AFML-TR-71-6

THERMAL DEGRADATION OF COPOLYMERS

N. Grassie, B. J. D. Torrance, and J. D. Fortune

University of GlasgowGlasgow, Scotland

TECHNICAL REPORT AFML-TR-71-6

January 1971

This document has been approved for public release and sale;its distribution is unlimited.

Air Force Materials LaboratoryAir Force Systems Command

Wright-Patterson Air Force Base, Ohio

AFLC-WPAFB-MAY 71 93

B3EST AVAILABLE COPY

NOTICE

When Government drawings, specifications, or other data are used for any purpose

other than in connection with a definitely related Government procurement operation,

the United States Government thereby incurs no responsibility nor any obligation

whatsoever; and the fact that the government may have formulated, furnished, or in

any way supplied the said drawings, specifications, or other data, is not to be regarded

by implication or otherwise as in any manner licensing the holder or any other person

or corporation, or conveying any rights or permission to manufacture, use, or sell any

patented invention that may in any way be related thereto.

Copies of this report should not be returned unless return is required by security

considerations, contractual obligations, or notice on a specific document.

AFML-TR-71-6

THERMAL DEGRADATION OF COPOLYMERS

N. Grassie, B. J. D. Torrance, and J. D. Fortune

University of GlasgowGlasgow, Scotland

This document has been approved for public release and sale;its distribution is unlimited.

FOREWORD

This report was prepared by the University of Glasgow, Glasgow,Scotland, on Air Force Contract No. AF 61(052)-883. The contractwas initiated under Project 7342, "Fundamental Research on Macro-molecular Materials and Lubrication," Task 734203, "FundamentalPrinciples Determining the Behavior of Macromolecules," and wasadministered by the Air Force Materials Laboratory, Wright-PattersonAir Force Base, Ohio, with Dr. Ivan J. Goldfarb (AFML/LNP) asProject Scientist.

This report covers work from 1 October 1965 to 31 January 1970.

This technical report has been reviewed and is approved.

R. L. VAN DEUSENActing Chief, Polymer BranchNonmetallic Materials DivisionAir Force Materials Laboratory

ii

ii

ABSTRACT

This report is principally concerned with thethermal degradation of two copolymer systems, namelymethyl methacrylate/methyl acrylate and methylmethacrylate/butyl acrylate. The photo-degradationof the methyl methacrylate/methyl acrylate systemwas also studied insofar as it is relevant to thethermal reaction. In view of the similarity ofthe monomers, reliable values of reactivity ratioswere not available. An accurate method, making useof nuclear magnetic resonance spectroscopy, wasdevised and applied to these and other acrylate/methacrylate systems.

For both systems a series of copolymerscovering the whole composition range was synthesised.Degradations were carried out under vacuum eitherin a dynamic molecular still or using a new tech-nique developed in these laboratories. Thermalmethods of analysis, such as thermogravimetricanalysis (TGA) and thermal volatilisation analysis(TVA), showed that the copolymers became more stableto thermal breakdown as the acrylate content wasincreased. These techniques allowed a suitabletemperature range to. be chosen in which to studythe reactions isothermally. The gaseous degradationproducts, liquid products, chain fragments and residuewere each examined separately, using, among othertechniques, infra-red spectroscopy, gas-liquidchromatography, mass spectrometry and combined gaschromatography - mass spectrometry. The complexnature of the pyrolysis of these copolymer systemsis reflected by the variety of products obtained.

The main gaseous products were found to becarbon dioxide and smaller amounts of hydrogen fromthe methyl acrylate copolymer and carbon dioxide andbut-l-ene from the butyl acrylate copolymer. Themost important liquid products are methyl methacrylatefrom both systems and n-butanol from the butylacrylate copolymers with high butyl acrylate contents.Quantitative measurements enabled the build-up ofthese products to be followed as degradation proceeds,

S• iii

and mass balance tables were drawn up for eachcopolymer studied. The large chain fragmentswere only briefly examined. Molecular weightmeasurements on the residue indicated that break-down by random scission processes becomes moreimportant relative to breakdown by depolymerizationprocesses as the acrylate content is increased.Degradation schemes and mechanisms are postulatedto account for the formation of all of the im-portant products, although no really satisfactoryroute for alcohol evolution has been found.

iv

TABLE OF CONTENTS

FIGURES ix

TABLES xvi

CHAPTER 1. INTRODUCTION 1

CHAPTER 2. REACTIVITY RATIOS FOR THE COPOLYMERIZATIONOF ACRYLATES AND NETHACRYLP TES BY NUCLEARMAGNETIC RESONANCE, SPECTROSCOPY 3

INTRODUCTION 3EXPERIMENTAL 3

Monomer Purification 3Polymer ization 4Copolymer Analysis 4

RESULTS AND DISCUSSION 4

CHAPTER 3. THERMAL DEGRADATION OF COPOLYMERS OF METHYLMETHACRYLATE AN]) METHYL ACRYLATE. A. PRODUCTSAND GENERAL CHARACTERISTICS OF THE REACTION 11

INTRODUCTI1ON 11EXPERI M[ NT AL 11

Preparation of Copolymers 11Degradation Techniques 12Molecular Weights 14Spectroscopic Measurements 14

RESULTS AN]) DISCUSSION 14

Thermal Volatilization Analysis 14Changes in Molecular Weight 17Volatile Products of Analysis 19Unsaturation in the Residual Copolymer 19Absence of Methanol among the ReactionProducts 23

Production of Methyl Acrylate andChain Fragments 24

Production of Carbon Dioxide 24

CONCLUS I ONS 2 5

v

CHAPTER 4. THIERMAL DEGRADATION OF COPOLYMERS OF METHYLMETHACRYLATE AND METHYL ACRYLATE. B. CHAINSCISSION AND THlE MECHANISM OF TtlE REACTION .26

INTRODUCTION 26EXPERIMENTAL 26RESULTS AND DISCUSSION 26

Chain Scission and Carbon DioxideProduc ti on 26

Volatilization, Chain Scission, andZip Length 28

Chain Scission and CopolymerComposition 32

Activation Energy for Chain Scission 33Chain Scission and the Production ofPermanent Gases 33

CONCLUS IONS 41

CHAPTER 5. PHOTODEGRADATION OF COPOLYMERS OF METHYLMETHACRYLATE AND METHYL ACRYLATE ATELEVATED TE14PERATURES 43

INTRODUCTION 43EXPER IMENTAL 43

Copolymers 43Molecular Weights 43Photodegradation Techniques 44

RESULTS 44

Influence of Temperature on Rate ofVolatilization 44

Molecular Weight Changes 47Volatile Products of Degradation 49Rates of Volatilization 49Chain Scission and the Productionof Carbon Dioxide 49

Chain Scission and Volatilization 51Chain Scission and CopolymerComposition 51

DISCUSSION 54

vi

CHAPTER 6. THERMAL DEGRADATION OF COPOLYM1'hTS OF METHYLMETHACRYLATE AND n-BUTYL ACRYLATE. A.EXPER IMENTAL METHODS 57

SYNTHESIS OF COPOLYMERS 57DEGRADATION APPARATUS 58PRODUCT ANALYSIS 60PRESSURE MEASUREMENT 61MOLECULAR WEIGHT MEASUREMENTS 61TIIEWI4AL METHODS OF ANALYSIS 61GAS-LIQUID CHIROMATOGRAPHY 64MASS SPECTROMETRY 65SPEC TROME TRI C MEASUREMIENTS 65SOL-GEL ANALYSIS 66

CHAPTER 7. THERMAL DEGRADATION OF COPOLYMERS OF METIHYLMETHACRYLATE AND n-BUTYL ACRYLATE. B.THERMAL ANAL YS IS 68

INTRODUCTION 68THERMAL VOLATILIZATION ANALYSIS 68

Collection of T.V.A. Data 68Interpretation of T.V.A. Data 79

THERMOGRAVIMETRIC ANALYSIS 80

Determination of Kinetic Parametersfrom T.G.A. 80

Dynamic T.G.A. 82Isothermal T.G.A. 84

DISCUSSION OF RESULTS 90

CHAPTER 8. THERMAL DEGRADATION OF COPOLYMERS OF METHYLMETHIACRYLATE AND n-BUTYL ACRYLATE. C.IDENTIFICATION OF THE PRODUCTS OF REACTION 92

INTRODUCTI ON 92ANALYSIS OF PRODUCT GASES 92

Gas Chromatography 92Mass Spectrometry 92Infra-red 98

ANALYSIS OF LIQUID VOLATILES 101

vii

Gas Liquid Chromatography 101Combined Gas Chromatography-MassSpectrometry 101

CHAPTER 9. TH•MMAL DEGRADATION OF COPOLYMERS OF NETHYLMETHACRYLATE AND n-BUTYL ACRYLATE. D.QUANTITATIVE ANALYSIS OF LIQUID AND GASEOUSPRODUCTS 115

INTRODUCTI ON 115ANALYSIS OF THE GASEOUS DEGRADATION

PRODUCTS 115

Analytical Techniques 115

Gas Chromatography 115Infra-red Spectroscopy 118Treatment of Results 118

DISCUSSION OF RESULTS 138

CALCULATION OF MASS BALANCE DATA 145

ANALYSIS OF THE LIQUII) DEGRADATIONPRODUCTS 145

Analytical Techniques 145Treatment of Results 164Discussion of Results 172

CHAPTER 10. THERMAL DEGRADATION OF COPOLYMERS OF METIIYLMETHACRYLATE AND n-BUTYL ACRYLATE. E. THERESIDUE AND CHAIN FRAGMENTS 177

INTRODUCTI ON 177INVESTIGATION OF THE RESIDUE 177

Spectroscopic Techniques 177

Infra-red 177U.V.-Visible 180N. 11. R. 180

Elemental Analysis 183Molecular Weight Measurements 183Sol-Gel Analysis 185

viii

Effect of Copolymer Composition onGel Formation 189

Effect of Molecular Weight on GelFormation 189

Effect of Degradation Temperatureon Gel Formation 189

Degradation at a Lower Temperature 192

Carbon Dioxide Production andChain Scission 192

Mass Spectrometry 193

INVESTIGATION OF CHAIN FRAGMENTS 193

CHAPTER 11. SURVEY OF RESULTS AND) GENERAL DISCUSSION 198

INTRODUCTION 198METHYL METHAC RYLATE/METHYL ACRYLATE

COPOLYMERS 198METHYL METHIACRYLATE/BUTYL ACRYLATE

COPOLYMERS 202

Production of Alcohol 202Production of But-l-ene 209Production of Carbon Dioxide 209PrQduction of n-Butyl Methacrylate 210Production of n-Butyl Acrylate 211Other Products 211

