t! v. £5.31... 25 .5 er!) . f a. ..: .11.: . 2r. . a thmvfi 38:1." . f a... . lwm. 3.3;. .4»... 2. a . Lu»... Sn. . .. l .. .1. . z. .. ,. .... ... 32.... 5:... .....§...... 5 of. i. 5-3521... 3‘ 1.2!. 2. in... .3 22.... . s . fix £1» 9.13... c}?! \1 g 105 £55. . 34:3...“ . gruff. «3.9. . I f 14-. €291»... , . 1.2.7 420‘ I .8.) 5313‘; .. II. I“... 2.. a. u at“: I. .z¢¢£.§\.at« . .V .I a. 3:... .!Y( t x f . 1:33.41! .1! ll.‘ 1.— : THESIS ’5 2060 This is to certify that the dissertation entitled POLYMERIC ENULSIFIERS BASED UPON REVERSIBLE HYDROPHOBES presented by BERNHARD DRESCHER has been accepted towards fulfillment of the requirements for Ph.D. degree in ab. Date filili‘bi) l 1 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan §tate University Wm Major professor CHEMICAL ENGINEERING 0-12771 um —.. “#3 _. —_——_— -4 ‘- —— 1 V — — ~- ~—‘4 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 11m Wm." POLYMI POLYNIERIC EMULSIFIERS BASED ON REVERSIBLE FORMATION OF HYDROPHOBIC UNITS By Bernhard Drescher A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemical Engineering 2000 POLlMl This 1 synthesis, and ofmemacn'li. Of this dissc‘ POIImerizatic determine lht 311d emulsifi PTD\'ided gui: A se W‘ll’lmethac C0P01.\TII€riz é’lI‘Col) met} and the initi; ABSTRACT POLYMERIC EMULSIFIERS BASED ON REVERSDBLE FORMATION OF HYDROPHOBIC UNITS By Bernhard Drescher This research has provided a fundamental characterization of the design, synthesis, and emulsification properties of reversible emulsifiers based upon a backbone of methacrylic acid and oligomeric grafts of poly(ethylene glycol). The two broad goals of this dissertation were 1) to identify the underlying relationships between the polymerization reaction conditions and the resulting polymer structure, and 2) to determine the effects of the polymer composition and molecular weight on the solution and emulsification properties of the copolymers. Meeting these two objectives has provided guidelines for the design and synthesis of this new class of emulsifiers. A semi-batch reactor was successfully designed to synthesize a series of poly(methacrylic acid-g-ethylene glycol) of predictable compositions by a free radical copolymerization of methacrylic acid with the macromonomer methoxy poly(ethylene glycol) methacrylate. The relationship between the number average molecular weight and the initiator concentration was established. The most important design conclusion from the solubility studies is that P(MAA- g-EG) copolymers with a PEG content of more than 10 wt.% are unsuitable for emulsifiers because they are too hydrophobic under acidic conditions, and precipitate out of solution. However, the pH-dependent solution properties of copolymers with an EG mutant of 3.5 and ntentiomctric titr: of It camoxylic emulsions could I basic. and back to The dynun C“Winners to st;- maclcsccnce after system. The LS i“dependent of b al'tragg droplet c. haddlllon,[he a T116113 the Clbylene Ely. methl‘lene- M. constraints Place complex With 21 indicate that the weight of the composition, th. ltSs P01ar [h an was found that, is content of 2.5 and 5 wt.% make them very appropriate for reversible emulsifiers. The potentiometric titrations confirmed that the presence of the complex decreases the acidity of the carboxylic acid moieties. Visual emulsification studies confirmed that the emulsions could be formed, broken, and reformed by adjusting the pH from acidic, to basic, and back to acidic conditions. The dynamic drop tensiometry studies led to the conclusion that the ability of the copolymers to stabilize emulsions arises primarily from their ability to prevent droplet coalescence after they are formed, and not by reducing the interfacial free energy of the system. The LSCM technique revealed that the average droplet size was essentially independent of both the copolymer composition and molecular weight, and the number average droplet diameters were found to be in the range spanning from 8 to 12 microns. In addition, the average droplet size was found to be essentially independent of time. The 2-D NOESY NMR results also indicate that, in the complex, the protons in the ethylene glycol repeat unit are spatially closer to the oc—methyl than they are to the methylene. Molecular modeling results suggest that this effect could arise from constraints placed on the PMAA backbone in order to accommodate the formation of a complex with a one-to-one repeat unit ratio. The polarity-sensitive fluorescence results indicate that there is a marked dependence of the aggregation behavior on the molecular weight of the P(MAA-g-EG) copolymers. For example for a given copolymer composition, the aggregates formed by the higher molecular copolymer may appear to be less polar than those formed by the lower molecular weight copolymers. In addition, it was found that, at a given composition, the critical aggregation concentration decreases as the molecular weight is increased. meinen Eltem gewidmet iv lwould like to h tremendous emt tits’mgn State L'm‘ lwould like It may doctoral comm. gret progress in my timers and colloid i“ MI appreciate, lwould like tr 3. l . ‘l‘Dl- Worden for 3 Dr" . LMMIJOhIlSOII a ‘i'tn- , r “142] \XRfim 1\ Che-.- f . 2 Kb 01' the” pat fitllow‘ The Students firel'atio M I Wool filth . gTOup‘ Ind; \v 333'; .- F M‘- Khanh ,\' Acknowledgments I would like to express my appreciation to my major professor Alec B. Scranton for his tremendous encouragement, support, dedication and guidance throughout my stay at Michigan State University and the University of Iowa. I would like to thank Dr. John Klier of the Dow Chemical Company for serving on my doctoral committee and for allowing me to gain industrial experience and make great progress in my research during an internship at the Dow Chemical’s functional polymers and colloids division. His collaboration, support and encouraging discussions are greatly appreciated. I would also like to acknowledge the support from Sharon Harris and David Hayes of the Dow Chemical Company. I would like to thank the members of my doctoral committee, Dr. Smith, Dr. Ofoli and Dr. Worden for their guidance and helpful suggestions. I would also like to thank Dr. Kermit Johnson and Dr. Long Li for their help during the NMR experiments with the Varian VXR-SOO. I would also like to acknowledge Dr. Joanne Whallon and Dr. Shirley Owens for their patience and help in the practical part of Laser Scanning confocal microscopy. I am grateful to Dr. Vladimir Aseyev and Dr. Wiencek for their support and for allowing me to use their Light scattering facilities at the University of Iowa. The students in the polymer lab have made the last years enjoyable and educational. I would like to thank the current and former students in Dr. Scranton’s research group, including Kiran Baikerikar, Julie Jess-op, Arvind Mathur, Vijaykumar Narayanan, Khanh Nguyen, Katy Padon and Andy Wijaya. LN OF TABLES... LIST OF FIGURES. L [mourn 1.] Importance 1.3 Block-Gm f1 ’- BACKGROL‘. 3.1 Needfor Rt 2.3 Cleat'able 1 2.3 Complemn 2.3.] In: 2.3.3 a 2.3.3 T}; 2'4 ReSPOIlSit Nt’m'orl 516”?!le 0f 2'6 Ref? Fences TABLE OF CONTENTS LIST OF TABLES ............................................................................ LIST OF FIGURES... ........................................................................ 1. INTRODUCTION ................................................................... 1.1 Importance and Impact of Reversible Emulsions 1.2 Block-Graft Copolymers as Novel Reversible Emulsifiers 2. BACKGROUND ..................................................................... 2.1 Need for Reversibility In Emulsion Technology 2.2 Cleavable Surfactants 2.3 Complexation of Oligomer With Compilmentory Polymers 2.3.1 Introduction 2.3.3 Experimental Investigation of PMAA:PEG Complexes 2. 3.3 Theory of Oligomer:Polymer Complexation 2.4 Responsive Poly(methacrylic acid) Ethylene Glycol Hydrogel Networks 2.5 Kinetics of Free Radical Polymerizations 2.6 References 3. OBJECTIVES ....................................................................... 4. SYNTHESIS AND CHARACTERIZATION OF POLYMERIC EMULSIFIERS CONTAINING REVERSIBLE HYDROPHOBES: POLY(METHACRYLIC ACID-G-ETHYLENE GLYCOL) ................................................... 4.1 Introduction 4.2 Materials and Methods 4.3 Gel Permeation Chromatography 4.4 I H Nuclear Magnetic Resonance Spectroscopy 4.5 Results and Discussion 4.5.1 Gel Permeation Chromatography results 4.5.2 IH-NMR Spectroscopy 4.5.3 average number of grafts per chain vi ..... ix OOOOXi Page 13 I7 17 19 28 29 33 40 46 48 48 49 51 53 54 54 57 60 (L 1.6 Conclusions 4.7References 5 80mm I GLYCOL) C o 5.1 1ntr0duction 5.2 Dmomic L1; 5.3 Detennit: Solution : 5.3.1 San 5-4 PotentimI copoltmt' 5.4.1 Ex"; 5.4.2 51 aqueous 5.4.3. C . aqueous 5-5 Conclusion 5'6R‘i19fences t Etrusmc; POLYlETHW 6.1 Introductiot 6.2 Ret'ersibilit 6.3 Effect of 1 emulsifit .4 ConcIusion DYNAMIC I: Z1 IerdItc‘tiO I,“ Etp er intent 7.2.13“ 7-32.11. 7 C 7.2.3 R, GEMS-Or. 1 elelenCes DROPLET S E'EG) BU 8.11 8') erduCIiC .. a ’ (1‘3 r 0101 5 4.6 Conclusions 4. 7 References 6. SOLUTION BEHAVIOR OF POLY(METHACRYLIC ACID-G-ETHYLENE GLYCOL) COPOLYMERS IN WATER ............................................ 5.1 Introduction 5.2 Dynamic Light Scattering Theory 5.3 Determination of Aggregate Size of P(MAA-g-EG) in aqueous Solution using Dynamic Light Scattering. 5.3.1 Sample Preparation 5.3.2 Data Analysis 5.4 Potentiometric titration of aqueous solutions of P(MAA-g-EG) copolymers 5.4.1 Experimental methods 5.4.2 Solubility behavior during titration of P(MAA-g-EG) in aqueous solution 5.4.3. Conformational changes during titration of P(MAA-g-EG) in aqueous solution 5.5 Conclusions 5.6 References EMULSIFICATION STUDIES OF POLY(METHACRYLIC ACID)-GRAFTED- POLY(ETHYLENE-GLYCOL) COPOLYMERS ............................ . ........ 6.1 Introduction 6.2 Reversibility of emulsions stabilized by P(MAA-g-EG) 6.3 Effect of molecular weight and composition of the pH-dependent emulsification properties of P(MAA-g-EG) 6.4 Conclusions DYNAMIC INTERFACIAL TENSION MEASUREMENTS ........................ 7.1 Introduction 7.2 Experimental 7.2.1 Sample Preparation 7.2.2 Measurement Procedure 7.2.3 Results and Discussion 7.3 Conclusions 7.4 References DROPLET SIZE ANALYSIS OF EMULSIONS STABILIZED WITH P(MAA- G-EG) BLOCK COPOLYMERS USING LASER SCANNING CONFOCAL MICROSCOPY ...................................................................... 8.1 Introduction 8.2 Background on laser scanning confocal microscopy vii Page 72 73 73 73 80 80 80 87 87 88 91 93 107 109 109 109 III 114 120 120 121 121 122 123 128 140 141 141 142 8.3 Experiment. 8.3.1 Exit 8.5 References 9 MOLECIIM commm EFFECT ...... 9-1 1118 nuclear 9'2 11w TWO-din 9-3 Formation grafts to 9.3.1 So 9.3.2 Sp 9.3.3 p“ 9142i 9.3.5 R. 9'4 Condusutu' 9'5 References 10 A POLAR DEPENDEV WATER... 10‘] lm’Oduci 10'2 P-Wne a 10.3 Experim 10.4 Results c 10.4,] 8.3 Experimental 8.3.1 Experimental Considerations 8.3.2 Sample Preparations 8.3.3 Image collection and processing 8.3.4 Data reduction 8.3.5 Results 8.4 Conclusions 8.5 References 9 MOLECULAR CHARACTERIZATION OF THE P(MAA-G-EG) COPOLYMER COMPLEX USING THE NMR NUCLEAR OVERHAUSER EFFECT ............................................................................. 9.1 The nuclear Overhauser Effect 9.2 The two-dimensional NMR NOES Y experiment 9.3 Formation and Disruption of the complex between polyethylene glycol grafts to the poly(methacrylic acid) backbone 9.3.1 Sample Property 9.3.2 Spin-lattice relaxation time T1 analysis 9.3.3 Pulse width to? calibration 9.3.4 2D-NOES Y Experiment 9.3.5 Results and Discussion 9.4 Conclusions 9.5 References 10 A POLARITY SENSITIVE FLUORESCENCE STUDY OF THE PH- DEPENDENT AGGREGATION 0F P(MAA-G-EG) COPOLYMERS IN WATER ............................................................................. 10.1 Introduction 10.2 Pyrene as a photophysical Probe to characterize Polarity 10.3 Experimental Methods and techniques 10.4 Results and Discussion 10.4.1 Excitation Spectra 10.4.2 Emission Spectra 10.4.3 The critical aggregation Concentration 10.5 Conclusions 10.6 References 11. CONCLUSIONS ....................................................................... viii Page 143 143 144 144 146 148 149 161 162 163 168 171 171 173 173 174 174 I77 186 188 188 I89 192 I93 193 196 198 200 212 214 Tak4l Inkll Thk43 ThkSJa TURSJb Thk7t Table 4.1 Table 4.2 Table 4.3 Table 5.1a Table 5.1b Table 7.1 LIST OF TABLES Chapter 4 Gel permeation chromatography results for the moments of the molecular weight distribution of poly(methacrylic acid-g-ethylene glycol) copolymers ........................... 1H-NMR results for the MAA to EG repeat unit ratio in the poly(methacrylic acid-g-ethylene glycol) copolymers. Average number of grafts per Chain calculated based upon a combination of the GPC and the NMR experimental results .............................................. Chapter 5 List of Light scattering analysis results for samples P(MAA-g-EG) 5/ 10, 5/5, 5/0.5, 2.5/ 10, 2.5/1 and 2.5/0.5. List of Light scattering analysis results for samples P(MAA-g-EG) 0/05 and 5/1 ................................... Chapter 7 Overview of complete interfacial tension data in the system methyl laurate/water at the presence of P(MAA- g-EG) copolymers. The aqueous polymer solutions were titrated to the pH values as indicated in the table. In the adjacent columns, the relative standard deviations for the measured interfacial tension results are given ............... ix Page 69 7O 71 105 106 139 Table 9.1 Table 9.2 Table 9.1 Table 9.2 Chapter 9 Exponential data analysis of spin-lattice relaxation times T1 for observed resonances in the chemical shift range of 0 to 4 ppm for P(MAA-g-EG) 2.5/5 in D20 ................. Cross-peak distance analysis for pairs CHz—CH3, EG— CH2, EG—CH3 for P(MAA-g-EG) 2.5/5 in D20 at 25°C. As a control, three different stepsizes between the intensity levels were employed to ensure accurate reading ............................................................ Page 184 185 Eunll 55:12 Ennll low 53%: 44 Chapt Schent promo Copol methu Repre‘ I comp. comm oxyge nape. Chap Sunpi Chap Calib' mono nuns. 1151: Nmr' acuh persu Cone confi fitto Num acull Wei g TeaCl of [1 least PM;I ml Pp”) "Her are CI n \ Figure 1.1 Figure 1.2 Figure 2.1 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 LIST OF FIGURES Chapter 1 Schematic depiction of copolymer 1 under acidic (complex- promoting) conditions and under basic conditions. Copolymer 2 -not shown here- features additional lauryl methacrylate grafts incorporated into the backbone ............ Representation of a 1:1 MAAzEG repeating unit ratio complex formed between PMAA and PEG segments, each containing eight repeating units. Hydrogen, carbon and oxygen atoms are shown as white, black and gray, respectively .......................................................... Chapter 2 Simplified general water treatment scheme ..................... Chapter 4 Calibration curves prepared using commercially available monodisperse poly(acrylic acid)-sodium salt standards with molecular weights of 7,500, 16,000, 28,000, 62,900, 115,000, and 272,900 .............................................. Number average molecular weight Mn of poly(methacrylic acid) as a function of the weight percentage of sodium persulfate initiator fed into the reactor. The curves correspond to the upper and lower limit of the 99% confidence interval determined by nonlinear least squares fit to Equation (4.1) ................................................. Number average molecular weight MD of poly(methacrylic acid-g- ethylene glycol) copolymers as a function of the weight percentage of sodium persulfate initiator fed into the reactor. The curves correspond to the upper and lower limit of the 99% confidence interval determined by nonlinear least squares fit of Equation (4.1) to the data for pure PMAA (as in Figure 4.2) .......................................... 1H-NMR spectrum of P(MAA-g-EG) 5/10 from 0.8 to 4.2 ppm. The relative values for the integrated areas for the intensities of interest are also shown. Proton assignments are given in the text ................................................ xi Page 6 7 39 65 66 67 3111.5 MAA l TUnCllt‘ 111011012. offingl Chapu 35:15.1 8105.01 ' thcelo' hasthtl 11161.2: the IF. mode? to L’- i q Forku I onthcfl depen. also: @353 0 En 011010;: b) Tl, lunchE noml fund}1 [11116 (I lime Vfluc 5533 5.4 13511111 smnpj fullCIi (lCVl'd duun Specul Speci .5416 5.5 11)er Nfluu Ean56 Hydr 501L111 5?er 5'7 Hl'dr SOhul Figure 4.5 Figure 5.1 Figure 5.2 Figure 5.3 Figure 5.4 Figure 5.5 Figure 5.6 Figure 5.7 Page MAA to EG repeat unit ratio in the graft copolymer as a function of the MAA to EG repeat unit ratio in the monomer feed. The diagonal line corresponds to equality of the polymer and monomer composition ..................... 68 Chapter 5 a) Oscillation of scattered field vector is perpendicular to the electric field vector of the incident light. Emitted light has the same radiation frequency. b) Graphical representation of the angular distribution of the intensity of scattered light for vertically polarized incident light ........................................................ 96 For large particles the intensity of scattered light depends on the angle of observation. Therefore the angular dependence of the scattered light gives information about size of a macromolecular particle (R) ........................... 97 a) Erratic raw data of scattered light as detected by a photomuliplier tube. b) The normalized time correlation function le(t)l as a function of time. Initially this function is 1 (not normalized ). For very long times the correlation function decays to 0, (not normalized 2). Relaxation time of a correlation function, 1:, is the time required for the time correlation function to drop to 1/e of its initial value ................................................................. 98 Dynamic light scattering studies on angle dependence for sample P(MAA-g-EG) 2.5/5. The normalized correlation functions ln( le(t) 1-1) as a function of lag time show deviation from a straight line. This corresponds to a distribution of relaxation decay times. In this case, each species contributes its own exponential, according its specific diffusion coefficient ...................................... 99 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g—EG) 5/5 ........................... 100 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g-EG) 2.5/5 ........................ 101 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g-EG) 5/1 ........................... 102 xii 2,..53 PTOjCClCl F::e 5.9 2.7262 {5‘35 63 .21; . L. (D the nun copolynl correspo as given .‘vlodifie aeid-g-e molecul solution into the “1.9/0 01 Chapte Results the PIN W85 ft) solutiot solution the 013 phase 80111110 Rehuix by 111g adjuste solutio aqueot Relatil bl 10\ adlustt SOIutic aqUCOI Relatil by Po ratio c the a poll'm Solum Figure 5.8 Figure 5.9 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Projected aggregate sizes of P(MAA-g-EG) as a function of the number average molecular weight of the individual copolymers. If two modes are present, the open symbols correspond to the first mode for the same group of polymers as given in the legend .............................................. Modified pH-titration curves of different poly(methacrylic acid-g-ethylene glycol) copolymers with a number average molecular weight of about 25000 each in a 1 wt.% aqueous solution. The ethylene glycol grafting level incorporated into the poly(methacrylic acid) backbone are 2.5, 5, and 10 wt.% of total polymer mass ....................................... Chapter 6 Results of the reversibility test of emulsions stabilized by the P(MAA-g-EG) 5/ 10 copolymer. The initial emulsion was formed by mixing 33 vol.% of an aqueous polymer solution with 66.6 vol.% methyl laurate. The polymer solution had a concentration of 0.5 wt.%. The volume of the oil phase, emulsion phase and un-emulsified water phase are shown as a function of the adjusted polymer solution pH .......................................................... Relative amount of oil released from an emulsion stabilized by high molecular weight polymers as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution .................................................... Relative amount of oil released from an emulsion stabilized by low molecular weight polymers as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution .................................................... Relative amount of oil released from an emulsion stabilized by P(MAA-g-EG) copolymers with an MAAzEG repeat unit ratio of about 10:1as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The 'polymer concentration was 0.5 wt.% in the aqueous solution ............................................................... xiii Page 103 104 115 116 117 118 I Yrs-$6) J" " y. ‘ 14 T: I} "5‘51, 1.3:: 1.5 Poem ileum Relatiu by Pl.\l ratio ot‘ the act polyme solutior Chapltl mem tension Force 1‘ Interfa and 0.} titrated 0.25 to Interfa moleCL wool) dillere Interfe molee. COpol} ditlere Imierfa presen of 25 111181113 Presen of 10 HVQI-aE Figure 6.5 Figure 7.1 Figure 7.2 Figure 7.3 Figure 7.4 Figure 7.5 Figure 7.6 Figure 7.7 Relative amount of oil released from an emulsion stabilized by P(MAA-g-EG) copolymers with an MAAzEG repeat unit ratio of about 10:1as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution ............................................................... Chapter 7 Drawing of the main part of the DVT 10 drop volume tensiometer .......................................................... Force balance for the drop volume experiment ................. Interfacial tension measured for the system methyl laurate and 0.5 wt.% aqueous solution of P(MAA-g-EG) 5/5 at a titrated pH of 2 for five different flow rates ranging from 0.25 to 2 mL/hr ...................................................... Interfacial tension versus level of grafting for low molecular weight (ca. 17,000 g/mole) P(MAA-g-EG) copolymers in the system methyl laurate/water at four different pH levels .................................................. Interfacial tension versus level of grafiing for low molecular weight (ca. 120,000 g/mole) P(MAA-g-EG) copolymers in the system methyl laurate/water at four different pH levels .................................................. Interfacial tension in the system water/methyl laurate in the presence of P(MAA-g-EG) copolymers with an EG content of 2.5 wt.% at different solution pH versus the number average molecular weight of the copolymers ................... Interfacial tension in the system water/methyl laurate in the presence of P(MAA-g-EG) copolymers with an E0 content of 10 wt.% at different solution pH versus the number average molecular weight of the copolymers ................... xiv Page 119 131 132 133 134 135 136 137 3. E13 .4 oo Esra 8.3 Egnlo 5§e85 I. '1‘: .-‘ ":Lire 8.6 ll... ltrue 8,7 lezel dfiue nuflec Chapl Pnnci Exuat [hocu ‘Tho” “our of 30( The n to7oE Proce. ofopt steti laurat 2.5/ll 1116 21 pietur inere; H1510 meth: moth: 136) : that. H1310 uunh nunh EG) that. Figure 7.8 Figure 8.1 Figure 8.2 Figure 8.3 Figure 8.4 Figure 8.5 Figure 8.6 Figure 8.7 The effect Of concentration on the surface activity of three different P(MAA-g-EG) 5/10, 2.5/5 and 10/1 varying in molecular weight and level of grafting ......................... Chapter 8 Principle of Laser Scanning Confocal Microscopy ........... Exitation Spectrum of DiO (3,3 Dioctadecyloxacarbocyanine perchlorate) ...................... “Dio”-Fluorescence image of methyl laurate droplets in water as measured with the Zeiss LSM 210 at a resolution of 300 pixels per inch at an absolut magnification of 2000. The image width/height ratio was corrected from 512/512 to 768/512 using Adobe Photoshop 3.0 ......................... Procedure to identify the actual droplet size from a series of optical sections .................................................. Z-series of “Dio”-Fluorescence images through methyl laurate droplets stabilized in water with P(MAA-g-EG) 2.5/ 10 copolymers at a solution pH of 2 as measured with the Zeiss LSM 210. The stepsize was set to lmicron. The pictured are organized left to right and top to bOttom with increasing depth .................................................... Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g- EG) 2.5/10 measured after preparation, and 7, 14 days after that .................................................................... Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g- EG) 2.5/l measured after preparation, and 7, 14 days after that ................................................................... XV Page 138 151 152 153 154 155 156 157 Tamil Histtl mett met‘t EU) that. Fight 8.9 Hist metl metl EG) that 1%qu 8.10 Nun stab funt 5157113 9.2 A 1 Perm 3‘1 Euu94 A C) Eun95 7. g. T o b t1 Flute 9.5 Figure 8.8 Figure 8.9 Figure 8.10 Figure 9.1 Figure 9.2 Figure 9.3 Figure 9.4 Figure 9.5 Figure 9.6 Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g- EG) 5/5 measured after preparation, and 7, 14 days after that ................................................................... Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in 3 667/333 by volume methyl laurate/water emulsion stabilized with P(MAA—g- EG) 5/1 measured after preparation, and 7, 14 days after that ................................................................... Number average droplet diameter of methyl laurate stabilized in water by P(MAA-g-EG) copolymers as a function of time ..................................................... Chapter 9 Nuclear spin energy levels for a two spin system ............. A plot of the rate of growth of the magnetization vectors during the T1 recovery experiment of 0.125 g P(MAA-g- EG) 2.5/5 in 1.25 ml of D20 using a VARIAN VXR 500 spectrometer ......................................................... Signal Intensity versus pulse width for a set of pulse width in the range of 26 to 66 us. The 1:12 pulse width was calculated to be 11.8 us ............................................ Applied pulse sequence for 2D NOESY NMR experiment .......................................................... 2D lH-NMR contour plot of a solution of 0.125 g P(MAA- g-EG) 2.5/5 in 1.25ml D20 at a solution pH value of ~2. The level spacing represents the spacing relative intensity of successive contour levels and is set to be 1.27. The ratio between one intensity level and the next lowest is equal to the stepsize .......................................................... 2D lH--NMR contour plot of a solution of 0.125 g P(MAA- g-EG) 2.5/5 in 1.25ml D20 at a solution pH value of ~11. The level spacing represents the spacing relative intensity of successive contour levels and is set to be 1.11. The ratio between one intensity level and the next lowest is equal to the stepsize .......................................................... xvi Page 158 159 160 178 179 180 181 182 183 Chal nus 10.1 Pitt COP-V cons mon | valu Egut 10.2 Pyre COPt 000‘ I IUOTI salt: hgut103 Plo: , — _ nu. 10.4 lure 10.5 Sht Figure 10.1 Figure 10.2 Figure 10.3 Figure 10.4 Figure 10.5 Figure 10.6 Figure 10.7 Chapter 10 Pyrene excitation spectra for 20:1 P(MAA-g-EG) copolymer 2.5/0.5 solutions of various concentrations with constant pyrene concentration (6 x 10'8 M). Emission was monitored at 397 nm. The solution pH were titrated to a value of 7 ............................................................ Pyrene excitation spectra for 20:1 P(MAA-g-EG) copolymer 2.5/0.5 solutions of various concentrations with constant pyrene concentration (6 x 10'8 M). Emission was monitored at 397 nm. The solution pH were titrated to a value of 2 ............................................................ Plot of the pyrene excitation ratio (1333113333) as a function of pH for 10:1 P(MAA-g-EG) 5/0.5 copolymer solutions with 6 x 10'8 M pyrene. The emission was monitored at 397 nm. The polymer concentrations for each series were 0.004, 0.1 and 0.5 wt.%, respectively ........................... Plot of the pyrene excitation ratio (133/1333.3) as a function of pH for 10:1 P(MAA-g-EG) 5/0.5 and 5/10 copolymer solutions with 6 x 10'8 M pyrene. The emission was monitored at 397 nm. The polymer concentrations for each series was 0,5 wt.% ................................................ Pyrene emission Spectra for 20:1 P(MAA-g-EG) 2.5/0.5 copolymer solutions at pH 2. Excitation was at 322 nm and pyrene concentration in each sample was 6 x 10'8 M. The polymer concentration in the aqueous solution was varied between 0 and 0.5 wt.% .................................. Pyrene emission spectra for aqueous solutions of pyrene with added electrolytes. Excitation was at 322 nm and pyrene concentration in each sample was 6 x 10'8 M. . A plot of the pyrene emission ratio 1333/13725 (II/I3) as a function of pH for 10:1 copolymer 5/O.5 and 5/10 solutions with a pyrene concentration of 6 x 10'8 M. Excitation was at 322 nm. The pyrene emission ratio for titrated water is shown as a reference ............................................... xvii Page 202 203 204 205 206 207 208 Figure 10.8 Figure 10.9 figurelOlO Figure 10.8 Figure 10.9 Figure10.10 A plot of the pyrene emission ratio (excitation at 322 nm) versus concentration for 10:1 P(MAA-g-EG) copolymers 5/10and 5/0.5 ........................................................ A plot of the pyrene emission ratio (excitation at 322 nm) versus concentration for 20:1 P(MAA-g-EG) copolymers 2.5/10, 2.5/5 and 2.5/0.5 ........................................... Plot of critical aggregate concentration determined by the ratioing technique for the 10:1 and 20: MAAzEG copolymers as a function of their molecular weight ........... xviii Page 209 210 211 1.1. NI The adx on demand cou example, in the not controlled and new enterit 11: reversible . cluring ability “51¢- Resersi Wit oil by 211 I111. Other a! Mucous Phot. biochemical re Potential app} 55131111315 and bmkm an d 16 Chapter 1 INTRODUCTION 1.1. IMPORTANCE AND IMPACT OF REVERSIBLE EMULSIONS The advent of emulsions that may be repeatably formed, broken, and re-formed on demand could have a tremendous impact is on a wide variety of applications. For example, in the medical arena, reversible emulsifiers could lead to the development of new controlled release devices, new methods for solubilization of chemotherapy drugs, and new enteric coatings. In the more mundane area of cleaning and degreasing agents, the reversible emulsifiers could lead to advanced aqueous degreasers that provide the cleaning ability of emulsifying cleaners while minimizing the production of oily-water waste. Reversible emulsifiers could even make an impact in the realm of pumping heavy crude oil by allowing low viscosity, oil-in-water emulsions to be formed and broken at will. Other areas in which reversible emulsifiers could find application could include aqueous photoresists, agents for oil recovery or environmental remediation, and biochemical reactor systems, to name a few. And this is surely only a partial list of the potential applications of reversible emulsifiers. Once they are available, creative scientists and engineers are bound to find new uses for emulsions that can be formed, broken, and reformed on demand. 11 BLOCK-GI Block to are widely use hydrophobic blo extends into the include the eth} tndsion is acl polymeric mate unphipathic blt WNW phaut adStuIce from AS disc maO‘Omolecule 311513 from a ‘- bonding, ionic c(”lllllexes ma; erratum pol- M i”1120an 1351311 of lever 1“ this turn 35 a {€512 these 8113161113 I Winning an .I 1.2 BLOCK-CRAFT COPOLYMERS AS NOVEL REVERSIBLE EMULSIFIERS Block copolymers containing alternating hydrophilic and hydrophobic segments "2 In these polymers, the are widely used as stabilizers for aqueous emulsions. hydrophobic block is designed to penetrate into the oil phase, while the hydrophilic block extends into the aqueous phase. Prominent examples of block copolymer emulsifiers include the ethylene oxide/propylene oxide diblock and triblock systems. Stability of the emulsion is achieved via steric stabilization of the colloidal particles by absorbed polymeric materials. It has been found that the most effective steric stabilizers are amphipathic block or grafi copolymers. The stabilizing moieties that reach out into the dispersion phase have to be mutually repulsive in order to effectively keep the colloids at a distance from each other. As discussed in more detail in Chapter 2, complexation and aggregation of macromolecules in solution has attracted much attention.3 Interpolymer complexes may arise from a variety of interaction forces such as van der Waals attraction, hydrogen bonding, ionic or hydrophobic interactions.4’5’6‘7 It is noteworthy that the interpolymer complexes may possess unique properties that are distinct from those of the individual constituent polymers. The ability to control the formation and rupture of a complex is of great importance both from the scientific and practical View, and is imperative for the design of reversible emulsifiers in this project. In this dissertation we are primarily interested in interpolymer complexes that form as a result of hydrogen bonding between complementary pairs of repeat units. In these systems, each pair of complementary repeat units contains a proton donor group containing an electron-deficient hydrogen and a proton-accepting group containing a lone pair of clgctroDS Ltigsgcritet'ia 151 Sl‘stem. each PM no and “Ch huge- HydrOE acrylic acid) ha reversible emult In a pit approach for t spontaneously 2 tie graft Cs“ poly(methacryl topoluners, th backbone unde Studies reveale inch repeat un Biscuit the 13011111613, the hI’dIOphobic ( ctilttpletely h3- lhis new apphl Ollie emulsit‘ ml’Oll/e chem: 1 pair of electrons. One complex-forming pair of complementary polymers that meets these criteria is poly(methacrylic acid) (PMAA) and poly(ethylene glycol) (PEG). In this system, each PMAA repeat unit may serve as the proton donor by virtue of the carboxylic group, and each PEG repeat unit may serve as the proton acceptor by virtue of the ether linkage. Hydrogen-bonded complexes between PEG and PMAA (or the closely related d8’9’m’ll and will serve as the basis for the acrylic acid) have been extensively studie reversible emulsifiers in this dissertation. In a previous paper from our laboratory, Mathur et a1ll first reported a new approach for designing emulsifiers in which the hydrophobic blocks are formed spontaneously and reversibly by complexation of two hydrophilic segments of a comb- type graft copolymer. These polymeric emulsifiers contain a backbone of poly(methacrylic acid), and grafis of poly(ethylene glycol). In the comb-type grafi copolymers, the PEG grafts may form hydrogen-bonded complexes with the PMAA backbone under acidic conditions, but not under basic conditions. Molecular modeling studies revealed that the complexes could be formed in a one-to-one stoichiometry with each repeat unit of the PEG graft complexes with a repeat unit of the PMAA backbone. Because the complexes are considerably more hydrophobic than the constituent polymers, the emulsifiers form alternating blocks of hydrophilic (uncomplexed) and hydrophobic (complexed) segments when the complexes are formed, but are rendered completely hydrophilic when the complexes are broken. Unlike traditional emulsifiers, this new approach allows emulsions to be broken at will, with the hydrophobic segment of the emulsifier becoming completely water soluble in a reversible manner that does not involve chemical degradation. Grafi Cl oligolethylene copolymerizatic of the resulting (shown pictoria the constituent Ill? p01}Iner res 511108 the poly ill solvent 151 meletely hyc copolymer 51.311 11115 (11: Our laboratory More 91111113 dc“13nd. Hun initial studies SIlllllesis am complexatiOn. 311d appamus relationships 1 and 11) 10 Char 9398151, 51315 Graft copolymer emulsifiers with a poly(methacrylic acid) backbone and oligo(ethylene glycol) grafts may be synthesized using a one-step free radical copolymerization reaction as shown in Figure 1.1 Under acidic conditions the PEG grafis of the resulting copolymers may form 1:1 hydrogen-bonded complexes with the PMAA (shown pictorially in Figure 1.2), and the complex is considerably more hydrophobic than the constituent PMAA and PEG homopolymers.8"2"3 Therefore, in the complexed state, the polymer resembles a hydrophobic/hydrophilic multi-block copolymer, (see Figure 1). Since the polymer complexes can be reversibly disrupted and reformed by a change in pH, solvent type or temperature, the polymers may be reversibly switched between a completely hydrophilic grafi copolymer state to a hydrophobic - hydrophilic multi-block copolymer state. This dissertation builds upon the foundation provided the previous research from our laboratory in which it was demonstrated that polymer complexation could be used to produce emulsifying agents that allow emulsions to be formed, broken, and reformed on demand. However, the research reported in this thesis goes considerably beyond the initial studies from our laboratory and provides new information and guidelines for the synthesis and characterization of reversible emulsifiers based upon polymer complexation. The broad goals of this dissertation were: i) to develop a reaction scheme and apparatus to reliably and repeatably produce grafi copolymers and to identify the relationships between the reaction conditions and the resulting molecular architecture; and ii) to characterize the molecular structure, the emulsification properties (droplet size, capacity, stability, interfacial tension), and aqueous solution behavior (aggregation, wrjomiation) experimental n The litc description of of poljmer co purl are descr trough 10. 1 acid'lgraftedp solution behat' and potentiorr ”9011mm is Lfluction is d1 emulsions and 1‘2le Study 0 9' The hydro ”Sing Dinette Ellen in Chap conformation) of the copolymers. As the following chapters will show, a broad array of experimental methods was used to obtain the desired data. The literature review in Chapter 2 contains the motivation for this research and a description of previous work in the area of cleavable surfactants, as well as a discussion of polymer complexes and their behavior in aqueous solutions. The objectives of this work are described in Chapter 3, while the body of the work is contained in Chapter 4 through 10. The synthesis and characterization of the polymer system co-(polyacrylic acid)-grafied-poly(ethylene glycol) P(MAA-g-EG) is described in Chapter 4. The solution behavior of P(MAA-g-EG) copolymers is studied using dynamic light scattering and potentiometric titrations in Chapter 5. The emulsification behavior of these copolymers is presented in Chapter 6. A series of experiments on interfacial tension reduction is described in Chapter 7. The oil droplet size distribution of oil in water emulsions and their stability is analyzed in Chapter 8. The results of a two-dimensional NMR study of aqueous solution of P(MAA-g-EG) copolymers are presented in Chapter 9. The hydrophobicity of P(MAA-g-EG) copolymers in aqueous solution was probed using pyrene fluorescence in Chapter 10. Finally, conclusions and recommendations are given in Chapter 1 1. Basic Cor Q50 Hyc SChematic dt (5011111110113 all Basic Conditions Poly(ethy|ene glycol) grafts /i/\e \ WOODCOWOOOQOWOOOOOW Poly(methacrylic acid) backbone 00(1) Acidic Conditions Hydrophobic chain segments due to complexation “1.1 \g}........ // Ooooé 'OOOOODSS‘W Hydrophilic segments of /poly(methacrylic acid) backbone Figure 1.1 Schematic depiction of copolymer P(MAA-g-EG) under acidic (complex—promoting) conditions and undre basic conditions. Riltt ud P 330% Figure 1.2 Representation of a 1:1 MAAzEG repeating unit ratio complex formed between PMAA and PEG segments, each containing eight repeating units. Hydrogen, carbon and oxygen atoms are shown as white, black and gray, respectively. Piinna 1.. New York Laschews Beltturov. 1001-17 ( Antipina. 50811111. 5 Abe, K., Chen, H. Turin, N. Oi'ama. I Ol’ama. I Ohm. H. Mathur, Saito, S_ H’i’mker. 10. 11. 12. 13. 