00001.tif

PDF

Download

Open document

Flip pages

Contact Us

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back o f the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600 SYNTHESIS A N D C H A R A C T ER IZ A T IO N O F M ACROCY CLIC A N D C A TEN A TED POLYSTYRENE-POLY(2-VINYLPYRIDINE) BLOCK CO PO LY M ERS by Y aodong G an A D issertation P resented to the FACULTY O F THE G RA D U A TE SC H O O L UNIVERSITY O F SO U TH ER N C A L IFO R N IA In P artial F ulfillm ent of the R equirem ents for the D egree D O C TO R O F PH ILO SO PH Y (C hem istry) A ugust 1995 C o p y rig h t 1995 Y aodong G an UMI Number: 9617099 UMI Microform 9617099 Copyright 1996, by UMI Company. Ail rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. UMI 300 North Zeeb Road Ann Arbor, MI 48103 UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90007 This dissertation, written by YAODONG GAN under the direction of h.Is . Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean o f G raduate Studies D a te ^1x14,^1995. DISSERTATION COMMITTEE /../isteiCzx Z..jrr. Chairperson i , . L , * .... D edicate to m y p aren ts, sister, w ife an d son. i i Acknowledgments I w o u ld like to sincerely th an k Dr. Thieo E. H ogen-E sch for h is consistent encouragem ent, m otivation, gu id an ce a n d frien d sh ip th ro u g h o u t this w ork. I w ish to th an k Dr. Surya P rakash for his collaboration an d advice for the accom plishm ent of this dissertation. I w o u ld also like to acknow ledge the follow ing people for directing a n d h elp in g m e in the v ario u s projects of m y dissertation: Dr. A u th u r W . A dam son, John J. A klonis, Dr. R obert Bau, Dr. L arry R. D alton, D r. M. G ersh en zo n , Dr. G eorge A. O lah, Dr. R onald Salovey, Dr. L aw rence A. Singer, Dr. W illiam P. W eber, Dr. Joachim Z oller, Dr. Sylvia Law , Dr. Gia Kim , Dr. Q im o Liao, Dr. F u w en Shen, Ms. R osalie R ichards, M r. T im La, M s. Joyce L aq u in d an u m , M r. B. Y eom , M r. Jim M eritt, a n d the m em bers of Prof. H ogen-E sch's g ro u p p a st an d p resen t, especially M r. D ahai D ong a n d M r. Jam es A. Lee for their friendship and technique support. Finally, I w o u ld like express m y deep appreciation to all m y fam ily m em ber in clu d in g m y g ran d m o th er, sister, uncles a n d aunts for th eir su p p o rt an d u n d e rstan d in g , specially m y p a re n ts a n d w ife for th eir love, patience a n d dedication w ith o u t w hich this w o rk w o u ld not have been successfully accom plished. TABLE OF CONTENTS D ED IC A TIO N ....................................................................................................... ii A CK NOW LEDG M ENTS..................................................................................iii LIST OF TA BLES............................................................................................ ix LIST OF SCHEM ES.......................................................................................... x LIST OF FIG URES...........................................................................................xi LIST OF ABBREVIATIONS......................................................................xiv A BSTRA CT......................................................................................................... xv CHAPTERS 1. FLUORESCENCE STUDIES OF LINEAR AND MACROCYCLIC POLYSTYRENE ............................................. 1 A bstract................................................................................................1 Introduction......................................................................................... 2 Experim ental...................................................................................... 4 M aterials........................................................................................... 4 Fluorescence measurements........................................................4 UV absorption measurements.................................................... 7 R esults..................................................................................................7 Emission o f PS macrocycles ..................................................... 8 Iellm o f macro cyclic P S............................................................10 Absorptivities o f cyclic P S.......................................................10 D iscussion..........................................................................................12 Effects o f concentration............................................................ 12 i v Excimer emission intensities o f PS macrocycles...................13 Excimer emission wavelength shifts o f macrocyclic P S 16 Dependence ofle/Im o f linear and macrocyclic PS on DP..18 Dependence o f absorption intensity o f cyclic PS on DP 19 Acknowledgm ents ......................................................................... 21 R eferences.........................................................................................21 2. SYNTHESIS AND CHARACTERIZATION OF MACROCYCLIC POLYSTYRENE-b-2-VINYLPYRIDINE BLOCKCOPOLYM ERS.................................................................. 41 A bstract.............................................................................................. 41 Introduction...................................................................................... 42 E xperim ental....................................................................................45 High Vacuum Techniques......................................................... 45 Synthesis o f living P2VP-b-PS-b-P2VP.................................45 Fluorescence M easurements ....................................................48 Results and Discussion................................................................. 48 Synthesis and Isolation o f macrocyclic PS-b-P2VP.............. 48 Characterization o f PS-b-P2VP macrocycles........................52 Thermal analysis by using DSC and TGA............................ 54 Morphology o f PS-b-P2VP macrocycles............................... 56 Excimer Emission Intensities Differences between Linear and M acrocyclic PS-b-P2VP copolym ers.................57 A cknow ledgem ent.......................................................................... 59 R eferences.........................................................................................59 V 3. EFFECTS OF LITHIUM BROMIDE ON THE GLASS TRANSITION TEMPERATURES OF LINEAR- AND MACROCYCLIC-POLY(2-VINYLPYRIDINE) AND - POLYSTYRENE ..............................................................................66 A bstract...............................................................................................66 Introduction.......................................................................................67 Experimental Section .................................................................. 68 Synthesis.........................................................................................68 Sample P reparation....................................................................69 M easurem ents................................................................................70 SEC A nalysis.................................................................................70 Results and Discussion..................................................................71 Tg o f linear, unpurified cyclic and purified cyclic P2VP as a function o f molecular weight............................... 71 Effects o f LiBr content on the Tg o f isobaric linear and macrocyclic P2VP...................................................73 LiBr effects on Tg o f P2VP with different DP, and the width o f Tg transition ................................................74 Comparison ofTg behaviors between purified P2VP and PS macrocycles ..................................................................76 Acknowledgm ents ......................................................................... 77 R eferences.........................................................................................78 v i 4. SYNTHESIS AND CHARACTERIZATION OF A CATENATED POLYSTYRENE-POLY(2-VINYLPYRIDINE) — A NOVEL TYPE OF BLOCK COPOLYM ER.................... 87 A bstract...............................................................................................87 Introduction.......................................................................................87 Experim ental ..................................................................................93 M aterials.........................................................................................93 Synthesis.........................................................................................94 Isolation..........................................................................................94 C haracterization...........................................................................96 Results and Discussion..................................................................96 Synthesis o f catenated PS-b-P2VP and reaction conditions........................................................................................96 Isolation o f PS-h-P2VP catenane ...........................................98 Characterization o f PS-h-P2VP catenane............................ 100 A cknowledgm ents ....................................................................... 103 R eferences.......................................................................................103 5. ANIONIC POLYMERIZATIONS OF TRIMETHYLVINYLSILANE AND PHENYLDIMETHYLVINYLSILANE TOWARDS THE SYNTHESIS OF WELL-DEFINED POLYFLUORODIMETHYLVINYLSILANE AND POLY VINY LA LCO HOL..............................................................113 A bstract............................................................................................ 113 v i i Introduction.................................................................................. 113 E xperim ental..................................................................................118 M aterials.....................................................................................118 P olym erizations........................................................................ 119 Protodesilylation Reactions...................................................119 Oxidation R eactions................................................................120 Results and Discussion............................................................. 121 P olym erization.......................................................................... 121 Protodesilylation...................................................................... 123 Oxidation reactions.................................................................124 A cknow ledgem ents.....................................................................125 R eferences..................................................................................... 125 v i i i LIST OF TABLES Chapter-Table 1-1 Molecular weight (MW) and molecular weight distributions (MWD) of isobaric linear and macrocyclic PS................................ 26 1-2 Absorption intensities (A) and molar absorptivity coefficients (e) of isobaric linear (L) and macrocyclic (C) PS (DP=22) in cyclohexane solutions............................................................................ 27 2-1 SEC, NMR, DSC and TGA characterization: apparent molecular weights, molecular weight distributions, glass transition temperatures and decomposition temperatures of isobaric linear (L) and macrocyclic (C) polystyrene-poly(2- vinylpyridine) block copolymers (PS-b-P2VP)..............................62 3-1 Apparent molecular weights (MW) and glass transition temperatures (Tg) of the unpurified and purified cyclic (C) P2VP and linear (L) P2VP of identical degree of polym erization 80 3-2 Apparent molecular weights (MW) and glass transition temperatures (Tg) of linear polystyrene (PS) and cyclic PS with identical degree of polymerization synthesized by D. Dong and reprinted with his permission.....................................81 4-1 Characterization results of isobaric linear (L) and macrocyclic (C) polystyrene (PS), Poly(2-vinylpyridine) (P2VP), PS-b-P2VP and PS-cat-P2VP............................................107 5-1 Anionic polymerization of phenyldimethylvinylsilane (PDMVS) in toluene initiated with t-BuLi..................................... 129 5-2 Anionic synthesis of polytrimethylvinylsilane (PTMVS) and polystyrene-b-PTMVS (PS-b-PTMVS) copolymers in toluene by using t-BuLi as initiator............................................ 130 LIST OF SCHEMES Chapter-Scheme 1-1 Diagram showing the emitted energy of less stable configuration hv" is greater than that of the stable configuration hv'.................. 17 1-2 The structures of o,o'-bridged cyclic diphenyldiacetylenes (In), p,p'-bridged cyclic diphenyldiacetylenes (IIn) and p,p'-bridged cyclic tolans (Illn )................................................................................. 20 1-3 The structure of paracyclophanes........................................................21 3-1 Synthesis of linear and macrocyclic P2VP's and purification of cyclic P2V P........................................................................................72 4-1 Cyclization of P2VP2-, 2Li+ with EX2 in the presence of m acrocyclic polystyrene......................................................................108 4-2 Isolation of catenated PS-b-P2VP copolymer..................................109 5-1 Unexplored reactions of PPDM VS................................................... 114 5-2 Synthesis of PPDMVS, PFDMVS and PVA....................................116 5-3 Synthesis of a alcohol via oxidative cleavage of the C-Si bond reported by Tam ao....................................................................118 LIST OF FIGURES Chapter-Figure 1-1 Emission spectra of linear and macrocyclic polystyrene (Mp=2200) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings. The scale of the fluorescence intensity of linear PS in the inset has been enlarged five times compared with that of isobaric macrocycli PS. ^,ex=253 nm ............................................................................................ 29 1-2 Emission spectra of linear and macrocyclic polystyrene (Mp=3400) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings. ^,ex=253 nm .......................30 1-3 Emission spectra of linear and macrocyclic polystyrene (Mp=43000) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings. Xex=253 nm .......................31 1-4 Excimer relative emission intensities of linear and cyclic PS in cyclohexane (0.001 M) as a function of molecular weight.............32 1-5 Excimer emission peak wavelengths of linear and cyclic PS in cyclohexane (0.001 M) as a function of molecular weight.............33 1-6 The ratio of excimer to monomer emission intensities (Iex/Imo) of linear and cyclic PS in cyclohexane (0.001 M) as a function of molecular weight......................................................34 1-7 Absorption spectra of linear and macrocyclic polystyrene (Mp=43000) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings...............................................35 x i 1-8 Absorption spectra of linear and macrocyclic polystyrene (Mp=3400) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings................................................36 1-9 Absorption spectra of linear and macrocyclic polystyrene (Mp=2200) at 298 °K in cyclohexane solution at concentration of 0.001 M in units of benzene rings. The inset shows the enlarged spectra from 220 nm to 300 nm .........................................37 1-10 Absorbance at 259 nm versus concentration for cyclic and linear PS (DP=22)..................................................................................38 1-11 Molecular model of macrocyclic polystyrene 14-mer synthesized by using lithium naphthalide as initiator and C H 2B r2 as coupling agent.................................................................. 39 1-12 Molecular model of macrocyclic polystyrene 14-mer synthesized by using lithium naphthalide as initiator and CH2Br2 as coupling agent showing electron clouds........................39 2-1 Size exclusion chromatograms of linear PS precursor (3), linear PS-b-P2VP precursor (2) and crude macrocyclic PS-b-P2VP (1) (DP=66)...................................................................... 63 2-2 Thermogravimetric curves of linear P2VP-PS-P2VP block copolymer and purified PS-b-P2VP macrocycle (DP=66)...............64 2-3 Bridging and looping chain arrangements in block copolym er m icrodom ains......................................................................65 3-1 Tg of linear (O), unpurified cyclic (A) and purified cyclic P2VP (©) as a function of 1/Mp: (O, A) precipitated in hexanes only; (©) precipitated in water followed by x i i disolution in THF and reprecipitation in hexanes........................... 82 3-2 Effects of LiBr on the Tg of matched linear (O) and macrocyclic (0 ) P2VP of Mp=4200 and linear PS (A) of M p=4000.................. 83 3-3 Effects of LiBr on the Tg of linear P2VP (O) Mp=4200; ( 0 ) M p=6600........................................................................................... 84 3-4 DSC thermograph of cyclic P2VP )DP=42) containing (a) 0.75 wt% LiBr; and (b) 15 wt% LiBr....................................... 85 3-5 Tg's of linear (A) and cyclic (0 ) polystyrene as a function of M W ....................................................................................................... 86 4-1 Topologically interesting m acrocycles...............................................88 4-2 Size exclusion chromatogram of crude catenated PS-b-P2VP copolym er........................................................................ 110 4-3 Size exclusion chromatogram of purified catenated PS-b-P2VP copolym er........................................................................ I l l 4-4 NMR spectrum of catenated PS-b-P2VP copolymer..................... 112 5-1 Proton NMR spectra of (a) PPDMVS, (b) product after protodesilylation at 40 °C and (c) product after protodesilylation at 60 °C (PFDM VS)............................................131 5-2 Fluorine-19 NMR spectrum of PFDMVS (by using CFCI3 as the internal reference, 5=0.00)..................................................... 