SUMIIARY 212

REFEREMCES 213

ix

FIGURES

1. Nuclear magnetic resonance spectrum of 1-4 1NHA-PrAcopolymer. 5

2. Nuclear magnetic resonance spectrum of 1-5 HNA-MAcopolymer. 7

3. r, versus r2 plots for (a) IPIA-•A; (b) MivIA-EA;(c) MMA-PrA; (d) I&IA-BuA; (e) IMIA-EMA. 9

4. Degradation apparatus. 13

5. Thermal volatilization analysis thermograms ofpoly(methyl methacrylate) and methyl methacrylate-methyl acrylate copolymers. 15

6.1 Molecular weight changes during degradation of 26/1methyl methacrylate-methyl acrylate copolymer. 18

7. Ultraviolet absorption spectra of 2/i methyl meth-acrylate-methyl acrylate copolymer in chloroformsolution: (A) undegraded; (B) after 50%.volatilization 20

8. Infrared absorption spectra of films of 2/1,methyl methacrylate-methyl acrylate copolymer:(-) undegraded; (---) after 50% volatilization. 21

9. Production of carbon dioxide during degradation of7"7/1 methyl methacrylate-methyl acrylate copolymer. 27

10. Relationship between carbon dioxide production andchain scission during degradation of methyl meth-acrylate-methyl acrylate copolymers: (o) 112/1;(e) 26/1; (o) 7-7/1; (m) 2/1. .29

11. Relationship between volatilization and chainscissions during degradation of 26/1 methyl meth-acrylate-methyl acrylate copolymers: (o) 282°C;(.) 294 0 C; (b) 326'C. 31

12. Dependence of rate of production of carbon dioxideon size of polymer sample for methyl methacrylate-methyl acrylate copolymers. 34

x

13. Relationship betweer, rate of chain scission andmethyl acrylate content for methyl methacrylate-methyl acrylate copolymers. 35

14. Arrhenius plot of chain scissions duringdegradation of methyl methacrylate-methylacrylate copolymers. 36

15. Extent of volatilization in 30 min at varioustemperatures: (o) poly(methyl methacrylate);(0) methyl methacrylate-methyl acrylate copolymer(26/1); (A) methyl methacrylate-methyl acrylatecopolymer (7-7/1). 46

16. Change in molecular weight with volatilizationfor photodegradation of methyl methacrylate-methyl acrylate copolumers at 170OC: (0) 112/1;(W) 26/1; (A) 7.7/1; (A) 2/1. 48

17. Volatilization-time curves for the photodegradationof poly(methyl methacrylate) and methyl methacrylate-methyl acrylate copolymers at 170'C: (o) PTZIA;(') MMh4A/MA 112/1; (to,4') MMA/MA,26/I; (4,) NN.4A/MA,7.7/1; (4A) ,rA/MA,2/1. 50

18. Relationship between chain scissions and volatil-ization in the photodegradation of methyl meth-acrylate-meth1l acrlate copolymers at 1700C:(W) 26/1; (A) 7".7/1; (A ) 2/1. 52

19. Time dependence of chain scission in the photo-degradation of methyl methacrylate-methylacrylate copolymers at 1700C: (Pi) 112/1;(a) 26/1; (•) 7-7/1; (A) 2/1. 53

20. The sealed tube technique. 59

21. Molecular weight plots. 62

22. Schematic diagram of the differential condensationT.V.A. (D.C.T.V.A.) 63

23. D.C.T.V.A. of a poly(methyl methacrylate)standard 69

24. D.C.T.V.A. of a 0-4 mole percent n-butyl acrylatecopolymer 70

xi

250 D.C.T.V.A. of a 3"9 mole percent n-butyl acrylatecopolymer. 71

26. D.C.T.V.A. of a 16-3 mole percent n-butyl-acrylatecopolymer. 72

27. D.C.T.V.A. of a 50.0 mole percent n-butyl acrylatecopolymer. 73

28. D.C.T.V.A. of a 52.4 mole percent n-butyl acrylatecopolymer. 74

29. D.C.T.V.A. of an 82.2 mole percent n-butyl acrylatecopolymer. 75

30. D.C.T.V.A. of a 93-4 mole percent n-butyl acrylatecopolymer. 76

31. Programmed (5 0 C/min.) T.G.A. traces of a series ofcopolymers. The figures in brackets give copolymercomposition as mole percent n-butyl acrylate. 81

32. First order plot for a 3-9 mole percent n-butylacrylate copolymer, 83

33. Activation energy plot from initial rate data fora poly(methyl methacrylate) standard. 85

34. Activation energy plot from initial rate data fora 0-4 mole percent n-butyl acrylate copolymer. 86

35. Activation energy plot from initial rate data fora 3-9 mole percent n-butyl acrylate copolymer. 87

36. Activation energy plot from initial rate data fora 16-3 mole percent n-butyl acrylate copolymer. 88

37. Activation energy plot from initial rate data fora 52"4 mole percent n-butyl acrylate copolymer. 89

38. G.L.C. trace for a 50.0 mole percent n-butylacrylate copolymer degraded at 313 0 C for 11 hours.Column run isothermally at 40'C for 3 min. thenprogrammed at 50 C/min. 93

39. Room temperature g.l.c. of a 16-3 mole percentn-butyl acrylate copolymer degraded at 313'Cfor 16 hours. 94

xli

40. Room temperature g.l.c. of the permanent gas fractionfrom the degradation of a 52-4 mole percent n-butylacrylate copolymer pyrolyzed at 3130C for 24 hours. 95

41. (a) But-l-ene standard. (b). Product gases froma 16-3 mole % n-butyl acrylate copolymer degradedfor 12 hours at 3130C. 97

42. Infra-red spectrum of a 93.'4 mole percent n-butylacrylate copolymer degraded for 12.5 hours at 3130C. 99

43. G.L.C. trace of the liquid degradation products fromthe breakdown of a 50.0 mole percent n-butylacrylate copolymer heated at 3130C for 4 hours. 102

44. G.L.C. trace from G.C.M.S. of liquid degradation

products. 103

45. Mass spectrum of component 1 in Fig. 44 (n-butanol). 104

46. Mass spectrum of component 2 in Fig. 44 (methylmethacrylate) 105

47. Mass spectrum of component 3 in Fig. 44 (toluene) 106

48. Mass spectrum of component 4 in Fig. 44 (n-butylacrylate. 107

49. Mass spectrum of component 5 in Fig. 44 (n-butylmethacrylate) 108

50. Infra-red calibration curves for but-l-ene andcarbon dioxide. 119

51. Gas pressure plots for a 16-3 mole percentn-butyl acrylate copolymer degraded at 3130C. 120

52. Gas pressure plot for a 3"9 mole percent n-butylacrylate copolymer degraded at 3130C. 122

53. Gas pressure plots for a 50-0 mole percent n-butylacrylate copolymer degraded at 3130C. 123

54. Gas pressure plots for an 82.2 mole percent. n-butylacrylate copolymer degraded at 3130C. 124

55. Gas pressure plots ,for a 93"4 mole percent n-butylacrylate copolymer degraded at 3130C. 125

Xiii

56. Gas pressure plots for a 50.0 mole percent n-butylacrylate copolymer degraded at 3320C. 126

57. Gas pressure plots for a 93"4 mole percent n-butylacrylate copolymer degraded at 3320'. 127

58. Gas evolution plots for copolymers degraded at3130C. 139

59. Gas evolution plots for copolymers degraded at3130C. 140

60. Gas evolution plot for a 93"4 mole percent n-butylacrylate copolymer degraded at 3130C. 141

61. Gas evolution plot for a.50-0 mole percent n-butylacrylate copolyrmer degraded at 332°C. 142

62. Gas evolution plot for a 93.4 mole percent n-butylacrylate copolymer degraded at 332 0 C. 143

63. Residue and liquid volatile plots for a 3"9 molepercent n-butyl acrylate copolymer degraded at 313'C. 153

64. Residue and liquid volatile plots for a 16.3 molepercent n-butyl acrylate copolymer degraded at 3130C. 154

65. Residue and liquid volatile plots for a 50.0 molepercent n-butyl acrylate copolymer degraded at 3130C. 155

66. Residue and liquid volatile plots for an 82.2 molepercent n-butyl acrylate copolymer degraded at 3130C. 156

67. Residue, chain fragment, and liquid volatile plotsfor a 93-4 mole percent n-butyl acrylate copolymerdegraded at 3.130C. 157

68. Liquid volatile and chain fragment plots for a 50-0mole percent n-butyl acrylate copolyrmer degradedat 33200C. 158

69. Residue plot for a 50.0 mole percent n-butylacrylate copolymer degraded at 3320C. 159

70. Liquid volatile and chain fragment plots for a 93-4mole percent n-butyl acrylate copolymer degradedat 3320C.

xiv

71. Residue plot for a 93"4 mole percent n-butylacrylate copolymer degraded at 332°C. 161

72. G.L.C. trace of the liquid degradation productsfrom a 50-0 mole percent n-butyl acrylatecopolymer pyrolyzed at 313'C for 16 hours. 162

73. n-Butanol production as a percentage of' thetotal liquid products at 313'C. 166

74. n-Butanol production as a percentage of thetotal liquid products for a 93-4 mole percentn-butyl acrylate copolymer. 167

75. Plots of methyl methacrylate production at 313'C. 168

76. Plots of methyl methacrylate and n-butanolproduction at 313'C. 169

77. Plots of n-butanol production for a 93-4 molepercent n-butyl acrylate copolymer. 170

78. Infra-red spectra of an 82-2 mole percentn-butyl acrylate copolyiner. 178

79. U.V. spectra of an 82.2 mole percent n-butylacrylate copolymer degraded for the times shown. 181

80. N.M.H. of a 50-0 mole per'cent n-butyl acrylatecopolymer degraded at 3130C for 16 hours.(Run in carbon tet.rachloride solution.) 182

81. Molecular weight versus percentage

volatilization plots. 184

82. Effect of composition on gel formation at 329 0 C. 186

83. Effect of molecular weight on gel formationat 3290C. 187

84. Effect of temperature of degradation on gelformation in poly(n-butyl acrylate) TT 870,000. 188

n

85. Carbon dioxide evolution versus chain scissionplot for a 52-4 mole percent n-butyl acrylatecopolymer degraded at 2370C. 195

xv

86. Mass spectra (20ev) of products from thedegradation of a 52-4 mole percent n-butylacrylate copolymer for seven hours at 2370C. 196

