)......139 C.2.3 Optical Properties for the Emulsi fier Sodium Dodecyl Benzene Sulfonate (SDBS).....141

PAGE 6

v C.2.3.1 Calibration of the Refractometer.............141 C.2.3.2 Estimation of Refractiv e Index for SDBS...........141 C.2.3.3 Estimation of Absorption Coefficients for SDBS.......143 C.2.3.4 Use of the Model Molecule for Estimation of the SDBS Optical Properties.........143 C.2.3.5 Obtaining the Optical Properties of PTSA..........145 Appendix D: Behavior of Surfactant Micellar Solutions in Different Environments...........152 Appendix E: Transmission Spectra of Emulsions Under Different Emulsification Conditions.. ............159 Appendix F: Experimental Results fo r Determining the Stability of the Emulsion......................168 Appendix G: Calculated Particle Size Distribution of Small and Large Droplet Populations.....173 Appendix H: Preliminary Results from the Emulsion Polymerization Reactions............177 H.1 Materials and Methods. H.2 Experiment al Setup and Reactor Configuration.......178 H.3 Experimental Procedure. .180 H.4 Sensor Array to Monitor the Emulsion Polymerization Reaction........181 H.5 Results Obtained From the Polymerization Reaction...184 H.6 Temperature Control System....186 Appendix I: Protocol for Oper ating the Dilution System............198 ABOUT THE AUTHOR....End Page

PAGE 7

vi LIST OF TABLES Table 1.1 Sale of the Emulsion Polyme rization Products in the Year 1999............3 Table 4.1 Comparison of the Physical Propertie s of Styrene, Butyl Methacrylate and Decane....62 Table 4.2 Experimental Conditions.... Table 5.1 Results Obtained by Deconvoluti ng the Uv vis Transmission Spectra of the Emulsion Before Stability....83 Table 5.2 Effects of the Manipulated Vari ables on the Nano-droplet Population of the Dispersed Phase........89 Table 5.3 Summary of E xperimental Results.....90 Table 5.4 Comparison of the Online and O ffline Experimental Results along with the Results Obtained for the Styrene Emulsion..............92 Table 5.5 Comparison of the Surfactant Distribution Over Each Population of the Dispersed Phase........ Table 5.6 Comparison of the Area Required to be Stabilized per Surfactant Molecule to Achieve a Dispersed Phase of Given Size Characteristics.....98 Table C.1Refractive Index of SDBS Solution at Di fferent Concentrations.....142 Table C.2 Concentration of PTSA Solution.............145 Table C.3 Concentration of PTSA Solutions (further diluted)......... Table C.4 Concentration Vs Refractive Index for PTSA Solutions.............147 Table H.1Recipe and Experimental Conditions for the Emulsion Polymerization Reactions.........178 Table H.2 Preliminary Results Obtain ed by the Deconvoluting the Spectral Signal Obtained at the End of the Reaction (for 100% Conversion).......186

PAGE 8

vii LIST OF FIGURES Figure 1 Representation of the La tex Mixture for Emulsion Polymerization...9 Figure 2 Change in the Shape of the Simulated Transmission Spectra of the Nano-droplet Populati on of Styrene in Water Emulsion as a Function of its Standard De viation and Constant Mean Diameter (Dn) ..........47 Figure 3 Change in the Shape of the Simulated Transmission Spectra of the Nano-droplet Populati on of Styrene in Water Emulsion as a Function of its Mean Diameter (Dn). Figure 4 Change in the Shape of the Si mulated Transmission Spectra of the Large Droplet Population of Styrene in Water Emulsion as a Function of its Mean Diameter (Dn) .....48 Figure 5 Change in the Shape of the Simulated Transmission Spectrum of Large Droplet Populati on as a Function of its Standard Deviation ....48 Figure 6 Comparison of the Simulated Tr ansmission Spectra of Decane in Water Emulsion (with SDBS Emulsifier) for Droplet Po pulations of Different Sizes..49 Figure 7 Effect of Particle Size for Decane in Water Emulsion with SDBS as the Emulsifier on th e Simulated Transmission Spectra for the Latex Consisting of Sma ll Droplets (Amplified Lower Wavelength Region)....50 Figure 8 Comparison of the Effect of Pa rticle Size for Decane in Water Emulsion with SDBS as the Emulsifier on the Simulated Transmission Spectrum of the Emulsifier Distributed on the Droplets (Amplified Lower Wavelength Region) .....51 Figure 9 Comparison of the Effect of the Particle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size (Small Droplet Size 10 to 30nm in Diameter) ..............52 Figure 10 Comparison of the Effect of th e Particle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size of Range 10 to 30 nm (Amplified Lower Wavelength Region) ...53

PAGE 9

viii Figure 11 Comparison of the Effect of th e Particle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size (Bigger Droplets of Size Range 50nm to 100nm in Diameter Dn)...54 Figure 12 Comparison of the Effect of th e Particle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size of Range 50 to 100 nm (Amplified Lower Wavelength Region)..55 Figure 13 Comparison of the Shape of the Normalized Experimental Spectrum with that of the Simulate d Transmission Spectrum of Large and Small Droplets Different Mean Diameters (Dn) added in Equal Proportions........56 Figure 14 Comparison of the Shape of the Simulated Transmission Spectrum of Large Droplets about 3 microns and Small Droplets of 30nm Mean Diameter (Dn) adde d in Different Proportions and the Normalized Experimental Spectrum. Figure 15 Comparison of the Shape of the Simulated Transmission Spectrum of Large Droplets about 3 microns and Small Droplets of 50nm Mean Diameter (Dn) adde d in Different Proportions and the Normalized Experimental Spectrum. Figure 16 Experiment al Design Strategy......74 Figure 17 Results Obtained from Replic ate Experiments at the Center-point..75 Figure 18 Schematic of the Reaction Ve ssel Assembly for the Experiments Performed..76 Figure 19 Dilution System Assembly for Spectroscopy Measurements...77 Figure 20 Reactor for the Experiments with Model Molecules Figure 21 Reactor Assembly Setup with Temperature Control Jacket for the Reactor and Surface Tensiometer Probes..78 Figure 22 Dilution System Assembly to Acquire Transmission Spectra..79 Figure 23 System Setup for Experiments with Model System..79 Figure 24 Temperature Control System for th e Sample Holder of the Spectrometer...80 Figure 25Entire Setup for Experiments with Model System.....80

PAGE 10

ix Figure 26 pH Meter to Measure the pH of the Suspending Medium....81 Figure 27 Comparison of the Experimental and Estimated Spectrum for Experiment 1..99 Figure 28 Comparison of the Experimental and Estimated Spectrum for Experiment 2..99 Figure 29 Comparison of the Experimental and Estimated Spectrum for Experiment 4 Figure 30 Comparison of the Experimental and Estimated Spectrum for Experiment 5 Figure 31 Comparison of the Experimental and Estimated Spectrum for Experiment 6 Figure 32 Comparison of the Experimental and Estimated Spectrum for Experiment 7 Figure 33 Comparison of the Experimental and Estimated Spectrum for Experiment 8 Figure 34 Comparison of the Online Experiment al and the Estimated Spectra at Low Temperature.103 Figure 35 Comparison of the Offline Experime ntal and the Estimated Spectra at Low Temperature.103 Figure 36 Effect of pH on the Mean Diam eter Dn of Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio in Emulsion Recipe. 95% Confidence Intervals are denoted by +. Continuous Lines Suggest Trend Figure 37 Effect of pH on the Standard Deviation of Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio in Emulsion Recipe. 95% Confidence Intervals are denoted by +. Continuous Lines Suggest Trend....104 Figure 38 Effect of pH on the Weight Per cent of Oil in the Na no-droplet Population at Different Conditions of Surfactant to Oil Ratio in Emulsion Recipe. 95% Confidence Intervals are denoted by +. Continuous Lines Suggest Trend....105

PAGE 11

x Figure 39 Effect of Surfactant to Oil Ratio in Emulsion Recipe on the Mean Diameter Dn of the Nano-droplet Population at Different C onditions of Temperature and pH. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend.. Figure 40 Effect of Surfactan t to Oil Ratio in Emulsion Recipe on the Standard Deviation of the Nano-droplet Populati on at Different Conditions of Temperature and pH 95 % Confidence Interval s are denoted by +. Continuous Lines Suggest Trend.106 Figure 41Effect of Surfactant to Oil Ratio in Emulsion Recipe on the Weight Percent of Oil in the Na no-droplet Population at Different Conditions of Temperature and pH 95 % Confidence Interval s are denoted by +. Continuous Lines Suggest Trend..106 Figure 42 Effect of Temperature on the Mean Diameter (Dn) of the Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend.107 Figure 43 Effect of Temperature on the St andard Deviation of the Nano-droplet Population at Different C onditions of Surfactant to O il ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend. Figure 44 Effect of Temperature on the Weight Percent of Oil in the Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend.....108 Figure 45 Effect of the Different Emulsifi cation Conditions on the Mean Diameter Dn of the Large Droplet Population........109 Figure 46 Effect of the Different Emulsification Conditions on the Standard Deviat ion of the Large Droplet Population.....109 Figure 47 Schematic of the Transmission Measurement. Figure 48 Comparison for Extinction for La rge Particles (Monomer Droplets Mean Size 3.2 microns Std Dev 0.2) with the Features of the Scattering Component of the Experimentally Observed Spectrum..127 Figure 49 Comparison of Simulated Transmissi on Spectra for Particle Populations with Different M ean Sizes but the Constant Standard Deviation 0.2......128

PAGE 12

xi Figure 50 Spectral Mass Balance for 30 nm and 3.2 microns Mean Diameter Population....129 Figure 51 Spectral Mass Balance for 40 nm and 3.2 microns Mean Diameter Population Figure 52 Spectral Mass Balance for 45 nm and 3.2 microns Mean Diameter Population Figure 53 Spectral Mass Balance for 50 nm and 3.2 microns Mean Diameter Population....131 Figure 54 Spectral Mass Balance for 60 nm and 3.2 microns Mean Diameter Population Figure 55 Particle Size Distributions with Different Mean Sizes....... ............132 Figure 56 Refractive Index of Decane at Different Temperatures.. Figure 57 Refractive Index of Styrene....150 Figure 58 Absorbance of Styrene Figure 59 Absorbance of Sodium Dodecyl Benzene Sulfonate......151 Figure 60 Refractive Index of Sodium Dodecyl Benzene Sulfonate.. Figure 61 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at High Temperatures (60 Deg C)...152 Figure 62 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at High Temperatures (60 Deg C). Am plified Lower Wavelength Region Figure 63 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatures (22 Deg C)....154 Figure 64 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatures (22 Deg C). Amp lified Lower Wavelength Region.....154

PAGE 13

xii Figure 65 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatures (22 Deg C). Am plified Higher Wavelength Region Figure 66 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10..155 Figure 67 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10. Amplified Lower Wavelength Region...156 Figure 68 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10. Further Amplification of Lower Wavelength Region.156 Figure 69 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2....157 Figure 70 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplifi cation of Lower Wavelength Region...157 Figure 71 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplification Lower Wavelength Region...158 Figure 72 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplifi cation of Higher Wavelength Region..158 Figure 73 Effect of Surfactant Concentr ation at Low pH (pH 2) and High Temperature (60 Deg C) Figure 74 Effect of Surfactant Concentr ation at Low pH (pH 2) and High Temperature ( 60 Deg C). Amplified Lower Wavelength Region..160 Figure 75 Effect of Surfactant Concentration at Low pH (pH 2) and Low Temperature (50 Deg C) ............160 Figure 76 Effect of Surfactant Concentr ation at Low pH (pH 2) and Low Temperature ( 50 Deg C). Amplified Lower Wavelength Region..161 Figure 77 Effect of Surfactant Concentration at High pH (pH 10) and High Temperature (60 Deg C) Figure 78 Effect of Surfactant Concentra tion at High pH (pH 10) and High Temperature ( 60 Deg C). Amplified Lower Wavelength Region.162

PAGE 14

xiii Figure 79 Effect of pH at Low Surfactant Concentration (S/O Ratio = 0.0154) and High Temperature (60 Deg C).....162 Figure 80 Effect of pH at Low Surfactant Concentration (S/O Ratio = 0.0154 ) High Temperature ( 60 Deg C). Amplified Lower Wavelength Region...163 Figure 81 Effect of pH at High Surfactant Concentration (S/O Ratio = 0.046) and High Temperature (60 Deg C). Figure 82 Effect of pH at High Surfact ant Concentration (S /O Ratio = 0.046) and High Temperatur e (60 Deg C). Amplified Lower Wavelength Region....164 Figure 83 Effect of Temperature at High Surfactant Concentrations (S/O ratio = 0.046) and High (pH= 10)...164 Figure 84 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and High pH (pH = 10)...165 Figure 85 Effect of Temperat ure at High Surfactant Concen tration (S/O Ratio= 0.046) and High pH (pH = 10). Amplified Lower Wavelength Region.165 Figure 86 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and Low pH (pH = 2)......166 Figure 87 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and Low pH (pH = 2). Amp lified Lower Wavelength Region..166 Figure 88 Effect of Temperature at Low Surfactant Concentration (S/O Ratio = 0.0154) and Low pH (pH =2).........167 Figure 89 Effect of Temperature at Low Surfactant Concentra tion (S/O Ratio=0.0154) and Low pH (p H =2 ). Amplified Lower Wavelength Region....167 Figure 90 Actual Offline Spectrum of Emulsi on at Room Temperature at Different Times...169 Figure 91 Comparison of Normalized Offline Spectrum of Emulsion at Room Temperature at Di fferent Times. Normaliza tion from 230 to 820nm.170 Figure 92 Comparison of Normalized Off line Spectra of Emulsion at Room Temperature Different Time s when Stability is Achieved. Normalization from 230 to 820nm..170

PAGE 15

xiv Figure 93 Comparison of Normalized Off line and Online Spectrum after Stabilization of the Emul sion. Normalization Wavele ngth from 230 to 820 nm...171 Figure 94 Comparison of the Offline and On line Normalized Spectra with Statistics.171 Figure 95 Comparison of the Online and Off line Normalized Spectra with Statistics Amplified Lower Wavelength Region.......172 Figure 96 Normalized Droplet Size Distribution for Experiment 1..173 Figure 97 Normalized Droplet Size Distribution for Experiment 2..174 Figure 98 Normalized Droplet Size Distribution for Experiment 4..174 Figure 99 Normalized Droplet Size Distribution for Experiment 5..175 Figure 100 Normalized Droplet Size Distribution for Experiment 6..175 Figure 101 Normalized Droplet Size Distribution for Experiment 7..176 Figure 102 Normalized Droplet Size Distribution for Experiment 8..176 Figure 103 Schematic of Densitometer Setup.188 Figure 104 Surface Tensiometer Setup (Schematic)...189 Figure 105 Signal Flow-path for the Temperature Control System Figure 106 Water Flow for the Temperature Control System.191 Figure 107 Schematic of the Dilution System (As Represented by Sacoto33)....192 Figure 108 Reactor Setup for the Polymeriza tion Reaction with the Dilution System...193 Figure 109 Change in the Shape of the Tran smission Spectra of the Reaction Mixture as the Reaction Progresses. Figure 110 Estimated Conversion of Styren e to Polystyrene Using Spectroscopy. Figure 111 Particle Size Distribution of the Polystyrene Particles Formed as a Result of the Polyme rization Reaction After the Reaction is Completed.196 Figure 112 Comparison Between the Estimated and the Measured Spectra of the Styrene Emulsion at Time Zero (Before the beginning of the Reaction)..197

PAGE 16

xv LIST OF SYMBOLS A( ) Absorption coefficient as function of wavelength c Concentration in g/cc CMC Critical Micellar Solution D Diameter of the particles D Appreciable Decrease with the increase of the effect dI The change in the intensity of the light Extotal Total extinction Exsmall Extinction of small particles Exlarge Extinction of large particles f(D) Particle Size Distribution I The intensity of the transmitted light I0 The intensity of the incident light I Appreciable Increase with the increase of the effect k Imaginary part of the refractive index l Path length m Complex Refractive Index n Real part of the refractive index Np Number of particles N Not Appreciable

PAGE 17

xvi NPE Not Possible to Express les surfmolecuN Number of surfactant of molecules Nsmpa Number of particles of the size characteristics of the swollen or empty micelles for dispersing the estimated amount of oil msmspN Number of moles of surfact ant required to stabilize small particle population smspN Total number of surfactant molecules required to stabilize the surface area of small particle population P Principal value of the integral OD Optical Density Qext Extinction Coefficient absorbanceQ Absorption efficiency scatteringQ Scattering efficiency R Radius of one droplet of the dispersed phase stabreqS The surface area of the small particle population S/O Surfactant to Oil ratio T Temperature TR Reduced temperature Tc Critical temperature V Volume of one droplet of the dispersed phase V(T) Molar volume of decane hydrocarbon at temperature T

PAGE 18

xvii Other Symbols 0 Optical Density 0 Wavelength I Initial Wavelength f Final wavelength extinction The extinction coefficient Gem Free Energy Interfacial Tension s Mean diameter of small particle population s Standard Deviation of the small particle population % Weight percent (T) Density of the hydrocarbon at temperature ‘T’ Acentric factor

PAGE 19

xviii Identification of Nucleation Locus in Emulsion Polymerization Processes Vineet Shastry ABSTRACT Particle Nucleation is the forcing f unction in the Emulsion Polymerization processes and it plays an impor tant role in dictating the final properties of the latex produced. Identification of the main nucleation sites and characterizing them in terms of their size and composition is important fo r elucidating the mech anism of particle nucleation. This research focuses on identifying the most likely nucleation locus in emulsion polymerization processes by characterizing the initial conditions of the reaction mixture. In order to achieve this objective, a me thodology was devised, which used a non-reacting model emulsion system instead of the orig inal emulsion. The model emulsion system selected has the same dispersion properties as that of the monomer emulsion system, but different optical properties. The model em ulsion system enabled the study of the distribution of the emulsifier using Uv vis spectroscopy. This approach also eliminated the time constraint associated with sampling during a polymerization reaction. A quantitative deconvolution using the turbidity equation, was done on the transmission Uv vis spectra of the

PAGE 20

xix emulsions. This enabled the characterization of the emulsions in terms of their particle size distribution, particle number and th e composition of the droplet populations comprising them. The studies conducted provi de the experimental evidence for a previously unidentified nano-droplet populat ion of size range 30 to 100nm in diameter. To further support this experimental evidence, calculations were performed to obtain the emulsifier distribution over the nano-droplet population. The calculations suggest the probability of existence of the nano-droplet population to be much higher than th e probability of the existence of the swollen micelles. The results, depending upon the emulsification conditions, indicate the presence of about 15 % to 80% of the dispersed pha se in the nano-drople t population. The large interfacial area offered by the nano-droplet popul ation due to their high particle numbers and high percentage of the dispersed oil pha se in them, make them the most probable particle nucleation loci in emul sion polymerization processes. Designed experiments were performed to experimentally observe the changes in the nano-droplet populations. The effects of th e process variables, namely pH, surfactant concentration and temperature, on the size a nd compositional characteristics of the nanodroplet population were investigat ed. The results suggested that the surfactant to oil ratio was the dominating factor governing the size and the weight percent of the dispersed phase in the nano-droplet population.

PAGE 21

1 CHAPTER 1: INTRODUCTION 1.1 Motivation Particle nucleation is the forcing function in emulsi on polymerization processes since it plays a significant role in the development of most of the properties of the final latex. The nucleation mechanism in emulsion polymerization is not fully understood and therefore it still remains an area of active research1. The main objective of this research effort is to identify the actual nucleati on locus for the understanding of the nucleation mechanism in emulsion polymerization. This research proposes the use of optical techniques for the emulsion characterization in terms of its size and composition. This dissertation describes the model emulsion sy stems and the experiments based on them that provide relevant information for iden tifying the nucleation m echanism. A brief description of the pr ocess is provided in this ch apter in addition to giving a brief idea regarding the importance of Em ulsion Polymerization processes. Explanation and reasoning for the main thrust of this res earch effort and the cont ributions as a result are described.

PAGE 22

2 1.2 Significance of the Emulsion Polymerization Process Emulsion Polymerization is a process of great industrial importance. It finds applications in the manufacture of a wide range of products such as paints, latex, adhesives, production of coatings and other synthetic materials2. It also has applications in bio-separations through f unctionalization of the latex pa rticles, and it offers great promise for the synthesis of nano-materials2. This process is a technologically and comme rcially important synthesis process. It grew rapidly in use as a result of the intens e research into its a pplication for producing synthetic alternatives to natural r ubber latex during th e Second World War2. It is a basic process of a massive global indus try that continues to expand due to the versatility of the reaction and greater realization of the ability to control the proper ties of the polymer latexes produced. Millions of tons of synthetic polymer latexes are prepared by this process for use as commodity polymers in a wide variety of applications, such as: synthetic rubber, high – impact polymers, latex foam, carpet backing, binders for non woven fabrics, adhesives, additives for cons truction materials such as Portland cement, mortar, concrete and sealants3. This industry has grown into a multi -million dollar industry today. From the reports on the International Instit ute for Synthetic Rubber Production4 almost 10.4 million metric tons of synthetic r ubber latex was consumed world wide in the year 1998. Western Europe recorded the greatest consump tion growth rate of 5.4% amounting to approximately 2.7 million metric tons. Nort h America showed a growth rate 3.1% reaching nearly 3.1 million metric tons consum ed in 1998. The global market in the year

PAGE 23

3 1999 for the synthetic latex polymers was estimated to be about 13,500 million dry pounds5. Worldwide Consumption of main polymer systems in the year 1999 as per the reports of Polymer Latex GMBH6 is as follows Table 1.1 Sale of the Emulsion Polyme rization Products in the Year 1999 Polymerization Product In Million tons Polyvinyl chloride 24 Polystyrene 12 Synthetic rubber 10 Synthetic latex 6.0 Other Emulsion products 4.6 As per the forecast of the market research conducted by the Paint and Research Association7 UK, the consumption of the enviro nment friendly adhesive products until the year 2004 is expected to show an overa ll growth from 3 to 5% per year by volume. According to the market report published by the Freedonia8 group in October 2002, the emulsion polymer demand in the United States is expected to grow 2.8% annually to five billion pounds in 2006. Market value is expected to rise 4.4% per year to 4.3 billion dollars. Given the magnitude of these numbe rs, the impact of the Emulsion Polymer industry on the global economy can be appreciated.

PAGE 24

4 1.3 A Brief Description of Emul sion Polymerization Process The main advantage of the Emulsion Poly merization process is that it can be adjusted to tailor the properties of the pr oduct polymer and of the latex. A few important terms that are used throughout the dissert ation for the description of the emulsion polymerization process are defined as below. 1.3.1 Important Definitions 1. Nanodroplets: Nano-droplets are the small droplets of the monomer with the size ranging form 30 nm to 100nm in diameter. 2. Time zero condition: Is the initial condi tion of the reaction mixture before the initiator is adde d to it (beginning of the reaction). Essentially at time zero, a liquid – liquid emulsion system with monomer as the dispersed oil phase in water (continuous phase) is present in the reactor. 3. Micelles: Aggregated molecules of surfactants of size range 1 nm to 3 nm in diameter. 4. Swollen Micelles: Micelles contai ning monomer; size range 5 to 10 nm in di ameter. 5. Particle/ polymer particle: Swollen mi celles with either dead or growing polymer chain; size ra nge 20 to 80nm in diameter. 6. Monomer droplets: Droplet s of monomer having size range one to a few microns. 7. Particle number: Total number of any particulate entity such as micelles, swollen micelles, particles, monomer droplets, nano-droplets etc. 8. Oligomeric radicals: The initiator radicals react wi th the dissolved monomer to form radicals which in turn reac t with the other dissolved monomer thus increasing

PAGE 25

5 their chain length. Such radicals that have three to five m onomer units (very short chain length) are called o ligomeric radicals. 9. Forcing Function: Forcing function of an even t is the factor primarily responsible for a particular process phenomenon. 1.3.2 Emulsion Polymerization Process Emulsion polymerization is a free radical initiated chain polymerization in which a monomer, or a mixture of monomers is polymerized in the presence of an aqueous solution of a surfactant to form a pr oduct known as latex. La tex is defined as a colloidal dispersion of polymer particles in aqueous medium. The main ingredients for conducting the emulsion polymerization reac tions in addition to the monomer and water, include surfactants, initiators and sometimes chain transfer agents. When the surfactant is added to water and the concentration of the surfactant is above the critical micellar concentration (C MC), the hydrophobic ends of some of the surfactant molecules come together to fo rm micelles. These mi celles are generally referred to as empty micelle s as they only contain surfactant molecules. These micelles are about 5 nm1 in diameter and their concentration as a function of th e concentration of the surfactant above CMC is given by1,9,10 Number of micelles = agg s A c s addedn CMC a N N r S * * 42

PAGE 26

6 where, 1. [Sadded] is the molar concentration of the su rfactant added to the continuous phase. 2. Nc is the particle number. 3. [CMC] is the critical micellar concentration of the surfactant. 4. NA is the avogadros number. 5. as is the area occupied by one surfactant molecule. 6. nagg is the mean aggregation number of the micelle. The rest of the surfactant that does not form micelles but remains in dissolved form is termed free emulsifier. The monomer is then added to this aqueous solution of the surfactant. Some of the monomer enters th e micelles. Such micelles are called swollen micelles. The average size of the monomer swollen micelles is 10nm9. The rest of the monomer remains as monomer droplets. In ot her words, the otherwise immiscible oil phase (in this case monomer) is dispersed in to the water phase. An emulsion of oil in water is thus obtained. Some of the emul sifier surrounds the monomer droplets to maintain the stability of the emulsion. The emulsion polymerization reaction is in itiated by the addition of the initiator. The initiation mechanism or the mechan ism of primary free radical formation is helpful in addressing the number of free oligomeric radicals that will be formed. The primary free radicals react with the dissolved monomer in the aqueous phase to form the oligomer radicals which play an important role in nucleation as per the existing (micellar and homogeneous coagulative and droplet) nucleation theories11.