132 x i i i LIST OF ABBREVIATIONS BuLi.................... butyllithium DBX..................... bis (bromomethyl) benzene, or dibromo-xylene DD^-.................... l,l-dilithio-l,l,4,4-tetraphenyl-butane PS......................... polystyrene P2VP.................... poly (2-vinylpyridine) PTMVS............... poly (trimethylvinylsilane) PPDMVS poly (phenyldimethylvinylsilane) PFDMVS poly (fluorodimethylvinylsilane) THF..................... tetrahydrofuran M, MW.............. molecular weight M p...................... peak molecular weight Mn...................... number average molecular weight M w..................... weight average molecular weight MWD................. molecular weight distribution (M w/M n) DP....................... degree of polymerization SEC..................... Size Exclusion Chromatography DSC.................... Differential Scanning Calorimetry TGA................... Thermogravimetric Analysis Tg....................... glass transition temperature Td....................... thermal decomposition temperature <G>.................... ratio of SEC peak molecular weight for two polymers having the same DP, but different geometries x i v ABSTRACT Macrocyclic polystyrene-b-poly(2-vinylpyridine) (PS-b-P2VP) copolymers have been synthesized by sequential anionic polymerization of styrene and 2-vinylpyridine followed by coupling using 1,4- bis(bromomethyl)benzene (1,4-DBX). The products were characterized by SEC, NMR. DSC, TGA and spectrofluorimetry. It has been found for the first time that the thermal decomposition temperatures of low degree of polymerization (DP) macrocyclic polymers are lower than those of linear polymers of same DP and composition. A 50/50 PS-b-P2VP (Mp=104) "catenane" is prepared by end to end coupling of a P2VP dianion lithium salt with 1,4-DBX in THF in the presence of a PS macrocycle (Mp=4500). The catenane was isolated by precipitation- extraction procedures that were optimized using a 50/50 PS-b-P2VP macrocycle as a model for the solubility characteristics of the catenane. The catenane copolymer was characterized by SEC, NMR and emission (fluorescence) spectroscopy. For PS macrocycles, the emission intensities increase and the excimer emission maxima are blue shifted with decreasing DP and the ratios of excimer and monomer emission Ie /IM increase with lower DP, whereas those of the isobaric linear PS are essentially constants. Increases in absorptivity of the PS macrocycles with decreasing DP are also observed for lower DP cycles whereas these differences gradually disappear for the higher MW macrocycles (DP>200). XV With decreasing DP, the glass transition temperatures (Tg's) of P2VP, PS and PS-b-P2VP macrocycles slightly decrease, however the Tg differences between macrocyclic and linear of these polymers strongly increase. The Tg's of both linear and cyclic P2VP increase linearly with added LiBr. The slopes of the Tg vs LiBr concentration plots are higher for macrocyclic P2VP than for the isobaric linear chains, and increase with decreasing DP. The anionic polymerizations of trimethylvinylsilane (TMVS) in toluene at -20 °C and phenyldimethyl-vinylsilane (PDMVS) at 0-20 °C proceed in high yields (>95%) giving narrow molecular weight distribution (MWD<1.15) polymers. The reaction of PPDMVS with HBF4.Et20 in toluene at 60 °C resulted in the formation of polyfluorodimethylvinylsilane (PFDMVS) as shown by Iff and 19p NMR. x v i CHAPTER 1 Fluorescence Studies of Linear and Macrocyclic Polystyrene Abstract Although the excimer emission maxima (X=333 nm) and intensities of narrow molecular weight distribution (MWD) atactic polystyrene (PS) are independent of degree of polymerization (DP), this is not the case for PS macrocycles. The emission intensities increase and the emission maxima are blue shifted with decreasing DP and the ratios of excimer and monomer emission Ie/Im also increase with lower DP. For example, the excimer emission intensity is increased about five fold and the emission is blue shifted by 21 nm for a narrow MWD PS macrocycle of a DP of 22 and the Ie/Im ratio is increased about five fold. Increases in absorptivity of the PS macrocycles with decreasing DP are also observed for the lower DP cycles whereas these differences gradually disappear for the higher MW macrocycles (DP>200). These observation are consistent with conformations of the lower DP PS macrocycles in which the majority of phenyl rings is present along the periphery l of the roughly planar macrocycle and oriented in parallel fashion perpendicular to the plane of the ring. Introduction A comparison of linear and macrocyclic polymers is of intrinsic interest. Because of topological restraints and the absence of end groups macrocyclic polymers are expected to show differences in chain dynamics that may be expressed in unusual rheological thermomechanical or mechanical properties. Compared with their linear isobaric precursors, the "matching" macrocycles are expected to possess smaller hydrodynamic size and higher overall segment densities and this also may be reflected in properties such as adsorption. Anionic polymerization followed by end to end coupling is a useful strategy toward the synthesis of macrocyclic vinyl polymers First, dianion precursors having narrow molecular weight distributions (MWD)'s are generally obtained. Furthermore, macrocycles may be synthesized over a large range of molecular weights (1,500-70,000). Also, the dianion precursors separately may be both protonated and coupled so that the linear and macrocyclic polymers have identical molecular weights (MW's) and MWD1s. Finally the formation of macrocyclic block copolymers and matching linear block copolymers is possible. Macrocyclic block 2 copolymers may display interesting morphological and other properties compared with the isobaric linear copolymers of the same composition 7. Differences in properties of matching linear and macrocyclic polymers in addition may be of help in evaluating the content of linear chains in macrocycles of samples of unknown purity. Thus most synthetic pathways leading to the formation of macrocycles in principle may result in linear impurities and this may be revealed in their properties 8-12 _ Fluorescence spectroscopy of polystyrene (PS) has been a topic of study for several decades 13-15^ ancj reports on the fluorescence of polystyrene block copolymers have also appeared recently 16-18_ However, the fluorescence behavior of macrocyclic PS and PS-containing block copolymers has never been reported. In the following we report interesting and unprecedented differences in fluorescence of matching isobaric linear and macrocyclic polystyrene and the corresponding matching polystyrene-b-poly(2-vinylpyridine) copolymers (PS-b-P2VP) over a molecular weight (MW) range from 2200 to 43000. 3 Experimental Materials: The synthesis of linear and macrocyclic polystyrene (PS) was carried out in THF at -78 °C using lithium naphthalene as initiator and is reported elsewhere 8-10# Some experimental details are described in Chapter 2 of this thesis and refs.11-12. Some of the characterization data of these polymers are given in Table 1. The matching linear and cyclic PS samples (10.0 mg) were purified by dissolution in 0.5 ml of THF and precipitation in 10 ml of hexane. This treatment was repeated several times in order to completely remove hexane soluble contaminants such as naphthalene or coupling reagents. The samples were dried in a vacuum oven at 60 °C for two days. The purification yields were about 80%. Fluorescence Measurements: Fluorescence spectra were obtained using a 90° geometry Perkin-Elmer LS-5 fluorescence spectrophotometer with a Perkin-Elmer 561 recorder, and checked on a Perkin-Elmer 650-10S, which consists of a xenon lamp, two double-grating monochromators and an photomultiplier tube. Spectrophotometric grade cyclohexane was purified by vacuum line distillation from Na-K alloy and checked for fluorescence prior to use. Interfering emission of the solvent was not observed and 4 solvent impurities were not detectable by gas chromatography. The purified PS white powders of MW's ranging from 2,200 to 43,000 were dissolved (0.100 gram/liter) in cyclohexane by stirring at 45 °C (water bath). All solutions were prepared right before the fluorescence measurements. However, they could be stored in well-capped flasks under nitrogen atmosphere in the dark for several weeks, without spectral deterioration (identical emission wavelengths and intensities). The quartz cell (1x1x4 cm) was flushed with nitrogen and capped prior to the fluorescence measurements (at least 3/4 volume of the cell was filled with solution). The emission was scanned at 25 °C from 260.0 nm to 460.0 nm at an excitation wavelength of 253 nm. Reproducibility was found to be within 1.0 nm. The reproducibility of the emission spectra was checked by using two instruments (a Perkin-Elmer 650-10S and a Perkin-Elmer LS-5 fluorescence spectrophotometers) and this always gave the same results. We calibrated both fluorescence spectrometers using a mercury lamp, a standard ovalene block (#2) (purchased from Perkin-Elmer) and a 1,2:5,6-dibenzanthracene solution in heptane. A narrow (3 nm) emission slit was chosen for optimum resolution. As a result, a sharp decrease in emission intensity is observed compared to the case where the emission slit was 5 nm. However, for cyclic and linear PS-P2VP block copolymer 5 samples, which exhibited much low emission intensities than PS homopolymers, the emission slit was set at 5 nm to obtain a clearer emission signal. Low concentrations (< 0.100 gr/1) were used in all cases. At higher concentrations, the shape of the emission band is distorted because of self-absorption of emitted light 19. Previous studies ^0 carried out above a PS concentration of 1.00 gr/1 have shown that the emission band is shifted to longer wavelengths with increasing concentration. Also, reproducible fluorescence spectra of very dilute (0.10-0.30 gr/1) polystyrene cyclohexane solutions have been reported 13,18,20_ jn these cases, a polystyrene concentration of 0.100 g/1 in cyclohexane was sufficiently small that self-absorption of emission is negligible. At extremely low concentrations of PS (0.010 g/1), the PS solutions gave an unacceptably low signal to noise ratio. In our experiments, fluorescence intensities of both monomer and excimer peaks at concentrations between 2 x 1 0 - 3 m and 8x10“^ M (chromophore units) were shown to increase linearly with concentration, and the emission maxima were highly reproducible with no self-absorbance. Therefore, PS concentrations of 0.10 g/1 in cyclohexane solutions were chosen for most measurements. 6 UV absorption measurements: Absorption spectra of linear and cyclic PS in cyclohexane solutions (0.100 gr/1) were obtained using a Milton Roy Spectronic 3000 Array with a Epson Lx-810 printer. All solutions (MW from 2,200 to 43,000) were prepared in the same way as for the fluorescence measurements. Concentrations of lxl0-3 M, 5 x 10- 4 M, 2 . 5 x 10~4 M and 1.25x10“^ M (styrene units) of linear and cyclic PS (MW=2200) were used to check compliance with Beer's law. Spectrophotometric grade cyclohexane was used as the reference. All of the measurements were repeated at least twice and some of the samples were run five times by replacing the solutions in the cell. Absorbances were found to be within 3%. Good consistency with Beer's law was found for both linear and cyclic PS of DP=22 (figure 10). The absorption bands for both linear or cyclic PS in the 240-280 nm range were always independent of MW and concentration. The molar absorption coefficient of 2.2K cyclic PS (6=520) was observed to be 2.5 times that of the absorption intensity of the isobaric linear PS (6=203). Results Narrow distribution (Mw/Mn<1.30) linear and macrocyclic PS in the range of DP= 22-430 were synthesized in THF at -70 7 °C by end to end coupling using CH2Br2 or 1,4- bis(bromomethylbenzene) as reported elsewhere 8-10. por ^he cyclic PS synthesized using 1,4-bis(bromomethylbenzene) (DBX) as coupling agent, an additional phenyl chromophore is present in the polymer backbone although this chromophore has the same emission as a PS pendent phenyl group 21. The use of dibromomethane (CH2Br2 ) as coupling reagent results in a structurally more regular vinyl macrocycle without introducing any additional chromophores (diagram 1). Low MW (DP < 50) PS macrocycles obtained with this coupling agent were used for these studies since the number of chromophore units of the macrocycle is the same as that of linear polymer (Table 1). The high molecular weight samples (such as 22K and 43K) were made by using DBX as coupling agent, because the effects of the structure of the coupling reagent should be negligible above a DP of 200. Emission of PS macrocycles Two emission peaks are observed for 0.100 g/1 solutions of all linear PS samples, one having a maximum at about 280nm and a second having a broad structureless maximum at about 333nm. The 280 nm peak is due to "monomer" fluorescence arising from a single phenyl unit. The 333 nm emission is due to emission from an excited-state complex (excimer fluorescence) 13,18,20. The excimer emission of low MW (DP = 22) macrocyclic PS shows a large enhancement compared with that of the matching linear precursor (Figures 1 and 2). This enhancement is especially large (-5.5 times) for the lowest molecular weight macrocycle (DP = 22). This enhanced excimer emission of the macrocycles decreases as DP increases and above a DP of about 300, the emission intensities of macrocyclic and linear PS are essentially identical. (Figure 3). In contrast, excimer emission of linear PS is essentially MW independent decreasing slightly below a DP of 100 but is constant above it. This was also observed by Ishii and coworkers who concluded that phenyl rings in low MW linear PS moved relatively independently, while those in the high MW polymers formed clusters - * - 3 . The increase of excimer emission intensities of low DP PS macrocycles is accompanied by an excimer emission blue shift (Figures 1 and 2). The emission spectra of linear and macrocyclic PS (DP=22) scaled to similar emission intensities illustrate the differences in emission wavelength (Figure 1, inset) . Figure 5 shows the wavelength of excimer emission maxima of linear and cyclic PS in cyclohexane (0.001 g/lOOml) as a function of molecular weight. The excimer emission at 333-334 nm of linear PS is independent of DP between 22 and 430. However, for macrocyclic PS of lower DP, the excimer emission is significantly blue shifted. At 9 high MW (DP=430) the emissions of linear and macrocyclic PS are identical ( X . max = 334 nm) but the emission wavelength maxima of macrocyclic PS dramatically decrease with decreasing DP reaching a blue shift of 21 nm at a DP of 22. Ie/Im of Macrocyclic PS The MW dependence of the ratios of excimer to monomer emissions (Ie/Im) for linear and macrocyclic PS in cyclohexane is given in figure 6. A trend similar to that in figure 4 is observed. As shown in previous studies 13-15 on linear PS, below a MW of 20,000, Ie/Im slightly decreases with decreasing MW. Above a MW of 20,000, Ie/Im is essentially MW independent 1®. However, for the case of cyclic PS, there is an increase in Ie/Im from 5 to about 26 with decreasing MW as the DP decreases from 430 to 22. This will be discussed below. Absorptivities of cyclic PS The near UV absorption spectra of linear and macrocyclic PS in cyclohexane (0.100 g/1) showed three absorptions at 253.8, 259.0, 268.4 nm. The positions of these maxima did not change with MW for either linear or macrocyclic PS in the 2,200 to 43,000 MW range. For high MW linear and macrocyclic PS (43K), the spectral shape and intensities of the three absorption bands were identical 10 (Figure 7). Although the absorption spectra of linear PS are essentially independent of MW, the absorption intensities of macrocyclic PS increased with decreasing MW (Figure 8). For very low MW (2.2K) cyclic PS, the absorption intensities of the three bands increased dramatically, being about 2.5 times higher than those of the isobaric linear PS. Furthermore, the absorbance of linear PS at 232 nm is nearly zero for DP of 22, but that of the isobaric macrocyclic PS approximately equals that of the 259 nm maximum of the linear PS (Figure 9). The absorption intensities (at 259 nm) of both cyclic and linear PS (2.2k) increased linearly with polymer concentration (Figure 10), showing compliance with Beer's law, and confirming the increased absorptivities of macrocyclic PS compared with linear PS. The absorbances of all bands (at 253.8, 259.0 and 268.4 nm) of 2.2k cyclic and linear PS solutions exhibited the same linear relationships. Molar absorptivity coefficients (£) at 253.8, 259.0, 268.4 nm of linear and macrocyclic PS (DP=22) in cyclohexane are shown in Table 2. 