1.2 REFERENCES Piirma, 1., Polymeric Surfactants, Surfactant Science Series Vol. 2, Marcel Dekker, New York, 1992. Laschewsky A., Adv. Polym. Sci, 124, 1—86 (1995). Bekturov, E. A., Bimendina, L.A., Interpolymer Complexes, Adv. Polym. Sci, 43, 100-147 (1980). Antipina, A. D., Baranovskii, V. Yu., Papisov, I. M. Kabonov, V. A., Vysokomol. Soedin. Ser. A, A14, 941 (1972); Polym. Sci. USSR (Engl. Transl.) 14, 1047 (1972). Abe, K., Koide, M., Tsuchida, E., Macromolecules, 10, 1259 (1977). Chen, H. L., Morawetz, H., Macromolecules, 20, 474 (1982). Turro, N. J ., Arora, K. S., Polymer, 2 7, 783 (1986). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 471 (1987). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 1839 (1987). Ohno, H. Matsuda, H., Tsuchida, E., Makromol. Chem, 182, 2267 (1981). Mathur, A. M., Drescher, B., Scranton, A. B., Klier, J ., Nature 392, 367 (1998). Saito, S. and Sakamoto, T., Colloids & Surfaces, 23, 99-104 (1987). Hemker, D. J ., Garza, V., and Frank, C. W., Macromolecules, 23, 4411-4418 (1990). 2.1 N Must I emulsion proc (one notable which it is ad Instance, it is delivery! but the digestil'e unterials, gu Solution pH could find m- 501ubilizatiol. Olller and deglfiasi: luhllcants an emUlslfi' the Contains 8 Co. Chapter 2 BACKGROUND 2.1 NEED FOR REVERSIBILITY IN EMULSION TECHNOLOGY Most of the effort in emulsion technology has been focussed on stability of the emulsion products."2 The deliberate breaking of the emulsions has had much less study 3’4’5). There are many examples in (one notable exception is in the petroleum industry which it is advantageous to have a stable emulsion that may be broken on demand. For instance, it is desirable for many medical formulations to be stable at the time of oral delivery, but to be unstable (and therefore release the active drug) at a designated point in the digestive tract. This problem is addressed using environmentally responsive materials, such as hydrogels whose drug delivery characteristics are controlled by solution pH changes.6 In addition to these enteric formulations, reversible emulsions could find medical applications for new controlled release devices, and new methods for solubilization of chemotherapy drugs. Other potential applications for reversible emulsions fall in the area of cleaning and degreasing. In various industrial production processes, it is necessary to clean oils, lubricants and fuels from parts or surfaces, this is typically done using detergents that emulsify the organic phase in water. The effluent stream from these cleaning processes contains a colloidal mixture of the oil phase in water that ultimately must be disposed of. A reversible emulsifier would provide the cleaning ability of a detergent when the emulsion is formed but would allow the oil phase and the water phase to be recovered gparately whet recycled and ‘ separating the . dilution. as 30 raters no long The de- gtod example rater emulsio general schenn have dense ma an oil skimmr emillol’ed to i W floccular. that were ad. Wilght P01)'m alllomerates fl111211101] 0r 1 1110109931 tre celluloSe and on Streams m 1 agents are Us deftlulgifiers ; llkj'lbenmc separately when the emulsion is broken. This could allow one or both of the phases to be recycled, and would certainly reduce the volume of oily water waste. The value of separating the oil and water has been long recognized,7 especially since the principle of dilution, as accomplished by flushing polluted waters down a sewer or into territorial waters no longer complies with environmental laws,8 let alone moral principles.9 The de-emulsification process in a typical municipal sewage treatment plant is a good example to help visualize the problems associated with the separation of oil-in- water emulsions stabilized by common surfactants.7 A simplified flow chart of this general scheme is shown in Figure 2.1. The effluent from a municipal sewer system may have dense materials, which are separated in a settling basin coupled with a trash trap and an oil skimmer. After the gross trash and floating oil have been removed, flotation is employed to bring suspended solids and dispersed oil onto the surface. Over the years, new flocculants such as organic polymers have more and more replaced iron and alum that were added to provide flocculation.7 These flocculants, usually high-molecular weight polymers derived from ethylene, propylene and similar polymers tend to form agglomerates with dispersed oil drops, which may be filtered out more efficiently. Once filtration or flocculation has reduced the oxygen demand to reasonable levels, then biological treatment can remove most of the rest of the impurities, such as suspended cellulose and fibrous trash.lo In general, however, biological processes do not work well on streams containing oil.7 Hence, over the years, chemical or biological'1 demulsifying agents are used to reduce the residual oil fraction as much as possible. The biological demulsifiers are essentially small organisms that metabolize the surfactants such as linear 2 alkylbenzene sulphonates.l Most of the emulsion-breaking chemicals used in the 10 breaking of of amounts of 111 amounts of oi Basic streams really a speci: oil drops (acic rill react witl For example. aiding enoug destabilize th completed. ti biological 5).. COIlsldering 11 ”13le pres< Clearly rever Could gleatly To ill reWSible en llgttre 2.1.14 Promoting Cc from one am the Clean Vt" breaking of oil-in-water emulsions either adjust the pH or supply large ions.7 Large amounts of lime, alum, ferric chloride and clay are used to form flocs or adsorb small amounts of oil.7 Acidic waters may be neutralized with lime stone, lime or ammonia. Basic streams are treated with waste acids or sulfuric acid. None of these chemicals are really a specialized demulsifying chemical. They serve to neutralize the charges on the Oil drops (acids, bases) or help to clump the oil drops together (polymers). Materials that will react with the emulsifier to change their solubility will tend to break the emulsion. For example, an oil-in-water emulsion stabilized by sodium stearate can be broken by adding enough calcium ion to form sodium stearate, which will dissolve in the oil, thus 7 destabilize the emulsion. At last, after a final filtration or flotation step has been completed, the main flow is recycled to be treated by an activated sludge or other biological system. ‘3 This example showed how challenging a wastewater treatment is, considering that all kinds of different surfactants such as cationic, anionic and non-ionic may be present. A crucial step in the cleanup process is the demulsification process. Clearly reversible emulsifying agents whose emulsification ability is readily controlled could greatly simplify this process. To illustrate the dramatic reduction in oily water waste that can be realized using reversible emulsifiers, consider the aqueous degreasing process shown schematically in Figure 2.1.14 In this scheme and oily part or surfaced would be cleaned under emulsion- promoting conditions, then the emulsion would be broken and the oil and water separated from one another. Since the same emulsifier in this water phase can be activated again, the clean water (containing the emulsifier) may be used recycled and used to clean 11 mother part or since the water In a sin nods usually inilhits. As 01 point. Anotht heaty crude o crude oils and it difficult or 1 such as found 101551 1116 V1: alternative. "1 51162000 pp”- fululsion into that may Tang lhswater has Revers re316110113 SUI 5011161111168 PC 0”lilélnding n - 1 them and C l Pnpamive C} another part or surface. The environmental and commercial benefit would be enormous, since the water is used in a closed loop and is not discharged as a waste. In a similar fashion, other applications may be conceived.14 For instance, drilling muds, usually oil-in-water (O/W) emulsions, are used in order to lubricate and cool drillbits. As outlined before, the used drill mud needs to be disposed or recycled at some point. Another possible application might be the specific emulsification of water in heavy crude oil or bitumens with the help of such a reversible surfactant. Most heavy crude oils and bitumens have Viscosities of the order of thousands of centipoise, making it difficult or impossible to pump them through pipelines under difficult circumstances, such as found in Alaska. In order to avoid the costly approach of heating the pipelines to lower the Viscosity, oil-in-water emulsions may be pumped as a less expensive alternative. Typical transport emulsions consist of 70 wt.% oil and 30 wt.% water and 500-2000 ppm of a surfactant formulation.” This formulation lowers the viscosity of the emulsion into the range of 50-200 cP as compared to the Viscosities of heavy crude oils that may range between 10,000 to 100,000. In current practice, at the destination point, the water has to be separated from the emulsion using techniques as discussed above. Reversible emulsifiers would also be useful for facilitating heterogeneous reactions such as emulsion polymerizations. Polymerizations and reactions are sometimes performed in micelles or vesiclesls derived from surfactants because of the outstanding reactivity control. Micelles may mimic “tiny reactors”, which solubilize, orient and compartrnentalize reactants. However, the use of micellar media in preparative chemistry has been hindered by the difficulty in isolating products emulsified by the surfactants. Normal extraction procedures are generally precluded because of the 12 boredom of U emulsifiers. The pre would be very community to conditions. E) C leave WWW due lenreen the dhtl'uctible g; in 5011111011 01- deflHIction Of Since then, r various new t; Jaeger clear-able surf in lite SlTUCtL HOWCVW un formation of troublesome emulsions. Again this problem would be solved by reversible emulsifiers. The previous paragraph elucidated certain aspects why a real reversible surfactant would be very useful. It comes as no surprise that efforts have been made in the research community to design surfactants that are at least destructible under a certain set of conditions. Examples of these cleavable surfactants are given in the next section. 2.2 CLEAVABLE SURFACTANTS Cleavable or destructible surfactants are designed that lose their surface-active property due to various mechanisms. In general, there exists a destructible group between the hydrophobic and hydrophilic portions of cleavable surfactants. The destructible groups are labile under the influence of certain reactants such as acid or base in solution or under illumination. For the first time, in 1980, Cuomo et al.'6 reported the destruction of unsymmetrical disulfide surfactants with dithioerythritol at the S-S linkage. Since then, many researchers have devoted their efforts to prepare and characterize various new types of cleavable surfactants. Jaeger and collaborators”18 have investigated a variety of acid-sensitive cleavable surfactants based upon glycerol and containing a hydrophobic 1,3-dioxane ring in the structure. These surfactants are stable under neutral and alkaline conditions. However, under acidic conditions, they can be decomposed into non-surface—active species because their hydrophobic and hydrophilic groups are no longer linked through the acid-sensitive acetal bond. In the presence of 1 M hydrogen chloride, the 1,3-dioxane ring'9 is opened and an aldehyde and the water soluble glycerol-rest are formed. In that 13 manner.a\'ar1 the cleavage s udactant wht desuuctibility nfllimole con 11 HCI aqueo 1.4 millimole Kida e tithe Starting ltuboxyl-atoi 133% h}'droc alk)'lidene-2-. SOdlum mt glucollll‘dnos 110m The t 032 and 0.8o The s Elllconcqé“ those SUI’fac carbonyl C011 madam W. nTufted as 1 110m [he 111in manner, a variety of other surfactants bearing an 1,3-dioxane ring or a dioxolane ring20 as the cleavage site were developed. Similarly, Wang et 211.” reported a sulfonate type surfactant whose 1,3—dioxane ring decomposed in 2 M hydrogen chloride at 50°C. The destructibility rate of these surfactants is reported between 1.5 and 10 hours for a 10 millimole concentration of sulfonate-type surfactants with varying alkylchain lengths in 2 M HCl aqueous solution. The critical micelle concentration is reported between 2 and 7.4 millimole for these products. Kida et a1.22 synthesized three new surfactant series with N-acetyl-D-glucosamine as the starting material. The first type of surfactant, sodium methyl 4,6-O-alkylidene-2- (carboxyl-atomethylamino)-2-deoxy-D-glucopyranoside, exhibited a full decomposition in 2% hydrochloric acid after 30 hours. However the second surfactant, methyl 4,6-0- alkylidene-2-deoxy-2-(trimethylammonio)-D-glucopyranoside iodide, and the third, sodium methyl-2-acetamide-4,6-O-alkylidene-3-O-[ l -(carboxylato)-ethyl]-2-deoxy-D- glucopyranoside, showed a decomposition percentage of ca. 50% and 20% after 120 hours. The critical micelle concentrations for these compounds were reported as 0.45, 0.32 and 0.86 mol/l, respectively. The same author23 also synthesized new amido nonionic surfactants based upon gluconc-l-S-lactone, that features an acid-sensitive acetal bond. Under acidic conditions, these surfactants decompose into non-surface-active gluconamide derivatives and carbonyl compounds. In an acidic aqueous solution, circa 80 % and 50% of the present surfactant was cleaved at an adjusted solution pH 1 or pH 3, respectively. The CMC is reported as 16 millimoles, while the air-water interfacial tension is reduced to 29 mN/m from the initial (surfactant free) value of 72 mN/m. l4 [n a 0L reported the Inerlomethllc afiehydesrelc mmmm matinelianl charlie surl later. llle Sl; into the range Jaeger vesicles.27 Tl: Hynlyss real mfinxyfluorc ShOWCd poorl; Jaeger mm a clea burylslloxyk: " lfii Thelma W Complete M NaOD so Smdks Jaeger Elk UndEr In a new approach to develop nonionic acid-sensitive surfactants, Yue et al.24 reported the successful synthesis of a series of cleavable poly(ethylene glycol) monomethyl ether based surfactants having a noncyclical acetal linkage. The amount of aldehydes released during the cleavage process in a pH 3 buffer solution of sulfuric acid was determined with a gas chromatograph. The acid-catalyzed hydrolysis was complete in a time frame of circa 2 to 3 hours. Compared with corresponding cyclic acetal-linked cleavable surfactants,25 these noncyclic acetal-linked surfactants hydrolyze three times faster. The surface tension in the system air-aqueous solution of surfactants was lowered into the range of 24 to 34 mN/m for all candidates. Jaeger et a1.26 prepared double chain surfactants specially designed to form vesicles.” These cationic surfactants were either based on a bromo ketal or chlorosilane. Hyrolysis reactions in 1M hydrochloric acid were evaluated chromatographically with carboxyfluorescin as the entrapped agent. The cleavable natures of these products showed poorly, as it takes respectively 24 and 48 hours to hydrolyze them completely. al.28’29 also reported the first examples of destructible surfactants based Jaeger et upon a cleavable silicon-oxygen bond. The surfactant [2-(n-dodecylmethyl-tert- butylsiloxy)ethyl]trimethylammonium nitrate hydrolyzed with the catalysis of an acid or base. The hydrolysis of 0.02 M of that surfactant in a solution of DC] in D20 at a pH -0.5 was complete after a maximum of 2 hours. The hydrolysis of the same amount in a 0.2 M NaOD solution of the surfactant was about 65% complete, monitored by 1NMR studies. 30.3I Jaeger and collaborators presented an example of cleavable surfactants that is stable under acidic conditions but may be hydrolyzed with a base. In a series of 15 experiments. Dodecylpheny Dodeeylpheny respectively. West z nimethylamn: ll DCl. llo dllierent. 11 l rapidly in 0.1 Photo though cleax Ween the l The photolal Working con PhOlOdeStruc lilmZ-und; each exhibit . lamp leads t lllt DOlar Sl p“fill-limes \ DESI; hllephlllc- Wh‘merizal; experiments, the authors showed that base-catalyzed elimination of [4-[2-[(4- Dodecylphenyl)sulfonyl]ethoxy]phenyl]triammonium nitrate and potassium 4-[2-[(4- Dodecylphenyl)sulfonyl]ethoxy]-benzenesulfonate gives vinyl sulfone and phenols, respectively. West and co-authors32 reported another base-labile surfactant, (2-acetyldecyl)- trimethylammonium iodide, that proved to be stable for extended periods of time in 0.01 M DCl. However, in neutral and basic media, the compound's behavior was quite different. It decomposed slowly in water, more rapidly in 0.1 M NaHCO; (ca. 1 day), rapidly in 0.1 M Na2CO3 (ca.90 min.) and completely within 5 minutes in 0.1 NaOD. Photochemistry has also been used to cause the disruptions of amphiphiles through cleavable linkages.33 Azo groups are useful candidates as the photolabile link between the hydrophilic headgroups and hydrocarbon tails of the surfactant molecule.“35 The photolabile group has to be reasonably stable under the heat and light of normal working conditions. Dunkin and collaborators36 have synthesized a series of such photodestructible surfactants as sodium 4-alkylphenylazosulfonates and sodium 4-(2- cyano-2-undecylazo)benzoate. These surfactants were surface active in aqueous solution, each exhibiting a distinct micelle concentration. UV-radiation with the use of a mercury lamp leads to a complete photoscission in the case of the azosulfonates. Subsequently, the polar sulfonate group stays in solution while the hydrophobic phenol group precipitates out. Destabilization of the organized architecture by disruption of the surfactants hydrophilic-hydrophobic balance has also been achieved by redox reactions and by polymerization. Saji and co-workers37 reported nonionic surfactants synthesized from 4- 16 [l-lheX)'lphel reduction. 1: potential ele: dispersions r azobenzene-t such as com reduce the St surfactants an Franlu {liposomesl Stabilized wi Ph0l0p01}'me destabilizario 23C 23'] lntrod Com much anentl 3nd Smaller poll-Illa Syg rall . Ordered hell [(4-hexylphenyl)azo]phenol, which lose their hydrophilic-hydrophobic balance upon reduction. In acidic solution, the nitrogen-nitrogen bond is cleaved with controlled- potential electrolysis that leads to the formation of aniline derivatives. Aqueous dispersions of hydrophobic pigments that are solubilized with the help of these azobenzene-based surfactants may be spread on base metal surfaces. Metal substrates, such as copper and stainless steel, having a lower standard potential than the surfactant, reduce the surfactant to an aniline derivative. This hampers the performance of the surfactants and that leads to the deposition of the pigments on the metal substrate. Frankel et al.38 reported polymerizations between lipids that form bilayer vesicles (liposomes). The authors show that vesicles of phosphatidylethanolamine (PE) can be stabilized with other lipids such as phosphatidylcholines (PC). As the PE was photopolymerized, the PE and the PC phase separated which leads to vesicle destabilization detectable by the release of the aqueous contents. 2.3 COMPLEXATION OF OLIGOMERS WITH COMPLEMENTARY POLYMERS 2.3.1 Introduction Complexation and aggregation between macromolecules in solution has attracted much attention.39 Macromolecular complexes between high molecular weight polymers and smaller complementary oligomers are important in many biological and synthetic polymer systems. The equilibrium between highly ordered helical conformations and random coils is prevalent in many biological proteins and nucleic acids.40 The highly ordered helical state is stabilized in solution by the formation of intramolecular hydrogen l7 bonds.4O X-ray stranded helical interactions. autho 540.41.42.43; Complexation h“ for the developr controlled by ch some control or complementary sensitive mate energy, In three “Ell as hydrc bEN‘een a p] Elcceptor gro- chamlenzec dim“)? on lawn-“018m, the effect of are Characre der Waals ; comprexm Van der we bonds.40 X-ray diffraction studies40 have shown that oligonucleotides can form double- stranded helical complexes with polynucleotides due to similar hydrogen bonding interactions. To elucidate the nature of this complexation equilibrium, several authors”4 "42’43‘44 have analyzed these systems both experimentally and theoretically. Complexation between synthetic oligomers and complementary polymers may be used for the development of new materials.45’46’47'48’49 The complexation equilibrium may be controlled by changing the conditions such as temperature and pH, and therefore provides some control over the material properties. The complexation of synthetic oligomers with complementary polymers may be exploited for such applications as environmentally “’45 and materials which convert chemically energy to mechanical sensitive materials energy. In these systems, complexation may occur as a result of hydrogen bonding, as well as hydrophobic, ionic, or van der Waals interactions. Hydrogen bonding occurs between a proton donor group containing an electron deficient proton, and a proton acceptor group, which contains a lone pair of electrons.50 A hydrogen bond may be characterized as a proton shared by two lone electron pairs. Hydrogen bonds are distinctly directional and specific, and are more localized than any other type of weak intermolecular interaction.49 The hydrophilic interaction in aqueous solutions arises from the effect of the hydrophobic solute on the structure of water.51 Hydrophobic interactions are characterized by an increase in strength with an increase in temperature. Finally, van der Waals attractions arise from dipole-induced dipole or dispersion interactions. The complexation between isotactic and syndiotactic poly(methyl methacrylate) occurs due to van der Waals interactions.52 In polymer systems, stable macromolecular complexes may 18 form even if natureof the In th poly(ethylenu reviewed. Pl while PhlA; bonded corr strength as tl interactions. complexes, COrrlplement 232 Exp. Bail. Wll'limethar [he Viscosir COml‘OSitior We solur from 3.8 to conditions l fofined Wit compreXeS (termed at form even if the interaction energy per segment is relatively small due to the cooperative nature of the complexation process. In this section, the literature pertinent to the complexation of oligomeric poly(ethylene glycol) (PEG) with polymeric poly(methacrylic acid) (PMAA) will be reviewed. PEG is a Lewis base by virtue of the lone pair of electrons in the ether linkage, while PMAA is a Lewis acid. Therefore, the two species are able to form hydrogen- bonded complexes with one another. Furthermore, since the complexes increase in strength as the temperature is increased, they are believed to be stabilized by hydrophobic interactions. Topics in this review include experimental investigations of PMAAzPEG complexes, and theoretical modeling of the complexation of synthetic oligomers with complementary polymers. 2.3.2 Experimental Investigation of PMAA:PEG Complexes Bailey et a1.53 studied the association of poly(acrylic acid) (PAA) and poly(methacrylic acid) (PMAA) with PEG using viscometry. These authors measured the viscosity of aqueous solutions of PEG and FAA as functions of solution pH, composition, and temperature. The formation of a complex was indicated by an increase in the solution viscosity. The viscosity was found to decrease as the pH was increased from 3.8 to 7.0 and plateaued between pH 7 and 12. These results indicated that acidic conditions favored complex formation. The authors concluded that PAA/PEG complexes formed with a 2:3 carboxylate:ether (COOH/EO) molar ratio, while PMAA/PEG complexes formed in a 1:1 COOH:EO ratio because maxima in solution viscosity occurred at these compositions. Finally, the PAA:PEG complex strength decreased in 19 strength with 16! with temperature stabilized perhar exhibited a muc complexes more Antipina riscometry and mired with PE( relatively short no change in so 3000 and 3000 acidic P01ycar1: disSOCiate. PEr SOlution pH. A when the PEG PrOnounCed [he melem Wit} the “Operative than lenglh or meaSUIEmems PAA:PEG 53'5“ [haw-a e PEG and MW strength with temperature, while the strength of the PMAA/PEG complexes increased with temperature. This result indicated that the PMAA complexes were hydrophobically stabilized, perhaps due to the (Jr-methyl group. Furthermore, complexes involving PMAA exhibited a much larger effect on viscosity than those of PAA, indicating that PMAA complexes more strongly with PEG than PAA. Antipina et a1.54 studied the complexation of PAA and PMAA with PEG using viscometry and potentiometry. High molecular weight PAA and PMAA (100,000) were mixed with PEG chains of various molecular weights in dilute solution. Addition of relatively short PEG chains of molecular weight 1000 to solutions of PMAA resulted in no change in solution pH. However, addition of longer PEG chains of molecular weight 2000 and 3000 led to gradual increases in solution pH, indicating that a fraction of the acidic polycarboxylic acid protons were bound in a complex and were unable to dissociate. PEG of molecular weight greater than 6000 lead to a rapid increase in the solution pH. Addition of PEG to PAA, on the other hand, lead to an increase in pH only when the PEG molecular weight was greater than 6000, and the pH increases were less pronounced than with PMAA complexes. Again, PMAA was found to form stronger complexes with PEG. The chain length dependence of the complex stability illustrated the cooperative nature of the complexation process. The authors identified a critical PEG chain length of 2000 for PMAA complexes and 6000 for PAA complexes. Viscosity measurements agreed with those of Bailey12 except that viscosity minima for the PAA:PEG system were observed at 1:1 COOH2EO ratio instead of a 2:3 ratio. Ikawa et al55 studied the nature of the hydrophobic stabilization in complexes of PEG and poly(carboxylic acids) using viscomety, potentiometry, and turbidimetry. From 20 these studies.t PlllA system hrdrophobics PM. Furthe from the dept PlliA syster from water It system. Othe hlilh pH. but dissociation; 058d; molecular vr Wrentiometr mOlecular “ molecular V dependenCe that in Wate Cletermine 1 PM di complexari tmpemur Show“ to aP‘Efll'sis is these studies, they found an abrupt transition in complex yield as a function of pH for PMAA systems, and only a gradual transition for PAA. They attributed this difference to hydrophobic stabilization through the (Jr-methyl group that exists for PMAA but not for PAA. Further evidence of hydrophobic stabilization of PMAAzPEG complexes arose from the dependence of critical chain length of PEG on solvent concentration. For PMAA systems, the critical chain length of PEG increased when solvent was changed from water to a water/methanol mixture, while it remained constant for the PAA:PEG system. Other conclusions drawn by these authors are that the complex is not formed at high pH, but only at low pH, the complex stoichiometry is 1:1, and the degree of dissociation plays an important role in complex formation. Osada and collaboratorss‘5‘57'58 investigated the effects of temperature and molecular weight in the complexation of PMAA and PEG using viscometry and potentiometry. A critical PEG molecular weight of 1000 was identified, since PEG of molecular weight lower than 1000 did not form complexes with PMAA, and PEG with molecular weight 3000 and greater gave substantial complexation. The temperature dependence of complexation was found to be opposite in ethanol/water mixtures from that in water, due to disruption of hydrophobic interactions. Potentiometry was used to determine the bound fraction of MAA repeating units, using the assumption that the PMAA dissociation constant was independent of the degree of ionization or complexation. They found that the bound fraction was dependent upon chain length, temperature, and solvent. Since the dissociation constant of polymeric acids has been shown to be dependent on the degree of ionization and complexation,59 therefore the analysis is only approximate. 21 ln a seri spectroscopy inn and PAA. By ex the intramoleculr tagged PEG pro Pth to PEG tagged case. ind Fully tagged PE bl" a gradual dl't PM was adde allEminent of P, intermolecular m0nomer ratio Cycle formatio Frank Cfiects Of 501‘ intensity in a and complex SolurjOnS‘ ag dependeme mOIar ratio : viscomemc In a series of papers,""""'62 Frank and collaborators reported fluorescence spectroscopy investigations of the complexation of pyrene end-labeled PEG with PMAA and PAA. By examining the excimer to monomer intensity ratio of sparsely tagged PEG, the intramolecular end-to-end contact was characterized, while experiments with fully tagged PEG provided information about intra and intermolecular contacts. Addition of PMAA to PEG greatly reduced the excimer to monomer intensity ratio in the sparsely tagged case, indicating a decrease in the intramolecular mobility due to complexation. Fully tagged PEG exhibited a sharp rise in excimer to monomer intensity ratio followed by a gradual drop at a MAAzEG ratio of 2:1. The rise and drop was not observed when PAA was added. The initial increase in excimer to monomer ratio was explained by alignment of PEG along the PMAA chain through hydrogen bonds, leading to increased intermolecular excimer contact at chain ends. The decrease in intramolecular excimer to monomer ratio was explained by loss of chain flexibility and disruption of intramolecular cycle formation. Frank and collaborator358‘59’60 used the fluorescence technique to investigate the effects of solvent and molecular weight. The observance of high intramolecular excimer intensity in a methanol/water solvent indicated that intramolecular cycles remained intact and complexation was weaker than in water. Similar results were observed in basic solutions, again indicating the lack of complexation. The fluorescence results showed a dependence upon molecular weight. Intermolecular excimer formation exhibited at 1:1 molar ratio for PAA molecular weight above 4800, while for PAA of molecular weight 1850, the increases were much slower with a plateau at PAA:PEG of 3:1. In contrast to viscometric results discussed earlier, these fluorescence results indicated complex 22 formation in th Further investi inlonnation on lonned pyrene- groups. Deere indicated that tl the excimer-ion Fluorom were reported interaity Chang Wronment, c. addition of pp 611VProminent l2 Mintensity weigh 0f 24,1 fraction of b manUllOnS damn chrom PMAA PM. cortrpltexafion formation in the PEGzPAA system even when the PEG molecular weight is 1850. Further investigation of the PMAA/pyrene end-labeled PEG complex provided information on the hydrophobic interaction. Excimer emissions were attributed to pre- formed pyrene-pyrene pairs arising from hydrophobic interactions between pyrene groups. Decreases of intermolecular excimer formation with addition of PMAA indicated that the compact structure of the PMAAzPEG complex caused dissociation of the excimer-forming pyrene complex. Fluorometric studies of equilibrium and kinetics of PAA association with PEG were reported by Chen and Morawetz."3 Dansyl chromophores, which exhibit an intensity change and spectral shift when moved from a hydrophilic to hydrophobic environment, were attached to the PAA backbone. Large peak intensity changes upon addition of PEG to dansylated PAA revealed that the chromophores end up in an environment largely devoid of water. Complex formation as measured by fluorescence peak intensity depended strongly on PEG chain length. At pH 3.7, PEG with a molecular weight of 24,000 gave a sharp increase in intensity, whereas PEG 3400 did not. The fi‘action of bound PAA was measured using a model of linearly additive intensity contributions from complexed and uncomplexed PMAA. The authors found that the dansyl chromophore could not be readily used to study complex formation in dansylated PMAA. PMAA itself forms a compact globular structure in solution, so that subsequent complexation gives little or no spectral intensity change. Complexes of poly(methacrylic acid) (PMAA) with free (ungrafted) poly(ethylene glycol) (PEG) have been well characterized in aqueous solution using a wide variety of physical, spectroscopic, and compositional techniques.64’65’66’67’68 It has 23 been concluded stabilized by h) cooperative in r. Experiments pe molecular weig addition invest: to ethylene gl} neutralization d stabilization, ll’lr ln previr incoI’Olymers l Characterized}! ethibit Comple addition of be differences as [Ogether mm molecular w e by Cl'Cling (h. In addition, . prepared Sim comm, the Complexes. Whilibrmm r been concluded that these polymers form hydrogen-bonded complexes which are stabilized by hydrophobic interactions in water. Since polymer complexes are generally cooperative in nature, the complex stability increases with increasing molecular weight. Experiments performed with PMAA of high molecular weight revealed a critical PEG molecular weight (2000 g/mol) below which complexation does not occur.69‘70 In addition, investigators have shown that complexes form in a 1:1 methacrylic acid (MAA) to ethylene glycol (EG) repeat unit ratio, and that the complexes can be broken by neutralization of the PMAA, or by the addition of alcohols. Due to the hydrophobic stabilization, the strength of the complex increases with increasing temperature. In previous work done by Peppas and collaborators, the complexation equilibrium in copolymers containing PMAA backbone chains and complementary PEG grafts was characterized."72 The grafted systems, like the ungrafted systems discussed above, exhibit complexes with a 1:1 repeat unit ratio, and the complexes are disrupted by the addition of bases or alcohols. However, the grafted systems exhibit some important differences as well. Covalently linking the complementary PEG and PMAA constituents together removes the critical chain length effect (complexes were formed for all molecular weights) and the polymer complexes can be repeatably broken and reformed by cycling the system between complex-promoting and complex-breaking conditions.73 In addition, the graft copolymers allow non-stoichiometric complexes to be readily prepared simply by using nonstoichiometric repeat unit ratios during synthesis. In contrast, the free (ungrafted) systems tend to rearrange and form only stoichiometric complexes. An illustrative statistical mechanical description of the complexation equilibrium revealed that there is an entropic driving force for loop formation, especially 24 for long grafts. ‘ addition. the t ungrafted (free ‘r blocks form spr ola graft copr Characterized P.‘ al/‘Oh’lrnethatcrg Under acidic co Pllhh (shom hldIOphobic m melexed star ls“ Flame r.r bl“ achange in between a Cor multi-brmk or Two mullsifrcatior designed to Condjrims. l Cothplexing lacldlc) C0nd for long grafts, and that several small loops are favored over a single large loop.74 In addition, the theoretical studies revealed that the conformational averages for the ungrafied (free) case asymptotically approach those for the grafted case as the segmental binding free energy, polymer concentration, and graft length are increased. Mathur et al.75 reported novel polymeric emulsifiers in which the hydrophobic blocks form spontaneously and reversibly by complexation of two hydrophilic segments of a graft copolymer. To illustrate this general approach the authors selected the well- characterized PMAA — PEG complexation system. Graft copolymers were synthesized with a poly(methacrylic acid) backbone and oligo(ethylene glycol) grafts, as shown in Figure 1.1. Under acidic conditions the PEG grafts may form 1:1 hydrogen-bonded complexes with the PMAA (shown pictorially in Figure 1.2), and the complex is considerably more hydrophobic than the constituent PMAA and PEG homopolymers.76’77’ Therefore, in the complexed state, the polymer resembles a hydrophobic/hydrophilic multi-block copolymer, (see Figure 1.1). Since the polymer complexes can be reversibly disrupted and reformed by a change in pH, solvent type or temperature, the polymers may be reversibly switched between a completely hydrophilic graft copolymer state to a hydrophobic - hydrophilic multi-block copolymer state. Two classes of copolymers were designed to exhibit dramatically different emulsification properties in response to changes in pH. The first class of polymers were designed to produce stable emulsions under acidic conditions, but not under basic conditions. The strategy for these copolymers was to use non-stoichiometric ratios of the complexing repeat units to ensure an amphipathic character under complex-promoting (acidic) conditions. Specifically, representative copolymers were synthesized with MAA to 25 EG rcpeat unit: Under compler hydrophobic se may extend (or The se basic conditior stoichiometric completely hy Moreover. lor the copolymei dimipted). ' exPected to p and PEG gr. ch83 0f COp( addition of 1; hl'dmphobrc As d undi‘r acidic We basic . of lhe emUlsj Roofs Alrh. eInulsified u Copolymer ( EG repeat units ratios of 10:1 (copolymer 1a) and 20:1 (copolymer 1b) (see Figure 1.1). Under complex-promoting conditions, this molecular architecture leads to relatively short hydrophobic segments (~ 20 repeating units), and much longer hydrophilic segments which may extend (or loop) into the aqueous phase to provide steric stabilization. The second class of copolymers was designed to produce stable emulsions under basic conditions, but not under acidic conditions. In this case, block-graft copolymers with stoichiometric ratios of the MAA and EG repeat units were used to render the polymer completely hydrophobic (water insoluble) under acidic (complex-promoting) conditions. Moreover, long-chain alkyl grafts were covalently bonded to the PMAA backbone so that the copolymer assumes an amphipathic nature under basic conditions (when the complex is disrupted). Therefore, under basic conditions the pendent hydrophobic segments are expected to partition onto the oil-water interface while the hydrophilic PMAA backbone and PEG grafts provide electrostatic and steric stabilization. Accordingly, the second class of copolymer was synthesized using a 1:1 MAA to E0 repeat unit ratio, with the addition of lauryl methacrylate (LM) (10 mol.% of the MAA repeat units) to impart the hydrophobic grafts to the final polymer. As designed, the PMAA-g-PEG copolymers (1a and 1b) form stable emulsions under acidic conditions (100% of the oil is emulsified), but do not form stable emulsions under basic conditions. Molecular architecture plays an important role in the performance of the emulsifiers since the percent oil emulsified depends upon the MAA to EG repeat unit ratio.75 Although both polymer la and 1b exhibit the same general trend in which the oil is emulsified under acidic conditions but not under basic conditions, the 20:1 MAA to EG copolymer (1b) exhibits the sharp decrease in emulsification capacity at a noticeably lower 26 pll than the 10 hydrophilic (un lower degree < interface as the variations in II emulsifier. ln cont vith 100% of conditions. T objectives \vhe Performance ft- hmopolwner lads ‘0 no st; “able indefini im901181“ role In cor ht‘dmpmrrc c addition‘ the e ratio of back cornonomerS fomted and b .‘lb‘l0r{§m_,.er~ 1h N . OOIandl73 S pH than the 10:1 (la) system. This general trend likely arises from the greater number of hydrophilic (uncomplexed) MAA repeat units in the 20:1 copolymer (lb). Therefore, a lower degree of dissociation is required to remove the copolymer from the oil-water interface as the MAA to EG repeat unit ratio is increased. This result suggests that simple variations in the molecular architecture may be used to tailor the transition pH of the emulsifier. In contrast, the PMAA-g-PEG-co-LM copolymer (2) exhibits the opposite trend with 100% of the oil emulsified under basic conditions, and 0% emulsified under acidic conditions. Therefore the general emulsification properties are in agreement with our objectives when we designed the molecular structures. Comparison of the emulsification performance for the copolymers Figure 1.1a, 1.1b, and 1.2 with the results for the PMAA homopolymer reveals the pivotal role of the reversible hydrophobes. The homopolymer leads to no stable emulsions while the block/graft copolymers form emulsions that were stable indefinitely. For the first class of copolymers, the molecular architecture plays an important role in the performance of the emulsifiers. In conclusion, reversible hydrophobes formed by the complexation of two hydrophilic constituents may be used to design reversible emulsifying agents. In addition, the emulsification behavior of these copolymers may be controlled by varying the ratio of backbone acid to ethylene glycol repeat units and by adding hydrophobic comonomers such as alkyl methacrylates. These copolymers allow emulsions to be formed and broken on demand, and could therefore find a wide range of application. Moreover, this approach may yield superior emulsifiers since theoretical work by Noolandi78 suggests that multi-block copolymers may be more efficient as 27 compatibilizers into the bulk p extended to a structures. l 233 Theoryr Statistic are more comp changes in botl must be consic freedom afforr the [fartslationr mm in s; the free (uncor degrees of f” oligomer mol.