PAGE 27

7 Emulsion Polymerization reactions can be di vided in to stage I, stage II and stage III. The stage I of the emulsion polymeri zation reaction is characterized by the disappearance of micelles and formation of part icles (or particle nucleation). The stage II is characterized by the growth of the part icles and the presence of monomer droplets, while the stage III is characterized by the disappearance of the monomer droplets. The three main proposed and widely accep ted (to different degrees) particle nucleation mechanisms are: 1. Micellar nucleation 2. Homogeneous and coagulative nucleation 3. Droplet Nucleation An elaborate discussion on the merits a nd limitations of the proposed nucleation theories is presented in Chapter 2. Particles ar e primarily formed in the stage I as per the dominant nucleation mechanism. The particle s grow in size as the reaction progresses inside them. The monomer droplets serve as th e supply reservoirs of the monomer to the particles in which the re action is taking place. The emulsifier gets redistributed for maintaining the stability of th e latex mixture as the reaction proceeds, since the interfacial area between the particul ate entities changes. Due to the growth in the particle size and the change in the particle number, the in terfacial area between the particles and the continuous phase changes continuously for th e stage I of the reaction. The number of particles remain constant but the particles continue to gr ow in size as the reaction progresses to the stage II. This causes a ch ange in the interfacial area between the

PAGE 28

8 dispersed phase and the continuous phase even during the second stag e of the reaction. The monomer droplets present in the emul sion provide the growing particles with monomer. During the second st age, the entry of another radical inside the polymer particle could either terminat e a growing chain or could he lp in the propagation of an already existing dead polymer chain inside the pa rticle in case branch ing or chain-transfer reactions are present. Each particle acts as a bulk reactor where the polymerization reactions occur. The second stage of the emulsion polymerization reaction is thus characterized by the growth of the polymer pa rticles and decreasing size of the monomer droplets. The third stage in the emulsion polymeri zation reaction is characterized by the disappearance of the monomer droplets. As more monomer is used, the monomer droplets providing monomer to the growing pol ymer particles gradually decrease in size and finally disappear. This marks the beginni ng of the last stage of the reaction. The reaction inside the polymer particles continues until it finally ends due to the lack of any further availability of the monomer. Figure 1 is a representation of the reaction mixture typically present inside the reactor.

PAGE 29

9 Figure 1 Representation of the Late x Mixture for Emulsion Polymerization 1.4 Significance of the Research Proper understanding of partic le nucleation or particle formation is important since it governs1: 1. The number of particles formed in an unseeded emulsion polymerization reaction 2. Particle size and hen ce the particle size distribution 3. The generation of particle s depends upon the nucleation conditions. It is therefore ne cessary to determine the operating cond itions required to avoid or to generate new crop of particles. 4. Rate of emulsion polymerization, and ther efore the reaction dynamics, is proportional to particle number. Monomer droplets Swollen micelles Empty Micelles Polymer particles

PAGE 30

10 1.5 Outline of the Dissertation Different nucleation theories proposed to date, are discussed along with their merits and limitations in the chapter on liter ature review of this dissertation. In this chapter are also described, the initial efforts made at USF for understanding the nucleation mechanism in emulsion polymeri zation. A brief histor y of the research, inclusive of the experimental work and deve lopment of the data analysis techniques is presented. In the Chapter 3 is delineated the problem statement. This chapter focuses and justifies the strategic approach followed in addressing the problem of identifying thr nucleation locus in emulsion polymerization. The arguments presented to justify the approach are supported by the results obtaine d via computer simulations. This chapter presents the main hypothesis of this work a nd discusses its relevanc e in understanding the nucleation mechanism. In the Chapter 4, is outlined the experime ntal work undertaken in this research effort. The experiments with the model systems are explained and justified. Chapter 5 is focused on the results of initial condition experiments performed with the model system. Discussion on the signif icance of the results and its implications pertaining to the understanding of the initial conditions fo r an emulsion polymerization reaction is presented.

PAGE 31

11 In the Chapter 6 the conclusions of this research and makes recommendations for future work are summarized. 1.6 Contributions Some of the contributions of this research effort are: 1.6.1 Identification of the Main Locus for Nucleation in Emulsion Polymerization There exists a controversy pertaining to th e locus of nucleation. Theories that are accepted to different degrees, propose different particulate entities as the reaction locus for the reaction to occur. Micellar nucleation theory pr oposes the monomer swollen micelles as the main nucleation locus1,2 while the homogeneous nucleation considers the precipitated oligomeric radicals as the main nucleation locus. The homogeneous and coagulative nucleation mechanism consider s the precursor particles as the main nucleation locus,1, 2 while the droplet nucleation theory attributes the main reaction locus entity to the monomer droplet s. Thus, identification of th e main nucleation loci in emulsion polymerization reactions constitutes a contribution from this research effort. 1.6.2 Characterizing the Emulsifier Distribution as a Function of Initial Emulsification Condi tions in an Emulsion Polymerization Reaction The emulsifier concentration has a great influence over the size and compositional characteristics of the dispersed phas e along with the other initial emulsification conditions such as temperature, rate of shear, pH of the suspending medium. Emulsifier distribution as a function of the initial c onditions specified above fo r a stable emulsion, determines the size and the compositional characteristics of the actual nucleation locus.

PAGE 32

12 1.6.3 Implementation of Spectroscopic Tec hniques to Identify the Nucleation Mechanism A rapid and minimallyinvasi ve measurement technique enabling continuous characterization of the particulate system pres ent inside the reactor with minimal sample preparation is implemented. This technique can thus serve as an important tool for the continuous monitoring of the emulsion pol ymerization process in the industry. 1.6.4 Emulsion Characterization Characterization of the dispersed phase in terms of its size and the compositional characteristics as a function of emulsificati on conditions is of fundamental importance to understand the emulsification process. Implem enting the spectroscopic techniques enable the comprehensive characterization of the dispersed phase.

PAGE 33

13 CHAPTER 2: LITERATURE REVIEW 2.1 Introduction A number of theories have been proposed to explain the nucleation mechanism in emulsion polymerization in the past. The part icle nucleation phenom enon continues to be an area of active debate1. Various researchers have propos ed different theories based upon their experimental evid ences and mathematical modeling for explaining the nucleation mechanism in emulsion polymerizat ion. The objective of th is chapter is to highlight the controversy in the proposed ma in nucleation locus postulated by different theories put forth to date. The controversy on the main nucleation locu s can be attributed to the limitations of the existing experimental capabilities av ailable for the researchers for completely characterizing the reaction mixt ure present inside the react or at early times of the reaction. This chapter provides a brief backdrop of the theories that have been proposed for the nucleation mechanism and the main nuc leation locus suggested by each of them. A discussion on the merits and limitations of each of the proposed theories is also presented. The efforts undertaken over a pe riod of years to id entify the nucleation mechanism at the University of South Flor ida, by the research group of the Polymer Synthesis and Characterization La boratory are briefly described.

PAGE 34

14 2.2 Proposed Nucleation Mechanisms and their Limitations Harkins’ mechanism as explained by Mohammed S. El-Aasser2, states that the major source of particle nuc leation is the monomer swollen surfactant micelles. According to Harkins, radicals generate d in the aqueous phase enter the monomerswollen micelles and initiate polymerization to form monomer swollen polymer particle nuclei. The monomer swollen particles grow by polymerization of monomer supplied to them by the monomer droplets by diffusion th rough the aqueous phase. The particle nucleation stage ends with the disappearance of the micelles. The major locus of polymerization was postulated to be th e monomer swollen micelles. The above mentioned nucleation mechanism is calle d the micellar nucleation. The monomer swollen micelles supposed to be the main reac tion sites (reaction loci ) as per the micellar nucleation theory are around 5 to 10 nm in diameter. The micellar nucleation mechanism generally results because of low monomer solubility. The quantitative development of the Hark ins’ theory of emulsion polymerization kinetics was published by Smith and Ewart in 1948 with a first atte mpt at verification by Smith2. Smith and Ewart’s equations12 predicted the rate of pol ymerization being directly proportional to the number of particles. The number of particles were predicted to be proportional to the 2/5th power of initiator concentration and 3/5th power of surfactant concentration. The predicted orders of initiator concentration and surfactant concentration were verified experimentally for styrene12. The congruence of the order of concentration

PAGE 35

15 of the initiator and the surfactant predic ted by the Smith Ewart’s theory with the experimental observation was widely quoted as evidence for the micellar theory for nucleation1. Gilbert1,2,13, El Aaser3 cite Roe in pointing out, that the same two exponents can be predicted by a homogeneous – nucle ation mechanism or any other model for nucleation mechanism considering complete su rface coverage by surfactant as one of the governing events for the cessation of particle formation. The equations derived by Roe, which are identical to Smith –Ewart equatio ns without consideri ng micellar entry are reported by El Aaser3. According to Roe (as cited by El Aaser3), particle generation occurs at each interactio n between dissolved free radical and dissolved monomer molecule and continues until the surfactant is de pleted to a level not sufficient to stabilize new particles through adsorption. Roe demons trated that though the micelles are an important source for particle form ation, they were not necessary. Alexander Dunn12 reproduces these derivations of Smith and Ewart’s equations for the rate of polymerization for Case 2 s cenario. Case2 scenario is observed when the number of radicals per particle is equal to 0.5. When the Case 2 kinetics apply, the rate of polymerization of a particle is constant, independent of part icle size or rate of the entry of the radicals if the concentration of the monomer inside the pa rticles is constant. Alexander Dunn12 further states that the effect of the ionic strength of the aqueous phase was not considered. Other implicit assumptions namely

PAGE 36

16 1. “No difference in the area occupied by an emulsifier at the polymerwater interface and at the airwater interface.” 2. “Independence of the area occupied by an emulsifier to the presence of the monomer or other swelling agen ts in the particles were subsequently found to be incorrect.” Gardon14 showed that the validity of the Smith and Ewart’s theory is confined to specifiable intervals of convers ion, to a certain range of mo nomer/ water ratio, and to the surfactant concentrations whose uppe r and lower limits are given. Gardon14 explicitly states the important assump tions underlying the Smith and Ewart theory namely, monomer swollen latex particle s being the main locus of r eaction, initiation of polymeric chains by the entry of the radicals from the water phase into the particles, instantaneous chain termination and nucleation by radical absorption by the particles containing the growing chain or the monome r swollen micelles. Gardon14 calculated the conversion at which the particle nucleation is complete along the predictions for particle size distributions. Based on the assumptions of the Smith Ewart theory, the relationship between the final number of particles formed as a function of the rate of radical production per cubic centimeter of water, frac tion of monomer in th e particles and bulk rate of polymerization were derived. Gilbert1 and Dunn12 refer to Fitch and Gardon to po int out, that quite a wide range of exponents of the surfactant concentration and initiator concentration proportional to

PAGE 37

17 the number of particles at the end of the reaction is actually observed and also that the experimental data do not always obey a simp le power law when plotted accurately. El Aaser3, cites the investigation of vinyl chloride emulsion polymerization and vinyl acetate emulsion polymerization by Jacobi and Priest respectively, as the first to report on homogeneous nucleation. The preci pitated oligomeric radicals form the spherical particles, adsorb surfactant to fo rm primary particles thus becoming the main loci of nucleation. Hansen, Ugelstad, Fitc h and Tsai put forward the theory of homogeneous nucleation commonly referred to as the HUFT theory as reported by Gilbert1. Since the coagulation events are include d in the extension of the HUFT theory, it is also referred to as the homogeneous a nd coagulative nucleati on theory. Gilbert1 gives a detailed illustration of the sequen ce of events for the HUFT theory. Works of Napper and Gilbert1,2,15 at University of New South Wales, Sydney Australia, propose that particle nucleation invol ves at least two mechanistic steps as opposed to the single step process for micellar nuc leation or homogeneous nucleat ion. The first step is the formation of the “precursor particles” due to homogene ous nucleation and the second step is the formation of the mature particles by the aggregation of the precursor particles. Gilbert refers to the work of Feeney1 et al to show that carrying out emulsion polymerization in a polyacrylamide gel can isol ate the precursor particles. By doing so, it can be ensured that the small particles coul d not coagulate. Particles of radius 5nm measured by small angle neutron studies were found to be isolated by these methods and could be stored indefinitely. Compositional in formation of the particles could thus be

PAGE 38

18 obtained. However, the quantitative informa tion on the number of precursor particles, which is critical in deciding the most likel y nucleation locus, cannot be obtained using this method. In quantifying the homogeneous nucleat ion for mathematical formulation purposes, the competition between entry (capture) of newly formed radicals and the radicals that form the precursor particles is taken into account along with the knowledge of the aqueous phase chemistr y in emulsion polymerization. Nucleation below the CMC of the surfact ant is better explained by HUFT theory. Gilbert argues that homogeneous and coa gulative nucleation is still the dominant mechanism for the emulsion polymerization carr ied out at surfactant concentration above CMC. Gilbert and Napper16 briefly summarize the principal mechanism that may be operative in particle nucleation in a given sy stem. They inferred th at it is impossible to use the polymerization rate da ta alone to make any mech anistic deductions concerning nucleation. The polymerization rate data may be used in conjunction with other observable experimental data to provide mech anistic information on particle formation. They consider the experimental observations so as to obtain information on nucleation. The observations considered are polymerization rate, particle formation rate, particle size distribution for the case study of polymer ization of styrene.

PAGE 39

19 The objective of the measurement of th e appropriate quantit ies of the small particles was not attainable by the experimental techniques available to Gilbert and Napper4. Dependence of the particle number on the concentration of the surfactant and initiator and the data on the ch ange in the number of partic les as a function of time in conjunction with other measurements can provide very useful information on the mechanism of particle formation. Particle size distribution (PSD) data was used to provide means of refuting mechanistic suppositions. Apart from the mean and standard deviation of the full particle size distribution, Gilbert and Napper1,2 suggest that the sign of the skewness of the particle size distribution towards the sma ller sizes is indicative of the dominant mechanism. Gilbert and Napper1,16 have reported the PSD at times during the inception of the second stage to gather the mechanistic information of the nucleation process that essentially and predominantly occurs dur ing the first stage of the reaction. The nucleation mechanism theory that coul d explain the experimentally observed positive skewness of the particle size distri bution during the inception of the second stage is the approach proposed by Gilbert for identifying nucleation mechanism. Gilbert1,13,15 cites the positive skewness of the particle size distribution at the end of the nucleation stage, plotte d as a function of particle vo lume as the evidence for the theory of homogeneous and coagulative nucl eation. The positive skewness at the end of

PAGE 40

20 the nucleation stage will indicate the high proba bility of the existence of smaller particles at the end of the nucleation stage. Positive sk ewness of the distribution indicates that the rate of nucleation is an increasing function of time. The high probabi lity of most of the particles being small at the end of the nucleati on stage indicates that most of the particles were formed during the later times of the nucleation period. Had the particles formed earlier, they would have grow n as the reaction progressed th roughout the first stage. Thus the probability of obtaining larger particles at the end of the nucleation stage would be higher. A negatively skewed dist ribution would thus be obtained. Gilbert’s two step mechanism of homogeneous and coagul ative nucleation is the only theory that predicts nucle ation rate as an increasing f unction of time. The particle size distributions for the pa rticle generation by micellar entry or simple homogeneous nucleation3 is characterized by negative skewness which in turn is indicative of nucleation rate as a decreasing function of time. This implies that most of the particles are formed in the later times of the 1st stage of the reaction. Mice llar theory predicts a decrease in the rate of nucleation suggesting that most of the particles are formed during the early times in the 1st stage of the reaction. Similarl y the single step homogeneous nucleation mechanism too predicts nucleation rate as the decreasi ng function of time. According to Gilbert and Napper16, the calculated and observed particle size distribution of the system and the calculated change in th e number of particles as a function of time refute single step micellar and homogeneous nucleation mechanism for particle formation of styrene system under consideration. As pe r them, the behavior of nucleation rate as

PAGE 41

21 increasing function of time can be explaine d by sequential coagulation only. The positive skewness of the particle size distribution e xpressed in terms of unswollen volume is indicative of homogeneous nucleation followed by coagulative nucleati on, rather than a single step micellar or homogeneous nucleation. PJ Feeney15 presents a detailed theory fo r nucleation kinetics in emulsion polymerization systems based on the coagulation of precursor particles. The coagulative theory presented combines extended Muller – Smoluchowski coagula tion kinetics with DLVO theory15. Feeney15 provides the mathematical expressions for the time evolution of the nucleation rate, particle number and particle size distribution. With physically reasonable values for the parameters for the coagulation kinetics, Feeney obtained agreement with the early time evolution of PSD which is essentially sensitive to the assumptions pertaining to the nucleation mechanisms since different nucleation mechanisms having different mechanistics w ould predict different shapes of the PSDs. Feeney also obtained excellent agreement w ith the data on the dependence of particle number on surfactant and init iator concentrations. Feeney15 observed positive skewness of the particle size distribution for differen t surfactants, which according to him is a conformation of coagulativ e nucleation theory for nucleation mechanism. The observation of positively skewed distri bution of the particles at the inception of the second stage is thus cited as the primary experimental evidence for supporting the homogeneous and the coagulative nucleation mechanism as the actual mechanism of

PAGE 42

22 nucleation in emulsion polymerization1,2. The primary precursor particles and their sizes are typically around 10 to 15 nm in diameter3 are supposed to be the main nucleation sites by this nucleation mechanism theory. Jorge Herrera Ordonez and Roberto Olayo9,10,17 in their recent works question the evidence put forth by Gilbert. In doing so they cite Morrison in pointing out that the differences in the growing rates of differen t size particles can caus e a positively skewed distribution of the particles at the end of the first stage. According to Morrison, as reported by Ordonez and Olayo9,10, neither the homogeneous nucleation nor the homogeneous and coagulative nucleation is ab le to produce the experimentally observed concentration of the polymer particles in emulsion polymerization above the CMC. Furthermore Herrera-Ordonez9,10 states that the extension of the Derjaguin -LandauVerwey -Overbeek (DLVO) theory to model the behavior of the very small latex particle like those in interv al I can be debated. Herrera-Ordonez7 refers to Hansen in doubting the possibility of the completely covered particle s being sufficiently unsta ble to coagulate. Herrera-Ordonez10 in presenting his arguments, refers to the work undertaken by Gianneti to argue that the homogeneous and coagulativ e nucleation is not the only mechanism due to which the particle formation above the cr itical micellar concen tration takes place. HerreraOrdonez7,8 refer to the modeling of particle size distribution done by Gianneti by using two methods namely the zero – one approach as originally formulated by Gilbert and Napper and the generating function approach. His results were contrary to the results of Gilbert and Napper, since neither model was able to fit for increasing nucleation

PAGE 43

23 rates9,10. Instead these data fitt ed nicely when a prevaili ng decreasing rate of the formation of mature particles was assumed. Gi annetti argued that most of the mechanistic information inferred from early – time par ticle size distribution obtained from the experiments performed by Lichti9,10 et al is lost because of the stochastic broadening of the PSD that occurs after nucleation stops This reasoning was based on the observation, that the nucleation time was much smaller th an the sampling time needed to obtain the experimental particle size distribution data. Herrera-Ordonez and Roberto Olayo9 propose a detailed mathematical model for the kinetics of styrene emul sion polymerization. The mode l predicts that micellar nucleation dominates over homogeneous nucle ation and that the evolution of the nucleation rate reaches a maximum, wher e desorbed radicals have an important contribution. The results of this model deve loped were discussed and compared against the experimental data9,10. The theoretical results, whic h were obtained without the coagulation of the particles being taken into account were congruent with the experimental data of the evolution of the styrene monomer convers ion and the rate of polymerization. They concluded that if co agulative nucleation takes place above the critical micellar concentr ation, it is not significant9,10. They further examine the same mathematical model for describing the emul sion polymerization of methyl methacrylate monomer above the CMC of the surfactant17. On the basis of the model results they argued that the observed bimodal PSD and the rate polymerization PSD, need not necessarily be ascribed to the secondary nucleation. According to them, the predicted

PAGE 44

24 PSDs in the early time during the first stages of the reaction can also be caused due to the differences in the growing rate of differe nt size particles as predicted for styrene emulsion polymerization9,10 apart from sequential coagulation as advocated by Gilbert2. The monomer concentration in the particles and the micelles during the first stage of the reaction show considerable size de pendence if there are very small (< 10nm)1. The values can be considerably smaller than the normal values predicted by Morton’s equation1. Ordonez and Olayo9,10 did not consider the above mentioned size dependent compositional characteristic of the particles wh ile developing their model for identifying the dominant nucleation mechanism. C. S. Chern and C.H. Lin18 used a water insoluble dye as a probe to study particle nucleation in semi-batch emulsions polymeri zation of methyl methacrylate. From the results obtained from their experimental effo rts that involved the determination of the quantity of a water insoluble dye incorporated into the latex particles, Chern and Lin18 conclude that homogeneous nucle ation plays a key role in the particle formation period when the surfactant concentration is below CMC. When the surfactant concentration is above the CMC, mixed modes of particle nucleation (micellar and homogeneous nucleation) are operative in the polymerization system. The hypothesis on droplet nuc leation (monomer droplets being the most likely reaction loci) was dismissed by Mohammed S and El Aasser3on the grounds of the low

PAGE 45

25 availability of the interfaci al area thereby decreasing it s likeliness of occurring in conventional emulsion polymerization. Howeve r, the droplet nucleation was accepted to be the primary nucleation mechanis m in mini-emulsion polymerization19. Though it was initially expected that all the droplets w ould successfully compete for radicals thereby becoming polymer particles, it was found that on ly a relatively small fraction (less than 20%) succeeded in converting th emselves into polymer particles by this process. The mechanism of disappearance of droplets in miniemulsion pol ymerization other than by becoming polymer particles still remains an unanswered question20. Collision between droplets and existing particles and diffusion are the two main possibilities that have been widely cited20. Increase in the interfacial area of a given particulate entity and the continuous phase, increases the probability of the oligom eric radicals coming in contact with that particular particulate entity. The interfacial area made available by a particular entity would increase the chances of that particul ate species to become the main locus for nucleation. Since the size of the monomer dropl ets is much larger as compared to the other particulate entities (lik e the swollen micelles) the surface area offered by them is considerably lesser than the smaller particulat e species thereby decreasing their likeliness of being the main locus of particle nucleation. From the literature cited in this secti on we can appreciate the fact that there remains a great deal of controversy regard ing the main locus of nucleation and its

PAGE 46

26 composition. The controversy on the most likely nucleation locus is expected to seriously hamper the elucidation of nucleation mechanism in emul sion polymerization process. The efforts undertaken at USF for addressing the nucleation mechanis m are discussed in the next section. 2.3 Direction of Ongoing Research at USF The controversy pertaining to the main nuc leation loci can be attributed to the unavailability of the experiment al capabilities that allow th e complete characterization of the particle (droplet) populat ions comprising the emulsion mixture present inside the reactor at the incepti on of the polymerization reaction. Co mplete characterization of the reaction mixture at the incepti on of the reaction (also referr ed as time t = zero condition in this dissertation) involves obtaining quant itative information on the populations of the in terms of particle number and their composition1 and qualitative information regarding the true size distribution of each of them, simultaneously. This information is the key for refuting, accepting or proposing the mechanisms for nucleation2. It was for these reasons that a sensor array was developed and impl emented as a part of our ongoing research work to enable the monitoring of the cri tical parameters for emulsion polymerization reaction continuously, simultaneously and in real time throughout the reaction. Relevant information required for identifying the most likely nucleation locus was intended to be obtained by extrapolating the data on cri tical parameters to time zero.

PAGE 47

27 The critical parameters of the reacti on mixture monitored throughout the reaction using this sensor array were: 1. Composition of the particles 2. Free emulsifier concentration and 3. Size distribution of the partic le and droplet population/s Online densitometer was employed in obtaini ng the online estimate of the conversion of the reaction mixture. The inli ne estimate of the interfaci al tension was obtained by implementing the inline surface tensiometer. Jaime Vara21 has described in great detail the implementation of the densitometer and the surface tensiometer. The sensor array developed at our laboratory incorporat ed the spectroscopic measurements for characterizing the latex mixture throughout th e reaction apart from the densitometer and interfacial tension measurements. This sensor array is described in detail in Appendix H. A discussion supporting the choice of the sp ectroscopic techniques over other available techniques for investigating the number a nd size distribution char acteristics of the particle populations compri sing the emulsion / latex pres ent inside the reactor is presented henceforth. Giannetti10 underscores the need to have fast and reliable measurements that would provide information on the actual partic le size distribution da ta of the particle populations present in the latex for understanding the nucleation mechanism. According to Gilbert1, the complex nature of the process such as number of phases present etc, forces a need to introduce parameters that cannot be determined by prior information.

PAGE 48

28 To obtain information on particle size distribution a number of techniques have been used by a number of researchers. A ngular light scattering, small angle neutron scattering and transmission el ectron microscopy are the most widely used ones. These techniques are unable to provi de an estimate of particle number and hence cannot provide adequate information required for investig ation of nucleation mechanism in emulsion polymerization. Information on particle number of a particular part iculate population is very pertinent in deciding the reaction locus, since the probability of a particular particle population becoming the main reac tion loci increases with its increasing particle number. Microscopic techniques (including Transmission Electron Microscopy) are inadequate in characteri zing liquid –liquid systems due to instability in the “fix” stage. Boundaries for the pa rticles of size range below 10 nm appear fuzzy due to inadequate differences between the refractive index. This makes the results of the particles with size range below 10nm unreliable1. The above discussion underscores a need of an experimental technique that can completely characterize the reaction mixtur e present in the re actor throughout the reaction, quickly, continuously and in real time throughout the reaction. The development of techniques for characterizi ng the reaction latex mixture fr om its online transmission Uv vis spectrum provide the qualitative and quantitative information necessary to identify the most likely nucleation locus. Garcia Rubio22 has demonstrated the change in the shape of the transmission Uv vis spectrum as a function of particle size for well-characteriz ed polystyrene standards.