1 1 Discussion Effects of concentration In agreement with previous studies 20f excimer emission band shapes of PS were found to be concentration independent below 2 x 1 0 - 3 m (styrene units) 18,25. The same is the case for the relative emission intensities of excimer and monomer 33. Emission intensities were also observed to increase linearly with PS concentration in the 8xl0 ~3 m - 2xl0 ~3 m (styrene units) range. These data are consistent with that of Hirayama and indicate that below 2xl0 _ 3 m excimer formation is intramolecular only 22 _ Data of Lindsell et al. on PS-Polybutadiene block copolymers show that PS excimer emission is independent of block length thus indicating the absence of long range intramolecular excimer formation 16,21. This is in agreement with model studies and indicates that in our case also excimer formation is relatively short range. Above of PS concentration of 1 0 _ 2 m in styrene units, excimer emission increases relative to monomer emission in agreement with Hirayama's findings indicating enhanced intermolecular excimer formation 2 1,2 2 # In solution at low polymer concentration only intramolecular excimers exist while at higher concentration or in films intermolecular excimers would also be present 18,25,32^ thus leading to a higher Ie/Im. Such a 1 2 concentration dependence of Ie/Im has been reported for linear PS in several solvents including cyclohexane ^3 for concentrations up to 30 weight percent. Torkelson and Lee observed a greater concentration dependence of Ie/Im in PS block copolymers than in PS homopolymers at concentrations (e.g. 0.10-1.00 weight %) comparable to the copolymer CMC Therefore, above the CMC the higher local concentration of PS units in the block copolymer micelle would result in a higher Ie/Im than for PS solutions. A similar rationale might be used as one of the reasons for the higher Ie/Im values of the PS rings. As the DP of the PS rings decreases, Ie/Im is increased, since macrocycles possess a smaller hydrodynamic size and hence a higher local concentration of phenyl rings compared with the isobaric linear chains. However this effect is not large as the effective overall concentration of phenyl groups in macrocycles is only about 20-30 percent larger. Excimer Emission Intensities of PS Macrocycles Of considerable interest is the much greater emission intensity of the macrocyclic PS at low DP (< 34) compared to the linear polymers (Figures 1,2 and 4). This result was completely unexpected and is apparently unprecedented. First, it should be noted that the absorptivity of the PS macrocycles is increased. This enhancement being about 2.5 13 fold for the macrocycle of DP of 22. This should be reflected in a corresponding enhancement in emission. However, for this macrocycle the emission enhancement is greater (x5) and the emission is blue shifted (21 nm). It is clear that this can not be due simply to the greater local concentration of phenyl rings in PS macrocycles than in isobaric linear PS due to the smaller (20-30%) hydrodynamic size of the cycles. Since the low MW PS cycles (2.2K and 3.4K) do not contain any additional chromophoric group due to the coupling unit, the detailed conformation of the low DP macrocycles is apparently responsible for the enhanced excimer emission and blue shifts. The mechanism of PS excimer formation was studied by Nakahira et al. 23^ Longworth 24a^ Klopffer 24b^ David et al. 25a, b^ and Ishii et al. 13,26_ excimer formation by direct intramolecular interaction appears to involve parallel rings with a separation distance of about 3.73 A or less 26_ However, the actual distance between phenyl groups is strongly affected by the steric effects nearby PS main chain. Ishii et al. studied excimer emissions in 1,2- dichloroethane (DCE) solutions of isotactic and atactic PS, poly(o-methylstyrene) (POMS), poly(m-methylstyrene), and poly(p-methylstyrene) 13,26_ They found that the calculated rate constants for excimer formation (ka) in the isotactic polymers except for POMS were almost the same and about 14 three times larger than those in the corresponding atactic polymers. They ascribed this to more numerous "trapping sites" for excimer formation in the relatively stiff isotactic PS helices. Apparently excitation migration occurred intramolecularly 23 _ Exciton transfer in atactic PS occurs along the chain without being "trapped". On the other hand, the emission rate constant ka of isotactic POMS was almost identical to that of the atactic POMS, which was very small compared with values for the other methyl- substituted PS. Ishii concluded this was due to the closest distances between the POMS parallel aromatic rings being larger than the limit for exciton diffusion as a result of steric crowding by the methyl group as shown by X-ray analysis 27,28_ Thus it would appear that in isotactic PS helices, excimer formation is enhanced by the proximity of the phenyl rings stacked in roughly parallel fashion around the helix. A similar interpretation could apply for the case of macrocyclic PS. The serious conformational restraints in this case may cause crowding of the phenyl groups relative to linear PS especially for very low MW cyclic polymers where ring strain is pronounced. This would increase the fraction of excimer-forming sites, thus resulting in an increase in excimer emission band and a decrease in monomer emission. It should also be noted that since each 15 macrocyclic PS sample has a molecular weight distribution ranging from 1.10 to 1.24, the lower MW cycles in each sample should play a correspondingly larger rule in the emission enhancement (see below). Excimer Emission Wavelength Shifts of Macrocyclic PS The preferred chromophore spatial alignment for excimer formation is generally assumed to be of a parallel sandwich type, the interchromophore distance being between 3.0 and 3.7 A 22,24,29_ However, strict adherence to this is not necessary 30a,b_ Hirayama measured the fluorescence spectra of a variety of diphenyl and triphenyl alkanes in cyclohexane and discussed the formation of the corresponding excimers in relation to molecular conformation 32 _ Intramolecular excimer formation in Ph(CH2 )nph occurred only when n=3, and this is consistent with either of two possibilities: (A) The excimer is of a parallel sandwich type with a certain degree of overlapping of the two 7 t molecular orbitals; or (B) Without satisfying the criterion requirements in (A), Excimer formation is possible with a mutual distance somewhat greater than 3.2 A, or with only partial overlap of MO1s without an exact face to face parallel configuration. Upon absorption of radiation, the excited molecule still possesses its So geometry and solvation as shown 1 6 in scheme 1. For linear PS very shortly after excitation, the geometry of the excited state complex reaches an optimum excited dimer state Si, characterized by a plane parallel relative orientation of the two phenyl rings. Due to the ring strain in the low DP macrocycle, the parallel geometry of the excimer state is not achievable (see below). This less stable "deformed" excimer is expected to have a higher energy and thus to emit at a lower wavelength. (S-| with S c geometry) (S i with distorted > ----------- excim er geom etry) % (S-| with parallel Energy excim er geom etry) absorbed (h v ) Energy Energy emitted hv' emitted hv" (linear PS) (cyclic PS) ' * .’ i ( S 0 with S t geometry) ( S 0 with S c 3 geometry) S c h e m e 1: Diagram showing the emitted energy of less stable configuration hv" is greater than that of the stable configuration hv'. The strain present particularly in the low MW macrocycle could account for the blue shifts in the emission maximum. For higher MW macrocycles, ring strain is less pronounced and both blue shifts and emission intensities 17 decrease. High DP (e.g. DP=430) macrocyclic PS and linear PS of any DP give essentially identical absorption and emission spectra. The rather pronounced asymmetry of the excimer emission of the DP=22 macrocycle is consistent with this since the overall emission signal is the sum of a distribution of macrocycles whereas the smaller cycles contributing disproportionately to the emission. Furthermore these cycles also give the more pronounced emission blue shifts. Molecular modeling of low DP PS macrocycle (DP=14) tends to support the above interpretations 31. The structure of an atactic cyclic PS 14-mer shows an elliptical donut-shaped molecule with most phenyls are oriented in parallel fashion outward from the periphery of the ring and perpendicular to the plane of the ring (Figures 11 and 12). As shown in figure 11 the relative orientations of most of the neighboring phenyls are not quite parallel and with an interchromophore distance of less than 3.7 A. Thus stacking is imperfect so that most excimers have only partial orbital overlap, especially in the bending sections of the molecule containing the CH2 coupling unit. Dependence of Ie/Im of Linear and Macrocyclic PS on DP The sharply increased absorptivity (2.5 times) of the low DP (DP=22) cycle would enhance both excimer and monomer 1 8 emission by roughly this factor (see above). However this is not the case since the excimer emission is enhanced about 5 times and the monomer emission is about the same. This indicates that the higher value of Ie/Im of the cycle is due to enhanced excimer formation at the expense of monomer emission, and this is consistent with the enhanced fraction of excimer-forming sites in the macrocycles as discussed above. Dependence of absorption intensity of cyclic PS on DP The increase in integrated absorption intensity due to the parallel orientation of chromophores is defined as "hyperchromism" 34-36^ ancj attributed to the interaction between the transition dipoles 35 _ Toda and his coworkers reported extensive studies on cyclic acetylenes. They observed hyperchromism (12-37% enhancements in absorption intensities) in the longest-wavelength absorption maxima of o,o'-bridged cyclic diphenyldiacetylenes (In)/ p,p'-bridged cyclic diphenyldiacetylenes (Iln) and p.p'-bridged cyclic tolans (IIIn) as compared with the absorption intensities of the respectively open-chain analogs (Scheme 2). This was reasonably ascribed to the enhanced "coplanarity" of the two phenyl groups as a result of ring formation. Maximum extinction coefficients (E-values) were attained in I5, II1 5, III13 consistent with molecular models indicating rigid and 19 planar structures due to the presence of bridging chains of appropriate lengths 36 # ,0(CH2)nC C=C-C=C (In) (Hn) (IIIn) Schem e 2: The structures of o,o'-bridged cyclic diphenyldiacetylene (In), p,p'-bridged cyclic diphenyldiacetylenes (IJ) and p.p'-bridged cyclic tolans (IIIn). The detailed reason(s) for the pronounced hyperchromism of low DP (DP=22) macrocyclic PS compared with the isobaric linear PS is still under investigation. However, the phenyl rings in this small ring are most likely "stacked" along the periphery of the ring (see above), and it is this stacking that is most likely responsible for the observed hyperchromicity 35 _ Cram, Allinger, and Steinberg 37 reported that paracyclophanes (Scheme 3), show a red shift in the long- wavelength absorption spectrum for m<3 and n<4 . They interpreted this finding as being due to a ground state transannular resonance effect between the two phenyl rings that are held face to face. Later, Hirayama 32 showed that was not the case for the absorption spectrum of 1,3- diphenylpropane, thus indicating the absence of a ground- 2 0 state association of the two phenyl groups. Since in this case the two phenyl groups were located rather "far" from each other, thus intramolecular excimer interactions occurred during the lifetime of the monomer excited state, resulting in enhancement of excimer emission. The emission intensity in this case depended primarily on collision- induced conformational changes. S ch em e 3: T h e stru ctu re o f p a ra cy c lo p h an e s Acknowledgments This work was supported by NSF-DMR grant 9101558, Polymers Program. We wish to thank Drs. A. Adamson and L. Singer and M. Vala for helpful discussions, and Dr. A. Adamson for the use of the fluorescence equipment. The technical assistance of Mr. Tim La and Ms. Rosalie Richards is also much appreciated. References 1. Rempp, P.; Hild, G. and Kohler, A., Eur. Polym. J., 1980, 16, 525. 2 1 2 . 3. 4 . 5. 6 . 7. 8. 9. 10 . Rempp, P.; Hild, G. and Strazielle, C., Bur. Polym. J., 1983, 19 (8), 721. Rempp, P.; Lutz, P. and Strazielle, C., Polymer Preprints., 1986, 27 (1), 190. Roovers, J.; Macromolecules, 1985, 18, 1359 . Hocker, H. and Geiser, D., Macromolecules, 1980, 13(3), 653. Deffieux, A., private communication. Lescanec, R. L.; Hajduk, D. A.; Kim, G. Y.; Gan, Y.; Yin, R.; Gruner, S. M.; Hogen-Esch, T. E. and Thomas, E. L., Macromolecules, 1995, 28, 3485-89. (a) Toreki, W.; Hogen-Esch, T.E. and Butler, G. B., Polym. Preprts., 1987, 28 (2), 343. (b) Toreki, W. and Hogen-Esch, T.E., Polym. Preprts., 1988, 29(2), 416. (c) Toreki, W. and Hogen-Esch, T.E., Polym. Preprts., 1989, 30(1), 129. (d) Sundararajan, J. and Hogen-Esch, T. E., Polym. Preprts., 1992, 33(1), 162. Hogen-Esch, T.E.; Sundararajan, J. and Toreki, W., Makromol. Chem., Macromol. Symp., 1991, 47, 23-42. (a) Sundararajan, J., Doctoral Dissertation, University of Southern California, 1991. (b) Dong, D., Doctoral Dissertation, University of Southern California, 1995. 2 2 11. Gan, Y.; Zoller, J. and Hogen-Esch, T. E., Polym. Preprts., 1993, 34(1), 69. 12. Gan, Y. Zoller,; J.; Yin, R. and Hogen-Esch, T. E., Makromol. Chem., Macromol. Symp., 1994, 77, 93-104. 13. (a) Ishii, T.; Handa, T. and Matsunaga, S., Makromol. Chem., 1976, 177, 283. (b) Ishii, T. Handa, T. and Matsunaga, S., Makromol. Chem., 1977, 178, 2351. (c) Ishii, T.; Handa, T. and Matsunaga, S., Macromolecules, 1978, 11(1), 40-6. 14. Vala, M. T.; Haebig, J.; and Rice, S. A., J. Chem. Phys., 1965, 43, 8 8 6 . 15. Torkelson, J. M.; Lipsky, S.; and Tirrell, M., Macromolecules, 1981, 14, 1601. 16. Lindsell, W. E.; Robertson, F. C.; and Soutar, I., Eur. Polym. J., 1981, 17(3), 203-8. 17. Phillips, D.; Roberts, A. J.; Rumbles, G.; and Soutar, I., Macromolecules, 1983, 16, 1597. 18. Torkelson, J. M. and Lee, S. E.. Polym. Prepr., 1985, 26(2), 271-2. 19. Harris, D. C., "Quantitative Chemical Analysis", Third Edition, W. H. Freeman and Company, NY, 1991. 20. Nishihara, T. and Kaneko, M., Die Makromolekulare Chemie, 1969, 124, 84-90. 21. (a) Atik, S. S.; Nam, M. and Singer, L. A., Chem. Phys. 23 Letts., 1979, 67(1), 75-80. (b) Singer, L. A., personal communication. (c) Adamson, A., personal communication. 22. (a) Hirayama, F., "Energy Transfer and Quenching in Plastic Scintillators", thesis, University of Michigan, 1963. (b) Hirayama, F., J. Chem. Phys., 1965, 42(9), 3163. 23. Nakahira, T.; et al. Makromol. Chem., Rapid Commun., 1980, 1(7), 413-18. 24. (a) Longworth, J. W., Biopolymers, 1966, 4, 1131. (b) Klopffer, W., "Organic Molecular Photophysics", Vol.l, Birks, J.B., Ed., Wiley, New York, N.Y.,1973, Chapt. 7. 25. (a) David, C.; Piens, M.; and Geuskens, G., Eur. Polym. J., 1973, 9, 533. (b) David, C.; Lavoreille, N. P.; and Geuskens, G., Eur. Polym. J., 1974, 10, 617. 26. Ishii, T.; Handa, T.; and Matsunaga, S., J. Polym. Sci., Polym. Phys. Ed., 1979, 17, 811-823. 27. Corradini, Nuovo Cimento suppl., 1960, 15, 96. 28. Odajima, A.; Sauer, J. A.; and Woodward, A. E., J. Polym. Sci., 1962, 57, 107. 29. Chandross, E. A. and Dempster, C. J., J. Am. Chem. Soc., 1970, 92, 3586. 30. (a) Wang, U. and Morawetz, H., Makromol. Chem., Suppl., 24 1975, 1, 283. (b) Johnson, G. E.; J. Chem. Phys., 1975, 62, 4697. 31. Casonava, collaborated research. 32. Polacki, Z., Polym. Photochem., 1986, 7, 325-336. 33. (a) Torkelson, J. M.; Lipsky, S.; Tirrell, M.; and Tirrell, D. A., Macromolecules, 1983, 16, 326. (b) Major, M. D. and Torkelson, J. M., Macromolecules, 1986, 19(11), 2801-6. 34. Pandolfe, W. D. and Bird, G. R.. Photogr. Sci. Eng., 1974, 18(3), 340-6. 35. Tinoco, I., J. Am. Chem. Soc., 1960, 82, 4785-90. 36. Toda, F.; et al. Bull. Chem. Soc. Jap., 1971, 44(7), 1914-16. 37. Cram, D. J.; Allinger, N. L.; and Steinberg, H., J. Am. Chem. Soc., 1954, 76, 6132. 25 TABLE 1: Molecular weight (MW) and molecular weight distributions (MWD) of isobaric linear and macrocyclic PS. Sample Mp(L)a D (L)b D (C)c < G > d PS 2 , 2 0 0 1 .17 1 . 2 0 0.79 PS 3, 400 1 .16 1 .17 0.75 PS 8 , 500 1.06 1 . 1 1 0.73 PS 2 2 , 0 0 0 1 . 04 1 . 1 0 0.77 PS 43,000 1 .2 0 1.24 0 .77 a. SEC peak molecular weight (MW) for linear precursor, b. D(L) = Mw/Mn of linear precursor, c. D(C) = Mw/Mn of fractionated cyclic polymer, d. Ratio of Mp(C) to Mp(L). 26 TABLE 2: Absorption intensities (A) and molar absorptivity coefficients (£) of isobaric linear (L) and macrocyclic (C) PS (DP=22) in cyclohexane solutions. Sample Cone, in Ph unit [Ml Absorption intensities 253.4nm 259.Onm (A) at 2 6 8 .Onm PS (L) 0 . 0 0 0.003 0.004 0 .004 PS (L) 1 .25xl0~4 0 . 021 0 . 026 0 .018 PS (L) 2 . 50xl0~4 0.047 0.056 0 . 042 PS (L) 5 . OOxlO- 4 0.082 0 . 1 0 1 0 .074 PS (L) 10.OxlO- 4 0.158 0.195 0 .143 e (L) 167 203 147 PS (C) 1 ,25xl0-4 0.065 0.073 0 .054 PS (C) 2 .50xl0~4 0 .115 0.129 0 .093 PS (C) 5.OOxlO- 4 0.226 0 .254 0 .184 PS (C) 10 . OxlO- 4 0.440 0.4 90 0.352 e (C) 465 520 377 27 Ph Ph Ph Ph Diagram 1: Polystyrene 20-mer synthesized by cyclization using CH2Br2 and B r C H 2 -^~^-cH 2 B r in THF at -70 °C at 10'5-10'6 M. 28 1 0 0 a o - 3 h a < 60 - cn C 9J 4) e 41 O CO 41 ha 40 - 4 1 420 340 380 300 260 Em ission W avelength (nm) Figure 1: E m ission Spectra o f Linear (— ) and M acrocyclic (----------- ) Polystyrene (M p=2200) at 298 o k in C yclohexane Solution at Concentration o f 0.001 M in Units o f B enzene R ings. T he Scale o f the Fluorescence Intensity o f Linear PS in the Inset Has B een Enlarged Five Tim es Com pared with That o f Isobaric M acrocyclic PS. X e \= 2 5 3 nm. 29 1 0 0 <« u cn >% w 2 a o - < 60 - s 4{ c 40 o J3 2 at > 20 42C 300 380 Emission W avelength (nm) Figure 2: E m ission Spectra o f Linear (— ) an d M acrocyclic (-------) P olystyrene (M p=3400) at 298 °K in C ycloh exan e S olution at C oncentration of 0.001 M in U n its o f B enzene Rings. Xex=253 nm . 30 1 0 0 80 - >* Ul <Q u < >» 6 0 - V) C at it o e Ot U « «> I* 40 - 260 300 340 380 Emission Wavelength (nm) Figure 3: Em ission Spectra of Linear (— ) and M acrocyclic (-------) Polystyrene (Mp=43000) at 298 °K in C yd oh exan e Solution at Concentration of 0.001 M in Units of B enzene Rings. Xex=253 nm. 31 >• m c c c e □ Cyclic PS • Linear PS E iu > o c S 10- E o x Ul 10000 20000 30000 40000 0 50000 Total MW Figure 4: Excimer relative emission intensities of linear and cyclic PS in cyclohexane (0.001 M) a9 a function of molecular weight 32 340 o> 3 30- □ Cyclic PS A Unear PS E 320- 310 10000 20000 30000 0 40000 50000 Total MW Figure 5: Excimer emission peak wavelengths of linear and cyclic PS in cyclohexane (0.001 M) as a function of molecular weight 33 30 20 - o E □ Cyclic PS # Linear PS M O 10 - 20000 10000 30000 40000 0 50000 Total MW Figure 6: The ratio of excimer to monomer emission intensities (lex/lmo) of cyclic and linear PS in cyclohexane (0.001 M) as a function of MW 34 8.4- A b s o p b a 0 . 