- mns and loo the complem Consecmive c which the firs uncomprexfed‘ Kaban the compatibilizers than di- or triblock copolymers since less of the material would be lost into the bulk phase as micelles or mesophases. Finally, this general approach could be extended to a wide variety of other complexing systems that result in amphipathic StI'UCtlll‘C S . 2.3.3 Theory of OIigomer:Polymer Complexation Statistical thermodynamic descriptions of the complexation of macromolecules are more complicated than those for small molecules. For macromolecular complexation, changes in both the conformational and configurational contributions to the chain entropy must be considered. The conformational contribution arises from the internal degrees of freedom afforded by bond rotation, while the configurational contribution results from the translational degrees of freedom determined by the number of ways the chains may be arranged in space. A polymer chain exhibits more conformations and configuration in the free (uncomplexed) state than in the complexed state since rotational and translational degrees of freedom are lost upon complexation. The conformation of the complexed oligomer molecule may be described by a combination of three possible structures: tails, trains and loops. A tail is a sequence of repeating units which is bound (complexed) to the complementary polymer molecule only at one end; a train is a sequence of consecutive complexed repeating units; and a loop is a sequence of repeating units in which the first and the last end are complexed to the polymer, while all the others are uncomplexed. Kabanov and collaborators52 reported a statistical thermodynamic description of the complexation of oligomers with complementary polymers. The total free energy of 28 complexation v between complt upon complex; from the Specit~ number of pest entirely entrop complexation. that is, they rt oligomer conti model agreed r the equilibriur complertes fon- 2-4 RESP As shot Wtalemry boo fimher W h 3” hl'dmphirr moluble hem P 01m. ettvironmem }_ including C complexation was divided into contributions, one arising from the specific interactions between complexing functional groups and a second arising from configurational changes upon complexation. The first contribution contained an enthalpic component arising from the specific interaction energy, and an entropic component due to the change in the number of possible configurations upon complexation. The second contribution was entirely entropic, and could be calculated from configurational states before and after complexation. The authors assumed that the oligomers bond in an all-or-nothing fashion, that is, they neglected the possibility of loops and tails and assumed that only one oligomer conformation was possible in the complexed state. Calculations based on this model agreed at least qualitatively with experimentally observed trends. For example, the equilibrium bound fraction that depended strongly upon chain length and stable complexes formed even with weak segmental interactions. 2.4 RESPONSIVE POLY(METHACRYLIC ACID) ETHYLENE GLYCOL HYDROGEL NETWORKS As shown in the previous section, various aspects of the complexation of free (not covalently bound) PMAA or PAA with PEG have been studied. Based on these studies, further work has been focused on crosslinked “hydrogel” systems. In general, hydrogels are hydrophilic polymer networks, capable of imbibing large amounts of water, yet insoluble because of the presence of crosslinks or entanglement.79 Polymeric hydrogels which show swelling behavior sensitive to the surrounding environment have been studied as promising candidates for a variety of applications, . . . so including controlled release devrces, membranes, and sensor components. 29 Polyelectrol variation in poly(acryla changes in factors sucl the swelling applied on poly(vinyl Hyr presence 0 membrane; oxidase w. extended E the enzym COllapge‘ C Polyelectrolyte gels are typical candidates of hydrogels which demonstrate swelling variation in response to changes in pH, while hydrophobic polymeric associates such as poly(acrylarnide)81 and poly(isopropylacrylamide)82 exhibit transitions as a result of changes in temperature. Apart from effects caused by solution pH or temperature, other factors such as presence of an electric field or enzymatic reaction products may change 83’“ reported that an the swelling behavior of polymeric hydrogels. Hirose and coworkers applied oscillating electric field was in resonance with a deformation pattern of a poly(vinyl alcohol)-poly(sodium acrylate) composite hydrogel. Hydrogels have been found to be very useful to control drug release in the presence of enzymatic reactions such as glucose.85 Ito et al.86 tested porous cellulose membranes that were made pH sensitive with grafted PAA chains. The enzyme glucose oxidase was also bonded to the cellulose. At neutral pH, the PAA chains were fully extended and blocked the pores in the cellulose membrane. Added glucose reacted with the enzyme to produce gluconic acid which protonated the PAA and caused the chains to collapse, opening up the pores for drug delivery. In the system PMAA and PEG, the polyelectrolye and the glycol may be crosslinked to form pH-responsive networks. Here, the complex of poly(methacrylic acid) (PMAA) and poly(ethylene glycol) (PEG) under acidic conditions is considerably less hydrophilic than the individual polymers. Therefore a hydrogel containing a backbone of PMAA and crosslinking grafts of PEG exhibits a relatively low degree of swelling under complex-promoting conditions (low pH when the acid is protonated), and a high degree of swelling when the complex is broken (high pH when the acid is neutralized). 30 lhe r polylmethacry collaborators. perfonnance 1 changes in pl complex-tom reversibly con complex was titrating the ad 0“ Changes 0‘ Composition Ot‘erhauser l P0ll’lmethacr preview), n the CIhllene The equilibrium swelling properties of self-associating networks of poly(methacrylic acid-g-ethylene glycol) have been investigated by Peppas and collaborators. For the first time, Klier et a1.” reported the synthesis and swelling performance of poly(methacrylic acid-g-ethylene glycol) networks with respect to changes in pH and temperature. The authors showed that by covalently linking the complex-forming constituents in a crosslinked graft copolymer, the swelling could be reversibly controlled in accordance with the state of the complexation. In this paper, the complex was formed when the acid was protonated (low pH) and could be broken by titrating the acid to a high solution pH. The dependence of equilibrium network swelling on changes of such variables such as external pH, temperature, as well as the network 1.88 performed Nuclear composition and structure were considered. Scranton et a Overhauser Effect (NOE) studies to show complexation between dilute solutions of poly(methacrylic acid) and poly(ethylene glycol) at lower PEG molecular weights than previously reported. NMR relaxation time studies and NOE experiments revealed that the ethylene groups of PEG and the (Jr-methyl groups of the PMAA align in close proximity, apparently arising from hydrophobic interactions. Peppas and Klier reported dynamic swelling studies on P(MAA-g-EG) hydrogels which showed that the swelling dynamics were lowest in copolymer networks with an equimolar ratio of ethylene oxide to methacrylic acid.89 Solute release studies to investigate the effect of pH on release behavior showed lowest solute release rates under complex-fonning conditions, thereby confirming the relationship between complexation, swelling and solute permeability. In order to explain the increase in complexation of oligomers grafted to a complementary polymer (as compared to complexation of two ungrafted complementary 31 chains). 3 get Recently, Bel membranes a" dynamic svveli controlled by t deviation froni oscillatory pli selling or n resulted in lot retoned thatt media due Ii demonstrated Significant la. the Smallest C0mpleued 3, 0f magnitude than proxy“ the gels. Final Plblas‘g‘E COMl'mEr chains), a general statistical mechanics framework was reported by Scranton et a1. Recently, Bell and Peppa890’9' studied the pH-sensitive swelling of P(MAA-g-EG) membranes and demonstrated the effect of PEG graft length on the equilibrium and dynamic swelling properties. These investigators concluded that the swelling force was controlled by the relaxation process of the macromolecular chains indicating a significant ”‘93 observed that under deviation from pure F ickian swelling behavior. Bell and Peppas oscillatory pH conditions network collapse (complexation) occurred more rapidly than swelling or network expansion (decomplexation), and that longer PEG graft lengths resulted in lower diffusion coefficients. In a more recent study, Lowman and Peppasg‘1 reported that the diffusion coefficients were lower in acidic compared to neutral or basic media due to the formation of interpolymer complexes in the gels. The authors demonstrated that the ratio of the solute radius to the network mesh size also was a significant factor in the overall behavior of these gels. While the diffusion coefficient of the smallest solute, proxyphylline, only changed by a factor of five between the complexed and uncomplexed state, the diffusion coefficients varied by almost two orders of magnitude for the solute FITC-dextran, which has a molecular radius ten times greater than proxyphylline. These results are explained in terms of mesh size characteristics of the gels. l. 95 Finally, Dorski et a recorded the glucose-responsive swelling behavior of P(MAA-g—EG) networks containing glucose oxidase immobilized onto the graft copolymer. 32 2.5 Free rt three steps; ir polymerizatior the size of the by a simple so lnitiati PIOpav: Termi Here, R' re l ’ 1" repIESeI-n disproportion glitn by- 2.5 KINETICS OF FREE RADICAL POLYMERIZATIONS Free radical polymerization reactions occur through a mechanism that includes three steps; initiation, propagation and termination. In a classical description of the polymerization kinetics, it is assumed that the reaction rate constants are independent of the size of the grong polymer chain, and the polymerization mechanism is represented by a simple set of reactions.96 Initiation: Kr R'+M——>M,‘ (2'1) Propagation: Kp . . M,‘;'+M—>M,,+1 (2 2) Termination: (2.3) Kr M n + M m -—> dead polymer chain Here, R’ represents the initiating radical, M represents a monomer molecule and M ,I represents a growing polymer chain of n monomer units. The total termination constant, K,, includes contributions due to termination by combination and disproportionation. Reaction 2.2 is a bimolecular reaction and the rate of propagation is therefore given by: 33 where [M‘] st system. lhe ra When System Will it termination re The frequenc concentration balanced by r} Oltadica]s in 1 Since mum“ that Steadl‘SIate z c0"Celttration experimentall R], = KP ~[M’]-[M] (2.4) where [M *] stands for the sum of the concentrations of all monomer-ended radicals in the system. The rate of termination may be written as: R, = 2.1g -[M"]2 (2.5) When a free-radical polymerization is first started, the number of radicals in the system will increase from zero as the initiator begins to decompose. The frequency of termination reactions will also increase from zero as the initiator begins to decompose. The frequency of termination reactions is proportional to the square of the total concentration of radicals in the system. Eventually, the rate of radical generation will be balanced by the rate at which radicals undergo mutual annihilation, and the concentration of radicals in the system will reach a steady state. Since the steady state is reached soon after the polymerization start, it can be assumed that it applies to the whole course of the polymerization. By applying the steady-state assumption to the equations 2.4 and 2.5 and solving for the total concentration of monomer-ended radicals, we find an expression in terms of experimentally accessible quantities: K I . . 112 (2.6) [M.]=[f K. [1]) 34 “herefi the l the initiator e specific tempt From rate of polyrtt independent c value. Here. R, is th Lhe“Hal initia the ““196th Mica] PIOdut Thek where f, the fraction of the radicals generated that are captured by monomers, is called the initiator efficiency. The specific rate constant of decomposition of the radical at a specific temperature is Kd, From this three-step mechanism for the reaction, the following equation for the rate of polymerization can be derived under the assumptions that kinetic constants are independent of chain length, and the concentration of radicals attains a quasi-steady-state value. R. 112 (27) RP = KP '[M][f] I Here, R, is the rate of initiation and depends on the method of initiation that is used. A thermal initiator, which is a thermolabile compound, decomposes into two free radicals at the temperature of the reaction mixture. This can be expressed in terms of the rate of radical production as: Rl'=2.f.Kd.[1] (2.8) The kinetic chain length v is the average number of monomers that react with an active center from its formation until it is terminated. It is given by the ratio of the polymerization to the rate of initiation and, again, under steady-state conditions where R. =Ri,: R __ KI, ~[M] (2.9) R, '2-(f-K.-rIr-K.)'” 35 from ec molecu the mo] Since I indeper, Chains i initiatio loaned molecu] lhc cont deviatio COIttroll radical-r bro rad SignlilCe aPPFOac] radicals SOleriOn from equation (2.4), (2.5) and (2.6). In this classical picture of the polymerization reaction, the average polymer molecular weight can be found by taking the kinetic chain length and multiplying it by the molecular weight of the monomer unit M0 for a homopolymer. Mn =M0~v (2.10) Since RD and R. are constant throughout the entire reaction (the kinetic constants are independent of conversion and molecular weight), the molecular weight of dead polymer chains is constant throughout the entire reaction. The lifetime of a growing radical from initiation to termination is on the order of a second. Therefore long polymer chains are formed almost instantaneously from the beginning of the reactions. Although the molecular weight of dead chains may be approximately constant throughout the reaction, the concentration of such chains increases with conversion. The limitations of this classical model have been realized. A major reason for deviations from this classical model is the fact that the termination reaction is diffusion- controlled at all stages of the reaction. This arises from the nature of the free radicals and radical-radical reactions. Because radicals are uncharged and have an unpaired electron, two radicals can approach each other and react to form a bond without encountering a significant energy barrier. Activation energies for radical-radical reactions are low, 7 As a result, the rate of reaction of two approaching zero for small alkyl radicals.9 radicals is limited by the rate at which two reacting partners can diffuse through the solution to approach each other; hence the reaction is diffusion controlled. 36 Since dithaion-cont also diffusion the radical in rate of the ter termination rt radical chain: concentration conversion. l classical picr Often increast acceleration, Altho exPerimental independent . theeharacter: each of “hit and a Small r the COUISe ( conversim] ls became the chain length experiencEd Since the termination reaction of two small alkyl radicals is known to be diffusion-controlled, it is not surprising that termination in free-radical polymerizations is also diffusion controlled. Therefore, any variable that changes the diffusive behavior of the radical in the reaction mixture also changes the rate of the termination reaction. If the rate of the termination reaction decreases, the rate of the polymerization increases since termination removes radicals from the system. The diffusive characteristics of growing radical chains depends on variables such as the average polymer chain length and the concentration of polymer chains. Therefore, the rate of termination depends on conversion. Consequently, a polymerization system is much more complicated than the classical picture described previously. Experimentally observed polymerization rates often increase at a threshold value of conversion. The type of phenomenon is called auto acceleration, the gel effect or Trommsdorff effect in crosslinking systems. Although the rate of termination has been found to depend on conversion, experimental evidence has indicated that the rate of propagation is essentially independent of conversion.98’99 The reason for this difference stems from differences in the characteristics of the two types of reactions. While termination involves two radicals, each of which is attached to a long polymer chain, propagation involves one such radical and a small monomer molecule. While the long-chain radicals may lose mobility during the course of the reaction, the monomer molecules remain mobile until very high conversion is achieved. The rate of propagation does not depend strongly on chain length because the chemical environment near the free radical is essentially independent of chain length. The number of accessible monomers and the numbers of collisions experienced by a free radical are fairly independent of chain length. It should be noted 37 that the rate temperature some point (polymer pla could becom the reactionr In the polymerizati these model conversion a NEIL the fur the difl‘eren Com'ergjcm C mudels use p forms! they that the rate of propagation can be greatly reduced by vitrification. If the reaction temperature is lower than the glass transition temperature of the polymer being formed, at some point in the reaction the glass transition temperature of the reaction mixture (polymer plasticized by monomer) could exceed the reaction temperature and the system could become glassy. In a glassy system, the mobility of the monomers is very low and the reaction essentially stops. In the past, there has been significant work on the modeling of linear free-radical polymerization reactions taking into account diffusion-controlled tennination.'°0"0"'°2 In these models, a fimctional form for the termination rate constant as a function of conversion and molecular weight is first proposed based on some physical arguments. Next, the fianctional form is fit to experimental data by some adjustable parameters and the differential equations describing the polymerization reaction are solved with a conversion and molecular weight dependent termination constant. Although different models use different assumptions, different physical arguments, and different functional forms, they all give results, which agree well with experimental conversion versus time and molecular weight versus conversion curves. 38 Trash Figure 2.1 Simplified general water treatment scheme. 39 10. 11. 12. 13. 14. 15. 16. 2.6 REFERENCES Becher, P., EncyCIOpedia of Emulsion Technology, Vol. 1, Marcel Dekker, New York, 1983. Clint, J. H., Surfactant Aggregation, Chapman and Hall, New York, 1991. Tarnbe, D. E.; Paulis, J .; Sharma, M. M., J. Coll. Interface Sci, 1 71, 463 (1995). Abdel-Azim, A. A. A., Zaki, N. N., Polym. Adv. Techn, 9(2) 159 ( 1998). Nurxat—Nuraje, Chen, W. H., Chen, W., Li, Z. P., Wang, H. Q., J. Disp. Sci. Techn., 20(5), 1501 (1999). Dumitriu, S., Popa, M., Dumitriu, M., J. Bioact. and Compatible Polymers, 5, 89 (1990). Lissant, K. J ., Demulsification — Industrial Applications, Surfactant Science Series, Vol. 13, Marcel Dekker Inc., New York, 1983. United States Code: The Clean Water Act (CWA); 33 U.S.C. s/s 121 et seq. (1977). Batty, J., Lim, R., Archives of Environmental Contamination and Toxicology, 36(3), 0301 (1999). ' Margesin, R., Schinner, F., Appl Microbiol Biotechnol, 49(4), 482 (1998). Wu, G., Yan, X., Bulletin of Environmental Contamination and Toxicology, 60(2), 329 (1998). Sales, J. et al., Bulletin of Environmental Contamination and Toxicology, 63(1), 94 (1999). Rocha, C., Infante, C., Appl. Microbiol. Biotechnol, 47(5), 615 (1997). Mathur, A. M., Synthesis and Characterization of Polymeric Pseudocrown Ethers and Reversible Block/Graft Copolymeric Emulsifiers based upon Polymer Complexation, Dissertation, Michigan State University, East Lansing, 1996. Fendler, J. H., Membrane Mimetic Chemistry, Wiley-Interscience, New York, NY, 1982. Cuomo, J., Merrifield, J. H., Keana, J. F. W., J. Org. Chem. 45, 4216 (1980). 40 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Jaeger, D. A. et al., J. Am. Chem. Soc., 110, 5123 (1988). Jaeger, D. A., Gilick, T. G., J. Am. Chem. Soc., 64, 1550 (1987). Piasecki, A., Sokolowski, A, Burczykm, B. Kotlewska, U., J. Am. Oil Chem. Soc., 74(1), 33 (1997). Yamamura, S., Nakmura, K., Tanaka, K., J. Am. Oil Chem. Soc., 66, 1165 (1989). Wang, G.-W., Yuan, X.-Y., Liu, Y.-C., Lei, X.-G., J. Am. Oil Chem. Soc., 71(7), 727 (1994). Kida, T., Yurugi, K., Masuyama, A., Nakatsuji, Y., Ono, D., Takeda, T., J. Am. Oil Chem. Soc., 72(7), 773 (1995). Kida, T., Morishima, N., Masuyama, A., Nakatsuji, Y., J. Am. Oil Chem. Soc., 71 (7), 705 (1994). Yue, C., Harris, J. M., Hellberg, P.-E., Bergstrom, K., J. Am. Chem. Soc., 73(7), 841 (1996). Burczyk, B., Sokolowski, A., J. Am. Chem. Soc., 74(7), 895 (1997). Jaeger, D. A., Golich, T. G., J. Am. Chem. Soc., 64(11), 1550 (1987). Jaeger, D. A., Russell, S. G. G., Shinozaki, H., J. Org. Chem, 59, 7544 (1994). Jaeger, D. A., Ward, M. D., Dutta, A. K., J. Org. Chem, 53, 1577 (1988). Jaeger, D. A., Ward, M. D., J. Org. Chem, 47, 2221 (1982). Jaeger, D. A., Finley, T., Walter, M. R., Martin, C. A., J. Org. Chem, 51, 3956 (1986). Jaeger, D. A., Martin, C. A., Golich, T. G., J. Org. Chem, 49, 4545 (1984). West, C.A., Sanchez, A. M., Hanon-Aragon, K. A., Salazar, I., Menger, F. M, Tetrahedron Letters, 3 7(51), 9135 (1996). Kunitake, T., Nakashima, N., Shimomura, M., Okahata, Y., Kano, S., Ogawa, T., J. Am. Chem. Soc., 102, 6642 (1980). Dunkin, I. R., Gittinger, A., Sherrington, D.,C., Whittaker, P., J. Chem. Soc., Chem. Commun, 2245 (1994). 41 35. 36. 37. 38. 39. 40. 41. 42. 43. 45. 46. 47. 48. 49. 50. 51. Nuyken, 0., Meindl, K., Wokaun, K., Mezger, T., J. Photochem. Photobiol., A, 81, 45 (1994). Dunkin, I. R., Gittinger, A., Sherrington, D. C., Whittaker, P., J. Chem. Soc. Perkin Trans. 2, 1837 (1996). Saji, T., Ebata, K.-I., Sugawara, K, Liu, S., Kobayashi, K., J. Am. Chem. Soc., 116, 6053 (1994). Frankel, D. A., Lamparski, H., Liman, U., O’Brian, D. F., J. Am. Chem. Soc., 111, 9262 (1989). - Bekturov, E. A., Bimendina, L.A., Interpolymer Complexes, Adv. Polym. Sci, 43, 100-147 (1980). Poland, D. Scheraga, H. A., Theory of Helix-Coil Transitions in Biopolymers, Academic Press, New York, 1970. Poland, D., Cooperative Equilibria in Physical Biochemistry, Oxford University Press, Oxford, 1978. Magee, W. S., Gibbs, J. H., Zimm, B. H., Biopolymers, I, 133 (1963). Damle, V. N., Biopolymers, 9, 353 (1970). ' Blake, R. D., Biopolymers, 11, 913 (1972). Klier, J., Peppas, N. A., Proc. Adv. Biomed. Polym., I, 107 (1989). Klier, J., Peppas, N. A., Proc. Intern. Symp. Contr. Rel. Bioact. Agents, 16, 107 (1989). Tsuchiba, E. Abe, K., Interactions Between Macromolecules in Solution and Intermacromolecular Complexes, Adv. Polym. Sci, 45, 1 (1982). Bekturov, B., Bimendina, L. A., Interpolymer Complexes, Adv. Polym. Sci, 43, 100 (1980). Osada, Y., Conversion of Chemical into Mechanical Energy by Synthetic Polymers, Adv. Polym. Sci, 82, l (1987). Vinogradov, S. N., Hydrogen Bonding, Van Nostrand Reinhold, New York, 1971. Tanford, C., The Hydrophobic Effect, Wiley and Sons, New York, 1973. 42 52. 53. 55. 56. 57. 58. 59. ,' 61. 62. 63. 67. 68. 70. Kabt Bailc‘ Anti] Sore Hum (197. Osad 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 66. 67. 68. 69. 70. Kabonov, V. A., Papisov, I. M., Vysokomol. Soyed., A14, 941 (1972). Bailey, F. E., Lundberg, R D., Callard, R. W., J. Polym. Sci, 2, 845 (1964). Antipina, A. D., Baranovski, V. Y., Papisov, I. M., Kabanov, V. A., Vysokomol. Soyed., A14, 941 (1972). Ikawa, T., Abe, K., Honda, K., Tsuchida, E., J. Polym. Sci, Polym. Chem, 13, 1505 (1975). Osada, Y., and Sato, M., J. Polym. Sci, Polym. Lett., 14, 129 (1976). Osada, Y., J. Polym Sci, Polym Chem, 15, 255 (1977). Osada, Y., J. Polym Sci, Polym. Chem, 1 7, 3485 (1979). Nagasawa, M., Murase, T., Kondo, K., J. Phys. Chem, 69, 4005 (1965). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 474 (1987). Char, K., Frank, C. W., Gast, A. P., Tang, W. T., Macromolecules, 20, 1833 (1987). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 1839 (1987). Chen, H. L., Morawetz, H., Eur. Polym. J., 19, 923 (1983). Tsuchida, E. and Abe, K., Interactions Between Macromolecules in Solution and Intermolecular Complexes, Adv. Polym. Sci, 45 , 1-119 (1982). Osada, Y., Conversion of Chemical into Mechanical Energy by Synthetic Polymers, Adv. Polym. Sci, 82, 1-46 (1987). Kabanov, V.A. and Papisov, I.M., Vysokomol. Soyed. A21, 243-281 (1979). Scranton, A.B., Klier, J. and Aronson, C.L., “Complexation of Polymeric Acids with Polymeric Bases”, in Polyelectrolyte Gels, R.S. Harland and R.K. Prud’homme Eds., ACS Symposium Series 480, American Chemical Society, Washington DC, 171-189 (1992). Bekturov, E. A., Bimendina, L.A., Interpolymer Complexes, Adv. Polym. Sci, 43, 100-147 (1980). Turro, N. J ., Arora, K. S., Polymer, 27, 783 (1986). 43 71. 74. 75. 76. 77. 78. 79. 80. 83. 85. 86. 87. 88. ' 89. ' H1106 Klier. (1990 3ch (1991 Oyam Oyam Matht Sailo. Hemk 800131 Klier. Ph.D. 88pm and P D01, 1 Shigu Ishiha 110, Y 1. K1 1701311 SC'an P0136 0615. 171.1 P917136 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. ' 88. 89. Klier, J., Scranton, A.B., and Peppas, N.A., Macromolecules, 23, 4944-4949 (1990). Scranton, A. B., Klier, J ., Peppas, N.A., J. Polym. Sci: Polym. Phys., 29, 211-224 (1991). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 471 (1987). Oyama, H. T., Tang, W. T. Frank, C. W., Macromolecules, 20, 1839 (1987). Mathur, A. M., Drescher, B., Scranton, A. B., Klier, J ., Nature 392, 367 (1998). Saito, S. and Sakamoto, T., Colloids & Surfaces, 23, 99-104 (1987). Hemker, D. J ., Garza, V., and Frank, C. W., Macromolecules, 23, 4411-4418 (1990). Noolandi, J., Makromol. Chem Theory Simul., 1(5), 295-298 (1992). Klier, J ., Self-Associating Networks of Poly(methacrylic acid-g-ethylene glycol), Ph.D. Dissertation, Purdue University, Lafayette, Indiana, 1989. Peppas, N. A., Barr-Howell, B. D., in: Peppas, N. A. (ed.), Hydrogels in Medicine and Pharmacy, CRC, Boca Raton, Florida, 28-55, 1986. Tanaka, T. Polymer, 20, 1404 (1979). Hirotsu, S., Hirokawa, Y., Tanaka, T., J. Chem. Phys., 87, 1392 (1987). Doi, M., Mitsumoto, M.,Hirose, Y., Macromolecules, 25, 5504 (1992). Shiga, T., Hirose, Y., Okada, A., Kurauchi, T., J. Appl. Polym. Sci, 47, 113 (1992). Ishihara, K., Matsui, K., J. Polym. Sci: Polym. Lett. 24, 413 (1986). Ito, Y., Casolaro, M., Kono, K., Imanishi, Y., J. Contr Rei, 10, 195 (1989). J. Klier, A. B. Scranton, and N. A. Peppas, Self-associating networks of poly(methacrylic acid-g-ethylene glycol), Macromolecules, 23 (1990) 4944-4949. Scranton, A. B., Klier, J ., Aronson, C. L, Complexation of polymeric acids with polymeric bases, in: Harland, R. S. and Prud’homme, R. K. (Eds), Polyelectrolyte Gels, ACS Symposium Series 480, American Chemical Society, Washington DC, 171-189, 1992. Peppas, N. A., Klier, J., J. Control. Rei, 16(1-2), 203 (1991). Scranl 2913i . Peppa . Peppa . Khmc 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. Scranton, A. B., Klier, J ., Peppas, N. A., J. Polym. Sci, Part B: Polym. Phys., 29(2), 211(1991). Peppas N. A., Khare, A. R., Adv. Drug Del. Rev., 11(1-2), l (1993). Peppas, N. A., J. Bioact. Compat. Polym., 6(3), 241 (1991). Khare A. R., Peppas, N. A., Polym. News, 16(8), 230 (1991). Lowman A. M., Peppas, N. A., J. Biomat. Sci. Polym. Ed., 10(9). 999 (1999). Bell C. L., Peppas, N. A., J. Biomater. Sci, Polym. Ed., 7(8), 671 (1996). Odian, 6., Principles of Polymerization, Wiley and Sons, New York, 1981. Lowry, T. H., Richardson, K. S., Mechansim and Theory in Organic Chemisty, Harper and Row, New York, 1986. O’Driscoll, K. F., Mahabadi, H. K., J. Polym. Sci, Polym. Chem, 14, 869 (1986). DeShrijer, F., Smets, G., J. Polym. Sci, A-I, 4, 2201 (1966). 100. Mita, I., Horie, K., Macramol. Sci, Rev. Macromol. Chem. Phys., C27, 91 (1987). 101. Olaj, O. F., Zifferere, G., Gleixner, G., Macrololecules, 20, 839 (1987). 102. Sch, 8. K., Sundberg, D. C., J. Polym. Sci, Polym. Chem, 20, 1345 ( 1982). 45 The pre' detelopment 01 afidress this n: complexes in b. 6111 build upor ”131 reversibIe 1 8990151116 arc]! Momed a s 3mm"Hie aci methacrylate. ' HIkblock’lgraftj one go L I Clalauenzafi 0 n The compound Rafielers fOr tween the rea 1' ‘1 . Chapter 3 OBJECTIVES The previous discussion illustrates that there is a need for an investigation into the development of reversible emulsifiers. The general objective of this research project is to address this need by investigating the potential of reversible hydrophobic polymer complexes in block-graft copolymers for the design of reversible emulsifiers. This thesis will build upon the previous research by Mathur and collaborators, who demonstrated that reversible emulsifiers could be designed using complementary polymers in a grafi copolymer architecture. As illustrated in the previous chapter, these researchers performed a series of preliminary studies of copolymers containing an acrylic or methacrylic acid backbone, and grafts of methoxyoligo(ethylene glycol) acrylate or methacrylate. The main focus of their initial studies was to establish the reversibility of the block/graft surfactants. One goal of this dissertation research is the to provide a fundamental characterization of the block-graft copolymer system poly(methacrylic acid-g-ethylene glycol) (P(MAA-g-EG)) with respect to the molecular architecture and molecular weight. The composition and degree of length of the polymer chain are the essential design parameters for the reversible emulsifiers, and a systematic study of the relationship between the reaction conditions and the polymer structure is important. A second goal of this project is to investigate the relationship between the solution characteristics (including the emulsification properties) of the P(MAA-g-EG) copolymers and the 46 nierlying p01 mersible emu To met follows: 1. 4...) OJ To des Poly(m weight 10 cha speetro: permea' To esti acidic c To stud; C0p01yn To iden methyl T0 Char by P(M To prox constiiih NMR 5; To 011211 llene underlying polymer structure. These studies will provide guidelines for the design of reversible emulsifiers. To meet these general goals, the specific objectives of this research project are as follows: 1. To design a semibatch polymerization reactor that will be used to synthesize poly(methacrylic acid-g-ethylene glycol) copolymers with varying molecular weight and level of grafting; To characterize the copolymer composition using nuclear magnetic resonance spectroscopy and investigate the copolymer molecular weights using gel permeation chromatography; To estimate the aggregate size of P(MAA-g-EG) copolymers in water under acidic conditions using dynamic light scattering; To study the pH-controlled reversibility of emulsions stabilized by P(MAA-g-EG) copolymers; To identify the effect of the solution pH on the interfacial tension in the system methyl laurate/water/P(MAA-g-EG); To characterize the droplet size distribution and stability of emulsions stabilized by P(MAA-g-EG) copolymers using laser scanning confocal microscopy; To provide insight into the spatial arrangement of the complementary polymer constituents of in complexes of P(MAA-g-EG) using two dimensional NOESY NMR spectroscopy; and To characterize the hydrophobicity of P(MAA-g-EG) in aqueous solution using pyrene fluorescence. 47 SY EMU T1 and prOp cOmpositi series of For this Chapter 4 SYNTHESIS AND CHARACTERIZATION OF POLYMERIC EMULSIFIERS CONTAINING REVERSIBLE HYDROPHOBES: POLY(METHACRYLIC ACID-G-ETHYLENE GLYCOL) 4.1 INTRODUCTION The broad objective of this dissertation is to analyze the emulsification behavior and properties in the system of P(MAA-g-EG) copolymers as a function of their composition and molecular weight. To meet this objective, it is necessary to synthesize a series of grafi copolymers polymers of various molecular weights and level of grafting. For this reason, a series of poly(methacrylic acid-g-ethylene glycol) copolymers of varying molecular architectures were synthesized in a custom-built semi-batch reactor by a free radical copolymerization of methacrylic acid with the macromonomer methoxy poly(ethylene glycol) methacrylate. The copolymers were characterized using lH-NMR to determine the polymer composition and gel permeation chromatography (GPC) to characterize the polymer molecular weight distribution. The GPC and NMR results were combined to estimate the number of grafts per chain for each of the polymer samples. In addition, for a constant MPEGMA content, the average number of grafts per chain decreased as the initiator concentration was increased. These studies provide important information about the relationship between the reaction conditions and the molecular architecture of the resulting reversible copolymeric emulsifiers. 48 free r 163ch 1000. averag: macror macror reactan monon PUInps 183C101 motor measu, sealed blanke Polimt ACE ( a COM: Ace 01 km“ 111" m? m 01 SS 0C 11C 4.2 MATERIALS AND METHODS The poly(methacrylic acid-g-ethylene glycol) copolymers were synthesized by free radical copolymerization of methacrylic acid (MAA, Aldrich) with the macromonomer methoxy poly(ethylene glycol) monomethacrylate 1000 (MPEGMA 1000, Polysciences, Warrington, PA). In this nomenclature, the “1000” denotes the average molecular weight of each oligomeric ethylene glycol chain in the macromonomer. This corresponds to an average of 22.7 ethylene glycol repeat units per macromonomer. Polymerization was initiated using sodium persulfate (Aldrich). All reactants were high purity and were used as received. The c0polymers were synthesized using a semi-batch reactor in which the monomer was continuously fed into the well-mixed system using programmed syringe pumps (KD Scientific, Boston MA) with gas-tight syringes (SGE, Austin TX). The reactor contents were mixed using a Trubore® stirring system driven by a heavy-duty air motor (Cole Palmer Instrument Co., Chicago, IL). The speed (RPM) of the stirrer was measured with an optical tachometer TACH-4-AR (Monarch Inst., Amherst, NH). The sealed reactor was equipped with a reflux condenser and was kept under a nitrogen blanket to eliminate the presence of oxygen, which can inhibit free radical polymerizations. All glassware components used in the reactor were purchased from ACE Glass Inc. (Vineland NJ). The temperature of the reaction vessel was maintained at a constant value of 75°C using PID temperature controlled heating unit purchased from Ace Glass. Stock solutions of the monomer MPEGMA 1000 were prepared by dissolving known amounts of the monomers into a 50/50 (v/v) mixture of 200 proof ethanol and 49 HPLC dbsol rehig [repa wlfiel trou COVCI SOrn flou: Cepoj their fiSpe are: corn} 10 C0 £9561 cOne HPLC grade water under stirring. Similarly, initiator solutions were produced by dissolving solid sodium persulfate in HPLC grade water, and the solution was stored in a refrigerator at 40°F until further use (solutions were used within three days of their preparation). Before each run, 200 ml of a 50/50 (v/v) ethanol/water mixture were poured into the 500 ml reactor, and the system was sealed except for one opening through which reactants were fed. The dissolved oxygen was removed by bubbling nitrogen through the solvent for about 5 min, and then the nitrogen inlet valve was opened to cover the reactor contents with a nitrogen blanket before the system was completely sealed. A 100 ml syringe and two 50 ml syringes were filled with pure methacrylic acid, 50 ml of MPEGMA solution and 30 ml of initiator solution, respectively. Each feed pump was programmed to deliver a predetermined constant volumetric flow rate selected for the specific reaction. For example, for the P(MAA-g-EG) 5/10 copolymer, the flow rate of the methacrylic acid feed, the MPEGMA solution feed and the initiator solution feed were selected as 1.000 ml/min, 1.111 ml/min and 0.500 ml/min, respectively. The feedlines were degassed before the monomer feed was began, and the automatic shut off time was 45 min for all syringe pumps. When the pumping was complete, the system was removed from the heating source, and the reactor was allowed to cool to room temperature in a post-polymerization step. Using this reactor, an array of polymers were produced with theoretical ethylene glycol contents of 0, 2.5, 5, 10, 20, 30 wt.% and varied chain lengths due to initiator concentrations of 10, 5, 1, 0.5, 0.1 wt.% of total monomer mass. In the remainder of this dissertation, the copolymers will be specified with the ethylene glycol content followed by the initiator concentration. For example, a copolymer labeled as P(MAA-g—EG) 5/ 10 50 is polymerized concentration t lire ge 1090A instrun autosampler. l Laboratories) \ packing, An system was us lnstmment 501' Caliber for H1 the calculation “umber averar llre m monodisperSe 0011mm 3)'Sten 161100, 28.06 samiles Were l~hfil polwach methyl 81’ 011p mvestigallOH. is polymerized with an E6 content of 5 wt.% of the total monomer, and an initiator concentration of 10 wt.% of the total monomer. 4.3 GEL PERMEATION CHROMATOGRAPHY The gel permeation chromatography (GPC) studies were performed using an HP 1090A instrument (Hewlett-Packard, Palo Alto CA) equipped with a factory-installed autosampler. For these experiments a set of two PL aquagel-OH MIXED 8pm (Polymer Laboratories) columns were used in series following a guard column containing the same packing. An HP 1047A refractive index detector equipped with a thermostated optical system was used in these studies. The data was collected and analyzed online using the Instrument software Chemstation for LC (Hewlett—Packard) and the Software package PL Caliber for HP Chemstation version 4.01 (Polymer Laboratories). This software allowed the calculation of the various moments of the molecular weight distribution, including the number average and weight average molecular weights. The molecular weight calibrations were performed with commercially available monodisperse poly(acrylic acid)-sodium salt standards (Polymer Laboratories). The column system was calibrated using a series of standards with molecular weights of 7500, 16,000, 28,000, 62,900, 115,000, and 272,900. The reported polydispersities of these samples were 1.34, 1.41, 1.60, 1.74, 1.67, and 1.51 respectively. Considering the fact that poly(acrylic acid) differs from poly(methacrylic acid) only by nonexistence of the or- methyl group both constituent polymer chains have comparable radii of gyration in the formation of an expanded coil. Although a grafted copolymer is subject of our investigation, the molar ratio of the monomer poly(methacrylic acid) to the co-monomer 51 MPEGMA is 1 reasonable cho voltune of 100. phosphate buf phosphate (All (Aldrich, St. L pH corrected tr solution. The CXperiments. molecular Wei P01)Ttomials. For the EEG) COPOlyr phOSphate but‘ Using Gelman transferred inn fit “the CalitI “11101311163. 