PAGE 49

29 C. Bacon23 and E Stiemle24 explored the potential of thes e spectroscopic techniques to a great degree by developing prototypes for obtaining multi angle multi wavelength measurements to obtain information on size distribution characteristics of particle populations. Andres Cardenas25 explored the limits and applic ability of this measurement technique including reflecta nce spectroscopy. S. Fisher26 and S. Thennadil27 modeled the colloidal systems from the perspective of particle-particle interactions. From the Uv vis transmission spectrum of the reaction mixtur e, information on the number of particulate populations, number of particles contained in each population and the composition of the particles of each population can be obtained. J. Mehta28 S. Marathe29. D. Imeokparia30, S. Shetty31, P. Vinnik32 obtained information on the initiat ion efficiency using labeled initiators to measure the rate of radical en try per particle. Obtaining the transmission Uv vis spectrum of the particulate mixture inside the reactor continuously and in real time was made possible by the development of the continuous sampling and parallel dilution system developed and patented by Un iversity of South Florida, Tampa33. Characterization of the particulate systems on th e basis of their partic le number, particle size distribution, chemical composition from their transmission Uv vis spectrum was made possible by the algorithms which were based on Mie scattering theory developed by Dr Luis H Garcia Rubio34,35. Paul Sacoto33 gives the information in detail on the development and the implementation of the continuous sampling and parallel dilution system.

PAGE 50

30 Maria Celis de Arce36 successfully employed the sp ectroscopy technique for the measurement of the droplet size distribution as a function of the o il phase concentration and emulsifier concentration. She concluded that the results obtained from the light scattering interpretation models are applicable in the qualitative analysis of the liquidliquid emulsions. Maria Celis s uggested that, from the singl e scattering models it could be safely inferred that sample integrity can be preserved even after successive dilutions36. This inference thus enables the use Uv vis spectroscopy as a tool for characterizing the reaction mixture to address the proble m of nucleation mechanism in emulsion polymerization.

PAGE 51

31 CHAPTER 3: PROBLEM DEFINITION 3.1 Introduction The issues associated with the main theo ries put forth until date for explaining the nucleation mechanism in emulsion polymerizati on have been discussed in the literature review section of this disse rtation. In the light of these issues, the importance of identifying the most likely nucleation site to elucidate the nucleation mechanism has been highlighted. This chapter focuses on the preliminary studies for formulating the hypothesis for identifying the most likely nuc leation locus and on th e direction for the necessary experimental work. The relevanc e of the proposed hypotheses for identifying the nucleation mechanism is discussed. 3.2 Approach Elucidating the nucleation mechanism in emulsion polymerization is the longterm goal of this research effort. In order to achieve this objective, this research focuses on exploring the “time zero” condition of the emulsion polymerization reaction for identifying the most likely nucleation. The limitations of the experimental techniques available to the researchers for reaction mixtur e characterization at time zero have been described briefly in the earlier chapter. Implementation of the UV Vis spectroscopic techniques in order to overcome the limitati ons of the existing experimental methods for the complete characterization of the r eaction mixture has also been discussed

PAGE 52

32 Characterizing the reaction mixture throughout the reaction in term s of the critical parameters and extrapolating the interval I rate data to time zero for investigating the time zero condition was the approach undertaken by the researchers to date. However, the dynamics of the reaction and the sampling issu es presented considerable difficulties. In order to overcome this difficulty, this resear ch takes the approach of characterizing the emulsion present inside the reactor at time zero, using a non-reacting “model emulsion system”. A model emulsion system is an emulsion system comprised of constituents displaying similar dispersion behavior of the monomer emulsion system under similar emulsification conditions but havi ng different optical properties. The rationale for taking the approach fo r performing the experiments with the model emulsion system is provided in the de tail in the experimental section of this dissertation (Chapter 4) along w ith the details of the system. The key issues necessary to understand the nucleation mechanism namely, 1. Plausible nucleation loci 2. Emulsifier distribution as a f unction of emulsification conditions 3. Size dependent compositional characteristics of the plausible nucleation locus were addressed with the above mentioned m odel system and characterization techniques.

PAGE 53

33 3.2.1 Preliminary Studies Prior to performing the experiments with the model system mentioned above, simulation studies relating the size distribution and compositi onal characteristics of the styrene latex mixture to the characteristic s of its transmission Uv vis spectra were undertaken. The Uv vis spectrum for the styren e latex with known op tical properties, but different size distribution characteristics were simulated and compared with the experimentally obtained spectra. This was done to explore the possi bility of the number of droplet populations likely to be present and to determine their characteristics. The details pertaining to these si mulations are outlined in 3.2.2. Simulations were performed with the help of the algo rithms developed in-house34,35. The inferences drawn from the results from the simulation studies lead to th e formulation of the main hypotheses in this research effort. 3.2.2 Algorithms Implemented The simulation programs34,35 enabled the prediction the spectra on the basis of the latex properties. These algorithms34,35 incorporated the turbidity equation (equation 3.1) to calculate the transmission spec tral features of latex with known optical properties. The turbidity equation relating the transmission sp ectrum of each of the particulate population to its particle size distribution, compositi on and particle number is based on the Mie scattering theory.

PAGE 54

34 The turbidity equation is given by37, dD D f D D m Q l Next p) ( ) ), ( ( ) 4 ( ) (2 0 0 0 (3.1) where 1. 0(0) is the optical density as a function of wavelength 0 2. Np is the number of particles 3. l is the path length 4. m(0), is the complex refractive in dex as a function of wavelength 0 m(0) is generally expressed as37 ) ( ) ( ) ( ) (0 0 0 0 0 0 n k i n n where n(0) is the real part of the refractive index and k(0) is the imaginary part of the refractive index 5. Qext is the extinction coefficient (function of the complex refractive index and the size of the particles 6. D is the diameter of the particles and f( D) is the particle size distribution of the particles. From the turbidity equation, it can be in ferred that the shape of the simulated spectrum is sensitive to the pa rticle size distribution of the particles comprising the latex along with their particle number and composition. In fact, it is evident, that for the particles of particular composition at given concentrati ons, the shape of the spectrum is determined by the size distribution only. Th e size dependence of the light scattering efficiency and the absorption efficiency of a part icle is presented in Appendix A.

PAGE 55

35 Figures 2,3,4,5 depict the change in the simulated Uv vis transmission spectrum of styrene-in-water emulsion. Noticeable chan ge in the shape of the simulated spectrum can be observed with the change in the mean diameter and standard deviation of the particle population comprising the latex. T hus, from the simulation studies it can be demonstrated, that the change in the size dist ribution characteristics of the latex mixture will appreciably reflect upon its measured transmission spectrum. Similar simulations were carried out fo r decane-in-water emulsion system with SDBS (Sodium Dodecyl Benzene Sulfonate) as the emulsifier. Decane being a linear hydrocarbon, does not show an absorption sign al whereas, the emulsifier SDBS has a distinct signature in the UV region of the spectrum. Performing simulations for such a system enables the study of the changes in the spectral features of the emulsifier as a function of the size distribution characteri stics of the emulsion. Figures 6 through 12 depict the results of the simulations for the model emulsion system. The emulsifier spectrum signal was found unique to the size of the particle on wh ich it is distributed (refer Figures 6, 7, 8). Similar observations can be made for the particles with same sizes, but different structures (r efer Figures 9,10,11,12). It is evident from the results of the si mulation studies described above, that the shape and the spectral features of simulated transmission Uv vis spectrum of a particle population is extremely sensitive to th e size, composition and the structural characteristics of the particles comprising it. A comparison of the simulated spectrum

PAGE 56

36 with the experimentally observed spectru m for the styrene-in-water emulsion SDS (Sodium Dodecyl Sulfate) as the emulsifier revealed that no simulated spectrum of a single particle population matche d the shape of the measured spectrum. This indicated the presence of two or more droplet populations that may have contributed to the experimentally observed transmission spectra l signal of the styrene emulsion latex. Simulations were done in order to expl ore the size charac teristics of the constituent populations of the styrene emulsion. The parameters of the simulations were set to generate the transmission spectra of the styrene latex emulsion, such that the simulated spectral features (unique to the si ze characteristics of th e particle populations) were compatible with the observed spectral characteristics of the experimental transmission spectrum. The simulated spectra with spectral features similar to the measured spectrum were then added in pr oportion, to match its shape (refer Figures 13,14,15). Inferences can thus be drawn on the size characteristics of the particle populations comprising this emulsion. The next section describes the theoretical considerations and the mathema tical calculations needed to pe rform the above procedure. 3.2.3 Simulations and Experime ntal Analysis Performed In order to compare the shapes of the e xperimental spectrum and the calculated spectrum (a combination of the simulated sp ectra of the large and the small particles added in proportion), the following procedur e was performed. The simulated spectra having the spectral features compatible with those of the experimental spectrum were

PAGE 57

37 normalized. The appropriate simulated spect rum was normalized by its respective area. Normalization of the spectrum by its ar ea is done as shown in equation 3.2. f i o od ) ( ) ( ) ( (3.2) where 1. I = Initial wavelength 2.f =Final wavelength The experimental spectrum is similarly normalized. The simulated spectra of the small and large particle populat ions were then added in pr oportion. The shape of the resulting spectrum was then compared with that of the experimental spectrum. The procedure described above is continued until the shapes of the calculated spectra and the experimental spectra were in agreement with each other. Such addition of the normalized simulated spectra (in the requi red proportions) for matching the shape and the spectral features of the normalized experimental spect rum provides relative concentrations of the populations that comprise the emulsion. A comp arison of the shapes of the spectra done in this manner eliminates the effect of th e particle number. The procedure for carrying out the above calculations is simila r to as given in Appendix B. It has been demonstrated by Alupoei38 that the uncertainty associated with the number of particles in a population is a majo r contributor to the propagated experimental error and it may bias the conclusions relativ e to changes in the spectra due to other

PAGE 58

38 variables. It was for this reason that the comp arison of the shapes of the experimental and calculated spectra was done by eliminating the effect of the number of particles. The results are shown in Figures 13,14 and 15. Performing these calculations and comparisons led us to the following inferences. 3.3 Inferences Drawn from the Simulation Studies 1. A particle population of the nano-droplets with a mean droplet size of 30 to 100 nm is present during the beginning of th e reaction apart from the large monomer droplet population of size range 1 to a few microns in diameter. 2. At least 40 to 50% of monomer is present in the nano-droplets. 3. High interfacial area offered by the nanodroplets owing to small size and high particle number make them a st rong candidate for becoming the main reaction loci. 4. There is a need to determine the true parameters of the population of the nanodroplets owing to the spectral contribution of the large particles to the absorption component of the experimental spectrum. 3.4 Hypothesis Proposition The results inferred from the simulation studies provide the basis for proposing the following hypotheses that are necessary to understand the nucle ation mechanism in emulsion polymerization.

PAGE 59

393.4.1 Hypothesis No 1: Nano-droplets are the most likely nucleation loci in emulsion polymerization reactions. 3.4.2 Hypothesis No 2: The initial emulsification condi tions govern the size of the nanodroplets. 3.4.3 Hypothesis No 3: Early time characterization of nano-droplet population and particles in emulsion polymerizati on, in terms of their si ze, size distribution and composition can be accomplished using Uv vis spectroscopy techniques and model emulsion systems. The relevance of the hypothesis for elucidating the nucleation mechanism is presented in the next section. 3.5 Relevance of the Proposed Hypotheses for Identifying Nucl eation Mechanism 3.5.1 Hypothesis No 1: Nano-droplets are the most likely nucleation loci in emulsion polymerization reactions. The chapter on literature review descri bes each of the proposed (and widely accepted to different degrees) theories on nucleation mechanism and the different nucleation locus proposed by each of them. The main nucleation locus in emulsion polymerization process thus continues to remain a subject of active debate. It is very important that the main r eaction locus be identified for studying the nucleation mechanism because it will be at this reaction site that the nucleation

PAGE 60

40 phenomenon will actually occur resulting in the birth of the new particle. The nucleation mechanism will thus be largel y governed by the characteristics of the reaction site. Simulation studies indicated the presence of appreciable amount of dispersed phase in the nano-droplet population and the large interfacial area offered by them. The hypothesis on the nano-droplets being the main nucleation locus is th erefore presented. 3.5.2 Hypothesis No 2 : The initial emulsificatio n conditions govern the size of the nanodroplets. Initial emulsification conditions are charac terized by the emulsifier concentration, the pH of the suspending medium, the temper ature at which the emulsion is prepared and the shear rate of the mixer for a vessel with given mixing characteristics. These conditions have a strong influence on the in itial size distribution of the nano-droplets. The initial emulsifier distribution is such, th at the free energy due to the large interfacial area present between the two immiscible phases is decreased, thereby imparting stability to the emulsion. The primary function of the em ulsifier is to reduce the interfacial tension (commonly denoted by ‘’) and hence, the free energy required to disperse a liquid. The expression for the free energy to disperse a liqui d of volume ‘V’ with drops of radius R in a solvent is given by39 Gem = *3* R V (3.3) Change in the emulsification conditions cause changes in the dissociation characteristics of the emulsifi er and emulsifier efficiency for stabilizing the interfacial

PAGE 61

41 area between the continuous phase and the dispersed phase. In other words, the size, the number and the composition characteristics of the droplet populat ions comprising the emulsion are affected by the emulsification conditions. Since the nano-droplets are smaller in size and higher in number than the monomer droplets, the surface area required to be stabilized for the nano-droplet population is expected to be much higher th an that of the monomer droplets. It is therefore expected that most of the emulsifier shall be utilized for stabilizing the nanodroplet population. The generation of the el ectrical double layer ar ound the nano-droplets due to the surfactant distribution before th e beginning of the reaction and around the polymer particles after the initiation of the reaction, imparts stability to the reaction mixture by preventing coalescence/coagulati on between the droplets/particles. This electric double layer also affects the rate of radical entry by offering resistance to the entering radicals. Thus the density of the emulsifier distributed over the nano-droplets which is a function of initial emulsificati on conditions, has a prof ound effect on the rate of nucleation and hence th e nucleation mechanism.

PAGE 62

423.5.3 Hypothesis No 3: Early time characterization of nano-droplet population and particles in emulsion polymerizati on, in terms of their si ze, size distribution and composition can be accomplished using Uv vis spectroscopy techniques and model emulsion systems. According to Lichti9,10, a very fast and reliable measurement technique for characterizing the latex in terms of size and composition is necessary to understand nucleation mechanism. Limitations associated with the currently used experimental techniques by different researchers caus e considerable difficulty in early time characterization of the reac tion latex and hence valuable information necessary to understand the nucleation mechanism is lost. Early time characterization of the emulsi on mixture present inside the reactor is relevant in undertaking the studies for elucid ating nucleation mechanism. This is because it provides the information on the most likely nucleation site s for the emulsion polymerization reaction. Once, the likely nucl eation sites are identi fied, characterizing them on the basis of their size, size dist ribution, number and composition will provide valuable insight into identifying the mechanism of nucleation. The compositional characteristics of the droplets during the early stages of the reaction influence the rate of nucleation. The ra te capture efficiency of these particles is a function of the concentrati on of the monomer in them1,9. In the case of micellar

PAGE 63

43 nucleation (assuming it to be the true nucle ation mechanism) according to Olayo et al9,10 the radical capture efficiency is a function of the monomer concen tration in micelles. According to the homogeneous a nd coagulative nucleation theories2,13,16, the precursor particles are formed due to the propagation of the initiator radical until they precipitate and later coagulate with other su ch particles to achieve stabilit y. This theory states that the rate of coagulation (and hence the ra te of nucleation) is dependent upon the concentration of the monomer in the precursor particles. Thus, the concentration of the monomer in the precursor pa rticles (assuming homogeneous and coagulative nucleation to be the true nucleation mechanism) or in the micelles (assuming the micellar theory for nucleation is true) presents a ve ry relevant issue that needs to be addressed in order to understand the nucleation mechanism. It has been pointed out in the earlier section of this chapter, that Olayo et al9,10 used the Morton’s equation as a function of size to calculate the concentration of the monomers in the particles as a function of size. Gilbert1 expresses his reticence to use the Morton’s equation to predict the concentration of the monomer inside the particles depending upon the radius for the early reacti on times. According to him, the monomer concentration inside the particles show noticeable dependence on the radius of the particle and hence will be significantly less than th e value predicted by Morton’s equation. According to Gilbert1, the lack of knowledge of the size dependence of the concentration of the monomer inside the par ticles makes it difficult to use stage I rate data to gain information about particle formation.

PAGE 64

44 Thus, if the nano-droplets are the true si tes for nucleation (proposed as the main hypothesis in this dissertation), then the radi cal capture efficiency of these nano-droplets (thus converting the nano-droplets to part icles) and their nucl eation rate will be dependent upon their size and the concentra tion of the monomer in them. Hence, it is important that the compositional characteristics of the likely reaction sites as a function of their size and the changes they undergo over time as the reaction progresses be investigated. This research effort therefore takes th e approach of characterizing the latex mixture present in the reactor before the re action is initiated (esse ntially a liquid-liquid emulsion system at early times) in order to understand the nucleation. The results from the simulation studies s howed that the shape of the simulated transmission Uv vis spectra of styrene in water emulsion changed appreciably with the change size distribution charac teristics of the droplet popu lations that comprise the emulsion (refer Figure 2 through Figure 5). Similar observations were made for the simulation studies involving Decane in wate r emulsion with Sodium Dodecyl Benzene Sulfonate (SDBS as the emulsifier). The conclusions from the works of Maria Celis36 stated that the sample integrity is maintain ed throughout successive dilutions. It was thus inferred that the spectroscopi c tools could be implemented to characterize the reaction mixture during early times to obtain the information required for identifying the nucleation mechanism.

PAGE 65

45 In order to establish the relationship betw een the size of the r eaction site and its compositional characteristics, it is necessary to characterize the dr oplet populations in terms of their size and compos itions. In order to achieve this, the information on the particle size distributions (PSD s) of the reaction loci in th e early stages of the reaction needs to be coupled with emulsifier and oil phase mass balance. Use of the model emulsion system comprised of molecules w ith the similar dispersion properties but different optical properties is therefore considered. The study of the distribution of the emulsion component of interest was undertaken by selecting the appropriately labeled compound. Information on the composition of the droplet populations co mprising the emulsions and their size characteristics can thus be obt ained by taking this approach. 3.6 Conclusion Simulation studies indicate the nano-dr oplet populations to be the prime candidates for being the most likely nuclea tion loci. The size and the compositional characteristics of the nano-droplet populations are expected to change with the change in the emulsification conditions. The Uv vis spectroscopic techniques can be used to characterize the liquid-liquid emulsion present in the reactor at time zero in terms of number, size distribution and composition of each of the droplet populat ions comprising it. Appropriate model

PAGE 66

46 emulsion system can be employed in orde r to achieve this objective. Relevant information on size and compositional character istics of the most likely nucleation loci that is necessary to elucidate the nucl eation mechanism in emulsion polymerization processes can be obtained by following this approach. Details pertaining to the e xperiments with the model systems and justification for their consideration are provided in the experimental section.

PAGE 67

47 Figure 2 Change in the Shape of the Simulate d Transmission Spectra of the Nano-droplet Population of Styrene in Wate r Emulsion as a Function of its Standard Deviation. and Constant Mean Diameter (Dn) Figure 3 Change in the Shape of the Simulated Transmission Spectra of the Nano-droplet Population of Styrene in Water Emulsi on as a Function of its Mean Diameter (Dn) 200 300 400 500 600 700 800 900 0 0.5 1 1.5 2 2.5 3 3.5 Std dev 0.6 Std dev 0.4 Std dev 0.3 Std dev 0.2 Mean size 50 nmNormalized Optical DensityWavelength (nm) 200 300 400 500 600 700 800 900 0 0.2 0.4 0.6 0.8 1 1.2 1.4 150 nm 100 nm 50 nm 30 nm Std dev = 0.2 Wavelength (nm)N o rm a l i z e d O p t i c a l D e n s i t y

PAGE 68

48 Figure 4 Change in the Shape of the Simu lated Transmission Spectra of the Large Droplet Population of Styr ene in Water Emulsion as a Function of its Mean Diameter (Dn) Figure 5 Change in the Shape of the Simulated Transmission Spectra of the Large Droplet Population of Styrene in Water Emulsion as a Function of its Standard Deviation 200 300 400 500 600 700 800 900 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 5 microns 4 microns 3 microns 2 microns Std dev = 0.2Normalized Optical DensityWavelength (nm)200300400500600700800900 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 Std dev = 0.5 Std dev = 0.3 Std dev = 0.4 Std dev = 0.2Dn= 3 micronsN o r m a l i z e d O p t i c a l D e n s i t yWavelength (nm)

PAGE 69

49 Figure 6 Comparison of the Simulated Tran smission Spectra of Decane in Water Emulsion (with SDBS Emul sifier) for Droplet Populations of Different Sizes 200 220 240 260 280 300 320 340 360 380 400 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 Normalized Optical DensityWavelength (nm) 100 nm 75 nm 50 nm 10 nm

PAGE 70

50 Figure 7 Effect of Particle Size for Decan e in Water Emulsion with SDBS as the Emulsifier on the Simulated Transmission Spectra for the Latex Consisting of Sma ll Droplets (Amplified Lower Wavelength Region) Note: The emulsifier signal looks very di fferent when the emulsifier is distributed over small droplets than when it is in solu tion. Note the change in the spec tral features of the emulsifier. Emulsifi er spectral features are uni que to the size of the droplets over which they are distributed. 200 220 240 260 280 300 320 340 360 380 400 0 0.2 0.4 0.6 0.8 1 1.2 1.4 x 103 Normalized Optical DensityWavelength (nm) Beer-Lamberts Law 20 nm 15 nm 10 nm

PAGE 71

51 Figure 8 Comparison of the Effect of Part icle Size for Decane in Water Emulsion with SDBS as the Emulsifier on the Simulated Transmission Spectrum of the Emulsifier Dist ributed on the Droplets (A mplified Lower Wavelength Region) Note: Emulsifier spectral features are unique to the size of the droplets over which they are distributed. 200 220 240 260 280 300 320 340 360 380 400 0 1 2 3 4 5 6 x 10-3 Wavelength (nm)Normalized Optical Density 50 nm 30 nm 10 nm

PAGE 72

52 Figure 9 Comparison of the Effect of the Particle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size (Small Droplet Size 10 to 30nm in Diameter) 200 220 240 260 280 300 320 340 360 380 400 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Wavelength (nm)Normalized Optical Density Beer Lamberts Law Homogeneous Core Shell

PAGE 73

53 Figure 10 Comparison of the Effect of the Pa rticle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size of Range 10 to 30 nm (Amplified Lower Wavelength Region) Note: The emulsifier spec tral features are unique to the particle structure. 200 220 240 260 280 300 320 340 360 380 400 0 0.2 0.4 0.6 0.8 1 1.2 1.4 x103 Wavelength (nm)Normalized Optical Density Core Shell Beer Lamberts Law Homogeneous

PAGE 74

54 Figure 11 Comparison of the Effect of the Pa rticle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size (Bigger Droplet s of Size Range 50nm to 100nm in Diameter Dn) 200 220 240 260 280 300 320 340 360 380 400 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 Normalized Optical DensityWavelength (nm) Beer Lamberts Law Homogeneous Core Shell

PAGE 75

55 Figure 12 Comparison of the Effect of the Pa rticle Structure on the Simulated Transmission Uv vis Spectrum of the Latex for the Same Droplet Size of Range 50 to 100 nm (Amplified Lo wer Wavelength Region) Note: The emulsifier spectral features are unique to the particle structure. 200 220 240 260 280 300 320 340 360 380 400 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 x 10 3 Normalized Optical DensityWavelength (nm) Beer Lamberts Law Homogeneous Core Shell

PAGE 76

56 Figure 13 Comparison of the Shape of the Norm alized Experimental Spectrum with that of the Simulated Transmission Spectrum of Large and Small Droplets of Different Mean Diameters (Dn) added in Equal Proportions Figure 14 Comparison of the Shape of th e Simulated Transmission Spectrum of Large Droplets about 3 mi crons and Small Droplets of 30nm Mean Diameter (Dn) added in Different Pr oportions and the Normalized Experimental Spectrum 200 300 400 500 600 700 800 900 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Extinctions of small particles Size 30 nm Extinctions of big particles Size 3 microns Experimental spectrum Extinctions 50% 30nm small particles + 50% 3 microns big particles Extinctions 40% 30 nm small particles + 60% 3 microns big particlesNormalized Optical DensityWavelength (nm) 200 300 400 500 600 700 800 900 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 Extinctions of small particles Size 30 nm Extinctions of big particles Size 3 microns Experimental spectrum Extinctions 50% 30nm small particles + 50% 3 microns big particles Extinctions 40% 30 nm small particles + 60% 3 microns big particlesNormalized Optical DensityWavelength (nm) 200 300 400 500 600 700 800 900 0 0.005 0.01 0.015 0.02 0.025 Experimental spectrum Normalized extinction spectra 30nm + 3 microns Extinction spectra 50nm + 3 microns Extinctions 100nm + 3 microns Std dev = 0.2Normalized Optical DensityWavelength (nm) 200 300 400 500 600 700 800 900 0 0.005 0.01 0.015 0.02 0.025 Experimental spectrum Normalized extinction spectra 30nm + 3 microns Extinction spectra 50nm + 3 microns Extinctions 100nm + 3 microns Std dev = 0.2Normalized Optical DensityWavelength (nm)

PAGE 77

57 Figure 15 Comparison of the Shape of the Simulated Transmission Spectrum of Large Droplets a bout 3 microns and Small Droplets of 50nm Mean Diameter (Dn) a dded in Different Proportions and the Normalized Experimental Spectrum 200 300 400 500 600 700 800 900 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 Experimental spectrum Extinctions 42% 50 nm small 58% 3 microns large Extinctions 50% 50nm small 50% 3 microns large Wavelength (nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 78

58 CHAPTER 4: EXPERIMENTAL WORK UNDERTAKEN 4.1 Introduction Simulation studies presented in Chapter 3 indicated that the reaction mixture at time zero could be characterized using Uv vi s spectroscopic techniques. The preliminary studies suggested the presence of significant amount of the dispersed phase in the nanodroplet population. The nano-dr oplet population offered high interfacial area (owing to their small size and high particle number). Thes e inferences resulted in the formulation of the hypothesis proposing the nano-droplets to be the main nucleation loci in emulsion polymerization processes. In order to test the above hypothesis, experimental efforts were undertaken to characterize the reaction mixt ure at time zero condition in terms of number, size characteristics and composition of each of its comprising dr oplet populations. Experiments were performed using a non-reac ting model emulsion system having similar dispersion characteristics as that of the orig inal monomer in water emulsion but different optical properties. These experiments were perf ormed to investigate the effects of initial emulsification conditions on the size and the compositional characteristics of the dispersed phase. Spectroscopic techniques were used to characterize the dispersed phase of the model emulsion system.