2 - n c e 253.0 0.0 2* 22 8 230 2 7 0 290 3G6 Figure 7: Absorption Spectra o f Linear (— ) and M acrocyclic (---------) Polystyrene (M p=43000) at 298 OK in Cyclohexane Solution at Concentration o f 0.001 M in Units o f Benzene Rings. A,ex=253 nm. 35 0.1- A b s o r b a 0.2- n c e 259.0 295.8 0.8 260 250 280 210 279 290 3 G 0 239 228 Nanoaeters Figure 8: Absorption Spectra of Linear (— -) and Macrocyclic (-------- ) Polystyrene (Mp=3400) at 298 ok in Cyclohexane Solution at Concentration o f 0.001 M in Units of Benzene Rings. >.ex=253 nm. 36 A b s 0 a a . r b a a a n c Cyclic PS e Linear PS 220 230 240 250 250 Manometers Figure 9: Absorption Spectra of Linear (— ) and Macrocyclic (-------- ) Polystyrene (Mp=2200) at 298 °K in Cyclohexane Solution at Concentration of 0.001 M in Units of Benzene Rings. The Inset Shows the Enlarged Spectra from 220nm to 300 nm. A.ex=253 nm. 37 Absorbance (259 nm) 0.6 0 .4 - 0.2- 0.0 2 6 4 8 0 10 1 2 a Cyclic PS (2.2K) ♦ Linear PS (2.2K) Polym er C o n cen tratio n (mg/100 ml) Figure 10: Absorbance at 259nm versus concentration for cyclic and linear PS (DP=22). 38 Figure 11: Molecular model o f macrocyclic polystyrene 14- synthesized by using lithium naphthalide as initiator and CH2Br2 as coupling agent. Figure 12: M olecular model o f m acrocyclic polystyrene 14-mer synthesized by using lithium naphthalide as initiator and CH2Br2 as coupling agent showing electron clouds. Chapter 2 SYNTHESIS AND CHARACTERIZATION OF MACROCYCLIC POLYSTYRENE-b-2-VINYLPYRIDINE BLOCKCOPOLYMERS Abstract Macrocyclic polystyrene-b-2-vinylpyridine (PS-b-P2VP) block copolymers have been synthesized by sequential anionic polymerization of styrene and 2-vinylpyridine (2-VP) followed by coupling using 1,4-bis (bromomethyl) benzene (1,4-DBX). A small portion of the living ABA precursors were protonated to serve as isobaric linear precursors. The products were characterized by SEC, NMR. DSC, TGA and spectrofluorimetry. The SEC data showed that the macrocycles were comparable in purity to the homopolymer macrocycles prepared previously. As was the case for the macrocyclic homopolymers, the Tg differences between macrocyclic and linear PS-b-P2VP strongly increase with decreasing degree of polymerization (DP). It has been found for the first time that the thermal decomposition temperatures (Td) of low DP macrocyclic polymers are lower than those of linear polymers of same DP and composition. The emission (fluorescence) of macrocyclic PS-b-P2VP copolymers in THF was compared with those of their linear 41 precursors of the same molecular weight (MW) and MW distribution. The excimer emission intensity of the cyclic PS-b-P2VP was always about twice that of the matching isobaric linear block copolymer, and the emission intensity of both copolymers was much lower than that of the mixed solutions of PS and P2VP of the same concentrations and DP's as that of the blocks, indicating significant intramolecular energy transfer of the styrene excimers to the P2VP blocks. Introduction Chemists' fascination with molecular architectures has continued to expand as the sophistication of synthetic chemistry has grown. Such effects have allowed the development and testing of theoretical concepts and provided new and largely unforeseen areas of applications. The shapes of molecules are a reflection of their structures. A related property is topology, which is concerned with the connectivity of atoms, independent of geometry, atomic sizes, rigidity, etc. and consequent shape. Topological isomerism in polymers is associated with molecules having the same structure but different topologies, such as linear, branched or star, and macrocyclic polymers. 42 Macrocyclic polymers having no ends, represent one extreme of the topological spectrum. These topologically different macromolecules are expected to differ in their properties and thus represent an interesting class of compounds. We recently reported the synthesis of macrocyclic poly (2-vinylpyridine) (P2VP) and polystyrene (PS) by end to end coupling of the dianion precursors in highly dilute (1 0“ 5_io~6 m) thf solutions using 1 ,4-bis(bromomethyl) benzene (DBX) 1 r ^ . The macrocycles were of relative narrow distribution (D < 1.30) and of a wide MW range (2,000 < Mn < 60,000). A small portion of the dianion precursor solution was segregated and reacted with methanol affording a matched linear polymer of the same DP and structure. These matched linear and cyclic polymers, showing significant differences in hydrodynamic volume as demonstrated by size exclusion chromatography (SEC) and viscosity measurements, which had also reported by several other groups 1~^ . The SEC data in particular indicated a high (>95%) purity of the macrocyclic P2VP prepared in this way. This conclusion was further supported by dramatic differences in glass-transit ion temperatures (Tg) of macrocyclic- and linear P2VP. Thus in contrast to the linear P2VP the Tg values of macrocyclic P2VP remain constant over a large MW range (Mn > 4,000) 12,14 accord with calculations 43 carried out by DiMarzio and Guttman for macrocyclic polystyrene using a lattice model These interesting results have prompted us to try to synthesize PS-b-P2VP macrocyclic block copolymers. Such polymers consisting of incompatible blocks are topologically homogeneous but also have blocky character. This novel combination of structural features may be expected to result in unusual properties in a number of areas. First they are unprecedented and are of great interest as probes in investigating the role of polymer topology in block copolymer microdomain formation. Thus a comparison between cyclic and linear block copolymers of identical chain length and composition could test current models predicting microdomain formation in block copolymers. An added attraction of PS-b-2VP macrocyclic polymers is that microdomain formation readily occurs and is convenient to study by election microscopy. Furthermore, the Tg values of PS and P2VP are virtually identical facilitating analysis by differential scanning calorimetry (DSC), especially of the smaller rings. Finally, macrocyclic block copolymers are of interest in that the radius of gyration of either one of the blocks in solution may be determined by the use of isorefractive solvents and compared with that of the same block present in the isobaric linear block copolymer. 44 Experimental Hiah Vacuum Techniques In order to obtain monodisperse, well-defined polymers, high vacuum, breakseal techniques and ultra-pure solvents and reagents are employed supplemented with the use of teflon stopcocks suitable for vacuum work. The experimental procedures are similar to that employed for the synthesis of macrocyclic PS and P2VP, and are described elsewhere 1/2,9,10. Synthesis &f living P2VP-b-PS-b-P2VP The following is a typical procedure used to make solutions of living P2VP-b-PS-b-P2VP in THF: A reaction apparatus was cleaned carefully and annealed in a annealing oven. During this process, the reaction vessel was heated up to 5 65 °C so that a trace amount of possibly remained polymer residue would decompose completely. After the ampoules of initiator, monomers and wash carbanion solution were attached to the main vessel, the entire apparatus was evacuated, checked for leaks, and flamed. The reaction vessel was then cooled in dry ice/isopropanol and THF was distilled in through the vacuum line until the total volume was approximately 150 ml. The breakseal containing carbanion wash solution was broken and the contents were 45 allowed to mix with THF. The entire apparatus was then rinsed with this dilute carbanion solution to remove any reactive impurities from the glass, as well as the solvent. This red solution was then poured into a side bulb and any traces of carbanion remaining in the apparatus were then rinsed out with pure THF distilled back from the side bulb. When all of the red carbanion had been rinsed into the side bulb, the solvent was carefully distilled back into the reaction flask, and the dry carbanion residue remained in the side-bulb. The ampoule of initiator solution wrapped in dry ice was then broken and introduced to the reaction flask. During this step, the temperature of the initiator should be close to the temperature of THF solvent to avoid splash. As the THF was warmed to 0 °C, a cold dauber (-78 °C) was used to assist in washing traces of initiator left in the lower portions of the ampoule and the upper portions of the apparatus down into the solution. Care was taken to make sure that there were no traces of initiator on the tip of the monomer inlet tube, as clogging, due to premature polymerization, would occur. Once all of the initiator was washed down, the solution was cooled to -78 °C, the ampoule of ~ 30% styrene/THF solution was broken and the contents were added dropwise to the rapidly stirred solution. The color of the reaction solution changed from dark green to yellow-orange. In this case, the level of the cooling 46 solution (dry ice/isopropanol) was kept a little below the solvent level inside the flask in order to prevent the formation of a high molecular weight polymer ring. Efficient magnetic stirring was maintained in order to ensure uniform dispersion of monomer drops into the solution. The system was opened to the vacuum line several times during the polymerization in order to maintain an optimum vacuum. Upon completion of the addition of styrene/THF monomer, the solution was allowed to stir for a few minutes and degassed one more time. The ampoule of containing a 10% DPE/THF solution (1 . 2 eq.) was then broken and the contents were allowed to drop into the stirring reaction solution. After the color of reaction solution was changed instantly from yellow-orange to red, the ampoule of ~ 30% 2- vinylpyridine/THF monomer solution was broken and the contents were added dropwise to the rapidly stirred red solution. The coupling reaction and fractionation of PS-b-P2VP macrocycles are essentially similar as those procedures applied in the case of PS or P2VP macrocycles 1^2,9,10_ por the purpose of fluorescence measurements, the PS-b-P2VP copolymers (10.0 mg) were purified by dissolution in 0.5 ml of THF and reprecipitation with 10 ml of hexane. This treatment was repeated several times in order to completely remove traces of contaminants such as naphthalene or coupling 47 reagents. The samples were dried in a vacuum oven at 60 °C for two days. The major characteristic data of these samples (#1,#5,#6) were listed in Table 1. Fluorescence Measurements Fluorescence spectra were obtained using a Perkin-Elmer LS-5 fluorescence spectrophotometer with a Perkin-Elmer 561 recorder, and checked on a Perkin-Elmer 650-10S. Spectrophotometric grade THF was purified by distillation over Na-K alloy and checked for fluorescence prior to use. Interfering emission was not observed in the solvent and solvent impurities were not detectable by gas chromatography. The purified PS-P2VP block copolymers of MW's ranging from 5,700 to 29,000 were dissolved (0.100 gram/liter) in THF, by placing them in a 45 °C water bath stirring until all solids were dissolved. Results and Discussion Synthesis and Isolation of macrocyclic PS-b-P2VP The synthesis of the PS-b-P2VP macrocycles was essentially the same as that used for macrocyclic PS and P2VP homopolymers, which had been documented previously 1/2,9,10_ Lithium naphthalide in THF at -78 °C was used as initiator to prepare a monodisperse two-ended living PS dianion. A small 48 portion of which was separated in a few cases for protonation and subsequent SEC analysis. A slight excess (-1.2 equiv) of a THF solution of DPE was then added to the flask in order to convert the highly reactive PS anion to the more stable 1,1- diphenyl alkyl anion. This procedure is known to prevent side reactions in the subsequent initiation of 2-VP. The THF solution of 2-VP monomer to the polystyryllithium was then added 15 minutes after addition of the styrene at -78 °C . The color changed from orange red to wine red characteristic for the "living" 2-picolyl carbanion. Because of the much greater stability of this anion the coupling reactions could be carried out at room temperature without significant side reactions such as proton abstraction from the solvent. Then, a small portion of the dianion precursor was separated and protonated (by CH3OH) in order to obtain a linear block copolymer with the same DP as the macrocycle. The remaining portion of the two-ended precursor was coupled at very low anion concentrations (1 0-^M) under high vacuum conditions. An advantage in this synthesis is that the highly reactive polystyryl anion is transformed into a much less reactive picolyl anion that is, however, reactive enough for coupling with 1.4-bis (bromomethyl) benzene (1,4-DBX). Even in this case, the stability of the extremely dilute (1 0_^-1 0~ ^M) P2VP dianion precursors is limited presumably as a result of partial dissociation of the P2VP~Li+ ion pairs into the 49 highly reactive free anions - * - 1. Therefore the coupling reaction was carried out by simultaneous addition of a concentrated (~1 0~2m) precursor dianion and a dilute (~10-3m) solution of 1,4-DBX at such a rate that the anion concentration was maintained at about 10-^-10-®M. Under these conditions the residence time of the anions in the reaction solution does not exceed a few seconds. It should be pointed out that premature termination of the dianion under these conditions forms a monoanion that further couples with 1,4-DBX to form high DP products that do not contaminate the cycles (1) . In order to further prevent side reactions, during several later trials the reactor was redesigned to be able to carry all reactions including the coupling reaction at low temperatures (-78 °C). This is especially crucial for a demanding synthesis of this type. The macrocyclic PS-P2VP block copolymers were fractionated by slow addition of non-solvent to the crude product dissolved in THF. The isolation of this macrocycles from the crude reaction product is a different and more time consuming process than for the homopolymers. Instead of using chloroform-methanol, THF was used as the solvent and hexane as the non-solvent. This generally gave satisfactory separation. Because of the presence of the 2-VP block, pure methanol could not be used as the nonsolvent as methanol is a good solvent for P2VP. The higher MW chain-extension 50 "polycondensate" produced by intermolecular coupling was fractionally precipitated while mostly the more soluble macrocycles remained in supernatant solution which were recovered and analyzed by SEC. On the other hand, a new fractionation method was attempted by slowly evaporating the solvent (THF) of crude products with efficient stirring to let the higher MW polymers gradually precipitate. This sometimes gave even better separation results than the traditional method (addition of non-solvent to the crude solution) . Since extensive fractionation tends to give low yields, we kept the number of fractionation steps as low as possible. In case of high MW shoulders were present, additional non solvent was added to completely remove the high MW product. Care was taken to ensure that the SEC elution volume maximum and molecular weight distribution of the macrocycle in the crude product was not substantially altered by the fractionation procedures. Thus, the peak maximum (Mp) of fractionated macrocycles was always identical to that in the crude product indicating no inadvertent fractionation of the macrocycle itself occurred.. Due to limitations in the size of the reaction vessels and in order to maximize the quantities of macrocyclic polymer products, relatively concentrated "living" polymer precursor solutions were used. Probably as a result of this, the samples listed in Table 1 51 are not extremely monodisperse. Another probable reason for the relatively wide MW distributions of the linear precursors is the fact that these polymers were analyzed before precipitation. Both the Tg and Td data indicated that the macrocycles do not contain major fractions of linear polymers. From the SEC data, we estimate less than 5% linear impurities. Characterization of PS-b-P2VP macrocycles The SEC chromatograms of the linear PS precursor, the PS-b-P2VP diblock precursor and the crude macrocyclic PS-b- P2VP with a DP of 66 are shown in Figure 1. The peak of the linear PS-b-P2VP diblock copolymer is shifted to higher molecular weights without the presence of any remaining PS homopolymer. The fractionated macrocyclic PS-b-P2VP shows an appreciably lower apparent MW than that of the linear precursor. Also, the peak molecular weight (Mp) of the fractionated copolymer are essentially identical to that of the crude reaction product. The ratio denoted by <G>, of apparent peak molecular weights (Mp) of cyclic- and linear block copolymers is quite low (0.71) . This substantial difference in elution volumes, which is also observed for the PS and P2VP macrocycles, is strong evidence for the presence of highly pure macrocycles. 