1 MPEGMA is far greater than one in all studied cases. Hence, poly(acrylic acid) was a reasonable choice as a standard for our polymer system. For all experiments an injection volume of 100. rd and a flow rate of 1.000 ml/min were used. The carrier solvents was a phosphate buffer solution prepared by dissolving 1.204 grams of sodium dihydrogen phosphate (Aldrich, St. Louis, MO) and 21.247 grams of sodium nitrate phosphate (Aldrich, St. Louis, MO) in 1.000 liters of HPLC grade water. The buffer solution was pH corrected to pH=7.0 from pH=5.3 by addition of about 1.6 ml 5N sodium hydroxide solution. The columns were maintained at a temperature of 35°C throughout the experiments. As shown in Figure 4.1, the calibration data for the logarithm of the molecular weight versus the elution time was fit to both first order and second order polynomials. For the GPC analysis of the graft copolymers, 0.25 wt.% solutions of the P(MAA- g-EG) copolymers were prepared by dissolving 12.5 mg of the polymers in 5 ml of the phosphate buffer solution described above. After dissolution, the samples were filtered using Gelman Acrodisc 0.2um PTFE filters (Pall Corporation, East Hills, NY) and were transferred into the HP autosarnpler screw cap vials. For most samples, the second order fit of the calibration data was used to find the apparent molecular weight of the graft copolymers. For very high molecular weight samples (apparent M.1 greater than 560,000 grams/mole), the first order fit was used. 52 Prr a 500 Ml spectrum, erperimer. the deuter recovery e pulse repe‘ All speetn were can firnetions 1 Software vi To EG) COpOl 198? Petri Were dried White p0“- demented p01.‘Tner w “’33 filtere‘ 791R tube 4.4 1H NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY Proton nuclear magnetic resonance ('H-NMR) spectroscopy was performed using a 500 MHz Varian VXR-500 NMR spectrometer (Varian, Palo Alto CA). For each spectrum, 64 scans with a spectral window size of 8000 Hz were averaged. The experiment was performed with sample spinning at a rate of 20Hz with the field locked to the deuterium resonance of the solvent (deuterated pyridine). Based upon Tl inversion recovery experiments, a 11.8 yrs pulse (corresponding to 90° flip angle) was chosen. A pulse repetition time of five times the maximum Tl relaxation time of 3 sec was selected. All spectra were recorded for a sample temperature of 75°C. Fourier Transformations were carried out with zero filling, and the F ID was not treated with any apodization functions before Fourier transformation. The data reduction was done using the Varian software version 5.1. To prepare samples for the NMR studies, the reactor solutions of the P(MAA-g- EG) copolymers (dissolved in a 50/50 mixture of ethanol and water) were spread onto large Petri dishes and were dried in a hood for several weeks at room temperature, then were dried in a vacuum oven at 40°C for two days. The dried polymer was ground into white powder, then 0.16 grams of polymer were dissolved in 1.00 ml of 100 atom% deuterated pyridine-d5 (ISOTEC, Miamisburg, OH) in a seal-capped 5m] vial. The polymer was allowed to completely dissolve for 24 hours. After this time, each solution was filtered using 45 um Gelman Acrodiscs and was transferred into a 5 mm diameter NMR tube (Kontes Scientific, Vineland NJ). 53 4.5.1 summ; design pereen numbe The se repeat the nu: the pc rejecti (Consi. disearr 4.5 RESULTS AND DISCUSSION 4.5.1 Gel Permeation Chromatography Results The experimental results from the gel permeation chromatography studies are summarized in Table 4.1. In this table, the first column corresponds to the sample designation (recall that the first number in this designation corresponds to the weight percent of the ethylene glycol repeat units in the monomer mixture, while the second number corresponds to the weight fraction of the initiator relative to the total monomer). The second column in the table indicates the ratio of methacrylic acid to ethylene glycol repeat units in the monomer mixture during synthesis. Columns three through six contain the number average molecular weight, M.,, the weight average molecular weight, M.,, and the polydispersity of the sample. After application of the statistical Q test for data rejection, the results for two samples with unreasonably high apparent molecular weights (considerably out of the range of the calibration) and unusually high polydispersities were discarded. The data in Table 4.1 illustrate the effect of the initiator concentration on the molecular weight of the poly(methacrylic acid-g-ethylene glycol) copolymers. As expected, the molecular weight of the polymers increases as the initiator concentration is decreased, and is proportional to the inverse square root of the initiator concentration (as illustrated in Figures 4.2 and 4.3 discussed below). The data also illustrate that there is no consistent trend for the polydispersity as a function of the MAA:EG repeat unit ratio. For repeat unit ratios ranging from infinity (pure PMAA) to 4.6 (a graft copolymer with a graft occurring about every 100 backbone repeat unit), the polydispersity remains essentially constant, with an average value of 3.4. These values of the polydispersity are 54 typical for f] polymer occr The coneentratior plotted as a samples. In while the lint expected ins concentration weight is pro falls into the the molecular typical for free radical chain polymerizations, and do not suggest that chain transfer to polymer occurs to a significant degree. The relationship between the polymer molecular weight and the initiator concentration is illustrated in Figure 4.2 in which the number average molecular weight is plotted as a function of the initiator concentration for the pure poly(methacrylic acid) samples. In this figure, the points correspond to the pure PMAA data from Table 4.1 while the lines correspond to the 99% confidence interval for best fit of the data to the expected inverse square root relationship between molecular weight and initiator concentration. Figure 2 illustrates that, as expected, the number average molecular weight is proportional to the inverse square root of the initiator, and that all of the data falls into the 99% confidence interval. The best-fit mathematical relationship between the molecular weight and the initiator concentration if given below: _ 77,000 (4.1) M n,I’MAA tm’ where MmpMAA and win, denote the number average molecular weight of poly(methacrylic acid) (grams/moles) and the weight percentage of the initiator (dimensionless) fed to the reaction system. Figure 4.3 illustrates the effect of the oligo(ethylene glycol) macromonomer on the GPC results for the polymer molecular weight as a function of initiator concentration. The Figure contains five series of data points, with each series corresponding to a different MAA:EG repeat unit ratio during synthesis. This data are superimposed on the 99% confidence interval determined from the GPC results for pure PMAA. Note that the inverse square root dependence between the molecular weight and the initiator 55 concentratior the addition from the res 99% confide trend suggest the hydrodyn The t oligo(ethylen the presence grafts after 8] 0f long-chair Sittthesis. I] polymers con addition, Cha increasing the molecular We Observed Valr miltst’er to p0 during sl’nlheg the excluded oligo(ethylene backbone. A ladius‘ concentration is generally followed by all of the data. However, the Figure illustrates that the addition of the oligo(ethylene glycol) macromonomer leads to a systematic deviation from the results obtained for pure PMAA with the data consistently falling above the 99% confidence interval at low initiator concentration (high molecular weight). This trend suggests that the presence of the oligo(ethylene glycol) grafts leads to an increase in the hydrodynamic volume of the dilute polymer chains. The apparent increase in the hydrodynamic volume upon the addition of the oligo(ethylene glycol) grafts could potentially arise from structural differences caused by the presence of the macromonomer during synthesis, or from conformation effects of the grafts after synthesis. For example, this trend could possibly arise from the introduction of long-chain grafts as a result of chain transfer to the ethylene glycol grafts during synthesis. If this were the case, the effect should be the most pronounced for the polymers containing high ethylene glycol content (low MAA:EG repeat unit ratio). In addition, chain transfer to the grafts would reduce the primary chain length while increasing the branching, and therefore should have little effect on the number average molecular weight, but would result in an increase in the polydispersity index. The observed values for the polydispersity index (shown in Table 4.1) suggest that chain transfer to polymer is not extensive. Even if there were no additional chain transfer during synthesis, the presence of the grafts could increase the hydrodynamic radius due to the excluded volume introduced by the grafts. The volume occupied by the oligo(ethylene glycol) grafts cannot be occupied by other grafts, or by the PMAA backbone. As a result, the chains will tend to expand, increasing the hydrodynamic radius. 56 452 'H-N.‘ A rep shown by the (correspondin pyridine-d5. characteristic copolymer. 1 methyl peak correspondint, intense peak triads.2'3~"5 1 atactic mono] L4. {:1 (£2. (D —— 4.5.2 lH-NMR Spectroscopy A representative 1H-NMR spectrum of a P(MAA-g-EG) copolymer solution is shown by the trace of Figure 4.4. This spectrum corresponds to the P(MAA-g—EG) 5/ 10 (corresponding to a MAA:EG repeat unit ratio of 9.7 to 1 during synthesis) dissolved in ' et al., the spectrum contains several pyridine-d5. As described previously by Klier characteristic features that can be used to determine the composition of the graft copolymer. The set of peaks that appear between 1.7 and 2.0 ppm corresponds to the or- methyl peak of the methacrylic acid and the MPEGMA units. The resonance corresponding to the Ot-methyl protons are split into three spectral features. The most intense peak occurs at 1.78 ppm, and may be attributed to syndiotactic monomer triads.2'3’4'5 The peak of the next highest intensity appears at 1.83 ppm, and arises from atactic monomer triads. Finally, a relatively weak peak that corresponds to isotactic triads2 is observed at 1.9 ppm. In this investigation we use the NMR spectrum to determine the copolymer concentration, therefore all three peaks will be integrated together to determine the total number of (tr-methyl protons in the copolymer. The second region of interest in the 1H-NMR spectrum lies between 2.4 and 2.8 ppm where peaks arising from the methylene groups in the methacrylic acid backbone are observed. Starting at high field, four consecutive broad peaks arising from the methylene are observed: two relatively intense peaks (at 2.47 and 2.61) and two relatively weak peaks (at 2.71 and 2.74 ppm). Again, in this contribution, all four of these peaks are integrated to determine the total number of methylene protons in the copolymers. The third region of interesting in the lH NMR spectrum lies between 3.5 and 3.7 ppm where a single, relatively narrow peak attributed to the ethylene segments in the grafts is 57 obse are i ofel 1116 r metl 10 n ther first 0pc: fract 10 El mole an at 31635 areas 111111 r the d3 of the “1111 r3 observed. This peak is not split since the resonance of the oligo(ethylene glycol) grafts are insensitive to tacticity. Three sets of very weak resonance arise from trace quantities of ethanol solvent that remain in the sample. Specifically, a triplet observed at 1.20 ppm, the quartet at 3.75 ppm and a triplet at 3.35 ppm correspond to the methyl group, the methylene group and the hydroxy group of the residual ethanol.° Finally, it is interesting to note that a single sharp peak is consistently observed at 3.23 ppm, and is assigned to the methoxy endgroup of MPEGMA.6 The NMR data collected for the P(MAA-g-EG) copolymers are shown in Table 4.2. In this table, the first column corresponds to the sample designation (recall that the first number in this designation corresponds to the weight percent of the ethylene glycol repeat units in the monomer mixture, while the second number corresponds to the weight fraction of the initiator relative to the total monomer). The second column lists the MAA to EG repeat unit ratio in the monomer mixture fed to the reactor. Note that each MAA molecule contains one MAA repeat unit, while each MPEGMA macromonomer contains an average of 22.7 EG repeat units. Columns three, four and five list the relative peak areas for the (rt-methyl, methylene and ethylene peaks, respectively. These NMR peaks areas are normalized such that the sum of the integrals for the three regions of interest is 100. Finally, the last two columns list the experimental values for the MAA to EG repeat unit ratio in the graft copolymer. The data the in the last two columns are calculated from the data in columns three, four and five. For example, if P represents the ratio of the area of the (tr-methyl peaks to the area of the oligo(ethylene glycol) peak, the MAA:EG repeat unit ratio in the copolymer can be calculated using the following equation: 58 This MAA repeat MPEGMA nI the area of MAA:EG re equation: Agair per MM re] MPEGM A n‘I Exalnination Calculated Us Slightly smar “insistent di broilCler than mwwsm. ettpen'memaI 118111ng (1‘11 "MAA : i . 1 (4'2) This equation accounts for the facts that there are three OL-methyl protons per MAA repeat unit, there are 4 ethylene protons per EG repeat unit, and that each MPEGMA molecule contains an a-methyl group. Similarly, if Q represents the ratio of the area of the methylene peaks to the area of the oligo(ethylene glycol) peak, the MAA:EG repeat unit ratio in the copolymer can be calculated using the following equation: sang“; (43> n ,1.“ 22.7 Again, this equation accounts for the facts that there are two methylene protons per MAA repeat unit, there are four ethylene protons per EG repeat unit, and that each MPEGMA molecule contains a (it-methyl group. Examination of the data in Table 4.2 reveals that the MAA to E6 repeat unit ratio calculated using the area of the peaks for the methylene backbone group is consistently slightly smaller than the value obtained using the area of the a-methyl peaks. This consistent difference likely arises from the fact that the methylene peak is significantly broader than the oc-methyl peak, therefore more of the signal could be lost in the baseline. However, the agreement between the two numbers is generally good, and is within the experimental accuracy of the technique. Therefore, we will average the values obtained using the oc-methyl and the methylene in our further discussions. 59 repé nor the I ratio diag This of rr for e eye; as a an it incor Quiel Well 4.5.3 Chain the ( alwa- The data in Table 4.2 illustrate that the experimental values for the MAA to EG repeat unit ratio in the copolymer are in good agreement with the known values for in the MAA:EG ratio in the monomer feed. The relationship between the compositions of the monomer feed and the final copolymer is illustrate in Figure 4.5, which contains a plot of the MAA:EG repeat unit ratio in the copolymer as a function of the MAA:EG repeat unit ratio in the monomer feed. Note that the data in Figure 4 generally lie close to the diagonal line that corresponds to equality of the polymer and the monomer compositions. This result is not surprising since the polymerization is practically a homopolymerization of methacrylic acid (for example, the 19.9:1 system contains 450 MAA monomer units for every MPEGMA molecule; the 9.7:] system contains 220 MAA monomer units for every MPEGMA molecule, etc). Therefore, the oligomeric macromonomer are dispersed as a dilute solution in the reaction mixture, and nearly every active radical center will be an MAA-terminated radical. It is extremely unlikely that a growing radical chain will incorporate two macromonomers in close succession (a PEGMA-tenninated radical will quickly become an MAA-terminated radical by propagation), therefore the grafts will be well distributed along the polymer chains. 4.5.3 Average number of grafts per chain The experimental results obtained from the gel permeation chromatography can be combined with the lH NMR spectroscopy results to estimate the number of grafts per chain for each of the polymer samples. Using the number average molecular weight from the GPC studies and the MAA:EG repeat unit ratio from the NMR experiments, the average number of grafts per chain may be calculated using the following equation: 60 Int C1111 res; obt are rear wei C011 dec The (let: em] the 83151 cm _ (4.4) "MPEGMA = Mn chain ”MAA n ' 22:7 ' M MAA 'I' M MPEGMA Hi In this equation, M: is the number average molecular weight as given by gel permeation chromatography MM,“ and Murrow represent the molecular weight of the methacrylic acid monomer and methoxy poly(ethylene glycol) monomethacrylate macromonomer, respectively. Table 4.3 contains results for the calculation of the number of grafts per chain obtained using Equation 4.4. As expected, for a constant initiator concentration, the average number of grafts per chain increases as the amount of the MPEGMA in the reaction mixture is increased (recall that the first number in the sample designation is the weight fraction of MPEGMA in the monomer mixture). In addition, for a constant MPEGMA content, the average number of grafts per chain decreases as the initiator concentration is increased. This trend arises from the fact that the molecular weight decreases with increasing initiator concentration. The average number of grafts per chain is expected to play an important role in determining the emulsification performance of the graft copolymers. Since the formation of the reversible hydrophobic blocks requires the presence of the grafts, an effective emulsifier should have an average of at least one graft per chain. It is noteworthy that three of the samples synthesized with high initiator concentrations (samples designated 2.4/10, 5/25, and 10/ 18.5) contain an average of less than one graft per chain. In these systems, the primary chain length is so low that many propagating radicals do not encounter a MPEGMA molecule before they terminate. These copolymers are not 61 expected to function well as emulsifiers since many of the chains will be the homopolymer PMAA. Several of the copolymers synthesized with intermediate initiator concentrations contain an average of one to two grafts per chain. In the complexed state, these chains would resemble diblock, triblock, tetrablock, or pentablock systems, depending upon the positioning of the graft or grafts along the backbone. Finally, many of the copolymers contain an average of three or more grafts per chain. These copolymers will resemble multiblock copolymers when the grafts are folded down and complexed with the backbone. In a future publication we will correlate the average number of grafts per chain with the effectiveness of the copolymer as a reversible emulsifier. 4.6 CONCLUSIONS In this chapter, I presented the synthesis and characterization of a series of poly(methacrylic acid-g-ethylene glycol) copolymers of varying molecular architectures. A semi-batch reactor was designed to synthesize the copolymers by a free radical copolymerization of methacrylic acid with the macromonomer methoxy poly(ethylene glycol) methacrylate. The resulting copolymers were characterized using lH-NMR to determine the polymer composition, and gel permeation chromatography to characterize the polymer molecular weight distribution. The GPC results revealed that, for a given MAA:EG repeat unit ratio, the number average molecular weight is proportional to the inverse square root of the initiator concentration. In addition, for repeat unit ratios ranging from infinity (pure PMAA) to 4.6 (a graft copolymer with a graft occurring about every 100 backbone repeat unit), the polydispersity was found to remain essentially constant, with a value typical for free 62 radical chain polymerizations. The experimental polydispersity values suggest that chain transfer to polymer does not occur to a significant extent. The GPC results also indicate that the addition of the oligo(ethylene glycol) macromonomer leads to a systematic deviation from the results obtained for pure PMAA with the data consistently falling above the 99% confidence interval at low initiator concentration (high molecular weight). This trend suggests that the presence of the oligo(ethylene glycol) grafts leads to an increase in the hydrodynamic volume of the dilute polymer chains, possibly due to the excluded volume introduced by the oligomeric grafts. The 1H NMR results illustrated that the experimental values for the MAA to EG repeat unit ratio in the copolymer are in good agreement with the known values for in the MAA:EG ratio in the monomer feed. The GPC and NMR results were combined to estimate the number of grafts per chain for each of the polymer samples. The average number of grafts per chain we found to range from 0.3 to 18. As expected, for a constant initiator concentration, the average number of grafts per chain was found to increase as the amount of the MPEGMA in the reaction mixture was increased. In addition, for a constant MPEGMA content, the average number of grafts per chain decreased as the initiator concentration was increased. 63 5.6 - 5.41 5.2 - 5.01 4.83 4.6L (109(Mp) 4.4 - I PL poly(acrylic standards) , first order polynomial fit 40 .1 - - - - second order polynomial fit 4.2 -l 3.8 - I ' l . . , . . . 14.5 15.0 15.5 16.0 16.5 17.0 17.5 retention time [min] Figure 4.1 Calibration curves prepared using commercially available monodisperse poly(acrylic acid)-sodium salt standards with molecular weights of 7,500, 16,000, 28,000, 62,900, 115,000, and 272,900. 64 Nor Wei con non 1 75k 150k- :0" g 125k- Ev 2‘100k- z . O) '5 75k- E to 3 501(1 0 2 1 g 25k~ 0 I I ' f l I ‘ l 0 2 4 6 8 10 Initiator concentration [wt.%] Figure 4.2 Number average molecular weight Mn of poly(methacrylic acid) as a function of the weight percentage of sodium persulfate initiator fed into the reactor. The curves correspond to the upper and lower limit of the 99% confidence interval determined by nonlinear least squares fit to Equation (4.1). 65 XL co; the 1111: 1 75k 150k- PMAA (nongrafted) Copolymer composition + MAA:EG = 20:1 A MAA:EG = 10:1 v MAA:EG = 5:1 D MAA:EG = 2:1 or .1D .4 .3 8 31 74' x l l Molecular Weight Mn [g/mole] 75k - 50k - 25k - 0 . l . , . . . l . . 0 2 4 6 8 1 0 Initiator concentration [wt.%] Figure 4.3 Number average molecular weight Mn of poly(methacrylic acid-g- ethylene glycol) copolymers as a function of the weight percentage of sodium persulfate initiator fed into the reactor. The curves correspond to the upper and lower limit of the 99% confidence interval determined by nonlinear least squares fit of Equation (4.1) to the data for pure PMAA (as in Figure 4.2). 66 0 0 3 5 3 0 2.5 2.0 I. 5 1 0 "I 1.7—r L—T—__l L__+_J 9.26 33 40 57.34 Figure 4.4 lH-NMR spectrum of P(MAA-g-EG) 5/10 from 0.8 to 4.2 ppm. The relative values for the integrated areas for the intensities of interest are also shown. Proton assignments are given in the text. 67 22— 20.4 18‘ 10 wt.% 5 wt.% 1 wt.% - 0.5 wt.% OD molar ratio MAA:EG in polymer o m a a) co 8 fi 3‘ 6‘: I 4 l a l l l l J l 4 l .,.,.,.,.,.,.,.,.,.,., 6 810121416182022 feed molar ratio MAA:EG O N A Figure 4.5 MAA to EG repeat unit ratio in the graft copolymer as a function of the MAA to EG repeat unit ratio in the monomer feed. The diagonal line corresponds to equality of the polymer and monomer composition. 68 Table 4.1 Gel permeation chromatography results for the moments of the molecular weight distribution of poly(methacrylic acid-g-ethylene glycol) copolymers. saw“ MAA:EG M., PDI Designation F eed Ratio M‘" 0/10 Na 17000 59000 3.4 0/ 5 Na 33000 1 17000 3.6 00.6 Na 40000 142000 3.5 0/1 Na 66000 248000 3.7 0/0.5 Na 122000 564000 4.6 2.5/10 19.9 17000 58000 3.5 2.5/5 19.9 44000 153000 3.5 2.5/1 19.9 116000 323000 2.8 2.5/0.5 19.9 126000 304000 2.4 5/25 9.7 9600 32000 3.3 5/10 9.7 26000 125000 4.9 5/5 9.7 25000 94000 3.8 5/2.5 9.7 61000 189000 3.1 5/1 9.7 1 19000 303000 2.5 5/0.5 9.7 165000 364000 2.2 10/18.5 4.6 7700 23000 3.0 10/ 10 4.6 15000 54000 3.7 10/ 5 4.6 25000 96000 3.8 10/4 4.6 31000 120000 3.8 10/1.7 4.6 80000 244000 3.0 10/1 4.6 32000 271000 3.3 20/10 2.0 24000 1 l 1000 4.7 20/10 2.0 25000 1 15000 4.6 20/5 2.0 25000 102000 4.1 20/3.25 2.0 41000 169000 4.1 20/0.5 2.0 83000 315000 3.8 69 Table 4.2 lH-NMR results for the MAA to EG repeat unit ratio in the poly(methacrylic acid-g- ethylene glycol) copolymers. Sample MAA:EG integral of Integral of integral of MAA:EG MAA:EG designation ratio in a-methyl methylene ethylene ratio in polymer ratio In monomer peaks peaks peak from a-mothyl pofll’ygor methylene 0/10 oo 61.4 38.6 - oo 00 0/5 00 62.3 37.7 - 00 00 0/2.6 00 63.6 36.5 - 00 00 0/1 00 62.8 37.2 - do 00 0/0.5 00 62.0 38.0 - oo oo 2.5/10 19.9 61.0 35.2 3.8 21.4 18.5 2.5/5 19.9 59.9 36.4 3.7 21.3 19.4 2.5/1 19.9 59.5 37.0 3.5 23.0 21.4 2.5/0.5 19.9 59.8 36.4 3.7 21.3 19.4 5125 9.7 58.9 35.6 5.4 14.4 13.0 5/10 9.7 57.3 33.4 9.3 8.2 7.1 5/5 9.7 57.0 35.3 7.8 9.8 9.0 5/2.5 9.7 57.4 35.2 7.5 10.2 9.3 5/1 9.7 56.5 35.8 7.7 9.7 9.2 5/0.5 9.7 57.5 35.6 7.0 11.0 10.1 10/18.5 4.6 55.3 32.8 12.0 6.1 5.4 10/10 4.6 52.6 33.1 14.4 4.8 4.5 10/5 4.6 52.8 31.8 15.4 4.5 4.0 10/4 4.6 53.6 32.1 14.3 5.0 4.4 10/1.7 4.6 53.6 32.3 14.1 5.0 4.5 10/1 4.6 54.0 33.1 12.9 5.6 5.1 20/10 2.0 45.3 26.3 28.4 2.1 1.8 20/10 2.0 45.1 26.5 28.4 2.1 1.8 20/5 2.0 45.7 27.8 26.5 2.3 2.0 20/3.25 2.0 46.2 27.6 26.2 2.3 2.0 20/0.5 2.0 43.5 26.2 30.3 1.9 1.6 7O Table 4.3 Average number of grafts per chain calculated based upon a combination of the GPC and the NMR experimental results. Sample M., nun/nee Average number of Designation from GPC from NMR J[alts per chain 0/ 10 17000 00 0 0/5 33000 00 0 012.6 40000 00 0 0/1 66000 00 0 0/0.5 122000 00 0 2.5/10 17000 20.0 0.4 2.5/5 44000 20.3 1.1 2.5/1 1 16000 22.2 2.6 2.5/0.5 126000 20.3 3.1 5/25 9600 13.7 0.3 5/10 26000 7.7 1.6 5/5 25000 9.4 1.3 5/2.5 61000 9.7 3.0 5/1 1 19000 9.4 6.1 5/0.5 165000 10.5 7.6 10/18.5 7700 5.7 0.6 10/10 15000 4.7 1.5 10/5 25000 4.3 2.6 10/4 3 1000 4.7 3.0 10/1.7 80000 4.8 7.7 10/1 82000 5.3 7.2 20/ 10 24000 1 .9 4.9 20/10 25000 1.9 5.] 20/5 25000 2. 1 4.7 2013.25 41000 2.2 7.7 20/0.5 83000 1 .8 18.4 71 4.7 REFERENCES Klier J ., Scranton A. B., Peppas N. A., Macromolecules 23, 4944 (1990). Klesper E., Gronski W., Polymer Letters, 7, 727 (1969). Klesper E., Johnsen A., Gronski W., Wehrli F. W., Makromol. Chem. 176, 1071 (1975). Klesper E., Johnsen A., Gronski W., Polymer Letters, 8, 369 (1970). Lyeria J .R., IBM J. Res. Develop, 111 (1977). Pouchert C. J ., The Aldrich library of NMR spectra. Milwaukee, Wisconsin, USA: Aldrich Chemical Co., 1983. 72 Chapter 5 SOLUTION BEHAVIOR OF POLY(METHACRYLIC ACID-G-ETHLENE GLYCOL) COPOLYMERS IN WATER 5.1 INTRODUCTION In this chapter, the aqueous solution behavior of poly(methacrylic acid-g-ethylene glycol) copolymers is investigated. The chapter begins with a brief discussion of the theory behind dynamic light scattering to provide background necessary for the interpretation of the experimental results. In the first experimental section, dynamic light scattering studies are reported to characterize the aggregate size of these copolymers in the compact coil conformation. In a second experimental section, the solution behavior (as characterized using potentiometric titration) of a series of copolymers with various grafting levels of PEG is discussed as a function of various degrees of neutralization. In particular the pH-induced conformational transition from a compact coil to an extended form has been investigated. 5.2 DYNAMIC LIGHT SCATTERING THEORY Data obtained from dynamic light scattering measurements can provide information about motion and dynamics of individual macromolecular chains in 1,2 solution. When light impinges on matter, the electric field of the light induces an oscillating polarization of the electrons in the molecules.3'4 The molecules then serve as 73 secondary sources of light and thereby emit scattered light. The frequency shifts, the angular distribution, the polarization, and the intensity of the scattered light are determined by the size, shape and molecular interactions in the scattering material.4 Thus, from the light-scattering characteristics of a given system, detected with the aid of electrodynamics, and analyzed using the theory of time-dependent statistical mechanics, information about the structure and molecular dynamics of the scattering medium can be obtained.‘ If the scattering center moves relative to the light source, the frequency of the scattered light is shifted from the incident frequency by an amount proportional to the velocity component of the scatterer perpendicular to the direction of the light beam. Such very small, time-dependent frequency changes, are detectable with the aid of modern photomultipliers. These translational fluctuations are often rather fast, i.e. in order of 10'6 seconds. They form the basis of quasi-elastic light scattering caused by translational or rotational degrees of freedom.4 According to classical theory,4 the electrons and nuclei in a given particle oscillate about their equilibrium positions in synchrony with the electric vector of the incident radiation. If the incident light wave is being transmitted along the x-axis and the electronic field vector is in the y-direction, then a fluctuating dipole will be induced in the 2 Each oscillating dipole is itself a source of particle along the y-direction. electromagnetic radiation. The net result of this interaction of light and a scattering particle is that some of the energy that was associated with the incident ray will be radiated in directions away from the initial line of propagation (see Figure 5.1a). Thus, 74 the intensity of light transmitted through the particle along the incident beam direction is diminished by the amount of radiation in all other directions by the dipoles in the particle. Classical electromagnetic theory shows that the intensity of light radiated by a small isotropic scatterer iszz'4 10 =%¥—10(1+c0520) (5.1) where 19 is the light intensity at a distance r from the scattering entity and 0 is the angle between the direction of the incident beam and the line between the scattering center and detector. In this equation, 10 is the incident light intensity, 7» is the wavelength of the incident light and 01 is the excess polarizability of the particle over its surroundings. The polarizability relates the magnitude of the induced dipole moment to the intensity of the incident electric field strength. Figure 5.1b shows a graphical representation of the angular distribution of the intensity of scattered light for vertically polarized incident light. In large molecules, defined as molecules with a size that is greater than a twentieth of the incident wavelength,5 dipoles are induced in different parts of the same macromolecule. The scattered light from the different points in the molecule reaches the detector with different phases. If two beams, which are out of phase, interfere, the intensity of the resulting radiation is smaller than the sum of intensities of light scattered by all the individual mass points of that particle.4’5 This phenomenon is called intramolecular interference. The scattered intensity registered by the detector is that resulting from the individual partial intensities and the corresponding phase shifts (see Figure 5.2). The phase shift increases with increasing angle of observation.6'7 75 The molecules in the illuminated region are perpetually translating, rotating and vibrating by virtue of thermal interactions.2 Because of this motion the positions of the charges are constantly changing so that the total scattered electric field at the detector will fluctuate with time. Dynamic or quasi-elastic light scattering from polymers in solution arises from concentration fluctuations within the scattering volume. The fluctuations of the intensity of scattered light are more frequent as molecules move faster in a solution. Implicit in these fluctuations is important structural and dynamical information about the positions and orientations of the molecules. Since molecular motion is erratic, a recording of the scattered field will look very much like a noise pattern, as shown in Figure 5.3a. Hence time dependent correlation functions adapted from the theory of stochastic process are used to describe dynamic variables “hidden” in the apparent random fluctuation data. The motion of the macromolecules in solution is analyzed by the construction of a time correlation function gI (t) of the scattered electric field. In fact, in the dynamic light scattering experiments, the time autocorrelation function of the scattered light intensity 62 (t) is measured: G, (r) = (1(O)I(t)) (5.2) In this time correlation function, the scattering intensity at time zero is compared in a discrete manner with that at an overall delayed time rd varying normally between 10'6 to 103 seconds. About 105 comparisons are needed to get a properly averaged correlation function. The sampling time interval, At, can range anywhere between 12.5 ns and Is.5 76 The autocorrelation function is a measure of the similarity between two noise signals I(t=0) and I(t=At). When At=0 these two signals are completely in phase with each other and (I (0)1 (1)) is large; whereas when At increases the noise signals get out of phase with each other and the autocorrelation function is small.4 Thus, with growing delay time, the average product of photon counts from the time intervals shows behavior of an exponentially decaying function which at long delay times eventually approaches a base line, as shown in Figure 5.3b. When At is extremely short the product will have a value close to the averaged square of the static LS intensity, <12>. However the base line represents the squared static LS intensity, 2. These intensity fluctuations are related to fluctuations of the corresponding electric field. And therefore the correlation function of the scattered light intensity G2 (t) has to be converted into the correlation function g,(t) of the scattered electric field. 02 (t) can be converted into a correlation function g1(t) of the scattered electric field using Siegert’s relationship:8 g2(t)=1 + file/(OF (5.3) with G __(___,1) (5.4) g. —1 lg ()|= fl Glob ) where 02(00) is the experimentally determined baseline. g2(t)=Gz(t)/Gz(oo), and B is an arbitrary constant. In order to present measured correlation functions G;(t) in a more suitable fashion for comparison the following normalization procedure is commonly used:5 77 Gz(t)—Gz(°°) = 02(1) —1 (5-5) G2(°°) Gz(°°) 162 (t )1 = In many applications the autocorrelation function decays like a single exponential. The correlation time 1: represents the characteristic decay time of the property. Finally, the translational diffusion coefficient of the spheres is related to the decay constant 1:.4’8 For monodisperse particles and small in diameter compared to the wavelength of the light (but also hard spheres of any size) it can be shown that the following equations hold:4’5 102(01 =e-r/r’ : e—r'r = e-Iyqzt (56) igl(t)i :e—t/r :e-rr ze-m’r (5,7) _ 1 _ 1 (5.8) T — F — qu q = 47010 sing (5'9) 2, 2 Here 1: and t’ are relaxation times, D and D’ are translational diffusion coefficients, and I‘ and I" are relaxation rates. Here q is the magnitude of the scattering vector and can be defined by the scattering geometry. The symbols no and lo refer to the refractive index of the solvent and the incident wavelength, respectively. For small, dilute, non-interacting spheres the Stokes-Einstein Equation gives the following relationship between the diffusion coefficient and hydrodynamic radius:2’4 78 __ kT (5.10) 637701811 where T and k represents the temperature of the solution and Boltzmann constant, respectively, and 110 represents the viscosity of the pure solvent. The hydrodynamic radius, Rh, is defined by the Stokes-Einstein equation. The hydrodynamic radius differs from the radius of gyration because of the hydrodynamic interactions among the segments of a macromolecule. The hydrodynamic radius of a macromolecule, K.,, is the radius of a non-interacting hard sphere having the same diffusion coefficient as the macromolecule under investigation. In many cases simple single exponential behavior is not observed for the g1(t) vs. time curve and the presence of superimposing multiple exponentials is apparent. These deviations may result from a degree of polydispersity, because in this case each species contributes its own exponential, according to its specific diffusion coefficient. In these 9,10 cases a cumulative expansion is made: 1 5 l 5.11 F212 —§I‘3r3 +... ( ) lng,(t)=lnA—I‘lt+ Analysis of the cumulative expansion of the correlation function is usually performed by fitting a polynomial up to the third order to the function In(g2(t)-1). The polynomial coefficients are converted into the coefficients of the cumulated expansion of the field correlation function. Under the assumption that the scattering particles behave as hard spheres in dilute solution and within the Rayleigh-Debye4 theory the particle radius distribution function is calculated from the decay time distribution function using the following equation:8 79 kT 2 (5.12) 5.3 DETERMINATION OF AGGREGATE SIZE OF P(MAA-G-EG) IN AQUEOUS SOLUTION USING DYNAMIC LIGHT SCATTERING 5.3.1 Sample Preparation All glassware was thoroughly cleaned with acetone, which was evaporated using a dried air stream before use. To prepare each sample, 0.25 g of polymer was placed into 50 mL of HPLC grade water (taken from a freshly opened container) to give a 0.5 wt.% solution. The P(MAA-g-EG) copolymers used were all copolymers with a grafting level of 2.5 and 5 wt.% and the initiator concentration of 10, 5, 2.5 and 0.5 wt.%. In addition, the homopolymer PMAA polymerized with an initiator concentration of 0.5 wt.% was used. The polymers were dissolved in the HPLC grade water under constant stirring. After all polymer solids were dissolved, the solutions were titrated with 4M HCl to a solution pH value of 2.0. The sample vials remained unopened for two weeks until the dynamic light scattering experiments were carried out. 5.3.2 Data Analysis Vertically polarized Laser light (532nm) from a He-Ne Laser (coherent, enterprise series, Palo Alto, CA) was scattered from the 0.5 wt.% copolymer solutions. The copolymeric solution volume was typically about 1 ml and was therrnostatted at 25°C in a polished borosilicate glass cuvette. Prior to filling, the cuvettes were inspected to be free of cracks and were thoroughly cleaned with acetone to remove all dust particles. 80 Light scattered from the sample was detected with a photomultiplier tube (PMT). The Laser power was regulated with a potentiometer such that the count rate did not exceeded a limit of about 150 kHz to protect the PMT from damage. For all measurements the count rates varied in the limits of 70 to 150 kHz. The PMT was mounted onto a computer controlled lever arm that could rotate into different angles relative to the direction of the incident light. For these studies, the dynamic light scattering was measured at angles of 30°, 45°, 60°, 90°, 135°. The signal pattern from the PMT was analyzed using the ALV-SOOO/E software package8 (ALV-Laser Vertriebsgesellschaft, Langen, Germany). The two main tasks of the software are to calculate the correlation function of the scattered light and to analyze the data using a nonlinear fit model to calculate the decay rate distribution function. The ALV-SOOO/E Multiple Tau Digital Correlator was used to perform real-time computation of correlation functions with a fixed range of simultaneous lag times that varied between 12.5 nano-seconds and 1000 milli-seconds. The total acquisition time for each measurement run was 15 or 30 seconds. Typically five to 10 of these correlation functions were averaged. It was conclusive from the real time correlation function studied that there is greater deviation in scattering for light observed at lower angles such as 30° and 45°. Hence for these angles acquisition times of 30 seconds each were averaged, while for any higher scattering angle, runs of 15 seconds acquisition time were averaged. Charged particles such as polyelectrolytes have a tendency to attract dust particles. If dust particles pass through the scattering volume, irregular patterns sweep through the online detector signal. If this occurred, the measurement was discarded. 