PAGE 79

59 Details of the experimental work performed with the model emulsion system and the rationale for taking this approach ar e presented in this chapter. 4.2 Rationale for Using Model Emulsion System Results inferred from the simulation studi es presented in Chapter 3 (Figures 13 through 15) indicated a presence of more th an one population of th e dispersed phase of the emulsion at time zero of emulsion polymer ization process. For the identification of the most likely nucleation locus, it is important that each population of the dispersed phase be characterized in terms of its size, numbe r and composition. Such characterization of the dispersed phase requi res information on the distribution of each component (oil and emulsifier) in the disp ersed phase population. Surfactant distribution determines the composition and the feasibility of the existence of the droplet population having particular size characte ristics. Hence, the study of surfactant distribution is important for identification of the most likely nucleation locu s. For obtaining the information on surfactant distribution, it is neces sary to identify the surfactant signal and determine its contribution to the measured spectra. Performing experiments with monomer emulsion systems makes it difficult to identify the surfactant signals since the monomers (e.g Styrene) have strong absorption peak in the UV region overlapping the signal of the surfactant. It was for this r eason that the experiments with model emulsion systems consisting of components with simila r physico-chemical pr operties but different optical properties as proposed in th e Chapter 3 of this dissertation.

PAGE 80

60 Under similar emulsification condi tions (temperature, pH, surfactant concentration), dispersed phases having simila r physico-chemical pr operties (viscosity, vapor pressure, density, interfacial tension) display similar dispersion characteristics39. The components of the model emulsion syst em selected to imitate the dispersion characteristics of the reaction latex mixtur e at time zero therefore have their physicochemical properties similar to the respective components of the reaction latex mixture present inside the reactor at time zero. Th e distribution of the components of emulsion system (namely emulsifier, oil phase and c ontinuous phase) under consideration can be studied by the use of appropriately labeled compounds. The optical properties of the selected components for the model system are different from those of the respective compone nts of the reaction latex mixture such that good contrast for the spectroscopy measurement is provided. The optical properties of the compounds of interest are repor ted in Appendix C. The inves tigation of the distribution of the emulsion component of interest (emuls ifier or oil) is po ssible by identifying and evaluating its spectral contribu tion to the measured spectrum. The primary advantage of using the non-reacting model emulsion system is the removal of the time constraint for immediate sampling of the monomer emulsion mixt ure, present inside the reactor at time zero.

PAGE 81

61 4.3 Experiments with Model System In this section, the materials comprisi ng the model system are specified. The experimental design strategy with the objective of systematically investigating the effect of the different manipulated variables under co nsideration on the emulsion characteristics is explained. A justification for selecting th e temperature, pH of the suspending medium and the surfactant concentration as the ma nipulated variables describing the initial emulsification conditions for studying their effects on the emulsion characteristics is provided. 4.3.1 Materials and Methods 4.3.1.1 Materials The model emulsion system employed mode l molecules with emulsifier (Sodium Dodecyl Benzene Sulfonate) having a distinct absorption peak in the Uv vis signal of its transmission spectrum. Decane was used as the transparent oil phase (dispersed phase) to mimic the dispersion behavior of the monome rs owing to its similar physico-chemical properties with styrene, butyl methacrylate et c. The reagents were obtained from Sigma Aldrich. In Table 4.1,the comparisons of the approximate values of the physical properties of decane40,41,43,46,47 styrene40,41,42,44 and butyl methacrylate21,40,41,45,48 are shown.

PAGE 82

62 Table 4.1 Comparison of the Physical Propertie s of Styrene, Butyl Methacrylate and Decane Physical Properties Styrene Butyl Methacrylate Decane Density (g/cm3) 0.906 0.889 0.73 Vapor pressure (mm Hg) 5 4.9 1.4 Interfacial Tension (dynes/cm) 27.7 28 30.4 Viscosity (centipoise) 0.675 0.832 0.863 Phosphate buffer saline (PBS) was used as the continuous phase. The emulsions were prepared in a buffer of known pH as the suspending medium. This was done to explore the dispersion characte ristics of the oil phase as a function of pH since the emulsifier efficiency was expected to change as a function of the pH of the suspending medium. The pH of the PBS buffer was decr eased by the addition of HCl (pH 1) and increased by the adding NaOH of pH 13. 4.3.1.2 Methods The emulsion was prepared under differen t conditions as described in section 4.3.6 and the transmission Uv vis spectrum of the diluted emulsion was recorded. The buffer was prepared at room temperature in such a way that it w ould possess the desired pH at elevated temperatures. Three replicat e measurements of the transmission spectrum were obtained for each of the experiments. The mean transmission spectrum for each experiment performed was obtained. The 95% confidence intervals for the optical densities at each wavelength were calculated. In comparing the shape of the spectra, the

PAGE 83

63 optical density for any particular wavelengt h lying outside the upper and lower limits of the confidence intervals was considered to be an outlier since it wa s considered to be statistically different. The spectra l results are shown in Appendix E. Three replicate experiments were performed at the room te mperature. The results of the transmission spectra so obtained were compared and are shown in Figure 17. The emulsion was characterized in terms of the particle size distribution, particle number and particle composition for each popul ation of the dispersed phase comprising it from its transmission Uv vis spectrum, us ing turbidity equation (equation 3.1). The algorithms developed in-house performed cons trained optimization for characterizing the emulsion present inside the reactor from its transmission spectrum. The mass balance on the oil was the constraint that was implemented for resolving the spectrum to characterize the emulsion. The transmission Uv vis spectr a of the emulsion could only be partially resolved (from 280 to 820 nm) for characterizati on purposes since the spectral features of the emulsifier changed considerably in the lower wavelength region. In order to resolve the complete Uv vis transmission spectrum of the emulsion in the model system, more work needs to be done on the estimates of the optical properties of the surfactant. The work performed on for estimating the optical properties of the su rfactant during this research is reported in Appendix C.

PAGE 84

64 4.3.2 Equipment and Experimental Setup The emulsion was prepared in a 500 ml gl ass reactor. A temperature controller supplying the necessary amount of heat to the heating jacket electrically, maintained the desired temperature of the emulsion inside the reactor. Two baffles for breaking the vortex caused as a result of agitation were at tached to the covering lid. The covering is placed on the reactor such that the stirrer rod passes thr ough the central opening of the lid. The stirrer rod fits into the chuck of th e motor. The emulsion is kept under constant agitation with the help of a stirrer rota ting 500 RPM. The RPM was verified with a stroboscope for each experiment. A sample slip-stream was drawn continuously from the reactor with the help of the sample pump and sent to the dilution system where it was diluted with the suspending medium. The temp erature of the diluent was maintained the same as the emulsion with the help of another temperature c ontrol system. This temperature control system c onsisted of an electrically powered heating mantle for maintaining the temperature of the diluent. The temperature sensing thermocouple for this temperature control system was immersed in the diluent (suspending medium is used as the diluent). The sample emulsion stream was drawn fr om the reactor and sent to the dilution system where it came in contact with the di luent. The transmission spectrum of this diluted emulsion was recorded at desired sample times. The flow rates of the diluent stream and the sample slip-stream were such that the transmission spectrum of the diluted emulsion was within the linear range of the spectrometer (opt ical density below 1). The

PAGE 85

65 complete protocol for operating the diluti on system is presented in Appendix I. The spectrometer used was a Hewlett Packard sp ectrometer model number HP8452A with a resolution of 2 nm. The sample cell holder of the Hp spectrometer was maintained at the desired temperature with the help of a nother electrical temperature controller manufactured by Perkin Elme r (C5700820). All the hoses and tubes were insulated with glass fiber insulation to minimize the heat loss. The pH of the suspending medium and the diluent was monitored with the help of the pH meter manufactured by Fisher (Fisher accumet model number 610). The custom-made aluminum surface-tensiometer probes through which the nitrogen is bubbled are placed in the reactor. The in terfacial tension of the emulsion is measured with the surface-tensiometer. Figures 18 and 19 show the schematic of the entire experimental setup. 4.3.3 Experimental Design Strategy The experiments performed with the mode l systems were designed to explore the effects of the initial emulsification cond itions. The variables describing the initial emulsification conditions are listed as follows: 1. Surfactant concentration, 2. Temperature, 3. pH. The effect of the above mentioned variab les on the size distribution characteristics of the dispersed phase of the emulsion in th e model system are studied. These variables are expected to affect the emulsifier beha vior in a micellar solution. Change in the

PAGE 86

66 emulsifier behavior is expected to affect characteristics of the dispersed phase of the emulsion since the characteristics of the di spersion (interfacial area stabilized, composition of the dispersed phase, size dist ribution of the droplets of the dispersed phase population, decision on the continuous ph ase etc) are decided by the emulsifier characteristics. The justification for the se lection of the above-m entioned variables for studying their effect on the dispersion characte ristics is provided in the next section. 4.3.4 Relevance of the Experimental Variables 4.3.4.1 Effect of Surfactant Concentrat ion on Initial Distribution of the Particle Populations Preparation of the emulsion involves usage of shear force to achieve the dispersion of one liquid phase into another. The reduction of the shear force requirement for achieving dispersion and maintaining the stability of the emulsion is the primary function of the surfactant (emulsifier). The emulsifier concentration a ffects its dynamic characteristics46 thereby affecting the characteristics of the stable di spersed phase. The stability of the dispersed phase is expected to increase with the increase in the emulsifier concentration and hence is expected to reflect upon compositional, particle number and the particle size distribution characteristics of the dispersed phase. Thus the emulsifier concentration governs the characteristics of the dispersed phase and the characte ristics of the most likely nucleation locus. It is therefore importa nt to explore the eff ect of the surfactant concentration on the initial di stribution of droplet populati ons of the dispersed phase.

PAGE 87

67 4.3.4.2 Effect of Temperature on Initial Distri bution of the Particulate Populations Change in temperature alters the inte rfacial tension. The surface tension of a micellar solution decreases with increasing temperature since the effectiveness of the surfactant is dependent on the temperature. This change in the surfactant kinetics causes a change in the adsorption characteristics of emulsifier on the dispersed phase and the viscosity of the emulsion 39. The characteristics of the dispersed phase of the reaction mixture are thus expected to change by the changes in temperature thereby potentially affecting the choice of the most likely nucleation locus. Hence it is necessary to investigate the effect of the temperature on the size and composition characteristics of the dispersed phase. 4.3.4.3 Effect of pH on Initial Distribut ion of the Particulate Populations The change in pH of the emulsion affect s the dissociation characteristics of the emulsifier. The interfacial area between th e continuous and the dispersed phase is dependent upon the dissociation characteristics of the emulsifier. Hence the change in the pH of the suspending medium is going to affect the size and the composition characteristics of the dispersed phase, which in turn is very critical in deciding the most likely nucleation locus. It is therefore necessary to investigate the effect of pH on the characteristics of the droplet populations of the dispersed phase. In order to observe the effects of surfactant co ncentration, temperature and pH of the suspending medium on the emulsifier characteristics (since the em ulsifier characteristics will affect the characteristics of the dispersion), pre liminary experiments were performed on the

PAGE 88

68 surfactant micellar solutions. The surfactan t micellar solutions of Sodium Dodecyl Benzene Sulfonate (SDBS) was studied spectro scopically to observe the micelle forming behavior of the surfactant molecules at diffe rent conditions of surf actant concentration, temperature and pH. The concentration of SDBS was well above its critical micellar concentration and hence, was considered as a dispersion solution of a ggregated surfactant molecules (micelles). From the turbidity equa tion (equation 3.1) it can be inferred that the change in the shape of the transmission Uv vis spectrum of the micellar solution reflects upon the change in the size and number charact eristics of the micelles formed. Changes in the aggregation behavior of the surfactant molecules for forming micelles is thus indicated by the change in th e shape of the transmission Uv vis spectrum of the micellar solution. The differences in the shapes of the transmission Uv vis spectra with the change in the surfactant concentrat ion, temperature and the pH of the suspending medium are reported in Appendix D. From the results obta ined form the preliminary experiments, it can be inferred that the surfactant concentra tion, temperature and pH of the suspending medium affect the emulsifier behavior. The changes in the emulsifier behavior as a function of the mentioned variables are e xpected to affect the size, number and composition characteristics of the droplet populations of the disp ersed phase of the emulsion. 4.3.5 Experimental Design Strategy It has been mentioned in the earlier s ection that the surfactant concentration, temperature and pH of the suspending medi um are the three variables describing the

PAGE 89

69 initial emulsification conditions, whose effect s on the characteristics of the dispersed phase need to be investigated. The expe riments were therefore designed as a 2f factorial design (f = 3). However, the design of th e experiments was not symmetric since the extremes of each of the variab les in order to see significant differences in the shapes of the transmission spectra of the emulsions were needed to be explored. This resulted in a skewed design of experiments with two leve ls and three variables. The experimental conditions for each experiment are enumerated in Table 4.2. Figure 16 is a pictorial representation of the experimental design strategy. The upper and the lower levels of the contro l variables of interest are explained as follows: 4.3.5.1 Surfactant Concentration The higher level of the surfactant co ncentration was chosen to be 0.046 surfactant/oil ratio in the emulsion recipe while the lower limit of the surfactant concentration was chosen to be 0.0154 surfact ant/oil ratio. In both the recipes, for the emulsion prepared, surfactant concentrati on was higher than the critical micellar concentration. These levels of surfactant concen trations were selected such that there was enough concentration difference between the higher and the lower levels of the surfactant concentrations and yet both of the extremes were above CMC. The presence of micelles in both the recipes was thus en sured and hence the change in the emulsion characteristics in the presence of the micelles could be studied. The higher lim it of the surfactant

PAGE 90

70 concentration was almost 1.3 times that of the surfactant concentr ation used in the standard recipe of the emul sion polymerization experiments1. The lower limit was less than half the surfactant concentration of the standard recipe mentioned above. 4.3.5.2 Levels of pH The pH levels of the suspending media were decided based upon the dissociation characteristics of the surfactant. The pKa va lue of the surfactant was expected to be closer to that of the Dodecyl Benzen e Sulfonic Acid. The pKa value for 1-[(4butylphenyl) sulfanyl] trioxida ne, an organic compound, having very similar structure to the Dodecyl Benzene Sulfonic Acid was calcula ted to be 6.91 with an error of 0.41. This pKa was calculated using the so ftware developed by the ACD50,51 laboratory. Structural similarity between Dodecyl Benzene Sulf onic acid and 1-[(4-butylphenyl) sulfanyl] trioxidane justified the expectation for the cl oseness in their pKa value. The lower limit of the pH of the suspending me dium was selected to be 2 (m uch lower than its expected pKa value). This was done so that the eff ect of the non-dissociat ed surfactant on the emulsion formation could be observed. The highe r limit of the pH was chosen to be 10. The surfactant was expected to dissociate comp letely at 10 pH. Thus, the effect of the completely dissociated surfactant on the emulsion formation can be observed. The comparison of the emulsifier capability wh en dissociated partially to that when dissociated completely can thus be studied.

PAGE 91

71 4.3.5.3 Temperature The higher and lower limits of temperature were initially chosen to be 60 degree Celsius and 50 degree Celsius respectively because this is the temperature range typically maintained for emulsion polymerization reacti ons. However, little difference in the shape of the spectrum was observed within this temp erature range of intere st for the first six experiments in the design. Therefore, the lower limit of the emulsion preparation was chosen to be 22oC (room temperature) thus skewing the experimental design strategy. 4.3.6 Experimental Procedure 150 ml of the suspending medium is adde d to the 500ml glass reactor placed in the heating jacket of the temp erature controller that was used to control the temperature of the emulsion in the reactor. To the susp ending medium was added 30 ml of surfactant solution. This surfactant solution was made by the addition of the surfactant in required quantity as per the recipe to 30 ml of the suspending medium. The thermocouple for sensing the temperature inside the reactor wa s then lowered inside the reactor along with the surface tensiometer probes. The contents in the reactor were then heated until the desired temperature set point was reached. 160 ml of decane was then added to the reactor and this mixture was then subjected to continuous agitation to form emulsion. After ensuring that the stability of the emulsion was achieved as described in the next paragraph, it was sampled for analysis purpos es. The sample slip-stream of the emulsion inside the reactor was diluted by pumping it into the dilution system with a rotary

PAGE 92

72 peristaltic pump. This dilution of the sample stream enabled the measurement of its transmission spectrum. To determine the time required to stab ilize the emulsion, another experiment in which the emulsion was prepared as above was performed. The pH of the suspending medium of the emulsion was maintained at 7 pH The Surfactant to Oil ratio in the recipe was maintained at 0.0307 as per the standard recipe. The experiment was carried out room temperature. The emulsion was sampled off line at varying time intervals. The shapes of the transmission spectra of the emulsion sampled at different times were compared by comparing the spectra normalized by their respective areas (refer Appendix F). It was observed that the shape of the transmission spectrum of the emulsion changed as a function of time indicating the change of the particle size distributions of the particulate populations comprising the emulsi on. After about two hours fifteen minutes of the addition of decane and beginning of the agitation action, the sh ape of the spectrum remained constant indicating that the stability of the emulsion had been achieved. The change in the shape of the spectra of the emulsi on as a function of tim e is represented in Appendix F. The deconvolution results of the Uv vi s transmission spectra of the emulsion before stabilization are presented in Chap ter 5 along with the pertinent discussion. 4.4 Data Analysis Cardenas, Shastry and Garcia-Rubio52 describe the transmission spectroscopic techniques as an analysis tool for charact erization of emulsion latex in great detail. The

PAGE 93

73 measured transmission spectra of the emul sions is deconvoluted using the software developed in-house34,35 to obtain information on the particle composition, particle number and the particle size distribution of each of the particle popul ation present in the emulsion. The algorithms were based on the tu rbidity equation (equation 3.1). The optical properties of the oil phase at appropriate te mperatures were used as the inputs for executing these algorithms. The estimation of the optical properties is presented in Appendix C. In the section 4.3.1.2, the analysis of the transmission Uv vi s spectral data of the emulsion using the turbidity (equation 3.1) to obtain the inform ation on the particle number, particle size distribution and composition of the particle composition comprising it has already been described. The emulsion spectra were analy zed from the wavelength region of 280 to 820 nm to determine the size distribution charac teristics of the droplet populations comprising the emulsion. Weight fraction of the disperse d phase in the small nano-droplet population was calculated from the transmission sp ectrum for each emulsification condition along with the number of particles and the size dist ribution characteristics of each of the droplet population. The results are report ed in the next chapter.

PAGE 94

74 Table 4.2 Experimental Conditions Experiment Temperature in oC pH Surfactant/ Oil ratio Experiment 1 50 2 0.0154 Experiment 2 60 2 0.0154 Experiment 4 60 10 0.0154 Experiment 5 50 2 0.046 Experiment 6 60 2 0.046 Experiment 7 22 10 0.046 Experiment 8 60 10 0.046 Figure 16 Experimental Design Strategy T = 60 C S/O=0.0154 pH=10 T = 60 C S/O=0.046 pH=10 T = 60 C S/O=0.046 pH=2 T = 60 C S/O=0.0154 pH=2 T = 50 C S/O=0.0154 pH=2 T = 22 C S/O=0.0307 pH=7 T = 22 C S/O=0.046 pH=10 T S/O pHT = 50 C pH=2 T = 60 C pH=10 T = 60 C S/O=0. pH=10 T = 60 C pH=2 T = 50 C S/O=0.0 pH=2 T = 22 C S/O=0. pH=7 T = 22 C S/O=0. pH=10 T S/O pHT = 50 C S/O=0.046 pH=2 T = 60 C S/O=0.0154 pH=10 T = 60 C S/O=0.046 pH=10 T = 60 C S/O=0.046 pH=2 T = 60 C S/O=0.0154 pH=2 T = 50 C S/O=0.0154 pH=2 T = 22 C S/O=0.0307 pH=7 T = 22 C S/O=0.046 pH=10 T S/O pHT = 50 C pH=2 T = 60 C pH=10 T = 60 C S/O=0. pH=10 T = 60 C pH=2 T = 50 C S/O=0.0 pH=2 T = 22 C S/O=0. pH=7 T = 22 C S/O=0. pH=10 T S/O pHT = 50 C S/O=0.046 pH=2

PAGE 95

75 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 Figure 17 Results Obtained from Re plicate Experiments at Center-pointNormalized Optical Density

PAGE 96

76 Figure 18 Schematic of the Reaction Ve ssel Assembly for the Experiments Performed

PAGE 97

77 Figure 19 Dilution System Assembly for Spectroscopy Measurements

PAGE 98

78 Figure 20 Reactor for the Experiments with Model Molecules Figure 21 Reactor Assembly Setup with Temperature Control Jacket for the Reactor and Surface Tensiometer Probes

PAGE 99

79 Figure 22 Dilution System Assembly to Acquire Transmission Spectra Figure 23 System Setup for Experiments with Model System Note: Temperature controllers for the diluent a nd the reactor can be seen along with the reaction setup assembly, surf ace tensiometer probes an d the dilution system.

PAGE 100

80 Figure 24 Temperature Control System for th e Sample Holder of the Spectrometer Figure 25 Entire Setup for Experiments with Model System

PAGE 101

81 Figure 26 pH Meter to Measure th e pH of the Suspending Medium

PAGE 102

82 CHAPTER 5: RESULTS AND DISCUSSIONS 5.1 Introduction This section reports and di scusses the results obtained from the experiments performed as described in Chapter 4. The simu lation studies described in the Chapter 3 of this dissertation indicate the pr esence of two or more populat ions of the droplets present in the cuvette. The results inferred from the simulation studies led the formulation of the proposed hypothesis on the presen ce of small nano-droplet popul ation of size range 30 to 100nm in diameter in the emulsion, along with the large monomer droplet population of size range 2 to 3 microns. The spectral data of the emul sions prepared at different conditions of temperature, surfactant concen tration and pH of the suspending medium were analyzed to determine the number of populations of the dispersed phase and their size characteristics. This analysis was done by sp ectroscopy. In order to obtain information of the characteristics of the di spersed phase, the spectral signal from the wavelength range, 280 to 820 nm was analyzed. In this chapter, the effects of each of the emulsification condition on the size charac teristics of the particle population are presented and discussed. Before recording the transmission Uv vis spectra of the emulsions for performing the analysis, it was ensured that the emulsion had achieved stability.

PAGE 103

83 This was made possible by pre-determining the time required for the emulsion to achieve stability as describe d in section 4.3.6. The spectral da ta reported in Appendix F was deconvoluted using the turbidity equati on (equation 3.1) as de scribed in section 4.3.1.2. The results are reported in Table 5.1. It ca n be observed from the results that the size distribution characteristic s of the small particle population remain fairly constant before the emulsion was stabilized while th e mean diameter of the large particle population changed significantly. The standard deviation of the larg e particle population was also observed to be constant. The percen tage of the dispersed phase in the small particle population seemed to be changing unt il emulsion attained stability. The weight percent of oil in the disper sed was constant after the st ability of the emulsion was achieved. Table 5.1 Results Obtained by Deconvoluting th e Uv vis Transmission Spectra of the Emulsion Before Stability Time Mean diameter of small Particle Population Std dev Small particle Population Mean diameter of Large particle Population Std dev Large particle Population Wt% of Oil in Small Particles 15 minutes 32 0.1 5202 0.6 54.4 20 minutes 30 0.1 6490 0.5 51.9 45 minutes 31 0.1 5512 0.6 44.1 1 hour 32 0.1 4875 0.5 57.2 1 hour 45 min 32 0.1 5713 0.6 42.8 2 hours 15 minutes 30 0.1 3759 0.6 70.7

PAGE 104

84 Table 5.2 lists the effect of the manipula ted variables on the characteristics of the dispersed phase populations namely, number of populations of the dispersed phase, their size characteristics and the percentage of the disperse d phase contained in each of the population. In Table 5.3 the summary of th e experimental results is reported. Table 5.4 reports the comparison between the results obtained with offlin e and online spectra of the decane-in-PBS buffer emulsion. Results obtained by deconvoluting the transmission spectrum of the styrene-in-wat er emulsion before the beginning of the polymerization reaction are also presented in Table 5.4. The spectrum of styrene was deconvoluted for the wavelength region 200 to 820 nm. Excellent fit was obtained between the estimated and the measured spectra of the styrene in water emulsion (refer Figure 113). The comparison between the expe rimental spectra and the calculated spectra are presented in Figures 27 th rough 33. Figures 36 through 44 elucidate the effect of different emulsifica tion conditions on the mean diam eter, standard deviation of the nano-droplet population along with the wei ght percent of the dispersed phase in them. Change in the size characteristics of the large droplet populat ion as a function of emulsification conditions are shown in Figur es 45 and 46. The continuous lines have been placed to suggest the main trend in the data. For each of the experiments, an analysis was performed on the surfactant distribution over the particle populations of the dispersed phase. The results are tabulated in Table 5.5. Table 5.6 reports the ar ea required to be stabilized per molecule for a particle population of the given size characteri stics to exist. The smaller the area

PAGE 105

85 required for the stabilization of the area per molecule of th e surfactant, higher is the probability of the stability of the particle and hence, higher the probability of its existence. 5.2 Effects of the Manipulated Variables 5.2.1 Effect of Surfactant Concentration The effect of the surfactant to oil ratio on the mean and the standard deviation of the nano-droplet population are shown in Fi gures 39 and 40 respectively for different conditions of pH and temperatur e. The effect of the surfactan t to oil ratio on the wt% of the dispersed oil phase in the nano-droplet popul ation at different conditions of pH and temperature is shown in Figure 41. It was obser ved that the mean diameter of the small particle population was influenced by the su rfactant concentrations The mean diameter of the small particle populations at high surfactant to oil rati o was around 30 nm in diameters whereas for low surfactant to oil ratio, it was around 100 to 110 nm in diameter. The number of partic les of the small particle popul ation was found to be much higher for the emulsion recipe having higher surfactant to oil ratio than the one having lower surfactant to oil ratio. The percentage of the dispersed phase in the nano-droplet population is greatly influenced by the surfactan t to oil ratio. At low surfactant to oil ratios, for different conditions of pH and te mperatures, only 18% of the dispersed phase was present in small particles while for th e emulsion recipes having higher surfactant to oil ration, more than 70 to 80% of the disp ersed phase was presen t in the nano-droplet population. The surfactant to oi l ratio however did not have much effect on the size

PAGE 106

86 characteristics namely the mean diameter a nd the standard deviation of large droplet population (refer Figures 45 and 46). 5.2.2 Effect of Temperature The effect of the temperature on the mean and the standard deviation of the nanodroplet population are shown in Figures 42 and 43 respectively for different conditions of surfactant to oil ratios and pH. The effect of temperature on the wt% of the dispersed oil phase in the nano-droplet populat ion at different conditions of pH and surfactant to oil ratios is shown in Figure 44. The effect of temperature on the mean diameter of the nanodroplet population at both high and low surfact ant to oil ratio for the suspending medium with low pH is negligible. However for high differences in temperature, as shown in Figure 42 for a suspending medium with high pH the mean diameter decreased slightly with the increase in temperatur e but was still within the limits of 95% CI. This indicated that the change in the mean diameter is negligible as a functio n of temperature for suspending medium with high pH. The standard deviation of the sm all particle population was also within the limits of 95% CI for high and low surfactant to oil ratio recipes for low pH of the suspending medium (refer Figure 43). At high pH of the suspending medium a higher estimate of the standard devi ation of the estimate of small particles was observed at low temperatures. The temperature effect on the percenta ge of the dispersed phase in the mass fraction of the small partic le population is almost negligible for the emulsion recipe having high surfactant to o il ratio for both high and low pH of the suspending medium.