52 <G> values of PS-b-P2VP macrocycles are listed in Table 1. As shown in the table the <G> values appear to decrease slightly with increasing DP as was also observed for macrocyclic PS and P2VP ^he <q> values (0.68-0.72 in the 6,000-60,000 MW range) are somewhat smaller compared with macrocyclic PS and P2VP. In all cases the linear and macrocyclic block copolymers are relatively monodisperse (ratio of weight-to number-average MW, Mw/Mn < 1.4). The NMR spectra of macrocyclic PS-b-P2VP polymers clearly indicate the presence of the PS and P2VP blocks. As was demonstrated for the PS and P2VP homopolymers there is considerable peaks broadening for all the signals indicating the formation of the polymer. In most cases, the block lengths of PS and P2VP were kept virtually identical. For example, the first stage of the polymerization (PS block) of sample #2 showed that the MW of this sample was around half of the MW of the corresponding linear PS-b-P2VP block precursor (Figure 1) . The composition (PS/P2VP) of the linear PS-b-P2VP block copolymers determined by -*-H NMR integration of the H- 6 resonance of P2VP relative to the aromatic resonance of PS and that of H3 -H5 of P 2 VP also indicated equal block lengths for this sample, which is in good agreement with SEC result. Although there are SEC and NMR results indicating the formation of PS-b-P2VP macrocycles, these evidences cannot be 53 considered conclusive. More convincing characterization evidences for these novel macrocycles are provided in the next section. Thermal analysis by using DSC and TGA DSC analysis has been of great value in the characterization of macrocyclic PS and P2VP 12,14_ As illustrated in Table 1, the Tg values of the cyclic block copolymers are almost a constant above a DP of about 66 then slightly decrease below a DP of 66 with decreasing Mp. Whereas the linear PS-b-P2VP copolymers show the expected decrease of Tg with decreasing DP. Thus, the lowest MW (DP=57) PS-b-P2VP macrocycles have Tg values that are 16 °C higher than that of for the isobaric linear block copolymers. This trend is similar to the cases for the PS and P 2 VP macrocyclic homopolymers 12,14_ -phe higher Tg of the macrocyclic PS-b-P2VP with decreasing DP is interpreted as due to the decreased mobility of the lower DP macrocycles and with the absence of end groups. As was found for the case of PS and P2VP macrocycles below a DP of 42 there is a significant decrease of Tg with decreasing DP. The origin of this effect is currently not well understood but may be due to the greater translational mobility of these smaller cycles. The Tg values for the cyclic block copolymers are essentially identical with that of the corresponding PS and 54 P2VP macrocycles. This is hardly surprising since the Tg1 s of linear and macrocyclic PS and P2VP are nearly identical. Because of this the Tg values of the linear and cyclic block copolymers are rather insensitive to composition. Figure 2 and Table 1 show the relationship between thermal decomposition temperature (Td) measured by thermal gravimetric analysis (TGA) and DP. The Td values for linear PS-b-P2VP block copolymers are around 400 °C, and the Td values for cyclic PS-b-P2VP are especially low in the case of low DP samples. With increasing DP, the Td differences between the linear and cyclic polymers decrease very rapidly. The difference in Td between low DP macrocyclic and linear polymer are probably due to ring strain effects. Low molecular weight cyclic polymers are readily decomposed aided by the high strain. For the large rings, ring strain is no longer a factor. The purity of the macrocycles may be estimated provided that the only contaminant is a linear polymer of the same DP. The very different Td vs DP profiles of macrocyclic and linear PS-b-P2VP may be still another sensitive criterion for the purity of the low DP samples. The difference in Td between low DP linear and macrocyclic PS-b-P2VP copolymers is 30-35 °C. The difference in Tg1s in this case is only 14- 16 °C (Table 1) . 55 We also investigated the Td values for some macrocyclic P2VP homopolymers (i.e. sample #8 is shown in Table 1) . Almost the same Td trend was observed in this case indicating that the Td differences were due to the nature of the rings and were not relative to the blocky character of the polymers. Morphology of PS-b-P2VP macrocycles The lamellar morphology of microphase-separated cyclic PS-b-P2VP copolymers was studied by the group of Prof. E. Thomas (MIT) 15 . The focus of this study is to probe the effect of loops versus bridges on the morphological characteristics of microphase-separated block copolymer system. While linear ABA block copolymers can be arranged as bridging and looping chain conformations, only the latter option is available to macrocyclic AB diblocks 2 (Figure 3). Thus, compared with linear block copolymers of the same DP and block composition the macrocyclic block copolymer have a significant smaller domain size. In fact, the effect of looped versus bridged chain conformations in the microphase- separated state through an analysis of the relative domain spacings of the lamellar formed by the cyclic diblock and linear triblock systems have been directly observed The periodicity of the ordered microstructure, D, for the cyclic PS-b-P2VP (MW=56,000, PS%=50%) is 202 A and for the isobaric 56 linear precursor P2VP-b-PS-b-P2VP is 221 A. This observation is attributed to the differences in the architecture between the two systems Excimer Emission Intensities Differences between Linear and Macrocyclic PS-b-2VP copolymers In order to check for the occurrence of intramolecular energy transfer, equimolar solutions of mixtures of isobaric PS and P2VP were compared with that of PS of the same MW of the same total concentration (0.01 g/1) in THF. The emission intensities of the mixtures were roughly half indicating the absence of major intermolecular energy transfer processes. In this case, the emission intensities increased proportionally with PS concentration increasing from 8x10“^ M to 2x10~3 m (chromophore units), which eliminated intermolecular effects in this concentration range and more detail was described in Chapter 1. However the emission of the linear PS-b-P2VP (sample #1 in Table 1, Mp=3,100-2,600) in THF solution was much less (~6 times) than that of a solution containing a mixture of the two homopolymers of roughly the same molecular weights (PS=4,000, P2VP=4,000). This indicates significant intramolecular energy transfer in the linear PS-P2VP block copolymer, causing quenching of the PS singlet excited state. 57 Similar intramolecular quenching of PS emission in block copolymers has been demonstrated by Lindsell, Robertson, and Soutar for the case of PS-polybutadiene block copolymers -*-6 and by Phillips et al. for the case of styrene- containing block copolymers ^ . The emission intensities of the PS-P2VP macrocycles were found to be about twice that of the matching linear block copolymers. These differences appear to be related to a less efficient intramolecular energy transfer from the PS to the P2VP blocks in the macrocycles. These intriguing differences between the linear and macrocyclic block copolymers are of considerable interest. The presence of low MW impurities resulting from the end to end coupling process is unlikely since the emission spectra did not change upon repeated purification of the cycles by reprecipitation. The presence of the bis-1,4-methylenebenzene coupling unit is not likely to be the reason either since the emission of this 1, 4-substituted unit is similar in wavelength and intensity to that of the pendent phenyls. Furthermore if the coupling unit somehow were responsible for the differences then the difference in emission between linear and macrocyclic PS-b- P2VP of the higher molecular weight would be much smaller. Therefore we conclude tentatively that the greater emission intensities of the cycles are related to a less efficient energy transfer process in the macrocyclic block copolymers 58 from the PS excimers to the P2VP blocks. The appreciably lower degree of quenching in the macrocyclic PS-b-P2VP is likely due to the much lower conformational mobility of both blocks in the macrocycle. Molecular modeling studies appear to support this. Thus there are few conformation where the P2VP and PS are in close proximity. In fact it is rather surprising that there are no significant DP effects in this case since these effects would be expected to be more pronounced for low DP macrocycles. Acknowledgement: This research was supported by NSF-DMR Polymers Program. We would like thank Joyce Laquindanum for her help with the TGA measurements and Dahai Dong for his design of a low-temperature reactor. References 1. (a) Toreki, W.; Hogen-Esch, T. E. and Butler, G. B., Polym. Preprs, 1987, 28(2), 343. (b) Toreki, W. and Hogen-Esch, T. E., Polym. Preprs, 1989, 30(1), 129. (c) Hogen-Esch, T. E.; Sundararajan, J. and Toreki, W., Makromol. Chem. Macromol. Symp., 1991, 47, 23-42. 2. (a) Sundararajan, J. and Hogen-Esch, T. E., Polym. Preprs, 1991, 32(3), 604. 59 (b) Gan, Y.; Zoller, J.; Hogen-Esch, T. E., Polym. Preprs, (ACS, Div. Polym. Chem.) 1993, 34(1), 69. (c) Gan, Y.; Zoller, J.; Yin, R. and Hogen-Esch, T. E., Macromol. Symp., 1994, 77, 93-104. 3. (a) Hild, G.; Kohler, A. and Rempp, P., Eur. Polym. J., 1980, 16, 525. (b) Hild, G.; Strazielle, C. and Rempp, P., Eur. Polym. J., 1983, 19, 721. (c) Rempp, P.; Strazielle, C. and Lutz, P., Encyclopedia of Polym. Sci. & Engineering, 1987. 4. (a) Geiser, D. and Hocker, H., Macromolecules, 1980, 13, 653. (b) Hocker, H., Angew. Makromol. Chem., 1987, 100, 87. 5. Vollmert, B. and. Huang, J, Makromol. Chem., Rapid Commum., 1980, 1, 333; Ibid, 1981, 2, 467. 6. Roovers, J. and Toporowski, P. M., Macromolecules, 1983, 16, 844. 7. McKenna, G. B.; et al., Macromolecules, 1987, 20, 498- 512. 8. DiMarzio, E. and Guttman, C., Macromolecules, 1987, 20, 1403. 9. Sundararajan, J., Ph.D thesis (Chemistry), University of Southern California, April, 1991. 10. Sundararajan, J. and Hogen-Esch, T.E., Polym. Prepr., 1992, 33(1), 162-3. 60 11. (a) Tardi, M.; Rouge, D.; Sigwalt, P., Eur. Polym. J., 1967, 3, 85. (b) Tardi, M.; Sigwalt, P., Eur. Polym. J., 1972, 8 , 151. (c) Chang, C. J.; Kiesel, R. F.; Hogen-Esch, T. E., J. Am. Chem. Soc., 1975, 97, 2805. 12. (a) Gan, Y., Ph.D thesis (Chemistry), Chap.3, University of Southern California, May, 1995. (b) Gan, Y.; Dong, D.; and Hogen-Esch, T. E., Polym. Preprs., (Am. Chem. Soc., Div. Polym. Chem.) 1994, 35(2), 574. 13. Dong, D., Ph.D thesis (Chemistry), University of Southern California, July, 1995. 14. Gan, Y.; Dong, D.; and Hogen-Esch, T. E., Macromolecules, 1995, 28, 383-385. 15. Lescanec, R. L.; Hajduk, D. A.; Kim, G. Y.; Gan, Y.; Yin, R.; Gruner, S. M.; Hogen-Esch, T. E.; and Thomas, E. L., Macromolecules, 1995, 28, 3485-89. 16. Lindsell, W. E.; Robertson, F. C.; and Soutar, I., Eur. Polym. J., 1981, 17(3), 203-8. 17. Phillips, D.; Roberts, A. J./ Rumbles, G.; and Soutar, I., Macromolecules, 1983, 16, 1597. 61 Table 1: SECa, NMR, DSC and TGA Characterization: Apparent Molecular Weights, Molecular Weight Distributions, Glass Transition Temperatures and Decomposition Temperatures of Isobaric Linear (L) and Macrocyclic (C) Polystyrene-Poly(2-vinylpyridine) Block Copolymers (PS-b-P2VP). Mpb Mnb DC <G>d Composition e Tgf TdS (%PS) (°C) (°C) 5,700 (L#l) 5,200 1.32h 0.73 55% 81 405 4,200 (C#l) 3,700 1.17 97 374 6,600 (L#2) 6,200 1.39h 0.71 58% 85 407 4,700 (C#2) 4,600 1.18 (55%)i 99 372 7,700 (L#3) 7,200 1.32h 0.72 58% 89 399 5,600 (C#3) 4,800 1.14 101 375 13,000 (L#4) 15,000 1.34h 0.73 47% 98 9,500(C#4) 10,500 1.32 100 20,000 (L#5) 20,000 1.09 0.71 63% 100 14,000 (C#5) 13,000 1.29 101 29,000 (L#6) 28,000 1.28h 0.71 86% 100 21,000 (C#6) 22,000 1.16 102 56,000 (L #7) 59,000 1.31h 0.68 50% 100 402 38,000 (C #7) 40,000 1.36 101 406 6,500 (L#8) 6,300 1.30h 0.74 o% j 90 407 4,800 (C#8) 4,600 1.18 99 373 a. SEC measurements were carried out by using THF as eluting solvent. b. Apparent peak and number average molecular weights of linear and macrocyclic polymers. (Polymers with the same # are of the same DP). c. Polydispersity, D=Mw/Mn. d. Ratio of the apparent peak molecular weights of the cyclic and linear PS-b-P2VP (<G>=Mp,c/Mp,1). e. Calculated by NMR measurements. f. Glass transition temperatures measured by DSC. g. Decomposition temperatures measured by thermogravimetric analysis (TGA). h. Measured directly on crude reaction solution, i. Calculated from SEC data. j. P2VP homopolymers. 62 12 14 18 20 22 Figure 1: Size Exclusion Chromatograms of Precursor (3), Linear PS-b-P2VP Precursor (2) Macro-cyclic PS-b-P2VP (1) (DP=6 6). Linear PS and Crude 63 WEIGHT (%) 100 75 * 50- 25 - 7 5 0 6 5 0 1 5 0 2 50 3 5 0 5 5 0 5 0 4 5 0 Tem perature (C) Figure 2: Thermogravimetric Curves of Linear P2VP-PS-P2VP Block Copolymer (-----) and Purified PS-b-P2VP Macrocycle (----) (DP = 66) . 64 linear triblocks B A "Bridge" "Loop" linear diblocks ring diblocks B A Figure 3. Bridging and looping chain arrangements in block copolymer microdomains. 65 Chapter 3 Effects of Lithium Bromide on the Glass Transition Temperatures of Linear- and Macrocyclic-Poly(2- vinylpyridine) and -Polystyrene Abstract The glass transition temperatures (Tg's) of linear and cyclic Poly(2-vinylpyridine) (P2VP) and polystyrene (PS) are determined in the presence and absence of LiBr. In contrast to earlier conclusions the Tg values of narrow molecular weight distribution P2VP and PS macrocycles decrease rather than increase with decreasing DP. However in agreement with earlier results the Tg differences between macrocyclic and linear P2VP's and PS's strongly increase with decreasing DP. In the presence of varying amounts of LiBr, the Tg's of both linear and cyclic P2VP increase linearly with added LiBr whereas the case of PS the Tg values are independent of added LiBr. The slopes of the Tg vs LiBr concentration plots are higher for macrocyclic P2VP than for the corresponding isobaric linear chains, and the slopes of these plots of linear P2VP tend to increase with decreasing DP. These results 66 may be interpreted as due to the interaction of Li+ ions with the pyridine nitrogens of the chain thus suppressing segmental motion. Introduction We have reported previously that the values of the glass transition temperature (Tg) of macrocyclic poly (2- vinylpyridine) (P2VP) increase with decreasing molecular weight (MW) consistent with entropy calculations . This increase in Tg was interpreted as reflecting the increased conformational stiffness of the macrocycles as the degree of polymerization (DP) decreased - * - “3. Differences in Tg were reported as large as 4 0 °K between linear and macrocyclic P2VP around a DP of 40 ^. These differences in Tg were observed to disappear at a DP of around 200. However, the poly(2-vinylpyridine) macrocycles were synthesized by reacting the lithium dianion "living" precursors with 1,4-bis-(bromomethyl) benzene (1,4-DBX) thus generating LiBr and the polymers were isolated by precipitation in hexane. Therefore it is possible that the LiBr present in the polymer matrix may have increased the glass transition temperature of the P2VP macrocycle for instance by coordination of one or more pyridine groups to the Li cation. Such coordination processes are well 67 documented for polyethylene oxide (PEO) and were demonstrated to affect the DSC behavior and conductivity of the PEO matrix 5-7 _ Effects of added LiC1 0 4 on the Tg of PE0-P2VP blends were recently observed Therefore a study of the possible effects of LiBr on the Tg of linear and macrocyclic P2VP seemed of interest. In order to evaluate the occurrence of salt effects unrelated to the direct coordination of the P2VP pyridine nitrogens with Li ion, we also investigated the effect (s) of LiBr on the Tg of polystyrene (PS) since this polymer resembles P2VP both with regard to structure and thermal properties but does not specifically coordinate the Li cation. Experimental Section Synthesis: The anionic polymerization of 2VP, the end to end cyclization and the fractionation of cyclic P2VP has been described previously a series of matching monodisperse linear P2VP's and fractionated cyclic P2VP's with the same molecular weight were prepared. The cyclic P2VP samples prepared by this method contained a certain amount of LiBr, which was removed by precipitation in water followed by centrifugation and drying. The resulting polymer was redissolved in THF then precipitated in hexane. 68 Macrocyclic PS was synthesized by initiation of styrene in THF at about -70 °C using lithium naphthalene as initiator. Styrene was distilled into the reactor under high vacuum over about one hour. After completion of the polymerization an aliquot of the PS dianion precursor was withdrawn and protonated (by CH3OH) to serve as an isobaric linear PS sample. Cyclization of the remaining PS dianion with 1,4- bis-(bromomethyl) benzene (DBX) was performed under high vacuum at -70 °C and at low carbanion concentration (<3xl0- ^M). DBX (TCI America, Inc.) was recrystallized in CHCI3 several times prior as coupling agent. Both linear and macrocyclic PS samples were obtained by precipitation in excess methanol. Sample Preparation: Anhydrous LiBr (Aldrich) (15.00 mg) was dissolved in 10.0 ml methyl alcohol (MeOH). Aliquots of 0.0, 0.2, 0.4, 0.8, 1.6, 2.4, 4.0 ml of the LiBr/MeOH solution were transfered to 7 sample vials each containing 40.0 mg of P2VP, and methanol was added to each vial to give 4.0 ml of solution. After the solvent was removed, the samples were dried in vacuo for two days at 50- 60 °C. In case of PS, due to its poor solubility in MeOH, THF was used as the solvent in order to obtain the LiBr- containing PS samples. 69 Measurements: Differential scanning calorimetry (DSC) was carried out on a Perkin-Elmer System-4 thermal analyzer under nitrogen atmosphere. The instrument was calibrated with indium (m.p.=156.60 °C) before and after use. All samples (20-30 mg) were sealed in aluminum pans and preheated from 40-50 °C to 140-150 °C at the rate of 20 °C/min and cooled to 40-50 °C at the rate of -20 °C/min in order to assure identical thermal treatments, and then were heated to 140-150 °C at the rate of 20 °C/min. The DSC analyses were carried out 2-3 times. Reproducibility was found to be within 1 °C. The glass transition zone was determined as the temperature range between two intersection points of the base lines with the extrapolated sloping portion of the DSC curves, the value of Tg being defined as the mid-point of the heat capacity change. SEC Analysis: Size exclusion chromatography (SEC) was carried out by using THF as the eluting solvent at 1.0 ml/min. One percent by volume of triethylamine was added to the THF in order to prevent adsorption of P2VP onto the column. Molecular weight calibration was done using narrow molecular weight distribution P2VP and PS standards (Polysciences, Inc.). 