87. G.L.C. trace of short chain fragments from thedegradation of a 50-0 mole percent n-butylacrylate copoiymer at 313'C for 11-5 hours. 197

xvi

TABLES

1. Nuclear Magnetic Resonance spectral analyses. 8

2. Reactivity ratios. 10

3. Composition of volatile products of degradation. 16

4. Colour and absorption maxima for conjugatedethylene structures. 22

5. Data on sequence distribution in the copolymers. 23

6. Summary of carbon dioxide-chain scission data. 30

7. Zip lengths. 32

8. Chain scissions and permanent gases. 37

9. Rate exponents for various reaction mechanisms 40

10. Photodegradation data obtained at 170 0 C. 47

11. Molar composition of monomeric products (MMIA/MA) 49

12. Chain scission and the production of carbon dioxide. 49

13. Zip lengths for depolymerization 51

14. Copolymer composition and molecular weight data. 58

15. G.L.C. columns used for investigation ofdegradation products 64

16. Dependence of TVA rate maxima on the composition

of methyl methacrylate/n-butyl acrylate copolymers. 77

17. Orders of reaction. 82

18. Activation energies for degradation of methylmethacrylate/n-butyl acrylate copolymers. 82

19. Activation energy from initial rate measurements. 90

20. G.L.C. data for hydrocarbon products. 96

21. G.L.C. data for isomeric butenes. 96

.4

xvii

22. G.L.C. data for permanent gases. 96

23. Mass spectral data. 98

24. Assignments of the infra-red absorptionsshown in Fig. 42. 100

25. G.L.C. data for liquid products. 101

26. Mass spectral data for liquid products. 109

27. Mechanism of fission in the mass spectrometer. 113

28. Mass spectra of liquid products. 114

29. G.L.C. data for a 50-0 mole percent n-butylacrylate copolymer degraded at 313 0C for 11 hours. 116

30. G.L.C. data for a 16.3 mole percent n-butylacrylate copolymer degraded at 313 0 C for 16 hours. 117

31. Data for a 52"4 mole percent n-butyl acrylatecopolymer degraded at 313'C for 24 hours. 117

32. Gas evolution data for a 16-3 mole percent n-butylacrylate copolyrner degraded at 313'C. 121

33. Gas evolution data for a 3"9 mole percent n-butylacrylate copolymer degraded at 313'C. 128

34. Gas evolution data for a 50-0 mole percent n-butylacrylate copolymer degraded at 313 0 C. 129

35. Gas evolution data for an 82.2 mole percent n-butylacrylate copolymer degraded at 313 0 C. 130

36. Gas evolution data for a 93-4 mole percent n-butylacrylate copolymer degraded at 313'C. 132

37. Gas evolution data for a 50.0 mole percent n-butylacrylate copolymer degraded at 332 0 C. 134

38. Gas evolution data for a 93.4 mole percent n-butylacrylate copolymer degraded at 332 0 C. 136

39. Molar ratios of carbon dioxide and but-l-ene. 144

xviii i

40; Molar ratios of but-l-ene and n-butyl acrylate. 144

41. Data for the degradation of a 3"9 mole percentn-butyl acrylate copolymer at 313'C. 146

42. Data for the degradation of a 16-3 mole percentn-butyl acrylate copolymer at 313 0 C. 147

43. Data for the degradation of a 50-0 mole percentn-butyl acrylate copolymer at 313 0 C. 148

44. Data for the degradation of an 82-2 mole percentn-butyl acrylate copolymer at 313°C. 149

45. Data for the degradation of a 93-4 mole percentn-butyl acrylate copolymer at 313 0 C. 150

46. Data for the degradation of a 50.0 mole percentn-butyl acrylate copolymer at 332 0 C. 151

47. Data for the degradation of a 93-4 mole percent

n-butyl acrylate copolymer at 332°C. 152

48. G.L.C. data for liquid products. 163

49. G.L.C. data for liquid products. 163

50. Analysis of the liquid degradation products.. 165

51. Analysis of the liquid degradation products. 171

52. Mass balance data. 173

53. Yields of n-butyl acrylate and n-butyl methacrylate. 175

54. Change in elemental analysis during reaction. 183

55. Polymers examined by sol-gel analysis. 185

56. Sol-gel analysis data for poly(n-butyl acrylate). 190

57. Sol-gel analysis data for poly(n-butyl acrylate). 190

58. Sol-gel analysis data for a 93.4 mole percentn-butyl acrylate copolymer Mn 76,500, degradedat 329 0 C. 191

xix

59. Sol-gel analysis data for a 93-4 mole percentn-butyl acrylate copolymer,M[ 3,160,000, degradedat 329 0 C. d e191

60. Sol-gel analysis data for a 52-4 mole percentn-butyl acrylate copolymer,Mn 1,330,000, degradedat 313 0 C. 191

61. Data for the degradation of a 52-4 mole percentn-butyl acrylate copolymer at 237°C. 194

62. Sequence distribution data for copolymers ofn-butyl acrylate and methyl methacrylate. 203

63. Sequence distribution data for copolymers ofmethyl methacrylate and methyl acrylate. 205

CHAPTER 1

INTRODUCTION

The presence of a small amount of a second monomercopolymerised into a homopolymer can have a profound effectupon the stability of the homopolymer. In some cases theeffect is to enhance the stability, in others the copolymeris less stable. Since copolymers are being increasinglyapplied commercially and industrially it is important tounderstand, at a fundamental chemical level, the reasons forthese differences in stability so that they may be taken intoaccount, along with the other physical and chemical propertiesof the material, in choosing materials most effectively fornew applications.

The introduction of a few per cent of an acrylatecomonomer has long been recognised as an effective method ofstabilising poly(methyl methacrylate) although the precisechemical mechanism of this effect is not clearly understood.The degradation reactions of a number of polymethacrylatehomopolymers have been thoroughly investigated (1) so that aconsistent unified picture of the reactions which occur inthe whole series has emerged. The acrylate series has alsorecently similarly been studied (2) and since a variety ofboth acrylates and methacrylates are available the largenumber of possible copolymers provides a fertile field fordetailed study of comonomer stabilisation mechanisms and theinfluence on them of successive small changes in the structuresof the constituent comonomers.

The present report describes detailed investigations ofthermal degradation processes which occur in the methylmethacrylate/ methyl acrylate and methyl methacrylate/n-butyl acrylate copolymer systems and the photodegradationreactions which occur in the former.

In order to obtain the maximum benefit from•work ofthis kind it is essential to be able to prepare copolymersof precisely known composition. However, since the chemicalnatures of the constituent monomers are so similar, considerable

-• •

-2

difficulties have been encountered in deducing the structuresof the copolymers using conventional analytical methods.Although values for the reactivity ratios of the methylmethacrylate/ methyl acrylate system were available in theliterature it was considered necessary to derive morereliable values. N.M.R. was found to be particularlyvaluable for this purpose and Chapter 2 is devoted to adescription of the measurement of reactivity ratios for thecopolymerisation of a number of acrylate monomers with methylmethaerylate. Chapters 3 and 4 describe various aspects ofthe thermal degradation of the methyl methacrylate/nethylacrylate system and Chapter 5 demonstrates how the reactionsare modified by photoinitiation.

Although some thermal analysis data are presented forthe methyl methacrylate/ methyl acrylate system in Chapter 3a very much more sophisticated form of the thermal volatilisationanalysis (TVA) equipment had become available when the methylmethacrylate/ n-butyl acrylate system was being studied andChapter 6 is devoted to this. In Chapters 7 and 8 a detaileddescription of the thermal degradation of the methyl meth-acrylate/ n-butyl acrylate system is given and the reportcloses with a general survey and discussion of the results(Chapter 9).

CHAPTER 2

REACTIVITY RATIOS FOR THE COPOLYMERISATIONOF ACRYLATES AN) METHACRYLATES BY

NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY

I NTRODUC TI ON

Few values of reactivity ratios for pairs of acrylateand methacrylate monomers. have been reported. This isprincipally due to the fact that the similarities of thestructures of the components make analysis of the copolymersdifficult. Elemental analysis cannot be made accurate enoughfor the purpose (3) and the application of i.r. or u.v. spec-troscopic methods would require the two monomers to haveabsorbing structures which are significantly different.

The gas-liquid chromatographic analysis of pyrolysisproducts has been applied but may be of doubtful quantitativevalidity (3). Radiometric (4) and isotopic (5) methods'arethe only ones to have been used successfully but, being timeconsuming and requiring elaborate experimental technique,have only been applied to a few isolated systems.

This chapter shows how nuclear magnetic resonancespectroscopy can be successfully used for such monomer pairs,although this method also has its limitations. It haspreviously been used to determine the monomer content ofvinyl acetate-ethylene copolymers (6,7) but reactivity ratioswere not calculated. The method is clearly widely applicableto copolymer analysis and is particularly valuable for itsrapidity and simplicity.

EXPE2 IMENTAL

Monomer purification

Methyl methacrylate (IR A) (I.C.I. Ltd.), methyl acrylate (MA)(B.D.H. Ltd.), butyl acrylate (BuA) (Koch Laboratories), ethylacrylate (EA) (Light and Co.), ethyl methacrylate (EMA) (I.C.I. Ltd.)

-4-

and propyl acrylate (PrA) (prepared by alcoholysis of methylacrylate (8)) were washed with caustic soda solution toremove inhibitor, subsequently with distilled water, anddried over calcium chloride. The monomers were vacuumdistilled and stored at -18'C.

Polymerisation

Copolymers of MIA with MA, EA, PrA, BuA and EMA werestudied.

After degassing and distillation in vacuum, monomermixtures of known composition were polymerised in bulk undervacuum to about five per cent conversion. The pair NIIMA-MAwas polymerised at 65 0 C with 0-075 per cent w/v azoiso-butyronitrile as catalyst. All other pairs were polymerisedat 60'C with 0.18 per cent w/v catalyst.

The copolymers were precipitated three times fromchloroform solution by methanol and dried in a vacuum ovenfor 24 hours at 50 0 C.

Copolymer analysis

N.m.r. spectra were obtained using a Perkin-Elmer RIO 60Mc/s spectrometer with integrator using 20 mg copolymer samplesdissolved in 1 ml of CDCI 3 . Ten integrals were obtained foreach sample and the average used for the calculation ofcopolymer composition.

RESULTS AND DISCUSSION

In the n.m.r. spectra of the copolymers MLA-EA, MIJA-PrA,NMA-BuA and 'IMA-E1,1A the peak due to the -0-C113 protons (inMMA) was resolved from those due to the -0-CH2- protons (inthe second monomer) as shown for a representative example inFigure 1. The monomer compositions of the copolymers werecalculated from the ratios of the areas under those peaks, asmeasured by the integral curves, the areas being proportionalto the number of protons contributing to the peaks. Thus

I-OCHI2- c 2 (number of second monomer units in chain)

I-OCHc3 c 3 (number of 10,A units in cha in)

in which I-0CH2- and I-OC'13 are the integrals of the -OCH2-and -OC11 3 peaks respectively.