PAGE 107

87 5.2.3 Effect of pH Figures 36 and 37 depict the effect of pH on the mean diameter, standard deviation of the nano-droplet population respectiv ely at different conditions of surfactant to oil ratios. Figure 38 depicts the effect of pH on the weight percent of the dispersed phase in the nano-droplet population for diffe rent surfactant to o il ratios. Figures 36 suggests that the mean diameter of the na no-droplet population for the emulsion recipes with low surfactant to oil ratios remains unaff ected at high temperatures by the change in the pH of the suspending medium. Similar observation was made with respect to the effect of pH of the suspendi ng medium on the mean diameter at high temperatures for the emulsion recipe with high surfactant/oil ratio. Standard deviation of the small particle population remained unchanged as a function of pH of the suspending medium for the emulsion recipes with both high and low surf actant to oil ratios, at high temperatures (refer Figure 37). Higher variab ility in the estimated standa rd deviation of the small particle population was observed for th e suspending medium with low pH. The percentage of the dispersed phase in the sma ll particles for both the cases of surfactant to oil ratio in the emulsion recipe seemed to be unaffected by the pH of the suspending medium (refer Figure 38). The mean and the standard deviation of the large droplet population was unaffected by the emulsification cond itions (refer Figures 45 and 46).

PAGE 108

88 The results described above are conven iently tabulated as in Table 5.2. The effects of the manipulated variables viz temp erature, pH and surfactant concentration on the size distribution characteristics namely the mean ( s) and the standard deviation ( ) of the nano-droplet population and on the wei ght percent of the oil contained in them (represented as ‘%’) are summarized in Tabl e 5.2. The effect of an emulsification condition on the characteristics of the na no-droplet population is regarded to be appreciable when the observed changes ar e beyond the 95% CI of the estimates respectively. Due to the skewed design of expe riments, the isolation of a particular effect on the characteristics of the small particle population at given condi tions was not possible in certain cases and hence could not be expressed. The effects of the variables at the speci fied conditions on these characteristics are denoted as: 1. I = Appreciable increase with the increase of the effect; 2. D = Appreciable decrease with the increase of the effect; 3. N = Not appreciable; 4. NPE = Not possible to express

PAGE 109

89 Table 5.2 Effects of the Manipul ated Variables on the Nano-droplet Population of the Dispersed Phase Effect of pH Effect of Temperature Effect of Surfactant Concentration At low Surfactant Conc At high Surfactant Conc At low Surfactant Conc At high Surfactant Conc At low pH At high pH s N s N s NPE s N s D s D N N NPE N N N For High Temp % N For High Temp % N For High pH % NPE For High pH % N For High Temp % I For High Temp % I s NPE s NPE s N s N s D s NPE NPE NPE N N N NPE For Low Temp % NPE For Low Temp % NPE For Low pH % N For Low pH % N For Low Temp % I For Low Temp % NPE

PAGE 110

90 Table 5.3 Summary of E xperimental Results Experiment Conditions Mean diameter of small particle population Standard deviation of small particle population Mean diameter of large particle population Standard deviation of large particle population Wt% of decane in small particle population T=50 C S/O=0.0154 pH=2 109 0.1 3409 0.6 18.0 T=50 C S/O=0.0154 pH=2 109 0.1 3409 0.6 18.0 T=50 C S/O=0.0154 pH=2 108 0.1 3409 0.6 18.0 T= 60 C S/O=0.0154 pH=2 106 0.1 3732 0.6 12.0 T=60 C Surf=0.0154 pH=2 106 0.1 3634 0.5 15.4 T=60 C S/O=0.0154 pH=2 102 0.1 3586 0.6 14.1 T= 60 C Surf=0.0154 pH=10 98 0.1 3613 0.6 16.2 T=60 C S/O=0.0154 pH=10 95 0.1 3626 0.6 19.5 T=60 C S/O=0.0154 pH=10 93 0.1 3615 0.6 15.9 T=50 C S/O=0.046 pH=2 32 0.1 3449 0.6 76.5 T=50 C S/O=0.046 pH=2 32 0.1 3479 0.6 75.6 T=50 C S/O=046 pH=2 32 0.1 3479 0.6 75.6

PAGE 111

91 Table 5.3 (Continued) T=60 C S/O=0.046 pH=2 30 0.1 3773 0.6 79.9 T= 60 C S/O=0.046 pH=2 31 0.1 3779 0.6 77.3 T=60 C S/O=0.046 pH=2 32 0.1 3719 0.6 75.7 T=22 C S/O=0.046 pH=10 31 0.1 3660 0.5 77.2 T=22 C S/O=0.046 pH=10 33 0.1 3573 0.6 69.9 T=22 C S/O=0.046 pH=10 31 0.1 3660 0.5 77.2 T=60 C S/O=0.046 pH=10 28 0.1 3575 0.6 80.6 T=60 C S/O=0.046 pH=10 27 0.1 3544 0.6 80.8 T=60 C S/O=0.046 pH=10 29 0.1 3559 0.5 82.0

PAGE 112

92 Table 5.4 Comparison of the Online and Offlin e Experimental Results along with the Results Obtained for the Styrene Emulsion As mentioned earlier in this section, the transmission spectrum for the styrene in water emulsion was analyzed for the wave length region, 200 to 820nm. Excellent fit between the estimated and measured spect rum indicate that the estimates of the parameters of the droplet size distribution of each of the population comprising the emulsion are reliable. Figure 113 shows the fit between estimated and the measured spectrum. Experiment Name Experiment Conditions Conc is in Surf/oil ratio (% age) Mean diameter of small particle populati on Std deviation of small particle population Mean diameter of large particle population Std deviation of large particle population Wt% of decane in small particle population Online Data T= 22 C Surf/Oil =0.03 pH = 6.89 29 0.13 3759 0.58 70.7 Offline data T= 22 C Surf/Oil =0.03 pH = 6.89 32 0.13 3299 0.57 68.2 Spectrum of the Styrene in water emulsion before the reaction begins T = 60 C Surf/Oil =0.043 PH =7 37 0.10 3246 0.20 36.8

PAGE 113

93 5.3 Surfactant Stabilized Area Analysis Analysis on the surfactant distribution over the part icle populations of the dispersed phase enabled the feasibility studies for the existence of a particular particle population. In doing so, the area required to be stabilized per surf actant molecule was calculated. The following procedure was followed to calculate the area stabilized by the surfactant: It is known from the literature that the area st abilized per molecule of surfactant for large particles is 0.48 nm2 54. From the estimate of the partic le number of the large particle population and from the estimated size distributi on, the surface area of the large particles in the diluted sample present in the cuve tte was calculated. Assuming representative sampling, the estimated concentration of the su rfactant in the cuvett e was calculated from the surfactant/ oil ratio in the recipe. The number of molecules requi red to stabilize the large particle population were calculated by the formula: molecule t surfac one by stabilized Area population particle e l of Area Nles surfmolecutan arg (5.1) Number s Avogadro N MolWt AmSles surfmolecu' (5.2) where AmS is the amount of surfactant for stabilization for large dr oplet population. The remaining amount of surfactant was thought to be available for stabilizing the small particle population. The surf ace area of the small particle population that needs to be stabilized was calculated from the estimate of the particle number of the small particle population and from their estimated size distributi on characteristics. In order to obtain the

PAGE 114

94 area required to be stabilized per molecule of the surfactant for the small particle population, following calculations were performed: Weight of surfactant required to stabilize = Total Wt of surf actant – Weight of surfactant the small particle population utilized to stabilize large populatio n Wt Mol t Surfac particles small stabilize to wt Emulsifier Nmsmsptan (5.3) where msmspN = Number of moles of surfactant required to stabilize small particle population smspN number s Avogadro Nmsmsp' (5.4) where smspN Total number of surfactant molecules require d to stabilize the surface area of small particle population Avogadro’s number = 6.022*1023 Therefore, smsp stabreqN Particles Small of Area Surface Total S (5.5) where stabreqS is the surface area of the small particle population requi red to be stabilized by one molecule of the surfactant The results obtained from the calculations de scribed were performed for all the cases and are expressed in Table 5.5.

PAGE 115

95 Table 5.5 Comparison of the Surfactant Distribution Over Each Population of the Dispersed Phase Experimental conditions Total surface area in nm2 1015 Surface Area of small particles *1015 Surface Area for large particles *1014 Amount of surfactant large particles in grams/ml *10-7 Amount of surfactant on small particles in grams/ml *10-6 Area stabilized per molecule for large particles in nm2 Area stabilized per molecule for small particles in nm2 T =50 deg C pH = 2 Surf /oil =0.0154 7.9 7.5 3.7 3.4 8.6 0.48 0.4 T =60 deg C pH = 2 Surf /oil =0.0154 4.3 4.0 2.3 2.1 5.4 0.48 0.35 T =60 deg C pH = 10 Surf /oil =0.0154 9.3 8.9 3.8 3.5 10 0.48 0.4 T =50 deg C pH = 2 Surf /oil =0.046 344 344 3.3 3.0 82.5 0.48 1.88 T =60 deg C pH = 2 Surf /oil =0.046 449 449 2.94 2.7 91.7 0.48 2.21 T =22 deg C pH = 10 Surf /oil =0.046 435 434 4.02 3.7 100 0.48 1.94 T =60 deg C pH = 10 Surf /oil =0.046 420 419 2.9 2.6 82.5 0.48 2.29

PAGE 116

96 An interesting observation is that, for the emulsion recipe with low surfactant to oil ratio, the area stabilized pe r molecule of the surfactant fo r small particle population is in close agreement with the values reported in literature54,55. This suggests that the diluted sample obtained inside the cuvette is stable and hence, the information drawn from these measurements are reliable and meaningful. C onsistency in the results obtained for the replicate measurements and replicate samples satisfactorily address the issue of sample integrity and stability. Since the sample obtained from emulsion recipe with low surfactant/ oil ratio is stable, the sample obtained from the emulsion recipe with high surfactant/oil ratio can also be inferred to be stable. Calculations were also performed to test the probability of the existence of particle populations with size characteristics of swollen and empty mi celles that could be formed by the residual surfactant after stab ilizing large particle population. Since the weight of the oil phase that is dispersed in small particle populat ion has been estimated by spectroscopy, the amount of area needed to be stabilized by each surfactant molecule to achieve the dispersion droplets of require d size distribution was calculated using the following equations: Number of small particles of the size ra nge of swollen micelles or empty micelles is calculated by equation 5.6 (assuming th at these particles are monodispersed) Density particle one of Vol stabilized be to phase oil of Wt Nsmpa* (5.6)

PAGE 117

97 where Nsmpa is resulting the number of particles of the size char acteristics of the swollen or empty micelles for dispersing the estimated amount of oil Total area of small particles is given by equation 5.7 as smpaN r particles small of area Total * 42 (5.7) Area required to be stabilized per particle by one surfactant molecule can be calculated by equation 5.5. Results shown in Table 5.6 indicate th at a much higher area will have to be stabilized by one surfactant molecule for th e dispersed phase to be of size range of swollen micelles (5 to 10 nm diameter) than for the size range of nano-droplets (30 to 100 nm diameter) to achieve the same amount of oil dispersion. This would suggest that the probability of the presence of the swollen mice lles is much lower than the presence of the nano-droplets for the given quantity of the su rfactant available and the amount of oil required to be dispersed. In other words, th e required quantity of surfactant to disperse the same amount of oil phase into a particle population of size characte ristics of a swollen micelle or empty micelles is much higher than the quantity of surfactant available in the recipe.

PAGE 118

98 Table 5.6 Comparison of the Area Required to be Stabilized per Surfactant Molecule to Achieve a Dispersed Phase of Given Size Characteristics Experiment Area required to be stabilized per molecule for nanodroplets Area required to be stabilized per molecule for 10nm dispersed micelles Area required to be stabilized per molecule for 3nm dispersed micelles Experiment 1 0.4 4.54 15.13 Experiment 2 0.35 3 10.00 Experiment 4 0.4 4.03 13.44 Experiment 5 1.88 6.18 20.6 Experiment 6 2.21 6.47 21.58 Experiment 7 1.94 6.26 20.88 Experiment 8 2.29 6.53 21.77 5.4 Conclusion The above results suggest that the feas ibility of the small population of the dispersed phase having the size characteris tics of the nano-droplets (diameter around 30 nm) is higher than those of th e swollen or empty micelles.

PAGE 119

99 Figure 27 Comparison of the Experimental and Estimated Spectrum for Experiment 1 Figure 28 Comparison of the Experimental and Estimated Spectrum for Experiment 2 100 200 300 400 500 600 700 800 900 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 100 200 300 400 500 600 700 800 900 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Wavelength (nm) Experimental Spectrum Estimated SpectrumO p t i c a l D e n s i t y 100 200 300 400 500 600 700 800 900 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 1.05 Experimental Spectrum Estimated Spectrum Wavelength (nm)O p t i c a l D e n s i t y

PAGE 120

100 Figure 29 Comparison of the Experimental and Estimated Spectrum for Experiment 4 Figure 30 Comparison of the Experimental and Estimated Spectrum for Experiment 5 100 200 300 400 500 600 700 800 900 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 100 200 300 400 500 600 700 800 900 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Wavelength (nm) Experimental Spectrum Estimated SpectrumO p t i c a l D e n s i t y 100 200 300 400 500 600 700 800 900 0.8 1 1.2 1.4 1.6 1.8 2 Wavelength O D 100 200 300 400 500 600 700 800 900 0.8 1 1.2 1.4 1.6 1.8 2 Wavelength (nm) Experimental Spectrum Estimated SpectrumO p t i c a l D e n s i t y

PAGE 121

101 Figure 31 Comparison of the Experimental and Estimated Spectrum for Experiment 6 Figure 32 Comparison of the Experimental and Estimated Spectrum for Experiment 7 100 200 300 400 500 600 700 800 900 0.8 1 1.2 1.4 1.6 1.8 2 2.2 O D Wavelength 100 200 300 400 500 600 700 800 900 0.8 1 1.2 1.4 1.6 1.8 2 2.2 Wavelength(nm) Experimental Spectrum Estimated SpectrumO p t i c a l D e n s i t y 100 200 300 400 500 600 700 800 900 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 OD Wavelength (nm) 100 200 300 400 500 600 700 800 900 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 Wavelength (nm) Experimental Spectrum Estimated SpectrumO p t i c a l D e n s i t y

PAGE 122

102 Figure 33 Comparison of the Experimental and Estimated Spectrum for Experiment 8 100 200 300 400 500 600 700 800 900 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 OD 100 200 300 400 500 600 700 800 900 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 O p t i c a l D e n s i t yWavelength (nm) Experimental Spectrum Estimated Spectrum

PAGE 123

103 Figure 34 Comparison of the Online Experiment al and the Estimated Spectra at Low Temperature Figure 35 Comparison of the Offline Experime ntal and the Estimated Spectra at Low Temperature Optical Density 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength O D 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength(nm)Optical Density Measured Spectrum Estimated SpectrumOptical Density 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength O D 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength(nm) 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength O D 100 200 300 400 500 600 700 800 900 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength(nm)Optical Density Measured Spectrum Estimated Spectrum 100 200 300 400 500 600 700 800 900 0.5 1 1.5 2 2.5 3 3.5 Wavelength 100 200 300 400 500 600 700 800 900 0.5 1 1.5 2 2.5 3 3.5 Wavelength(nm) Optical Density Measured Spectrum Estimated Spectrum

PAGE 124

104 Figure 36 Effect of pH on the Mean Diam eter Dn of Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio in the Emulsion Recipe. 95 % Confidence Intervals are de noted by +. Continuous Lines Suggest Trend Figure 37 Effect of pH on the Standard Deviation of Nano-droplet Population at Different Conditions of Su rfactant to Oil Ratio in Emulsion Recipe. 95% Confidence Intervals ar e denoted by +. Continuous Lines Suggest Trend 1 2 3 4 5 6 7 8 9 10 11 12 20 40 60 80 100 120 140 pHMean Diameter (nm)O S/ O ratio = 0.0154 S/ O ratio = 0.046 1 2 3 4 5 6 7 8 9 10 11 12 0 0.05 0.1 0.15 0.2 0.25 0.3 pH Standard Deviation of nano-droplet populationOS / O ratio = 0.0154 S / O ratio = 0.046

PAGE 125

105 Figure 38 Effect of pH on the Weight Percen t of Oil in the Nano-dr oplet Population at Different Conditions of Surfactant to Oil Ratio in Emulsion Recipe. 95% Confidence Intervals ar e denoted by +. Continuous Lines Suggest Trend Figure 39 Effect of Surfactant to Oil Ratio in Emulsion Recipe on the Mean Diameter Dn of the Nano-droplet Populati on at Different Conditions of Temperature and pH 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend 1 2 3 4 5 6 7 8 9 10 11 12 0 10 20 30 40 50 60 70 80 90 100 pHWeight Percent of Oil in nano-droplet populationS/ O ratio = 0.0154 S/ O ratio = 0.046 O 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 10 20 30 40 50 60 70 80 90 100 110 120 Surfactant to Oil ratioMean Diameter (nm) oTemperature = 50 Deg C pH = 2 Temperature = 60 Deg C pH = 2 Temperature = 60 Deg C pH = 10

PAGE 126

106 Figure 40 Effect of Surfactan t to Oil Ratio in Emulsion Recipe on the Standard Deviation of the Nano-droplet Population at Different Conditions of Temperature and pH. 95% Confidence Interval s are denoted by +. Continuous Lines Suggest Trend Figure 41Effect of Surfactan t to Oil Ratio in Emulsion Recipe on the Weight Percent of Oil in the Nano-droplet Population at Di fferent Conditions of Temperature and pH. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Surfactant to Oil ratioStandard Deviation of nano-droplet population oTemperature = 50 Deg C pH = 2 Temperature = 60 Deg C pH = 2 Temperature = 60 Deg C pH = 10 0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05 10 20 30 40 50 60 70 80 90 Surfactant to Oil ratioWeight Percent of Oil in nano-droplet population oTemperature = 50 Deg C pH = 2 Temperature = 60 Deg C pH = 2 Temperature = 60 Deg C pH = 10

PAGE 127

107 Figure 42 Effect of Temperature on the Mean Diameter (Dn) of the Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend Figure 43 Effect of Temperature on the St andard Deviation of the Nano-droplet Population at Different C onditions of Surfactant to O il ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend 25 30 35 40 45 50 55 60 30 40 50 60 70 80 90 100 110 Temperature in Deg C Mean Diameter (nm) S / O ratio = 0.0154 pH = 2 S / O ratio = 0.046 pH = 2 S / O ratio = 0.046 pH = 10 O 10 20 30 40 50 60 70 80 0 0.05 0.1 0.15 0.2 0.25 0.3 Temperature in Deg CStandard Deviation of nano-droplet population S / O ratio = 0.0154 pH = 2 S / O ratio = 0.046 pH = 2 S/ O ratio = 0.046 pH = 10O

PAGE 128

108 Figure 44 Effect of Temperature on the Weight Percent of Oil in the Nano-droplet Population at Different Conditions of Surfactant to Oil Ratio and pH of the Emulsion Recipe. 95 % Confidence Intervals are denoted by +. Continuous Lines Suggest Trend 20 25 30 35 40 45 50 55 60 65 10 20 30 40 50 60 70 80 90 Temperature in Deg CWeight % of oil in nano-droplet population S/ O ratio = 0.0154 pH = 2 S/ O ratio = 0.046 pH = 2 S / O ratio = 0.046 pH = 10 O

PAGE 129

109 Figure 45 Effect of the Different Emulsificati on Conditions on the Mean Diameter Dn of the Large Droplet Population Figure 46 Effect of the Different Emulsi fication Conditions on the Standard Deviation of the Large Droplet Population 0 20 40 60 80 100 1000 1500 2000 2500 3000 3500 4000 4500 5000 Temperature in Deg CMean Diameter ( nm) Off line On line Expt7 Expt5 Expt1 Expt2 Expt6 Expt2 Expt8 0 20 40 60 80 100 1000 1500 2000 2500 3000 3500 4000 4500 5000 Temperature in Deg CMean Diameter ( nm) Off line On line Expt7 Expt5 Expt1 Expt2 Expt6 Expt2 Expt8 10 20 30 40 50 60 70 80 90 100 0 0.2 0.4 0.6 0.8 1 Temperature in Deg CStandard deviation of large particle population On line Off line Expt7 Expt1 Expt5 Expt2 Expt6 Expt4 Expt8 10 20 30 40 50 60 70 80 90 100 0 0.2 0.4 0.6 0.8 1 Temperature in Deg CStandard deviation of large particle population On line Off line Expt7 Expt1 Expt5 Expt2 Expt6 Expt4 Expt8

PAGE 130

110 CHAPTER 6: CONCLUSION AND RECOMMENDED FUTURE WORK This dissertation project has been focuse d on the identification of the most likely nucleation locus in emulsion polymerization processes. The preceding chapters have underscored the importance of their identif ication for elucidating the nucleation mechanism. Limitations associated with the e xperimental tools available to the past and current researchers in getting the relevant information for identifying the nucleation locus have been highlighted. The Uv vis spectrosc opic techniques coupled with the algorithms developed in-house have been identified as th e experimental tool fo r the characterization of the reaction mixture. These techniques pr ovide the relevant and necessary information of the reaction mixture for iden tifying the most likely nucleation locus. The approach and the thought process using a non -reacting model emulsion system have been delineated. The hypothesis on the most likely nuclea tion locus has been proposed and the experimental efforts undertaken to prove it ha ve been described. This chapter focuses on the conclusions that were based on the obtaine d experimental results and the calculations performed. Contributions as a re sult of this research work have been enumerated along with the recommendations for future work.

PAGE 131

111 6.1 Conclusions The model molecules having similar phys ico-chemical properties (viscosity, surface tension, vapor pressure and density) a nd hence the dispersion characteristics can be successfully used to mimic the actual beha vior of the dispersed phase. This approach enables the characterization of the dispersed phase in terms of the size and composition. The distribution of the emulsion component of interest can be studied using appropriately labeled compound. A nano-droplet population of size of 30 to 100 nm diameter was found to exist in emulsion along with the large particle (m onomer/ oil droplet population) of size characteristic one to a few microns. These nano-droplet populations were found to contain 12 to 80 % of the dispersed o il phase depending upon the emulsification conditions. High interfacial ar ea offered by the nano-droplet s and high content of oil phase in them make them a strong candidate for being the most likely nucleation locus for emulsion polymerization processes. Total number of the nano-droplets forme d, their size distribution characteristics and the weight fraction of the dispersed phase present in the nano-droplet population were found to be primarily influenced by the emulsifier concentration. pH of the suspending medium and the temperature did influence the size and the number characteristics of the nano-dr oplet population significantly.

PAGE 132

112 The size and the number characteristic s of the large droplet population remained significantly unaffected by the em ulsification conditions namely, surfactant concentration, pH of the suspending medi um and temperature. The nano-droplet population has a much narrower distribution (around 0.1) as compared to the large particle population (around 0.6). 6.2 Contributions 1. Identifying the existence of the nano-dr oplet population and characterizing them by their number and size distribution 2. Identification of the nano-droplet population as the main nucleation loci for emulsion polymerization reaction 3. Identifying surfactant to oil ratio as th e process variable that predominantly governs the size, composition and the numbe r characteristics of the nano-droplet population. The contributions enumerated above enable a better understanding of the emulsification process. This study provides us eful insight in addressing the issue of nucleation mechanisms in emulsion polymeri zation process by identifying a nano-droplet population as the likely nucleation loci and ch aracterizing them in terms of their size, composition and number.