70 Results and Discussion To-of linear, unpurified cyclic and purified cyclic P2VP as a function of molecular weight In order to evaluate the effect(s) of LiBr on the Tg of macrocyclic P2VP, the four P2VP macrocycles having DP's of 19-190, were first synthesized and then fractionated. The fractionation procedures involved repeated addition of hexane to THF solutions of the crude reaction product followed by filtration to remove high MW "polycondensation" products The SEC peak maxima of the macrocycles were essentially identical with those of the crude products. The THF solutions of the resulting narrow MW distribution P2VP macrocycles (Table 1) were then divided into two sets. One set was precipitated in hexane and then analyzed by DSC. The other identical set of samples was precipitated in water to remove LiBr followed by centrifugation and drying. In order to eliminate any solvent effects on Tg, the samples were then redissolved in a small amount of THF reprecipitated in hexane and analyzed by DSC. The isobaric linear P2VP samples obtained by protonation of the P2VP precursor dianion were also precipitated in hexane (Scheme 1) • Multiple reprecipitation in water didn't affect the Tg values of the P2VP macrocycles. A P2VP macrocycle of a DP 71 of 42 was used to carry out studies on the effects of added LiBr on Tg. Some of this sample containing LiBr was later reprecipitated in water. DSC analysis of this polymer sample showed that the Tg value was identical with that of the macrocycle before addition of LiBr. 0 2-VP 1) EBr, I2'--------- I ► I E + LiBr Vy'L i + . V A ,.p~ .,p 2-\ 2) Fractionalimr" DPrecipitate in H.O /(C y c lic P 2 V n \2 )Dissolve jn THF DMeOH / \ „ p 2)Precipitate in hexane /^Precipitate in hexane \3)Reprecipitate r in hexane C~X r\ , H I E + LiBr I E W W AA (Linear P2VP) (U npurified (Purified 2 Cyclic P2VP) Cyclic P2VP) I = initiator EBr2= B rC H2 — ^ ~ ~ ^ - C H 2B r Scheme 1: Synthesis of Linear and Macrocyclic P2VP's and Purification of Cyclic P2VP The Tg vs. MW profiles of the linear, the isobaric unpurified and purified P2VP macrocycles are shown in Figure 1. The unpurified cycles show increasing Tg values with decreasing DP as observed previously. DSC studies on P2VP with high contents of LiBr did not show any glass 72 transitions. The purified P2VP macrocycles (without LiBr) show essentially no variation of Tg with DP with a slight decrease of about 5 °C for the macrocycle having a DP of about 20. It is clear that the differences in Tg between the purified and unpurified P2VP macrocycles are related to the absence and presence of LiBr. Effects of LiBr content on the Tg of isobaric linear and macrocyclic P2VP However Tg differences between purified macrocyclic- and linear P2VP although smaller than reported before remain appreciable increasing with decreasing DP to a value of about 38 °C at a DP of 20. In order to document these effects further, LiBr blends of linear and macrocyclic P2VP of a DP of 42 were prepared and analyzed by DSC (Figure 2). In both cases a linear relationship is obtained between Tg and LiBr content. The slope for the macrocycle is clearly lower than that of the linear P2VP. In order to confirm that the observed variation in Tg is related to the presence of the 2-pyridyl groups, similar LiBr blends were prepared for a linear polystyrene of approximately the same DP (=40) (see Exp. Section). In this case LiBr has no effect on Tg. The above data confirm that the apparent increases in Tg with DP of the unpurified P2VP macrocycles are related to the presence of LiBr. Coordination of Li cation to one or 73 more pyridine nitrogens is expected to substantially reduce the mobility of the chain, and this reduction is expected to be larger for the linear than for the macrocyclic P2VP since the linear chain is conformationally less restrained and hence undergoes a greater loss of mobility upon coordination. The interaction between pyridine nitrogen and the lithium cation seems to only relate to the number of Li+ and 2VP monomer units and appears to be unrelated to the polymer chain length. The mole ratio of LiBr and P2VP is in proportion to their weight ratio. In this case, both the mole ratio or weight ratio will represent the effects of LiBr on the Tg of P2VP. LiB-r_e.ff.ects on Ta of P2VP with different DP. and the width of Tg transition The effect of addition of LiBr on the Tg values of linear P2VP's of different DP's is shown in Figure 3. The slope again is greater for the lower MW sample (DP=42) probably for the same reason as the corresponding effects for the matched linear-macrocyclic pair of Figure 2. The Tg of the longer less mobile chains is plausibly less affected by the presence of LiBr. It is of interest to note that the DSC plots of the P2VP-LiBr systems show other differences with the P2VP 74 samples without LiBr, the width of the thermal transitions increasing with increasing LiBr concentration (Figure 4). Thus the P2VP macrocycle (DP=42) in the presence of low concentration of LiBr shows a relatively narrow Tg transition of about 8 °C. In the presence of 15 weight % of LiBr, however, the width of the transition is increased to about 43 °C. Similar results were found for linear P2VP but not for linear polystyrene. This may indicate the presence of P2VP chain segments having a wide spectrum of mobilities consistent with a variety of microenviroments resulting from different degrees of coordination of the pyridine groups. Effects of even higher LiBr contents unfortunately could not be studied since the glass-transition region began to overlap with a melting-transition endotherm. In principle, the above results could be used to estimate the LiBr content of the unpurified macrocycles. For instance for the unpurified macrocycle (DP=42) a salt content of 8 weight % was determined from the apparent Tg value. This is considerably higher than calculated on the basis of the reaction stoechiometry (~4 Wt%). However it should be pointed out that the macrocycles are recovered by repeated precipitation of the high MW side products resulting from intermolecular ("polycondensation") reactions. During this fractionation process, hexane is added dropwise to a THF solution of the crude cyclization 75 product and higher MW P2VP is then first precipitated out. At this point, the resulting solution containing a large fraction of THF will dissolve most of the LiBr salt. It is possible therefore that the LiBr was inadvertently concentrated in the low MW macrocycles. This concentration effect may have been amplified by effective Li ion binding by the P2VP macrocycle that may resemble binding by crown ethers. For the same amounts of monomer and solvent, the formation of high MW cyclic P2VP (DP=190) uses less initiator and coupling agent, so that less LiBr is produced. As a result the Tg values of purified and unpurified P2VP (DP=190) are found to be almost identical. Comparison of Tg behaviors between purified P2VP and PS macrocycles Because of the structural similarity of P2VP and PS, the determination of the Tg behavior of PS macrocycles as a function of DP was of obvious interest - * - 0 . A number of PS macrocycles were prepared by D. Dong of this group using techniques similar to that used for macrocyclic P2VP (Table 2) 11. In this case the presence of LiBr in the PS macrocycles is not likely since LiBr is highly soluble in methanol, the precipitating solvent for polystyrene (see Exp. Section). 76 The corresponding Tg vs. MW plots for macrocyclic and linear PS show patterns that are very similar to that observed for the case of purified macrocyclic and linear P2VP (Figure 5). Starting at high DP there is essentially no change in the Tg of macrocyclic PS as the DP decreases until a DP of about 40, after which there is a rapid decrease to a Tg of 8 6 °C at a DP of 20. For both the P2VP and PS cases the difference in glass transition temperature, ATg, between linear and macrocyclic polymers increases with decreasing DP as concluded previously 2. It is not clear to what degree this increase in ATg is due to the absence of end groups in the macrocycles and to the increasing stiffness of the macrocycles as the DP decreases. The availability of even lower DP macrocycles would be helpful in this regard. Unfortunately the differentiation between linear and macrocyclic polymers for instance by SEC is difficult since the difference in hydrodynamic volumes appears to narrow appreciably below a DP of about 20 as indicated by SEC (Table 1). Thus other characterization methods will have to be found. This is currently under study. Acknowledgments This work was supported by NSF-DMR, Polymers Program. We wish to thank Professor R. Salovey for 77 the use of the DSC equipment and Mr. B. Yeom and Mr. F. W. Shen for their technical assistance with the DSC measurements and data analysis. References 1. (a) Toreki, W.; Hogen-Esch, T.E. and Butler, G.B. Polym. Preprts., 1987, 28(2), 343. (b) Toreki, W. and Hogen-Esch, T.E. Polym. Preprts., 1988, 29(2), 416. (c) Toreki, W. and Hogen-Esch, T.E. Polym. Preprts., 1989, 30(1), 129. 2. Hogen-Esch, T.E.; Sundararajan, J. and Toreki, W. Makromol. Chem., Macromol. Symp., 1991, 47. 23-42. 3. Sundararajan, J. and Hogen-Esch, T.E. Polym. Preprts., 1992, 33(1), 162. 4. DiMarzio, E. and Guttman, C.Macromolecules, 1997, 20, 1403. 5. (a) Fenton, D.E.; Parker, J.M. and Wright, P.V. Polymer, 1973, 14, 589. (b) Wright, P.V. Br. Polym. J., 1975, 7, 319. (c) Wright, P.V. J. Polym. Sci., Polym. Phys. Ed., 1976, 14, 955. 6 . Watanabe, M. et al, Polymer J., 1986, 18, No. 11, 809- 817. 7. Wintersgill, M.C. et al, Polymer, 1989, 30, June 78 (Conference issue), 1123-6 . 8 . Li, J.; Mintz, E.A. and Khan, I.M.Chem. Mater., 1992, 4, No. 6, 1131-4. 9. Li J. and Khan, I.M. Macromolecules, 1993, 26, No. 17, 4544-50. 10. (a) Rempp, P.; Hild, G. and Strazielle, C.Eur. Polym. J., 1983, 19(8), 721. (b) Roovers, J. Macromolecules, 1985, 18, 1359. (c) Hocker, H. and Geiser, D. Macromolecules, 1980, 13 (3), 653. (d) Deffieux, A., private communication. 11. (a) Sundararajan, J., Ph.D., dissertation, University of Southern California (1991) . (b) Dong, D., Ph.D., dissertation, University of Southern California (1995) . 79 TABLE 1: Apparent molecular weights (MW) and glass transition temperatures (Tg) of the unpurified and purified cyclic (C) P2VP and linear (L) P2VP of identical degree of polymerization. Mp(L)a D(L)b Tg(L) (°C) Mp(C)c D(C) Tg(C)unpurified (°C) Tg(C)purified (°C) <G>d 1900 1.12 58.0 1590 1.05 — 95.9 0.84 4220 1.08 85.3 3410 1.04 116.6 99.1 0.81 6610 1.21 89.9 4800 1.10 105.0 98.7 0.73 18970 1.16 97.3 14570 1.15 98.2 99.7 0.77 a. SEC peak molecular weight (MW) for linear precursor (L), b. D = M w /M n c. Apparent MW for fractionated cyclic polymer (C) determined by SEC. d. Ratio of Mp(C) to Mp(L). 80 TABLE 2: A p p a re n t m olecular w eights (MW) a n d glass tran sitio n tem p era tu re s (Tg) of lin ear polystyrene (PS) a n d cyclic PS w ith identical d eg ree of po ly m erizatio n synthesized b y D. D ong a n d re p rin te d w ith his perm ission. M p(L )a D(L)b Tg (L) (°C) e Tg (C) (°C) e <G>C Yield %d 1660 1. 35 63 70 0.82 70 2500 1. 11 68 88 0.79 85 4100 1.14 80 100 0 .74 65 8500 1.06 89 100 0 .73 45 22000 1. 04 98 100 0.78 45 43000 1.24 100 100 0 .79 10 73000 1 . 03 100 100 0 .81 15 a. SEC p e ak m olecular w eig h t (MW) for linear p recu rso r (L), b. D = M w /M n , c. Ratio of M p(C) to M p(L), d. D eterm ined w ith SEC resu lts of c ru d e cyclic PS sam ples, e. R eproducibility + 1.0 °C . 81 120 1 0 0 - o. 90 - 70 - 60 - 50 2 3 5 0 4 6 4 1/Mp X 104 Figure 1. Tg of linear (□ ), unpurified cyclic ( • ) and purified cyclic (■ ) P2VP as a function of 1/Mp: ( □ , # ) precipitated in hexanes only; (■ ) precipitated in H 20 follow ed by dissolution in THF and reprecipitation in hexanes. 82 140 120 - O o 100 o> H 80 - 60 : 40 20 5 10 15 Molar ratio (LiBr/2VP) X 10 o Figure 2. Effects of LiBr on the Tg of Matched Linear ( □ ) and Macrocyclic ( O ) P2VP of Mp=4200 and Linear PS ( ▲ ) of Mp=4000. 83 T g (oC) 100- 90 80 8 2 4 6 0 Molar ratio (LiBr/2VP) X 102 Figure 3. Effects of LiBr on the Tg of Linear P2VP ( A ) Mp=4200; ( O ) Mp=6600. 84 0 i i i i--------------•--1--------- ■ i i 60 80 100 120 140 160 ISO T em perature (oC) Figure 4. DSC Thermograph of Cyclic P2VP (DP=42) containing (a) 0.75 wt% LiBr; and (b) 15 wt% LiBr. 85 Tg (oC) 120 OC 90 80 70 60 50 1 00 200 500 0 300 400 MW x 10 ' 2 Figure 5. Tg's of linear ( □ ) and cyclic ( o ) po lystyrene as a function of MW. 86 CHAPTER 4 Synthesis and Characterization of a Catenated Polystyrene-Poly(2-vinylpyridine) --- A NOVEL TYPE OF BLOCK COPOLYMER Abstract The synthesis, isolation and characterization of a 50/50 polystyrene-poly(2-vinylpyridine) (Mp=10^) "catenane" is reported. This catenane is prepared by end to end coupling of a P2VP dianion lithium salt with 1,4- bis(bromomethylbenzene) in THF in the presence of a PS macrocycle (Mp=4500). The catenane was isolated by precipitation-extraction procedures that were optimized using a 50/50 PS-b-P2VP macrocycle as a model for the solubility characteristics of the catenane. The catenane copolymer was characterized by SEC, NMR and emission (fluorescence) spectroscopy. Introduction The synthesis of topologically interesting molecules has attracted considerable attention in recently years. 87 Compounds whose molecular graph cannot be drawn in a plane without intersecting lines are particularly fascinating ^. Among these, the synthesis of rotaxanes, polyrotaxanes, catenanes, polycatenanes have been discussed and various synthetic approaches have been reported (Figure 1). Typical examples of such systems are two (or more) interlocked rings (catenanes). The synthesis and study of such compounds frequently reveals interesting information on aspects of molecular architecture and intramolecular interactions and occasionally leads to new materials with unusual properties 5. 1. C atenane 2. R otaxane J n P olyrotaxan e 4. C aten ated block copolym er 5.. M acrocyclic block cop olym er Figure 1: Topologically interesting macrocycles. Polymeric catenanes 1 are comprised of two threaded macrocycles held together solely by topological constraints, i.e. lacking a chemical or physical bond between the two rings 2. Catenanes of this type are expected to display a motion of the two rings relative to one another, that is not possible with linear or macrocyclic polymers. Furthermore the intramolecular interactions of the two rings is topologically homogeneous. Thus each part of one of the rings at least in principle is expected to show the same time averaged interaction with any part of the other ring. This homogeneity of polymer-polymer interactions is absent in simple linear or even macrocyclic polymers. This relative large degree of independence, in spite of their physical linkage, makes them a unique and fascinating topic of study. Generally there are two major methods for catenanes synthesis: (A) statistical threading, in which the macrocycle is used in a large amount and threading is predominantly entropically driven, and (B) template threading, involving an enthalpic driving force such as the formation of charge transfer complexes 6,7. The template synthesis of low MW catenanes was intensively studied by Dietrich-Buchecker and Sauvage 8-10 and Stoddart et al.and recent work in the field of host- guest chemistry 12 aiso relevant. 89 Dietrich-Buchecker and Sauvage and their groups recently developed an efficient synthesis of interlocked rings based on the three dimensional template effect of a transition metal, able to gather and predispose coordinating molecular threads so as to lead to interlocked systems after cyclization 1*8, 9,13-20 _ However, these catenanes were oligomers of very low MW (MW<1000). For the case where the two rings differ chemically, the enthalpic Ah term in the free energy of the threading equilibrium of linear chains (L) with macrocycles (C) (Equation 1) 21-26 j_s expected to be positive (repulsive) 6'27_ However, the equilibrium can be shifted toward the "threaded state" (Ro) by increasing macrocycle concentration. K L + n C ^=-"" - R0 (1) Equation 1: Threading equilibrium Equilibrium (1) is driven to the right by masoaction according to K = [RO]/ [L] [C]n (2) 90 The statistical threading of linear polymers by crown ethers has been studied extensively by Gibson et al. 27,28_ It is found that the threading efficiency of low MW macrocycles and linear polymers is affected significantly by size and compatibility 29 _ Experiments by Schill et al. 30,31 ancj Harrison et al. 21 provided information about the ring-size constraints. While there are some discrepancies, both groups concluded that the cycle should contain a total of at least 22 atoms, in order to allow threading by a polymethylene chain. The synthesis of a rotaxane consisting of a 30-membered macrocycle threaded by 1,10-decanediol was reported by Agam and coworkers who concluded that for flexible chains, ring size is the most important geometric variable in determining the efficiency of threading 2 2 _ Polymeric "catenanes" consisting of two high MW polymeric macrocycles linked by catenation apparently have not yet been synthesized and characterized, although several attempts have been reported 33,34_ ^ synthesis of macrocyclic polymers by end to end cyclization, such catenated cycles (and perhaps even polycatenanes) are probably present. However, their isolation from such complicated mixtures seems problematic since this would involve separating isobaric macrocycles and catenanes. 91 For example, Wood and coworkers in their attempts to prepare polymeric catenanes cyclized polymer-supported linear polyesters in the presence of cyclic polyesters^. Gel permeation chromatography showed substantial quantities of high molecular weight material had been formed. However, due to difficulties in the separation and isolation, There was no convincing characterization for catenane formation although catenation would be probable because of the large size of the rings. The isolation and characterization of block copolymer catenanes consisting of structurally different polymer macrocycles A in principle is more readily carried out than that of the catenated homopolymers because the solubility of such a" block" copolymer would differ from that of its component macrocycles. We now wish to report the synthesis, isolation and characterization of a polystyrene (PS) -b- poly (2- vinylpyridine) (P2VP) catenane that we propose to write as PS-cat-P2VP of fairly high MW (Mp=10,000). The catenane was synthesized by the end to end coupling of a P2VP dianion (DP ~ 50) in the presence of macrocyclic PS (DP ~ 50). The reaction involving coupling of a PS dianion with 1,4- bis(bromomethylbenzene) (DBX) in the presence of a macrocyclic P2VP is doomed to failure because of side reactions 35. The techniques needed to separate the PS-cat- 92 P2VP from the reaction mixture were optimized by studying the separation of PS-b-P2VP macrocycles ( 5 . ) from the corresponding PS and P2VP macrocycles. The characterization of the PS-cat-P2VP was carried out by SEC and by - * - H NMR that was facilitated by the chemical shift of the 6-proton of the P2VP ring (8=8.25 ppm) that absorbs down field from the other P2VP and PS aromatic resonances. Experimental Materials: The experimental procedures for the synthesis, isolation and characterization of macrocyclic PS homopolymers and PS-b-P2VP copolymers have been described previously 36-40_ jn this case the macrocyclic PS (DP=45, MWD=1.09, ~2 grams) was prepared by end to end coupling of a PS dianion using 1,4-bis(bromomethylbenzene) (DBX)38,40 (tci America, Inc., 99%), that was recrystallized in CHCI3 several times. The sample was dried in a vacuum oven at 45 °C for 24 hours and on a high vacuum line (1 0 ~ 5 torr) overnight to eliminate any included solvent or moisture. A macrocyclic PS-b-P2VP (50/50) (MW=13.000) sample prepared previously was used as a model of a PS-cat-P2VP copolymer with regard to the isolation of PS-cat-P2Vp40. 