81 A typical plot of the normalized correlation functions of the scattered light intensity as a function of the lag time is shown in Figure 5.4. As can be seen in the figure, all correlation functions decay in a non-linear fashion. This non-linear decay is a clear indication that the samples are not mono-disperse. In the case of a mono-disperse sample such as a non-aggregated protein or a very narrow fraction of a synthetic polymers, the decay should resemble a straight line on this semi-log plot. Therefore, one conclusion that can be drawn from Figure 5.4 is that the distribution size of the P(MAA- g—EG) copolymer aggregates is not narrow. Each curve in Figure 5.4 corresponds to a distribution of relaxation times rather than a single relaxation time. It can be noted from Figure 5.4 that the quality of the data for the normalized correlation function 1G2(t)1-1 depends upon the scattering angle. Specifically, for the higher scattering angles 90° and 135° the data exhibits more scatter than the data for the three smaller angles. Physically, there are several independent reasons that may cause the deviation of the logarithmic normalized correlation function from a straight line. In general, these results may result from a degree of polydispersity, fluctuations resulting from internal molecular flexibility or slow motion of intermolecular aggregates.4 In addition, broadness of the decay time distribution may be introduced from the fitting process, if the sample has a bimodal hydrodynamic size distribution with two close peaks, or from the fitting uncertainty, if the sample has broad physical size.5 Standard regularization techniques were used to fit a distribution function for the relaxation times of the autocorrelation ftmction. For molecular size distribution analysis, the CONTIN routine,9"° accessible through the ALV software, was used. The fitting routine CONTIN limits the number of solutions by using a number of different 82 regularization parameters and tries to optimize those parameters by some statistical tests. The ALV-NonLin Data Analysis8 fits an integral type model function to the correlation function using a constrained regularization method. In this case, the following nonlinear fit model is used: r... (5.13) g2(t) —1 = [ Ie’"G(I‘)dI‘] Fm... where 6(1) denotes the decay rate distribution function. The fitting function was regularized to the order of zero. Zero order regularization corresponds to minimization of the area under the decay rate distribution function G(I‘). The number of grid points for the fitting was typically set to 50. The time limits of the decay time distribution function were set to include completely the lag time range where the correlation function decays from its initial value to the baseline to ensure that the fitting function contained all information of the correlation function. The ALV- NonLin data analysis results in a discrete distribution function of logarithmically equidistant spaced decay times t=1/I‘ which is the adequate solution for the logarithmically spaced lag time grid of the correlation function. Using Excel Spreadsheets the radii are calculated from the corresponding decay time values applying the Stokes-Einstein relation as discussed in section 5.2. The relevant constant parameters for the Stokes-Einstein equation are: Temperature [K] : 298.00 Viscosity (cP): 0.5329 Refractive Index (-): 1.332 Wavelength [nm]: 532 83 A brief overview of the data shown in Figures 5.5 through 5.7 illustrates the care that must be taken in analyzing dynamic light scattering data. In Figure 5.5, the distributions of hydrodynamic radii are shown for the P(MAA-g-EG) 5/5 sample in aqueous solution at pH 2. The figure contains five distributions corresponding to the five different angles used in the light scattering experiments. Comparing the distributions from the highest observation angle to the lowest, the maximum of the distribution shifts from a value of 128 nm (for an angle of 135°) to a value of 171 nm (for an angle of 30°). This shift in apparent size is typical for the angle-dependence of scattering from large particles. As explained in section 5.2, intramolecular interference is causing the angular distribution of the scattered intensity. Hence the signal is diminished at the higher angles of 90° and 135° and that may lead to a further uncertainty in the fitting process. As will be shown later in this chapter, aggregative size is best estimated by extrapolating the results to a zero angle (a hypothetical angle of 0°). In Figure 5.6 the hydrodynamic radius distribution for sample P(MAA-g-EG) 2.5/5 for all investigated angles is shown. In this case, the dynamic light scattering results clearly indicate that the distribution is bimodal since the distribution is bimodal at all angles. The figure illustrates that the halfwidth of the mode corresponding to a higher hydrodynamic radius is decreasing with decreasing scattering angle, however this effect should not be taken to accurately indicate the actual breath of the distribution. In Figure 5.7 the hydrodynamic radius distribution for sample P(MAA-g-EG) 5/1 for all investigated angles is shown. The hydrodynamic radius distribution appears to be single modal for observation angles of 135° through 60°. For the lowest angles the distribution seems to have three modes. This example shows that it is important to compare light 84 scattering data of the same sample detected at different angles. The appearance of three modes may be attributed to an artifact of the fitting procedure coupled with the broad aggregate size distribution. In general, fitting analysis does not fit well such distributions, ‘0 and the two “phantom” modes that appear for the two smallest angles are most likely an indication of the breadth of the distribution. If the modes were real, they should appear at all angles. In order to analyze the scattering angle dependent hydrodynamic radius distributions, in a final step, an extrapolation of the modes as a function of the scattering angle to a "zero" angle is made for each sample. Linear regression of the modes as a function of the scattering angles gives a linear fitting equation that is used to calculate the hydrodynamic radius at a theoretical angle of 0°. At a projected scattering angle of zero, which is physically not measurable since the incident light intensity will overcast the scattering, the effect of angle-dependent scattering will be minimized. The linear regression was used at least for the data collected under the angles of 30°, 45° and 60°. If the mode peaks for the higher hydrodynamic radii for the two highest scattering angles fit reasonably well to the projected line through the data points for the first three angles, all data points were fitted. In Table 5.1, the results of the linear regression of the modal peaks for the higher hydrodynamic radii as a function of the scattering angle are given. Tables 5.1a and 5.1b show the light scattering data for the investigated P(MAA-g-EG) copolymers. Table 5.1a shows data for the P(MAA-g-EG) copolymers 5/10, 5/5, 5/0.5, 2.5/10, 2.5/1 and 2.5/0.5. Table 5.1b shows data for the P(MAA-g-EG) copolymers 0/0.5 and 5/ 1. The first and second columns in each Table provide the designation of the polymer sample and the modality of the distribution for the P(MAA-g-EG) copolymers. 85 Column four and five show the modes of the radii distribution measured for each angle (column three) ranging from 30 to 135°. Column six gives the radius extrapolated to an angle of “zero” for the larger mode of the bimodal distribution (or the only mode of a single-modal distribution) based on the data from column five. The smallest absolute value for the correlation coefficient of the linear regression R (given in column seven) was used as the criteria for the best fit. The projected hydrodynamic radius at "zero" angle is used to describe the aggregate size. Further analysis in terms of distribution width or number average aggregate sizes would lead to over-interpretation of the dynamic light scattering data for the reasons explained previously. In order to probe any dependencies of the aggregate size on the molecular weight or composition of the investigated P(MAA-g-EG) copolymers, the apparent aggregate sizes (at zero angle) are plotted versus the number average molecular weight for each sample in Figure 5.8. In Figure 5.8, the solid black square symbols correspond to the P(MAA-g-EG) copolymers with an EG content of 2.5 wt.% of total polymer mass (20 PMAA repeat units for every PEG repeat unit). It can be concluded from the plot that, for this sample, the hydrodynamic radius increases with increasing number average molecular weight. It is interesting that, for this modest grafting level, the single data point for the homopolymer poly(methacrylic acid) falls on the best-fit line. With increasing chain lengths the single polymer occupies more space and therefore, the aggregate gains size, as well. The polymer containing a relatively large number of grafts exhibits a different trend. As shown in Figure 5.8, the light scattering data indicates that the P(MAA-g-EG) copolymers having a grafting level of 5 wt.%, exhibit an the aggregate size that decreases 86 slightly with increasing number average molecular weight. This effect may be explained by the fact that the increased grafting levels imparts a higher hydrophobicity into the P(MAA-g-EG) copolymers, and therefore may lead to a more compact coil in aqueous solution. Since the polymer concentration was constant at 0.5 wt% for all investigated solutions, the aggregates of the P(MAA-g-EG) copolymers having a grafting level of 5 wt.% exhibit smaller radii than the aggregates of the P(MAA-g-EG) copolymers having a grafting level of 2.5 wt.%. Again, this effect arises from the more hydrophobic nature of the more highly grafted chains under acidic conditions when the hydrophobic complexes are formed. 5.4 POTENTIOMETRIC TITRATION OF AQUEOUS SOLUTIONS OF P(MAA-G-EG) COPOLYMERS 5.4.1 Experimental Methods All the glassware was thoroughly cleaned with Alconox detergent in reverse osmosis (RO) de-ionized water followed by rinsing with methanol and drying with a purified air jet stream. To prepare the samples for potentiometric titrations, 0.5 g of dried finely ground copolymers were mixed with 50 ml Of HPLC grade water in a 2 02. glass vial to give a 1 wt.% solution. The mixture was shaken with a laboratory shaker for about three consecutive days to ensure dissolution. To perform the titrations, each solution was transferred to a 100 ml beaker and was well mixed by the use of a magnetic stirrer. The pH was measured using an Accumet pH meter 910 equipped with the Accumet electrode 13-620-285, both from Fisher Scientific (Pittsburgh, PA). Two 50 ml 87 burrets were filled with a 1M sodium hydroxide solution and a ca. 0.94 M hydrogen chloride solution. The base solution was formed by dissolving 8 g of NaOH pellets (J. T. Baker, Inc., Philippsburg, NJ) with 200 mL of HPLC grade water. The acidic solution was prepared by mixing 16.28 ml of an aqueous solution of hydrogen chloride, which was reported in the concentration range of 35.5 to 38 wt.% with 183.8 ml of HPLC grade water. Since sample 0/10 with an initial pH of 2.40 was titrated to neutrality from the initial pH of 2.40 with 3.3 m1 base and titrated back to an pH value of 2.42 with 3.5 ml acid, the molarity of the acid titration solution was determined to be 0.94 M. Under well stirred conditions, the copolymeric solutions were titrated with base to the neutral point at pH 7.00. After the neutral point was reached, the total amount of base added used was noted. Then the polymer solution was titrated back with hydrogen chloride to acidic conditions (pH ~2.4). All titrations were performed in the ambient atmosphere, therefore C02 from the air was allowed to dissolve and form carbonic acid in the aqueous solutions. 5.4.2 Solubility behavior during titration of P(MAA-g-EG) in aqueous solution In the following section the macroscopic solution properties Of copolymers that have the same initial composition but differ in average molecular weight are compared and contrasted. All copolymer samples that had an ethylene glycol content of 2.5 wt.% were readily dissolved after stirring. The solution had a clear appearance and no solids could be observed. The transparent appearance remained throughout the titration. In contrast to that, the samples of P(MAA-g-EG) 5/0.5, 5/1 and 5/5 with an ethylene glycol content of 5 wt.% appeared slightly bluish from the Tyndallll effect. The Tyndall effect 88 arises from the scattering of light as it passes through a colloid due to the presence of suspended particles, and indicates that the particle size is on order of 0.1 micron. The copolymer solution Of P(MAA-g-EG) 5/ 10 appeared turbid (pale white) but not bluish. During titration of the sample of P(MAA-g-EG) 5/10 at an apparent pH value of about 4.6 the solution becomes significantly clearer, and clearer yet as the pH value approaches 5.3. For the complete series of 5 wt.% EG it the solution clears completely during titration with base as the pH value closes in to neutrality. It was clear that the copolymers containing 10 wt.% EG were considerably more hydrophobic since the samples contained undissolved solids under acidic conditions in solution (except the sample 10/1, which appeared turbid without any solids at the bottom). The sample of P(MAA-g-EG) 10/ 10 had a large agglomerate of solids in an otherwise clear solution. Typically the turbidity Of the solution cleared during the titration, and all solids dissolved at a pH value in the range of 5.5 to 5.8. Other interesting observations during the titration are that there was a significant buffering effect from the presence of the copolymer with the apparent pH value of the solution remaining at a pH value of 5.71 for some time during the titration. In addition, the viscosity was found to increase dramatically at a pH Of 5.75. This thickening effect was confirmed by previous viscosity studies of P(MAA-g-EG) in aqueous solution.12 It was concluded that the tremendous increase of the viscosity Of the copolymers in aqueous Solution is associated with the uncoiling of the polymer chains due to charge-charge repulsion of the negatively charged dissociated carboxylic groups. In addition, the breaking of the complex that allows the grafts to extend may enhance this effect. This thickening behavior could be reversed by titrating back to acidic conditions. 89 The series of the samples containing 20 wt.% EG (corresponding to an MAA to EG repeat unit ratio of approximately 2 to 1) were very hydrophobic and precipitated out of aqueous solution. The P(MAA-g-EG) 20/ 10 and 20/5 copolymers agglomerated into one swollen gel-like mass. Titration of these samples was particularly challenging due to the diffusion time Of the hydroxyl ions into the mass (after the addition Of an aliquot of base, the pH in the external solution would rise with time as the ions diffused into the mass). Again, at a pH value of around 5.8, the polymer samples dissolved. During the back titration with acid, the complex would immediately form in the vicinity of the acid drop as it descended into the sample. For many polymers, these local precipitates were somewhat robust since they generally did not readily redissolve (under acidic conditions the dissolution kinetics were slow) but didn’t form large conglomerates because they were dilute. Since copolymers that have a PEG content of 10 wt.% and above precipitated out under conditions acidic, they are not good candidates for reversible emulsifiers. The precipitated polymers will typically remain as a solid phase will not partition as onto the interface when a nonpolar phase is added. However, the pH-dependent solution properties copolymers with an EG content of 2.5 and 5 wt.% makes them very appropriate for reversible emulsifiers. These polymers have significant hydrophobic character (which provides a driving force to move them to the oil/water interface), and they remain as suspended colloidal agglomerates under acid conditions. In addition, the high molecular weight copolymers, regardless of the incorporation of PEG, show promise as thickening agents. 90 5.4.3 Conformational changes during titration of P(MAA-g-EG) in aqueous solution In this section, the data from the potentiometric titrations of P(MAA-g-EG) will 13,14 to be analyzed using the classical methods developed by Katchalsky and collaborators describe polyelectrolyte systems. These investigators extensively studied the potentiometric titration of weak polyacids in aqueous solutions both from the experimental and theoretical points of view. One of the most noticeable features Of the dissociation equilibrium of these systems is the continuous decrease Of the apparent ionization constant K with increasing degree of dissociation, 01. Overbeek's pointed out that this effect should be connected with the increasing difficulty to remove protons from polyions with increasing charge which can arise from a pH-induced conformational change from a compact coil to an extended coil form in a limited range of degrees of dissociation”. Once the pH of a solution containing the ionic polymer of known concentration Cp (defined as the moles of ionizable functional groups per liter), is measured as a function of the volume of the titrant of a known concentration, it is possible to define the apparent equilibrium constant pKA.” In terms of experimental parameters the apparent dissociation constant can be calculated based upon the Henderson-Hasselbach equation originally proposed by Katchalsky and Spitnik. '8 a (5.14) 12K. = pH -log(1—_—a) 91 Here, the degree of dissociation, 01, is calculated from the directly observable degree of neutralization 01.1, and from the pH Of the polyelectrolyte solution. The degree of neutralization, 01, is expressed as: a-a +10‘0” (5.16) The degree Of neutralization 01,1 represents the ratio of the concentration in mole/l of the sodium ions, due to added NaOH, to the concentration of ionizable groups in monomeric equivalents in mole/l. For the P(MAA-g-EG) copolymers in this study, the pH-titration results are shown in Figure 5.9. In this figure, the apparent dissociation constant, pKA, is plotted as a function of the degree of dissociation, 01., for three different P(MAA-g-EG) copolymers at a number average molecular weight of about 25,000. The figure illustrates that, for all three polymers shown, as the degree of dissociation is increased from zero (fully protonated polymer backbone) to approximately 0.05 the pKA increases sharply and monotonically before reaching a flat plateau at a different 01 value for each polymer (the or value at which the plateau is Observed increases as the EG content is reduced from 20 Wt.% to 5 wt.%). The plateau occurs at an 01 value of 0.02 to 0.1 for the 20/5 copolymer, and at an 01 value of 0.15 to 0.25 for the 5/5 copolymer. Finally, the pKA value is observed to increase monotonically with increasing or (however with a lower slope that the portion of the curve). The general shape of the pKA vs 01 curves shown in Figure 5.9 are typical for synthetic polyelectrolytes, and indicate that there is a transition from a compact coil at 92 low pH to an extended coil as the degree of dissociation is increased due to the addition of the base”),20 In addition, the fact that the pKA transition is more pronounced for the 20/5 copolymer than either the 10/5 or the 5/5 copolymers, this indicates that the acidity of the carboxylic acid is reduced by the presence of the oligomeric ethylene glycol. Therefore, the presence of the complex (a hydrogen bonded complex in which the carboxylic hydrogen is participating) makes it more difficult to dissociate the polymeric acid. 5.5 CONCLUSIONS In this chapter, the solution properties of poly(methacrylic acid-g-ethylene glycol) block/graft copolymers in water were investigated using dynamic light scattering and potentiometric titration. The overall purpose of the study was to identify how the presence of oligomeric ethylene glycol grafts modifies the solution behavior Of P(MAA- g-EG) in water, and to determine which P(MAA-g-EG) copolymers are most appropriate for design of reversible emulsifiers. The most important design conclusion from the study is that P(MAA-g-EG) copolymers with a PEG content of more than 10 wt.% are unsuitable for emulsifiers because they too hydrophobic under acidic conditions, and precipitate out of solution. However, the pH-dependent solution properties copolymers with an EG content of 2.5 and 5 wt.% make them very appropriate for reversible emulsifiers. These polymers have significant hydrophobic character (which provides a driving force to move them to the Oil/water interface), and they remain as suspended colloidal agglomerates under acid 93 conditions. In addition, the high molecular weight copolymers, regardless Of the incorporation of PEG, show promise as thickening agents. The dynamic light scattering experiments provided some insight into the aggregate size distribution of P(MAA-g-EG) copolymers. The polymers with an EG grafting level of 2.5 wt.% exhibited a bimodal distribution under acidic conditions. The mode at the smallest dimension corresponded to a size smaller than 30 nm and may correspond to single chains, while the second mode occurred at a diameter in the range of 221 nm to 282 nm for the 2.5/ 10 and 2.5/5 copolymers and in the range of 572 nm to 832 nm for 2.5/l and 2.5/0.5 copolymers. This mode may be attributed to copolymer aggregates. The molecular weight dependence of the aggregate size may be attributed to the fact that aggregates of the longer chains occupy more space (most Of the copolymers contain only a few hydrophobic blocks under acidic conditions). The dynamic light scattering results for the polymers containing 5 wt.% EG exhibited a broad single modal distribution. The aggregates of P(MAA-g-EG) copolymers with a EG grafting level of 5 wt.% exhibited a smaller hydrodynamic radius than those containing 2.5 wt.% EG, and exhibited a decreasing hydrodynamic radius with increasing molecular weight. The first result may be explained by the fact that the increased grafting levels imparts a higher hydrophobicity (on average twice as many hydrophobic blocks per chain) into the P(MAA-g-EG) copolymers, and therefore may lead to a more compact coil in aqueous solution. Similarly, within the series of copolymers containing 5 wt.% EG, the longer chains contain on average more hydrophobic blocks per chain, and therefore exhibit more compact coiling. 94 Titrations of aqueous solutions of P(MAA-g-EG) at the same molecular weight were carried out to determine the effect of the presence of the PEG grafts on the potentiometric titrations. The potentiometric titration results for three different P(MAA- g-EG) copolymers (2.5, 5, and 10.wt% EG) at a number average molecular weight of about 25000 each were analyzed in terms of the dependence of the apparent dissociation constant pKA on the degree of dissociation 01. For each copolymer, a local maxima in the pKA curve was observed. This anomaly is more pronounced and the onset of the transition is shifted to higher 01 values with increasing PEG content, and indicates that the presence of the complex decreases the acidity of the carboxylic acid moieties. Essentially, the complex is “protecting” the hydrogen that in engaged in the hydrogen bond, making harder to dissociate the carboxylic acid. 95 EO incident - > / I I I / xx \ 1 xx / xx scattered / / b) incident Figure 5.1 a) Oscillation of scattered field vector is perpendicular to the electric field vector of the incident light. Emitted light has the same radiation frequency. b) Graphical representation Of the angular distribution of the intensity of scattered light for vertically polarized incident light. 96 ’ , ’ (compact coil) / _-.. incident V 1, / \ )1800l ‘ " J \\ \V\ R \ \ ’ I ‘ - —- ’ d d '1 900 V (expan e c01 ) Figure 5.2 For large particles the intensity of scattered light depends on the angle of observation. Therefore the angular dependence of the scattered light gives information about size of a macromolecular particle (R). 97 I(t) G2(t)= b) Normalised Correlation Function, I Gz(t) I = [Gz(t)-G(°o )] / Gzioo) 1'“ b ,.. 0.81r E 0.6 -~ :37 _ _ 0.4 “Ale 0.2 ‘— 0 . 25% 1 10 100 1000 10000 Time, usec Figure 5.3 a) Erratic raw data of scattered light as detected by a photomuliplier tube. b) The normalized time correlation function le(t)l as a function of time. Initially this function is 1 (not normalized <12>). For very long times the correlation function decays to 0, (not normalized 2). Relaxation time of a correlation function, 1:, is the time required for the time correlation function to drop to He of its initial value. 98 vvvvv - Q VW~VW‘V‘V‘V‘V"" -\‘L\“ AMA-“Aw; ......... 4 o\e N “A + 135° + 90° .4.— 60° —v— 45° :3. (5 c “I _.| -8-+ -10~ O —4 4 lag time [ms] Figure 5.4 ‘ Dynamic light scattering studies on angle dependence for sample P(MAA-g-EG) 2.5/5. The normalized correlation functions ln( I 620) I -1) as a function of lag time show deviation from a straight line. This corresponds to a distn'bution of relaxation decay times. In this case, each species contributes its own exponential, according its specific diffusion coefficient 99 1 35° OJ 1 10 100 1000 90° OJ 1 10 100 1000 60° OJ 1 10 100 1000 45° OJ 1 10 100 1000 O 30 OJ 1 10 100 1000 Hydrodynamic Radius Rh [nm] Figure 5.5 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g-EG) 5/5. 100 OJ 1 10 100 1000 OJ 1 10 100 1000 OJ 1 10 100 1000 OJ 1 10 100 1000 Hydrodynamic Radius Rh, nm Figure 5.6 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g-EG) 2.5/5. 101 135° OJ 1 10 100 1000 10000 1 10 100 1000 10000 Hydrodynamic Radius Rh, nm OJ Figure 5.7 Hydrodynamic radius distributions for a 0.5 wt.% aqueous solution of sample P(MAA-g-EG) 5/ 1. 102 900 - 800 - aggregate size at "zero angle" [nm] 8 O L I 2.5 Who/O EG I O 5 wt.% EG A pure PMAA linear fit for 2.5 wt.% EG linear fit for 5 wt.% E6 0 El 13 r l I fi 1 I I f I T I j l I T l I I 0 25k 50k 75k 100k 125k 150k 175k 200k number average molecular weight Mn [g/mole] Figure 5.8 Projected aggregate sizes of P(MAA-g-EG) as a function of the number average molecular weight of the individual copolymers. If two modes are present, the open symbols correspond to the first mode for the same group of polymers as given in the legend. 103 7.0 - 6.5 - 6.0 4 ' x“ ' D. 5.5 - L o’ if —-— 20/5 f; —0— 10/5 or: _____A__ 5/5 r T ' ' l l l ' 0.0 0.1 0.2 0.3 0.4 . 0.5 0.6 degree of dissociation a Figure 5.9 Modified pH-titration curves of different poly(methacrylic acid-g-ethylene glycol) copolymers with a number average molecular weight of about 25000 each in a 1 wt.% aqueous solution. The ethylene glycol grafting level incorporated into the poly(methacrylic acid) backbone are 2.5, 5, and 10 wt.% of total polymer mass. 104 Sample modality 6[°] Rhl [nm] ha[nm] Rflnm] Correlation Name Small LarJgg At 0=0 coefficient R 5/10 1 30 345 411 -0.9 45 321 60 240 90 105 135 91 5/5 1 30 171 177 09 45 164 60 143 90 125 13 5 128 5/0.5 l 30 175 155 -0.72 45 60 60 40 9O 7O 2.5/10 2 30 5 345 282 -0.6 45 l 164 60 l 73 2.5/5 2 30 10 47 221 -1.0 45 8 164 60 7 143 90 7 105 2.5/1 2 3O 27 643 832 -0.9 45 25 444 60 24 424 2.5/0.5 2 3O 19 563 572 -0.59 45 19 538 6O 19 548 Table 5.1a List of Light scattering analysis results for samples P(MAA-g-EG) 5/10, 5/5, 5/O.5, 2.5/10, 2.5/1 and 2.5/0.5. 105 Sample modality 9[°] Rhl[nm] Rfinm] RhZInm] Correlation Name Small Large At0=0 coefficientR 0/0.5 2 3O 17 550 679 -O.8 45 17 450 60 15 130 90 17 205 135 Sample modality 9[°] R..‘[nm] RhZInm] Rb‘lnm] Correlation Name Small Large At0=0 coefficient R 5/1 1 (broad) 30 64 - 56 -0.39 45 45 - 6O 45 - 9O 53 - 135 47 - Table 5.1b List of Light scattering analysis results for samples P(MAA-g-EG) O/O.5 and 5/1. 106 10. 11. 12. 13. 14. 15. 16. 17. 5.6 REFERENCES Tanford, C., Physical Chemistry of Macromolecules, Wiley, New York, Chapter 5, 1961. Rudin, A., The Elements of Polymer Science and Engineering, 2“d edition, Academic Press, San Diego, 1998. Chu, B., Laser Light Scattering: basic principles and practice; Academic Press, Boston, 1991. Beme, B. J ., Pecora, R., Dynamic Light Scattering with Applications to Chemistry, Biology, and Physics, Wiley, New York, 1976. Aseyev, V. 0., Manuscript of lecture on Light Scattering, University of Iowa, Iowa City, Iowa, March 2000. Brown, W. editor: Dynamic Light Scattering : The Method and some Application; Clarendon Press, Oxford, 1993. Schatzel, K., Quantum Optics 2, 287 (1990). Stieber, E, Peters, R., manual for software ALV-SOOO/E for Windows 95, ALV- Laser Vertriebsgesellschaft m. b. H., Langen, Germany, 1997. Provencher, S. W., Physics Communication, 2 7, 229 (1982). Provencher, S. W., Physics Communication, 27, 213 (1982). Hiemenz, P. C., Principles of Colloid and Surface Chemistry, Marcel Dekker, New York, 1977. Drescher, B., Reversible Block/Graft copolymer as Emulsifiers, Diplomarbeit, Universitiit Kaiserslautem, Germany, 1997. Katchalsky, N. S., Eisenberg, H., J. Polym Sci. 13, 69 (1954). Katchalsky, N. S, Gillis, J., J. Polym. Sci. 13, 43 (1954). Overbeek, J. T. G., Bull. Soc. Chim. Belges 57, 252 (1948). Mandel, M., European Polymer Journal 6, 807 (1970). Katchalsky, A., Pure Appl. Chem. 26, 327 (1971). 107 18. Katchalsky, A., Spitnik, P., J. Polym. Sci. 2, 432 (1947). 19. Sakurai, M. et al., Polymer Journal 25(12), 1247 (1993). 20. Ben-Naim, A., Hydrophobic Interactions, Plenum Press, New York, 1980. 108 L.- Chapter 6 EMULSIFICATION STUDIES OF POLY(METHACRYLIC ACID)- GRAFTED-POLY(ETHYLENE GLYCOL) COPOLYMERS 6.1 INTRODUCTION In this chapter, the influence of the molecular architecture on the emulsification behavior is investigated. Based upon the design guidelines provided by the solubility studies reported in Chapter 5, in this Chapter, the emulsification behavior is characterized for the copolymers that show promise as emulsifiers. In the first experimental section, a titration cycle is used to form, break and reform the oil/water emulsions using P(MAA-g- EG) copolymers containing 2.5 and 5wt.% ethylene glycol. The amount of oil released from the emulsions is investigated as a function of the solution pH. In the second experimental section, the effects of the molecular weight and composition of the polymers the pH dependent emulsification properties are investigated 6.2 REVERSIBILITY OF EMULSIONS STABILIZED BY P(MAA-G-EG) Studies were performed to evaluate the reversibility of emulsions that were stabilized by P(MAA-g-EG) copolymers under acidic conditions. A truly reversible emulsifier has the ability to form, break and re-form the emulsion on demand. A variation of the standard bottle test was performed to determine quantitatively the reversibility of emulsions stabilized by the P(MAA-g—EG) copolymers. For these studies, 109 stock solutions (at a concentration of 0.5 wt.%) of the following series of copolymers were prepared: 0/ 10, 0/5, O/l, 0/0.5, 2.5/ 10, 2.5/5, 2.5/ 1, 2.5/0.5, 5/ 10, 5/5, 5/1 and 5/O.5. The solutions were prepared in HPLC grade water under constant stirring, and the pH was adjusted to a valhe of 2.0 by adding small quantities of HCl. For the reversibility tests, 20 ml the solution under investigation was placed into a standard separatory funnel (Fisher Scientific, Pittsburgh), then 40 ml of the model oil phase methyl laurate (Aldrich, Inc., Milwaukee, WI) was added to the aqueous polymer solution. A “tissue tearor” laboratory homogenizer (Fisher Scientific) was used to vigorously and thoroughly mix the system. After mixing the funnel was sealed and the emulsion was allowed to settle for 24 hours. After settling, an emulsion phase appeared as a homogeneous opal white cream. Since the oil phase has the lower density, the buoyant force on the droplets caused them to rise to the top of the funnel until they attained a densely packed geometry stabilized by the emulsifier. Therefore a clear water phase (presumably an aqueous solution containing some P(MAA-g-EG)) was typically found at the bottom of the separatory funnel. The positions of the phase boundaries were marked on the outside of the separatory funnel. For all of the P(MAA-g-EG) copolymers studied, all of the oil phase was emulsified at pH 2.0. The emulsion was broken by adjusting the pH of the solution to higher (more basic) values. The solution pH was adjusted stepwise from the value of 2.0 to the pH values of 4.0, 5.8 and 7.0 and back to a pH of 2.0 using the Eppendorf precision pipettes (in all cases, the emulsion was completely broken at pH 7.0). To determine the amount of NaOI-l solution needed for the pH adjustment, 20 ml from the stock solutions were titrated to the target pH value. Subsequently, the same volumes of the titrant as measured 110 for the reference samples were used to modify the solution in the separatory funnel to equivalent pH values. After each pH adjustment, the content of the separatory funnel was thoroughly and vigorously blended with the laboratory homogenizer and the mixture was allowed to settle into phases after 24 hours. The positions of the new phase boundaries were recorded. After the complete breaking of the emulsion, the solution pH value was back-titrated to the initial pH value of 2.0 and the emulsion was again formed by agitation with the laboratory homogenizer. Figure 6.1 contains a bar chart showing the volumes of the various phases present as a function of the pH for an emulsion of methyl laurate stabilized by P(MAA-g-EG) 5/10. The figure illustrates that, at pH values of 2 and 4, the emulsion phase and small amount of an aqueous phase are present. At a pH value of 5.8, the emulsion is completely broken, and the oil and water phases separate completely from one another. After titrating back to the original solution pH of 2, the emulsion is reformed again after agitation. Similar results were obtained for all of the P(MAA-g-EG) copolymers investigated. Therefore, a first important result is that all investigated polymers were capable of emulsifying 100% of the oil present at a pH value of 2, and of these emulsions were reversible. 6.3. EFFECT OF MOLECULAR WEIGHT AND COMPOSITION OF THE PH- DEPENDENT EMULSIFICATION PROPERTIES OF P(MAA-G-EG) Figure 6.2 shows the relative amount of oil released as a function of pH for two polymers of similar molecular weight (~120,000 g/mole) but different composition. The error bars in Figure 6.2 through 6.5 correspond to the maximum deviation determined by 111 weighing the phase volumes. The figure contains data for the percentage of the oil that was released from the emulsion as a function of pH for the 2.5/1 and the 5/1 copolymers of P(MAA-g-EG). The figure illustrates that both of these copolymers emulsified 100% of the oil at pH 2, but that the 2.5/l copolymer released more oil as the pH is increased to 4, and finally to 6 and 7. For example, approximately 80% of the oil is released from the emulsion at pH 4 for the 2.5/1 copolymer, however the oil is still completely emulsified by the 5/1 copolymer at this pH. Both copolymers release nearly all of the oil at a pH of approximately 5.8. In Figure 6.3, the relative amount of oil released for emulsions stabilized by low molecular weight P(MAA-g-EG) copolymers 2.5/10 and 5/5 are shown as a function of pH. As discussed in Chapter 4, the copolymers P(MAA-g-EG) 2.5/10 and 5/5 have a repeat unit ratio MAA:EG of 20:1 and 94:1, respectively. Since these two polymer have similar molecular weights, this figure illustrates the effect of composition on the emulsification properties. The figure illustrates that the sample 5/5 completely emulsifies the oil present at both pH values of 2 and 4. However, at an adjusted pH value of 5.8 and higher, the oil bound in the emulsion is fully released. However, the data for the P(MAA-g-EG) 2.5/10 illustrates that the oil is already released to the fullest extent at a pH value of 4. As for the higher molecular weight polymers shown in the previous figure, the P(MAA-g-EG) copolymers containing the higher EG content show the ability to emulsify the oil at a higher pH. Together, Figure 6.2 and 6.3 illustrate the effect of the P(MAA-g-EG) composition on the pH-dependent emulsification properties of the copolymers. For both the high molecular weight polymers (~120,000 g/mole) and the low molecular polymer 112 (~20,000 gram/mole) the polymers containing the higher content of PEG grafts retain the ability for stabilize emulsions at a higher pH. Therefore, the increased hydrophobicity effectively shifts the critical pH value for breaking the emulsion to a higher value. This can be attributed to the fact that, as the number of hydrophobic (complexed) segment is increased, it takes a higher degree of dissociation of the backbone carboxylic acid groups to remove the copolymer from the oil/water interface. Figure 6.4 and 6.5 illustrate the effect of the copolymer molecular weight on the pH-dependent emulsification properties of the P(MAA-g-EG) copolymers. Each of these figures contains emulsification data for copolymers that possess approximately the same grafting level of EG but four different molecular weights ranging from 20 to 120 g/mole. In Figure 6.4 the relative amount of oil released from the emulsion is shown as a function of the pH value for polymers containing 5 wt.% EG. As shown in the figure, for all of the molecular weights, no oil is released as the pH is raised from 2 to 4. However, at a pH of 5.8, the amount of oil released from the emulsion varies from 100% for the shortest copolymer to ~80% for the longest copolymer. Interestingly, the data indicates that the amount of oil released from the emulsion decreases slightly as the molecular weight of the stabilizing copolymer is increased. Figure 6.5 illustrates the effect of molecular weight on the pH-dependent emulsification properties of copolymers containing 2.5 wt.% EG. Figure 6.5 corroborates the trend that the higher molecular weight copolymers are the most effective emulsifiers since, for example, only the copolymer P(MAA-g-EG) 2.5/0.5 with the highest molecular weight (~126,000 g/mole) emulsifies all of the oil at a pH value of 4. For higher pH values such as 5.8 and 7, the emulsions are essentially broken. The data in Figures 6.4 113 and 6.5 indicate that the emulsification ability of the copolymers increases slightly as the molecule weight is increased, but that this effect is not as dramatic as the dependence upon composition that was illustrated in Figures 6.2 and 6.3. 6.4 CONCLUSIONS An important result from this study is that P(MAA-g-EG) copolymers having an EG content of 2.5 to 5 wt.% can emulsify a 2:1 mixture of oil in water when present at 0.5 wt%. In addition, the P(MAA-g-EG) copolymers can be used to form, break and reform the emulsion repeatably. The studies reported in this chapter also provide important guidance on how to design emulsifiers tailored for specific applications since the effects of the copolymer composition and molecular weight are illustrated. The results indicated that both the level of grafting and the number average molecular weight of the copolymers influence pH-dependent emulsification properties. For a given molecular weight, the polymers containing a higher graft content retain the ability to stabilize emulsions at a higher pH. This can be attributed to the fact that, as the number of hydrophobic (complexed) segment is increased, it takes a higher degree of dissociation of the backbone carboxylic acid groups to remove the copolymer from the oil/water interface. Finally, the emulsification ability of the copolymers increases slightly as the molecular weight is increased, but this effect is not as dramatic as the dependence upon composition. 