PAGE 133

113 6.3 Recommendations for Future Work The results shown in Table 5.3 indicate em ulsifying efficiency per molecule of the surfactant increases with increase in surfactant con centration. Further work is recommended in this direction, since estimati on of the amount of surfactant distributed over each droplet population remains an unresolved issue as yet. Obtaining a good estimate of the optical properties of emulsifier is essential for quantifying the amount of emulsifier on each of the population of the disp ersed phase so that actual area stabilized by each emulsifier molecule on population of the dispersed phase can be determined experimentally. The hypothesis on emulsification efficiency as a function of emulsifier concentration can thus be addressed in fu ture for the complete characterization of emulsion and in depth understandin g of the emulsification process. Nucleation models consideri ng nano-droplets as the nucl eation loci need to be developed and tested. Effect of the size and compositional characteristics of the nanodroplets as the likely reaction loci on the initiation mechan ism needs to be explored. Subsequent changes undergone by the nano-drop lets (main reaction lo ci) as the reaction progresses and its effect on the particle gr owth and propagation reaction needs to be studied.

PAGE 134

114 REFERENCES 1. Gilbert Robert. Emulsion Polymeri zation; A Mechanistic Approach; Academic Press (1995) 2. Napper Donald; Gilbert Robert. Comprehensive Polymer Science Allen and Bevington, The Synthesis and Characteriza tion of Reactions and A pplications of Polymer;volume 4, Chain Polymerization Part II, (1989) 3. M.S. El Aaser. Emulsion Polymerizati on, Scientific Methods for the Study of Polymer Colloids and their Applicati ons; Kluwer Academic Publishers, (1990) 4. URL http://www.zirkle.com/servi ces/reports/chpt12rubber.pdf 5. URL http://www.klinegroup.com/bro chures/y359/brochure.pdf 6. URL http://www.polymerlatex.com/index.asp ?ln=2&tt=1&c=1977&ny=2&nt=3&nr=23 7. URL http://www.chemicals-yorkshire.com/assets/Polymers%20&%20Resins.pdf 8. URL http://www.mindbranch.com/page/ catalog/product/2e6a73703f636f64653d523135342d3 63030.html 9. Herrera Ordonez, Jorge; Olayo, Robert o. On the Kinetics Of Styrene Emulsion Polymerization above CMC .I. A Mathematical Model ;, Journal of Polymer Science : Part A: Polymer Chemistry; v38; 2201 2218 (2000) 10. Herrera Ordonez, Jorge; Olayo, Robert o. On the Kinetics Of Styrene Emulsion Polymerization above CMC .II. A Comp arison with Experimental Results; Journal of Polymer Science: Part A: Polymer Chemistry; v38; 2219-2231 (2000) 11. Peter Lovell.Free Radical Polymerizati on; Emulsion Polymerization and Emulsion Polymers; Chapter 1; Peter Lovell; Edited by P.A Lovell and M.S. El -Aaser; John Wiley and Sons Ltd (1997)

PAGE 135

115 12. Alexander Dunn. Harkins, Smith Ewart rela ted theories; Emulsion Polymerization and Emulsion Polymers; Chapter 4, Edited by P.A Lovell and M.S. El -Aaser; John Wiley and Sons Ltd (1997) 13. Feeney, P.J; Napper, Donald. H; Gilb ert, R.G. Coagulative Nucleation and Particle Size Distribution in Emul sion Polymerization; Macromolecules,17, 25202529 (1984) 14. Gardon, J. L. Emulsion Polymerization I; Recalculation and Extension of SmithEwart Theory; Journal of Polymer Science; Part A-1; volume 6; 623644 (1968) 15. Feeney P.J; Napper, Donald.H; Gilbert, R.G. The Determinant of Latex monodispersity in Emulsion Polymerizat ion, Journal Of Colloid and Interface Science, v118, No 2, (1987) 16. Gilbert, R. G. Particle Size Distributi ons and Molar Mass Dist ributions; Emulsion Polymerization and Emulsion Polymers ; Chapter 5, Modeling Rates, Edited by P.A Lovell and M.S. El -Aas er; John Wiley and Sons Ltd (1997) 17. Herrera Ordonez, Jorge; Olayo, R oberto. Methyl Methacrylate Emulsion Polymerization at Low Monomer Concen tration: Kinetic Modeling Of Nucleation, Particle Size Distributi on and Rate Of Polymerization. Journal Of Polymer Science: Part A : Polymer Chemistry; v39, 25472556 (2001) 18. Chern, C.S; Lin C.H. Particle Nucleati on Loci in Emulsion Polymerization of Methyl Methacrylate; Polymer ,v41, 44734481 (2000) 19. M.S. El Aaser and E. David Sudol. Features of Emulsion Polymerization Emulsion Polymerization and Emulsion Polymers; Chapter 2; Edited by P.A Lovell and M.S. El -Aaser; John Wiley and Sons Ltd (1997) 20. M.S. El Aaser and E. David Sudol. Miniemulsion Polymerization Emulsion Polymerization and Emulsion Polymers Chapter 20; Edited by P.A Lovell and M.S. El -Aaser; John Wiley and Sons Ltd (1997) 21. Jaime Vara, Masters of Science Thesis, Univ ersity of South Florida, Department of Chemical Engineering, (2000) 22. GarciaRubio, L.H. The Effect of Mo lecular Size on the Absorption Spectra of Macromolecules, Macromolecules, 20, 3070-3075 (1987) 23. Bacon C.P, Ph.D Dissertation, University of South Florida, Department of Chemical Engineering, (1999)

PAGE 136

116 24. Stiemle. E. Ph.D Dissertation, University of South Florida, Department of Chemical Engineering, (1999) 25. Cardenas A. Ph.D Dissertation, Universi ty of South Florida, Department of Chemical Engineering, (2001) 26. Fisher S Masters of Science Thesis, Univer sity of South Florida, Department of Chemical Engineering, 1998 27. Thennadil S. Ph.D Dissertation, Universi ty of South Florida, Department of Chemical Engineering, (2001) 28. Mehta, J; Garcia-Rubio, L.H. Initi ation Reaction and the Modeling of Polymerization Kinetics ACS Sym posium Series, Edited by T Provder, v 313, 202218 (1986) 29. Marathe S. Masters of Science Thesis, Univ ersity of South Florida, Department of Chemical Engineering, (1987) 30. Imeokoparia, D. Kinetic Study and Charact erization of branch ing in Benzoyl Peroxide initiated Solution Polyme rization of Vinyl Acet ate Ph.D Dissertation Department of Chemical Engineer ing University of South Florida, Tampa USA, (1991) 31. Shetty S, Ph.D Dissertation, University of South Florida, Department of Chemical Engineering 32. Vinnik P, Masters of Science Thesis, Univer sity of South Florida, Department of Chemical Engineering, (1997) 33. Paul Sacoto, Masters of Science Thesis, Univ ersity of South Florida, Department of Chemical Engineering, (1999) 34. Elicabe ,G; Garcia Rubio, L.H Advan ces in Chemistry Se ries; v227; ACS Washington DC, (1990) 35. Brandolin A; Garcia R ubio L.H; Provder T; Koehle r M. E and Kuo C, ACS Symposium Series; No 472; Chapter 2; (1991) 36. Maria Celis, Ph.D Dissertation, University of South Florida, Department of Chemical Engineering, (2000) 37. Kerker, M. The Scattering of Light and Other Electromagnetic Radiation, Pergamon Press, New York (1969)

PAGE 137

117 38. Alupoei C, Masters of Science Thesis; Department of Chemical Engineering, Univeristy of South Florida, (2001) 39. Philip Sherman Emulsion Science, Chapter 1 Principles of Emulsion formation (1968) 40. Library of physico-chem property data ; Handbook of Transport Property data; Viscosity, Thermal Conductivity and diffusion coefficients in liquids and gases, Carl L. Yaws; Gulf publis hing company, Houst on, Texas,Pg 45,49 (1995) 41. John A. Dean. Handbook of Organic Chemistry, McGraw Hill Book Company, Pg 1-136,1-163,1-344 (1987) 42. URL http://ntpdb.niehs.nih.gov/NTP_Reports/NTP_Che m_HS_HTML/NTP_Chem1/Radian100-425.html 43. URL http://surface-tension.de/index.html 44. URL http://ptcl.chem.ox.ac.uk/MSDS/ST/Styrene.html 45. URL http://physchem.ox.ac.uk/MSDS/BU/butyl_methacrylate_monomer.html 46. URL http://www.jtbaker.com/MSDS/englishhtml/d0136.htm 47. URL http://www.cheric.org/kdb/ kdb/hcprop/showprop.php?cmpid=10 48. http://www.intox.org/databank/documents/chemical/methymet/cie242.htm 49. URL http://www.online-tensiometer/einsatz/f10_2.html 50. Private Communications, Vine et Shastry, Scott MacDonald, ACD Laboratories 51. URL http://www.chem.wisc.edu/areas/reich/pkatable/index.htm 52. Cardenas; Shastry and Garcia-Rubio.In Si tu Spectroscopy of Monomer and Polymer Synthesis, Chapter 6, Spectrosc opic Techniques for Continuous Monitoring of Emulsion Polymerization Reacti on, pg 83-108, Kluwer Academic/ Plenum Publishers (2002) 53. URL http://www.protein-solutions.c om/psi_books/light_scattering/dynamic/ are_o_and_d_dc_temperature_dependent_.htm 54. Davies, J T. and Rideal, E. K. Interfaci al Phenomena; Academic Press, Pg 232,233 (1961)

PAGE 138

118 55. Drew, Meyers. Surfaces, Interfaces and Collo ids: Principles and Applications; VCH Publishers; Pg 227 (1991) 56. Bohren, Craig. F; Huffman; Donald R Absorption and Scattering of Light by Small Particles, Pg 29, 80, 81, 135; John Wiley and Sons Inc, (1983) 57. Brooks, B.T; Boord, C. E. The Chemistr y of Petroleum Hydrocarbons Reinhold Publishing Corporation, New York (1983) 58. Reid, R.C; Prausnitz, J.M; Sherwood T.K. The Properties of Gases and Liquids McGraw-Hill Inc, 3rd Edition (1977) 59. Emis, C.A; Oosterhoff, L.J; Gonda de Vries. Proc. Roy. Soc., A 297, 54-65 London (1965) 60. URL http://www.iapws.org/relguide/rindex.pdf 61. URL http://www.earthwardconsulting. com/density-calculator.htm 62. Private Communications, Vineet Sh astry, Roy, Singh, Binju Kulkarni Rhodia Chemicals (I) Ltd, Roha India

PAGE 139

119 APPENDICES

PAGE 140

Appendix A: Size Dependent Scattering and Absorption Particle Characteristics In this Appendix we elaborate upon the size dependence of the scattering and the absorption components of the transmission Uv vis spectrum of a pa rticle population as mentioned in Chapter 3. The li ght scattering efficiency and th e absorption efficiency is dependent upon the size of the particle. He nce the shape of the transmission Uv vis spectrum for a particle population of same co mposition but different size characteristics is expected to be different. I0 I | l | Figure 47 Schematic of the Transmission Measurement The change in the intensity of light through in finitesimally small distance dl is given by56 dl I dIextinction (A-1) where I is the intensity of the light dI is the change in the intensity of the light extinction is the extinction coefficient I0 is the intensity of the incident light I is the intensity of the transmitted light l is the path length 120

PAGE 141

Appendix A: (Continued) l extinction I Idl I dI00 l I Iextinction 0ln lextinctione I I 0 (A-2) extinction extinctionQ D24 (for single particle) Thus for a monodisperse particulate system23 ) 4 exp(2 0l Q D N I Iextinction p (A-3) Np = number of particles per unit volume Qextinction = Extinction efficiency For a polydisperse system 23,56 ) ) ( 4 exp(2 0dD D f D Q l N I Iextinction p (A-4) ) ) ( ( 0 le I I Where dD D f D Q l Nextinction p) ( 4 ) (2 (A-5) ( ) is also called optical density. Optical density is expressed as l c ODextinction 121

PAGE 142

Appendix A: (Continued) where c is the concentration of the dispersion c= Np ( number of particles per ml) For monodisperse system23,56 l N Q D l cp extinction extinction 24 For polydisperse system OD dD D f D Q N l D l cextinction p extinction ) ( 42 2 (A-6) where OD is the optical density Qextinction is a function of 1. Complex refractive index denoted by m 2. Size parameter denoted by x Qextinction56 is expressed as Qextinction = Qabsorbance+ Qabsorbance Qabsorbance and Qabsorbance are expressed as by Bohren and Huffman56 3 2 38 27 2 1 15 1 2 1 Im 42 2 4 2 2 2 2 2m m m m m x m m x Qabsorbance (A-7) 2 2 2 42 1 3 8 m m x Qscattering (A-8) 122

PAGE 143

Appendix A: (Continued) For the small particles of size ranges much le ss than the wavelength of the incident light (Uv vis light), the Rayleigh scattering is th e most relevant approximation. The absorption and the scattering efficiencies for Rayleigh approximations56 is given by equations A-9. 2 1 Im 3 4 1 2 1 Im 42 2 3 2 2m m x m m x Qabsorbance (A-9) From the equations A-7, A-8 and A-9 it is clear that as the particle size gets smaller, its scattering efficiency decreases much more than the absorption efficiency. Hence as the particle size becomes smaller, the absorption compone nt is expected to become more dominant while for larger particle s the scattering component is expected to be more dominant in the optical density observed. 123

PAGE 144

124 Appendix B: Spectroscopic Calculations B.1 Introduction This Appendix describes the preliminary calculations performed using simulated transmission Uv vis spectra th at lead to the postulation of the main hypothesis in Chapter 3 of this dissertation. The cal culations were performed in order to explore the possible characteristics of the Styrene in water em ulsion at time zero to understand the initial conditions prevalent in the reactor prior to the beginning of the emulsion polymerization reaction. The simulation algorithms34,35 developed in-house used the turbidity equation (equation 3.1) for simulating the transmissi on Uv vis spectra of the styrene droplet populations having different si ze distribution characteristi cs. In this Appendix are described the calculations performed with the simulated spectra, the pertinent explanations and the inferences drawn from these studies that l ead to the hypothesis postulation in Chapter 3. The transmission Uv vis spectra were gene rated for a wide range of mean sizes of nano-droplet populations of Styrene rangi ng from 30 nm to 60 nm. The generated simulated transmission Uv vis spectra for a particle population of given size distribution and known optical properties were such that the maximum optical density was one. The corresponding concentration of the particles in the populat ion was noted. Extinctions were obtained by dividing the spectra obtained by the respective concentrations. For large monomer droplets the extinctions we re obtained in the same manner.

PAGE 145

125 Appendix B: (Continued) The spectroscopic mass balance was perfor med with the help of the extinctions of the small nano-droplets a nd large monomer droplets. Th e extinctions of the nanodroplets and the large monomer droplets we re multiplied by appropriate fractions and then added together to obtain an extinction of the entire latex mixture of small and large particles. The shape of the extinction spectrum thus obtained is and compared with the shape of the extinction of experimentally observed spectrum of the latex mixture. The results are depicted in Figures 50 through 54. The above explanation on spectroscopic mass balance can be expressed as Extotal = xsmall*Exsmall + xlarge*Exlarge Where xsmall is the mass fraction of the small particles in the latex mixture xlarge is the mass fraction of the large particles in the latex mixture Extotal = total extinction Exsmall = extinction of small particles Exlarge = extinction of large particles For the particle populations of same compos ition, their transmission spectrum would be a function of their size distribution characteris tics alone. Therefore the relative spectral contribution of each population to the measur ed spectrum will be indicative of the respective concentrations of each populations. It was observe d that at least 40 to 50% of the monomer is present in the small nano-dr oplets of the monomer ranging from sizes 30

PAGE 146

126 Appendix B: (Continued) nm to 60 nm. Since the nano-droplets are large in numbers and offer a larger surface area there is a strong possibility of they are th e primary loci for nucleation. Details of the calculations performed are pres ented in the next section. B.2 Sample Calculations Spectroscopic Mass Balance: Vector name for experimental extinction: Exexp.txt Vector name for extinctions for big particles: Exbp.txt Vector name for extinctions for small par ticles of size n nanometers: Exn.txt Sample calculations: Vector for extinctions for the small particles of size 30 nm is given by: Ex30 = n30spc(:,2)/ 0.0000060536; Where n30spc(:,2) is the vector of the optical density from the simulated spectra 0.0000060536 is the concentration of th e monomer in the particles. Vector for extinctions for the big pa rticles (monomer droplets) of size 3.2 m is given by: Exbp = n1spbp (:,2)/ 0.00012865; Where n1spbp(:,2) is the vector of th e optical density from the si mulated spectra obtained from the program 0.00012865 is the concentration of the monomer in the particles. Mass balance:

PAGE 147

127 Appendix B: (Continued) Vector for extinctions for the latex mixture with small partic les of size 30 nm that best fits with the experiment al spectrum is given by C30 = 0.4*ex30(:,1) + 0.6 *exbp(:,1) thus suggesting that 40% of total monomer is in the nano-droplet popu lation of mean size 30 nm diameter. The following figure compares the spectral features of the scattering component of the experimentally observed spectrum with the simulated spectrum of latex of PSD 3.2 microns as the mean diameter and 0.2 as the standard deviation. Figure 48 Comparison for Extinction for La rge Particles (Monomer Droplets Mean Size 3.2 microns Std Dev 0.2) with the Features of the Scattering Component of the Experimentally Observed Spectrum 200 300 400 500 600 700 800 900 4000 4500 5000 5500 6000 6500 7000 7500 Wavelength (nm)E x t i nc t i on Calculated Spectrum Estimated Spectrum Experimental Spectrum

PAGE 148

128 Appendix B: (Continued) The similarity in the spectral featur es of the scattering component of the experimental spectrum and the calculated tr ansmission spectrum of the same latex with size characteristics 3.2 microns mean diameter and sta ndard deviation of 0.2 can be noted. This indicates that th e large monomer droplet populati on present inside the reactor at the beginning of the reaction has si milar size distribution characteristics. Figure 49 Comparison of Simulated Transmissi on Spectra for Particle Populations with Different Mean Sizes but the Constant Standard Deviation 0.2 From Figure 49 it is clear how the shape of the transmission spectrum of the latex changes with the size characte ristics of the latex. Figures 50 through 54 show simulated extinctions of the large pa rticle population added in proportion to the simulated extinctions of the small partic le population such that they are in accordance with the 200 300 400 500 600 700 800 900 0 0.5 1 1.5 2 2.5 x 10 5 30 nm 40nm 45 nm 50 nm 60 nm Wavelength (nm) Extinction

PAGE 149

129 Appendix B: (Continued) extinction of the latex obtained experime ntally. The Figures 50 through 54 thus demonstrate the results of the spectral mass balances performed. The results from the spectral mass balances thus performed indicate that about 40 to 50 % of the monomer is present within the nano-droplet population of mean size 30 to 60 nm in diameter. Figure 50 Spectral Mass Balance for 30 nm and 3.2 microns Mean Diameter Population E X T I N C T I O N Wavelength (nm) 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 Conc of monomer in small particles 40% experimental calculated Extinction

PAGE 150

130 Appendix B: (Continued) Figure 51 Spectral Mass Balance for 40 nm and 3.2 microns Mean Diameter Population Figure 52 Spectral Mass Balance for 45 nm and 3.2 microns Mean Diameter Population Wavelength (nm) Wavelength (nm) 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 4 x 10 Conc of monomer in small particles =40% experimental calculated 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 Concof monomer in small particles =41.38% experimental calculated Extinction Extinction

PAGE 151

131 Appendix B: (Continued) Wavelength (nm) Figure 53 Spectral Mass Balance for 50 nm and 3.2 microns Mean Diameter Population Wavelength (nm) Figure 54 Spectral Mass Balance for 60 nm and 3.2 microns Mean Diameter Population E X T I N C T I O N 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 Conc of monomer in small p articles =41.38% experimental calculated 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 18 x 10 4 Conc of monomer in small particles =46.15% experimental calculated Extinction Extinction

PAGE 152

132 Appendix B: (Continued) Figure 55 Particle Size Distributi ons with Different Mean Sizes B.3 Inferences Drawn from the Results Obta ined from the Spectral Manipulations From the spectral manipulations performed as described above, we conclude that the transmission spectral signal obtained at time zero can be mainly attributed to two particulate populations namely: 1. large monomer droplets ( mean size 3.2 mi crons and standard deviation 0.2) 2. nano-droplets (mean size 30nm to 100nm in diameter and standa rd deviation of about 0.2) The large monomer droplets contained ar ound 50 to 60 percent of the monomer and the number of large particles was in the tenth order of magnitude. 0 50 100 150 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 30 nm 35 nm 40 nm 45 nm 50 nm 60 nm D f(D)

PAGE 153

133 Appendix B: (Continued) The amount of monomer contained in the small nano-droplets was almost 40 to 50%. The number of nano-droplets in the reactor was ca lculated to be in th e order of magnitude of about sixteen. This would be indicative of the high interfacial area that is offered by these nano-droplets as opposed to any other pa rticulate entity presen t inside the reaction mixture in the reactor at time zero. Also the relative concentrat ion of the monomer content in them makes the nano-droplet population (of size range 30nm to 60nm in diameter) the prime candidate for being the most likely nucleation loci. Final number of the particles formed and the number of initial nano-droplets present in the reactor before the beginning of the reaction are of the same order of magnitude. This suggests that if nano-droplet s are the main locus of nucleation then, almost all of the nano-droplets get converted to the polymer particles. Thus governing the characteristics of the nano-droplets by investigating the effect of initial conditions on them is a study of practical and industrial importance.

PAGE 154

134Appendix C: Optical Properties C.1 Introduction This appendix describes the work undert aken to obtain the op tical properties for the oil phase, continuous phase and the emulsifi er. The optical properties are the inputs to the algorithms based on the turbidity e quation (equation 3.1) for characterizing the dispersed phase of the emulsion in terms of the number of populati ons of the dispersed phase, and the number of particles, size dist ribution and the composition of each of the dispersed phase population. The real and the imaginary parts of th e complex refractive index known as the refractive index and absorption coefficients respectively constitute the optical properties of a compound. Optical properties of the di spersed phase, continuous phase and the emulsifier are desired in order to interpret the Uv vis spectroscopy data of the emulsions at different emulsi fication conditions. The complex refractive index is expressed as56 ) ( ) ( ) ( ) (0 0 0 0 0 n ik n m C-1 n( 0) is the real part of the refrac tive index of the particles dispersed k( 0) is the absorption coefficient of the particles dispersed no( 0) is the real part of the refrac tive index of the suspending medium

PAGE 155

135 Appendix C: (Continued) The droplets dispersed in a liquid-liquid emulsion systems are primarily composed of dispersed phase and the emulsi fier. These dispersed droplets are suspended in the continuous medium. The optical properties for all th e components of the emulsion (dispersed phase, continuous phase and the em ulsifier), need to be determined as a function of wavelength and for the emulsifi cation conditions of interest. Since the scattering and absorption components are char acterized by the optical properties, it is essential to have good estimates of the optic al properties of to interpret the Uv vis spectroscopy data using the light scattering and absorp tion theories. The Uv vis spectroscopy data is interpre ted in terms of particle num ber, particle composition and particle size distribution of the particulate popula tions that comprise the emulsion at different emulsification conditi ons using the turbidity equa tion based on Mie scattering theory. The optical properties of all the co mpounds constituting the emulsion namely, decane (the dispersed oil phase), water (the continuous phase) and the emulsifier Sodium Dodecyl Benzene Sulfonate (referred to as SDBS henceforth) were obtained along with the ParaToluene Sulfonic Acid (referred to as PTSA henceforth) as a model molecule for the SDBS. The next section describes in detail the procedure and the rationale followed to obtain the optical pr operties of the above compounds.