-5 -

0

pm

0

Go

CH

I ~ C)

C))

Sr4

SV

-6-

If X is molar ratio MMA/second monomer in copolymer, then

I-OCH2 -/I-OC1 3 2/3X

No characteristic peaks could be distinguished for1MA-MA copolymers, MA and MN1A both contain -O-CH 3 groups andthe absorptions due to the tertiary protons in MA and themethylene protons overlap as illustrated in Figure 2. Thusthe total proton difference (8 from MMA, 6 from MA) betweenthe two monomers had to be used here to calculate copolymercomposition. Thus, if X = molar ratio MNA/MA in copolymer,then

Itot/I-O-CH 3 =Y=(8X + 6)/(X + 1)3

Therefore

x = (3Y - 6)/(8 - 3Y)

The ratios obtained by averaging ten determinations ofthe appropriate integrals when applied to the appropriateequations above gave the values for copolymer compositionsshown in Table l(a) and (b).

From the values of MI/N 2 and X shown in Tables l(a) and(b), reactivity ratios were calculated by the method Of Mayoand Lewis (9) using the equation

S= (m 1 /M 2 )U(l/X)[I + (MI /M 2 )rl] - 1}

r, versus r2 plots for the systems under investigation areillustrated in Figure 3. In estimating rl and r 2 values fromthe data the intersections of adjacent lines were discounted,and the reactivity ratios obtained are shown in Table 2.These values are seen to be in good agreement with thosecalculated from Q and e data (10) for these monomers (11,12)with the exception of the MMA-EA system.

The last column in Table 2 shows the only other resultsobtained directly for the radical copolymerisation of thesesystems. The values for MIA-MA were obtained by the use ofdeuterated monomer (5) (polymerisation temperature 130'C)while the MM.A-EMA system was studied by the use of radioactivetracers (4), and both are in reasonable agreement with thepresent results.

I-.%

00-

I • 7-•

C.5

a--

CC)

I

cc -OC

ato

Cs

- 8-

Table 1. Nuclear Magnetic resonance spectral analyses

(a)

Monomer pair Molar ratio in I-OCH 2- Molar ratiomonomer mixture 2 in copolymer

1 2 (M 1 /M 2 ) I-C1{ 3 (X)

MMA EA 4.05 0.08 8.762.02 0.15 4"451.0l 0.23 2.780.52 0.47 1.390.25 0.67 0.78

MMA PrA 4 0.09 7.412 0.20 3-331 0.32 2-00*5 0.63 1.060.25 0.97 0-69

MMA BuA 4 0.09 7-412 0.17 3.921-2 0.26 2.561 0.34 1-960-25 101 0-660.125 2"86 0-23

MIA EMA 4 0-15 4.442 0.29 2-391 0.67 1.00*5 1"27 , 0-530.25 2"53 0.26

(b)

Monomer pair Molar ratio in 31 Molar ratio

M M monomer mixture total in copolymerM1 2 (MI/M 2 ) I-0-CII3 (X)

MMA MA 5*2 7.77 7.71i0 7-32 1"940-85 7.33 2.00.38 6.86 0.760.20 6.78 0.64

2

C~l

(b) 2 2 (C)

3

1 2 3

Figure 3. rjversus-r 2 Plots for (a) I4MA-ILA; (b) I4DIA-EA;

- 10 -

Table 2. Reactivity ratios

Polymer- Monomer Other experiment-isation pair Experimental Q-e values al valuestempera- ° M M2 rI r 2 r r 2 r rture. 'C 1 2 12 1112

65 MIA MA 1.8±0.4 0"35±0"1 1"9 0-5 12"3±0"5 0"47±001(at 1300C)

60 •MA EA 203±0.12 0-24±0.12 1-43 0"73 - -60 MIA PrA 1-61±0.1 0.29±0.1 - N - - -60 MIA BuA 1.8±0-1 0.37±0-1 1.86 0-37 - -60 IA1A EMA 11-09±0-1 0.98±0-1 1.08 0-96 0.92 1-08

Thus nuclear magnetic resonance spectroscopy offers a generalmethod of analysis of copolymers with an accuracy comparablewith that achieved by other analytical methods. It has theadditional advantages that it is rapid and can cope with monomerpairs whose similarity in structure makes other analyticalmethods inapplicable. It is clearly of maximum use among theacrylates and methacrylates where one of the monomers does notincorporate the -0-CH2- structure. It is still useful, althoughless accurate when this is not so, provided the constituentmonomers have different numbers of protons as in the NI4A-MAsystem. It cannot be readily applied, however, when bothmonomers incorporate -0-C0H2- groups and the same number ofprotons, as for example, the system ethyl methacrylate-propylacrylate.

- ii -

CHAPTER 3

THERMA\L DEGRADATION OF COPOLYMERS OF METHYLMETHACRYLATE ANI) METHYL ACRYLATE. A. PRODUCTS

AND GENERAL CHARACTERISTICS OF THE REACTION

INTRODUCTION

A recent series of papers (13-16) has been concernedwith the study of degradation processes in copolymers. Ina summary of this and previous work (17) it was shown thatthis kind of investigation not only helps to clarify thedegradation processes which occur in the homopolymers of theconstituent monomers but also to demonstrate the mechanismwhereby a small amount of a comonomer may influence thestability of a polymer either favourably or adversely.

Acrylates are used commercially in this way to stabilisepoly(methyl methacrylate). The small concentrations usedhave no significant influence on the useful physical propertiesof the parent polymer. The mechanism of this stabilisationand the way in which the acrylate units influence the ultimatebreakdown process have not been reported upon. This and thefollowing chapter (18) represent the first stage of a studyof the degradation properties of methacrylate-acrylate systems.

EXPERIMEINTAL

Preparation of Copolymers

Methyl methacrylate (1,94A) (I.C.I. Ltd.) and methylacrylate (MA) (B.D.H. Ltd.) were purified by standard methodsinvolving washing with caustic soda solution to remove inhibitor,washing with distilled water to remove residual caustic soda,drying over anhydrous calcium chloride, and distilling undervacuum, the first and last 10% being eliminated. Thereafterthe monomers were stored in the dark at -18'C until used.The polymerisation initiator, 2,2'-azoisobutyronitrile waspurified by recrystallisation from methanol.

- 12 -

The reactivity ratios, ra = 18, rb = 0"35 (where A is10fl4A; B is MA), whose determination by nuclear magnetic reson-ance spectroscopic methods (19) was described in Chapter 2,were used to calculate the proportions of the monomers re-quired to prepare polymers of appropriate composition. Themonomers, purified as above, were distilled twice more undervacuum before distilling into dilatometers, containing 0.075%(w/v) of initiator, which were sealed off under vacuum.Copolymerisations were carried to approximately 5% conversionin a thermostat at 65 0 C. The resulting copolymers were pre-cipitated three times from chloroform solution by methanol,dried in a vacuum oven at 50'C for 24 hr. and ground to a finepowder (120 mesh). Copolymers with NNA and MA in the molarratios 112/1, 26/1, 7-7/1, and 2/1 were thus prepared.

Degradation Techniques

Degradations were carried out in the apparatus illustratedin Figure 4. Samples of the powdered copolymers (2-4 mg.)were spread evenly on the bottom of glass tube A, which was aQuickfit FG 35 flange with the end sealed and flattened. Itwas heated by immersion in an electrically heated and thermo-statted Wood's metal bath. The temperature of the insidesurface of the degradation tube was calibrated against thebath temperature by checking the melting points of pellets ofpure tin, bismuth and antimony (232, 271, and 327 0 C respectively).The water-cooled copper coil B, protected the grease of theflange from the heat of the bath.

The volume of the section bounded by taps Ti, T2 , and T3 ,including the Macleod gauge, and of the reaction vessel to TI,were measured by the standard method of expanding a knownvolume of gas, in a reservoir attached to the apparatus, intothese sections in turn and observing the change in pressure.The total volume to T2 and T 3 was of the order of 325 ml,varying with the various tubes, A, which were used from timeto time.

Permanent gases formed during degradation were estimatedfrom the increase in pressure observed during degradation withT2 closed and C immersed in liquid nitrogen. They wereanalysed for carbon monoxide, hydrogen, and methane by methodspreviously described (20). With C at -78'C the additionalpressure was due to carbon dioxide which was checked by ab-sorption on soda asbestos (B.D.11. Ltd., microanalyticalreagent) in D. Products not volatile at -78°C but volatileat room temperature were distilled into the calibrated

r~rd

UU

-- 14-

capillary E and stored at -18 0 C. They consisted of themonomers and were analyzed by standard gas-liquid chromato-graphy techniques by use of a Perkin-Elmer Fractometer.Methanol was sought in this fraction but never found.Products involatile at room temperature, which had condensedon the area of the reaction vessel cooled by B, were estimatedby weighing the reaction vessel before and after their re-moval by solvent.

Thermal volatilization analyses were carried out onequipment which has already been described in detail byMcNeill (21).

Molecular Weights

Number-average molecular weights were measured by usinga Mechrolab high-speed membrane osmometer. The molecularweights of the poly(methyl methacrylate) (PIANA) and t4e 112/1,26/1, 7-7/1 and 2/1 copolymers we e 82 x 104, 60 x 1060 x 10 4 , 42.5 x I04, and 37 x 10•, respectively.

Spectroscopic Measurements

Infrared spectra were obtained by using a Perkin-Elmer237 spectrophotometer. Materials of lower MA content wereused in the form of a film mounted between brass rings toprevent warping. For copolymers of higher acrylate contentsand for degradation residues, films were deposited on NaClplates from chloroform solution, the last traces of solventbeing removed by heating at 50'C in a vacuum oven for 24 hr.

Ultraviolet spectra were obtained by using a Unicam modelSP800 spectrophotometer with the polymer either in the form offilm or in chloroform solution.

RESULTS AND DISCUSSION

Thermal Volatilization Analysis

Thermal volatilization analysis thermograms for 25 mrg.samples of P11NM[A and the four copolymers are compared inFigure 5. Below 2000C the evolution of occluded volatilematerial is observed as the polymer melts and softens. Athigher temperatures there are two obvious trends in thecharacteristics of evolution of volatile products ofdegradation as the MA content of the copolymer is increased.

r~0

.11z Go 0coe

C'C-4

10

If))I-x

00

Li)

1.0.

r4-

ceciN I-

100

E ~ -cd000~10o

N ~%0.