114 El oil 7 7 7 7 l emulsion , 7 [1 water phase volume [mL] 2 4 5.75 2 adjusted pH value [-] Figure 6.1 Results of the reversibility test of emulsions stabilized by the P(MAA-g-EG) 5/10 copolymer. The initial emulsion was formed by mixing 33 vol.% of an aqueous polymer solution with 66.6 vol.% methyl laurate. The polymer solution had a concentration of 0.5 wt.%. The volume of the oil phase, emulsion phase and un-emulsified water phase are shown as a function of the adjusted polymer solution pH. 115 120- 100“ / '03 80- i 33 1 3 604 E z 4 '5 32’ 404+ Mn 3 +2.5/1 116k 9 201 -—<>—5/1 119k .- i i l 2 ' 3 ' A ' é ' é ' 7 adjusted pH of P(MAA-g-EG) solution H 1 Figure 6.2 Relative amount of oil released from an emulsion stabilized by high molecular weight polymers as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution. 116 120- .. T I g I fi-rwfihw I JI/ “L '03 80 q / a, / 8 a) 60 - / 712 = 1 1/ Mn 0 40- I’ +2.5/10 17k (1) / .2 / "A“ 5/5 25k 9 20 - f/ T f/ 0- . I l l ' l ' l ' I T l I 2 3 4 5 6 7 adjusted pH of P(MAA-g-EG) solution {-1 Figure 6.3 Relative amount of oil released from an emulsion stabilized by low molecular weight polymers as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution. 117 120- 100- § 80- (D 8 2 60- 9 E a: 40' .2 . 32' 93 20‘ .- t I W l adjusted pH of P(MAA-g-EG) solution [-] Figure 6.4 Relative amount of oil released from an emulsion stabilized by P(MAA-g-EG) copolymers with an MAA:EG repeat unit ratio of about 10:1 as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution. 118 120- 100- ;? 80 - 0) £8 + 2.5/10 % 6° " + 2.5/5 : + 2.5/1 :3) 40 1 —I—— 2.5/0.5 .2 . 35 e 20 _, .l o .. T ' l ' l 1 l ' l ' I 2 3 4 5 6 7 adjusted pH of P(MAA-g-EG) solution {-1 Figure 6.5 Relative amount of oil released from an emulsion stabilized by P(MAA-g-EG) copolymers with an MAA:EG repeat unit ratio of about 10:1 as a function of the adjusted pH value of the aqueous P(MAA-g-EG) copolymer solution. The polymer concentration was 0.5 wt.% in the aqueous solution. 119 Chapter 7 DYNAMIC INTERFACIAL TENSION MEASUREMENTS 7.1 INTRODUCTION The purpose of this study is to characterize the effect of the P(MAA-g-EG) copolymers on the interfacial tension between water and a model organic phase (methyl laurate). Interfacial tension is an important parameter in surfactant science because the ability of a surfactant to reduce the surface tension allows droplets (emulsions) to form more easily. This is because the surface tension creates a thermodynamic driving force for minimizing the total interfacial area and coalescing the droplets. In general, the higher the interfacial tension, the larger the driving force for coalescence. Characterizing the interfacial tension reduction observed for the P(MAA-g-EG) copolymers under acidic conditions is important because it will help to establish the manner by which the copolymers stabilize emulsions. If the copolymers lead to a significant reduction in the surface tension, then they work, in part, by making it easier to form the emulsion by lowering the thermodynamic barrier to emulsification. If the copolymers do not significantly lower the surface tension, it implies that the emulsion stability arises primarily by preventing coalescence of the droplets once they are formed. In this chapter, the effect of polymer architecture and molecular weight on the reduction of the interfacial tension in the system water/methyl laurate is presented. Previous interfacial tension studies1 were carried out with the Tensiometer Kriiss K122 employing the Wilhelmy plate method. The Tensiometer Kritss K12 is based upon 120 a force measurement during a dynamic experiment in which a plate is withdrawn from a two-phase liquid mixture. However, if the solution viscosity is also changing from one sample to another, the surface tension measurements based upon the Wilhelmy plate method can be compromised since the viscous drag forces are superimposed on the interfacial forces that are measured.2 For this reason, dynamic drop volume tensiometry technique was chosen for the studies in this dissertation. The dynamic drop volume tensiometry technique3 has been used successfully to study liquids with Viscosities up to 20 Paos (compared to viscosity of water: 9.8 x 10'4 Pats at room temperature).3 The dynamic drop volume technique relates the volume of a drop detaching from a capillary tip to the interfacial tension at the instant of detachment-3"“5 Dynamic interfacial tension has been proven to be a valuable method for comparing individual surfactants.3 The dynamic interfacial tension measurements are based upon a force balance on a growing oil droplet in an aqueous solution containing the emulsifier, and are much less sensitive to the solution viscosity. 7.2 EXPERIMENTAL 7.2.1 Sample preparation Aqueous stock solutions of P(MAA-g-EG) copolymers were produced by dissolving 1.1 g of a polymer sample into 220 g of HPLC grade of water under constant stirring for three consecutive days. The polymer solutions that were clear and apparently dissolved to the full extent were then decanted into 50 ml vials. The copolymers with higher levels of grafting - such as 10 wt.% - that did not dissolve to a full extend were 121 titrated to neutral conditions in order to dissolve all solid polymers. Quantities of 50 ml were taken from these dissolved stock solutions. All polymer solutions were then subject to titration with 4 M NaOH or HCI to adjust the pH value of the solutions to a value of pH 2, 4, 5.75 and 7, respectively. For the concentration tests, some solutions were diluted down to lower concentrations as stated in the subsequent section. 7.2.2 Measurement procedure In the following, the measurement procedure of a Kritss DVT-10 Dynamic volume tensiometer is described. Figure 7.1 shows a sketch4 of the main part of the device. Before conducting an experiment, all components that come in contact with the liquids are cleaned with isopropyl alcohol and acetone and rinsed with COpious amounts of HPLC grade water. A sample of denser liquid — in this case water containing P(MAA- g-EG) copolymers- was placed in a tube. The tube was filled halfway with 20 ml of the emulsifier solution. Located at the bottom of the tube is a small-bore capillary (OD 254 um) connected to a syringe pump. The syringe pump is used to force the lighter liquid methyl laurate through the capillary into the denser phase at a constant flow rate. An IR LED and photodiode are positioned above the capillary to detect drops that detached. A timer is started when the first drop in an experiment is detected. Time between subsequent drops is then measured. Since the flow rate is constant, the time between drops is easily converted to give the volume of each drop. Invariably eight drops were measured and averaged to allow statistics analysis of the experimental data. Before every measurement series, the feed lines syringes were degassed and the spherical shape of the drops was verified visually. As Figure 7.2 shows, when a droplet forms, a balance of 122 forces exists at the orifice until the drop grows large enough to detach. At the instant of detachment, the volume of the drop is directly proportional to interfacial tension between the two phases3. The separation force is determined by the buoyant lift due to the density difference. The adherence force due to the interfacial tension difference that prevents mixing of the two phases balances the separation force. Mathematically, this relation is expressed as: Vnmw '(pH ”p1.)'g = 0.1.”.d (7'1) In this equation, VDrop is the volume of the drop, p corresponds to the density of the heavy (subscript H) and light phase (subscript L). The symbol g represents the acceleration due to gravity. On the right hand side of the equation, the adherence force is given by the product of the interfacial tension 0 and the tip circumference n-d. Hence, the measured drop volumes of the lighter phase are converted into the interfacial tensions using equation 7.1. 7.2.3 Results and Discussion The first series of experiments was performed to identify the appropriate operating range for the oil flow rate in the dynamic interfacial tensiometer. These studies were performed because the accuracy of the data obtained with the dynamic interfacial tensiometer may be may be compromised if the flow rate of the pumped fluid is too high. At very high flow rates, the acceleration of the pumped fluid will to contribute to the detachment force and the recorded interfacial tension will be artificially low. Therefore, 123 in order to eliminate this potential error, data of one reference system were collected at different flow rates to ensure that the measured value does not depend upon the flow rate. In Figure 7.3 the interfacial tension between the oil methyl laurate and a 0.5 wt.% aqueous solution of P(MAA-g-EG) 5/5 at a pH of 2.0 for five different flow rates ranging from 0.25 to 2.00 mL/hr are shown. Figure 7.3 illustrates that the measured interfacial tension remains constant within the experimental accuracy for this range of flow rates (the slope of a fitted line through these data in Figure 7.3 was calculated to be near zero at a value of 0.09). Based upon these results, for all other experiments, the flow rate of the lower density fluid was set to l mL/hr. The data in Figure 7.3 do indicate that there is a reduction of interfacial tension in the system due to the presence of polymeric emulsifiers. The interfacial tension values in the range of 14.0 to 14.7 mN/m are considerably less than the value measured for the system methyl laurate and water in the absence of any polymer (21.8 mN/m). Figure 7.4 illustrates the effect of the P(MAA-g-EG) copolymers on the interfacial tension of the methyl laurate/water interface. The figure contains a plot of the measured interfacial tension as a function of the weight percentage of the oligomeric ethylene glycol in the graft copolymer (from zero to 30 wt.% of the grafts) at four difierent pH levels. All of the data in this figure correspond relatively low molecular weight polymers (number average molecular weight of 15,000 - 17,000 g/mole) and the polymer concentration is always 0.5 wt.%. Based upon the labeling scheme described in Chapter 4, the four different P(MAA-g-EG) copolymers are 0/ 10 (Mn = 17,000) 2.5/ 10 (Mn = 17,000), 10/10 (Mn = 15,000), and 30/10 (Mn = 15,000). Recall that the first 124 number in the ratio corresponds to the weight percentage of the grafts, while the second number corresponds to the weight percent of the initiator in the reaction mixture. The data in Figure 7.4 show some very interesting trends. First, note that all of the polymers exhibit the same trend for decreasing interfacial tension with increasing pH. For example, the interfacial tension is about 15.8 mN/m at a solution pH of 2 for the P(MAA-g-EG) copolymers with a level of grafting of 0 and 2.5. At a pH value of 4 the same polymers exhibit a smaller surface activity with values of 17.5 and 18.0 mN/m for copolymers with a level of grafting of 0 and 2.5, respectively. For the moderately acidic pH value of 5.75 and under neutral conditions, the interfacial tension values overlap and scatter between 19.2 and 21.1 mN/m. These trends illustrate for these copolymer solutions, the pH plays the predominant role in determining the interfacial tension, and the polymer composition has only a modest effect (note that there is no reduction in interfacial tension with decreasing pH if the copolymer is not present). This result suggests that under acidic conditions, all of the polymers tend to locate themselves at the oil/water interface. A second important result is that only a modest reduction in interfacial tension is observed for these P(MAA-g—EG) copolymers. The surface tension is reduced from a value of about 21.8 mN/m for pure water, to about 15 mN/m for the graft copolymers. Figure 7.5 illustrates the effects of pH and polymer composition on the observed interfacial tension for samples containing 0.5 wt.% of high molecular weight P(MAA-g- EG) copolymers (number average molecular weights 115,000 to 125,000 g/mole). Here, the data is compiled for P(MAA-g-EG) polymers with the following designations: 0/0.5 (M., = 121,000), 2.5/1 (Mn = 115,000), 2.5/0.5 (Mn = 125,000), and 5/1 (M., = 119,000). 125 As previously observed in Figure 7.4, the observed reduction in interfacial tension depends more strongly upon the pH of the polymer solutions than on the copolymer composition. Again, the interfacial tension is the lowest at the lowest pH value. At a pH of 2, the interfacial tension was measured to be in the range between 15.8 and 16.9 mN/m for high molecular weight polymers having a number average molecular weight in the range of 120,000 g/mole. For the moderately acidic pH value of 5.75 and under neutral conditions, the interfacial tension values overlap and scatter between 19.5 and 21.2 mN/m. These results are analogous to the values measured for the polymers having the same level of grafting and under the same acidic conditions but having a lower molecular weight. In order to clarify the apparent independence of the surface activity on the molecular weight, the interfacial tension is plotted as a function of the copolymer molecular weight in Figure 7.6 and Figure 7.7. Figure 7.6 shows the interfacial tension between water and methyl laurate as a function of the molecular weight of P(MAA-g- EG) copolymers containing 2.5 wt.% EG at pH values ranging from 2 to 7. Again, each symbol corresponds to a constant solution pH value. Note that for the data at pH 2 and 4, the measured interfacial tension seems to increase slightly with increasing molecular weight. At a pH value of 2, the interfacial tension in the system methyl laurate/water can be reduced to values around 15 to 17 mN/m- At a pH value of 4, the interfacial tension values are reduced into the ranges around 18-19 mN/m, at 5.75 and 7 only marginally to the range of 19-21 mN/m. In Figure 7.7 the analogous data is shown for the copolymer system with an EG content of 10 wt.%. Both Figure 7.6 and 7.7 indicate that the observed interfacial tension is more strongly dependent on the pH of the polymer solution 126 than on the polymer molecular weight. For a constant pH value a band value of interfacial tension can be expected. At a pH value of 2, the measured interfacial tension values vary in the area of 14.4 to 15.6 mN/m. At a pH of 4, the interfacial tension varies between 15.2 and 17.3 mN/m. This range is somewhat lower in comparison to the interfacial tension reduction for copolymers with an EG content of 2.5 wt.% at the pH of 4. Again, at the pH values of 5.75 and 7 only a minor reduction in interfacial tension is noticed. The effect of concentration on the surface activity of three different P(MAA-g- EG) varying in molecular weight and level of grafting are shown in Figure 7.8. The P(MAA-g-EG) copolymers discussed here are those denoted 5/10, 2.5/5 and 10/1 with the number average molecular weight of 26,000, 44,000 and 82,000, respectively. The concentrations of the copolymer aqueous solutions were varied from 0.001 to 0.5 wt.%. The pH values of all solutions were kept at a pH value of 2. It should be noted that the sample 10/1 and 5/10 appeared turbid. Nonetheless, both polymers show in Figure 7.8 similar surface reduction capacity in comparison to the dissolved copolymer 2.5/5 having a molar ratio MAA:EG of 20:1. Clearly, there is a reduction of the interfacial tension between water and methyl laurate from 21.8 mN/m for P(MAA-g-EG) 10/1 and from 21.5 mN/m for P(MAA-g-EG) 2.5/5 and 5/10 at the concentration of 0.001 wt.% to a value about 15.5 mN/m at a concentration of 0.5 wt.%. Invariably, the drop of interfacial tension is more rapid in the area of 0.001 wt.% to 0.0156 wt.%. The interfacial tension is lowered at a slower rate after the polymer concentration of 0.01 56. A compilation of all data is given in Table 7.1. For the cases in which a reliable measurement could not be obtained, the table does not have an entry (for example, 20/10 127 and 30/10 at pH 2.00 and 4.00). Reasons for a failed measurement were: 1) a highly too viscous aqueous polymer solution that hindered the formation of a circular drop and delayed the detachment; or 2) solution turbidity that hindered the LED detection. For each entry, the table contains the value of the measure interfacial tension and the relative standard deviation for the eight measurements. Notice that the data was generally very repeatable, with the relative standard deviation typically falling between 1 and 2% (that average standard deviation was 1.6%) It is interesting to compare these surface tension results to the solution property results presented in Chapter 5. Recall that in Chapter 5, the solution behavior was found to be dependent on the level of grafting, and that at a concentration of 0.5 wt.% in water, a P(MAA-g-EG) copolymer that has only 2.5 wt.% EG incorporated into its chain can be completely dissolved, even at acidic conditions. However, at a grafting level of 5 wt.% EG and a solution pH value of about 4 and below, the polymer solution becomes turbid (Tyndall effect) and at even higher concentrations, a turbid solution with precipitated solids is found. However, the interfacial tension data of copolymer P(MAA-g—EG) 10/ 10 at pH 2 and 4 is even lower than for those copolymers that are completely dissolved. Apparently, although some polymer is agglomerated such that some polymer is “lost” in precipitation, the surface reduction ability is not depleted. 7.7 CONCLUSIONS As a first important result, the studies reported in this chapter reveal that the copolymers P(MAA-g-EG) have the ability to moderately reduce the interfacial tension in the system methyl laurate/water (the maximum reduction was measured for the 128 polymer P(MAA-g-EG) 20/5 in which the interfacial tension was reduced from 21.8 mN/m in the absence of the polymer to a value of 14.3 mN/m at a pH of 2 with a polymer concentration of 0.5 wt.%). For comparison, the well-known surfactant sodium dodecyl sulfate is able to lower the interfacial tension in the system water/canola oil from 38 lem to about 5 mN/m". This illustrates that the P(MAA-g-EG) exhibit only a moderate ability to reduce surface tension implying that the ability of the polymers to stabilize emulsions arises primarin from their ability to prevent droplet coalescence after they are formed, not by making it easier to form droplets. In this manner, the P(MAA-g-EG) copolymers behave as stabilizers of emulsions, rather than as surfactants. This conclusion is further corroborated by the fact that the non-grafted poly(methacrylic acid) homopolymer exhibits a similar modest surface tension reduction as the graft copolymers (even though the PMAA homopolymer is not effective as an emulsifier). In this study we investigated the effects of the following variables on the surface tension reduction observed for the P(MAA-g-EG) copolymers in the methyl laurate/water system: pH, level of grafting, molecular weight and concentration. Of the first three variables, the pH was observed to have the largest effect on the observed interfacial tension, with the trend of decreasing interfacial tension with decreasing pH for a constant polymer concentration (no interfacial tension reduction is observed in the absence of the polymer). An important “transition” pH value appears to be equal to the pKA of the acid (~5.8) since the interfacial tension at pH values of 5.75 and 7.0 are both close to 21 mN/m (the value for the system with no copolymers), while the interfacial tension for pH values of 2 and 4 are significantly lower (~l4-16 mN/m). Compared to the pH, the level of grafting had little effect on the observed interfacial tension. Similarly, the molecular 129 weight of the P(MAA-g-EG) copolymers has relatively little effect on the interfacial tension (a slight increase in interfacial tension is observed with increasing molecular weight). The studies of the effect of concentration revealed that for all polymers, the interfacial tension reached a minimum plateau at a concentration of ~0.3 wt.%. 130 LIQUOIOVII———‘ LEO mung» Wm Figure 7.1 Drawing of the main part of the DVT 10 drop volume tensiometer 131 Balance of Forces at the tip Separation Force a VDrop . (pH _ pl. ). g - l I Adherence Force 0': 'fl'd = V[)rop(pH “p1.)'g V Figure 7.2 Force Balance for the drop volume experiment. 132 Interfacial Tension y [mN/m] 22 20 18 16 14 12 l I ‘ l r l j - P(MAA-g-EG) 5/5 at pH = 2 . I .. I I _ 'I T ' l ' l ' l 0.0 0.5 1.0 1.5 2.0 Flow rate of light phase [mL/hr] Figure 7.3 2.5 Interfacial tension measured for the system methyl laurate and 0.5 wt.% aqueous solution of P(MAA-g-EG) 5/5 at a titrated pH of 2 for five different flow rates ranging from 0.25 to 2 mL/hr. 133 21.04 A, V v 20.5 V A v 4 20.04 A ._.. 19.54 E ~ 1 2 19.0- .§. 18.5— }. 180 ' . S ' . - pH2 2 17.04 A pH 5.75 E 16.53 V PH7 3 160' E ' - I I E 15.5: c - 15.0- 14.5- _ 14.0 I I I f T I I 1 0 2.5 10 30 level of grafting [wt.%] Figure 7.4 Interfacial tension versus level of grafting for low molecular weight (ca. 17,000 g/mole) P(MAA-g-EG) copolymers in the system methyl laurate/water at four different pH levels. 134 A 214 v ‘ ‘1 ._. 20- X E Z v I pH2 E O pH4 : 194 o A pH5.75 3, q 0 V pH7 r: g 18- . .g u . g 2 17- . E. j I I 16- I l ' I ' I T I ' I ' l 0 1 2 3 4 5 level of grafting [wt.%] Figure 7.5 Interfacial tension versus level of grafting for low molecular weight (ca. 120,000 g/mole) P(MAA-g-EG) copolymers in the system methyl laurate/water at four different pH levels. 135 21~ A A E 20-I ‘ A V \ E V V : 19- I .5 ' ' ca 18- O ' 9:3 I pH2 Tu I pH4 '83 17- A pH 5.75 E V PH7 I . m 4 5 164 I .. I 15 I I r I I I ‘ l ' I ' I r 1 0 20000 40000 60000 80000 100000 120000 140000 number average molecular weight Mn [g/mole] Figure 7.6 Interfacial tension in the system water/methyl laurate in the presence of P(MAA-g—EG) copolymers with an EG content of 2.5 wt.% at different solution pH versus the number average molecular weight of the copolymers. 136 23- 221 r—121" ' V A g 20“ ‘ >- I I pH2 '5 1 A pH5.75 E 184 V pH7 c—d ' c g 17- . 1516- E . . I 15- ’ I ‘ I 14 I ' l ' T ' I 0 40000 80000 120000 1 60000 number average molecular weight Mn [g/mole] Figure 7.7 Interfacial tension in the system water/methyl laurate in the presence of P(MAA-g-EG) copolymers with an EG content of 10 wt.% at different solution pH versus the number average molecular weight of the copolymers. 137 II' I I I IIIII' I I I IrIII' I I I I 22-4 4 2‘ 1. i 204 .. J + 5/10 - ‘9‘ + 2.5/5 ' -~~~A--~ 10/1 16- lnterfacial Tension y [mN/m] a I 15- 14 III' I I I II‘rIrr I I I IIITI' I I fr 115-3 0.01 0.1 Polymer concentration [wt.%] Figure 7.8 The effect of concentration on the surface activity of three different P(MAA-g-EG) 5/10, 2.5/5 and 10/1 varying in molecular weight and level of grafting. 138 Table 7.1 Overview of complete interfacial tension data in the system methyl laurate/water at the presence of P(MAA-g-EG) copolymers. The aqueous polymer solutions were titrated to the pH values as indicated in the table. In the adjacent columns, the relative standard deviations for the measured interfacial tension results are given. pH = 2.00 pH = 4.00 pH = 5.75 pH = 7.00 P(MAA-g-EG) Ave. IFT RSD Ave. IFT RSD Ave. IF T RSD Ave. IF T RSD sample [mN/m] [%] [mN/m] [%] [mN/m] [%] [mN/m] [%] designation 0/10 15.8 1.8 17.5 1.7 20.9 1.0 20.6 1.8 2.5/10 15.8 3.9 18.0 3.9 19.9 3.1 21.1 1.0 5/10 15.7 1.9 18.0 1.1 20.6 1.2 20.1 1.2 10/10 14.4 1.9 15.4 1.7 20.5 3.0 20.8 1.2 20/10 20.0 1.3 20.8 1.2 30/10 19.2 1.3 20.5 1.4 0/5 16.2 1.8 18.0 1.4 21.0 1.1 20.9 1.4 2.5/5 15.5 1.4 18.1 1.1 20.8 2.1 19.5 1.7 5/5 14.6 1.5 16.2 1.4 20.1 1.4 19.2 1.3 10/5 14.7 1.6 15.2 1.7 19.7 1.1 20.8 1.5 20/5 14.3 1.6 19.5 1.5 20.9 1.9 30/5 20.1 1.3 20.5 1.7 0/1 16.5 1.4 18.7 1.2 2.5/l 16.7 1.7 19.0 1.8 20.0 1.3 19.5 1.3 5/1 15.8 1.5 17.8 1.1 20.8 1.2 20.8 0.7 10/1 15.6 1.5 17.3 1.2 20.8 1.4 20/1 15.1 1.7 19.6 1.6 20.9 1.5 30/1 19.1 1.6 22.0 2.8 0/0.5 16.3 1.9 17.6 1.6 21.2 2.3 20.9 1.8 2.5/0.5 16.9 1.4 18.5 1.1 20.9 0.9 20.1 1.7 5/0.5 19.2 4.0 18.5 0.8 20.6 1.7 20.3 1.2 10/0.5 16.5 1.3 20.5 1.4 21.0 2.0 20/0.5 19.8 1.4 21.4 1.3 30/0.5 19.2 1.2 139 7.8 REFERENCES Drescher B., “Reversible Block/Graft Copolymers as Emulsifiers”, Diplomarbeit (Dip1.-Ing.), Universitat Kaiserslautem, 1997. Kriiss Germany, “Processor Tensiometer K12”, User Manual, Kruess GmbH, Wissenschaftliche Laborgerate, Hamburg, Germany, 1992. Krtiss U. S. A., “Drop Volume Tensiometer DVT-10”, User Manual, Krtiss U.S.A., South Natick, Massachusetts, 1992. MacLeod C. A. and Radke C. J ., J. Coll. Inter. Sci, 160, 435 (1993). J oops P. and Van Uffelen M., J. Coll. Inter. Sci, 171, 297 (1995). Kriiss U. S. A., “Drop Volume Tensiometer DVT-10”, tutorial, Kriiss U.S.A., South Natick, Massachusetts, 1992. 140 Chapter 8 DROPLET SIZE ANALYSIS OF EMULSIONS STABILIZED WITH P(MAA-G-EG) BLOCK COPOLYMERS USING LASER SCANNING CONFOCAL MICROSCOPY 8.1 INTRODUCTION In this chapter, the quality and stability of the emulsions stabilized by P(MAA-g- EG) copolymers are characterized for four different molecular architectures using laser confocal microscopy to determine the droplet size distribution over time. These studies complement the more macroscopic characterization of the emulsions that were reported in Chapter 6, and provide important information about the size of the droplets stabilized by these copolymeric emulsifiers. Laser scanning confocal microscopy was selected to characterize the droplet size distribution because it offers important advantages over the other alternative techniques.l The unique feature of confocal microscopy is that is allows optical sectioning to produce images of the droplet cross-sections in a focal plane beneath the surface of the emulsion. Laser Scanning Confocal Microscopy (LSCM, also known as CSLM, Confocal Scanning Laser Microscopy) is now an established tool to provide high-resolution images and is predominantly used for a variety of biological specimens.2’3 We selected LSCM to characterize the droplets stabilized by the P(MAA-g-EG) copolymers because it allows the studies to performed in the native (aqueous) state of the emulsions, and therefore 141 allows the droplet size distribution to be monitored over time. In this chapter, we will describe the use of the LSCM technique to perform optical sectioning in order to analyze the droplet size distribution in an emulsion over time. To our knowledge, this the first time the technique has been used for this purpose. 8.2 BACKGROUND ON LASER SCANNING CONFOCAL MICROSCOPY The principle of Laser scanning confocal microscopy is illustrated in the schematic diagram in Figure 8.1. In LSCM, a preconditioned laser beam is passed through an objective lens that focuses it onto a tiny spot beneath the surface of a fluorescent specimen. The reflected light from the laser beam along with the fluorescent light emitted from the sample is captured by the same objective and directed to a dichroic mirror (beam splitter). This beam splitter allows the scattered laser light to pass through, but reflects the fluorescent light through a collecting lens and onto a photo multiplier tube (PMT). A confocal aperture (pinhole) placed in front of the photo detector, allows only the laser light from the focal plane to reach the detector. Light emitted from any plane above or below the focal plane falls outside the pinhole and is not allowed to reach the detector. A two-dimensional image in the focal plane is produced by a laser seaming process. As the laser scans across the specimen, the analog light signal, retrieved by the photo multiplier, is converted into a digital signal, producing a pixel-based image displayable on a computer monitor attached to the LSCM. The plane of focus (z-plane) is selected by a computer-controlled fine-stepping motor, which moves the microscope stage with the specimen up or down. Furthermore, scanning the object in z-direction, along the optical axis, at an equidistant step-size will 142 generate a series of two-dimensional grayscale images representing optical sections through a specimen.4 Modern instruments are built around a conventional light microscope, using laser light, sensitive photomultiplier tube detectors, and a computer to facilitate the collection and display of the images. For example, with the modern laser scanning confocal microscope (LSCM) LSM 210 manufactured by Carl Zeiss GmbH, (Gdttingen, Germany), the image is built up from the output of a sensitive photomultiplier tube, directly processed in a computer imaging system and then displayed on a high- resolution monitor, and recorded on hardcopy devices. In the LSCM, illustration and detection are confined to a single, diffraction limited, point in the specimen. 8.3 EXPERIMENTAL 8.3.1 Experimental Considerations The objective of this investigation is to analyze the droplet size distribution of the oil droplets that are dispersed in the continuous aqueous phase in the presence of the P(MAA-g-EG) emulsifiers. By coloring the dispersed droplet with a hydrophobic dye, the Laser scanning confocal microscope will be used to detect only the light due to fluorescence that is emitted from the droplets. The non-fluorescing water will provide a black background. By moving the point of focus into the sample at a series of equidistant gaps, a series of cross-sections throughout the sample will be compiled. By analyzing the shape and size of the droplet cross-sections in each frame, information about the droplet size distribution will be gathered. In the following sections, the sample preparation, data collection and analysis are presented. 143 8.3.2 Sample Preparation As described in the previous two chapters, aqueous solutions of P(MAA-g-EG) copolymers of varying in molecular weight and architecture were prepared at concentrations of 0.5 wt.%. The copolymers employed for this study were P(MAA-g- EG) 2.5/ 10, 2.5/ 1, 5/5, and 5/ 1. The model oil phase methyl laurate was colored with the fluorescing dye DiO (3,3 Dioctadecyloxacarbocyanine perchlorate). This dye is completely hydrophobic and has an absorption maximum at 498 nm, as illustrated in the UV-visible spectrum shown in Figure 8.2) and emitted a fluorescence signal in the green region of the visible spectrum around 530 nm. For the LSCM studies, the pH of the P(MAA-g-EG) emulsifier solution was adjusted to a value of 2. An emulsion was prepared by blending 66.7 vol.% of methyl laurate with HPLC grade water using a laboratory homogenizer (Fisher Scientific, Atlanta) for about 30 seconds. Using a pipette, small quantities of the emulsion were placed on a hanging drop slide also known as “well slide” (Fisher Scientific, Atlanta). The well slide was sealed with a cover slip (Fisher Scientific, Atlanta) that was affixed to the slide with stopcock grease. 8.3.3 Image collection and processing The emulsion was observed using the confocal Laser scanning microscope Zeiss LSM 210. Each slide bearing an emulsion stabilized with the P(MAA-g-EG) copolymers was analyzed three times to investigate the emulsion stability over time; immediately after the emulsion was prepared: 7 days after the emulsion was prepared; and finally 14 days after the emulsion was prepared. Images were collected in confocal fluorescence 144 mode, using the 488 nm line of a dual line argon-ion laser. The DiO fluorescence was detected using a BP 520-560 nm band pass barrier filter. With out-of-focus light thus removed, it is possible to see into a sample for a distance of several micrometers. For confocal investigation of the specimen we chose the objective lens with the highest numerical aperture available. The numerical aperture (NA) of an objective lens is a measure of its light-collecting ability and has an impact on the thickness of the optical section and controls the final resolution. The higher the NA value, the thinner the optical section will be. In our case the objective with the highest NA available was a 40X oil (immersol 418) neofluar objective with a NA equal to 1.3. With this objective, the optical section thickness is on the order of 300 nm. However, the vertical resolution is never as good as the lateral resolution. A useful feature of this Zeiss LSM 210 instrument is the ability to zoom an image with no loss in resolution using the same objective lens. This is achieved simply by decreasing the area of the specimen scanned by the laser by controlling the scanning mirrors and by placing the information of the scan into the 512 x 512 array of computer memory. Hence we could achieve an additional magnification by the factor 50. Therefore, the combined physical and digital magnification summed up to a total of 2000 for all carried out measurements. For this application the determination of the real diameter of a droplet of interest is possible with the use of the built-in mode of a Z-series.4 A Z-series is a sequence of optical sections collected at different levels from a specimen. Z-series were collected by coordinating the movement of the fine focus of the microscope with image collection, using a computer- controlled stepping motor to move the stage by a preset distance of invariably 1 micron. This is accomplished using a macro program that collects an image, moves the focus by 145 the predetermined distance, collects a second image, stores it, moves the focus again, and continuous in this manner until several images through the region of interest have been collected. Prior to storage, line averaging was engaged and set to four scans. Fifteen successive pictures were collected per series. The laser was operated at 1% intensity using filter 2 and only for the period of focusing and data collection in order to prevent heat effects or the evaporation of the water in the emulsion. 8.3.4 Data reduction In Figure 8.3 the first optical section of a Z-series for the emulsion stabilized with the P(MAA-g-EG) 2.5/ 10 copolymer is shown. In this figure, the circular cross-sections of the droplets containing the fluorescent dye are clearly visible for a focal plane several microns beneath the cover slip. The overall quality of the picture is good and the bright circles can be clearly distinguished and their diameters can be measured. In Figure 8.5 the complete Z-series of twelve optical sections is shown for the emulsion stabilized with copolymer P(MAA-g-EG) 2.5/ 10. In this figure, the upper left image corresponds to the first optical section, which is in the focal plane closest to the surface (~10 microns beneath the cover slip). The second and third images in the top row are optical sections in focal planes one micron and two microns beneath the first image, respectively. Similarly, the second row of images in the figure are optical section four, five, and six microns beneath the first image, and each successive row corresponds to a series of focal planes deeper into the sample. To establish the actual diameter of each droplet, it is necessary to follow the droplet through a series of optical sections in order to identify the focal plane that 146 traverses the center of the drop (the focal plane in which the cross-section of the droplet is the largest). This procedure is illustrated schematically in Figure 8.4. Droplets with diameters smaller than 2 microns could not be reliably measured since the probability of the focal plane passing through the center of the droplet is small. Figure 8.5 illustrates that the image quality is reduced as the depth of the focal plane is increased. This reduction in image quality arises from the reduction in the fluorescent light intensity due to diffraction and scattering of the light. Therefore, the images from the focal planes deep in the sample become darker and harder to distinguish (especially the smallest droplets). Due to this effect, it was not possible to see further than about 25 microns into the samples. For this reason, only the droplets that appeared in the first Z-section are included in the analysis of the droplet size distribution. As described above, these droplets are monitored in successive focal planes until the actual diameter is found, and since the most of the droplets have a diameter less than 20 microns, the droplets were well described by the first five or ten images. At least fifty distinct droplets were measured to characterize each distribution. It should be noted that while the LSCM is useful for characterizing emulsions stabilized by the P(MAA-g-EG) copolymers, it cannot be used to characterize all emulsions. This became very clear when we tried to use the technique to characterize an emulsion prepared with 0.5 wt.% of the commercial surfactant Tween 60 (Aldrich, St. Louis, Mo). Optical sections taken for this surfactant yielded no usable images since it was impossible to reliably distinguish the cross-sections of the droplets, presumably because there was an overwhelming number of small droplets on the order of 2 microns 147 or smaller. The droplet sizes observed for the P(MAA-g-EG) polymers happen to fall in the ideal size range for the LSCM technique. 8.3.5 Results The methyl laurate droplet sizes in the emulsions stabilized by the P(MAA-g-EG) copolymers 2.5/10, 2.5/1, 5/5 and 5/1 were estimated by a statistical analysis of the optical sections obtained by laser scanning confocal microscopy. For each analysis 80 to 100 droplets were used, and the measured real diameters were sorted into 15 classes ranging 2 microns to 60 microns (few droplets had diameters higher than 30 microns, and the overwhelming number of droplet fell in the range of 4 to 15 microns). A series of histograms showing the droplet size distributions for the four copolymers in this study are shown in Figure 8.6 through 8.9. These figures show the normalized frequency distribution of the droplet size for each sample at the three different times (day 1, day 7, and day 14). Each bar of the histogram spans four microns, and the class mark diameter was set at the median of each class. In addition, the mean a and standard deviation 0 of each distribution are marked in the figures. All four copolymeric emulsifiers exhibit similar trends for the droplet size distribution as a function of time. The mean droplet diameter shows little change of the course of 14 days, perhaps increasing very slightly (for the 2.5/ 10 copolymer, the mean diameter was found to be 11.2, 9.1 and 9.7 microns after 14 days, 7 days and on the day of emulsion preparation, respectively). Similarly, the standard deviation of the distributions was remarkably consistent over the fourteen day time span (e. g. the standard deviation is 7.9, 6.5, 7.4 after 14 days, 7 days and on the day of emulsion preparation 148 respectively for the 2.5/10 sample). These results indicate that the emulsion are very stable, in agreement with the more macroscopic emulsion stability studies performed with the naked eye (the emulsions are observed to be stable for months). Figure 8.10 shows a plot of the number average droplet diameter of methyl laurate stabilized in water by all four P(MAA-g-EG) copolymers in this study. The figure illustrates that the average droplet size is essentially independent of both the copolymer composition and molecular weight for the four copolymers investigated in this study. In addition, the average droplet size is essentially independent of time, with the possible exception of the 20:1 copolymer of high molecular weight, which shows a slight increase in the average droplet diameter. In general, the average droplet size is on the order of 10 microns. 8.4 CONCLUSIONS As a first important result, the technique of optical sectioning using confocal fluorescence spectroscopy can be successfully employed to determine droplet size distributions of methyl laurate/water emulsions stabilized with P(MAA-g-EG) reversible emulsifiers. The technique is most application to systems that have a droplet size on order of ten microns. If the droplet size is below about 2 microns, the distance between successive optical sections is too large to reliably assign the diameters. If the droplet size is above about 50 microns, it is difficult to measure the droplets since the image quality deteriorates for focal planes more than 25 microns into the sample. The LSCM technique was used to characterize the droplet size distribution of methyl laurate droplets dispersed in the continuous aqueous phase in the presence of four 149 P(MAA-g-EG) emulsifiers of different composition and molecular weight (the P(MAA- g-EG) 2.5/10, 2.5/ 1, 5/5 and 5/1 were tested). The average droplet size was found to be essentially independent of both the copolymer composition and molecular weight, and the number average droplet diameters were found to be in the range spanning from 8 to 12 microns. In addition, the average droplet size was found to be essentially independent of time, with the possible exception of the 20:1 copolymer of high molecular weight, which shows a slight increase in the average droplet diameter. 