PAGE 156

136 Appendix C: (Continued) C.2 Characterization of the Optical Properties C.2.1 Optical Properties for Decane (Dispersed Phase) The Sellimeir-Drude57 equation was used to calcul ate the refractive index of decane. The Sellimeir-Drude equation to calcu late the refractive index (n) for decane at temperature T is given as: 2 2 0 2 ) () ( ) ( ) ( 1 c V T B nT C-2 where B is a constant for a particul ar substance and is expressed as ) ( ) ( T K T B C-3 Where ‘K’ is a constant for a particular substance and ‘ (T)’ is the density of the hydrocarbon at temperature ‘T’. ‘Vo’ is the frequency of th e electrons in the effects of dispersions. ‘ ’ is the wavelength and ‘c ’ is the speed of light. Value of c= 3*1017 (nanometers/s) was a vector from 190 to 820 nm (with an increm ent of 2 nm ) The values of K and Vo obtained using Cauchy equations and with the help of the values of the refractive index as a f unction of wavelength present in the literature were as follows

PAGE 157

137Appendix C: (Continued) K= 1.16978*1031 Vo = 2.9738*1015 ) ( 10 7 420 14 T C-4 = T 20 C-5 20 = Density of decane at 20 deg C T = Density of decane at T Deg C In order to carry out the above calculations, we need an estimate of densties at different temperatures. We calculated the densities by obtaining the molar volume of the decane at different temperatures usi ng the Gunn Yamada equation58 ) ( ) ( ) (R RT f V T f T V C-6 Where V(T) is the molar volume of decane hydrocarbon at temperature T Where VR is the molar volume of the hydrocarbon at the reference temperature 20 deg C. TR is the reduced temperature evaluated as58 TR = T/Tc where Tc is the critical temperature At temperature T, ) 1 ( ) (2 1H H T f at temperature T C-7 4 3 2 1* 11422 1 02512 2 51941 1 33953 0 33593 0R R R RT T T T H C-8 2 2* 04842 0 09045 0 29607 0R RT T H C-9

PAGE 158

138Appendix C: (Continued) Thus at the reference temperature 20 deg C, ) 1 ( ) (2 1R R RH H T f C-10 4 3 2 1* 11422 1 02512 2 51941 1 33953 0 33593 0R r R r R r R r rT T T T H C-11 2 2* 04842 0 09045 0 29607 0R r R r rT T H C-12 is the acentric factor (a function of pressure). The values for the accentric factor ( ) can be calculated usi ng LeeKesler equation59 of state given as 6 1 6 1) ( 43577 0 ) ( 1 4721 13 ) ( 56875 1 2518 15 ) ( 169347 0 ) log( 288621 1 ) ( 09648 6 92714 5 log og Pc C-13 Tr = T/Tc where Tc is the critical temperature Tc for decane = 617.5 deg K and the value of molar volume of decane at 20 deg C ie VR = 0.00513 The density of decane at temperature T is obtained from its molar volume at that temperature from the following relationship ) ( ) ( T V wt Mol T Density C-14 Densities of decane at different temperatures from equation C-14 22 = 0.7284 50 = 0.706 60 = 0.698

PAGE 159

139Appendix C: (Continued) therefore values of from equation C4 at different temperatures 22 = 7.52*1011 50 = 1.1269*1013 60 = 1.504*1013 To calculate the refractive indexes at different temperatures as a f unction of wavelength, We substitute the values of K, Vo and the appropriate values of and for different temperatures in equation 2. The absorption coefficient at each wavelength is zero. C.2.2 Optical Properties for Water (Continuous Phase) The formulation used to calculate the re fractive index of wate r as a function of wavelength for different temperatures was or iginally developed by P. Scheibener et al60 In this exercise we use the modified version of the formulation as released by the International Association for the Properties of Water and Steam60. The refractive index of water as a function of wavelength and temp erature is expressed as equation C-15 2 12 2 n n 1 2 7 2 2 6 2 2 5 2 4 2 3 2 1 0* * * a a a a T a T a a aIR uv C-15 The values of the coefficients as present in the release are as follows a0 = 0.244257733 a4 = 1.5892057*10-3 a1 = 9.74634476*10-3 a5 = 2.45934259*10-3 a2 = -3.73234996*10-3 a6 = 0.900704920

PAGE 160

140Appendix C: (Continued) a3 = 2.68678472*10-4 a7 = -1.66626219*10-2 uv = 0.229202 IR = 5.432937 The values of T and are given as = specific gravity of water at that temperature in grams/ cubic centimeter T = T /TR where TR = 273.15 deg K T is absolute temperature of water = / 589 where wavelength in nanometers In order to obtain the density of water at different temperatures the following correlation was used60,61 ) 12963 68 ( 2 508929 ) 9865 3 ( ) 9414 288 ( 12T T T C-16 where T is the temperature of water in Deg C The value of obtained at different temperatures from equation C-16 can be substituted in equation C-15 to obtain the refractive index of water at different temperatures as a function of wavelength. Th e absorption coefficient at each wavelength is zero.

PAGE 161

141 Appendix C: (Continued) C.2.3 Optical Properties for the Emulsifier Sodium Dodecyl Benzene Sulfonate (SDBS) Unlike the decane and water, the SDBS s hows strong absorbtion peaks. Therefore in order to obtain a complete set of op tical properties for SDBS, the absorption coefficient at each wavelength needs to be estimated along with the estimation of the refractive index. C.2.3.1 Callibration of the Refractometer The refractometer (Bausch and Lomb, A bbe-3L, Bench Model) was calibrated using standards of known refractive inde x. The following relationship between the observed and the actual refr active index was obtained RIobserved = RIactual*0.9956 + 0.0019 C-17 C.2.3.2 Estimation of Refractive Index for SDBS Five solutions of SDBS in DI water with different concentrations were prepared and the refractive index was observed for each of them. The observed values of the refractive index were converted to the actual values of the refrac tive index by using the above calibration equation. The values reported are as follows

PAGE 162

142 Appendix C: (Continued) Table C.1 Refractive Index of SDBS Solution at Different Concentrations Concentration of Sodium Dodecyl Benzene Sulfonate on g/cc Actual refractive index (ns) 0.25 1.3731 0.2375 1.3706 0.2256 1.3645 0.2143 1.3661 0.2036 1.364 A graph of ns Vs C was plotted so as to obtain a stra ight line. The values of the slope and the intercept of the graph of ns Vs C was noted. The actual RI for SDBS was obtai ned by the following relationship nSDBS = slope* density of SDBS+ RIDI water the value of the refractive i ndex for SDBS was obtained as 1.5256. This value of the RI was in very close agreement with the refrac tive index of the sodium salts of the linear alkyl benzene sulfonic acid whic h was obtained to be as 1.511462.

PAGE 163

143 Appendix C: (Continued) C.2.3.3 Estimation of Absorption Coefficients for SDBS Since SDBS shows strong absorption peaks we need to estimate the absorption coefficients to completely define its optical properties. The absorption coefficients are obtained by obtaining the extinc tion coefficients of SDBS. The standard procedure of obtaining the extinctions coefficients would be by acquiring a series of absorption spectra of the SDBS at different known concentrations from the Beer Lamberts law56 A( ) = ( )*c*l C-18 where 1. A( ) is the absorption coeffici ent as function of wavelength 2. ( ) is the extinction coefficien t as afunction of wavelength 3. c is the concentration in g/cc 4. l is the path length (1 centimeter) The values of extinction are calculated at di fferent concentrations to remove any error. The absorption coefficients can then be obtained from the extinction data. C.2.3.4 Use of the Model Molecule for Estima tion of the SDBS Optical Properties Due to the long chain attached to the para position relativ e to the sulfonated carbon of benzene ring, SDBS has a strong micelle forming tendencies even at low concentartions (CMC of SDBS = 8*10-3 M). These micelles will hence tend to scatter

PAGE 164

144Appendix C: (Continued) light thus affecting the absorp tion signal of the SDBS in the smaller wavelength region. These would give us erroneous values of the extinctions and hence the absorption coefficients of the SDBS. The error in the optical properties of the SDBS could cause significant differences in the estimated of particulate properties from the Uv vis transmission spectroscopic signal and the actu al particulate properties of the emulsion. This necessitates the use of a model molecule that would give the same Uv vis signal as the SDBS, had SDBS not scattered light in the lower wavelength region. Since the micelle forming tendency is primarily attr ibuted to the long al kyl (dodecyl) chain attached to the benzene ring, considered a nother organic molecule having the same structure except the long chain was consider ed. This molecule was the para-toluene sulfonic acid (PTSA). PTSA has the smallest hydrocarbon (methyl group) attached at the para position of the carbon in the benzen e ring to which the sulfonic acid group is attached. Also it showed similar Uv vis sp ectral features as the SDBS. The micelleforming tendency is considerably low due to the absence of the long hydrocarbon chain. Thus the absorption signal obtained in the sm aller wavelength region is considerably free from the scattering signal that would interfer e the absorption signal due to the formation of micelles. We hence use PTSA as the mode l molecule to estimate the optical properties (absorption coefficients) for SDBS.

PAGE 165

145 Appendix C: (Continued) C.2.3.5 Obtaining the Optical Properties of PTSA Absorption spectrum of dilute PTSA solu tions were obtained at five different concentrations. The concentrations were note d. The optical density at 260 nm was plotted as a function of concentration of PTSA to ensu re that we are in the concentration range of the PTSA such that the absorption at 260nm varied linearly with the concentration as predicted by the Beer Lamberts law56. Table C.2 Concentration of PTSA Solution Filename Concentration in grams/ml PTSAW3.txt 9.263*10-5 PTSAW4.txt 7.7198*10-5 PTSAW5.txt 6.4335*10-5 PTSAW6.txt 5.361*10-5 PTSAW7.txt 4.4675*10-5 We noticed that the Beer-Lamberts law wa s being followed from the wavelength range 232nm to 820nm.. But from 190 to 230 nm wave length region, the relationship between the optical density and the concentration wa s not linear indicating the Beer-Lamberts law was not being followed. Hence in order to obtain a better estimate of the extinctions for the smaller wavelength regions, we diluted the solutions even further.

PAGE 166

146 Appendix C: (Continued) Table C.3 Concentration of PTSA Solutions (further diluted) File name Concentration in g/ml PTSAW6.txt 5.361*10-5 PTSAW7.txt 4.4675*10-5 PTSAW8.txt 3.722*10-5 PTSAW9.txt 3.102*10-5 PTSAW10.txt 2.585*10-5 The extinctions for each wavelength range we re estimated using the software developed inhouse. The complete extinction file for PTSA for the entire wavelength region was then obtained by joining together the appropr iate values of extinctions for smaller wavelength region an (below 232 nm ) and higher wavelength region (from 232 to 820 nm). The PTSA is highly hygroscopic. It has a great affinity for water and as such captures moisture from the atmosphere. This co uld lead to having errors (overestimation) in the measured weight of PTSA and hence the concentration. Thes e errors are however taken care of by the algorithms developed in house by weighting the er ror as a function of concentration and hence obtaining actual con centration of PTSA. The file containing the extinction coefficients as a function of wave length is required to obtain the absorption

PAGE 167

147 Appendix C: (Continued) coefficients. The corresponding files for th e extinction of SDBS were obtained by calculating the extinction using the mol ecular weight of SDBS separately. The refractive index of PTSA was obta ined in a manner similar for SDBS by measuring the refractive index of the PTSA solution at different concentrations. However as mentioned earlier PTSA being hygr oscopic, there could be an overestimation of the PTSA concentration. The actual concen trations of PTSA solutions were obtained ny plotting the difference of the actual RI a nd the reference RI of the PTSA Vs the measured concentration of the PTSA. The inte rcept on the x axis gave an estimate of the error in concentration of the PTSA. From this estimate of the error in concentration of PTSA solution, the actual concentration of the PTSA solutions were obtained. The data final concentration of the PTSA in solution Vs the refractive index is depicted as follows: Table C.4 Concentration Vs Refractive Index for PTSA Solutions Wt % of PTSA solution Actual refractive index of the solution 69.56 1.4479 64.8 1.4384 60.3 1.4273 56.08 1.4203 52.11 1.4128

PAGE 168

148Appendix C: (Continued) The refractive index for PTSA was obtained by the extrapolation of the above data to 100% thereby yielding the va lue of the refractive inde x of pure PTSA as 1.5092 With the estimates of the refractive index for PTSA, SDBS and with the estimates of the extinctions for PTSA and SDBS, the opt ical properties for SDBS can be obtained using the Kramer’s Kronig transforms. The Kramers Kronig transforms relate th e real and the imaginary parts of the complex refractive index. They are expressed as d P n2 2) ( * 2 1 ) ( C-19 d n P2 2) ( * 2 ) ( C-20 is the frequency P is the principal value of the integral. If either of the ) ( n or) ( is know, the other can be calculated. To obtain the real part of the complex refr active index of the mono mer, equation (C-19) is divided into three parts and is expressed as 2 12 2 2 2 2 0 2 2) ( ) ( ) ( ) ( ) ( ) ( 2 1 ) ( d d d n C-20 The above three integrals are numerically integrated34,35 Optical properties for styrene are represented in figures 58 and 59.

PAGE 169

149 Appendix C: (Continued) Figure 56 Refractive Index of Decane at Different Temperatures 100 200 300 400 500 600 700 800 900 1.38 1.4 1.42 1.44 1.46 1.48 1.5 1.52 1.54 20 Deg C 50 Deg C Wavelength (nm)R e f r a c t i v e I n d e x22 Deg C 60 Deg C

PAGE 170

150 Appendix C: (Continued) Figure 57 Refractive Index of Styrene Figure 58 Absorbance of Styrene 100 200 300 400 500 600 700 800 900 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 100 200 300 400 500 600 700 800 900 0.6 0.8 1 1.2 1.4 1.6 2 2.2 Wavelength (nm)R e f r a c t i v e I n d e x 100 200 300 400 500 600 700 800 900 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Wavelength(nm) 100 200 300 400 500 600 700 800 900 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Wavelength(nm)A b s o r b a n c e

PAGE 171

151 Appendix C: (Continued) Figure 59 Absorbance of Sodi um Dodecyl Benzene Sulfonate Figure 60 Refractive Index of S odium Dodecyl Benzene Sulfonate 100 200 300 400 500 600 700 800 900 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 100 200 300 400 500 600 700 800 900 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 Wavelength (nm)A b s o r b a n c e 100 200 300 400 500 600 700 800 900 1.4 1.45 1.5 1.55 1.6 1.65 R.I 100 200 300 400 500 600 700 800 900 1.4 1.45 1.5 1.55 1.6 1.65 R e f r a c t i v e I n d e xWavelength (nm)

PAGE 172

152 Appendix D: Behavior of Surfactan t Micellar Solutions in Different Environments This appendix depicts the change in the shape of the transmission Uv vis spectrum of the micellar soluti on of Sodium Dodecyl Benzene Sulfonate as a function of temperature, pH of the suspending medium and the surfactant concentration (initial emulsification conditions). The change in th e shape of the transmission spectrum is indicative of the change in the micelle form ing characteristics of the surfactant under different emulsification conditi ons. This appendix shows th e results of the measured transmission spectra as mentioned in Chapte r 4 for identifying the variables constituting the initial emulsification condi tions, whose effect on the char acteristics of the dispersed phase of the emulsion needed to be investigated. Figure 61 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at High Temperatur es (60 Deg C) 200 300 400 500 600 700 800 900 -0.02 0 0.02 0.04 0.06 0.08 0.1 10 p H 2 p HWavelength (nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 173

153 Appendix D: (Continued) Figure 62 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at High Temperatur es (60 Deg C). Amplified Lower Wavelength Region Wavelength (nm) 240 260 280 300 320 340 0 0.005 0.01 0.015 0.02 0.025 10 p H 2 p HN o r m a l i z e d O p t i c a l D e n s i t y

PAGE 174

154 Figure 63 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatures (22 Deg C) Figure 64 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatures (22 Deg C). Amplified Lower Wavelength Region 200 300 400 500 600 700 800 900 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 200 300 400 500 600 700 800 900 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 200 300 400 500 600 700 800 900 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Wavelength (nm) 10 p H 2 p H 240 250 260 270 280 290 300 0.005 0.01 0.015 0.02 240 250 260 270 280 290 300 0.005 0.01 0.015 0.02 Wavelength (nm)10 p H 2 p H Appendix D: (Continued) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 175

155 Figure 65 Effect of pH on Su rfactant Micellar Solution of Different Concentrations at Low Temperatur es (22 Deg C). Amplified Higher Wavelength Region Figure 66 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10 310 320 330 340 350 360 370 0 1 2 3 4 5 6 7 8 9 10 x 10 -4 10 p H 2 p H 200 300 400 500 600 700 800 900 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 22 De g C 60 De g CAppendix D: (Continued) Wavelength (nm) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 176

156 Figure 67 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10. Amplified Lower Wavelength Region Figure 68 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 10. Fu rther Amplification of Lo wer Wavelength Region 240 260 280 300 320 340 360 380 0 0.005 0.01 0.015 0.02 22 Deg C 60 Deg C Wavelength (nm) 240 245 250 255 260 265 270 275 280 285 0.005 0.01 0.015 0.02 0.025 22 De g C 60 De g C Appendix D: (Continued) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y Wavelength (nm)

PAGE 177

157 Figure 69 Effect of Temperature on Su rfactant Micellar Solu tion of Different Concentrations at pH 2. Figure 70 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplification of Lower Wavelength Region 2 00 300 400 500 600 700 800 900 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 Wavelength (nm) 22 De g C 50 De g C 60 De g C 250 260 270 280 290 300 310 320 330 340 0 0.005 0.01 0.015 0.02 22 deg C 50 deg C 60 deg C 250 260 270 280 290 300 310 320 330 340 0 0.005 0.01 0.015 0.02 22 Deg C 50 Deg C 60 Deg C Wavelength (nm) Appendix D: (Continued) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 178

158 Figure 71 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplification Lower Wavelength Region Figure 72 Effect of Temperature on Surf actant Micellar Solution of Different Concentrations at pH 2. Amplification of Higher Wavelength Region 240 245 250 255 260 265 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 22 deg C 50 deg C 60 deg C 240 245 250 255 260 265 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 22 Deg C 50 Deg C 60 Deg C Wavelength (nm) 2 60 270 280 290 300 310 320 330 340 350 -2 0 2 4 6 8 10 12 14 x 10 -3 22 De g C 50 De g C 60 Deg C Appendix D: (Continued) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t yWavelength (nm)

PAGE 179

159 Appendix E: Comparison of the Transm ission Spectra of Emulsion Under Different Emulsification Conditions This appendix depicts the transmission emulsion spectra obtained under different emulsification conditions. The change in th e shape of the transm ission emulsion spectra indicates change in the char acteristics of the dispersed phase population of the emulsion as a function of initial emulsification c onditions (namely surfactant concentration, temperature and the pH of the suspending me dium). These measured transmission spectra of the emulsions (prepared under different emulsification conditions) were deconvoluted using the turbidity equation as mentioned in Chapter 3. The experimental procedure to obtain these spectra was described in Chapte r 4. The results obtained as by deconvoluting these spectra are reported in Chapter 5. Figure 73 Effect of Surfactant Concentr ation at Low pH (pH 2) and High Temperature (60 Deg C) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 Higher surfactant concentration Lower surfactant concentration OD Wavelength (nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength (nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 180

160 Appendix E: (Continued) Figure 74 Effect of Surfactant Concentration at Low pH (pH 2) and High Temperature (60 Deg C). Amplified Lower Wavelength Region Figure 75 Effect of Surfactant Concentration at Low pH (pH 2) and Low Temperature (50 Deg C) 300 400 500 600 700 800 1.65 1.7 1.75 1.8 1.85 x 10 -3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength (nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 x 10 -3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 181

161 Appendix E: (Continued) Figure 76 Effect of Surfactant Concentr ation at Low pH (pH 2) and Low Temperature ( 50 Deg C). Amplified Lower Wavelength Region Figure 77 Effect of Surfactant Concentration at High pH (pH 10) and High Temperature (60 Deg C) 300 400 500 600 700 1.65 1.7 1.75 1.8 1.85 x 10 -3 Higher surfactant concentration Lower surfactant concentration OD Wavelength (nm) 300 400 500 600 700 1.65 1.7 1.75 1.8 1.85 x 10 -3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength (nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 x 10 -3 Higher surfactant concentration Lower surfactant concentration OD Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 x 10 -3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 182

162 Appendix E: (Continued) Figure 78 Effect of Surfactant Concentra tion at High pH (pH 10) and High Temperature ( 60 Deg C). Amplified Lower Wavelength Region Figure 79 Effect of pH at Low Surfactan t Concentration (S/O Ratio = 0.0154) and High Temperature (60 Deg C) 300 400 500 600 700 800 1.65 1.7 1.75 1.8 1.85 x 10 -3 Higher surfactant concentration Lower surfactant concentration OD Wavelength(nm) 300 400 500 600 700 800 1.65 1.7 1.75 1.8 1.85 x 10-3 S/O ratio = 0.046 S/O ratio = 0.0154 Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 x 10 -3 10 pH 2 pHOD Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 x 10 -3 p H = 10 pH = 2 Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 183

163 Appendix E: (Continued) Figure 80 Effect of pH at Low Surfactant Concentration (S/O Ratio = 0.0154 ) High Temperature (60 Deg C). Amplif ied Lower Wavelength Region Figure 81 Effect of pH at High Surfactant Concentration (S/O Ratio = 0.046) and High Temperature (60 Deg C) 250 300 350 400 450 1.62 1.64 1.66 1.68 1.7 1.72 1.74 x 10 -3 High pH Low pH OD Wavelength(nm) 250 300 350 400 450 1.62 1.64 1.66 1.68 1.7 1.72 1.74 x 10 -3 pH =10 pH = 2 Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 High pH Low pH OD Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 p H = 10 pH = 2 Wavelength(nm)N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 184

164 Appendix E: (Continued) Figure 82 Effect of pH at High Surfact ant Concentration (S /O Ratio = 0.046) and High Temperatur e (60 Deg C). Amplified Lower Wavelength Region Figure 83 Effect of Temperature at High Surfactant Concentrations (S/O ratio = 0.046) and High (pH= 10) 250 300 350 400 450 500 550 600 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 1.78 1.8 1.82 x 10 -3 High pH Low pHOD Wavelength(nm) 250 300 350 400 450 500 550 600 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 1.78 1.8 1.82 x 10 -3 pH = 10 pH = 2 Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 High temperature Low temperature (22 Deg C) (60 deg C) OD Wavelen g th ( n 200 300 400 500 600 700 800 900 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 x 10 -3 Tem p erature = 60 De g C Temperature = 22 Deg C Wavelength(nm)N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 185

165 Figure 84 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and High pH (pH = 10) Figure 85 Effect of Temperat ure at High Surfactant Concen tration (S/O Ratio= 0.046) and High pH (pH = 10). Amplified Lower Wavelength Region 240 260 280 300 320 340 1.6 1.65 1.7 1.75 1.8 1.85 1.9 x 10 -3 High temperature (60 deg C) Low temperature (22 deg C)Wavelength(nm) OD 240 260 280 300 320 340 1.6 1.65 1.7 1.75 1.8 1.85 1.9 x 10 -3 Temperature = 60 Deg C Temperature = 22 Deg C Wavelength(nm) a nd high pH (blow up) 200 300 400 500 600 700 800 900 1.6 1.65 1.7 1.75 1.8 1.85 1.9 1.95 x 10 -3 High temperature (60 deg C) Low temperature (22 deg C) OD Wavelength(nm) 200 300 400 500 600 700 800 900 1.6 1.65 1.7 1.75 1.8 1.85 1.9 1.95 x 10 -3 Temperature = 60 Deg C Temperature = 22 Deg C Wavelen g th ( nm ) Appendix E: (Continued) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 186

166 Appendix E: (Continued) Figure 86 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and Low pH (pH = 2) Figure 87 Effect of Temperature at High Surfactant Concentration (S/O Ratio = 0.046) and Low pH (pH = 2). Amplified Lower Wavelength Region Effect of temperature at high Surfactant concentration and low p 300 400 500 600 700 800 1.65 1.7 1.75 1.8 1.85 1.9 1.95 x 10 -3 High temperature (60 deg C) Low temperature ( 50 deg C)OD Wavelength(nm) 300 400 500 600 700 800 1.65 1.7 1.75 1.8 1.85 1.9 1.95 x 10 -3 Temperature = 60 Deg C Temperature = 50 Deg CWavelength(nm) 240 250 260 270 280 290 300 310 320 330 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 1.78 1.8 1.82 x 10 -3 High temperature Low temperature ( 50 deg C) OD Wavelength(nm) 240 250 260 270 280 290 300 310 320 330 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 1.78 1.8 1.82 x 10 -3 Temperature = 60 Deg C Temperature = 50 Deg C Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 187

167 Appendix E: (Continued) Figure 88 Effect of Temperature at Low Surfactant Concentration (S/O Ratio = 0.0154) and Low pH (pH =2) Figure 89 Effect of Temperature at Low Surfactant Concentra tion (S/O Ratio=0.0154) and Low pH (p H =2 ). Amplified Lower Wavelength Region 200 300 400 500 600 700 800 900 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 x 10 -3 High temperature Low temperature ( 50 deg C)Wavelength(nm) OD 200 300 400 500 600 700 800 900 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 x 10 -3 Temperature = 60 Deg C Temperature = 50 Deg CWavelen g th ( nm ) 240 260 280 300 320 340 360 380 400 420 1.6 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 x 10 -3 High temperature Low temperature ( 50 deg C)Wavelength(nm) OD 240 260 280 300 320 340 360 380 400 420 1.6 1.62 1.64 1.66 1.68 1.7 1.72 1.74 1.76 x 10 -3 Temperature = 60 Deg C Temperature = 50 Deg C Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 188

168 Appendix F: Experimental Results for Dete rmining the Stability of the Emulsion In order to determine the time needed for the emulsion to ach ieve stability, the Uv vis transmission spectra of the emulsion wa s recorded over a period of time. The size distribution characteristics of the partic le populations comprising the emulsions are reflected upon the shape of its Uv vis transm ission spectra. Therefore a constant shape of the Uv vis transmission spectra will indicate that the emulsion has achieved stability and the characteristics of the particle size dist ribution of the particle populations comprising it are constant. Once the emulsion stability is achieved, meaningful analysis of the transmission spectrum of the emulsion can be performed to characterize the dispersed phase. In Chapter 5 is reported the analysis of the transmission spectra of the emulsion at different times before the stability is achieve d. In this appendix, th e spectral results for the experiment to determine the time required for the emulsion to achieve stability are reported.

