Chapter 2 Free Radical Polymerization of Vinyl Chloride in the presence ..... 90. Table 2.9 Molecular weights and molecular weight distributions for PVCs ...... transportation, electronics, medical blood bags and prosthetic devices, wire and ...... MI. -. -. = -. -. = 1. 1. 0. 0. 0 and . Defining conversion as. 0. AXM. MII = (. ) 0. 0. M.
KINETICS AND MECHANISM OF VINYL CHLORIDE POLYMERIZATION: EFFECTS OF ADDITIVES ON POLYMERIZATION RATE, MOLECULAR WEIGHT AND DEFECT CONCENTRATION IN THE POLYMER
by KUN SI
Submitted in partial fulfillment of the requirements For the degree of Doctor of Philosophy
Dissertation Advisors: Dr. Morton Litt and Dr. Jerome Lando
Department of Macromolecular Science and Engineering CASE WESTERN RESERVE UNIVERSITY May, 2007
CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
Kun Si ______________________________________________________ candidate for the Ph.D. degree *.
Morton Litt
(signed)_______________________________________________ (chair of the committee)
Jerome B. Lando
________________________________________________
Adin J. Mann
________________________________________________
Regan Silvestri
________________________________________________
________________________________________________
________________________________________________
December 05, 2006
(date) _______________________
*We also certify that written approval has been obtained for any proprietary material contained therein.
DEDICATION
To my parents, Jinxing Si and Yuzhen Liu, who never went to college and sacrificed many things to give their all four children an opportunity of higher educations. With their unconditional love, support, understanding and patience, I can make this dissertation possible and finish the final step of my education.
謹獻給我的父母，司金星和劉玉真，他們沒有機會接受高等教育 卻送四個子女進了大學。 因他們的大愛，理解，支持與耐心，我得以 完成此論文並結束學業。
TABLE OF CONTENTS List of Tables…………………………………………………………………….
x
List of Figures……………………………………………………………………
xviii
List of Schemes…………………………………………………………………..
xxx
Acknowledgements………………………………………………………………
xxxi
List of Abbreviations & Acronyms……………………………………………
xxxii
Abstract………………………………………………………………………..
xxxv
Chapter 1 Introduction……………………………………………………........
1
1.1 Introduction……………………………………………………….........
2
1.2 A brief History of PVC………………………………………………..
3
1.3 Methods of Vinyl Chloride Polymerization……………………………
8
1.3.1 Polymerization in Bulk…………………………………………….
8
1.3.2 Polymerization in Suspension……………………………………..
10
1.3.3 Polymerization in Emulsion……………………………………….
11
1.3.4 Polymerization in Solution…………………………………...........
12
1.4 Structure and Properties of PVC………………………………….........
12
1.5 Structural Defects in PVC……………………………………………...
13
1.5.1 Short-chain Branching……………………………………………..
14
1.5.2 Long-chain Branching……………………………………………..
18
1.5.3 Unsaturation……………………………………………………….
20
1.5.4 Head-to-head Structures…………………………………………...
21
1.6 Stereoregularity of PVC………………………………………………..
21
iv
1.7 Dissertation Goals and Summary………………………………………
23
1.8 References……………………………………………………………...
27
Chapter 2 Free Radical Polymerization of Vinyl Chloride in the presence of Organic Additives……………………………………………………….
29
2.1 Introduction…………………………………………………………….
30
2.2 Experimental…………………………………………………………...
32
2.2.1 Materials…………………………………………………………...
32
2.2.2 Polymerization Apparatus…………………………………………
34
2.2.3 Bulk Polymerization of Vinyl Chloride…………………………...
35
2.2.4 Suspension Polymerization of Vinyl Chloride…………………….
37
2.2.5 GPC Characterization of PVC Samples…………………………...
38
2.2.6 Dynamic Thermal Stability Test of PVC Samples………………..
39
2.2.7 Static Thermal Stability Test of PVC Samples……………………
40
2.3 Results and Discussion…………………………………………………
40
2.3.1 Molecular Weight and Molecular Weight Distribution of PVCs….
40
2.3.1.1 Polymer aggregation in solution……………………………..
40
2.3.1.2 GPC data for bulk polymerized PVCs……………………….
43
2.3.1.3 PVC/DMT01b and PVC86k aggregation in solution………..
48
2.3.1.4 GPC data for suspension polymerized PVCs………………...
51
2.3.1.5 PVC dynamic thermal stability………………………………
53
2.3.2 Two-phase Literature Model for Bulk Polymerization of Vinyl Chloride…………………………………………………………...
v
54
2.3.3 Bulk Polymerization of Vinyl Chloride at 55 oC initiated by AIBN……………………………………………………………..
60
2.3.4 New Kinetic Model for Vinyl Chloride Bulk Polymerization initiated by AIBN…………………………………………………
65
2.3.5 Kinetic Parameters for Bulk Polymerization of Vinyl Chloride at 55 oC……………………………………………………………..
72
2.3.6 Kinetics and Mechanism of Vinyl Chloride Polymerization in the presence of Additives…………………………………………….
79
2.3.7 Bulk Polymerization of Vinyl Chloride at 55 oC in the presence of Pyridine, its Derivatives, Pyrazine, Benzothiazol, or Imidazolidinone…………………………………………………..
88
2.3.8 Bulk Polymerization of Vinyl Chloride at 55 oC in the presence of Phosphine Oxides…………………………………………………
92
2.3.9 Bulk Polymerization of Vinyl Chloride at 55 oC in the presence of Dimethoxybenzene, Carbonates, Carboxylates or Lactones……...
96
2.3.9.1 Polymerization in the presence of 1,4-Dimethoxybenzene…..
98
2.3.9.2 Polymerization in the presence of Ethylene Carbonate……...
100
2.3.9.3 Polymerization in the presence of 2-Coumaranone………….
102
2.3.9.4 Polymerization in the presence of γ-Butyrolactone…………..
106
2.3.9.5 Polymerization in the presence of Dimethyl Terephthalate….
110
2.3.9.6 Polymerization in the presence of Trimethyl 1,3,5Benzenetricarboxylate………………………………………
115
2.4 Conclusions…………………………………………………………….
119
vi
2.5 References……………………………………………………………...
121
Chapter 3 Differential Scanning Calorimetry and Thermogravimetric Analysis of Poly(vinyl chloride)s………………………………………….
125
3.1 Introduction…………………………………………………………….
126
3.2 Experimental…………………………………………………………...
128
3.2.1 Materials…………………………………………………………..
128
3.2.2 DSC Measurement of PVC Samples………………………………
129
3.2.3 TGA Measurement of PVC Samples……………………………...
129
3.3 Results and Discussion…………………………………………………
130
3.3.1 Crystallinity of PVCs prepared in the presence of Additives …….
130
3.3.1.1 Crystallinity of PVCs prepared in the presence of 2Coumaranone………………………………………………..
137
3.3.1.2 Crystallinity of PVCs prepared in the presence of 2,6Dichloropyridine…………………………………………….
140
3.3.1.3 Crystallinity of PVCs prepared in the presence of 1,4Dimethoxybenzene………………………………………….
143
3.3.1.4 Crystallinity of PVCs prepared in the presence of Dimethyl Terephthalate………………………………………………..
146
3.3.1.5 Crystallinity of PVCs prepared in the presence of γButyrolactone………………………………………………..
150
3.3.1.6 Crystallinity of PVCs prepared in the presence of Trimethyl1,3,5-Benzenetricarboxylate………………………………..
vii
152
3.3.2 Effect of Annealing Time and Temperature on the Crystallinity of PVCs……………………………………………………………...
157
3.3.2.1 Effect of Annealing on the Crystallinity of PVC/AIBN416…
158
3.3.2.2 Effect of Annealing on the Crystallinity of PVC/DMB0001a & DMB001b………………………………………………...
167
3.3.2.3 Effect of Annealing on the Crystallinity of PVC/DMT01d….
181
3.3.2.4 Effect of Annealing on the Crystallinity of PVC/EC01c…….
186
3.3.2.5 Effect of Annealing on the Crystallinity of PVC/SBP124…...
190
3.3.2.6 Effect of annealing on the crystallinity of PVC/SDMT01a….
194
3.3.3 Kinetics of PVC Crystallization……………………………...........
202
3.3.4 Thermal Degradation Behavior of PVC Samples…………………
208
3.3.4.1 Evaluation of Degradation Activation Energy of PVC samples by single-heating-rate TGA Curve…………………
215
3.3.4.2 Evaluation of Degradation Activation Energy of PVC samples by FR, WFO and KAS methods……………............
219
3.3.4.3 Evaluation of Degradation Kinetic Parameters (Ea, A, n) of PVC samples by Coats-Redfern method………....................
229
3.4 Conclusions……………………………………………………….........
235
3.5 References……………………………………………………………...
238
Chapter 4 Correlation between Structural Defects and the Dehydrochlorination of Poly(vinyl chloride)s……………………………
243
4.1 Introduction…………………………………………………………….
244
viii
4.2 Experimental…………………………………………………………...
246
4.2.1 Materials…………………………………………………………...
246
4.2.2 Dehydrochlorination Measurements of PVC samples…………….
247
4.2.3 1D & 2D NMR Measurements of PVC Samples………………….
248
4.3 Results and Discussion…………………………………………………
249
4.3.1 Two-parameter Kinetic Model for PVC Dehydrochlorination……
249
4.3.2 Dehydrochlorination of PVCs prepared in Bulk without Additives
257
4.3.3 Dehydrochlorination of PVCs prepared in Bulk in the presence of Additives………………………………………………………….
259
4.3.3.1 Dehydrochlorination of PVCs prepared in the presence of 2Coumaranone……………………………………………….
259
4.3.3.2 Dehydrochlorination of PVCs prepared in the presence of 2,6-Dichloropyridine………………………………………..
262
4.3.3.3 Dehydrochlorination of PVCs prepared in the presence of 1,4-Dimethoxybenzene……………………………………...
264
4.3.3.4 Dehydrochlorination of PVCs prepared in the presence of Dimethyl Terephthalate…………………………………….
266
4.3.3.5 Dehydrochlorination of PVCs prepared in the presence of γButyrolactone……………………………………………….
268
4.3.3.6 Dehydrochlorination of PVCs prepared in the presence of Trimehtyl-1,3,5-Benzenetricarboxylate……………………..
270
4.3.4 Dehydrochlorination of PVCs prepared in Suspension……………
272
4.3.5 Activation energies for PVC Dehydrochlorination at early stage…
275
ix
4.3.6 Identification of PVC Structural Defects using 1D &2D NMR…..
281
4.3.7 Determination of PVC Structural Defects using 1D 1H NMR…….
294
4.3.8 Correlation between the Dehydrochlorination Rate and the Structural Defects of PVCs……………………………………….
300
4.3.9 PVC tacticity………………………………………………………
303
4.4 Conclusions…………………………………………………………….
305
4.5 References……………………………………………………………...
306
Chapter 5 Conclusions……………………………………………………….....
311
Appendix………………………………………………………………………...
314
Bibliography…………………………………………………………………….
332
LIST OF TABLES Chapter 2 Table 2.1
A selected recipe for suspension polymerization of vinyl chloride at 53 oC…………………………………………………………….
Table 2.2
38
Molecular weights and molecular weight distributions for PVCs prepared by bulk polymerization at 55 oC in the presence of organic additives…………………………………………………...
Table 2.3
45
Suspension polymerization of vinyl chloride at 53 oC initiated by diisobutyl peroxydicarbonate in the presence of γ-butyrolactone,
Table 2.4
dimethyl phthalate or dimethyl terephthalate ……………………..
52
Bulk polymerization of vinyl chloride initiated by AIBN at 55oC...
61
x
Table 2.5
Molecular weights and molecular weight distributions for PVCs prepared by bulk polymerization at 55 oC initiated by AIBN……..