150 Beam Splitter Collecting Lens Objective Lens Confocal Focal Plane Unit with Pinhole Out-of-Focus Area Figure 8.1 Principle of Laser Scanning Confocal Microscopy. 151 ~12” Extinction [-] 1 O 290 390 490 590 690 790 wavelength [nm] Figure 8.2 Exitation Spectrum of DiO (3,3 Dioctadecyloxacarbocyanine perchlorate). 152 Figure 8.3 “Dio”-Fluorescence image of methyl laurate droplets in water as measured with the Zeiss LSM 210 at a resolution of 300 pixels per inch at an absolut magnification of 2000. The image width/height ratio was corrected from 512/512 to 768/512 using Adobe Photoshop 3.0. 153 Optical Sections and Z series Figure 8.4 Procedure to identify the actual droplet size from a series of optical sections 154 Figure 8.5 Z-series of “Dio”-Fluorescence images through methyl laurate droplets stabilized in water with P(MAA-g-EG) 2.5/10 copolymers at a solution pH of 2. as measured with the Zeiss LSM 210. The stepsize was set to lmicron. The pictures are organized left to right and top to bottom with increasing depth. 155 0.50 0.45 _ 0,40 d 0 9'- 9” mas/10 - day 14 0.35 0.30 0.25 0.20 0.15 0.10 0.05 ‘1 0.00 . 1 1b: . 1 . 0.50 3:: 9'0 9. 9+0 035 2.5110 - day 7 0.30 0.25 0.20 0.15 0.10 \ 0.05 S 0.00 1 W '4'0 ..... 50 ..... 6'0 0.50 0.45 d d 33;) -'° - 9*“ r111 2.5/10-day0 0:30 0.25 0.20 0.15 0.10 0.05 0.00 1 v 1 u 211W class mark diameter dI W1 normalized distribution of droplets \\\\ \\\\ Figure 8.6 Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g-EG) 2.5/10 measured after preparation, and 7, 14 days after preparation. 156 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0.50 0.45 0.40 0.35 0.30 0.25 0.20 O. 15 0.10 0.05 0.00 0.50 0.45 0.40 0.35 0.30 0.25 0.20 O. 15 0.10 0.05 0.00 normalized distribution of droplets g-o Q gm m 2.5/1- day 14 'I-I'm—I-i-I-I-I'T'I-I—l—l—F“ 10 20 30 40 50 60 g-o 9. gm 2.5/1 - day 7 figm 10 20 30 40 50 60 9'0 9 9+0 // 2.5/1 - day 0 Z 771/ 172171 1b. u u uzbu r r 1 Wm class mark diameter d. Figure 8.7 Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g-EG) 2.5/1 measured after preparation, and 7, 14 days after preparation. 157 0.45 8.33 m 5/5 - day 14 0.30 "'1 0.25 I? 0.20 >39; 0.15 . .339: 0.10 >qu >34: 0.05 a: :ozo: . 000 9.4! A. 9.0 . . ...... normalized distribution of droplets class mark diameter d1 Figure 8.8 Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g-EG) 5/5 measured after preparation, and 7, 14 days after preparation. 158 0.50 :2: .- . 0.35 m - day 14 0.30 0.25 0.20 0.15 0.10 0.05 ‘ 0.00 . 1 u r W 0.50 0.45 0.40 5‘- 9”“ 0.35 m 5/1 - day 7 0.30 0.25 0.20 0.15 0.10 S ‘ N Q. 0.00 \ 1 \ m 11. rrr.3b.‘FF-TFW [lf7 //// 1// A d' I q normalized distribution of droplets in \\\\ 0.50 0.45 040 9" 0'35 5/1 - day 0 0.30 0.25 0.20 0.15 0.10 0.05 2 0.00 . . . . .1 .IZZUZQO class mark diameter 11' Figure 8.9 Histogram of the normalized frequency distribution for the methyl laurate droplet diameter in a 66.7/33.3 by volume methyl laurate/water emulsion stabilized with P(MAA-g-EG) 5“ measured after preparation, and 7, 14 days after preparation. 159 22 ' r r F 1 I ' 1 1 T r T ' I ' f 20fl MAA:EG M _ - D 20:1 17000 4 18: 0 20:1 116000 1 E 16- A 10:1 25000 _ 1 V 10:1 119000 ' a) 14~ O - E 1 '4 £2 12- 4 9 . 1:1 . g 10- 8 0 e — 6 - E a, 8- _ c) 4 S 64 _ m . > cu 4- - 21 _ O ' l ' 1 ' 1 ' 1 ' I fit ' r ' I 0 2 4 6 8 10 12 14 16 time[days] Figure8.10 Number average droplet diameter of methyl laurate stabilized in water by P(MAA-g-EG) copolymers as a function of time. 160 8.5 REFERENCES Paddock, S.W. (editor), Confocal Microscopy Methods and Protocols, Humana Press (1999). Stevens, J .K., Mills, LR. and Trogadis, J .E. (editors), Three-Dimensional Confocal Microscopy: Volume Investigation of Biological Systems, Academic Press, San Diego (1994). Gu, M., Principles of Three-Dimensional Imaging in Confocal Microscopes, World Scientific Publishing Co, Singapore (1996). Whallon, J. H., Introduction to Laser scanning confocal microscopy, Laser Scanning Microscope Laboratory, Michigan State University, East Lansing (1993). 161 Chapter 9 MOLECULAR CHARACTERIZATION OF THE P(MAA-G-EG) COPOLYMER COMPLEX USING THE NMR NUCLEAR OVERHAUSER EFFECT In this chapter, an experimental characterization of the conformation of the P(MAA-g-EG) copolymer under acidic conditions is presented. Specifically, the spatial arrangement of the pendent oligo(ethylene glycol) grafts in relation to the poly(methacrylic acid) backbone is investigated using 2-dimensional NMR nuclear Overhauser effect spectr0300py (N OESY) studies. The nuclear Overhauser effect (N OE) is an NMR relaxational phenomenon that is manifested by a change of the intensity of a peak at one frequency by irradiation of a peak at a different frequency. This phenomenon can be used to estimate the inter-proton distances for the protons that experience the NOE magnetization exchange (for this exchange to occur, the must be within about five angstroms of one another). The following section provides the theory behind the nuclear Overhauser effect. This is followed by the theoretical description of the 2-dimensional NMR experiment employed in this study. Finally, the actual NOESY experimental results for the P(MAA-g-EG) copolymers are presented and discussed. 162 9.1 THE NUCLEAR OVERHAUSER EFFECT In NMR spectroscopy, the nuclear Overhauser effect (NOE) refers to a phenomenon in which irradiation of a resonance frequency for one proton (or other NMR active nuclei) results in the change in the intensity of a peak corresponding to other protons within the sample.I The importance of NOE lies primarily in the fact that the resonances that change their intensities arise spins close in space to those directly affected by the perturbation. The NOE has its origin in the population changes brought about by a particular form of relaxation, namely dipole-dipole—relaxation. The phenomenon derives its name from Albert Overhauserz, who in 1953 predicted the saturation of electrons in a metal would produce a large polarization of the nuclear spins. Subsequently, Solomon3 reported the first experimental observation of NOE and presented a theoretical description that is still widely used today. Aanet and Boum4 first illustrated the usefulness of the technique for structural elucidation, and Balarams reported its application to large, biological molecules. Today, NOE is used extensively for the structural determination of small organic molecules“7 and large molecules8 such as proteinsg. A theoretical description of the nuclear Overhauser effect in a system of two spins, I and S, which are dipole-dipole coupled, but not scalar coupled was given by Solomon.3 This description is used to explain intramolecular proton-proton NOE in organic and biological molecules. The system of two weakly coupled spins has four possible energy levels. The highest energy level corresponds to both spins I and S in their high-energy states aligned opposite to the external magnetic field. This state is denoted as 0t0t. In this nomenclature, the first symbol corresponds to the nuclear spin 163 state of I, while the second symbol corresponds to the nuclear spin state of S. The lowest level corresponds to both spins aligned with the field, and is denoted BB. Finally, the intermediate energy levels correspond to the spin aligned with one opposed to the field, and are denoted 0tB, B0t. In the case of homonuclear dipole-dipole coupling, the intermediate energy levels have nearly the same energy. The four possible energy states are illustrated schematically in Figure 9.11, and are numbered one through four. Three types of nuclear magnetic transitions can be identified for this two spin system: zero quantum, single-quantum, and double-quantum transitions. Zero quantum transitions occur between the two intermediate energy levels as the two spins exchange orientations in the magnetic field. Using the nomenclature defined by Solomon3, these are transitions from 0tB to Ba and vice versa. In homonuclear systems, the zero-quantum transitions involve very little energy change. Double-quantum transitions involve simultaneous flipping of both I and S, and therefore involve transitions between the highest and the lowest energy states, 0t0t to BB and vice versa. Finally, in single-quantum transitions, only one spin changes its orientation in the field, e.g. 0LB to BB. Zero- quantum, single-quantum and double-quantum transitions are characterized by the rate constants W0, W. and W2, respectively. Note that the change in energy accompanying the W0 process for a homonuclear two-spin system is much smaller than that accompanying the W1 process, which in turn is smaller than for the W2 process. These transitions are illustrated in Figure 9.1. At equilibrium with the external magnetic field, the nuclear spin states are populated according to their Boltzmann distributions.lo If the system is pulsed and a spectrum is acquired, equilibrium NMR peak intensities of lo and So would be obtained. 164 After a pulse, the system relaxes via the three types of transitions discussed earlier. However, only the single-quantum transitions are observed directly in the NMR spectrum.'0 The zero-quantum and double-quantum transitions are not observed directly, but their effects on the relaxation rate can be determined. After a pulse, the return of the peak intensity to equilibrium may be described by the following equations: d1. 9.1 -d—t,=—pt(11_IIO)—UU(IJ_I?) ( ) d] (9.2) jaw, 4041.11.- -1.-°> p.- = 2W.’ + W0 +W2 (9'3) 0’] = all z W2 _W0 (9.4) Here, the spin-lattice relaxation rate constants, pi, describe the exponential return of the peak intensity to equilibrium without explicit regard to the relaxation of other species. The coupling of two spin species is described by the term 0'in. This term represents the influence of themagnetization of a second spin, j, on the evolution of the magnetization of the spin under consideration, i. If the sample is irradiated at the Larmor frequency of spin S, S will become saturated, i.e. it will have equal populations in its 0t and B spin states. Since S is coupled to I by dipole-dipole interactions, the saturation of S will also affect the distribution of I among its spin states. Since the intensity of the NMR signal is proportional to the difference in populations of the two energy levels, the NMR 165 signal for I will be affected. Upon radiation of S, the NMR peak due to I may increase or decrease depending on whether zero-quantum or double-quantum transitions are more efficient. Double-quantum transitions increase the intensity of l by attempting to establish a Boltzmann distribution between levels one and four, while zero-quantum transitions decrease the intensity of I by equilibrating levels two and three. Zero- quantum transitions involve very little energy change, and therefore require very low- frequency fluctuating magnetic fields, while double-quantum transitions require magnetic field fluctuations at about twice the larmor frequency. Therefore, positive proton-proton NOE enhancements are observed for mobile small molecules, while negative NOE enhancements are observed for polymers or large biomolecules.l The fractional enhancement, or NOE factor, of spin species j is defined as: 1,. - If (95) O 1]. f}: In the case of two like spins (i.e. two proton spins) the steady-state NOE may be expressed as a function of the single- and double-quantum relaxation rates: W, —W, (9.6) 2W," + W0 + W, f,(°°)= While only limited information can be obtained from steady-state NOE measurements, much more information can be obtained from the rate of NOE evolution. For example, distance information cannot be obtained from the steady-state NOE experiment, but can be determined from the initial rate of evolution of the NOE. The initial rate of evolution is proportional to rls’6, where rig is the distance between the spins 166 I and S. For qualitative NOE 1H NMR measurements, no attempt is made to determine a specific interproton distance based on the magnitude of the observed NOE. Rather it is simply assumed that observation of a NOE between two protons implies that they are separated by a distance of around 4 A.'2 Alternatively, more detailed structural information can be obtained from quantitative NOE measurements. In a multispin system, the magnitude of the observed NOE is the result of a complex set of internuclear interactions. Consequently, it is difficult to extract a precise interproton distance from a NOE measurement made with any single irradiation time. However, the size of the NOE builds up exponentially as the irradiation time is increased, and the rate of increase, Oij, is easier to interpret than any single NOE measurement. In order to get an estimate for an unknown interproton distance runk, the NOE buildup rate between two protons of known distance must be measured so that this known distance rkmwn can be used as a "molecular ruler". If the NOE buildup o is measured for both the known and unknown distances, and if each pair of protons behaves as an isolated spin pair, then the two-spin approximation can be employed to estimate unknown distance according to equation 9.7: 1/6 (9.7) rank = (chow, ] rknown 0' am]: Good estimates of the cross-peak buildup rates are difficult to obtain. Clore and Gronenbomll have proposed that no accuracy is lost if distances are estimated from a 2D NOE spectrum obtained with a single mixing time. Since the cross-peak intensity is zero at a mixing time of zero seconds, the ratio of buildup rates oknown/ounk in equation 9.7 can 167 be approximated by the ratio of the cross-peak intensities corresponding to the NOE for the known and unknown distance obtained for a single mixing time. Several authors have used pre-steady-state NOE measurements to obtain information about interproton distances. This technique is particularly useful for the study of biological macromolecules because it allows the conformational studies to be performed in solution. Since NOE is a through-space effect that is present only if the nuclei are within several A of one another, these experiments provide a valuable tool for 2 NOE measurements allow establishing the three-dimensional structure of proteins.1 connectivity between protons on topologically remote parts of a protein to be established. The NOE growth rate has been used to study complexation systems.”"l4 Klier used truncated driven NOE to analyze complexation between poly(methacrylic acid) and oligomeric poly(ethylene glycol). These experiments provided information about the existence, stability, and stereochemistry of the complexes in dilute solution. The initial NOE growth rate was proportional to the fraction of bound repeating units; therefore NOE provided a very sensitive measure of the existence of the complex. Klierls observed an increasing initial grth rate as the molecular weight of the oligomer was increased, and found no NOE in complex-breaking solvents such as diluted sodium hydroxide and methanol. 9.2 THE TWO-DIMENSIONAL NMR NOESY EXPERIMENT In a 2D NMR experiment, a pulse sequence is employed instead of a single pulse, and the response is detected as a function of two variables. The first time variable t] is related to the spacing between pulses, and, for a given value of ti, a FID is detected at the 168 end of the pulse sequence as a function of a second time variable t2. The time variable t. is systematically changed by a fixed increment (Atl) and a F ID is acquired in t; for each t. value in turn. Hence a sequence of FIDs corresponding to different t. values is acquired, which can be thought of as a data matrix D(t1,t2) in the two time variables t1 and t2. Each row of the matrix corresponds to a FID corresponding to a specific t; value. A Fourier transform of any row in the time variable t; yields a conventional NMR spectrum. Consequently, the intensity of a given signal as a function of t. forms an interferogram or pseudo-F ID. The amplitude of the resonances in the spectra varies regularly as a function of t1. A trace constructed along the center of the resonance peaks in successive spectra follows a cosine wave. A second Fourier Transform, this time with respect to t1, extracts the frequency of this variation and results in a 2D spectrum. In other words, the 2D NMR spectrum is obtained from a two-dimensional Fourier transform of the data matrix D(t;,t2), in both time variables to yield a spectrum of intensity as a function of two frequency variables S((1)1, 002). For a given signal in the 2D NMR spectrum, the frequencies 0); and (02 reflect the modulation frequency of that signal during the time periods t1 and t2, respectively. Two-dimensional exchange spectroscopy correlates spins related to each other due to magnetization exchange arising from chemical exchange or from the Nuclear Overhauser effect. The most common form of 2D-exchange spectroscopy is lH two- dimensional Nuclear Overhauser effect spectroscopy (2D NOE or NOESY) which has the pulse sequence shown in Figure 9.2. The NOESY pulse sequence consists of an initial delay called presaturation (presat) that allows the magnetization to return to equilibrium from the previous cycle. The first 90° pulse flips the magnetization vector 169 into the xy plane and labels all protons according to their characterization precession frequencies 0);, determined by their chemical shifts, during the evolution period. After a time t;, the second 90° pulse flips the magnetization along the -z axis, and magnetization exchange occurs via the Nuclear Overhauser effect during a so-called mixing period 1:...- The third 90° pulse flips the magnetization back into the xy plane, and the final precession frequency (0213 detected during the detection period. Those spins that do not exchange magnetization have a final precession frequency equal to the initial frequency. Thus the NOESY spectrum contains a set of autocorrelation peaks along the diagonal of the spectrum defined by (05002. That is, all the signals observed in the 1D spectrum lie along the diagonal of the 2D NOESY spectrum plotting frequency 0). as a function of (02. Such a plot is referred to as a contour plot with regions of equal intensity comprising an envelope. Useful information only emerges if spins exchange magnetization during the mixing period. Here, two spins with unique chemical shifts, l and S, are considered with S having a somewhat slower precession frequency. If t; is arbitrarily chosen such that the angle of one spin I is equal to 11/2, the 2 component of I has zero intensity at the start of “cm, while the corresponding S vector, which did not precess as far during t1, is still largely inverted. Cross relaxation during I... thus results in a transient NOE enhancement at I. For values of t; where I is more inverted than S, a transient enhancement will occur in the reverse sense, appearing at S. Thus, by the end of the mixing period 1",, the intensity of the I vector which was initially given just by —cos(u)lot.) has acquired an additional dependence on —cos(0)sot1). The intensity of I at the end of rm has acquired a new modulation at (103. The extent i.e. amplitude, of this new modulation depends on the size and sign of the transient 170 enhancement. On Fourier transform with respect to t. this yields a major peak at F1=0)1, and a minor peak at Fl=0)s. F; or F2 denote the ordinate of the 2D-NOESY plot after Fourier transform with respect to t. or t2, respectively. The major peak is located on the diagonal, while the minor peak is a cross peak connecting I and S at F1=ws and F2=0)1. Similar conclusions apply in reverse to other intensities encountered in the spectrum. In a contour plot of a 2D spectrum, for spins that exhibit NOE interactions, the final frequency is different from the initial frequency and the magnetization exchange gives rise to signals not on the diagonal known as off diagonal cross-peaks. As a requirement, nuclei that show NOE interactions have to be spatially close in order to be detected in an NOESY experiment. NOESY data can be used either for qualitative or quantitative analysis. For qualitative analysis, the only information that is required from the NOESY spectrum is whether or not a NOE exists between two protons. The NOE is dependent on the inverse sixth power of the interproton distance (see equation 9.7), so NOE interactions are only observed between spatially close pair of protons. 9.3 FORMATION AND DISRUPTION OF THE COMPLEX BETWEEN POLYETHYLENE GLYCOL CRAFTS TO THE POLY(METHACRYLIC ACID) BACKBONE 9.3.1 Sample Preparation The intensity of a the NMR signal is proportional to the concentration of the protons present, therefore, relatively high concentrations (~8 wt.%) of the P(MAA-g-EG) copolymer were used to obtain an adequate signal to noise ratio in a reasonable period of time. Preliminary solubility experiments at this high concentration were performed in the following manner. Approximately 0.1252 g of solid P(MAA-g-EG) polymers denoted 171 2.5/5, 5/5, 10/5 were brought in contact with pure 1.25 ml deuterated water (D20) (Aldrich, St. Louis, MO) under constant stirring for one day. While 2.5/5 dissolved completely to give a viscous solution, the samples denoted 5/5 and 10/5, having a higher grafting level, did not appear to dissolve to the full extent (5/5) or at all (10/5) at their natural solution pH. Hence, further 2D NOESY studies were performed using the 2.5/5 sample. For the NOESY experiment, 0.1252 g of P(MAA-g-EG) 2.5/5 was dissolved in 1.25 ml of D2O under slow stirring. The vial was sealed with a rubber seal and flushed with N2 for 10 minutes to evacuate oxygen. The 2D NOESY spectrum of the pure polymer in D2O, at the pH of ~2 (without any base added), was acquired. At this acid pH, the hydrogen bonded complexes of the oligo(ethylene glycol) grafts with the poly(methacrylic acid) backbone will be formed. For comparison purposes, it was also desired to perform the NOESY experiments under basic conditions in which the complex would be broken. As a reference, 1.2514 g of P(MAA -g-EG) 2.5/5 was dissolved in 12.5 mL of HPLC grade water. To determine the amount of NaOD to add to the NSR sample, the reference sample was titrated to a highly basic pH (~11) using a 11.57 M NaOH solution using an Eppendorf pipette (Hamburg, Germany). Based upon these results, the NMR sample was titrated to pH 11 by adding 75 ul of 11.57 M NaOD. Tire solution was subject to vigorous mixing before the second 2D NOESY spectrum was acquired. 172 9.3.2 Spin-lattice relaxation time T. analysis The spin -lattice relaxation time T. is a measure of how fast the 2 component of a the magnetization vector, M2, grows back to the initial size, Mo, after being perturbed from equilibrium with a radio frequency pulse.12 After a 1t/2-pulse, T. is the time it takes for M2 to grow back to a value of (1-1/e) of the original value it had immediately before the pulse. At a value of 5 times the relaxation time T., the percent recovery of M2 is 99.3% and, therefore, considered fully recovered. The relaxation time T. is derived from the results of a built-in inversion recovery sequence called “dotl” of the VARIAN 5.02 software. A plot of the rate of growth of the magnetization vectors during the T. recovery experiment of P(MAA-g-EG) 2.5/5 is shown in Figure 9.2. The complete T. data analysis for all peaks in the range of 0 to 4 ppm is given in Table 9.1. The largest T. recorded was T.=2.97 at a chemical shift of 3.6 ppm corresponding to the ethylene glycol protons. 9.3.3 Pulse width 1V2 calibration The pulse width (PW) in microseconds was measured for the a—methyl resonance in the range of 1.5 to 2.0 ppm. An array of different pulse widths ranging from 26 to 66 microseconds separated by a delay of the 5-fold maximum relaxation time T. was employed and the corresponding peak intensities were detected. The 0t-methyl peak intensity as a function of consecutively applied pulse width is presented in Figure 9.3. The peak intensity sinusoidal curve crossed the zero mark and completed a full cycle at a pulse width of ca. 47 microseconds. Hence, the 11/2 pulse was determined to be 11.8. 173 9.3.4 2D NOESY Experiment The 2D NOESY experiment was performed using a VARIAN 500 with an equivalent magnetic field strength of 500.619 MHz. The observed transmitter power was set to 59 dB. The pulse sequence of the 2D NOESY experiment is shown in Figure 9.4. The presaturation time was set to 1 second, the 11/2 pulses were set to the calibrated duration of 11.8 us, each. The mixing time was set to 0.1 s and the acquisition was stopped after 0.141 5, since the FID relaxed to zero amplitude after that time. The instrument was set to acquire 2048 complex points in the first t2 dimension with a sweep width of 7238.5 Hz. A total of 16 transients were averaged for each detection period t2. For the acquisition in the other dimension the same sweepwidth of 7238.5 Hz was selected to give a square fiequency window. The data size of detection period t2 was set to 1024 number of complex points. The time increments for both time-domains were chosen to be the reciprocal of the sweepwidth, which is 1/7238.5 5. 9.3.5 Results and Discussion The two-dimensional NOESY contour plot for P(MAA-g-EG) 2.5/5 at a solution pH of ~2 is shown in Figure 9.5. Recall that in a NOESY spectrum, the diagonal peaks correspond to the normal NMR spectrum, while the presence of off-diagonal peaks indicates that the two corresponding protons experience dipole-dipole coupling. In the NOESY spectrum in Figure 9.5, three peaks on the diagonal are clearly distinguished: the ethylene glycol peak at 3.9 ppm, the (rt-methyl peak from the methacrylic acid at 1.3 ppm, and the methylene peak from the methacrylic acid at 2.2 ppm. In addition, the figure shows off-diagonal peaks that correspond to NOE between the (Jr-methyl and the ethylene 174 glycol peak (cross-peaks at F .=1 .3, and F2=3.9 and vice versa), as well as the methylene and the ethylene glycol peak (cross-peaks at F .=2.2, and F 2=3.9 and vice versa). Finally, there are off-diagonal peaks that indicate the presence of NOE interaction between the or- methyl and the methylene of the poly(methacrylic acid). These peak locations and intensities are summarized in Table 9.2. Note that each peak was measured using three different level spacings in order to accurately locate the top of peak. The peak was best measured by the level spacing that resulted in the highest peak intensity. In Figure 9.6 the two-dimensional NOESY contour plot of the neutralized polymer (basic conditions) is shown. The most striking feature about the NOESY spectrum for the neutralized polymer is the absence of the off-diagonal peaks at 3.9 ppm. Under basic conditions, there is clearly there is no NOE interaction between the ethylene glycol and either the (rt-methyl or the methylene. Only the cross-peaks stemming from the methylene a-methyl NOE interaction can be seen. Since the a—methyl and the methylene are both bonded to the same carbon atom, they are intrinsically in relatively close approximation, and therefore exhibit an NOE enhancement even when the complex is broken. A clear result of the NOESY experiments is that shift from acidic to basic conditions leads to a transformation in the spacing of the ethylene glycol and backbone protons from close proximity under acidic conditions to further than 5 Angstroms under basic conditions. We attribute this change to the disruption of the complex. After the complex is broken, the ethylene glycol protons of the extended grafts too far to experience dipolar coupling with the methacrylic acid protons. Recall that the NOE intensity is related to the negative inverse sixth power of the interproton distance. 175 For a more quantitative analysis of the interproton distance of ethylene glycol protons to those from the methylene or (Jr-methyl groups, it is necessary to relate the intensity of the cross-peak of interest to a the intensity of a cross-peak for two protons with a known interproton distance. For our analysis, the (at-methyl, methylene cross-peak at the frequency pair of (1.3 ppm, 2.2 ppm) was used for this purpose, and the proton- proton distance was assumed to be 3.7 Angstroms for this peak (this was estimated based upon the bond geometry of a single methacrylic acid repeat unit).'6"7 The distance between the protons associated with the other off-diagonal cross-peaks in the NOESY spectrum are then calculated based upon equation 9.7. Using this analysis, the inter- proton distance between the a—methyl and the ethylene glycol was estimated to be 2.8 Angstroms while the inter-proton distance between the methylene and the ethylene glycol was estimated to be 3.5 Angstroms. The data in Table 9.2 illustrate that the nuclear Overhauser effect is more pronounced between the a-methyl and the ethylene glycol than it is between the methylene and the ethylene glycol (this leads to the smaller estimate for the interproton distance reported above). This suggests that in the complex the protons in the ethylene glycol repeat unit are spatially closer to the a—methyl than they are to the methylene, and that the (it-methyl plays a larger role in the hydrophobic stabilization of the complexes. A similar conclusion has been drawn based upon one-dimensional NOE growth rate studies.'8 To understand why the (It-methyl may be closer to the PEG than the methylene, it is useful to examine the molecular model of the complex shown in Figure 1.1. The figure illustrates that in order to form a complex with a one-to-one repeat unit ratio, the PMAA 176 backbone assumes a conformation in which all of the carboxylic acid groups (which participate in the complex) are on the same side of the backbone chain. This tends to locate the methylene protons on the far side of the backbone chain relative to the ethylene glycol repeat units. In contrast, location of the a-methyl group is less constrained by the backbone conformation, and the molecular model shows that in the backbone conformation that favors a one-to-one complex, some of the a-methyl groups lie close to the ethylene glycol repeat units. In addition, based upon the NOE results, we cannot rule out the possibility that inter-molecular aggregates form in which the a-methyl groups of one chain fall in proximity of the ethylene glycol repeat units of another chain. 9.4 CONCLUSIONS In this investigation, the complexation of PMAA with its grafted oligomeric EG was studied using 2D lH-NMR NOESY. A highly concentrated solution of P(MAA-g- EG) with an MAA:EG ratio of 20:1 was investigated. One clear result from the NOE experiments is that a shift from acidic to basic conditions, leads to a transformation in the spacing of the ethylene glycol and backbone protons from close proximity under acidic conditions to further than 5 Angstroms under basic conditions. This effect can be attributed to the disruption of the complex and the extension of the oligo(ethylene glycol) grafts away from the backbone. The NOE results also indicate that, in the complex, the protons in the ethylene glycol repeat unit are spatially closer to the a—methyl than they are to the methylene. Molecular modeling results suggest that this effect could arise from constraints placed on the PMAA backbone in order to accommodate the formation of a complex with a one-to—one repeat unit ratio. 177 Figure 9.1 Nuclear spin energy levels for a two spin system. 178 _1_ X "vL A U f :1- (soc) u + 02E] 03 0 Figure 9.2 A plot of the rate of growth of the magnetization vectors during the T1 recovery experiment of 0.125 g P(MAA-g-EG) 2.5/5 in 1.25 ml of D20 using a VARIAN VXR 500 spectrometer. 179 .4.-._.— \ "7 \/ A / ‘H\ v: " \\ Ml ,/'/ E f F \7 :2» ?> \ Figure 9.3 Signal intensity versus pulse width for a set of pulse width in the range of 26 to 66 us. The 1V2 pulse width was calculated to be 11.8 us. 180 11.8us 11.8 us 11.8w 141 ms PW PW mix pw Figure 9.4 Applied pulse sequence for 2D NOESY NMR experiment. 181 I ' I I" .1 r- v'l _ H l n I IJ 'ltl' “ - o ...... éifi _ H .h: ~.' “iii '. ' r"? r N 5- ;E‘E , . 8 r1 ‘.'r'll; ' h I! ~'.' II r- l' 'i i _ M lll' x ,1; l}; l . 1 ‘} g!) r " “l" r; I- 0.,: I I u- ' I 1"" t N W il" l' I i ‘: H." u '- 0 g .4 r0 1' to a B: ' h 9 Figure 9.5 2D lH-NMR contour plot of a solution of 0.125 g P(MAA-g-EG) 2.5/5 in 1.25ml D20 at a solution pH value of ~2. The level spacing represents the spacing relative intensity of successive contour levels and is set to be 1.27. The ratio between one intensity level and the next lowest is equal to the stepsize. 182 ”H A - E " n. 3 H in -m — 1— 'l .l + 0" .1 1—1 2“ 3‘ 4—4 5““ 6 Figure 9.6 2D lH-NMR contour plot of a solution of 0.125 g P(MAA-g-EG) 2.5/5 in 1.25ml D20 at a solution pH value of ~1 1. The level spacing represents the spacing relative intensity of successive contour levels and is set to be 1.1. The ratio between one intensity level and the next lowest is equal to the stepsize. 183 Table 9.1 Exponential data analysis of spin-lattice relaxation times T1 for observed resonances in the chemical shifi range of 0 to 4 ppm for P(MAA-g-EG) 2.5/5 in D20. Sample designation Chemical shifi [ppm] Relaxation time T] [s] ethylene glycol 3.6 2.97 methylene 2.74 0.42 methylene 2.71 0.41 methylene 2.61 0.48 methylene 2.48 0.51 minor unassigned resonance 2.04 0.43 (at-methyl 1.91 0.24 a-methyl 1.83 0.3 l (at-methyl 1.79 0.32 (It-methyl l .68 0.33 184 Table 9.2 Cross-peak distance analysis for pairs CH2—CH3, EG—CH2, EG—CH3 for P(MAA-g—EG) 2.5/5 in D20 at 25°C. As a control, three different stepsizes between the intensity levels were employed to ensure accurate reading. Cross peak Rel. frequency Level Spacing Intensity level Connection modulation F1, F2 [Ppm] CH2 — CH3 1.3, 2.2 1.1 28.4 EG — CH2 3.9, 2.2 1.1 2.0 EG — CH3 3.9, 1.3 1.1 8.6 CH2 — CH3 1.3, 2.2 1.2 26.6 EG — CH2 3.9, 2.2 1.2 2.1 EG — CH3 3.9, 1.3 1.2 7.4 CH2 — CH3 1.3, 2.2 1.27 25.5 EG - CH2 3.9, 2.2 1.27 1.9 EG - CH3 3.9, 1.3 1.27 7.4 185 10. 11. 12. 13. 14. 15. 16. 17. 9.5 REFERENCES Neuhaus, D., Williamson, M. P., The Nuclear Overhauser Effect in Structural and Conformational Analysis, VCH Publishers, New York, 1989. Overhauser, A. W., Phys. Rev., 89, 689 (1953). Solomon, 1., Phys. Rev., 99, 599 (1955). Anet, F. A. L., Bourn, A. J. R., J. Amer. Chem. Soc., 87, 5250 (1965). Balaram, P., Bothner-By, A. A., Dadok, J ., J. Amer. Chem. Soc., 94, 4015 (1972). Chazin, W. J ., Colebrook, L. D., Magn. Reson. Chem, 23, 597 (1985). Noggle, J. H., Schirmer, R. E., The Nuclear Overhauser Effect, Academic Press, New York, 1971. Dobson, C. M., Evans, P. A., Biochemistry, 23, 4267 (1984). Poulson, F. M., Hoch, J. C., Dobson, C. M., Biochemistry, 19, 2597 (1980). Sanders, J. K. M., Hunter, B. K., Modern NMR Spectroscopy - A guide for Chemists, Oxford University Press, Oxford, 1987. Glore, G. M., Gronenbom, A. M., J. Magn. Reson, 61, 158 (1985). Bruch, M. D. (editor),NMR Spectroscopy techniques, 2"d editition, Marcel Dekker, New York, 1996. Clore, G. M., Gronenbom, A. M., J. Magn. Res, 53, 423 (1983). Kushnir, T. Navon, G., Bull. Magn. Res, 6, 60 (1983). Klier, J ., Self-associating Networks of Poly(Methacrylic Acid-g-Ethylene Glycol), Ph.D. Dissertation, Purdue University, 1989. Zumdahl, S. S., Chemistry, 2nd ed., D C Heath and Company, Lexington, MA, 1989. McMurray, Organic Chemistry, Brooks/Cole, Pacific Grove, CA, 1992. 186 18. Scranton, A., Klier, K., Aronson, C. L., Complexation of Polymeric Acids with Polymeric Bases, in: Harland, R. S., Prud’homme, R. K., Polyelectrolye Gels, ACS Symposium Series, 480, Washington, 1992. 187 Chapter 10 A POLARITY-SENSITIVE FLUORESCENCE STUDY OF THE PH- DEPENDENT AGGREGATION OF P(MAA-G-EG) COPOLYMERS IN WATER 10.1 INTRODUCTION In this chapter, the aqueous aggregation behavior of poly(methacrylic acid-g- ethylene glycol) was investigated as a function of pH for polymers that vary in molecular structure using the polarity-sensitive fluorescence of pyrene. This analysis is based upon the fact that solvents of differing polarities lead to measurable and consistent changes in pyrene’s vibronic coupling and, hence, in its fluorescence spectrum.l In this chapter we utilize the polarity sensitivity of pyrene to reveal the polarity of microenvironments in aqueous solutions of P(MAA-g-EG) copolymers. The first section in this chapter provides the background of this method and how chromophores have been employed by other researchers. The rest of the chapter contains the results and discussion of the hydrophobicity domains of P(MAA-g-EG) in aqueous solution as determined by pyrene fluorescence. 188 10.2 PYRENE AS A PHOTOPHYSICAL PROBE TO CHARACTERIZE POLARITY Pyrene exhibits a fine vibrational structure in its fluorescence spectrum, and five major bands, from shorter to longer wavelength, are readily identified. The first band (I) ascending from lowest to highest wavelength is attributed to an actual overlap of many vibronic transitions but has been conveniently called the 0-0 band and nominally represents the transition between the lowest energy levels of the So and S. singlet states.2 It is this electronic transition that leads to the polarity-sensitive fluorescence of pyrene since this emission band is considerably intensified relative to the other bands in strongly polar solvents.2 There are numerous references in the literature that describe the use of pyrene as a polarity-sensitive fluorescence probe. This polarity-sensitive response, an example of the Haml effect, was characterized for pyrene by Nakajima3 when he investigated the effect of a nonpolar (cyclohexane) and a highly polar (N-methylformamide) solvent on the vibronic band intensities in the fluorescence spectrum of pyrene. The environmental effect on the vibronic band intensity in pyrene monomer fluorescence was utilized by Kalyanasundararn and Thomas4 in studies of micellar systems. The strong perturbation of the vibronic band intensities was used as a probe to accurately determine critical micelle concentrations and also to investigate the extent of water penetration in micellar systems. Bakeev et al.5 have used l-pyrenecarboxylaldehyde as a fluorescence probe to study the complexation of amphiphilic polyelectrolytes with surfactants of the same charge in aqueous solutions. In addition, the determination of the critical micelle concentration using pyrene as a solubilized fluorescent probe for diblock copolymersf’7 189 block electrolyte solutions}9 and triblock copolymers,l0 has been recently reported. The ability of pyrene to form excimers in aqueous solution has also been used to characterize micellar systems.ll Excellent review articles on the use of luminescence probes as sensors for probing the microenvironment in surfactant systems and investigating 1.12 [”1114 polymer association in water have been published by Turro eta and Winnick eta Other aromatic luminescent probes solubilized in aqueous media have also been used as probes for the study of micellar solutions.ls For example, the intrinsic excimer fluorescence of polystyrene-block-poly(ethylene propylene) has been used to study the block copolymer micelles and to determine the critical micelle concentration (CMC).l6 ”"8 and dypyme'9 for Investigators have also used 1,6-Diphenyl-1,3,5-hexatriene (DPH) the investigation of micelle formation by the Pluronic (BASF) class of triblock copolymeric surfactants. These investigators have utilized the environment-sensitive fluorescence of the solubilized fluorescent probe to obtain estimates for the CMC and to study micelle formation in the copolymer systems. The polarity-sensitive response of pyrene provides a valuable method for characterizing surfactant aggregation because the CMC of many of the block copolymer systems is too small to be determined by light-scattering techniques.9 In the presence of micelles, the pyrene probe partitions between the aqueous and micellar phases. This partitioning leads to a number of interesting changes in the (0,0) band in both the excitation and the emission spectra of pyrene. As the pyrene moves from the aqueous phase to the micellar phase, there is a red shifl in the excitation spectrum, a change in the vibrational fine structure of pyrene fluorescence (the ratio 11/13 decreases) and an increase in the fluorescence decay time (manifested by an increase in intensity). In the case of the 190 emission spectrum of pyrene, the ratio of the peak at ~372.5 nm (11) to the peak at ~383 nm (13) is very sensitive to the polarity of the environment.7'8’9 As pointed out earlier, the 11 peak, which arises from the (0,0) transition from the lowest excited electronic state, is a “symmetry forbidden” transition that can be enhanced by the distortion of the n—electron cloud. In contrast, the electronic transition that results in the 13 peak is not forbidden and thus is relatively solvent-insensitive.8 Hence, the ratio (II/I3) can serve as a measure of the polarity of the environment experienced by pyrene. The CMC values estimated using cl‘u‘3‘9 to be in agreement with those obtained from this ratio have been reporte conventional techniques such as surface tension measurements. Mathurzo’21 investigated the system PMAA-g-EG for the first time using the environment-sensitive fluorescence of pyrene to study the formation of the aggregates with hydrophobic interiors. Pyrene not only served as a fluorescent probe but also as the model “oil” phase for these studies. In these studies, Mathur utilized the environment- sensitive fluorescence of solubilized pyrene to study the aggregate formation behavior of block/grafi copolymers of poly(methacrylic acid-g-ethylene glycol) as a function of pH and concentration. In this investigation, changes in the excitation and emission spectra of solubilized pyrene are evaluated to quantify the partitioning of the probe between the hydrophobic domains and the bulk aqueous phase. This method was used to determine the critical aggregate concentration where all pyrene is emulsified. Based on this literature we have used the polarity-sensitive fluorescence of solubilized pyrene to study the aggregate formation behavior of selected block/graft copolymers of poly(methacrylic acid-g-ethylene glycol) as a function of pH and concentration. Steady-state fluorescence measurements of the excitation and emission 191 spectra of solubilized pyrene, as it partitions between the hydrophobic domains and the bulk aqueous phase, were used to characterize the formation of hydrophobic domains in aqueous solution. In particular, we were interested in determining the effect the copolymer molecular weight on the formation of the hydrophobic domains. In addition, the concentration dependence of this fluorescence response was used to determine the critical aggregate concentration for chosen P(MAA-g-EG) polymers. 