PAGE 189

169 Figure 90 Actual Offline Spectrum of Emul sion at Room Temperat ure at Different Times Appendix F: (Continued) 200 300 400 500 600 700 800 900 0 0.5 1 1.5 2 2.5 15 min 20 min 1 hr 55 min 2 hrs 2 hrs 15 min 2 hrs 30 min Wavelength (nm)O p t i c a l D e n s i t y

PAGE 190

170 Appendix F: (Continued) Figure 91 Comparison of Normalized Offline Spectrum of Emulsion at Room Temperature at Different Times. Norma lization from 230 to 820nm Figure 92 Comparison of Normalized Offline Sp ectra of Emulsion at Room Temperature at Different Times when Stability is Achieved. Normalization from 230 to 820nm 200 300 400 500 600 700 800 900 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 10 -3 2 hrs 15 min 2 hrs 30 min Wavelength (nm)N o r m a l i z e d O p t i c a l D e n s i t y 300 400 500 600 700 800 1.6 1.8 2 2.2 2.4 2.6 2.8 3 x 10 -3 15 min 20 min 1 hr 55 min 2 hrs 2 hrs 15 min 2 hrs 30 min Wavelength (nm)N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 191

171 Appendix F: (Continued) Figure 93 Comparison of Normalized Offlin e and Online Spectrum after Stabilization of the Emulsion. Normalization Wavelength from 230 to 820 nm Figure 94 Comparison of the Offline and Onlin e Normalized Spectra with Statistics 200 300 400 500 600 700 800 900 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 10 -3 N o rm a l i z e d O p t i c a l D e n s i t yWavelength (nm) On-line Off-line 200 300 400 500 600 700 800 900 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 x 10 -3 For mean and confidence intervals Offline spectrum after 2 hrs15 min Corresponding online Replicate measurements Wavelength (nm)N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 192

172 Figure 95 Comparison of the Online and Off line Normalized Spectra with Statistics Amplified Lower Wavelength Region Appendix F: (Continued) 250 300 350 400 450 500 1.58 1.59 1.6 1.61 1.62 1.63 1.64 1.65 1.66 x 10 -3 Corresponding online Replicate measurements Offline spectrum after 2 hrs15 min For mean and confidence intervals Wavelength (nm)N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 193

173 Appendix G: Calculated Particle Size Di stribution of Small and Large Droplet Populations In this Appendix are shown the calculate d droplet size distributions of the dispersed phase of the emulsion for th e experiments performed under different emulsification conditions. The particle size distributions shown in this appendix are normalized by the height. The comparison be tween the breadth and the mean of the distribution of the large and small nano-dr oplet population can hence be made. The parameters completely describing the depict ed size distributions of the droplets are reported in Chapter 5. The change in the dr oplet size distribution of the nano-droplet population as a function of the initial em ulsification conditions can be greatly appreciated. Figure 96 Normalized Droplet Size Distribu tion for Experiment 1 1.0 0.5 0 40 60 80 100 120 140 160 180 200 220 D (nm) Nano-Droplet Population 0 2 4 6 8 10 12 x 104 D (nm) Large Droplet Population f(D)

PAGE 194

174 Figure 97 Normalized Droplet Size Distribution for Experiment 2 Figure 98 Normalized Droplet Size Distribution for Experiment 4 1.0 0.5 0 1.0 0.5 0 0 50 100 150 200 250 300 D (nm) of Nano-Droplet Population 0 5 10 15 x 104 D (nm) of Large-Droplet Population 40 60 80 100 120 140 160 180 D (nm) of Nano-Droplet Population 0 5 10 15 x104 D (nm) of Lar g e Dro p let Po p ulation f(D) f(D) Appendix G : (Continued)

PAGE 195

175 Figure 99 Normalized Droplet Size Distribution for Experiment 5 Figure 100 Normalized Droplet Size Distribution for Experiment 6 1.0 0.5 0 1.0 0.5 0 15 20 25 30 35 40 45 50 60 65 D (nm) of Nano-Droplet Population 0 2 4 6 8 10 12 x 104 D (nm) of Large Droplet Population 15 20 25 30 35 40 45 50 D (nm) Nano-Droplet Population 0 2 4 6 8 10 12 14 x104 D (nm) Large Droplet Population f(D) f(D) Appendix G: (Continued)

PAGE 196

176 Appendix G: (Continued) Figure 101 Normalized Droplet Size Distribution for Experiment 7 1.0 0.5 0 1.0 0.5 0 10 20 30 40 50 60 70 D (nm) Nano-Droplet Population 0 2 4 6 8 10 12 x 104 D (nm) Large Droplet Population 15 20 25 30 35 40 45 50 55 D (nm) Nano-Droplet Population Figure 102 Normalized Droplet Size Distribution for Experiment 8 0 2 4 6 8 10 12 14 x 104 D (nm) Large -Droplet Population f(D) f(D)

PAGE 197

177 Appendix H: Preliminary Results from th e Emulsion Polymerization Reactions As indicated in the introduction and the problem definition sections, the overall objective of the project is the elucidation of the particle nucleation mechanisms in emulsion polymerization reactions. The prelimin ary work leading to this dissertation was accomplished through a study of emulsion polymerizations under a variety of experimental conditions. This A ppendix describes the efforts th at lead to the formulation the approach followed in this dissertati on. The reactor configuration, measurement strategy, and experimental conditions used for this study are describe d in this Appendix. Relevant results are presented and discussed. The experiments described in this Appendix were jointly performed. Vara21 reports the results of the non-spectroscopic measurements. The results from the spectroscopic measurements that served as the preliminary work for formulating the approach described in this dissertation are report ed in this Appendix. In this section is describe d the sensor array developed in the Polymer Synthesis and Characterization laboratory, USF fo r monitoring the emulsion polymerization reaction to understand the nucle ation mechanism. The entire experimental set up, reactor configuration and the proce dure in which the reaction was conducted is described. H.1 Materials and Methods Purified styrene was used as the monome r in this study. Water was the suspending medium. Sodium Lauryl Sulfonate was the em ulsifier used. Potassium persulphate was used as the initiator. Argon gas was used to ma intain inert environment. In order to carry

PAGE 198

178 Appendix H: (Continued) out the reaction, the above mentioned reactant s were added in the proportions as per the recipe in Table H.1. The experimental conditio ns for carrying out the reaction are listed. Table H.1 Recipe and Experimental conditi ons for the Emulsion Polymerization Reactions Components Vendor Quantities Nanopure disuntiled water 662 grams Styrene Alderich Chemical Co. 228 grams Sodium Lauryl Sulfate Sigma Chemical Co. 10 grams Sodium bi Carbonate J. T. Baker Chemical Co. 1.0 grams Potassium persulfate Sigma Chemical Co. 1.0 grams Temperature 60 Deg C RPM 300 The emulsion polymerization latex characteristic s such as density, in terfacial tension and transmission Uv vis spectrum at different times were monitored with the help of the sensor array for understanding th e nucleation mechanism as desc ribed in the later part of this Appendix. H.2 Experimental Setup and Reactor Configuration Batch emulsion polymerization reactions we re conducted in a one liter jacketed glass reactor manufactured by Kontes, Glass Co. Table H.1 gives a typical the recipe used for carrying out emulsion polymerization r eactions for styrene. Since the reaction is a thermally initiated exothermic reaction it is necessary to have an efficient

PAGE 199

179 Appendix H:(Continued) temperature control system to control the rate of the reaction. Water of appropriate temperature is circulated thr ough the jacket of the reactor to control the temperature of the reaction mixture inside th e reactor. The reaction mixture is drawn from the bottom of the reactor. Access to the contents insi de the reactor throughout the reaction for monitoring purposes is possible through the several ports of the reactor top. Th rough the center most port of the reactor top, passes the stirrer that maintains constant agitation inside the reaction mixture throughout the r eaction process. The stirrer use d, is a three bladed stirrer coated with teflon tape rotati ng at a speed of about 300 rpm. Through the other port of the reactor top, the thermocouples recording the temperature insi de the reaction mixture are inserted. The condenser, a jacketed tube fo r condensing the vapors that may be escaping from the reactor is attached to the reactor top through a speci ally provided port. The two metallic probes of the surface tensiometer mon itor the interfacial te nsion of the reaction mixture as function of time. A baffle is intr oduced from a port very close to the surface tensiometer probes to protect the nitroge n bubbles from bursting. To prevent the stopping of the reaction by the oxygen from th e air, typically a blanket of argon is maintained over the entire reaction mixture throughout the reaction. A slip-stream is taken from the bottom of the reactor to the densitometer for conversion measurements. The latex mixture coming out of the densitometer is returned to the reactor for most part. Some of it is passed on to the dilution system where the

PAGE 200

180 Appendix H:(Continued) reaction mixture is diluted for spectrophotometric measur ements. Figures 103 and 104 show the schematic of the experimental setup of the densitometer and the surface tensiometer employed in the sensor array. H.3 Experimental Procedure After setting up the reactor and the meas urement system, the ingredients were added in the following sequence. The water is ad ded first in the reactor, to which is then added the surfactant. The mixture is kept unde r constant agitation as we add monomer to it. A blanket of argon is maintained ove r the reaction mixture by bubbling argon through the hollow probes of the surface tensiometer continuously. A three-way valve helps to change the gas flow in the hollow probes of the surface tensiometer from Argon to Nitrogen whenever a measurement of interfaci al tension is desire d. The temperature of the reaction mixture in the r eactor is maintained at 60 De g Celsius by circulating hot water from the water bath in the jacket of the reactor. Once the emulsion is stabilized, and the de sired temperature for the experimentis reached, the potassium persulphate was adde d as the initiator. The polymerization reaction begins with the addition of the initia tor. Different parameters characterizing the reaction latex such as conve rsion, interfacial tension a nd the transmission Uv vis spectrum are monitored with the help of the sensor array described below.

PAGE 201

181 Appendix H: (Continued) H.4 Sensor Array to Monitor the Em ulsion Polymerization Reaction The sensor array developed at USF laborat ory consists of densitometer that gives an online estimate of the conversion th roughout the reaction, surface tensiometer for measuring the interfacial tension of the r eaction latex mixture and the dilution system with spectroscopic sensor for obtaining the Uv vis transmission spectrum of the reaction latex mixture. The densitometer has a vibrating hollow U tube, the frequency of the vibration changes with the change in the density of th e fluid inside it. As the conversion of the monomer to polymer increases, the density of the latex mixture in the reactor also changes thereby causing a change in the fre quency of the vibration. This change in frequency of the vibration is then expressed in terms of conversion21. Information on conversion was obtained using densitometer an d validated using traditional gravimetric techniques. Experimental data on the conversion of th e monomer to polymer with respect to time is necessary for providing the information regardi ng the composition of the particles in homo-polymerization. Composition of the reaction site has an effect on the rate of entry of radicals and hence on the process dynamics in general. Monitoring of the conversion of the polymerization reaction, con tinuously and in real time is necessary to understand the thermodynamics and the kine tics of the emulsion polymerization

PAGE 202

182 Appendix H: (Continued) processes. Since the conversi on is dependent upon the rate of the reaction, which in turn depends upon the number of particles, ther efore the information on conversion thus reflects upon the number of particles formed by nucleation. The online surface tensiometer used to obtai n estimates of the interfacial tension is a differential bubble tensiometer. It cons ists of two hollow probes of unequal diameter through which is bubbled the nitrogen gas. Th e difference in the pressure inside the bubbles from the respective probes is a functi on of the interfacial tension of the latex mixture21. Experimental data on the change in surf ace tension of the emulsion in the reactor with respect to time provides us with the information on the free emulsifier concentration. The free emulsifier concen tration keeps changing causing the surface tension of the latex particulate system to go high as the reaction progresses indicating more and more surfactant is being used up to stabilize the growing pa rticles. The change in the interfacial tension is indicative of the change in th e emulsifier distribution. The knowledge of the emulsifier distribution as a function of particle size and composition will provide an insight of the colloidal chem istry and the thermodynamics of the process to understand the nucleation mechanism.

PAGE 203

183 Appendix H: (Continued) The online Uv vis spectrum is obtained w ith the help of the continuous sampling and parallel dilution system devel oped at USF (patent number 5,907,108, (1998)). The dilution system enables the acquisition of the transmission Uv vis spectrum of the reaction latex mixture conti nuously and in real time. The spectrometer employed for acquiring the Uv vis spectrum was a HP 8452A spectrometer. The total time required in diluting the concentrated reaction latex mixtur e and acquiring its Uv vis spectrum is less than a minute in which dilution to the fifth order of magnitude was achieved. The above mentioned sensor array has been described in great detail in the works of Paul Sacoto33 and Jaime Vara21. Change in the shape of the transmissi on Uv vis spectrum of the emulsion latex provides the information on composition of the latex and the particle size distribution of that particulate system. Information on th e particle size distribution is obtained by deconvoluting the transmission Uv vis spectra34,35. Information on the early particle size distribution especially prior to the beginning of the reaction is critical in the identification of the likely nucleation loci. The information on composition can be obtained spectroscopically. Monitoring of the particle size distribut ion, particle number and the particle composition of all the particle population compri sing the reaction mixture is possible throughout the reaction using Uv vis spectroscopic techniques. The process of emulsion polymerization can thus be bette r understood with the information thus obtained.

PAGE 204

184 Appendix H: (Continued) H.5 Results Obtained From the Polymerization Reaction Different researchers have tried to extrapolate the critical information characterizing the reaction mixture during the first stage of the reaction to time zero has been the proposed approach for elucidating the nucleation mechanism. However measurements to characterize the “time zero condition” were not performed by them. This led to the development of the sensor arra y described earlier for mo nitoring the critical parameters of the emulsion polymer ization reaction. However, the difficulty in sampling the reaction latex at time zero (tim e zero was marked by the addition of the initiator) presents the prime hurdle for unders tanding the nucleation mechanism. It was for this reason that the experiments were pe rformed with the model system as described in the main body of this dissertation for studying the nucleation mechanism. Even so, the information obtained using this sensor array provides insight into the process kinetics and thermodynamics of th e emulsion polymerization process. Jaime Vara21 presents the results obtained by de nsitometry and the surfacetensiometer measurements for the polymerization reactions of styrene, butyl methacrylate and the copolymerization of styrene a nd butyl methacrylate. Vara21 reported the conversion and the change in the interfacial tension of the reac tion latex as a function of time as the reaction progressed. The interfacial te nsion did not change much during the first stage of the reaction indicating the adsorpti on of only the free emulsifi er on the droplet populations. The corresponding composition of the particles could be potentially obtained from the conversion estimates.

PAGE 205

185 Appendix H: (Continued) The information on the particle compos ition, number and populations can be obtained from the spectroscopic measurements The change in the composition and the particle size distribution of the populations present in the reactor is reflected upon its transmission Uv vis spectrum as the reaction progresses in time. This change in the shape of the spectrum of the reaction mixture, as th e reaction progresses is depicted in Figure 109. Note the decrease in the characteristic monomer peak at around 220 to 230 nm wavelength as the reaction progr esses. With proper calibration, the information on the conversion of the monomer in the reaction late x can be obtained. The results are depicted in Figure 110. Information on the droplet popu lations, at the beginning of the reaction (before the addition of the initiator at time t=0 and 0% conversion) was obtained by deconvoluting the spectral signal. The results are presented in Table 5.4. Strong characteristic monomer peak in th e region of interest of the Uv vis spectrum makes it impossible to identify th e surfactant signal. He nce the study of the surfactant distribution over the particle populations pr esent in the dispersed phase becomes a very difficult task. The informati on of the surfactant distribution over the dispersed phase at the beginning of the reaction is critical for elucidating the nucleation mechanism in emulsion polymerization. This difficulty posed by the monomer emulsion systems can be overcome by the usage of the model emulsion systems showing the same dispersion characteristics but at the same time also offering a good contrast for

PAGE 206

186 Appendix H: (Continued) spectroscopy measurements. The model emulsion systems and the experiments performed with them are described in the main body of the dissertation. The information on the particle size distri bution of the polystyrene latex present inside the reactor after 100% conversion is presented in Table H.2. This information was obtained by deconvoluting the entire Uv vis transmission spectrum of the polystyrene latex obtained as a result of the polymeriza tion reaction of styrene to polystyrene. The particle size distribution of the fina l latex is represented in Figure 111. Table H.2 Preliminary Results obtained by the dec onvoluting the spectral signal obtained at the end of the react ion (for 100% conversion)52 Parameter Estimates Number Average Diameter 87.4 nm Standard Deviation of the Particle Size Distribution 1.34 nm Particle Number (Number/ml) 6.0435*1010 Residual Sum of Squares 3.6*10-4 Standard Deviation of the residuals 0.0012 H.6 Temperature Control System The sensing element of this temperature control system is a J type thermocouple coated with teflon tape. It remains immersed inside the reactor containing the reaction mixture throughout the reaction. The thermoc ouple sends the voltage signal (analog

PAGE 207

187 Appendix H: (Continued) signals) to the A/D converter inside the com puter via multiplexer. Corrective signals are sent to the final control element by the te mperature control program such that the temperature inside the reactor is maintained at a desired set point. The final temperature of the flow-stream entering the reactor jacket dictates the effectiveness of the temperature control action. The temperature of this stream is governed by the valve position of the pneumatically operated control valve (the final control elem ent). The flow diagrams of the analog and digital signals from the sens ing element to the controller and from the controller to the final control element for th e desired control action are shown in figure 105. Figure 106 represents the schematic of the fl ow path of the water for the temperature control system.

PAGE 208

188 Appendix H: (Continued) Figure 103 Schematic of Densitometer Setup

PAGE 209

189 Appendix H: (Continued) Figure 104 Surface Tensiometer Setup (Schematic)

PAGE 210

190 Appendix H: (Continued) Figure 105 Signal Flow-path for th e Temperature Control System

PAGE 211

191 Appendix H: (Continued) Figure 106 Water Flow for the Temperature Control System

PAGE 212

192 Appendix H: (Continued) Figure 107 Schematic of the Dilution System (As represented by Sacoto33) Sensor Sensor Sensor Sensor Sensor Sensor qd(1) qd(2) qs(1),Cs(1) qs(2) Cs(n)Diluent q [=] ml/min C [=] g/ml Sample ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s Parallel Sampling Parallel Dilution Continuous Monitoring Sensor Sensor Sensor Sensor Sensor Sensor qd(1) qd(2) qs(1),Cs(1) qs(2) Cs(n)Diluent q [=] ml/min C [=] g/ml Sample ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s Parallel Sampling Parallel Dilution Continuous Monitoring Sensor Sensor Sensor Sensor Sensor Sensor qd(1) qd(2) qs(1),Cs(1) qs(2) Cs(n)Diluent q [=] ml/min C [=] g/ml Sample ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s Parallel Sampling Parallel Dilution Continuous Monitoring Sensor Sensor Sensor Sensor Sensor Sensor qd(1) qd(2) qs(1),Cs(1) qs(2) Cs(n)Diluent q [=] ml/min C [=] g/ml Sample ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s ) ( ) ( ) ( ) ( ) 1 ( n q n q n q n C n Cd s s s s Parallel Sampling Parallel Dilution Continuous Monitoring

PAGE 213

193 Appendix H: (Continued) Figure 108 Reactor Setup for the Polymeri zation Reaction with the Dilution System

PAGE 214

194 Appendix H: (Continued) Figure 109 Change in the Shape of the Tran smission Spectra of th e Reaction Mixture as the Reaction Progresses Time ( minutes ) Wavelength(nm) N o r m a l i z e d O p t i c a l D e n s i t y

PAGE 215

195 Appendix H: (Continued) Figure 110 Estimated Conversion of Styr ene to Polystyrene Using Spectroscopy Time(Seconds) Spectroscopic estimate o Gravimetric estimate % Conversion

PAGE 216

196 Appendix H: (Continued) Figure 111 Particle Size Distribution of the Polystyrene Particles Formed as a Result of the Polyme rization Reaction After the Reaction is Completed f(D) Diameter(cm)

PAGE 217

197 Appendix H: (Continued) Figure 112 Comparison Between the Estimated and the Measured Spectra of the Styrene Emulsion at Time Zero (Before the beginning of the Reaction) 100 200 300 400 500 600 700 800 900 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Wavelength(nm) Optical Density

PAGE 218

198 Appendix I: Protocol for Operating the Dilution System As mentioned in Chapter 4, this appendi x enumerates the steps for operating the dilution system for obtaining th e transmission UV Vis spectrum of the emulsion so that it can be characterized in terms of the size and the compositional ch aracteristics of the droplet population/s co mprising it. 1. Make the decision on the type of spectrometer to be used The options available were a. Hewlett Packard (Model HP 8452A) b. Ocean Optics (Model S2000, PC 2000) *. Incase of an Ocean Optics Spectrometer, please follow the configuration procedure mentioned in 12 and 13 and the experimental procedure from 14 through 26. The connections of the dilution sy stem as elaborated in st ep 3 are needed to be made regardless. 2. If we are using the Hp8452A spectrometer then make sure that the spectrometer is “On” for at least 45 minutes pr ior to starting the experiment. 3. Make the connections of the dilution system as follows: a. The hose whose one end is connected to the dilution branch and the other end is immersed in the diluent tank is fit into the cartridge of the diluent pump for pumping the di luent in that branch. The diluent is similarly pumped into each branch of the dilution system.

PAGE 219

199 Appendix I: (Continued) The hoses of size L/S 17 (Masterflex Cole Parmer) are used for pumping the diluent in the diltion branches. These hoses fit into each of these cartridges and then locked on th e pump head with 4 rollers. Make sure that the side locks are pull ed down so that the hoses are secured in place and do not change their positions even as the roller rotates. The diluent hoses communicate with the d iluent tank and the dilution system. b. The other ends of all the branches of the dilution system come together at a common meeting point leading to the “flow-through” cuvette. The other end of the “flow-through” cuvette is connected to the sink. c. For pumping the diluted sample from one branch is pumped into another branch for further dilution, the dilute d sample lines communicate from one branch of the dilution sy stem to the other throug h another cartridge pump on which fit the smaller cartridges. The L/S 13 (Masterflex Cole Parmer) hoses are placed in the smaller cartridge. These cartridges are now locked on the pump-head with 8 rollers. Make sure that the side locks are pull ed down so that the hoses are secured in place and do not change their positions as the rollers move. d. The concentrated sample that is desi red to be diluted is pumped into the first branch of the dilution system w ith the help of another sample pump. The sample is pumped by the manostat pump into the first branch of the

PAGE 220

200 Appendix I: (Continued) dilution system through the port provided. Make sure that this hose is secured tightly in place by securing it with the locks. Note: For the schematic of the dilution system refe r figure 124 in Appendix I of this dissertation. Figure 125 shows the photogra ph of the dilution system along with the reactor assembly. 4. After giving enough time for the Hp8452 to warm up go to the Hp8452 menu and hit “scan blank” 5. The flow-through cuvette is now placed in the spectromete r cuvette holder. 6. The Diluent pump and the diluted sample pump is now turned on so that only the Diluent passes through the fl ow cell. Make sure that the diluent flow in the dilution system is continuous and without any air bubbles. 7. The reference is now taken. 8. The sample pump is now turned on so that the sample now enters the dilution system. 9. Appropriate positions of the valve ar e maintained as per the need of the stage of the dilution is required such that the spectrum obtained is within the linear range of the spectrometer. 10. As the diluted sample shows up in th e flow cell another spectrum is taken. 11. The reference spectrum is subtracted from the above spectrum. The spectrum of the sample is thus obtained.

PAGE 221

201 Appendix I: (Continued) 12. If we are using an Ocean Optics spect rometer we first need to configure it. 13. Two types of spectrometers ar e available depending upon whether we are using a laptop or a desk top. Spectrometer Confi guration for a desktop computer. Two types of spectrome ter are available for desktop usage. 1. S2000 which is a stand alone spectrometer 2. PC2000 which is a plug in spectrometer. Procedure: a. Shut down the computer b. Select the Base address and th e IRQ settings by changing the appropriate switches provided on the A/D card for the S2000 spectrometer and in case of PC 2000 we select the switches provided on the card that carries the spectrometer. For the selection of the switches re fer to “Operating Manual and User’s guide” provided by Ocean Optics. c. The Base address and the IRQ settin gs selected should not coincide with the other devices inside the computer. d. Now place this hardware inside the slot provided in the computer. e. Now turn on the computer. f. Connect the light source to the spectrometer using the optic fiber. g. Go to Programs----OOIBase32------Spectrometer ------configure.

PAGE 222

202 Appendix I: (Continued) h Input the serial # Spectrometer type ,A/D card type i. Go to A/D interface and put in the value of the Base Address and the IRQ. a. Check for the signal. If we suntil are unable to find any signal then Restart the computer. *. Spectrometer Configuration for Laptop computer a. Spectrometer S2000 is connected to the A/D card DAQ700. b. This card is inserted in the laptop in th e space provided. c. Go to Programs --------OOIBase32 -----Spectrometer----Configure. d. Put in the spectrometer type and serial # and the type of interface e. Connect the light source to the spectrometer f. Restart the laptop. g. The spectrometer should now be configured. 14. The Flow cell is placed in the Cuvette holder from the Ocean Optics. 15. A Split fiber of 400 microns solarizat ion proof is employed to carry the light from the source to the cuvette holder carrying the flowcell One collecting arm of th fiber is connected to the UV source while the other arm is connected to the visible light s ource. The common arm which now carries both the UV and th e visible light goes to the cuvette holder. 16. One 400 micron straight fiber is conn ected to the spectrom eter which collects the UV Vis light coming from the flow cell and takes it to the spectrometer.

PAGE 223

203 Appendix I: (Continued) 17. Note that in the Scope Mode of the OOIBase32 software the signal obtained should be between 2500 to 3000 counts. 18. We can manipulate the signal obtained by playing with the windows of the source or by manipulatin g the integration time. 19. The Diluent pump and the diluted sample pump is now started such that only diluent passes through the dilution syst em and the flow-through cuvette. 20. The “lighted yellow bulb” icon on the OOIBase32 window is hit. This is the Reference. 21. The path of the light in between the collecting fiber and the delivery fiber is completely blocked and the signal goes to zero. 22. The “dark bulb icon” on the OOIBa se32 is now hit. This is the dark 23. The light path is now unblocked a nd the icon “A” for absorb ance is now hit so that now we are in absorbance mode. 24. The sample pump is now started and the material shows up in the spectrum in absorbance mode. 25. Care should be taken in maintaining the flowrate of the pumps so that we obtain absorbance within th e linear range of the spectrometer. 26. The sample spectrum is saved by: File----save experiment----giving the filename---and hitting save.

PAGE 224

ABOUT THE AUTHOR Vineet Shastry completed his Bachelo rs degree in Chemical Engineering from the University of Mumbai (Bomba y), India in 1996. Vineet Shastry, as an undergraduate worked on the techno-econom ic feasibility of portable desalination plant to solve the problem of drinking water in his district (county) in India. After completing his undergraduate course work in Chemical Engineering, he worked as a trainee Chemical Engin eer in the Sulfuric Acid Plant of M/s Albright and Wilson Chemicals, (India) Ltd. This i ndustrial experience motivated him to pursue his graduate studies. Vineet Shastr y started working on his Ph.D. degree in Chemical Engineering at the University of South Florida, Tampa Florida, USA under the guidance of Prof. Dr L H Garcia-Rubio in 1998.