Table 2.6
Chain transfer constant to monomer for vinyl chloride polymerization……………………………………………………..
Table 2.7
73
Summary of kinetic parameters for bulk polymerization of vinyl chloride initiated by 2,2’-azobisisobutyronitrile at 55 oC…………
Table 2.8
62
78
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of pyridine, its derivatives, pyrazine, benzothiazol, and imidazolidinone……………………………………………….
Table 2.9
90
Molecular weights and molecular weight distributions for PVCs prepared at 55oC in the presence of pyridine, its derivatives, pyrazine, benzothiazol, and imidazolidinone………………………
Table 2.10
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of phosphine oxides……………………………….
Table 2.11
100
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of ethylene carbonate……………………………..
Table 2.15
99
Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of DMB…….
Table 2.14
95
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of 1,4-dimethoxybenzene (DMB)…………………
Table 2.13
94
Molecular weight and molecular weight distributions for PVCs prepared at 55oC in the presence of phosphine oxides…………….
Table 2.12
91
Molecular weights and molecular weight distributions for PVCs
xi
101
prepared at 55oC in the presence of ethylene carbonate……........... Table 2.16
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of 2-coumaranone (CMN)…………………………
Table 2.17
117
Values of K, ktr2/ktr1 and CA for vinyl chloride polymerization at 55 o
C in the presence of 2-Coumaranone, dimethyl terephthalate, γ-
butyro-lactone or trimethyl 1,3,5-benzenetricarboxylate…………..
Chapter 3 Table 3.1
117
Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of TMB……..
Table 2.24
114
Bulk polymerization of vinyl chloride at 55 oC initiated by AIBN in the presence of trimethyl 1,3,5-benzenetricarboxylate (TMB)….
Table 2.23
113
Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of DMT…….
Table 2.22
110
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of dimethyl terephthalate (DMT)………………….
Table 2.21
109
Molecular weights and molecular weight distributions for PVCs prepared at 55oC initiated by AIBN in the presence of GBL………
Table 2.20
106
Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of γ-butyrolactone (GBL)…………………………
Table 2.19
105
Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of CMN…….
Table 2.18
101
DSC difference curve data for PVCs prepared in bulk at 55 oC in
xii
119
the absence of additives……………………………………………. Table 3.2
DSC difference curve data for PVCs prepared at 55 oC in the presence of 2-coumaranone (CMN)………………………………..
Table 3.3
165
DSC difference curve data for PVC/AIBN416 annealed at 125 oC for different periods of time………………………………………..
Table 3.12
161
DSC difference curve data for PVC/AIBN416 annealed at 100 oC for different periods of time……………………………………….
Table 3.11
158
DSC difference curve data for PVC/AIBN416 annealed for 120 minutes at different temperatures………………………………….
Table 3.10
154
Weight-average molecular weights and the tacticity of selected PVCs used for the crystallization kinetics studies…………………
Table 3.9
151
DSC difference curve data for PVCs prepared at 55 oC in the presence of trimethyl-1,3,5-benzenetricarboxylate (TMB)………..
Table 3.8
148
DSC difference curve data of PVCs prepared at 55 oC in the presence of γ-butyrolactone (GBL)……………………………….
Table 3.7
145
DSC difference curve data for PVCs prepared at 55 oC in the presence of dimethyl terephthalate (DMT)………………………..
Table 3.6
142
DSC difference curve data for PVCs prepared at 55 oC in the presence of 1,4-dimethoxybenzene (DMB)………………………..
Table 3.5
139
DSC data for PVCs prepared at 55 oC in the presence of 2,6dichloropyridine (DCPY)………………………………………….
Table 3.4
136
165
DSC difference curve data for PVC/DMB0001a annealed for 60 minutes at different temperatures…………………………………..
xiii
170
Table 3.13
DSC difference curve data for PVC/DMB0001b annealed for 60 minutes at different temperatures…………………………………..
Table 3.14
DSC difference curve data for PVC/DMB0001a annealed at 100 o
Table 3.15
C for different periods of time…………………………………….
C for different periods of time…………………………………….
C for different periods of time…………………………………….
189
191
DSC difference curve data for PVC/SBP124 annealed at 125 oC for different periods of time………………………………………..
Table 3.24
187
DSC difference curve data for PVC/SBP124 annealed for 60 minutes at different temperatures…………………………………..
Table 3.23
185
DSC difference curve data for PVC/EC01c annealed at 125 oC for different periods of time……………………………………………
Table 3.22
183
DSC difference curve data for PVC/EC01c annealed for 60 minutes at different temperatures…………………………………..
Table 3.21
180
DSC difference curve data for PVC/DMT01d annealed at 125 oC for different periods of time………………………………………..
Table 3.20
179
DSC difference curve data for PVC/DMT01d annealed for 60 minutes at different temperatures………………………………….
Table 3.19
176
DSC difference curve data for PVC/DMB0001b annealed at 125 o
Table 3.18
174
DSC difference curve data for PVC/DMB0001b annealed at 100 o
Table 3.17
C for different periods of time…………………………………….
DSC difference curve data for PVC/DMB0001a annealed at 125 o
Table 3.16
171
DSC difference curve data for PVC/SDMT01a annealed for 60
xiv
193
minutes at different temperatures………………………………….. Table 3.25
DSC difference curve data for PVC/SDMT01a annealed at 125 oC for different periods of time………………………………………..
Table 3.26
196
198
Observed maximum crystallization temperature ((Tc)max) and maximum crystallinity((Xc)max ) for some PVC samples annealed at (Tc)max for 60 minutes…………………………………………...
Table 3.27
Equilibrium values of crystallinity (Xc) and melting temperature (Tm) for some PVCs annealed at 125 oC………………………….
Table 3.28
200
202
Results of the exponent n+1, crystallization rate constant K/(n+1), crystallization half-time t1/2, and heat of fusion at infinity ΔH∞ for some PVC samples…………………………………………………
Table 3.29
206
The activation Energies (Ea) and the degradation temperatures for some PVC samples at different percentage of weight loss at a heating rate of 20 oC/min…………………………………………..
Table 3.30
Degradation temperatures of PVC/CMN001a under different degrees of degradation at heating rates of 5 , 10 and 15 oC/min......
Table 3.31
224
Degradation temperatures of PVC/EC01c with different degrees of degradation (α) at heating rates of 2, 5 , 10, 15 oC/min……………
Table 3.34
223
Apparent Activation Energies (Ea) for PVC/CMN001a calculated by three isoconversional methods………………………………….
Table 3.33
222
The derivative weight (dα/dt) of PVC/CMN001a under different degrees of degradation (α) at heating rates of 5 , 10 and 15 oC/min
Table 3.32
217
The derivative weight (dα/dt) of PVC/EC01c with different xv
227
degrees of degradation (α) at heating rates of 2, 5 , 10, 15 oC/min Table 3.35
Apparent Activation Energies (Ea) for PVC/EC01c calculated by three isoconversional methods……………………………………..
Table 3.36
228
229
Comparison of the kinetic parameters for PVC degradation obtained by different authors……………………………………….
233
Chapter 4 Table 4.1
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in bulk using 2,2’-Azobisisobutyronitrile (AIBN) as initiator……..
Table 4.2
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 2-coumaranone (CMN)……………….
Table 4.3
269
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of trimehtyl-1,3,5-benzenetricarboxylate (TMB)….
Table 4.8
267
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of γ-butyrolactone (GBL)……………….
Table 4.7
265
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of dimethyl terephthalate (DMT)………..
Table 4.6
263
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 1,4-dimethoxybenzene (DMB)……….
Table 4.5
260
Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 2,6-dichloropyridine (DCPY)………...
Table 4.4
258
Values of Ka and Kb for the dehydrochlorination of PVCs prepared
xvi
271
in suspension in the presence or absence of dimethyl terephthalate (DMT) using diisobutyl peroxydicarbonate (SBP) as initiator……. Table 4.9
Values of Ka and activation energies (Ea) of Ka for the dehydrochlorination of PVCs in the temperature range of 170-200 oC…….
Table 4.10
280
Hydrogen and carbon chemical shift assignments for PVC end groups, head-to-head addition segments and short branches………
Table 4.13
279
Values of Kb and activation energies (Ea) of Kb for the dehydrochlorination of PVCs in the temperature range 170-200 oC………
Table 4.12
278
Comparison of the activation energies (Ea) for PVC dehydrochlorination obtained by different authors…………………………
Table 4.11
274
292
Number of 4-chloro-2-butenyl end groups, 1,2-dichloroethyl end groups and chloromethyl branches per 1000 monomeric units in PVCs determined by 1D 1H-NMR………………………………..
Table 4.14
295
Number of internal allylic chlorines (I), external_1 allylic chlorines (E) and tertiary chlorines (T) per 1000 monomeric units in PVCs determined by 1D 1H-NMR……………………………..
Table 4.15
296
Number of 4-chloro-2-butenyl end groups, 1,2-dichloroethyl end groups and chloromethyl branches per macromolecule in PVCs determined by 1D 1H-NMR……………………………………….
Table 4.16
297
Number of internal allylic chlorines (I), external_1 allylic chlorines (E) and tertiary-carbon chlorinesa (T) per macromolecule
Table 4.17
in PVCs determined by 1D 1H-NMR……………………………..
298
Tacticity and the number-average molecular weight (Mn) of PVCs
304
xvii
LIST OF FIGURES Chapter 1 Figure 1.1
Evolution of world PVC production……………………………….
7
Figure 1.2
2003 World PVC applications……………………………………..
8
Figure 1.3
Schematic representations of (a) isotactic, (b) syndiotactic, and (c) atactic chains of PVC………………………………………………
22
Chapter 2 Figure 2.1
GPC traces of PVC (control) prepared by bulk polymerization at 55 oC, initiated by AIBN…………………………………………..
42
Figure 2.2
Thermal stability test for some PVC samples……………………..
47
Figure 2.3
GPC traces of PVC/DMT01b prepared by bulk polymerization at 55 oC, initiated by AIBN in the presence of 1.0 mole% of DMT….
48
Figure 2.4
GPC traces of PVC/DMT01b and PVC/control……………………
49
Figure 2.5
GPC traces of PVC86k prepared by bulk polymerization at room temperature, initiated by visible light………………………………
Figure 2.6
Dynamic Thermal Stability test of selected PVC samples as a function of time at 190 oC…………………………………………
Figure 2.7
53
The plot of Ln(R0p) as a function of Ln([AIBN]) for bulk polymerization of vinyl chloride at 55 oC………………………….
Figure 2.8
50
64
Plots of 1 / Pn0 and R p0 vs [AIBN]0.5 for bulk polymerization of vinyl chloride at 55 oC……………………………………………..
xviii
73
Figure 2.9
Plot of ( Pn0 ' − Pn0 ) /( 2 Pn0 − Pn0 ' ) vs R p0 for bulk polymerization of vinyl chloride at 55 oC……………………………………………..
77
Figure 2.10
DMB effect on polymerization of vinyl chloride (VC) at 55 oC…..
98
Figure 2.11
Plots of 1 / Pn0 and R 0p vs [CMN]/[VC] for bulk polymerization of vinyl chloride in the presence of CMN at 55 oC…………………..
Figure 2.12
Plots of 1 / Pn0 and R 0p vs [GBL]/[VC] for bulk polymerization of vinyl chloride in the presence of GBL at 55 oC……………………
Figure 2.13
108
Plots of 1 / Pn0 and R 0p vs [DMT]/[VC] for bulk polymerization of vinyl chloride in the presence of DMT at 55 oC…………………..
Figure 2.14
104
112
Plots of 1 / Pn0 and R 0p vs [TMB]/[VC] for bulk polymerization of vinyl chloride in the presence of TMB at 55 oC…………………..