10.3 EXPERIMENTAL METHODS AND TECHNIQUES Fluorescence spectroscopy was performed on aqueous copolymer solutions at pH values ranging from ~2 to ~7 and concentrations ranging from 0.0 wt.% (no polymer) to 0.5 wt.%. Typically a 100 ml sample with a polymer concentration of 0.5 wt.% was prepared and successive dilutions were made from this stock solution. The solubility limit of pyrene in water is reported at a concentration of 6x10'7 M.7 To prepare a saturated pyrene solution in water, 0.010 g of pyrene was added to 100 ml of HPLC grade water and stirred for 4 days. The solution was then filtered using a medium #2 filter paper (Whatman) to remove any undissolved pyrene. The concentration of this pyrene solution was assumed to be 6x10'7 M, which is the saturation concentration of pyrene in water. For all the studies reported in this chapter, the pyrene concentration was kept constant at a value of 6x10’8 M for each sample. This was to ensure that no excimer formation would occur since the concentration of pyrene was 10% of its saturation concentration. Sample concentrations varied from 0.0 wt.% (no polymer) to 0.5 wt.% and each 10 ml sample contained 6x10'8 M pyrene. The pH of the samples was varied 192 from a value of 2 to 7 by adding a few microliters of either a 1M HCl or IM NaOH solution. The fluorescence spectroscopy was performed using an Aminco-Bowman Series 2 Luminescence Spectrometer. Both the excitation and emission spectra were recorded for every sample. For the excitation spectra, emission was set at 397 nm while the excitation was varied between 290-360 nm. For emission spectra, excitation was set at 322 nm and the emission was collected between 370 and 450 nm. The slit width was kept at 0.2 nm and the spectra were collected at intervals of 0.2 nm at a rate of 1 run/s. The detector voltage was kept constant for the excitation and emission scans over all concentrations. 10.4 RESULTS AND DISCUSSION 10.4.1 Excitation Spectra Representative excitation spectra of solubilized pyrene for a series of different concentrations of copolymer 2.5/0.5 are shown in Figure 10.1. This figure contains a plot of the emission intensity (arbitrary units) at 397 nm as the excitation intensity is varied from 290 nm to 350 nm. The plot consists of four spectra representing aqueous solutions of this 20:1 (methacrylic acid (MAA) to ethylene glycol (EG) molar repeat unit ratio) copolymer with concentrations ranging fiom 0 wt.% to 0.5 wt.%, each at pH 7. The excitation spectra show the typical pattern with three peaks of increasing intensity located at 305.4, 318.6 and 333.8 nm. Note that for all polymer concentrations, the peak location does not change (and the ratio of the intensity at 333.8 and 338 nm does not change) for all polymer concentrations for these neutral conditions. 193 Figure 10.2 contains the excitation spectra for the same copolymer at a solution pH of ~2. This figure illustrates a characteristic red shift of all of the peaks in the excitation spectrum. For example, the maximum of the most intense peak shifts from 333.8 nm (corresponding to pyrene in the bulk aqueous phase) to 338 nm (representing pyrene in a hydrophobic environment in the presence of the copolymers). This red shift is observed for all three investigated polymer concentrations, and has been attributed to the partitioning of pyrene into hydrophobic domains.20 Under these acidic conditions, the ratio of the fluorescence intensity at 333.8 and 338 nm does change slightly with the polymer concentration. Mathurzo'z' assigned this shift as a measure of the relative amount of pyrene that partitions as the "oil phase" in the hydrophobic domains. In order to eliminate the effects of variable light intensity and detector sensitivity from measurement to measurement, a ratio of intensities was used. Specifically the ratio of the intensity at 338 (the wavelength of maximum intensity under acidic conditions) to the intensity at 333.8 nm (the wavelength of maximum intensity under neutral conditions) was used to describe the partitioning of pyrene into hydrophobic domains. Figure 10.3 illustrates the pH dependence of the intensity ratio at 338 nm to that at 333.8 nm (1333/13333). The data presented in this figure is for the copolymer 5/O.5 having a MAA:EG repeating unit ratio of 10:1 for concentrations of 0.004, 0.1 and 0.5 wt.%. There are two important observations that can be made from this figure. First, as the pH is raised from 2 to 7 the excitation ratio 1333/13333 decreases and approaches a value of ~0.4 at pH 7 for all concentrations. Second, this decrease is more pronounced as the concentration of the copolymer is increased. In addition, under acidic conditions the excitation ratio 1333/1333; increases from about 1.6 for a concentration of 0.004 wt.% to 194 about 1.9 when the copolymer concentration is raised to 0.1 wt.% or 0.5 wt.%, respectively. It was also observed that for a solution pH 7 the ratio remained constant at about 0.4, which was the value observed from the excitation spectrum of aqueous pyrene with no copolymer present. For the two higher concentrations, the ratio 1333/1333 of about 1.9 at a pH value of 2 corresponds to the total partitioning of pyrene in the hydrophobic domains, whereas the intensity ratio of about 0.4 corresponds to the situation of pyrene being surrounded by the polar environment of the bulk water. Since this value is already reached at a pH value of 5.75 for the lowest polymer concentration of 0.004 wt.% it can be concluded that all pyrene is released from the hydrophobic domain at that solution pH. At a solution pH of 5.75, a higher fraction of pyrene remains emulsified with polymer concentrations of 0.1 wt.% and higher. In Figure 10.4 illustrates the effect of molecular weight on the pH—dependent pyrene fluorescence. It contains the ratio of the intensity value at 338 nm to the value at 333.8 nm is plotted as a function of pH for two polymers with the same composition (a 10:1 MAA:EG repeat unit ratio), but different molecular weights (the 5/10 copolymer has a molecular weight of 26,000 while the 5/0.5 has a molecular weight 165,000. The figure illustrates that the ratio 1333/13333 assumes a value of about 1.95 at both solution pH 2 and 4 for copolymer 5/0.5. However, for the copolymer 5/10, the intensity ratio has a lower level of about 1.65 at pH 2 and 1.75 at pH 4. Moreover, at a given pH, the intensity ratio is always lower for the 5/ 10 copolymer than for the copolymer 5/0.5. These fluorescence results suggests that the aggregates formed by the higher molecular copolymer may be less polar than those formed by the lower molecular weight copolymers. This interpretation is consistent with the dynamic light scattering results in 195 Chapter 5 in which is was concluded that, for the 10:1 copolymers, the aggregates become more compact as the molecular weight is increased. The higher values of the 1333/13333 fluorescence ratio for the higher molecular weight may arise from the fact that water is more effectively excluded from these aggregates. 10.4.2 Emission Spectra The steady-state pyrene emission spectra for a 20:1 P(MAA-g-EG) copolymer 2.5/0.5 at a pH of 2 is shown in Figure 10.5. The figure contains spectra for copolymer solutions from 0 wt.% to 0.5 wt.% with the same pyrene concentration (6 x 10'8 M). The trend that can be readily observed is the change in the 13725/1333 peak intensity ratio (ratio of the peak near 372.5 nm and 383 nm, respectively) commonly referred to as the I3/I3 ratio. As in the case of the excitation spectrum, this trend may be explained by the partitioning of the pyrene between the aqueous and hydrophobic domains created by the copolymer aggregates under acidic conditions, and by the decreasing polarity in these hydrophobic domains. The increase in the ratio Il/I3 arises from forbidden vibronic band transitions that exhibit marked intensity enhancement under the influence of increasing solvent polarity. In Figure 10.6 the emission spectra of pyrene in water in the presence of 0.5 M of NaOH, HCl and NaCl in the absence of any polymer is shown. These studies were performed to determine the adding electrolytes on the pyrene spectra to ensure that electrolyte effects are not misinterpreted as polymer effects. The concentration of the electrolytes in this experiment is about 150% of the electrolyte concentration used to 196 titrate the aqueous solutions of the P(MAA-g-EG) copolymers from very acidic to neutral conditions. The insert in Figure 10.6, lists the values of the ratio 11/13 for each emission spectrum. As can be seen, the 11/13 value remains approximately constant at a value of ~1.6. This shows that the addition of electrolytes does not influence the fluorescence pattern of pyrene in aqueous solution. Hence, the change in the peak ratio is only due to hydrophobicity changes of the P(MAA-g-EG) polymers sensed by pyrene. Figure 10.7 illustrates the dependence of the 11/13 ratio for two 20:1 MAA:EG copolymers at a concentration of 0.5 wt.% as a function of pH to determine the influence of molecular weight. The figure contains data for solutions of polymer P(MAA-g-EG) 5/ 10 and 5/0.5 as a function of four different pH (2, 4, 5.75 and 7). As a reference, the corresponding values are shown for pyrene in water titrated to the corresponding pH. From Figure 10.7 it can be noted that, under acidic conditions, the emission ratio (II/I3) drops from a value of about 1.6 for polymer-free water to a value of about 0.7 for 10:1 MAA:EG copolymers at a concentration of 0.5 wt.%. For the polymer-free solution, the 11/13 ratio for pyrene was found to be essentially independent of pH. Notably, the peak intensity ratios of the 5/10 copolymer (with a molecular weight of 26,000 g/mole) have a consistently higher value than the corresponding ratios for copolymer 5/0.5 (with a molecular weight of 165,000 g/mole). This results is consistent with the result obtained from the excitation spectra. Values for the emission ratio 11/13 for pyrene in water reported in literature range from 1.58 to 1.9.4'6’8 The Il/I3 values for pyrene in a hydrophobic environment range from 0.55-0.6 in various hydrocarbon solvents,4 0.95 for polystyrene films,6 1.15-1.20 in polystyrene-b-poly(ethylene oxide) solutions,7 and between 1.05-1.12 in polystyrene-b- 197 8 poly(sodium acrylate) solutions. These values are in good agreement with the results reported in this study. 10.4.3 The Critical Aggregation Concentration A series of studies of the emission ratio as a function of polymer concentration was used to investigate the aggregation behavior and determine the critical aggreggation concentration. As discussed previously in chapter 1, the P(MAA-g-EG) copolymers exhibit an alternating hydrophobic/hydrophilic multiblock architecture under acidic conditions. Hence the hydrophobic segments of one or more chains may cluster together to form aggregates in a concentration dependent manner. A critical aggregation concentration can be defined in a manner similar to the critical micelle concentration of traditional surfactants. In this study, the polarity-sensitive fluorescence response of pyrene was used to identify the critical aggregation concentration for P(MAA-g-EG) copolymers with two different MAA to EG repeat unit ratios by measurement the I|/I3 ratio as a function of the copolymer concentration. A plot of the emission ratio (II/I3) as a function of copolymer concentration at a pH of 2 is presented in Figure 10.8. This figure contains data for P(MAA-g-EG) 5/10 and P(MAA-g-EG) 5/0.5 having a MAA:EG repeat unit ratio of 10:1 but with different molecular weights of 26,000 and 165,000, respectively. The data are represented by the symbols whereas the lines drawn through them illustrate trends. The sigmoidal nature of the trends in this figure is similar to plots of properties such as surface tension plotted as a function of surfactant concentration.6‘'7’22’2'‘23 For example, most surfactants exhibit a decrease in surface tension with an increase in the aqueous concentration. However, a 198 leveling-off is observed after a certain concentration and the CMC may be estimated from the intersection of tangents drawn along the two slopes.6"7’2°'2' A similar procedure can be adopted to estimate the critical aggregate concentration (CAC) from this data for the two c0polymers. This may be achieved by drawing tangents following the sigmoidal curve as shown in the figure and determining the CAC as the concentration corresponding to the point where the tangents intersect. From Figure 10.8, the estimated values for the CAC for the copolymers 5/10 and ms are 12 x 10'3 wt.% and 5 x 10'3 wt.% respectively. These results suggest that the higher molecular weight copolymer aggregates at a lower concentration. These values are in good general agreement with the CMCs reported for other di-block and tri-block copolymeric surfactants. For example, the CMCs reported in literature range from 4 x 10'4 wt.% to 6 x 10'4 wt.% for poly(ethylene oxide)-polystyrene tri- and diblock copolymers respectively.‘5 Figure 10.9 shows a plot of the emission ratio (II/I3) as a function of copolymer concentration at a pH of 2 for copolymers having a repeat unit ratio MAA:EG of 20: 1. In this case, the copolymers 2.5/10, 2.5/5 and 2.5/0.5 have number average molecular weights of 17,000, 44,000, and 126,000 g/mole. The copolymers exhibit the familiar sigmoidal trends and the technique used to estimate the CAC from data in Figure 10.8 can be similarly applied to determine the CAC from the data in Figure 10.9. Using this method the CAC for the 20:1 copolymeric emulsifiers with increasing molecular weight was found to be 9 x 10'3 wt.%, 7.3 x 10'3 wt.% and 4.5 x 10'3 wt.%, respectively. Again, these studies have shown that the CAC is reduced as the molecular weight is increased. Comparison of the CAC values for the 20:1 copolymers in this figure to those of the 10:1 199 copolymers in the previous figure reveals that the molecular weight has a larger effect than the composition on the CAC. As a summary, the experimental CMC values for both copolymer compostions are plotted as function of the copolymer molecular weight in Figure 10.10. The figure illustrates that the molecular weight has a marked effect on the CAC, and that the CAC decreases as the molecular weight is increased. In addition, the figure illustrates that, at a given molecular weight, the 10:1 copolymer was found to exhibit a higher CAC than the 20:1 copolymer even though it contains twice as many hydrophobic segments. This latter result is somewhat surprising. However, it should be noted, that this measurement is based solely the environmental polarity of pyrene, which relies the formation of smaller hydrophobic domains than may be required to stabilize oil droplets. The formation of the complex along the copolymer backbone will render the chain less flexible. It is possible that the enhanced flexibility of the 20:1 copolymer relative to the 10:1 copolymer makes it easier to form the small hydrophobic domains detected by the pyrene. 10.5 CONCLUSIONS In this chapter, the polarity-sensitive fluorescence response of pyrene has been used to characterize the aggregate formation behavior of the reversible block/graft copolymer poly(methacrylic acid-g-ethylene glycol) as a function of pH, composition, molecular weight and concentration. The results indicate that there is a marked dependence of the aggregation behavior on the molecular weight of the P(MAA-g-EG) copolymers. For example, the pyrene fluorescence data suggest that, for a given copolymer composition, the aggregates formed by the higher molecular copolymer may 200 be less polar than those formed by the lower molecular weight copolymers. In addition, it was found that, at a given composition, the critical aggregation concentration decreases as the molecular weight is increased. Surprisingly, it was found that the polymer containing the higher content of oligo(ethylene glycol) grafts exhibited the higher CAC at a given molecular weight. It was postulated that this result could arise from the decreases flexibility resulting from the increased content of double-strand complexes. This decreased flexibility could make it difficult to form hydrophobic domains small enough to envelop the relatively small pyrene molecules. 201 J .l 1.8 -l - - Polymer concentration - 1.6 - 0 wt.% - .5. - - - - - 0.004 wt.% - E 1-4" ------- 0.1 wt.% ‘ 3 ‘ ----- 0.5 wt.% 1‘ - 2‘ 1.2 - ' ‘ - Q ' If“ ' :9: 1.0 - I i ‘.‘ - .0 r 1‘ B ‘ I.’ “ ' ‘_' 0.8 -' I! 'n 5 "‘ .3: - '.' r ‘3 - g 0.6 - 2 ,{I -. l _ .2.) - I.“ LI "0 ‘.‘ \ E 0.4 'l I, ’.‘ II", ' ‘\‘ \ -I .I I. ’1’ ‘.‘ -' ,i' ‘x‘ 0.2 -I ;‘-=.:4”’ ~-.,’ ““3. -I 0-0 ' - I r l r T I ' 1 T I 290 300 310 ' 320 330 340 350 wavelength [nm] Figure 10.1 Pyrene excitation spectra for copolymer P(MAA-g-EG) 2.5/0.5 solutions of various concentrations with constant pyrene concentration (6 x 10'8 M). Emission was monitored at 397 nm. The solution pH were titrated to a value of 7. 202 2.0; Polymer concentration a . ,z . 0 wt.% .'I 1‘: 1.8 - ------ 0.004 wt.% 5f: 2 ,: - 1.6- --- -0.1 wt.% I: '1“: i ' - ----- -0£5“d36 ; 3i 3 intensity [arbitrary units] 0.43 I 4 I I ' ‘F 320 330 340 wavelength [nm] 1 I 290 300 310 Figure 10.2 Pyrene excitation spectra for copolymer P(MAA-g-EG) 2.5/0.5 solutions of various concentrations with constant pyrene concentration (6 x 10'8 M). Emission was monitored at 397 nm. The solution pH was titrated to a value of 2. 203 2 O I ' T ' l ' I 1 I ' 1 . -' .g a-I II . * d ..'_. 1.8 - _ H .1 II “at 1.6 1 ‘ A - \§ 1.4 e - ,9. 1.2 - - 9 . . z, 1.0 _ P(MAA-g-EG) 5/o.5 . _ '5 . —l— 0.5 wt.% . g, 0.8 - + 0.1 wt.% - g 0 6 - + 0.004 wt.% - 33 0.4 - A -+ . .1 0.2 . , . , . , . , . r . . . 2 3 4 5 6 7 pH value [-] Figure 10.3 Plot of the pyrene excitation ratio (1333/13333) as a function of pH for 10:1 P(MAA-g-EG) 5/0.5 copolymer solutions with 6 x 10'8 M pyrene. The emission was monitored at 397 nm. The polymer concentrations for each series were 0.004, 0.1 and 0.5 wt.%, respectively. 204 I ' I I l ' F V I ‘ l 2.0 .. I .. \ . T; 1.8 - O - (.5 . // . 8 _ O _ a 1.6 _a ' ' .9 1.4 - . - ‘5 . . E, 1'2 . + was 3 7;) 1 0 _ _O"— 5,10 _ c . .‘L’ ' O ' .E 0.8 A - .35 . . a: 0.6 4 - m 2 . 0.4 4 - l V l ' I ' l l I h 2 3 4 5 6 7 pH value [-] Figure 10.4 Plot of the pyrene excitation ratio (1333/13333) as a function of pH for P(MAA-g-EG) 5/0.5 and 5/ 10 copolymer solutions with 6 x 10'8 M pyrene. The emission was monitored at 397 nm. The polymer concentration for each series was 0.5 wt.%. 205 I r I ' I ' I ' I ' I Polymer conc. 0.6 4 ------ 0 wt.% 7 - - - - 0.004 wt.% 7’;- —--—-— 0.1 wt.% . E — 0.5 wt.% 3 I\\ ’6‘ 2‘ - ’ I “.3 ‘\ _ S 0.4 " ll \ II a.“ t ‘ ' 1 '9 “ ' s I o .E\ lg: '1 I .' ’4‘. . .0 \ g \~l '.‘ x -\ g 0.2 ‘ \ V ~ \ -I '93 . ‘ ‘ I ....... . \ E ‘ '. \ . ‘ - I": ': f ' 0-0 I ' F I I r I I I I i I 360 370 380 390 400 410 420 wavelength [nm] Figure 105 Pyrene emission spectra for P(MAA-g-EG) 2.5/0.5 copolymer solutions at pH 2. Excitation was at 322 nm and pyrene concentration in each sample was 6 x 10'8 M. The polymer concentration in the aqueous solution was varied between 0 and 0.5 wt.%. 206 009 Y I l l V j I l I I T T 1 ' 0.8 - - o.7- 'm‘nm NaOH 1.58 HCI 1 .58 NaCl 1 .56 0.6 - l u 9. 0.5- l I e I 0.4-- ' 0.3 - 0.2 - intensity [arbitrary units] —0.5 M NaOH * ----- 05MHG ~. 0-1- _ ----0.5MNaCI x.-. 0.0 . , . . , . , . , . , . T 350 360 370 380 390 400 410 wavelength [nm] r T .. 420 430 Figure 10.6 Pyrene emission spectra for aqueous solutions of pyrene with added electrolytes. Excitation was at 322 nm and pyrene concentration in each sample was 6 x 10'8 M. The insert in the right upper comer shown the intensity ratio 1333/1372; for all three electrolyte solutions. 207 1.8 I I I I I I I ' l r l m 1.6 - a _______________________________ a; ______________________ - .8 . ----- , 73- 1 4 _ ——a— was 3 _‘° ' —O—- 5/10 .2, ‘ “-3-" water 3 e 1.2 " u—I E . 8 a, 1.0 - J x o. - BM , 0'6 l ' I ' I ' I I I 2 3 ' 4 5 6 7 pH value [-] Figure 10.7 A plot of the pyrene emission ratio 1372,5/1333 (II/I3) as a function of pH for copolymer 5/0.5 and 5/10 solutions with a pyrene concentration of 6 x 10'8 M. Excitation was at 322 nm. The pyrene emission ratio for titrated water is shown as a reference. 208 1.6 J ‘ - 3 M": g 5/10 26000 :0. 1-4 ' 5/0.5 165000 " g . \ .9 1.2 - k 1 E . $7; 10g -—--——-fitof5/1O _ a:J ' - - - - fit of 5/0.5 g I 5/10 0.8 - o 5/0.5 -5, ..... , ..... ”5....-- - ‘ . - 0.6 W 1E-6 1E-5 1E-4 1E-3' 0.01 0.1 1 10 concentration [wt.%] Figure 10.8 A plot of the pyrene emission ratio (excitation at 322 nm) versus concentration for 10:1 MAA:EG P(MAA-g-EG) copolymers 5/10and 5/0.5. 209 1.8 I I l llllll' I VIII?" l I WIIIII I IIIIIII' T TIIIIIfl—fiw .........-.....-...._,,__‘ M: x. n ‘ 1.6 Q-,\ 2.5/10 17000 n . I \ 2.5/5 44000 ‘ _s 1 4 _ ‘ 2.5/0.5 126000 _ \m ' g- . .l o 1.2 - , _ “g in of 2.5/10 h - - - - -fit of 2.5/5 ‘ E ------ fit of 2.5/0.5 _ m 1.0 - c: I 2.5/10 515’ - o 2.5/5 ‘ — A 2 5/0 5 0.8 - ° ' - 0.6’ -1—memI—mm 1 E-6 1E-5 1E-4 1E-3 0.01 0.1 concentration [wt.%] Figure 10.9 A plot of the pyrene emission ratio (excitation at 322 nm) versus concentration for 20:1 MAA:EG P(MAA-g-EG) copolymers 2.5/ 10, 2.5/5 and 2.5/0.5. 210 .5 0) I :3 12- o I 20:1 MAA:EG n 0 10:1 MAA:EG '52 11‘ Linear fit of 20:1 MAA:EG .5 10‘ *-' 1 g 9- I C Q) 'l 8 8- 8 ‘ . Q) 7- g . e 6‘ 8 . < 5‘ .5 4- ': l l ' l ' I ' I r r ' 0 0.0 40.0k 80.0k 120.0k 160.0k Molecular Weight Mn [g/mole] Figure 10.10 Plot of critical aggregate concentration determined by the ratioing technique for the 10:1 and 20:1 MAA:EG copolymers as a function of their molecular weight. 211 10. 11. 12. 13. 14. 15. 16. 10.6 REFERENCES Ham J. S., J. Chem. Phys., 21, 756 (1953). Gillespie, G. D., in Structure Property Relations in Polymers: Spectroscopy and Performance, M. W. Urban and C. D. Craver Eds., ACS Advances in Chemistry Series, Washington DC, 236, pp 89 (1993). Nakajima A., J. Mol. Spec., 6], 467 (1976). Kalyanasundaram K. and Thomas J. K., J. Am. Chem. Soc., 99, 2039 (1977). Bakeev K. N ., Ponomarenko E. A., Shishkanova T. V., Tirrell D. A., Zezin A. B., and Kabanov V. A., Macromolecules, 28, 2886 (1995). Zhao C.-L. and Winnik M. A., Langmuir, 6, 514 (1990). Wilhelm M., Zhao C.-L., Wang Y., Xu K., and Winnik M. A., Macromolecules, 24, 1033 (1991). Astafieva I., Zhong X. F., and Eisenberg A., Macromolecules, 26, 7339 (1993). Astafieva I., Khougaz K., and Eisenberg A., Macromolecules, 28, 7127 (1995). Nivaggioli T., Alexandridis P., and Hatton T. A., Langmuir, 11, 730 (1995). Quina F., Abuin E. and Lissi B., Macromolecules, 23, 5173 (1990). Turro N. J ., Gratzel M., and Braun A. M., Angew. Chem. Int. Ed. Engl., 19, 675 (1980). Winnik F. M., in Interactions of Surfactants with Polymers and Proteins, E. D. Goddard and K. P. Anantpadmanabhan Eds., CRC Press, Ann Arbor, pp 367 (1992). Winnik M. A. and Winnik F. M., in Structure Property Relations in Polymers: Spectroscopy and Performance, M. W. Urban and C. D. Craver Eds., ACS Advances in Chemistry Series, Washington DC, 236, pp 485 (1993). Almgren M., Grieser F., and Thomas J. K., J. Am. Chem. Soc., 101(2), 279 (1979). Yeung A. S. and Frank C. W., Polymer, 31, 2101-2111 (1990). 212 17. 18. 19. 20. 21. 22. 23. Alexandridis P., Athanassiou V., Fukuda S., and Hatton T. A., Langmuir, 10, 2604 (1994). Alexandridis P., Holzwarth J. F., and Hatton T. A., Macromolecules, 2 7, 2414 ( 1994). Nivaggioli T., Tsao B., Alexandridis P., and Hatton T. A., Langmuir, 11, 119 (1995). A.M. Mathur, B. Drescher, A.B. Scranton, and J. Klier, Spectroscopy, in press. Mathur, A. M., Synthesis and Characterization of Polymeric Pseudocrown Ethers and Reversible Block/Graft Copolymeric Emulsifiers Based Upon Polymer Complexation, Dissertation, Michigan State University, East Lansing, MI, 1996. Rosen M. J ., Surfactants and Interfacial Phenomenon, Second Ed., John Wiley & Sons, New York, 1989. Clint J. H., Surfactant Aggregation, Chapman Hall, New York, 1992. 213 Chapter 11 CONCLUSIONS This research project has demonstrated the considerable potential of reversible hydrophobic polymer complexes in graft copolymers for the design of reversible emulsifiers. During the course of this project, it has been demonstrated that the copolymer system poly(methacrylic acid-g-ethylene glycol) (P(MAA-g-EG)) consisting of an poly(methacrylic acid) backbone with pendent poly(ethylene) glycol chains may be used as reversible emulsifiers which allow emulsions to be formed, broken, and reformed repeatably. This research has provided a fundamental characterization of how the solution and emulsification pr0perties of the copolymers depend upon the polymer composition and molecular weight. The study has relied upon a host of different experimental techniques to build the fundamental physical picture of how the reversible emulsifiers work. These techniques include: NMR spectroscopy, gel permeation chromatography, dynamic light scattering, potentiometric titration, drop volume tensiometry, 2D NOESY NMR spectroscopy, laser scanning confocal microscopy, and polarity-sensitive fluorescence spectroscopy. Before discussing the detailed conclusion from each of these studies, it is useful to provide a brief overview of the path taken in this research project. First, the relationship between the polymerization reaction conditions and the polymer structure was identified, and a semibatch polymerization reactor was designed to synthesize poly(methacrylic acid-g-ethylene glycol) copolymers with varying molecular 214 weight and level of grafiing. Using the well-characterized copolymers prepared with this reactor, the solution characteristics and important emulsification properties (i.e. stability, droplet size distribution, pH-controlled reversibility) were investigated and the effects of the polymer composition and molecular weight on these emulsification properties were established. Interfacial tension measurements and pyrene fluorescence were used to relate the underlying polymer structure to the emulsification phenomena, and to elucidate the mechanism by which the copolymers stabilize an emulsion. In addition, NOESY NMR was used to study spatial arrangements of the complementary polymer constituents under complex-forming and complex-breaking conditions. Finally, this research has accomplished its broadest objective and has provided specific guidelines for the design and synthesis of reversible emulsifiers. A semi-batch reactor was successfully designed to synthesize a series of poly(methacrylic acid-g-ethylene glycol) of varying molecular architectures by a free radical copolymerization of methacrylic acid with the macromonomer methoxy poly(ethylene glycol) methacrylate. Gel permeation chromatography was used to characterize the polymer molecular weight distribution. The GPC data show that for a given MAA:EG repeat unit ratio, the number average molecular weight is proportional to the inverse square root of the initiator concentration. For repeat unit ratios ranging from infinity (pure PMAA) to 4.6 (a graft copolymer with a graft occurring about every 100 backbone repeat unit), the polydispersity was found to remain essentially constant. The experimental polydispersity values suggest that chain transfer to polymer does not occur to a significant extent. 1H NMR was used to determine the composition of the P(MAA-g- EG) copolymers. 1H NMR results illustrated that the experimental values for the MAA to 215 EG repeat unit ratio in the copolymer are in good agreement with the known values for in the MAA:EG ratio in the monomer feed. The GPC and NMR results were combined to estimate the number of grafts per chain for each of the polymer samples. As expected, for a constant initiator concentration, the average number of grafts per chain was found to increase as the amount of the MPEGMA in the reaction mixture was increased. In addition, for a constant MPEGMA content, the average number of grafts per chain decreased as the initiator concentration was increased. This indicates that the poly(ethylene glycol) grafis are evenly distributed along the PMAA backbone The visual appearance and dissolution ability for P(MAA-g-EG) is dependent on the content of ethylene glycol incorporated into the copolymers. It is concluded from the observed dissolution during titration of aqueous P(MAA-g-EG) copolymer solutions that the solubility of the copolymers decreases with increasing ethylene glycol content. The homopolymer poly(methacrylic acid) and copolymers bearing only 2.5 wt.% of EG visually dissolved in water regardless of the degree of polymerization and solution pH. However, copolymers having an ethylene glycol content of 5 wt.% exhibit the Tyndall effect at strong acidic conditions proving the presence of small-dispersed polymer agglomerates that lead to the scattering of some light. If the mass percentage of ethylene glycol exceeds 10 wt.% of total polymer mass, un-dissolved solids appear. This is attributed to the fact that the ratio of the hydrophobic segments to the un-complexed PMAA backbone has exceeded a value that hinders the complete polymer from dissolution in water. The poor dissolution behavior of P(MAA-g-EG) copolymers that bear a MAA:EG repeat unit ratio of 5:1 (10 wt.% EG of total polymer mass) excludes them for further considerations as reversible emulsifiers. The aggregate sizes of P(MAA- 216 g-EG) copolymers with an MAA:EG repeat unit ratio of 20:1 and 10:1 at a solution pH of 2 is investigated using dynamic light scattering. The most important design conclusion from the solubility studies is that P(MAA- g-EG) copolymers with a PEG content of more than 10 wt.% are unsuitable for emulsifiers because they too hydrophobic under acidic conditions, and precipitate out of solution. However, the pH-dependent solution properties copolymers with an EG content of 2.5 and 5 wt.% make them very appropriate for reversible emulsifiers. These polymers have significant hydrophobic character (which provides a driving force to move them to the oil/water interface), and they remain as suspended colloidal agglomerates under acid conditions. The dynamic light scattering experiments revealed that the polymers with an EG grafting level of 2.5 wt.% exhibited increasing aggregate size with increasing molecular weight. This was attributed to the fact that aggregates of the longer chains occupy more space (most of the copolymers contain only a few hydrophobic blocks under acidic conditions. The aggregates of P(MAA-g-EG) copolymers with a EG grafting level of 5 wt.% exhibited a smaller hydrodynamic radius than those containing 2.5 wt.% EG, and exhibited a decreasing hydrodynamic radius with increasing molecular weight. The first result may be explained by the fact that the increased grafiing levels imparts a higher hydrophobicity (on average twice as many hydrophobic blocks per chain) into the P(MAA-g—EG) copolymers, and therefore may lead to a more compact coil in aqueous solution. Similarly, within the series of copolymers containing 5 wt.% EG, the longer chains contain on average more hydrophobic blocks per chain, and therefore exhibit more 217 compact coiling. The potentiometric titrations suggest that the presence of the complex decreases the acidity of the carboxylic acid moieties. Emulsification studies performed using a variation of the visual bottle test method revealed that the P(MAA-g-EG) copolymers having a grafting level of 2.5 and 5 wt.% can form, break and reform emulsion by adding adjusting the pH from very acidic (pH 2) to neutral conditions (pH 7) and back to acidic conditions. In addition, it was demonstrated that 0.5 wt.% aqueous solution of P(MAA-g-EG) copolymers having a MAA:EG repeat unit ratio of about 10:1 and 20:1 in aqueous solution could stabilize emulsions containing a 2:1 ratio of oil in water (by volume). It was demonstrated that the essential design parameters, the molecular weight and the level of grafiing, influence the breakability of an emulsion formed. An increase in the number of grafts per chain led to a need for an adjustment to higher solution pH values in order to release the oil from an emulsion originally stabilized at a pH of 2. An increase in molecular weight of copolymers with comparable levels of grafiing apparently leads to a minor decline of the amount of oil released as the solution pH is raised to a value higher than pH 4. Dynamic Drop volume interfacial tension studies illustrated that the copolymers P(MAA-g-EG) have the ability to moderately reduce the interfacial tension in the system methyl laurate/water to a maximum reduction of about 25% compared to the value in the absence of any copolymer. Similarly, also the non-grafted poly(methacrylic acid) homopolymer, which is not effective to prevent coalescence over time, reveals a modest surface tension reduction like the graft copolymers. These important results lead to the conclusion that the potential of the copolymers to stabilize emulsions arises primarily from their ability to prevent droplet coalescence after they are formed, and not by 218 reducing the interfacial free energy of the system. This investigation lead to the validation of the hypothesis that the P(MAA-g-EG) copolymers behave as stabilizers of emulsions, rather than as surfactants. Further, we studied the effects of the variables pH, level of grafiing, molecular weight, and polymer concentration on the surface tension reduction observed for the P(MAA-g-EG) copolymers in the methyl laurate/water system. Of the first three variables, the pH was observed to have the largest effect on the observed interfacial tension, with the trend of decreasing interfacial tension with decreasing pH for a constant polymer concentration. Apparently, there exists an “transition” pH value equal to the pKA of the acid (~5.8) since the interfacial tension at pH values of 5.75 and 7.0 are both close to the value for the system with no copolymers (~20-22 mN/m), while the interfacial tension for pH values of 2 and 4 are significantly lower (~14-16 mN/m). Unlike to the pH, the level of grafting and the molecular weight of the P(MAA-g-EG) copolymers has relatively little effect on the interfacial tension. Interfacial tension reduction with respect to the concentration of P(MAA-g-EG) copolymers showed that, for all polymers, the interfacial tension reached a minimum plateau at a concentration of ~0.3 wt.%. An important conclusion for the laser confocal microscopy studies is that the technique of optical sectioning using confocal fluorescence spectroscopy can be successfully employed to determine droplet size distributions of methyl laurate/water emulsions stabilized with P(MAA-g-EG) reversible emulsifiers. The technique is most applicable to systems that have a droplet size on order of ten microns. If the droplet size is below about 2 microns, the distance between successive optical sections is too large to reliably assign the diameters. If the droplet size is above about 50 microns, it is difficult 219 to measure the droplets since the image quality deteriorates for focal planes more than 25 microns into the sample. The LSCM technique was used to characterize the droplet size distribution of methyl laurate droplets dispersed in the continuous aqueous phase in the presence of four P(MAA-g-EG) emulsifiers of different composition and molecular weight (the P(MAA- g-EG) 2.5/10, 2.5/l, 5/5 and 5/1 were tested). The average droplet size was found to be essentially independent of both the copolymer composition and molecular weight, and the number average droplet diameters were found to be in the range spanning from 8 to 12 microns. In addition, the average droplet size was found to be essentially independent of time, with the possible exception of the 20:1 copolymer of high molecular weight, which shows a slight increase in the average droplet diameter. The 2—D NOESY NMR experiments clearly demonstrated that moving the system from acidic to basic conditions leads to a transformation in the spacing of the ethylene glycol and backbone protons from close proximity under acidic conditions to further than 5 Angstroms under basic conditions. This effect can be attributed to the disruption of the complex and the extension of the oligo(ethylene glycol) grafts away from the backbone. The NOE results also indicate that, in the complex, the protons in the ethylene glycol repeat unit are spatially closer to the (It-methyl than they are to the methylene. Molecular modeling results suggest that this effect could arise fi'om constraints placed on the PMAA backbone in order to accommodate the formation of a complex with a one-to-one repeat unit ratio. The polarity-sensitive fluorescence results indicate that there is a marked dependence of the aggregation behavior on the molecular weight of the P(MAA-g-EG) 220 copolymers. For example for a given copolymer composition, the aggregates formed by the higher molecular copolymer may appear to be less polar than those formed by the lower molecular weight copolymers. In addition, it was found that, at a given composition, the critical aggregation concentration decreases as the molecular weight is increased. Surprisingly, it was found that the polymer containing the higher content of oligo(ethylene glycol) grafts exhibited the higher CAC at a given molecular weight. It was postulated that this result could arise from the decreases flexibility resulting from the increased content of double-strand complexes. This decreased flexibility could make it difficult to form hydrophobic domains small enough to envelop the relatively small pyrene molecules. 221 llllliljllilllljlllllljil 30