93 Synthesis: Fifty ml of THF was distilled in vacuo from a diphenylmethyllithium solution into a 250 ml vessel containing 2.0 g of macrocyclic PS (DP=45). A few drops of lithium naphthalide/THF solution (0.09M) were slowly added, and the solution turned green indicating the absence of impurities. After introduction of an additional 0.405 mmoles of lithium naphthalide, 2.0 g of 2-VP was distilled into the reactor over a period of about one hour, and a small portion (5 ml) of the solution (linear precursor) was separated and reacted with MeOH to serve as standard. The remainder of the solution was segregated into an ampule equipped with a breakseal. The ampule was then attached to the cyclization apparatus and the cyclization of the P2VP dianion with an equivalent amount of purified 1,4-DBX (coupling agent) was carried out at room temperature (25 °C) as described earlier. The addition of equivalent amounts of living P2VP anions and DBX was carefully controlled so as to maintain a slight excess of the P2VP anion (10-^-10-^M) as seen by a light red color. The final reaction mixture was concentrated in vacuo, precipitated in hexane, filtered and dried in a vacuum oven overnight. Isolation: In order to optimize conditions required for separation of the PS-cat-P2VP from the corresponding macrocyclic PS and P2VP, a solution of mixtures of 94 macrocyclic PS (MW=4500), P2VP (MW=4200) and a macrocyclic PS-b-P2VP (MW=13,000, PS%=47%) of similar MW and composition as the catenane were used as model blends. Thus approximately equal amounts (1 0 mg) of each of the three polymers were dissolved in 15 ml of THF. The attempted separation of this blend by adding this solution (15 ml) to 2 0 0 ml cyclohexane or methanol gave stable emulsions, which could not be separated by centrifuging or prolonged settling. Thus the THF solution of the three polymers was first precipitated quantitatively in hexane so that a solid blend was obtained. After filtration of the precipitate and drying in vacuo overnight, the solid was extracted six times alternatively with 15 ml portions of cyclohexane and methanol. The SEC of the remaining solid indicate the absence of PS or P2VP macrocycles and only showed the peak corresponding to the macrocyclic block copolymer. For isolation of the PS-cat-P2VP from the reaction products, methanol (90 ml) was added to the dried crude reaction product in order to remove the P2VP homopolymers. The resulting milky solution was stirred and cooled to -20 °C and kept for 24 hours or centrifuged until the two phases separated. The solid polymer was filtered and dried overnight in vacuo, after which, it was dispersed in 90 ml of cyclohexane. The resulting suspension was centrifuged and the obtained solid was vacuum dried overnight. This 95 sample was extracted six times alternately by 2 0 ml portions of methanol and cyclohexane until the remaining solid neither dissolved in MeOH nor in cyclohexane. Characterization: Proton NMR was performed on a Bruker model AM-250 MHz FT-NMR. Fluorescence was carried out on a Perkin-Elmer LS-5 Fluorescence Spectrophotometer under an Ar atmosphere. The polymer concentrations of 0.10 g/ 1 in cyclohexane solutions were chosen for all the measurements.The emission spectra were recorded from 265 nm to 480 nm using an excitation wavelength of 258 nm. Each measurement was repeated 2-3 times. Wavelength reproducibility was found to be within 1 nm. SEC characterization was carried out on a Water model 6000A HPLC pump and model U6K injector were used. Detection was by a Perkin-Elmer LC-30 RI detector and LC-75 UV detector set at 268 nm. The eluting solvent was THF with 1% by volume of triethylamine in order to prevent adsorption of P2VP onto the column 36-42_ Results and Discussion Synthesis of catenated PS-b-P2VP and reaction conditions A macrocyclic PS sample of a MW of about 4,500 (45 methylene units, or 90-membered rings) was selected for the 96 catenation reaction since it is apparently large enough for the "threading" of a polystyrene chain 7 and relatively convenient to prepare in relatively large quantities 40 _ The PS macrocycle (DP=45) is expected to be highly extended in THF that is a good solvent without the occurrence of excessive entanglements or knotting. On the other hand, the ring is not very large so that the multi-threading by several chains can be avoided. The PS-cat-P2VP was synthesized through statistical threading followed by intramolecular cyclization. This process involves anionic polymerization of 2VP in the presence of PS macrocycles, producing threaded and unthreaded PS cycles and P2VP linear precursors (statistical threading). The intramolecular coupling with 1,4-DBX leads to a mixture of a PS-cat-P2VP and unthreaded PS and P2VP macrocycles and P2VP "polycondensates" 43-46_ Although the yield of catenane formation may be increased by optimizing the stoichiometry of the starting materials, the efficiency of threading from this method is expected to be low ^ . However, statistical threading is the only known method widely applicable for most polymers. The cyclic PS was mixed with the initiator (lithium naphthalide) before the polymerization of 2VP forming the linear P2VP dianion. Thus the P2VP precursor dianion may grow into the cavity of cyclic PS, possibly reducing the 97 repulsion between the PS and the P2VP thus the polymerization itself may be the driving force for the threading process. Upon formation of the PS-cat-P2VP, the dethreading process would be difficult since the PS rings are small compared with the radius of gyration of the precursor. Also the entanglement of PS rings with the P2VP backbone would reduce the freedom of macrocycles in this system. For maximum catenation yield the concentration of the PS macrocycle was kept as high as practical (~1 0-2] y i ) . The uncatenated PS and P2VP rings were recovered by fractionation using extraction (Experimental Section). The catenation reaction diagram is shown in scheme 1. Although significant quantities of macrocyclic PS and P2VP are also presented in the catenation product, the catenane is the only polymer in which both blocks PS and P2VP are present. This is due to the absence of the possibility for P2VP grafting onto PS rings. Isolation of PS-b-P2VP catenane The isolation-characterization of the catenane is a key and most challenging aspect of this research. Without it the synthesis of this novel block copolymer could be claimed but not substantiated. In fact during typical end to end cyclization the formation of some catenated polymers is 98 possible, even plausible but the isolation of such catenanes under these conditions is problematical. At present no technique we know of could unequivocally characterize catenated homopolymers. The isolation of the 50/50 PS-cat-P2VP from its blend with PS and P2VP macrocycles is possible at least in principle. Thus PS-cat-P2VP is expected to be insoluble in either methanol or alkanes due to the insolubility in these solvents of PS and P2VP respectively. In order to optimize conditions required for separation of the PS-cat-P2VP from the corresponding macrocyclic PS and P2VP, we carried out a series of experiments with solutions of mixtures of macrocyclic PS (MW=4500), P2VP (MW=4200) and macrocyclic PS- b-P2VP copolymers (MW=13,000, PS%=47%) of similar MW and composition as the catenane. Thus approximately equal amounts (1 0 mg) of each of the three polymers were dissolved in 15 ml of THF. The attempted fractionation of this blend by adding this solution to excess cyclohexane or methanol resulted in stable emulsions that were unsuitable for separation. An alternative method based on the selective extraction of a precipitated blend of the three polymers was more successful. The separation procedure is identical to that shown in scheme 2. Thus the THF solution of the three polymers was precipitated quantitatively in hexane. After 99 filtration of the precipitate and drying in vacuo overnight, the solid was extracted six times alternatively with 15 ml portions of cyclohexane and methanol. The SEC of the remaining solid indicated the absence of the PS or P2VP macrocycles and only shows the peak consistent with the macrocyclic block copolymer. A similar extraction method was also previously used by Agam and coworkers during their pioneering synthetic work on oligomeric polyrotaxanes from crown ethers and oligo (ethylene glycols) 32 _ ^he above successful procedure was then applied to the crude reaction mixture containing the PS-cat-P2VP. After the alternating extractions with cyclohexane and methanol, the extract solutions were found by SEC to contain the cyclic PS (Mp=4500, Mw/Mn=l.l) (in cyclohexane solution) and the cyclic P2VP (Mp=5500, Mw/Mn=1.15) along with the corresponding P2VP polycondensates (in methanol solution). The remaining solid product like the macrocyclic block copolymer was not soluble in either cyclohexane or methanol. Characterization of PS-b-P2VP catenane The SEC curve of the unfractionated reaction product is very similar to that of the corresponding PS or P2VP cycles 36-42 (Figure 2). The remaining uncatenated PS and P2VP cycles elute at about the same volume. The catenane if 100 present should elute together with the "polycondensation" products. The SEC chromatogram of the isolated product showed a relatively monodisperse (Mw/Mn=1.3) polymer, the peak molecular weight (Mp=1 0,0 0 0) of which approximately equals the sum of the molecular weighs of the macrocyclic PS and P2VP cycles (Figure 3). The composition of the polymer was further confirmed by - * - H NMR integration of the H- 6 resonance of P2VP relative to the aromatic resonance of PS and that of H3-H5 of P2VP (Figure 4). The calculated PS content (46%) is in excellent agreement with that calculated (45%) from the SEC chromatogram (Figure 4). The characterization data of PS-cat-P2VP are summarized in Table 1. Since PS exhibits strong excimer emission in contrast to P2VP, a fluorescence study of linear, cyclic and catenated PS-P2VP block copolymers was therefore performed in order to obtain further evidence for the catenane. For all our measurements, low concentrations (10“^M in phenyl units) solution were used so that the energy transfer from PS to P2VP would only occur intramolecularly. Fluorescence band at about 333 nm of the catenane further confirms the presence of the PS cycles in the catenane. The relative PS excimer emission intensities of the PS-cat-P2VP block copolymer are shown in Table 1, along with that of the 101 macrocyclic and linear PS-P2VP block copolymer of about the same MW and of linear PS. As shown in Table 1, the excimer emission intensity of the isolated PS-cat-P2VP is much greater than that of the cyclic P2VP indicating the presence of PS block in the catenane. Due to efficient quenching of the PS excimer by P2VP the excimer emission of the PS block in the PS-P2VP linear and macrocyclic block copolymer is significantly lower than that of the linear and cyclic PS homopolymers. Interestingly, the PS block emission of the catenane is even lower. The emission of the PS-b-P2VP macrocycle is double that of linear PS-b-P2VP of the same MW and composition. This is reasonably due to the rigidity of the macrocycles. Thus the severe restrictions on chain conformation inherent in polymer macrocycles would appear to prevent effective intramolecular quenching. Interestingly, the emission intensity of PS-cat-P2VP is found to be half that of linear PS-b-P2VP, and one fourth of the cyclic PS-b-P2VP. The very low emission intensity of the PS-cat-P2VP appears to be consistent with its unusual topology. Thus in contrast to the other two block copolymers all segments of the P2VP cycle have equal and conformationally unrestricted access to all segments of the PS cycle resulting in highly efficient intramolecular quenching of PS excimer emission. 102 Studies on these unprecedented catenated block copolymers and related macrocycles are continuing. Acknowledgments Support for this work was provided by NSF-DMR grant 9101558, Polymers Program. We wish to thank Professor A. Adamson for helpful discussions and the use of the fluorescence spectrophotometer, Dr. L. Singer for helpful discussions and Ms. R. Richards for her technical assistance with the fluorescence measurements. References (1) Dietrich-Buchecker, C.; Sauvage,J.-P. , Tetrahedron, 1990, 46 (2), 503-512. (2) Schill, G., Catenanes, Rotaxanes and Knots; Academic Press: New York, 1971 and references. (3) Wasserman, E., J. Am. Chem. Soc., 1960, 82, 4433-4434. (4) Frisch, H. L.; and Wasserman, E., J. Am. Chem. Soc., 1961, 83, 3789-3795. (5) Wu, C.; et al., Chem. Mater., 1991, 3, 569-72. (6) Gibson, H. W.; et al., Makromol. Chem., Macromol. Symp., 1992, 54/55, 519. (7) Shen, Y. X., personal communication. ( 8 ) Dietrich-Buchecker, C. 0.; Sauvage, J.-P.; Kern, J.-M, J. Am. Chem. Soc., 1984, 106, 3043-45. 103 (9) Dietrich-Buchecker, C. 0.; Sauvage, J.-P., Chem. Rev., 1987, 87, 795-810. (10) Dietrich-Buchecker, C. O.; Sauvage, J.-P., Angew. Chem., Int. Ed. Engl., 1989, 28, 189. (11) Ashton, P. R.; et al., Angew. Chem., Int. Ed. Engl., 1989, 28, 1396. (12) Cram, D. J., Angew. Chem.,Int. Ed. Eng., 1988, 27, 1009. (13) Dietrich-Buchecker, C. 0.; Sauvage, J.-P.; Kintzinger, J.-P., Tetrahedron Lett., 1983, 24, 5095-98. (14) Sauvage, J.-P. , Nouv. J. Chim., 1985, 9, 299-310. (15) Dietrich-Buchecker, C. 0.; et al., J. Am. Chem. Soc., 1990, 112, 8002-8. (16) Chambron, J.-C.; et al., J. Am. Chem. Soc., 1992, 114, 4625-31. (17) Sauvage, J.-P. and Weiss, J., J. Am. Chem. Soc., 1985, 107, 6108. (18) Dietrich-Buchecker, C. 0.; Khemiss, A. K.; and Sauvage, J.-P., J. Chem. Soc., Chem. Commum., 1986, 1376. (19) Guihem, J.; Pascard, C.; Sauvage, J.-P.; and Weiss, J., J. Am. Chem. Soc., 1988, 110, 8711. (20) Dietrich-Buchecker, C. 0.; Sauvage, J.-P.; Kern, J.- M.,J. Am. Chem. Soc., 1989, 111, 7791-7800. (21) Harrison, I. T. and Harrison, S., J. Am. Chem. Soc., 1967, 89, 5723. 104 (22) Gibson, H. W.; et al., Makromol. Chem., Macromol. Symp., 1991, 42/43, 395. (23) Lecavalier, P. R./ et al., Polym. Prepr., 1989, 30(1), 189. (24) Gibson, H. W.; et al., Polym. Prepr., 1990, 31(1), 79. (25) Lecavalier, P. R.; et al., Polym. Prepr., 1990, 31(2), 659. (26) Engen, P. T.; et al., Polym. Prepr., 1990, 31(2), 703. (27) Shen, Y. X.; et al, Macromolecules, 1992, 25, 2786. (28) Shen, Y. X. and Gibson, H. W., Macromolecules, 1992, 25(7), 2058. (29) Wu, C. ; et al., Polym. Commun., 1991, 32 (7), 204. (30) Schill, G. and Zollenkopf, H., Nachr. Chem. Techn., 1967, 79, 149. (31) Schill, G. and Zollenkopf, H., Ann. Chem., 1969, 721, 53. (32) Agam, G., Gravier, D., and Zilkha, A., J. Am. Chem. Soc., 1976, 98, 5206. (33) Wood, B. R. and Semiyen, J. A., Polymer, 1994, 35(7), 1542. (34) Rigbi, Z. and Mark, J. E., J. Polym. Sci., Polym. Phys.Ed., 1986, 24, 443-9. (35) Fontanille, M. and Sigwalt, P., Bull. Soc. Chim. Fr., 1967, 11, 4087. (36) Toreki, W. and Hogen-Esch, T.E., Polym. Preprts., 1989, 105 30(1), 129. (37) Sundararajan, J. and Hogen-Esch, T. E., Polym. Preprts., 1992, 33(1), 162. (38) Hogen-Esch, T.E., Sundararajan, J. and Toreki, W., Makromol. Chem., Macromol. Symp., 1991, 47, 23-42. (39) Gan, Y., Zoller, J. and Hogen-Esch, T. E. , Polym. Preprts., 1993, 34(1), 69. (40) Gan, Y., Zoller, J., Yin, R. and Hogen-Esch, T. E., Makromol. Chem., Macromol. Symp., 1994, 77, 93-104. (41) Toreki, W., Hogen-Esch, T.E. and Butler, G. B. , Polym. Preprts., 1987, 28(2), 343. (42) Toreki, W. and Hogen-Esch, T.E., Polym. Preprts., 1988, 29(2), 416. (43) Rempp, P.; Hild, G. and Strazielle, C.Eur. Polym. J. 1983, 19(8), 721. (44) Roovers, J. Macromolecules, 1985, 18, 1359. (45) Hocker, H. and Geiser, D. Macromolecules, 1980, 13(3), 653. (46) Deffieux, A., private communication. (47) Shen, Y. X. and Gibson, H. W., Macromolecules, 1992, 25(7), 2058. 106 TABLE 1: Characterization Results of Isobaric Linear (L) and Macrocyclic (C) Polystyrene (PS), Poly(2-vinylpyridine) (P2VP), PS-b-P2VP and PS-cat-P2VP. Sample Mpa Mw/Mn*3 lemc PS %d Cyclic PS 4, 500 1 . 1 0 2 0 0 Linear PS 4, 500 1 . 1 0 1 0 0 Cyclic PS-b-P2VP 5, 700 1.30 16 55% Linear PS-b-P2VP 5, 700 1.14 8 55% PS-cat-P2VP 10,300 1.30 4 4 6%e Cyclic P2VP 5, 500 1.15 0 . 10 Linear P2VP 5, 500 1.15 0 . 1 0 a. SEC peak molecular weight (MW), b. Molecular weight distribution determined by SEC, c. Relative fluorescence intensity at 333 nm (excimer emission), d. PS composition (mole %) in the block copolymers determined by NMR, e. Calculated value is 45% . 107 108 Cyclic PS (~10'2 M) (Mn~5,000) P2VP2\2Li+ (~10'2 M) (Mn~5,000) (Mn~10,000) EX' .EX Scheme 1: Cyclization of P2VP2", 2Li+ with EX2 in the Presence of Macrocyclic Polystyrene Schem e 2: Isolation o f catenated PS-b-P 2V P copolym er. Solid (PS) (P2VP) (PS-b-P2VP Catenane) 1) CH3OH/Stir 2) Settle for 24 h 3) Centrifuge Solid Solution (PS) (PS-b-P2VP Catenane) (P2VP) 1) Cyclohexane 2) Stir 3) Centrifuge Solution Solid Repeat 3 times Solid (PS) (PS-b-P2VP Catenane) (Pure PS-b-P2VP Calenane) 109 I L 1 ' I 1 I I I I I I I ! t 9 10 1 1 1 2 1 3 14 15 16 17 1 8 19 20 2 1 2 2 Elution Volume (ml) Figure 2: Size Exclusion Chromatogram o f Crude Catenated PS-b-P2V P Copolymer. no — |---- 1 -----1 -----1 -----1 ---- 1 ----- 1 -----1 j 1 -----1 ---- 1 -----[ ---- 1 -----1 -----1 -----j -----! -----1 -----1 ---- 1 -----1 -----1 -----1 11 13 15 17 19 21 23 Elution V olum e (m l) Figure 3: Size Exclusion Chromatogram of Purified Catenated PS-b-P2VP Copolymer. i l l - | - C H 2— C H -]J-C H 2— C M -|— 3 r ^ N £ j ® 10 s ppm Figure 4: NM R Spectrum o f Catenated PS-b-P2VP Copolymer. 