116
Chapter 3 Figure 3.1
DSC curves for PVC/AIBN416 recorded at the heating rate of 20 o
C/min………………………………………………………………
131
Figure 3.2
DSC curves for control PVCs prepared in the absence of additives
133
Figure 3.3
DSC Difference curves for control PVCs………………………….
134
Figure 3.4
DSC curves of PVCs prepared at 55 oC in the presence of indicated mole fraction (in ppm) of 2-coumaranone ………………
Figure 3.5
Figure 3.6
137
DSC difference curves for PVCs prepared in the presence of indicated mole fraction (in ppm) of 2-coumaranone………………
138
Effect of CMN on the crystallinity for the resulting PVCs………..
139
xix
Figure 3.7
DSC curves for PVCs prepared in the presence of indicated mole percent of 2,6-dichloropyridine……………………………………
Figure 3.8
140
DSC difference curves for PVCs prepared in presence of indicated mole percent of 2,6-dichloropyridine………………………………
141
Figure 3.9
Effect of DCPY on the crystallinity for the resulting PVCs……….
142
Figure 3.10
DSC curves forPVCs prepared in the presence of indicated mole fraction (in ppm) of 1,4-dimethoxybenzene……………………….
Figure 3.11
144
DSC difference curves for PVCs prepared in the presence of indicated mole fraction (in ppm) of 1,4-dimethoxybenzene……….
144
Figure 3.12
Effect of DMB on the crystallinity for the resulting PVCs………...
145
Figure 3.13
DSC curves for PVCs prepared in the presence of indicated mole percent of dimethyl terephthalate………………………………….
Figure 3.14
147
DSC difference curves for PVCs prepared in the presence of indicated mole percent of dimethyl terephthalate………………….
147
Figure 3.15
Effect of DMT on the crystallinity for the resulting PVCs………...
149
Figure 3.16
DSC curves for PVCs prepared in the presence of indicated mole percent of γ-butyrolactone………………………………………….
Figure 3.17
150
DSC difference curves ofor PVCs prepared in the presence of indicated mole percent of γ-butyrolactone…………………………
151
Figure 3.18
Effect of GBL on the crystallinity for the resulting PVCs…………
152
Figure 3.19
DSC curves of PVCs prepared in the presence of indicated mole percent of trimethyl-1,3,5-benzenetricarboxylate………………….
Figure 3.20
DSC difference curves ofor PVCs prepared in the presence of
xx
153
indicated mole percent of trimethyl-1,3,5-benzenetricarboxylate…
154
Figure 3.21
Effect of TMB on the crystallinity for the resulting PVCs………..
155
Figure 3.22
Plot of crystallinity as a function of molecular weights for PVCs prepared in the presence of CMN, DCPY, DMB, DMT, GBL and TMB………………………………………………………………..
Figure 3.23
DSC curves for quenched PVC/AIBN416 samples annealed at the indicated temperature for 120 minutes…………………………….
Figure 3.24
158
DSC difference curves for quenched PVC/AIBN416 samples annealed at the indicated temperature for 120 minutes…………….
Figure 3.25
157
159
Effect of annealing temperatures (Ta) on melting temperatures (Tm), glass translation temperatures (Tg) and crystallinity (Xc) for PVC/AIBN416 samples……………………………………………
Figure 3.26
DSC curves for PVC/AIBN416 samples annealed at 100 oC for the indicated periods of time followed by quenching………………….
Figure 3.27
C for the indicated periods of time………………………………..
164
DSC difference curves for quenched PVC/AIBN416 samples annealed at 125 oC for the indicated periods of time………………
Figure 3.30
163
DSC curves for quenched PVC/AIBN416 samples annealed at 125 o
Figure 3.29
163
DSC difference curves for PVC/AIBN416 samples annealed at 100 oC for the indicated periods of time followed by quenching….
Figure 3.28
161
164
Effect of annealing time on melting temperature (Tm) and crystallinity (Xc) for PVC/AIBN416 annealed at 100 oC, and 125 o
C, respectively…………………………………………………….
xxi
166
Figure 3.31
Melting temperature (Tm) and crystallinity (Xc) as a function of logarithm of annealing time for PVC/AIBN416 annealed at 100 o
Figure 3.32
C, and 125 oC, respectively……………………………………….
DSC curves for quenched PVC/DMB0001a samples annealed at the indicated temperature for 60 minutes…………………………..
Figure 3.33
175
DSC difference curves for quenched PVC/DMB0001a annealed at 125 oC for the indicated periods of time…………………………...
Figure 3.41
174
DSC curves for quenched PVC/DMB0001a annealed at 125 oC for the indicated periods of time……………………………………….
Figure 3.40
173
DSC difference curves for quenched PVC/DMB0001a samples annealed at 100 oC for the indicated periods of time………………
Figure 3.39
172
DSC curves for quenched PVC/DMB0001a annealed at 100 oC for the indicated periods of time……………………………………….
Figure 3.38
169
Effect of annealing temperatures (Ta) on melting temperatures (Tm) and crystallinity (Xc) for PVC/DMB0001a and DMB0001b..
Figure 3.37
169
DSC difference curves for quenched PVC/DMB0001b samples annealed at the indicated temperature for 60 minutes……………...
Figure 3.36
168
DSC curves for quenched PVC/DMB0001b samples annealed at the indicated temperature for 60 minutes…………………………..
Figure 3.35
168
DSC difference curves for quenched PVC/DMB0001a samples annealed at the indicated temperature for 60 minutes……………..
Figure 3.34
166
Effects of annealing time on melting temperature (Tm) and crystallinity (Xc) for PVC/DMB0001a annealed at 100 oC, and
xxii
175
125 oC, respectively……………………………………………….. Figure 3.42
DSC curves for quenched PVC/DMB0001b annealed at 100 oC for the indicated periods of time………………………………………
Figure 3.43
179
DSC difference curves for quenched PVC/DMB0001b samples annealed at 125 oC for the indicated periods of time………………
Figure 3.46
178
DSC curves for quenched PVC/DMB0001b annealed at 125 oC for the indicated periods of time……………………………………….
Figure 3.45
178
DSC difference curves for quenched PVC/DMB0001b annealed at 100 oC for the indicated periods of time…………………………..
Figure 3.44
176
180
Effects of annealing time on melting temperature (Tm) and crystallinity (Xc) for PVC/DMB0001b annealed at 100 oC, and 125 oC, respectively………………………………………………..
Figure 3.47
DSC difference curves for quenched PVC/DMT01d annealed at indicated temperature for 60 minutes………………………………
Figure 3.48
181
182
Plots of glass transition temperature (Tg), melting temperature (Tm) and crystallinity (Xc) of PVC/DMT01d as a function of annealing temperature……………………………………………...
Figure 3.49
DSC difference curves for quenched PVC/DMT01d annealed at 125 oC for the indicated periods of time…………………………...
Figure 3.50
183
184
Plots of glass transition temperature (Tg), melting temperature (Tm) and crystallinity (Xc) for PVC/DMT01d as a function of annealing time……………………………………………………...
Figure 3.51
DSC difference curves for quenched PVC/EC01c samples
xxiii
185
annealed at the indicated temperature for 60 minutes…………….. Figure 3.52
186
Plots of glass transition temperature (Tg), melting temperature (Tm) and crystallinity (Xc) of PVC/EC01c sample as a function of annealing temperature……………………………………………..
Figure 3.53
DSC difference thermograms for quenched PVC/EC01c annealed at 125 oC for the indicated periods of time…………………………
Figure 3.54
187
188
Plots of glass transition temperature (Tg), melting temperature (Tm) and crystallinity (Xc) of PVC/EC01c as a function of annealing time……………………………………………………...
Figure 3.55
DSC difference curves for quenched PVC/SBP124 samples annealed at the indicated temperature for 60 minutes……………..
Figure 3.56
189
191
Plots of the glass transition temperature (Tg), the melting temperature (Tm) and the crystallinity (Xc) of PVC/SBP124 as a function of annealing temperature…………………………………
Figure 3.57
DSC curves for quenched PVC/SBP124 annealed at 125 oC for the indicated periods of time…………………………………………..
Figure 3.58
192
193
Plots of the glass transition temperature (Tg), the melting temperature (Tm) and the crystallinity (Xc) of PVC/SBP124 as a function of annealing time…………………………………………
Figure 3.59
DSC curves for quenched PVC/SDMT01a annealed at the indicated temperature for 60 minutes………………………………
Figure 3.60
194
Plots of the glass transition temperature (Tg), the melting temperature (Tm) and the crystallinity (Xc) for PVC/SDMT01a as
xxiv
195
a function of annealing temperature………………………………. Figure 3.61
DSC difference curves for quenched PVC/SDMT01a samples annealed at 125 oC for the indicated periods of time………………
Figure 3.62
197
198
Plots of the glass transition temperature (Tg), the melting temperature (Tm) and the crystallinity (Xc) of PVC/SDMT01a samples as a function of annealing time……………………………
Figure 3.63
199
Plots of Ln(-Ln(1-Xt)) against Ln( t ) for PVC/AIBN416, DMB0001a, and DMB0001b isothermally crystallized at 100 oC up to 3600 seconds…………………………………………………
Figure 3.64
204
Plots of Ln(-Ln(1-Xt)) against Ln( t ) for PVC/AIBN416, DMB0001a, and DMB0001b isothermally crystallized at 125 oC up to 3600 seconds…………………………………………………
Figure 3.65
204
Plots of Ln(-Ln(1-Xt)) against Ln( t ) for PVC/DMT01d, EC01c, SDMT01a, and SBP124 isothermally crystallized at 125 oC up to 3600 seconds……………………………………………………….
Figure 3.66
205
Modified Avrami overall plot of Ln(-Ln(1-Xt)) against Ln( t ) for PVC/AIBN416, DMB0001a, DMB0001b, DMT01d, EC01c, SDMT01a, and SBP124 isothermally crystallized at 100 oC or 125 o
Figure 3.67
C for up to 3600 seconds………………………………………….
207
The overall variation of Tg and Tm with annealing temperature (Ta) for AIBN416, DMB0001a, DMB0001b, DMT01d, EC01c, SDMT01a, and SBP124……………………………………………
Figure 3.68
TGA and negative derivative TGA curves for purified PVC/EC01c
xxv
207
sample at the heating rate of 20 oC/min under nitrogen atmosphere Figure 3.69
208
TG and DTG curves for PVC/AIBN319, AIBN416, PVC40k, PVC86k at the heating rate of 20 oC/min under nitrogen atmosphere…………………………………………………………
Figure 3.70
215
TG and DTG curves for PVC/CMN00025b, DMB00005b, DMB0001a, DMT01b at the heating rate of 20 oC/min under nitrogen atmosphere………………………………………………..
Figure 3.71
Application of Coats-Redfern method (eq. 3.17) for degradation of PVC/AIBN319, AIBN416, PVC40k, and PVC86k……………….
Figure 3.72
218
Application of Coats-Redfern method (eq. 3.17) for degradation of PVC/CMN00025b, DMB00005b, DMB0001a, and DMT01b…….
Figure 3.73
216
218
Thermogravimetric (TG) and negative first derivative thermogravimetric (DTG) curves for PVC/CMN001a at heating rates of 5, 10 and 15 oC/min, respectively………………………….
Figure 3.74
220
The plots of natural logarithm of heating rate(β) versus the reciprocal absolute temperature for the PVC/CMN001a degradation at different weight loss(%)……………………………
Figure 3.75
Apparent activation energies (Ea) for the degradation of PVC/CMN001a and EC01c as a function of degree of degradation
Figure 3.76
221
224
Thermogravimetric (TG) and negative first derivative thermogravimetric (DTG) curves of PVC/EC01c at heating rates of 2, 5, 10, 15 oC/min………………………………………………
Figure 3.77
The plots of natural logarithm of heating rate(β) versus the
xxvi
225
reciprocal absolute temperature for the PVC/EC01c degradation at the different weight loss (%)………………………………………. Figure 3.78
Plots of Ln(Af (α)) as a function of Ln(1-α) for PVC/CMN001a and EC01c………………………………………………………….