%00

C"C)

U!011!00 10 94 Od GAIWDIýJ

- 16

First, the peak in the vicinity of 280 0C progressivelydecreases. Second, the m jaximum of the main peak moves tohigher temperatures. It is obvious from these thermogramsthat volatilization occurs at an appreciable rate at tem-peratures above 260'C, so that the temperature range 260-300°Cmay be most convenient for isothermal studies.

The volatilization thermo ram for PMIA, in Figure 5, has'previously been accounted for (21,22) in terms of the knowndegradation properties of this material. The lower tem-perature peak is the result of depolymerization initiated atunsaturated terminal structures. The peak at higher tem-perature is the result of depolymerization initiated by randomscission of the polymer molecules.

In the light of previous work on the influence of co-monomers on the degradation properties of PMMA (14-16) thereare two possible explanations of the progressive decrease inthe amount of chain terminally initiated depolymerization asthe MA content of the copolymer is increased. The first andmost obvious explanation is that, as in lA, NA copolymners withacrylonitrile (15), units of the second monomer block thepassage of chain depolymerization. Blocking of this kind iscertainly not complete since monomeric MA does appear amongthe reaction products. The constant ratio throughout therange of copolymer compositions studied, of one MA unit infour being liberated as monomer (see Table 3) may be somemeasure of the blocking efficiency in the temperature rangeof the randomly initiated reaction. The blocking efficiencyat the lower temperature at which the terminally initiatedreaction occurs should be very much greater as in acrylonitrilecopolymers (16).

Table 3. Composition of Volatile Products of Degradation

Composition, % by weight of total volatiles

Polymer CO2 Permanent gases Methanol lMMA MA Chain fragments

PMMA 9- -9112/1 i frace - - >96 - Trace26/1 1 Trace - 93 0.8 57.7/1 1 0.1 - 87 2-5 102/1 .3 0*4 - 64 7.0 25PMAa 7.5 1 15 1 - 0-76 75

a Data of 1Madorsky (25)

- 17 -

It is not possible to make quantitative deductions frommonomer ratio data alone, however, since the NA-terminatedradical which results from this kind of blocking must ultimatelyundergo some further reaction which may complicate the situation.

The second explanation of the suppression of the terminallyinitiated reaction is that, as in the copolymerization of 11iAand styrene (14,23), the presence of the second monomer favourscombination rather than disproportionation as the terminationstep. In this way the proportion of terminally unsaturatedstructures is drastically reduced even in copolymers containingquite small proportions of styrene. An assessment of thispossibility cannot be made for the M.IAMA system, however,since the relevant copolymerization data are not available.The tendency for the high temperature peak to move to highertemperatures with increasing NA content favours the firsttheory. Thus it can be accounted for in terpis of a progressivedecrease in the amount of monomer produced per initiation.This chain or zip length for depolymerization should not varywith copolymer composition if the second theory applies to thecomplete exclusion of the first.

Thus it is clear that the progressive suppression of thelow temperature peak in the volatilization thermograms inFigure 5 can be accounted for qualitatively in terms of blockingby MA units, although some contribution to the effect by a de-crease in the proportion of unsaturated chain ends formed duringcopolymerization cannot be entirely eliminated.

Changes in Molecular Weight

Figure 6 illustrates the changes which occur in themolecular weight of the 26/1 copolymer during degradation inthe temperature range 282-326°C. The other copolymers behavesimilarly. Although this behaviour is characteristic of arandom scission process it cannot be taken as evidence infavour of the random initiation reaction discussed above, sincethe large chain fragments which are seen in Table 3 to be amajor product of degradation of the copolymers must be formedin transfer reactions of the type which also lead to chainscission during the thermal degradation of poly(methyl acrylate)(PMA) .

All the copolymers remain completely soluble, even afterextensive degradation, which contrasts with the tendency togel formation in PHA (24).

A close association between chain scission and the productionof carbon dioxide is demonstrated in the following chapter (18).

-18-

CH

-0'

rd oca 0

toC)

00

caJ

CL 00

o

co %0

IDUISIJO ~ ~ ~ ~ ~ * 00%'L~e ji~~w

191

- 19 -

Volatile Products of Degradation

The compositions of the volatile products of degradationof the copolymers and the two homopolymers are compared inTable 3. Each of the results quoted represents an averageof 3-5 experiments carried to relatively high extents ofvolatilization in the temperature range 260-3000 C. The per-manent gas fraction is found to consist of hydrogen withtraces of methane and carbon monoxide.

The yields of chain fragments, carbon dioxide, and per-manent gases are approximately as one would predict from thecopolyrmer compositions and the behaviours of the homopolymers.There are, however, at least two major deviations between thepattern- of products predicted in this way and the actual ex-perimental results. First, no methanol is obtained from anyof the copolymers in spite of the high yield from PMA. Thus,because the yield of carbon dioxide is proportional to the MAcontent, it seems that carbon dioxide and methanol are producedin two completely separable reactions. This is perhaps sur-prising, since the methanol and carbon dioxide must both beassociated with decomposition of the ester group in the MAunit. Secondly, 11A is obtained from the copolymers in largeramounts than might be expected from the behaviour of PIA.This was previously observed by' Strassburger and his colleagues (26).It is obvious from the data in Table 3 that a fairly constantratio of about 1 in 4 of the MA units in the 26/1, 7.7/1 and2/1 copolymers are liberated as monomer. The amount producedfrom the 112/1 copolymer was too small to be detected.

Unsaturation in the Residual Copolymer

At higher extents of degradation, residual copolymer iscoloured yellow. For comparable extents of volatilization thedepth of colour increases with increasing MA content of thecopolymer. Ultraviolet spectra show an increase in absorbancein the 2750-3750 A region, as illustrated in Figure 7. Table 4shows the colour to be expected and the wavelength of absorptionfor various lengths of carbon-carbon conjugation which suggeststhat conjugated sequences up to six units in length are presentin the coloured material. The development of absorption at1630 cm" 1 in the infrared spectrum of degraded polymer, as inFigure 8, is confirmation of the appearance of ethylenicstructures.

-20-

1.2

0.8

.4

-oL.0

-0A

2750 3oo00 32•50

Wavelength, A

Figure 7. Ultraviolet absorption spectra of 2/1methyl methacrylate-methyl acryla-Lecopolymer in chloroform solution:(A) undegraded; (B) after 50" volatilization.

21-

%

I.I

.71

0.61 I I

0.5. l

i a I-0.4

II

-0.3 Not

U /

C 0.2

o

1600 180-j__ ___ __•

Wave Numbers, cM.

Figure 8. Infrared absorption spectra of films of 2/1, methylmethacrylate-methyl acrylate copolymer:(-) undegraded; (---) after 50' volatilization.

- 22 -

Table 4. Colour and Absorption Maxima forConjugated Ethylenic Structures

Length of Absorptionconjugation Colour maximum,

4 Pale yellow 29605 Pale yellow 33506 Yellow 36008 Orange 4150

The development of ethylenic conjugation is reminiscentof the degradation behaviour of polyvinyl esters which liberatethe corresponding acid in a chain process which progressesalong the polymer molecule. The ability of polymers todevelop this kind of conjugation, presumably by the liberationof hydrogen, has been demonstrated for polyethylene (27) andpolystyrene (28) under high energy and ultraviolet irradiation,respectively. It has been suggested (29) that once a radicalis produced on the polymer chain the reaction may proceed inthese polymers by a mechanism strictly analogous to the radicalmechanism which has been proposed for the degradation of poly(vinyl chloride) (30):

-CH1 2 -Ct-Ct2- CIH2- CHI2- Ctt2---

-CHi2-CH-C=I-CH-2- CH 2-Ct12 + H-

--- CH2 -CII=CC]1-CH-C11 2 CI --- + H12 [1]

A comparable reaction in IM1IA-MA copolyniers would involve theevolution of one or more of the products hydrogen, methane andmethyl formate and could account for at least some of thehydrogen and methane which have been shown above to appearamong the volatile products. In the present instance, theinitial radicals would be produced in the kind of transferreactions which lead to large volatile fragments and chainscission in the degradation of PHA (eq. [2]).

H! 9

R+ -CH -C-- -, + -CH 2 -c- [2]

CooCH3 COOCH 3

i3

- 23 -

Absence of Methanol among the Reaction Products

The complete absence of methanol from the volatileproducts of degradation of all four copolymers is an unexpectedfeature of the reaction in view of its prominence among theproducts from PMA. It would appear that while methanol pro-duction is a property of long sequences of methyl acrylateunits, it is not a property of isolated units and in this con-,nection it is clearly of interest to obtain information aboutthe distribution of MA units in the various copolymers underdiscussion. Harwood (31) has shown how the sequence dis-tribution of monomer units in copolymers can be predictedprovided reactivity ratios and monomer mixture or copolymercomposition are known. By using Hlarwood's methods, the datain Table 5 have been calculated for the four copolymers. Thefirst five columns in Table 5 are self-explanatory. The datain the last column represent the relative percentage concen-trations of MA units in the middle of MA triads whose immediateenvironment is thus comparable with the environment of a MAunit in a PMA molecule.

Table 5. Data on Sequence Distribution in the Copolymers

Copolymer IMA-M1MA MMA-MA MA-MA Fraction zmposition bonds in bonds in bonds in of MA in Z x % •MA

(MiA-MA) copolymer, copolymer, copolymer, middle units in% % % of MA triads copolymer

112/1 98.234 1-762 0.004 1 0.000020 0.000017

26/1 92-675 7P25 0-075 j 0-0004 0O-00157-7/1 77.95 21.1 0-95 1 0-0069 0-08

2/1 42.3 48.8 8.9 i 0.073i 2-43

In order to be able to discuss the whole question ofmethanol production in a completely satisfactory way it willclearly be necessary to study copolymers covering the wholecomposition range. However, the fact that no detectableamounts of methanol were obtained from any of the copolymersunder discussion allows a few observations to be made. Forexample, it is clear that methanol production is not even aproperty of pairs of MA units since measurable amounts ofmethanol would have been produced from the 2/1 copolymer, 8.9%of whose chain linkages are between pairs of methyl acrylateunits. Only 2-43% of all chain units are in an environmentcomparable with PiHA chain environment. Thus by comparisonwith PMA (Table 3) one would expect methanol to comprise about

- 24 -

0.36 % of the total products which is at the limit of theanalytical methods used. These results therefore indicatethat sequences of at least three MA units are necessary for theevolution of methanol. This conclusion is in accord with thethermal degradation behaviour of ethylene-MA copolymers (32).Thus block copolymers produce methanol in the quantities ex-pected from the MN content while random copolymers produce verymuch less.