112 Chapter 5 ANIONIC POLYMERIZATIONS OF TRIMETHYLVINYLSILANE AND PHENYLDIMETHYLVINYLSILANE TOWARDS THE SYNTHESIS OF WELL-DEFINED POLYFLUORODIMETHYLVINYLSILANE AND POLYVINYLALCOHOL Abstract The anionic polymerizations of trimethylvinylsilane (TMVS) in toluene at -20 °C and phenyldimethyl-vinylsilane (PDMVS) at 0-20 °C proceed in high yields (>95%) giving narrow molecular weight distribution (MWD<_1.15) polymers. The reaction of PPDMVS with HBF4 .Et2 0 in toluene at 60 °C resulted in the formation of polyfluorodimethylvinylsilane (PFDMVS) as shown by and • ' • ^ F NMR. Introduction Polytrimethylvinylsilane (PTMVS) and polyphenyldimethylvinylsilane (PPDMVS), their copolymers and polymer blends, have been demonstrated to possess excellent transparency, flexibility, permeability, and heat resistance, which have been widely applied and studied as materials for gas-separation membranes optical fibers 113 4, contact lenses 5,6^ coating of waterproofing fabrics ^, optical disks photoresists 9, etc. PTMVS and PPDMVS might also be useful as intermediates in the synthesis of monodisperse polyethylene, polyvinylalcohol, polyvinylether, polyacetylene, polyvinylhalide, polynitroethylene and polyvinylamine (Scheme 1). Furthermore, PTMVS or PPDMVS containing block copolymers might be converted to a series of novel block copolymers. However, such reactions of PTMVS and PPDMVS have not been explored. Hence, the synthesis of narrow MWD polymers and block copolymers of PTMVS and PPDMVS would be of interest . Schem e 1: U nexplored reactions o f PPD M V S. Silicon stabilizes adjacent carbanions by overlap with its vacant d orbitals 1®. Vinylsilanes such as TMVS and PDMVS are therefore possible candidates for "living" anionic polymerizations. The anionic polymerization of TMVS and PDMVS initiated with alkyl lithiums was first reported by N. S. Nametkin et al. H. From 1965 to 1982, they synthesized oligomers ^ , homopolymers H, block- 13 ancj graft- copolymers 14-16 under vacuum or argon atmosphere at temperatures above 0 °C up to 70 °C (mostly 35-40 °C) for 18-300 hours, by using butyllithiums, ethyllithium and metallic lithium as initiators and benzene, toluene, cyclohexane and heptane as solvents (they also tried DMF as promotor). These polymers were characterized by 1h and l^c NMR 17, size exclusion chromotography (SEC) 18,19^ DSC (180 °C) 20 ancj viscosity 21. Since the polymerization rates of PTMVS and PPDMVS are slow at low temperatures, higher reaction temperatures (25-60 °C) were chosen in order to obtain quantitative yields (96%). However, a MWD of 1.3 was measured later when GPC technique was well developed. This broader polydispersity of the resulting polymer was believed due to chain transfer and self-termination of the "living" chains, this occurrence of termination reactions at higher temperatures (25-60 °C) resulted from hydride elimination that causing broadened MWD 22,23. Later studies by Rickie (Dow Chemical Co.) confirmed such side reactions and their effects on MWD and yield 24,25^ especially for high 115 molecular weight PTMVS. He also found that such side reactions were enhanced by the addition of ethers 25 _ We now wish to report the synthesis of narrow MWD PTMVS and PPDMVS by anionic polymerization of the corresponding vinyl monomers, and their attempted transformation into other polymers (Scheme 2). In order to minimize the termination reactions and obtain better molecular weight distribution samples, we tried various solvent systems at lower temperatures (below 0 °C) for relatively long reaction times. High vacuum (10-6 torr) system and ultra-pure solvents and monomers were employed. Schem e 2: Synthesis o f PPD M V S, PFD M V S and PVA. BuLi / T oluene 30% H 202 /K 0 H H BF4-OEt2 / C H C I3 60 °C for 3 hours Bu F 116 Protodesilylation and oxidation of various of organosilicon compounds had been recently developed by Tamao 26-28f Fleming 29,30 ancj Nishiyama 31f a variety of alcohols could be synthesized via oxidative cleavage of the carbon- silicon bond, especially the "unactivated" alkyl-silicon bonds. In this case, the phenyldimethylsilyl group was used as a masked hydroxyl group. In these procedures, phenyl- silicon bonds are cleaved by acids (tetrafluoroboric acid diethyl ether complex) prior to the oxidation (Scheme 1). Protodesilylation of the phenyl group converted each of the silanes into the fluorosilanes, which further reacted with excess of m-chloroperbenzoic acid (MCPBA) or 30% hydrogen peroxide to give the corresponding alcohols . However, it wasn't possible to remove trimethylsilyl groups except with the most vigorous of reagents, such as concentrated sulphuric acid, and even then it was probable that the methyl groups would be cleaved off before the entire trimethylsilyl groups ^9. The reason that the phenyldimethylsilyl but not the trimethylsilyl group could be easily converted to the hydroxy group may be due to induction by the aromatic substituent. Although the synthesis of alcohol via various of organosilicon compounds had been successfully demonstrated 26-30^ none of these involved polymers. A successful transformation of a compound with similar structure similar 117 to PPDMVS was reported by Tamao and his coworkers in 1987 (Scheme 3). Thus, the synthesis of polyvinylalcohol by this route might thus be possible. (1) HBF4 OEt2 / CHC13 / 0 °C / 15 min (2) 30% H20 2 / 15% KOH / McOH / TH F/5 0 °C / 4h SiMe?Ph Scheme 3: Synthesis o f a alcohol via oxidative cleavage of the C-Si bond reported by Tamao. Experimental Materials: TMVS was purchased from Huls America Inc.. PDMVS was synthesized by adding vinyldimethylchlorosilane (Huls America Inc.) dropwise to a phenylmagnesium bromide/THF solution at -20°C followed by reflux for 2.5 hours. Stirring and adding speed were controlled in order to maintain a constant reaction temperature. The resulting milky-white solution was fractionally distilled and over 70% isolated yield of product (b.p.= 75-79 °C/10 mmHg) was obtained. The PDMVS was characterized by NMR and GC-MS. Both TMVS and PDMVS were distilled from CaH2 twice and from a potassium mirror under high vacuum in order to eliminate any moisture and other impurities. The monomers were stored in evacuated ampules at -20 °C in a freezer. Toluene was 118 purified by stirring over fresh K-Na alloy under vacuum overnight. Tetrafluoroboric acid-diethyl ether complex (85%) (Aldrich) was used as received. PPDMVS was purified by reprecipitation in THF using MeOH as the non-solvent. Polymerizations: Anionic polymerization was carried out by initiation of purified TMVS and PDMVS with n-, sec-, or tert-butyllithium in toluene at -70 °C to 20 °C. Reactions took 12 hours to several days. The colorless foaming solution was finally terminated with methanol and precipitated in cold MeOH. Polystyrene-b-TMVS block copolymers were synthesized via a similar procedure. The polymers and copolymers were characterized by SEC and by and 13C NMR (250 MHz). Protodesilylation Reactions: Plastic vessels were utilized during the protodesilylation to avoid the reaction between HBF4 .OEt2 and glass. In order to obtain reproducible and quantitative conversion, The low MW PPDMVS samples with identical degree of polymerization (DP=4 6) were used in all cases. The reaction of PPDMVS with HBF4 .Et2 0 was carried out in a polypropylene vessel. Purified PPDMVS (DP=46) was dissolved in chloroform, and HBF4 .Et2 0 was then added to the PPDMVS solution (1.0 gram in 15 ml) under argon. The reaction mixture was stirred at 60 °C for 3 119 hours. During this protodesilylation process, the reaction temperature was carefully monitored in order to prevent boiling of the solvent and swelling of the plastic vessel. After evaporation of the volatile materials, the resulting fluorosilane polymer was purified by reprecipitation in cold methanol and characterized by and ^NMR and by SEC. Oxidation Reactions: The PFDMVS obtained (0.57 g) was dissolved in 30 ml THF and transferred to a glass flask. A solution of 1.45 g KOH in 20 ml MeOH was introduced into the reaction vessel followed by the dropwise addition of 14 ml 30% H2 O 2 aqueous solution. The reaction mixture was refluxed (50 °C) for 4-19 hours. The resulting polymer was not water-soluble indicating that conversion was incomplete. The use of dioxane instead of THF/MeOH solvents following the same reaction procedure was also unsuccessful. Replacing H2O2 by three equivalents of m-chloroperbenzoic acid (MCPBA) with an excess of triethylamine in THF/MeOH and refluxing for 19 hours likewise did not give the desired result. 120 Results and Discussion Polymerization The polymerization of PDMVS at a concentration of 1.3 mol/liter was carried out in toluene at various temperatures by using t-BuLi as initiator (Table 1). At -25 °C, the polymerization was too slow to produce polymers and most of the monomer was recovered after 24 hours. The bulk polymerization at -25 °C for 12 hours was more successful giving PDMVS in quantitative yield. The MWD was not very narrow however (Mw/Mn=1.4). The polymerization at 0 °C, produced only oligomers (MW=400-1000) after 23 hours. At -20 °C in the presence of 10% (v/v) THF in toluene, after 14 hours, resulted likewise in low yields (3%) of oligomers (MW=400-1000). It has been shown previously that addition of a trace amount of THF increased the propagation rate but chain termination as well 25. on the other hand, at higher temperature (40°C), PPDMVS was obtained in high yield (80%) after 2 hours. SEC analysis showed a broadened (D=1.4 6) MWD which may due to increased termination at this temperature. A better MWD (=1.23) PPDMVS was produced in good yield (75 %) by using lower reaction temperature (20°C) and longer reaction time (24 hours). 121 The best results were obtained when PDMVS was polymerized at 0 °C for 2 days followed by raising the temperature to 20 °C for another 3 days. In this case, the MWD of the polymer was 1.1 (MW=7,400) and the yield was nearly quantitative (90-95%) (Table 1). For the case of the polymerization of TMVS initiated with t-BuLi in toluene at -78 °C, no polymerization occurred even after 4 days. At -10 °C after one hour, only low MW oligomers (MW=2600) were formed in low yields (6%). However, at a higher temperature (0-2 °C), TMVS could be polymerized quantitatively in 2 0 hours producing however PTMVS with a broad MWD (=1.56). Better results were obtained at -20 °C in toluene for four days which produced high yield (>80%) and narrow MWD (=1.1) (MW=10,000). The use of n- or sec-butyllithium instead of t-BuLi gave similar results. For the case of "living" polystyrene as initiator, the orange color of the PS anion instantly turned yellowish upon addition of TMVS indicating rapid initiation. The block lengths of the polystyrene and PTMVS could be obtained by controlling the respective monomer- initiator ratios. The PS-b-PTMVS copolymers were synthesized with a PS content ranging from 57% to 83% with MW is ranging from 6,800 to 22,000. Their compositions were measured by SEC and NMR (Table 2). 122 Cyclohexane containing 2% (v/v) THF was also found to be a good solvent system for the polymerizations of TMVS and for the synthesis of PS-b-TMVS. However, the high freezing point (6.5 °C) of cyclohexane limited its use to ambient temperature. Protodesilylation The protodesilylation of PPDMVS (6.7 wt%) in CHCI3 carried out by using HBF4.0Et2 (85%) at 0 °C for 15 minutes similar to that used for a organosilicon compounds 27 was unsuccessful, and the starting material was recovered quantitatively. At 20 °C after 3 hours, a white powder was obtained after removal of excess reactants and solvent. The polymer was redissolved in THF and purified by reprecipitation in cold methanol. Proton NMR integration showed 50% substitution of phenyl by fluorine. At 40 °C, after 3.5 hours, 90% substitution was obtained, and essentially quantitative substitution was obtained at 60 °C for 3 hours. The ^H NMR of the product is shown in Figure 1. The spectrum exhibited the documented structure of PFDMVS (5=0.1, doublet, 6H, Si(CH3 >2 ; 5=1.1, multiplet, 3H, - CH2CH-) and phenyl absorptions were now absent. Fluorine-19 NMR confirmed the structure of the dimethyl fluorosilyl group (Figure 2), showing a strong absorption at -158.6 ppm (silyfluorine). A small peak at -161 ppm 123 indicated that some of the CH3 groups were substituted by fluorine. Fluorine substitution tends to increase the reactivity of C-Si bonds (reactivity: -SiF3>-SiF2R>-SiFR2 ) 29-30. Oxidation reactions The oxidative cleavage of the PFDMVS carbon-silicon bonds was surprisingly difficult compared to that of low MW organic fluorosilanes that are easily oxidized by various oxidants 27,28,30. Thus the procedures demonstrated by Tamao and his coworkers in 1987 27 (Scheme 3) were completely unsuccessful. Even after refluxing PFDMVS with 14 ml 30% H2O2 in a MeOH/THF containing 10% KOH at 50°C for 19 hours, no oxidation of the starting polymer could be detected. The use of solvents such as ethylene glycol diethyl ether, ethoxyethanol and dioxane capable of dissolving the reactants, the use of higher temperatures (87 °C) and longer reaction times (19-23 hours) did not result in a water-soluble polymer and 1H NMR showed that most of the PFDMVS remained unreacted. One of the probable reasons for this is the poor solubility of the PFDMVS in this water containing medium. Although PFDMVS dissolved in a MeOH/THF (20/30 v/v) mixture very well, the polymer would precipitate out upon addition of 30% aqueous H2O2 even at refluxing temperature. Increasing the organic solvent content and 124 higher temperatures failed to produce satisfactory results. The use of organic oxidants such as m-chloroperbenzoic acid (MCPBA) in the presence of excess triethylamine at 50 °C in THF/MeOH for 19 hours, gave a polymer showing a small hydroxyl peak by ^H NMR but most of the fluorodimethylsilyl groups were still present. The reason (s) for this unsuccessful reaction are not obvious. The lack of solubility and the numerous bulky pendent groups apparently prevented the oxidation reaction. A similar lack of reactivity is observed in many reactions carried out on macromolecules. Conceivably the use of stronger oxidants and more vigorous reaction conditions will allow the quantitative oxidation of PFDMVS to polyvinylalcohol. Acknowledgements; Support was provided by NSF-DMR-Polymer Program. We wish to thank Drs. G. A. Olah, W. P. Weber and especially Dr. S. Prakash for helpful discussions. The technical assistance of Dr. Q. Liao is also much appreciated. References: 1. Koichi, 0.; et al., Eur. Pat. Appl., 26 Oct., 1983, EP 92417 Al, 25. 2. Alimova, L. Ya.; et al., Inzh.-Fiz. Zh., 1985, 48(1), 125 96-100. 3. Fujii, Y.; et al., Jpn. Kokai Tokkyo Koho, 15 May 1990, Heisei, JP 02126927 A2, 5. 4. Takamizawa, M.; et al., Jpn. Kokai Tokkyo Koho, 10 Feb. 1987, Showa, JP 62031808 A2, 3. 5. Takamizawa, M.; et al., Jpn. Kokai Tokkyo Koho, 9 Nov.1985, Showa, JP 60225115 A2, 3. 6. Takamizawa, M.; et al., Jpn. Kokai Tokkyo Koho, 18 Jul.1988, Showa, JP 63174014 A2, 6. 7. Sakamoto, I.; et al., Jpn. Kokai Tokkyo Koho, 17 Mayl986, Showa, JP 61098749 A2, 7. 8. Nagura, S.; et al., Jpn. Kokai Tokkyo Koho, 16 Sep.1988, Showa, JP 63223012 A2, 3. 9. Kushibiki, N.; et al., Jpn. Kokai Tokkyo Koho, 27 Apr. 1989, Heisei, JP 01110513 A2, 4. 10. Nametkin. N. S.; et al., Dokl. Akad. Nauk SSSR, 1981. 258 (4), 887-9. 11. Nametkin, N. S.; et al., Vysokomolekin Soedin, 1965, 7(1), 184. 12. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1980, 252(1), 123-5. 13. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1978, 238 (6) . 1358-60 . 14. Plate, N. A.; et al., Vysokomolekin Soedin, Ser. A, 1981, 23(3), 640-50. 126 15. 16. 17. 18. 19. 20 . 21 . 22 . 23. 24 . 25. 26. 27. 28. Nametkin, N. S.; et al., Vysokomolekin Soedin, Ser. A, 1969, 11(9), 2067-72. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1969, 185(1), 97-99. BulaiN, A. Kh.; et al., Vysokomolekin Soedin, Ser. A, 1981, 23(7), 1526-32. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1969, 186(6), 1336-7. Nametkin, N. S.; et al., Vysokomolekin Soedin, Ser. A, 1975, 17(5), 973-8. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1978, 239(4), 886-8. Nametkin, N. S.; et al., Jzv. Akad. Nauk SSSR, 1970, Ser. Khim.(2), 283-9. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1980, 251(4), 878-82. Nametkin, N. S.; et al., Dokl. Akad. Nauk SSSR, 1974, 215(4), 861. Rickie, G. K., J. Macromol. Sci., 1987, A24(l), 93-104. Rickie, G. K., J. Macromol. Sci.-Chem., 1986, A23(ll), 1287-97. Tamao, K., et al., Tetrahedron, 1983, 39, 983. Tamao, K., et al., Chemistry Letters, 1987, 171-4. Tamao, K., et al., J. of Organometallic Chem., 1984, 269, C37-39. 127 29. Fleming, I., et al., J. Chem. Soc., Chem. Commun., 1984, 29. 30. Fleming, I., et al., J. Chem. Soc., Chem. Commun., 1986, 1198-1201. 31. Nishiyama, H., et al. The 49th Spring Meeting of the Chemical Society of Japan, Tokyo, April, 1984, 1527-8. 128 TABLE 1: Anionic polymerization of phenyldimethylvinylsilane (PDMVS) in toluene initiated with t-BuLi 8. Entry Temp. (°C) Time (hours) MWa MWDb Yieldc (%) 1 -25 2 4 (no polymer) 0 2d -25 1 2 14,200 1.40 100 3 0 23 (Oligomers) - 4 e -2 0 1 4 4 0 0 -1 ,0 0 0 1.20 3 5 40 2 3,200 1.46 80 6 20 2 4 4,200 1.23 75 7 f 0 -2 0 4 4 -7 2 7,400 1.12 >94 a. SEC peak molecular weight (MW), b. MWD = Mw/Mn, c. Isolated yield, d. Bulk polymerization, e. 10% (v/v) of THF was added to toluene as the co-solvent, f. Two days at 0°C and three days at 20 °C, g. Monomer concentration is 1.3 M. 129 TABLE 2: Anionic synthesis of polytrimethylvinylsilane (PTMVS) and polystyrene-b-PTMVS (PS-b-PTMVS) copolymers in toluene by using t-BuLi as initiator. No. Temp. Time M W a MWDb Yield (%) PS (%)d 1 -78 o c 96 hr - - 0 - 2 -10 o c 1 hr 2,600 1.10 6 - 3 0-2 o c 20 hr 23,400 1.56 90 - 4 -20 o c 96 hr 10,000 1.10 82 - 5 e 20 o c 83 hr 6,800 1.10 56 83% 6 -20 °C 96 hr 22,000 1.18 80 57% a. SEC peak molecular weight (MW), b. MWD = Mw/Mn, c. Isolated yield of PTMVS, d. Polystyrene content in cases of block copolymers which were measured by both NMR and SEC, e. In cyclohexane containing 2% (v/v) THF. 130 (B) (C) T T T T T T T T T I T T T T ppm Figure 1: NMR spectrum of (a) PPDMVS, (b) product after protodesily1ation at 40°C and (c) product after protodesilylation at 60°C (PFDMVS). 131 . 004 ■200 PPM Figure 2: Fluorine-19 NMR spectrum of PFDMVS (CFCI3 as the re fe r e n c e , 5=0.00). CJ N >

Linked assets