Figure 3.79
227
230
Experimental and simulated TG and DTG curves for PVC/CMN001a thermal degradation at heating rates of 5, 10, 15 K/min………………………………………………………………
Figure 3.80
231
Experimental and simulated TG curves for PVC/EC01c thermal degradation at heating rates of 2, 5, 10, 15 K/min…………………
232
Chapter 4 Figure 4.1
Dehydrochlorination curves for PVC/AIBN416 at 190 oC under fast N2 flow………………………………………………………..
Figure 4.2
Kinetic curves for the dehydrochlorination of PVC/AIBN416 at 190 oC under fast nitrogen flow……………………………………
Figure 4.3
C under fast nitrogen flow………………………………………...
261
Kinetic dehydrochlorination curves for PVCs prepared in the presence of 2,6-dichloropyridine (DCPY)…………………………
Figure 4.6
257
Effect of CMN on Ka for the dehydrochlorination of the resulting PVCs……………………………………………………………….
Figure 4.5
254
Kinetic curves for the dehydrochlorination of control PVCs at 190 o
Figure 4.4
250
Effect of DCPY on Ka for the dehydrochlorination of the resulting
xxvii
263
PVCs……………………………………………………………….. Figure 4.7
Effect of DMB on Ka for the dehydrochlorination of the resulting PVCs………………………………………………………………..
Figure 4.8
Figure 4.17
276
Arrhenius plots of Ln(Ka) as a function of reciprocal temperature for some representative PVCs…………………………………….
Figure 4.16
273
Kinetic curves for the dehydrochlorination of PVC/AIBN319 in the temperature range of 170-200 oC………………………………
Figure 4.15
272
Kinetic dehydrochlorination curves for suspension PVCs at 190 oC under fast nitrogen flow……………………………………………
Figure 4.14
271
Effect of TMB on Ka for the dehydrochlorination of the resulting PVCs………………………………………………………………..
Figure 4.13
270
Kinetic dehydrochlorination curves for PVCs prepared in the presence of trimehtyl-1,3,5-benzenetricarboxylate (TMB)………..
Figure 4.12
269
Effect of GBL on Ka for the dehydrochlorination of the resulting PVCs……………………………………………………………….
Figure 4.11
268
Kinetic dehydrochlorination curves for PVCs prepared in the presence of γ-butyrolactone (GBL)………………………………..
Figure 4.10
266
Effect of DMT on Ka for the dehydrochlorination of the resulting PVCs………………………………………………………………..
Figure 4.9
264
278
Arrhenius plots of Ln(Kb) as a function of reciprocal temperature for some representative PVCs……………………………………..
279
150.9 MHz 13C NMR spectra of PVC/SBP160 in THF-d8 at 45 oC
282
xxviii
Figure 4.18
600 MHz 1D 1H (A) and 2D 1H-1H COSY (B) NMR spectra of PVC/DMT01b in THF-d8 at 50 oC with the corresponding correlation assignments……………………………………………
Figure 4.19
284
Contour Plots of 1H-13C gHMQC NMR spectrum of PVC/DMT01b sample in THF-d8 at 45 oC in the regions of (1H) 3.1-4.1 ppm and (13C) 34.0-50.0 ppm with one-bond correlation assignments…………………………………………………………
Figure 4.20
287
Contour Plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b sample in THF-d8 at 45 oC in the regions of (1H) 3.5-4.0 ppm and (13C) 25.0-75.0 ppm with long-range correlation assignments.…………………………..............................................
Figure 4.21
288
Contour plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b in THF-d8 at 45 oC in the regions of (1H) 2.5-4.2 ppm and (13C) 20.0-134.0 ppm with long-range correlation assignments…………………………................................................
Figure 4.22
289
Contour Plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b in THF-d8 at 45 oC in the regions of (1H) 3.4-4.9 ppm and (13C) 15.0-103.0 ppm with long-range correlation assignments.………………………………………………………..
Figure 4.23
290
Correlation between the dehydrochlorination rate (dHCle/dt) and the labile chlorine concentrations in PVCs………………………...
xxix
302
LIST OF SCHEMES Chapter 1 Scheme 1.1
Mechanistic sequel of head-to-head addition during the free radical polymerization of vinyl chloride…………………………...
15
Scheme 2.1
Schematic illustration of Ace-8648 Pressure Tube………………..
34
Scheme 2.2
Schematic of the polymerization set-up with a vacuum line and
Chapter 2
pump system………………………………………………………. Scheme 2.3
35
Chemical structures for 1,3-dimethyl-2-imidazolidinone, 2methyl-benzothiazole, 2-methylpyrazine, pyridine, 2,6-dichloropyridine, and 2,4,6-trimethylpyridine……………………………..
Scheme 2.4
89
Chemical structures for trimethylphosphine oxide, triethylphosphine oxide, tributylphosphine oxide, and triphenylphosphine oxide………………………………………….
Scheme 2.5
93
Chemical structures for 1,4-dimethoxybenzene, ethylene carbonate, 2-coumaranone, γ-butyrolactone, dimethyl terephthalate and trimethyl-1,3,5-benzenetricarboxylate………….
xxx
97
ACKNOWLEDGEMENTS
I would like to express my deepest gratitude to my advisors, Dr. Morton Litt and Dr. Jerome Lando, for their guidance, support and patience throughout the course of this research. I enjoyed every stimulating discussion with them and their instructions are essential for my growth as a scientific researcher. This dissertation would not have been completed without their mentoring, comments and a dozen times of correction of my manuscript. I would also like to thank Dr. Charles Wilkes for supervising this project and encouraging me in this study. The generous financial support of the Edison Polymer Innovation Corporation (EPIC) is gratefully acknowledged. A special appreciation goes to Dr. Virgil Percec for his first accepting me in his research group.
I would like to thank Dr. Ross Cozens, at OxyVinyl, for his invaluable help in PVC dehydrochlorination measurements and many fruitful discussions. My sincere gratitude goes to Dr. Dale Ray III and Prof. Peter Rinaldi for 1D and 2D NMR measurements. Dr. Ray helped me set up all the sequences for 2D NMR measurements conducted in this research and offered tremendous helps in processing and analyzing the NMR results.
I am also indebted to Dr. Liqun Zhang from Beijing University of Chemical Technology for kindly accepting me in his research group when I was stranded in China for 6 months in 2003. Thanks also go to Dr. Yushan Hu, Dr. Yi Jin, Dr. Jianhua Fang, Ms. Lin Yang and Ms. Songtao Mo for their generous helps during my difficult time. xxxi
I give my special thanks to our foreign student advisors, Ms. Edith Berger and Ms. Elise Lindsay, who, like our mom, give us unlimited loves and invaluable helps during our student years at Case. Ms. Edith Berger passed away on Nov 07, 2006, after an illness. Thank you so much for everything you have done for me, Ms. Berger. You issued me so many I-20s and marked so many notes in my file. I will miss you forever.
Many thanks go to our department technician Alex Schnittlinger for his technical assistance during the past years and everyone in Dr. Litt and Dr. Lando research groups for their help and support. Last, but not least, I am indebted to all my close friends for their great friendship during the years. I would like to list some of them here: Dr. Ling Cheng, Dr. Hu Duan, Ms. Tao Hui, Dr. Hong Guo, Ms. Jinxiu Sun, Dr. Yangyang Sun, Dr. Danling Wang, Dr. Hong Wang, Dr. Guozhang Xu, Dr. Hongli Yang, Ms. Jing Yang, Ms. Ling Yang, Dr. Wenxin Yu, Dr. Yan Zhang, Dr. Zhongfa Zhang, Dr. Yiqiang Zhao, Dr. Yuangang Zheng. These friends always gave me their helping hands during times of need and discomfort, making my life enjoyable while in USA.
LIST OF ABBREVIATIONS & ACRONYMS 1D
one-dimensional
2D
two-dimensional
AIBN
2,2’-Azobisisobutyronitrile
BHT
2,6-di-tert-butyl-4-methylphenol
Calc
calculated
CHX
cyclohexanone xxxii
CMN
2-coumaranone
COSY
homonuclear scalar coupling spectroscopy
DCPY
2,6-dichloro-pyridine
DEHA
diethylhydroxylamine
DMB
1,4-dimethoxybenzene
DMI
1,3-dimethyl-2-imidazolidinone
DMT
dimethyl terephthalate
DSC
differential scanner calorimetroy
Ea
activation energy
EC
ethylene carbonate
Exp
experimental
GBL
γ-butyrolactone
gHMBC
pulsed-field gradient heteronuclear multiple-bond correlation
gHMQC
pulsed-field gradient heteronuclear multiple quantum coherence
GPC
gel permeation chromatography
HCl
hydrogen chloride
HCle
evolved hydrogen chloride during PVC dehydrochlorination
HCls
hydrogen chloride dissolved in solid PVC during dehydrochlorination
H-H
head-to-head
HMBC
heteronuclear multiple-bond correlation
HMQC
heteronuclear multiple quantum coherence
MBTZ
2-methyl-benzothiazole
Mn
number-average molecular weight
xxxiii
MPZ
2-methylpyrazine
Mw
weight-average molecular weight
NMR
nuclear magnetic resonance
PFG
pulsed-field gradient
Pn
degree of polymerization
PVC
poly(vinyl chloride)
PYP
pyridine
Rp
polymerization rate
SBP
diisobutyl peroxydicarbonate
Ta
annealing temperature
TBPO
tributylphosphine oxide
Tc
crystallization temperature
TEPO
triethylphosphine oxide
Tg
glass transition temperature
TGA
thermogravimetric analysis
THF
tetrahydrofuran
Tm
melting temperature
TMB
trimethyl-1,3,5-benzene tricarboxylate
TMPO
trimethylphosphine oxide
TMPY
2,4,6-trimethylpyridine
TPPO
triphenylphosphine oxide
VC
vinyl chloride
Xc
crystallinity
xxxiv
Kinetics and Mechanism of Vinyl Chloride Polymerization: Effects of Additives on Polymerization Rate, Molecular Weight and Defect Concentration in the Polymer
Abstract by KUN SI
Poly(vinyl chloride), or PVC, has various defects which limit its processing temperature by lowering thermal stability. Possible structural defects in PVC are short and long chain branches, chloroallyl groups, end groups and head-to-head structures. Some structural defects connect with an active chloride, which easily loses HCl at elevated temperatures. In this thesis, a comprehensive study was performed on the vinyl chloride polymerization in the presence of small amount of organic additives. These additives were weakly basic compounds, high dipole carbonyl compounds, ether compounds and some heteroraromatics. The initial polymerization rates and the molecular weights of the resulting polymers increased in the presence of weakly ‘basic’ compounds such as dimethyl terephthalate (DMT), ethylene carbonate (EC), γbutyrolactone (GBL), tributylphosphine oxide (TBPO) and trimethyl-1,3,5-benzene tricarboxylate (TMB). A kinetic model was developed for vinyl chloride polymerization in the presence of these weakly basic additives, using the hypothesis that a hydrogenxxxv
bond complex formed between an additive and the terminal hydrogen of the propagating radical, and that the major termination reaction was between a small radical (possibly Cl• or combined Cl•) and the propagating radical. The kinetic model successfully explained the increase of the polymerization rate and the resulting polymer molecular weights. An optimal additive concentration exists to get maximum molecular weight increase and possibly reducing structural defects of the resulting polymers.