Apparently conflicting results have been reported forpyrQlysis-gas-liquid chromatography studies on MIIA-MA copolymers(26) when it was found that a 4/1 copolymer yielded methanol.However, in the preparation of this copolymer the conversion ofthe monomer mixture was carried to 1M01 so that in the laterstages of the copolymerization virtually pure PMA was beingformed, It was also reported, however, that an equivalent mix-ture of the two homopolymers gave a much higher yield of methanolthan a copolymer of the same overall composition.

Production of Methyl Acrylate and Chain Fragments

Although the principal products of degradation of PMA andPKiIA are so different, the reactions have been explained interms of a single chain depolymerization mechanism in whichthere are two competing steps, namely depropagation and intra-molecular transfer (1). In PMHA the former predominates sothat high yields of monomer are obtained, while in Pi'IA the latterpredominates to give long chain fragments as the principalproduct. It is clear from Table 3 that as one would expect,a high proportion of the 1M in the copolymers is accounted foras monomer among the products. Table 3 also shows that theyields of MA and chain fragments both increase in proportionto the concentration of MA in the copolymers. The constantratio, throughout the copolymer composition range studied, ofone MA unit in four being liberated as monomer is therefore ameasure of the relative probabilities of depropagation andtransfer occurring at a long chain radical terminated by a IMAunit. In copolymers containing higher proportions of MIA inwhich a higher proportion of the MA units will occur in groups,depropagation will be inhibited as in PMA, and these relation-ships will undoubtedly break down.

Production of Carbon Dioxide

Carbon dioxide is not produced in significant quantitiesduring the thermal degradation of PHIL[A. On the other hand,the data in Table 3 demonstrate that the yields of carbon dioxide

- 25 -

from the copolymers are approximately in proportion to the MAcontent so that liberation of carbon dioxide may be a propertyof individual MA units. In view of a possible association ofchain scission with MA units it seems that carbon dioxide mightalso be closely associated with chain scission. It is alsoclear that since chain scission plays such a vital part in thisdegradation process its mechanism will have to be clarified ifa satisfactory picture of the overall reaction is to be given.It has therefore been found convenient to devote the followingchapter to carbon dioxide production and chain scission leadingto a discussion of mechanism and kinetics,

CONCLUSIONS

The high rates of monomer production from PMNNA due to chainterminal initiation are suppressed in the copolymers due to theblocking of the depropagation process by the MA units. Thusdegradation occurs in the copolymers only at temperatures atwhich PINA molecules devoid of terminal unsaturation degrade.This involves random scission followed by depropagation. Whendepropagation reaches an isolated unit of MA, there is com-petition among deprogagation, intramolecular transfer, andintermolecular transfer which results in MA monomer, largechain fragments and chain scission, respectively. Depropagationis unlikely to occur from a radical chain end which comprises asequence of more than one 11A unit so that only the transferprocesses can occur. During the course of the reaction thecopolymers become yellow, the rate of coloration being greaterthe greater the MA content of the copolymer. This has beenassociated with the formation of carbon-carbon conjugation inthe chain backbone. It seems possible that hydrogen andmethane, which appear as minor products of the reaction, arebeing liberated from adjacent units in the polymer molecules ina chain reaction which is initiated by the intermoleculartransfer process mentioned above and which is therefore incompetition with chain scission.

The fact that methanol does not appear among the volatileproducts of degradation of any of the copolymers demonstratesthat at least Three adjacent units of NA are necessary in thepolymer chain for its formation.

. 26 -

CH1APTER 4

THERMAL DEGRADATION OF COPOLYMEMS OF METHYLMETHIACRYLATE ANJ) METHYL ACRYLATE. B. CHAINSCISSION AND THE MECHANISM OF THE REACTION

INTRODUCTION

It was suggested in the previous chapter (33) that theremay be a link between the chain scission and the production ofcarbon dioxide which occur during thermal degradation of methylmethacrylate-methyl acrylate (MIMA-MA) copolymers. In thepresent chapter it is demonstrated that over the range ofcopolymer compositions studied one molecule of carbon dioxideis liberated for each chain scission. It is therefore possibleto use carbon dioxide production as a measure of chain scissionand thus to study the relationship between chain scission andcertain other features of the reaction in a very much moresensitive way and to larger extents of reaction than is possibleby use of the more conventional measurement of molecular weight.

EXPER IMENTAL

The polymers referred to and the experimental methodsapplied in the present chapter are those described in theprevious chapter (33).

RESULTS AND DISCUSSION

Chain Scission and Carbon Dioxide Production

A typical series of curves for the production of carbondioxide over a range of temperatures is illustrated in Figure 9.During the initial period of low reaction rate, which is morepronounced the lower the temperature, the copolymer retains itspowdery appearance and has clearly not attained the temperatureof the reaction vessel. The point of transition to the higherrate can be seen to be associated with the conversion of thepolymer to a transparent film. This may be loosely terr.iedmelting. The mobility of the "molten" polymer thereafter will

->.

- 27 -

0

326*-0.

03 294 272

E 260'Eo .2

00

Time, min.

Figure 9. Production of carbon dioxide during degradationof 7"-/1 methyl methacrylate-methyl acrylatecopolymer.

- 28 -

ensure that its temperature rises rapidly to that of thereaction vessel and the subsequent high rate may be taken asa reliable measure of the rate at the temperature quoted. Byheating copolymer samples slowly it was confirmed by visualobservation that the "melting points" are all in the range240-260'C whereas pure P1MA "melted" at 160-170'C.

It has previously been shown (13) that the number of chainscissions N which has occurred per molecule of copolymer isgiven by equation [3],

N = [M (i - x)/MI] -1 [3]0

in which Mo and 1 are the original molecular weight and themolecular weight at an- extent of volatilization x, respectively.In Figure 10 chain scissions per molecule of polymer calculatedin this way are plotted against molecules of carbon dioxideproduced per molecule of polymer at a variety of temperaturesfor various extents of degradation of the four copolymers (seeTable 6, columns 5 and 6). The straight line in Figure 10has been drawn with a slope of 450. It is obvious that withinexperimental error and over the range of experimental con-ditions represented in Figure 10, each chain scission in allfour copolymers is associated with the production of onemolecule of carbon dioxide. Beyond about 45ýo volatilizationthere is some deviation from this relationship, but this isdue to the fact that polymer molecules are then disappearingfrom the system in appreciable quantities by complete un-zipping so that equation [3] no longer applies. Since themolecular weight decreases so rapidly during degradation andsince the measurement of low molecular weight is inaccurateowing to diffusion effects in the osmometer, the measurementof the production of carbon dioxide is very much more reliableat all but the lowest extents of degradation. For this reason,all chain scission data quoted in the following sections ofthis chapter were calculated from carbon dioxide productiondata.

Volatilization, Chain Scission, and Zip Length

The previous chapter (33) has shown the reaction to con-sist essentially of chain scission followed by three competingreactions, namely, depropagation to monomer, intramoleculartransfer giving large chain fragments, and intermoleculartransfer which may lead to further chain scission or, to alesser extent, to conjugated unsaturation in the polymerbackbone. It seems, therefore, that there should be at least

-29-

-20 .

0 16E

0'

a..

0 -12 13

0

0

00

CC

U 4 8021C',,j

Molecules of C0 2 /Molecule of polymer

Figure 10. Relationship between carbon dioxide production andchain scission during degradation of methylmethacrylate-methyl acrylate copolymers;(0) if112/1 ( 2• 6/1 (0) 7t,.-cl 7/1; • 2/1.

-30-

Table 6. Summary of Carbon Dio~xide-Chain Scission Data

DegradationfItern- Volatil- N fIoeue

perature, ization, M of Scissions/ Moeuspolymer. Oc% residue Imolecule C02 /molecule

112/1 294 6-6 66,000 j 7-46 8-2282 22-5 - --7294 31-5 58,000 I 6-21 . 11-6

26/1 282 26-0 15,000 28-6 .24-5282 .62:0 - I -52-4282 179 5 f - I 66.0294 2-3 4ý75,000 2-34 .2-8294 23-6 '1 40,000 118 17-6272 - 1190,000 2- 109294 12-5 115sOOO 4.8 6-0

3611:0 '110,000 4-686--7.7/1 282 16 4 27,000 12-1 .10-5

282 25-0 23,600 12-5 12-4282 31-4 13,000 21-3 17.8

2417-3 20,000 17-5 16-1294 32-2 - . 18-0294 50-3 - 30-0326 174.7 35-51

2/1 294 13-9 1115,000 20.3 22-0310 5-3 24,000 13 *6 11-2310 15*4 16,800 17-7 17.7310 27:3 .I.-35.5

1 310 45 2 - 48-0310 '53.4 - 60-0

a qualitative correlation between the number of monomer unitsliberated per chain scission, the zip length, and copolymercomposition arid the pattern of products described in theprevious chapter.

On plotting r~epresentative data from Table 6 as in Figure11, it is clear that over a large part of the reaction a con-stant amount of volatilization occurs per chain scission foreach copolymer, which is consistent with the reaction mechanismoutlined above. From the slopes of plots, as in Figure 11,the zip length for each copolymer can be calculated as inTable 7. Although there may be considerable experimental errorin these values of zip length, they are undoubtedly correct towithin 10%.

-31-

-80

-60

- -400

X

"- 200

S2 4 " 60• 80

Chain Scissions/Molecule of polymer

Figure 11. Relationship between volatilizalion and chainscissions during degradation of 26/1 methylmethacrylate- methyl acrylate copolymers:(.o) 282oC; (o) 294 0 C; (3) 326 0C.

-32-

Table 7. Zip Lengths

jMW lost/ Average wt. ZipSlope Initial !scission 'of monomer Length

Copolymer A M11, 1 - (moA/100) unit B (MoA/100B)

112/1 2"9 !600,000 17,320 100.0 17326/1 1-22 i600,000 7,320 99"5 74

7-7/1 1l. 6 8 1425,000 7,150 97*2 742/1 10-88 '370,000 3,260 95-3 34

Table 5 in the previous chapter indicates that in the112/1 copolymer the proportion of adjacent NAU units isextremely small, and indeed there must be relatively few MAunits even in the close vicinity of each other. Thus thereshould be expected to be relatively little intramoleculartransfer and this is confirmed by the absence of any sig-nificant amount of chain fragments. In the previous chapterit was shown that approximately one in four of the methylacrylate units areliberated as monomer so it may be assumedthat at least for single MA units there is a 1 in 4 chancethat depropagation will pass through them. For the 112/1copolymer this will lead to an average zip length ofapproximately 150 [112 + (112/4) + (1 1 2 / 4 L) + ... ] which isin satisfactory agreement with the value of 173 in Table 7.