Various methods were applied to evaluate the additive effects on the resulting polymer structures. Differential Scanning Calorimetry (DSC) was used to study the crystallization behavior of the resulting polymers. The thermal stability of the resulting polymers was evaluated by dynamic Thermogravimetric Analysis (TGA) and dehydrochlorination. A 2-parameter model was developed to describe the initial dehydrochlorination of the polymers at 170-200oC with rapid removal of HCl. The additive effect on the thermal stability of the resulting polymers was evaluated. 1D 600 MHz 1H NMR spectroscopy, combined with 2D correlation via homonuclear scalar coupling (COSY), pulsed-field gradient heteronuclear multiple quantum coherence (gHMQC) and heteronuclear multiple-bond correlation (gHMBC) NMR spectroscopy was used to identify and quantify the structural defects of the resulting polymers. Some additives were found to decrease the labile structure concentrations and thus increase the thermal stability of the resulting polymers. A linear correlation relationship was found between the rate of dehydrochlorination and the concentration of labile structural defects in the resulting polymers.
xxxvi
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Chapter 1 Introduction
-1-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §1.1 Introduction
Poly(vinyl chloride) (PVC), is one of the most widely produced polymeric materials in the world; the annual production of the PVC is second only to that of polyethylene1. In industry, PVC is generally synthesized by free radical polymerization of vinyl chloride (VC). Radiation-induced2, anionic3, and modified Ziegler-Natta catalyst4 methods have also been reported in the literature. Vinyl chloride differs from monomers like styrene, methyl methacrylate or vinyl acetate principally by the insolubility of the polymer in its monomer. Its key feature is that poly(vinyl chloride) is insoluble in its monomer, but slightly swollen by it. Polymerization of vinyl chloride also differs from the heterogeneous polymerization of monomers, such as acrylonitrile. Vinyl chloride partly swells its polymer but acrylonitrile does not. It also differs from the conventional emulsion polymerization of unsaturated monomers in which the polymer particles are swollen by their monomer.
Vinyl chloride is classified as a non-conjugated weak electron-withdrawing vinyl monomer with Q- and e-values of 0.056 and 0.16, respectively, determined from radical copolymerization5. This indicates that the VC monomer has low reactivity, but its radical is highly reactive. The propagating radical is so reactive that it tends to chain transfer to all substances in the polymerization system, such as monomer, initiators, solvents, and the resulting polymer. As a consequence, radical polymerization of VC produces anomalous units in the chain, which decrease its thermal stability. Generally, the termination reaction is dominated by chain transfer to monomer, if there are no other
-2-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ chain-transfer agents in the system. The overall polymerization of vinyl chloride can be described as follows:
Initiation:
Chain-transfer:
d ⎯⎯→
k
I
R• + Μ
2R•
i ⎯⎯→
k
M1•
Propagation:
M2•
+ M
p ⎯⎯→
k
M2•
kp
M3•
⎯⎯→
Mi
+ M
kp
⎯⎯→
Mi+1
Pi + M•
k
trp ⎯⎯→ ⎯ tri ⎯⎯→
M i• + S
k
Pi + P• P
Pi + I•
trs ⎯⎯→ ⎯
k
Pi + S•
Termination:
••••••• •
k
M i• + P M i• + I
M1• + M
trm ⎯⎯ ⎯ →
M i• + M
•
M i• + M j•
tc ⎯⎯→
M i• + M j•
td ⎯⎯→
M i• + M •
k
k
tm ⎯⎯→
k
Pi+j Pi + Pj Pi
Here I is the initiator, R• the primary radical, M the monomer, M• the monomeric radical, Mi• the growing radical with i monomer units, P dead polymer, P• the polymer radical generated by a chain transfer reaction, I• the initiator radical, S the solvent, S• the solvent radical, k’s corresponding rate constants.
§1.2 A brief History of PVC
The monomer, vinyl chloride, was first discovered in 1835 by Henri Victor Regnault6-9, a young Frenchman, born in Aix-la-Chapelle in 1810. Regnault came to -3-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ work for a short time in the winter of 1834-35 in Justus von Liebig’s laboratory at Giessen, Germany. He described the preparation of vinyl chloride as follows:
When one mixes an alcoholic solution of caustic potash with oil of the Dutch chemists (i.e. dichloroethane), one can observe after some time the appearance of a precipitate which continues to increase. When one takes the vessel containing the reaction mixture into one’s warm hands, the liquid begins to boil and a large amount of gas with an ether-like smell is evolved. This burns with a yellow flame with a green mantle10.
He examined the white precipitate and identified it as potassium chloride. The gas however proved less simple. He found, it can be condensed between –15o and –18oC, and it is soluble in alcohol and ether in all proportions and to a much lesser degree in water. The formula of the new compound was described as C2H3Cl. He named it ‘chloraldehydene’. In the same year11, he also prepared the bromine and iodine analogues of ‘chloraldehydene’: C2H3Br and C2H3I. It was Kolbe12 who first named Regnault’s compound as ‘vinyl chloride’ in 1854, although the origin of the term ‘vinyl’ was not discussed. It was not until 1870 that the structure of vinyl chloride was finally established by F.V. Kekule13.
It was not clear whether Regnault observed poly(vinyl chloride) in his early investigations. The first report of vinyl halide polymer was made in 1860 by A. W. Hofmann14. He noted the change of vinyl bromide monomer to a white mass without compositional change. Actually, Hofmann saw the polymerization of vinyl bromide but had little idea about the nature of the change. Since the polymer concept had not been -4-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ developed yet, he described the change as ‘metamorphosis’. He was followed by Eugen Baumann15 who described the preparation of poly(vinyl chloride) in 1872; almost 50 years before the macromolecular concept was developed. He detailed the sunlight induced change of vinyl chloride monomer to solid products, which he thought to be the isomers of the monomer. The properties described by him are those we ascribe, today, to poly(vinyl chloride), or PVC.
The further development of poly(vinyl chloride) is a tale of two continents as well as different reasons and objectives. Actually, it took 50 years for the issue of a German patent for the manufacture of vinyl chloride by reaction of acetylene and hydrogen chloride. In 1912, it was Frits Klatte16 who was assigned by his superiors at Chemische Fabrik Griesheim-Elektron to find uses for excess acetylene. It was no longer used for lighting because new efficient electric generators were developed which ended the acetylene lamp business. He reacted some acetylene with hydrochloric acid (HCl). Now this reaction will produce vinyl chloride, but at that time no one knew what to do with it, so he put it on the shelf, where it polymerized over time. He and his company GriesheimElektron patented the material in Germany in 1913. It was the first PVC patent in the world. They never figured out a use for their poly(vinyl chloride) product, and in 1925 their patent expired.
In 1926, an American chemist, Waldo Semon17, working at B.F. Goodrich, invented plasticized PVC, or vinyl, ‘by accident’, as he later claimed. Actually, his original assignment was to make an adhesive from a simple synthetic organic polymer. -5-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ His first attempts, using reclaimed crude rubber and a German prototype of synthetic rubber, were unsuccessful. When he used up all his supply of rubber, he began experimenting with synthetic organic polymers—including poly(vinyl chloride), a substance at that time considered no more valuable than refuse. Because the polymer was stiff at room temperature, Semon heated it in a solvent with a high boiling point, e.s., tritolyl phosphate, and got a jelly that was elastic and flexible after cooling, but was not adhesive. It demonstrated the plasticizing of PVC. Semon instinctively realized that he was halfway to a major breakthrough. He kept experimenting with poly(vinyl chloride) until he succeeded in plasticizing the polymer. PVC was always more durable than crude rubber. Semon's first breakthrough made it elastic as well as resilient; and his second breakthrough made it moldable into whatever shape was required. Semon's first applications included a golf ball and shoe heels, as well as a number of useful coatings for tool handles, wire, and other items. By the 1930s, BFGoodrich had begun to produce and market the first of the hundreds of commercial applications that would be found for plasticized PVC.
In the ’50s, five companies in the Unites States were producing PVC, and the number of producers increased to 20 by the middle ’60s. The number remained constant in the ’70s and ’80s. Currently, the largest PVC maker in the U.S. is Shintech, followed by Oxy Vinyls--a joint venture between Occidental Chemical and PolyOne--and Formosa Plastics and Georgia Gulf, which are tied for third18.
-6-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ PVC is now the second most used plastic in the world, next to polyethylene, the source of billions of dollars of revenue every year. The PVC industry employs more than 100,000 people in the United States alone. Figure 1.1 shows a graphic evolution of world PVC production from 1950 to 2003. PVC is a very versatile plastic. This is exemplified by the wide variety of end use applications that include toys, food packaging, furniture, transportation, electronics, medical blood bags and prosthetic devices, wire and cable insulation, water and sewer pipes, window frames, etc. Figure 1.2 illustrates the World PVC applications in 2003.
Evolution of World PVC Production 1950
0.22
1960
1.50
1970
6.00
1980
11.40
1985
13.90 18.00
1989 1996
22.45
1999
25.54*
2000
25.80* 26.23*
2001 2002
27.04*
2003 0.00
28.36* 5.00
10.00 15.00 20.00 25.00 PVC Production (Million Metric Ton)
Figure 1.1. Evolution of world PVC production (*source: CMAI).
-7-
30.00
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2003 World PVC Applications Coatings Others 7% Bottles 2% 3% Floorings 3% Profiles & Tubes 5% Pipe & Fittings 39%
Film & Sheet(rigid) 7% Wire & Cable 7% Film & Sheet 11%
Profiles(rigid) 16%
Figure 1.2. 2003 World PVC applications (source: CMAI)
§1.3 Methods of Vinyl Chloride Polymerization
PVC is commercially manufactured by four major processes: suspension, emulsion, bulk, and solution polymerization. Suspension polymerization is the most widely used procedure1, followed by emulsion and bulk polymerization. Solution polymerization is reserved for a few specialty copolymers, or where the application makes it appropriate, as in solution coatings.
§1.3.1 Polymerization in Bulk
The bulk polymerization of VC is the third most important manufacturing process for PVC. The advantage of bulk polymerization, in contrast to the common suspension or -8-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ emulsion polymerization is that the products are free of protective colloids, suspending agents, surfactants, buffers, water, additives, or solvents. There is, however, one great problem for technical application. This is to remove the heat generated during polymerization and to control the rate of reaction. The industrial-scale bulk polymerization is based on the Pechiney-Saint-Gobain process19, a two-stage process20, 21. In the first step VC is prepolymerized to approximately 10% conversion. Then the reacting mass is dropped into a second autoclave, and more monomer and initiator are added. The polymer beads grow larger and the mixture takes on the appearance of a dry powder. The reactor is specially designed to stir powdery material and is equipped with a condenser. To avoid agglomeration of the beads, it is very important to control the rate of agitation. PVC produced in this way is the purest product available on the market. Experimental procedures for bulk polymerization on a laboratory scale are relatively simple. Normally, the monomer is heated in the presence of a small amount of a monomer-soluble initiator, such as azobis(isobutyronitrile) or dicapryloyl peroxide, under a suitable condensing or pressure system until the desired conversion of monomer into polymer has been achieved. The VC can be recovered by distillation in an effective hood.
Some monomer-soluble initiators mentioned in many patents are di(2ethylhexanoyl) peroxide, 3,5,5-trimethylhexanoyl peroxide, di(t-butyl) peroxyoxalate, di(carballyloxyisopropyl peroxydicarbonate) di-2-butoxyethyl peroxydicarbonate, di-4chlorobutyl
peroxydicarbonate,
azobis(isobutyronitrile)
carbonitrile).
-9-
and
azobis-(cyclohexyl-
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §1.3.2 Polymerization in Suspension
As mentioned earlier, suspension polymerization is the most widely used method today, producing over 80% of the world’s PVC. The usual manufacturing procedure is a batch operation, although continuous processes have been described. A typical formula for such batch polymerization might be as follows:
Ingredient
Parts by weight
Vinyl chloride
100
Deionized water
200
Lauroyl peroxide
0.04
Poly(vinyl alcohol)*
0.01
*Degree of hydrolysis: 80%.
Here lauroyl peroxide is the initiator, which is monomer-soluble. Poly(vinyl alcohol) acts as a suspending agent, which is necessary for stabilizing the monomer droplets to avoid coagulation and to control the dimension of the particles. Sometimes small amounts of emulsifiers such as sulfonated oils are also used to increase the porosity of the polymer particles. The water/monomer ratio can be raised up to 4:1 and the polymerization temperatures range from 40-60 oC. When a pressure drop is observed, the polymerization is almost finished. The slurry from the autoclave is then discharged into an evacuated tank and unreacted monomer is pumped out, condensed, and after processing fed back into the polymerization. The water is removed by centrifugation
- 10 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ accompanied by washing with distilled water to remove all soluble electrolytes. Finally the PVC product is dried under reduced pressure at about 50 oC.
§1.3.3 Polymerization in Emulsion
Emulsion polymerization is mainly used for manufacturing vinyl chloride copolymers. Sometimes the polymerization product, in latex form, is used as it comes from the reactor. Sometimes it is converted to a dry powder. The homopolymer is never used in latex applications since the temperature required for fusion of the homopolymer particles is rather high.
In contrast to bulk and suspension polymerizations, where monomer-soluble initiators are used, in emulsion polymerization a water-soluble initiator is used. Some initiators used are: potassium persulfate, ammonium persulfate, sodium percarbonate, peracetic acid, hydrogen peroxide and cumene-hydroperoxide (water-soluble), etc. The ratio between water and VC is approximately 2:1. By analogy with suspension polymerization, the rate of agitation is very important for the preparation of useful emulsion resins. Generally, the speed of agitation is more moderate than in suspension polymerization. Sometimes it is desirable to keep the pH value constant during the reaction. Therefore, a buffer is added to the reaction system. The advantage of emulsion polymerization for latex production is obvious. However, there are other advantages of the process compared to suspension polymerization: a) the initiation and propagation steps can be controlled more independently than in suspension polymerization which has essentially the features of a bulk process; b) polymer particles have relatively little - 11 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ internal void volume as compared with the suspension polymerization product, and c) monomer can be added during the polymerization to maintain polymer composition constant. This is very important in copolymerization.
§1.3.4 Polymerization in Solution
Since PVC is not soluble in its own monomer, it is necessary to find a solvent for the polymer for solution polymerization. Such a solvent could be tetrahydrofuran, acetone, cyclohexanone, alkyl acetates, chlorinated alkyls, diethyl oxalate, etc. The azo and organic peroxo compounds used in bulk and suspension polymerization are suitable for initiation. Actually solution polymerization is very complex due to chain transfer to the solvent and the solubility of the polymer. Because of the cost of solvents and their recovery, this process is rarely used in industry.
§1.4 Structure and Properties of PVC
Poly(vinyl chloride) has a mainly ‘head-to-tail’ main chain structure, in which a chlorine atom is situated on alternate carbons of the polymer chain:
~~~CH2-CHCl-CH2-CHCl-CH2-CHCl-CH2-CHCl-CH2-CHCl-CH2-CHCl~~~ The structure of the polymer leads to a very rigid and relatively tough plastic in the unplasticized state. In the presence of a plasticizer, however, the dipole bonding between polymer chains is much reduced, leading to increased freedom of chain movement and thus to a flexible material. By suitable choice of plasticizer level and type a whole range of products from flexible elastomers to rigid compounds can be produced. - 12 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Materials can be made that are virtually unbreakable, that are weatherable with good property retention for over 30 years, that have stiff melts and little die swell for outstanding dimensional control in profile extrusion, or low viscosity melts for thin walled injection molding.
Many materials can be used as plasticizers for PVC; the important ones include diisooctyl phthalate, tritotyl phosphate, dibutyl sebacate and epoxidized soya bean oil. Generally, 40-60 parts of plasticizer per 100 parts of PVC polymer are used for most common applications. PVC containing this level of plasticizer is a flexible rubber-like material. From Figure 1.2 we find 45% of PVC is used in a plasticized form and 55% in the rigid form. The presence of chlorine in the molecule makes PVC particularly versatile since it makes it compatible with a wide range of other materials. The chlorine content also helps to make PVC flame retardant. PVC polymer is light, neutral, durable, robust, non-toxic and non-flammable.
Some chain branching is present in PVC; perhaps 6-10 branches per molecule. The branching makes the PVC imperfect. This is one example of a “structural defect”, which will be discussed in detail in the following section.
§1.5 Structural Defects in PVC
In spite of its enormous technical and economic importance, PVC still poses many problems: its rather low stability towards heat and light results in discoloration, HCl loss and serious corrosion phenomena, which require stabilization of the polymer for - 13 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ practically all technical applications. Many people agree that normal PVC with a head-totail structure should be quite stable to heat. It is generally assumed, therefore, that structural abnormalities in the polymer chains are responsible for the relative instability of PVC. Possible defect structures in PVC are branching, chloroallyl end groups and head-to-head structures. In addition to these abnormalities the steric order of the monomer units, i.e. the tacticity, may have some influence on the degradation process.
§1.5.1 Short-chain Branching
There are mainly four short-chain PVC branches; these have been studied in detail in the literature22-24. They are chloromethyl branches (MB), 2-chloroethyl branches (EBI), 1,2-dichloroethyl branches (EB-II), and 2,4-dichlorobutyl branches (BB). These have the following formulas.
-CHCl-CH2-CH-CHCl-CH2-
-CH2-CCl-CH2-CHClCH2-CHCl-CH 2CH 2Cl
CH2Cl (MB) -CH2-CCl-CH2-CHCl-
(BB) -CH2-CH-CH2-CHCl-
CH2-CH2Cl
CHCl-CH2Cl
(EB-I)
(EB-II)
The chloromethyl branches (MB) and 1,2-dichloroethyl branches are mostly generated by an intramolecular process first proposed by Rigo et al22 and proved by Starnes et al23, 24. This mechanism is outlined in Scheme 1.1:
- 14 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
P
VC kp
P
VC k p' ~CH2-CH-CH-CH2 VC kp''
~CH2-CH-CH-CH2-CH2-CH Cl
Cl Cl
k3
VC
kr k2
~CH2-CH-CHCl-CH2Cl
1,2-Cl shift
~CH2-CHCl-CH-CH2Cl
k-2
VC
VC k4
VC k5
VC k6
A2 +
+
~CHCl-CH-CH2-CHCl
CHCl-CH2Cl
CH2Cl ClCH2-CHCl
EB kp VC
k1
~CH2-CHCl-CH=CH 2 A1
~CH2-CH=CH-CH 2Cl ~CH2-CH-CH2-CHCl
Cl Cl
kp
MB
ClCH2-CHCl VC
VC
kp
VC
kp
P
Scheme 1.1. Mechanistic sequel of head-to-head addition during the free radical polymerization of vinyl chloride (VC), where P• is the propagating head-to-tail macroradical, and k’s are rate constants24.
An occasional head-to-head monomer addition can be followed by either an ordinary addition or a radical rearrangement. The latter implies 1,2-Cl migration forming a secondary radical, which presumably is more stable than the primary radical. Propagation will then result in a pendent chloromethyl group, shown as follows:
- 15 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ -CHCl-CH2-CH-CH-CH2
-CHCl-CH2-CH + CH=CH2
Cl Cl
Cl
Cl
-CH2-CH-CH-CH2CH2-CHCl
VC
Cl Cl
-CHCl-CH2-CH-CH-CH2 1,2-Cl shift
Cl Cl
-CHCl-CH2-CH-CH Cl CH2Cl
CH2=CHCl
-CHCl-CH2-CH-CH
-CH2-CH-CH-CH2CHCl Cl CH2Cl
Cl CH2Cl
The forming of 1,2-dichloroethyl branches was proposed by Starnes et al23, 24. This involves four steps, shown below:
-CHCl-CH2-CH-CH-CH2
-CHCl-CH2-CH + CH=CH2 Cl
Cl Cl
Cl
-CHCl-CH2-CH-CH-CH2 Cl Cl
1,2-Chlorine migration
2,3-Chlorine migration
-CHCl-CH2-CH-CH Cl CH2Cl -CHCl-CH2-CH
-CHCl-CH2-CH-CH Cl CH2Cl -CHCl-CH2-CH CHCl-CH2Cl
CH2=CHCl
CHCl-CH2Cl
-CHCl-CH2-CH-CH2-CHCl CHCl-CH2Cl
The first two steps are head-to-head addition of monomer to the propagating macroradical and subsequent rearrangement of the ensuing radical, via a 1,2-chlorine shift. The next step is more conjectural. It involves a 2,3-Cl shift, followed by normal
- 16 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ chain propagation, which generates the 1,2-dichloroethyl branch structure. Its formation can be regarded as having proceeded via a modified “billiard-ball” route in which the first chlorine that migrates is quickly replaced by another one. It has been pointed out that a tertiary carbon at the 1,2-dichloroethyl branch point is bound to a hydrogen, while the tertiary carbon bonds with a chlorine in the 2-chloroethyl branch.
The 2-chloroethyl branches are suggested to be formed by “backbiting” of normal propagating macroradical end groups, through a 1,3-hydrogen shift, indicated as following:
-CHCl-CH2-CHCl-CH2-CHCl
-CHCl-CH2-CCl
CH2=CHCl
1,3-Hydrogen migration Cl -CHCl-CH2C-CH2CHCl CH2-CH2Cl
CH2-CH2Cl
Finally, 2,4-dichlorobutyl branches are formed by “backbiting” of the normal propagating
macroradical end groups through a 1,5-hydrogen shift, involving a 6-
member ring transition state. It is shown as follows:
- 17 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ CH Cl
CH2
CH2 CCl
CHCl
H
CH2
1,5-H shift
CH
CH2
CH2
Cl
C
CHCl
Cl
CH2
CH
CH2
Cl
Cl CH2=CHCl Cl
~CHCl-CH 2-C-CH2-CHCl CH2-CH-CH 2-CH2 Cl
Cl
The tertiary carbon bonds with a chlorine in the 2, 4-dichlorobutyl branch.
§1.5.2 Long-chain Branching
The long-chain branches have the following structures:
Cl
H
~CH2-C-CH2~
~CHCl-C-CHCl~ CH2-CHCl-CH 2-CHCl~
CH2-CHCl-CH 2-CHCl~
LCB-II
LCB-I
The tertiary carbon at the LCB-II type point is bound to a hydrogen atom, while in the LCB-I type branching the tertiary carbon bonds with a chlorine atom. This labile chlorine is removed easily at elevated temperatures by dehydrochlorination. The propagating macroradical chain transferring to polymer initiates chain growth from the polymer to form long chain branches. For instance, the macroradical could attack the
- 18 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ methyne group in the polymer to form a 1,3-dichloroalkane chain end and a tertiary carbon radical bound with chlorine. This tertiary carbon radical initiates a type-I longchain branch. If the macroradical attacks a methylene group of the polymer, then a type II long-chain branch can be formed, shown as follows:
~CH-CH2-CH2 + ~CH2-C-CH2~
~CH-CH2-CH + ~CH2-CH-CH2~ Cl
Cl
Cl
Cl
Cl
Cl
Cl VC
~CH2-C-CH2~
VC
...
~CH2-C-CH2~
Cl
CH2-CHCl-CH 2-CHCl-CH2~ LCB-I ~CH-CH2-CH2 + ~CH-CH-CH~
~CH-CH2-CH + ~CH-CH2-CH~ Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
H VC
~CH-CH-CH~ Cl
...