As the MA content of the copolymers is increased so thata greater proportion of the MA units form adjacent sequences,monomer production is increasingly suppressed and intramoleculartransfer with the production of chain fragments plays an in--creasing part. Since the zip length values in Table 7 becomeproportionately greater compared with the monomer ratios inthe copolymers and since the zip length includes chain frag-ments based as it is on total weight loss, it is obvious thata radical terminated by a sequence of MA units is very muchmore likely to undergo intramolecular than intermoleculartransfer. The present experimental data only allow thesequalitative observations to be made. A more quantitativeanalysis will only be possible when a precise analysis of thechain fragments can be carried out.

Chain Scission and Copolymer Composition.

In view of the possibly important part which MA unitsmay play in the chain scission process it is important to

33-

investigate the dependence of rate of chain scission upon MAcontent as a step towards the establishment of the overallreaction mechanism. The rates of carbon dioxide evolutionafter "melting" of the polymer (see Figure 9)"were taken as ameasure of rate of chain scission. Owing to the considerablepotential experimental error and the possibility of the samplesize influencing measured rates, rates of carbon dioxide pro-duction were measured for samples of various sizes for eachcopolymer. Results are illustrated in Figure 12, whichdemonstrates no consistent trend with sample size. Thehorizontal lines indicate average values of rate. Figure 13illustrates the relationship between rate and MA content forthe 26/1, 7.7/1 and 2/1 copolymers at 294 and 272°C. Theslopes of the straight lines through these points are 0.4-2and 0-33 respectively. Although the limited number ofexperimental data makes' the probable error fairly high, takingthe two sets of data together it may reasonably be claimedthat the rate of chain scission is proportional to less thanthe half power of the MA content of the copolymer.

Activation Energy for Chain Scission

The data for the 26/1 and 7.7/1 copolymers in Figure 12are represented in Figure 14 in the form of an Arhennius plotfrom which a value of 23-5 + 2 k.cal/mole may be deduced forthe energy of activation of chain scission. This is alsothe value for the energy of activation for volatilization,since volatilization has been shown to be a linear functionof chain scission for each copolymer. It is low comparedwith the values for the initial volatilization of PMM4A (34)and PMA (25) which are 32 and 34 K.cal/mole, respectively.Both the low exponent in the relationship between rate ofchain scission and NA content and this low value of energy ofactivation are evidence in favour of the involvement of chainscission as an integral part of a radical chain process.

Chain Scission and the Production of Permanent Gases

It was reported in the previous chapter that permanentgases, particularly hydrogen, are produced during the reactionalthough in quantities very much less than carbon dioxide.It was also suggested that these permanent gases might beassociated with the colouration which occurs, the radicalresulting from intermolecular transfer acting as a centre ofinitiation for two competing reactions, the first resulting

-. .. ÷ ..

N (U2

No 0

ElE< CN- - r-4

00 0

ol 0lCO C4 IN0N

13 0 r-

oA00

13 0 0-

CC 4 C)C

0% K

%

m c

Co 0,-

00

4 (N4

Ol000

o C'cooOL X'Ulw'BLV'LLU o:) 0 L014:nPOd JO04D

- 35 -

40

0 02 0

Hý d

'0~ ~ p.U,

U, d

0Pi

0-Pc

N r.

-C Q)

G4DJ So

- 36 -

-7.6

-7.8 "\

26/1 -

0-8.2

-- 8.4

- ~ 1.6 1. 7 1. 1.9

1/T°AX 1 0a

Figure 14. Arrhenius plot of chain scissions during degradationof methyl methacrylate-methyl acrylate copolymers.

- 37 -

directly in chain scission and the second in the evolution ofhydrogen and methane in a chain process similar to that whichhas been suggested for the liberation of hydrogen chloridefrom poly(vinyl chloride) (30). If these conclusions arecorrect, they should provide a basis for the explanation ofany relationship which exists between carbon dioxide andpermanent gas production.

The data in Table 8 demonstrate that for a wide range ofdegradation temperatures and extents of reaction, the ratioof chain scissions to permanent gas production is constantfor each copolymer but that the proportion of permanent gasesincreases with the MA content of the copolymer. This is inaccordance with the mechanism outlined above. Thus the con-stant value for each copolymer demonstrates the closeassociation of chain scission and permanent gas productionthroughout the whole course of the reaction. Also, becausepermanent gas production and colouration are associated withMA rather than MMA units, the chain reaction in which permanentgases are liberated is more likely to occur, the greater theprobability that the transfer radical finds itself in thevicinity of MA units; that is, the greater the MA content ofthe copolymer.

Table 8. Chain Scissions and Permanent Gases.

ChainDegradation Volatil- scissions/temperature, ization, molecules of Average

Copolymer 0C % permanent gas ratio

26/1 200 1 20)220 2 25294 85 18i 20294 85 17326 80 20

7-7/1 326 80 121326 1 121

2/1 294 50 81310 52 71 8310 .60 8)

Mechanism and Kinetics of Chain Scission

In order to gain further information about the site ofthe initial chain scission and of the intermolecular transfer

- 38 -

process with which chain scission and carbon dioxide productionare associated, these two processes may be regarded as theinitiation and propagation steps of a chain reaction whoserate is measured by the rate of chain scission.

In view of the fact that the degradation reaction occursin the copolymers at the same temperature as the randomlyinitiated phase of the degradation of PiVR'IA, random initiationseems the most likely initiation process in this chain reaction.Thus the rate of initiation would be independent of the MAcontent of the copolymer and given by

rate of initiation ki[Mn] [4]

in which ki is the rate constant for initiation and [Mn] theconcentration of the copolymer which may be taken as constantthroughout the range of copolymer composition studied.

The propagation step might reasonably be expected toinvolve the abstraction of a tertiary hydrogen atom from amethyl acrylate unit by a polymer radical, P., thus

rate of propagation = k P[P.J[MA] [5]

Chain scission and the production of carbon dioxide can beexplained in terms of reaction of the resulting radical ina six-membered ring mechanism [eq. 6]

0I-7o 0

SI •CH2_C 2•

.2 12 2 1 2COOCII 3 CH03

I

The radical I will liberate monomer and chain fragments andultimately carry on propagation of the chain scission processby abstracting a further tertiary hydrogen atom.

-39-

In accordance with experimental findings, one moleculeof carbon dioxide is produced per chain scission in thismechanism if it can be assumed that the chain length isappreciable such that the number of initiation steps whichoccurs without production of carbon dioxide is negligiblecompared with the number of propagation steps. It may alsobe noted that a unit of 4i'fA is formed in each act of scissionwhich could explain the small proportion of 1I'IA which is foundamong the volatile products of degradation of PMA (35,36).

Termination might reasonably be expected to occur bymutual destruction of pairs of radicals, so that

rate of termination k t[P-]2 [7]

On applying stationary-state kinetics to this mechanism itcan easily be demonstrated that the rate of chain scissionshould be proportional to the first power of the IA, contentof the copolymer, which is not in accordance with the ex-perimental values of less than 0-5.

It is possible, however, that initiation may be specific-ally associated with MA units and that the propagation step israndom rather than specifically at MA units. Termination, onthe other hand, may be associated with the hydrogen andmethane loss reaction:

(c 3) 3)H H

SI I,-CH--C-- C2-U-- W12! 2

COUCH3 COOCH CO- 3

-CHI 2-C= cC = cH - C- 2 (CH14) [8]II ICOOCHt3 COOCHt3 C0Ctt3

Thus conjugated radicals of type II may be so stable as to beincapable of continuing propagation by abstraction of ahydrogen atom in which case termination would become effect-ively first order with respect to radical concentration.Eight different combinations of initiation, propagation andtermination are therefore possible, as in Table 9.

-. 4 ...

- 40 -

Table 9. Rate Exponents for Various Reaction Mechanisms

Initiation Propagation Termination n

Random Random Pairs of radicals 0Random Random Single radicals 0Random At MA units Pairs of radicals 1Random At MA units Single radicals 1At MA units Random Pairs of radicals 2At 1A units Random Single radicals 1At MA units At MA units Pairs of radicals 3/2AtMA units At MA units Single radicals 2

The exponent, n, in the relationship,

Rate of chain scission = K[MA]n [9]

is given for each of these combinations in the final columnof Table 9. Since the experimental value of n was found tobe less than 0"5, it is clear that a significant proportionof,,both initiation and propagation must be random althoughit is not possible to say, on the basis of the kineticevidence above, whether a propprtion of initiation or propa-gation or both are associated with 14A units. Since no bondswhich are thermally more labile than the bonds already presentin P4H1.A are introduced into the molecules by MA units andsince the introduction of MA units does not lower the reactiontemperature from that of the randomly initiated degradationof PMM1A, it seems most unlikely that there can be a significantproportion of preferred initiation at MA units. However,the tertiary hydrogen atoms in MA will be much more liable toparticipate in the propagation step than the chain methylenegroups whose reaction is implied by random propagation. Thefact that transfer of methylene hydrogen atoms plays a sig-nificant part in presence of tertiary hydrogen atoms is ex-plained by the very much higher concentration of the former.A reasonable mechanism for chain scission and the formationof carbon dioxide, following upon transfer at methylene groups,may be represented as in equation[10].

- 41 -

CH 3

CI3 33 31/• I3'-•C---C £jC~L=0 ----v . - CIt - C'- + CO2

1c '>~o 2COOCII3 COOCH

CH- CI1C31 3 1~ (H3

COOCH 3

CONCLUSIONS

These relationships between chain scission and certainother features of the reaction together with the generalreaction characteristics reportedin the previous paper leadto the following representation of the mechanism of theoverall reaction

Copolymer molecule

Chain scission(random)

Terminal chain radicals

2 Depropagation to first(2) •MA unit

Monomeric MM +MA-terminated radical

Intramolecular Intermolecular(3) Depropagatio (4) _ transfer

fer \5) .(predominantlySrandom)

MA monomer Chain fragments Chain radical

Chain scission Cd4, H2, un-and CO 2 saturation,

andcolouration

-42-

The evidence favours random scission in reaction (1) ratherthan preferential scission at NA units. It is the interplayof reactions (3),(4), and (5) which determines the principalcharacteristics of the overall reaction and their variation withc'opolymer composition. If the terminal MA resulting fromreaction (2) is single, as would be predominantly so in the112/1 copolymer, then only reactions (3) and (5) are possible.When the MA content is increased such that MA units occurincreasingly in groups, then reaction (4) becomes significantand, by comparison with reaction (5), proves to be much fasterthan reaction (3). Although tertiary hydrogen atoms areundoubtedly more reactive in transfer processes than methylenehydrogen atoms, the very much higher concentration of thelatter in the copolymers ensures that reaction (5) is pre-dominantly a random process.

Although for any given copolymer the relative amountsof reactions (6) and (7) remain constant throughout the reaction,reaction (7) becormes relatively more important as the pro-portion of MA in the copolymer is increased. This is notsurprising, since the principal permanent gaseous product,hydrogen, is to be expected from the MA units only, accordingto the mechanism proposed.

- 43 -

CHAPTER 5

PHOTODEGRADATION OF COPOLYMERS OF METHYLMETItACRYLATE AND METHYL ACRYLATE AT

ELEVATED TEMPERATURES

INTRODUCTION

Recent publications (15,18,33) have demonstrated how thepresence of the comonomers, acrylonitrile and methyl acrylate,influences the thermal degradation of poly(methyl methacrylate).In both systems the depolymerization reaction is initiated byrandom scission in the methyl methacrylate segments of thepolymer chains. Like pure poly(methyl methacrylate) thesecopolymers also depolynerize at elevated temperatures underthe influence of 2 537 X radiation. In the acrylonitrilepolymer (16) however, the initiation process consists of chainscission specifically at acrylonitrile units, and the principaldifferences between the thermal and photo reaction have beenaccounted for in terms of the different sites of initiationand the influence of the temperature (280°C and 160'C for thethermal and photo reactions, respectively) and viscosity of themedium upon the relative rates of the subsequent constituentprocesses comprising the total reaction. In the presentchapter the principal features of the photodegradation of themethyl methacrylate-methyl acrylate copolymer system are de-scribed and differences from the thermal reaction discussed.

EXPERIMENTAL

Copolymers

The four.copolymers studied were those whose preparationshave previously been described in chapters 3 and 4 (18,33).Their molar compositions are (MMIA/N1A) 112/1, 26/1, 7.7/1, 2/1.

Molecular Weights

Number-average molecular weights were measured by use ofa Mechrolab high-speed membrane osmometer.

- - 44 -

The amounts of material available for the measurement ofmolecular weights were of the order of only a few milligrams.Molecular weight measurements are therefore subject to con-siderable error and this accounts for the scatter of points inFigures 18 and 19.

Photodegradation Techniques

Photodegradations, except those involving the measurementof CO2 produced, were carried out as previously described (16),the polymer (2-4 mg.), in the form of a thin transparent film,being •rradiated in vacuo, through silica, by a source of2 537 A radiation.

Because of the very small amounts of CO2 involved it wasfound more convenient to make the CO2 evolution measurementsin the apparatus used for thermal studies with a suitablymodified reaction vessel. The exit tube was situated at theside of the reaction vessel, being replaced at the top of thereaction vessel by a silica window through which the polymerwas irradiated.

All product analyses were carried out exactly as describedfor the thermal reaction.

RESULTS

Influence of Temperature on Rate of Volatilization

The primary influence of ultraviolet radiation on-poly(methyl.methacrylate) is to cause chain scission, the radicals,

Ci3

" CH 2-•

COOCR3

ultimately appearing in the system (37). The overall charac-teristics of the photolysis depends upon the subsequent re-actions of these radicals, which in turn depends upon the tem-perature. At high temperatures, at which the polymer is inthe liquid state, monomer produced in the equilibrium,

CH YH ~ CH CHI3 3 I3

~.sCII -C-CII-C zz.ý-C -Cf'; + CHI -C2 22- 2 c 4g

COOH C0011 COOH 3 C001133 333

- 45 -

can easily escape, so that the reaction tends to the right andquantitative conversion to monomer occurs. On the other hand,at low temperatures when the polymer is in the form of a rigidsolid, monomer can not readily escape, appreciable depolymer-ization does not occur, and the polymer radicals subsequentlymutually destroy each other. At low temperatures, therefore,the photolysis is characterized by chain scission and at hightemperatures by monomer production. It is the high-temperaturereaction with which this chapter is concerned. Unfortunately,however, the range of temperature in which this reaction can bestudied is restricted, the lower limit (--150 0 C) being governedby the softening point of the polymer and the upper limit(-200'C) by the onset of thermal degradation. Clearly, inthis temperature range the viscosity of the polymer is changingrapidly with temperature, and since the viscosity could have aprofound influence on the above equilibrium it is important toobtain some assessment of the influence of the viscosity on theoverall reaction. The data in Figure 15 demonstrate that thereis no significant change in the rate of photodegradation of P11NAin the temperature range 150-170 0 C. Since the reaction consistsof photoinitiation followed by complete unzipping of the polymerchains the rate of the reaction should be governed by the rateof initiation and since photoinitiation should be associatedwith an activation energy close to zero, the constant rate isaccounted for and in turn implies that there is no significantviscosity effect which should be expected to cause an increasein rate with temperature. In turn, it may be concluded thatthe increase in the rate of photodepolymerization of the twocopolymers with temperature, illustrated in Figure 15, is notassociated with a viscosity effect but is a direct manifestationof the modification of the overall reaction by the presence ofthe comonomer.

As a result of these experiments, 170'C was chosen as asuitable temperature for a general study of the reaction sincean appreciable rate of reaction was obtained over the wholecopolymer composition range without interference from thermaldegradation. All subsequent data were obtained at this tem-perature and are summarized in Table 10. Chain scissions permolecule of polymer were calculated by using the formula

N = CL 0 (1 - x)/CL - 1

in which CLo and CL are the original chain length and the chainlength at an extent of volatilization x, respectively. Thenumber of chain scissions per unit length of chain, n, given by

n = N/CLo = [(i - x)/CL] - (1/CLo)

must be used as a comparative measure of the extent of chainscission, however, since the copolymaers have substantiallydifferent molecular weights.

-46

-40

30

02 20

~-10

14o 150 160 170 180I! I I ! -

Temperature, °C.

Figure 15. Extent of volatilization in 30 min at varioustemperatures: (o) poly(methyl methacrylate;(0). methyl methacrylate-methyl acrylate copolymer(26/1); (A) methyl methacrylate-methyl acrylatecopolymer (7"7/1).

- 47

TABLE 10. Photodegradation Data Obtained at 1700C

Chainlength

Volatil" of co- ScissionsCo- Time, ization, MV of polymer Per mole- Per chain unit

polymer min % residue (CL0 ) cule (N) (n=N/CL0 ) x 104

112/1 01 0 600000 6000 0 0101 5o4 250000 1.27 2-1130 20.7 - -60 41-4 2370004 047 0-7890 53o6 - -

120 60.8 - -26/1 0 0 600000 6030 0 0

5 1.7 376000 0-12 0"1915i 4-0 248000 1"32 2"19

23 13"4 229000 1"27 2"1130 j 30"4 172000 1"50 2"4945 34-2 180000 1819 1"9860 36-2 145000 , 1"65 2-7390 46.2 118000 1.68 2"79

120 54.-0 125000 1-21 2.017-7/1 0 0 425000 4370 0 0

.. 5 1-5 411000 0-02 0.0515 2-9 321000 0.27 0.6223 4"4 181000 1-25 2-8630 5o2 263000 0:53 1-21

S45 16"2 168000 I 1"12 2-5665 20.3 144000 lo35 3-0975 22.3 157000 1.10 2"5290 27-8 150000 1-05 2.40

120 29.8 ; 140000 1.13 2-58300 40-0 81000 1.96 4.47

2/1 0 0 370000• 3880 0 030 1"75 149000 1.46 3o7660 5.8 78000 3-08 7-95

150 13-9 49000 5-52 I 14-22300 28-2 43000 1 4.91 12o65600 39-6 47000 3.75 9-65

Molecular Weight Changes

The changes in molecular weight which occur duringphotodegradation are related to the extent of volatilizationin Figure 16. Like the results of thermal degradation theyare characteristic of a reaction in which random chain scission

-5 -,0 '.0

(NJ

rd~

0 -q

14

0 .6

-0

00

U-. Ad E00r-

0)0

to 1q

-~~ 0 E'

'0E

0

10 20 30 40 50 6H-, - r.. II

VT~~t.i z tinn 56

- 49 -

is involved. Even in the 2/1 copolymers there is no directevidence of the crosslinking which is typical of poly(methylacrylate), the residual material being completely soluble inall cases.

Volatile Products of Degradation

The pattern of volatile products is closely comparablewith that produced in the thermal reaction. The only significantdifference concerns the ratio of the monomers. The gas-liquidchromatographic analysis of the monomer fractions are presentedin Table 11, from which it is clear that approximately one inten of the MA units is liberated as monomer, compared with onein four in the thermal reaction.

TABLE 11. Molar Composition of Monomeric Products (MMA/MA)

Copolymer Products

112/1 1000/126/1 320/1

7.7/1 80/1

Rates of Volatilization

Volatilization versus time curves for the four copolymersand PI•HA are compared in Figure 17. It is obvious that, asin the thermal reaction, increasing concentrations of MA in-creasingly stabilize the copolymers.

Chain Scission and the Production of Carbon Dioxide

The data in Table 12 summarize the results of experimentsdesigned to determine the relationship between chain scissionsand CO2 production.

TABLE 12. Chain Scission and the Production of Carbon Dioxide

Copolymer Temp. Volatiliz- MWl of Scissions/ C02 / C02 /ol ation, % residue molecule molecule scission

26/1 160 11-6 244,000 1-062 4.84 4.5170 11-5 292,000 0.82 4-46 5&4

7.7/1 170 24.8 118,000 1.71 5-15 3-0170 4.6.2 70,000 2.26 10.6 4.7

2/1 170 7-4 75,000 3.57 7.47 2.1

-- 50-

[-70

L60

50I.

0!0

60-010 20 1006

_ III30 m~mn

Fiue1.Vltlzto-iecre0o h htdgaaino20yrehimtaryae n ehlmehcvaemtv

acrlat ooy.ir t100 o MA 0 M/A12

(crl) t MMA/NAr 26/1 ( 70") (o)/A 7.7/1 ; (Aj) ima/mA, 12/1 ;

- 51 -

In the thermal reaction a strict 1/1 ratio, was foundthroughout the polymer composition range which made it possibleto use C02 production as a direct measure of chain scission.A mechanism for chain scission was proposed which accountedfor these experimental observations. From the data in Table 12it is clear that .the C02 /chain scission ratio is considerablygreater than unity in the photo reaction. There appears to besome tendency for the ratio to fall as the 1A content of thecopolymer is increased, but the very small amounts of materialavailable make it difficult to obtain values of molecular weightof sufficient accuracy to study the effect with high precision.

Chain Scission and Volatilization

The relationship between chain scission and volatilizationis illustrated in Figure 18, the data in Table 10 being used.Once again, the scatter of the experimental points can be at-tributed to the difficulty of obtaining accurate values ofmolecular weig