VC
~CH2-C-CH2~
Cl
CH2-CHCl-CH 2-CHCl-CH2~ LCB-II
If the methylene radical loses one chlorine to monomer, an internal double bond can be formed:
Cl
Cl
Cl Cl
~CH2CH CH
~CH2
CH
CH
CH
CH2
CH
CH2
+
CH Cl
Cl
~CH2CH
~CH2
Cl
- 19 -
~CH2CH CH CH ~CH2
Cl CH +
CH2 Cl
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The chlorine on a carbon next to the double bond is an allylic chlorine, which functions as a labile chlorine in PVC dehydrochlorination at elevated temperaturess.
§1.5.3 Unsaturation
Unsaturated groups in PVC include terminal and internal double bonds. Internal double bond formation was described in the previous section. Terminal double bonds can be formed through the following mechanism:
~CH2-CH
~CH2-CH-CH-CH 2
CH=CH2
+
Cl Cl 1,2-Cl s hift
Cl
Cl
~CH2-CHCl -CH- CH2Cl
~CH2-CH-CHCl- CH2Cl VC Cl
~CH2
ClCH2
VC
VC Cl
~CH2
Cl
~ClCH
CH
CH
CH
CH2
CH
CH
CH
CH2
CH
CH
CH2
CH2
Cl
ClCH2
~CH2-CH=CH-CH 2Cl
Cl
Cl
~CH2-CHCl-CH=CH 2
+
+
ClCH2-CHCl
ClCH2-CHCl
We can see that there are two types of terminal double bonds. One is vinyl end group and the other can be called pseudo-terminal double bond. - 20 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §1.5.4 Head-to-head Structures
The PVC chain consists mainly of a head-to-tail arrangement of monomer units. However, a head-to-head PVC structure could be formed through the following mechanism: ~CH2-CH-CH2-CH Cl
Cl kp
VC
VC
~CH2-CH-CH2-CH Cl
~CH2-CH-CH-CH 2
Cl
Cl Cl
kp VC
VC kp''
~CH2-CH-CH2-CH-CH2-CH Cl
Cl head-to-tail
kp'
~CH2-CH-CH-CH2-CH2-CH Cl Cl head-to-head
Cl
Cl
The head-to-head structure could lower the thermal stability of PVC.
§1.6 Stereoregularity of PVC
When vinyl chloride monomer adds to a growing PVC chain, the terminal group takes one of two possible orientations relative to the penultimate group. If the backbone is in an all trans conformation, isotactic structures have adjacent chlorine atoms oriented at the same side of the carbon-carbon plane. Syndiotactic structures have the chlorine atoms alternating in their placement relative to the carbon-carbon-carbon plane. Atactic PVC is a mixture of the two stereoisomers. A schematic representation of isotactic, syndiotactic and atactic PVC structure is shown as in Figure 1.5. It was claimed that - 21 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ syndiotactic PVC can crystallize easily, and it is very important in understanding the properties of PVC.
H Cl
H
H Cl
H
H Cl
H
H Cl
H Cl
H
H
H
H
H Cl
H
H
H
H
(a) isotactic PVC H Cl Cl H
Cl H
H
H
H Cl
H
H
H
H
(b) syndiotactic PVC H Cl Cl H
H
H
H
H
H Cl
H Cl
H
H
H
Cl H
H Cl
H
H
H
Cl H
Cl H
H
H
H
(c) atactic PVC Figure 1.3. Schematic representations of (a) isotactic, (b) syndiotactic, and (c) atactic chains of PVC.
- 22 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §1.7 Dissertation Goals and Summary
PVC is one of the oldest plastics; 28 million tons per year are manufactured worldwide today. PVC owes its popularity to its versatility and low cost. The polymer itself is chemically inert and nonflammable, burning only in the presence of a source of ignition. It is compatible with many additives, including plasticizers, heat stabilizers, lubricants, fillers, and a wide range of other polymers. All of these features make PVC one of the world’s major bulk polymeric materials that have a huge impact on our everyday lives.
However, PVC still has a number of drawbacks. One serious problem for PVC is its rather low thermal stability. The dehydrochlorination of PVC starts at about 100 oC, and is the reason for its discoloration during extrusion, due to the formation of polyene sequences. As PVC is one of the relatively ‘old’ plastics, research in this field is not as attractive as that on new, pioneering polymers. Nonetheless, owing to its enormous commercial importance, the research and study of this polymer continues to attract the attention of thousands of polymer scientists and engineers. Many important basic problems that relate to PVC and vinyl chloride polymerization still remain to be fully understood or solved. Research on new methods of VC polymerization, the mechanism and modeling of VC free-radical polymerization, the relationship of the microstructure and stability of PVC, and the PVC degradation and stabilization will continue to be of interest.
- 23 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ A number of years ago, our group demonstrated that solvent interaction could be used to increase the stereoregularity in radical polymerization of poly(methacrylic acid) from 58% to 95% syndiotactic triads25,
26
. Hydrogen bonding between the pen-
penultimate unit and the carboxyl group in the radical was found to be the important interaction. Lack of hydrogen bonding interaction resulted in relatively low syndiotactic content as did too strong a hydrogen bonding interaction with solvent or monomer. A solvent such as 2-propanol with its moderate hydrogen bonding capability gave the highest syndiotacticity. It was also found that too bulky a hydrogen bonding solvent gave lower syndiotacticity.
This concept may be extended to the radical polymerization of vinyl chloride. Since poly(vinyl chloride) forms compatible mixtures with poly(methyl methacrylate) and poly(carbonate)27,
28
, the hydrogen on the chloromethylene(~CHCl~) group of
poly(vinyl chloride) is acidic enough to hydrogen bond with weak proton accepting groups. A terminal radical hydrogen should be more acidic than the internal hydrogens. These weak proton accepting compounds can be weakly basic compounds, high dipole carbonyl compounds, molecules having two connected noncarbon atoms, and some heteroaromatics.
Our hypothesis was that small quantities of those compounds, so-called ‘additives’, in the vinyl chloride polymerization system could preferentially form hydrogen bonds with the propagating chloromethylene hydrogen. This interaction might create steric hindrance that could direct the propagating radical to add vinyl chloride
- 24 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ monomer in a way that forms syndiotactic rich polymer PVC. Additionally, such interactions could reduce ‘backbiting’ and chain transfer to monomer or polymer. A more linear PVC structure could be formed that has higher molecular weight, higher thermal stability and/or higher crystallinity, depending on the additive.
The goal of this dissertation was to investigate the polymerization of vinyl chloride in the presence of small amounts of organic additives. The organic additives selected were weakly basic compounds, high dipole carbonyl compounds, high dipole ether compounds, and some heteroaromatics. Various methods were applied to evaluate the influence of the additives on the polymerization and the resulting PVC polymer structures. The following is a summary of this dissertation:
Chapter 1 was a general introduction to vinyl chloride polymerization and the history of poly(vinyl chloride). The microstructure of the PVC and its bulk polymerization, suspension polymerization, emulsion polymerization, and solution polymerization were briefly discussed.
Chapter 2 described in detail the free radical polymerization of vinyl chloride in the presence of various organic additives. Gel Permeation Chromatography (GPC) was used to determine the molecular weights and the molecular weight distributions of the resulting poly(vinyl chloride). A kinetic model was developed to describe the polymerization rate under the influence of additives and the molecular weights of the resulting polymers.
- 25 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ In Chapter 3, Differential Scanning Calorimetry (DSC) and Thermo-gravimetric Analysis (TGA) were used to characterize the crystallization and stability of the resulting polymers. The possible additive effect on the crystallinity of the resulting PVCs was evaluated.
In Chapter 4, the dehydrochlorination of PVCs prepared in the presence of various
additives
was
measured
and
the
possible
additive
effect
on
the
dehydrochlorination rate of the resulting polymers was discussed. A combination of 1D and 2D Nuclear Magnetic Resonance (NMR) spectroscopy was used to identify and quantify the structural defects of the resulting PVCs. The additive effect on the formation of some structural defects was investigated. A correlation was found between the dehydrochlorination rate and the labile structure concentration in the resulting polymers.
And finally, in Chapter 5, general conclusions were drawn based on the results of the GPC, DSC, TGA, NMR and dehydrochlorination of the resulting polymers.
- 26 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §1.8 References 1. Weiberg, E. L. Encyclopedia of PVC, 2nd Edition, Vol. 1, Ed. Nass, L. I., Heiberger, C. A., Marcel Dekker Inc., New York and Basel, 1986, p1-35. 2. Carenza, M.; Palma, G.; Talamini, G.; Tavan, M. J. Macromol. Sci., Chemistry 1977, A11(7), 1235-48. 3. (a): Endo, K.; Kaneda, N.; Waku, H. Polymer 2003, 44, 2655-60; (b): Wesslen, B.; Wirsen, A. J. Polym. Sci., Polym. Chem. Edition 1975, 13(11), 2571-80. 4. Florjanczyk, Z.; Kuran, W.; Sitkowska, J.; Ziolkowski, A. J. Polym. Sci., Part A: Polymer Chemistry 1987, 25(1), 343-51.
5. Odian, G. Principles of Polymerization 4th edition, John Wiley & Sons, Inc., New York, 2004, p493, p502. 6. Semon, W. L.; Stahl, G.A. History of Polymer Science and Technology, New York: Marcel Dekker; 1982. 7. Krause, A. Chem. Eng. 1965, 72, 72. 8. Regnaut, H. V. Liebigs Ann. 1835, 14, 22. 9. Regnaut, H. V. Ann. Chim. (Phys.) 1835, 59, 358. 10. Kaufman, M. The Chemistry and Industrial Production of Poly(vinyl chloride): The History of PVC, Gordon and Breach, New York, 1969.
11. Regnaut, H. V. Leibig Ann. 1835, 15, 60. 12. Kolbe, H. Lehrbuch der organischen Chemie, Braunschweig, 1854, I , 346. 13. Kekule, F. A. and Zincke, I., Ber. dt. chem. Ges. 1870, 3, 130. 14. Hofmann, A. W. Liebig Ann. 1860, 115, 271. 15. Baumann, E. Leibig Ann. 1872, 163, 308. - 27 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 16. Klatte, F. Hoechster Archiven, Frankfurt am Main, 1965, 10, 47-59. 17. Semon, W. L.; Stahl, G. A. J. Macromol. Sci., Chem. 1981, A15(6), 1263-78. 18. Tullo, A. Chemical & Engineering News, 2002, 80(20), p.20. 19. (a): Thomas, J. C. (Produits Chimiques Pechiney-Saint-Gobain), 1966, 5 pp., FR 1,427,935; (b): Thomas, J. C.; Fournel, F.; Soussan, S. (Produits Chimiques PechineySaint-Gobain), 1968, 7 pp., FR 1,522,403. 20. Toernell, B.; Uustalu, J. Journal of Applied Polymer Science 1988, 35(1), 63-74. 21. Chatelain, J. Br. Polym. J. 1973, 5(6), 457-65. 22. Rigo, A.; Palma, G.; Talamini, G.; Makromol. Chem. 1973, 53, 219. 23. Starnes, W. H. Jr.; Schilling, F. C.; Plitz, I. M.; Cais, R. E.; Freed, D. J.; Hartless, R. L.; Bovey, F. A. Macromolecules 1983, 16, 790-807. 24. Starnes, W. H., Jr.; Wojciechowski, B. J. Makromol. Chem. Macromol. Symp. 1993, 70/71, 1-11. 25. Lando, J. B.; Semen, J.; Farmer B. Macromolecules 1970, 3(5), 524-7. 26. Lando, J. B.; Litt, M.; Kumar, G.; Shimko, T. J. Polym. Sci. C 1974, 44, 203. 27. Jager, H.; Vorenkamp, E. J.; Challa, G.; Polymer Commun. 1983, 24, 290. 28. Belhaneche-Bensemra, N.; Belaabed, B.; Bedda, A. Macromol. Symp. 2002, 180, 203-15.
- 28 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Chapter 2 Free Radical Polymerization of Vinyl Chloride in the presence of Organic Additives
- 29 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.1 Introduction PVC with good stability has long been the objective of intense studies since the low stability of commercial PVC makes it difficult to process. The defect structures in the PVC backbone are considered to be the main factors in lowering its thermal stability. PVC degrades by releasing hydrogen chloride, with yellowing and subsequent blackening of the product. The present methods for enhancing the thermal stability of PVC involve modifying the polymer by copolymerization with another monomer, by adding some organic or inorganic compounds or blending with other polymers after polymerization. Those methods have nothing to do with the defect structures of PVC itself. The big challenge in the PVC community is to prepare PVC with fewer defects, thus having an enhanced thermal stability.
In the early 1960s, Burleigh first prepared so-called ‘crystalline’ poly(vinyl chloride) with a free-radical initiator using aliphatic aldehydes, especially nbutyraldehyde as a polymerization medium1-2. The crystallinity of the PVC obtained reached 42%, significantly exceeding the value for the commercial PVC (4-10%). Highly crystalline PVC was also obtained by the polymerization of vinyl chloride in the presence of acetaldehyde3, dialkyl phosphites4, triethylamine5, carbon tetrachloride6, and other compounds7. However the yields of the crystalline PVC were usually low and the degree of polymerization was only from 20 to 80. It was proposed that the high crystallinity was due to the ability of the additives to form donor-accepter or π-bond complexes with the propagating radicals, which promoted stereospecific addition of the vinyl chloride to the growing chains2, 4, 8. This mechanism was questioned and other authors pointed out that - 30 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ the highly crystalline PVC had a low degree of polymerization and was obtained when vinyl chloride was polymerized with solutes or additives9; low-molecular-weight PVC characteristically has an increased crystallinity and tacticity.
In 1981, Turska et al10 reported on vinyl chloride polymerization in the presence of small quantities of 2,5-dimercapto-1,3,4-thiodiazol (bismuthiol), epoxidized soybean oil, and Ergowax GS-1 (a mixture of mono and diesters of glycerol and stearic and palmitic acids). The thermal stability of the PVC was 20-30% higher than normal PVC with no evident molecular weight change. The syndiotacticity of the resulting PVC was also higher than similar commercial PVC, but dropped to its usual value, i.e. 54.5% of syndiotactic dyads, as conversion increased. No further studies of mechanism were performed since then. Recently, studies on the polymerization of vinyl chloride by the pseudo-living or ‘living’ radical polymerization method have been published11-14. Grishin et al.11, 12 studied the radical polymerization of vinyl chloride in the presence of catalytic amounts of stable nitroxyl radicals or their sources, such as C-phenyl-N-tert-butylnitrone (PBN), 2-methyl-2-nitrosopropane(MNP), or 1-tert-butyl-3-phenyl-1-oxytriazene(BPT). Studies of the polymerization kinetics of vinyl chloride and molecular weight characteristics of the resulting PVC show that the reaction occurs by a pseudoliving chain mechanism, thus opening the opportunity to prepare PVC with controlled chain structure and molecular weight. Percec and coworkers13, 14 prepared poly(vinyl chloride) by single electron transfer-degenerative chain transfer ‘living’ radical polymerization of vinyl chloride initiated with iodoform and catalyzed by nascent Cu0/tri(2-aminoethyl)amine or Na2S2O4 in water at 21 oC. The resulting PVC had a controlled molecular weight up to
- 31 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 120,000 (relative to polystyrene standards), a molecular weight distribution around 1.72.3, free of structural defects, and higher syndiotacticity (62%) than commercial PVC.
Poly(vinyl chloride) can form compatible blends with poly(methyl methacrylate) (PMMA) and polycarbonate (PC). It was proposed that the miscibility was due to a special hydrogen bonding interaction between the carbonyl groups (>C=O) of PMMA or PC and hydrogen from chloromethylene (~CHCl~) groups of PVC15, 16. The hydrogen on the chloromethylene groups of PVC may be acidic enough to hydrogen bond with compounds with weak proton accepting groups. Such weak proton accepting compounds can be weak base compounds, high dipole carbonyl compounds, molecules having two connected noncarbon atoms and some heteroaromatics. In this dissertation, such compounds were chosen as additives in the vinyl chloride polymerization system to investigate their influence on the resulting PVC. The interaction may reduce ‘backbiting’ and chain transfer to monomer or polymer, to form PVC with higher molecular weight, higher thermal stability and higher crystallinity.
§2.2 Experimental §2.2.1 Materials Vinyl chloride (VC) monomer, polymerization grade, provided by BFGoodrich, was used without further purification. 2,2’-Azobisisobutyronitrile (AIBN), purchased from Aldrich, was recrystallized three times from methanol and dried in vacuum at 45 oC. Tetrahydrofuran (HPLC grade, Fischer) was refluxed with 2,6-di-tert-butyl-4methylphenol (BHT) and distilled under nitrogen. Trimethylphosphine oxide (TMPO, - 32 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ mp: 140-141 oC) and triethylphosphine oxide (TEPO, mp: 48-50 oC) were purchased from Alfa-Aesar and used as received. Triphenylphosphine oxide (98%, Aldrich) (TPPO) and tributylphosphine oxide (95%, Aldrich) (TBPO) were recrystallized from ethanol. 2Methyl benzothiazole (99%, Aldrich) (MBTZ) was distilled under vacuum (65 o
C/1mmHg); 2-methylpyrazine (99+%, Aldrich), (MPZ) was distilled under reduced
pressure (95 oC/18mmHg). Ethylene carbonate (98%, Aldrich) (EC) was distilled under vacuum (95 oC/1mmHg, mp: 37-39 oC). 2,6-Dichloropyridine (98%, Aldrich), (DCPY) was recrystallized twice from ethanol and dried under vacuum (mp: 86-88 oC).
Pyridine (99%, Fisher) (PY) and 2,6-dimethylpyridine (2,6-lutidine) (99%, Aldrich) (DMPY) were refluxed with sodium hydroxide (NaOH) pellets and distilled at one atmosphere pressure. 2,4,6-Trimethylpyridine (2,4,6-collidine)(99%, Aldrich) (TMPY) was refluxed with NaOH pellets and distilled under reduced pressure (61oC/13mmHg). 2,2’-Bipyridyl (BPY) was recrystallized from hexane/ethanol and dried under vacuum (mp: 70-71oC). 1,3-Dimethyl-2-imidazolidinone (98%, Aldrich) (DMI) was distilled under vacuum (100 oC/1mmHg). ε-Caprolactone (98%, Aldrich), (CPLA) was distilled under vacuum (100 oC/13mmHg), and γ-butyrolactone (99+%, Aldrich) (GBL) was purified by passing through a neutral alumina column. 2-Coumaranone (97%, Aldrich) (CMN) was recrystallized from hexane/methanol (mp: 47-50 oC); dimethyl terephthalate (99+%, Aldrich) (DMT) was recrystallized from methanol (mp:140-142oC); trimethyl-1,3,5-benzene-tricarboxylate (98%, Aldrich) (TMB) was recrystallized from methanol (mp1: 132-134 oC; mp2: 142-145 oC); 1,4-dimethoxy-benzene (hydroquinone dimethyl ether) (99%, Aldrich) (DMB) was recrystallized from hexane/methanol (mp:
- 33 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 56-59 oC). Methanol was purchased from Fisher and used as solvent or precipitant without further purification.
§2.2.2 Polymerization Apparatus The polymerization was carried out in an ACE-8648 Pressure Tube with AceThred and Plunger Valve. The tube is a heavy-wall tube (max pressure: 10 atmospheres), with bushing and plunger valve that allows purging of the tube. The closed bottom plunger has a hole inside that when positioned in relation to O-Ring seal, will open the tube to the atmosphere; i.e., pull to close, push to open as shown in Scheme 2.1:
Scheme 2.1. Schematic illustration of Ace-8648 Pressure Tube.
The polymerization tube was connected with a high vacuum line, which had multiple polymerization tubes attached, shown in Scheme 2.2:
- 34 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Scheme 2.2. Schematic of the polymerization set-up with a vacuum line and pump system; a: Nitrogen flow; b1 to b7: two-way stopcock; c: threeway stopcock; A: vacuum pump; B: trap; C: vinyl chloride tank; D: recycle pressure tube; E1 & E2: polymerization pressure tubes (Ace8648); F: digital vacuum gauge.
§2.2.3 Bulk Polymerization of Vinyl Chloride For all bulk polymerization runs, the following procedure was used.
1. Put required amount of initiator AIBN and additive into the polymerization pressure tubes E1 and E2 and connect them to the vacuum line as shown in Scheme 2.2. Cool E1 and E2 to -205 oC in a liquid nitrogen bath and cool the trap B in liquid nitrogen. - 35 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2. Open b3, b4, b5, b6, close b1, b2, b7, switch 3-way-c to trap B and evacuate to high vacuum (~10μ) for about 15 minutes.
3. Then close b5 and open b1 to charge vinyl chloride into the pressure tubes E1 and E2. Close b1 when vinyl chloride reaches the marked level.
4. When vinyl chloride is completely frozen by liquid nitrogen, open b5 to evacuate the charged polymerization tubes E1 and E2, and keep the whole system at high vacuum for 10 minutes. Then switch 3-way-c to nitrogen flow to let nitrogen purge the polymerization tubes E1 and E2. When the pressure reaches one atmosphere, switch 3way-c back to the vacuum pump system.
5. Repeat the purging and pumping three times, then close b4 and b3 and remove liquid nitrogen bath to warm the vinyl chloride. After the vinyl chloride is melted, freeze it again in the liquid nitrogen bath and repeat procedure 4. After the freezing-pumpingmelting cycle has been performed three times, close b3 and b4 and seal the polymerization tubes E1 and E2 by pulling up the plunger, and then disconnect them from the vacuum line.
6. Open b7, and stop the vacuum pump. Keep the pressure tubes E1 and E2 sealed. Warm and weigh the tubes to calculate the exact amount of vinyl chloride added. Put the sealed pressure tubes into a 55 ±0.5oC oil bath for 1-2 hours. Mark the starting time, when poly(vinyl chloride) (white powder) is seen to precipitate from the vinyl
- 36 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ chloride monomer (colorless liquid). Stop the polymerization at low conversion by cooling in a liquid nitrogen bath.
7. Re-connect the pressure tubes into the vacuum line and recycle the un-reacted vinyl chloride monomer into pressure tube D. Start the vacuum pump to remove residual vinyl chloride from the polymer particles. Weigh the pressure tubes again to measure the amount of poly(vinyl chloride) produced and the conversion.
8. Dissolve the poly(vinyl chloride) in fresh tetrahydrofuran (THF) and precipitate the product into a large amount of methanol (10 times larger than THF). Repeat the dissolving and precipitating procedure twice to completely remove the additive and unreacted AIBN, and dry the final PVC in a vacuum oven at 60 oC for 2 days.
§2.2.4 Suspension Polymerization of Vinyl Chloride Suspension polymerization of vinyl chloride was performed in a 3L pressure reactor at OxyVinyls’ pilot plant. The polymerization temperature was 53±0.5 oC and the initiator used was diisobutyl peroxydicarbonate (SBP), synthesized by reaction of secbutyl chloroformate with H2O2 plus NaOH. The polymerization was stopped at a specified pressure drop or time limit by adding diethylhydroxylamine (DEHA). A recipe for the suspension polymerization of vinyl chloride in the presence of dimethyl terephthalate (DMT) is given in Table 2.1.
- 37 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.1. A selected recipe for suspension polymerization of vinyl chloride at 53 oC. Ingredient Parts Weight Note DM water
395
2Kg
Flush water
1
5.23g
Dimethyl terephthalate
1
16.15g
100
519.70g
ALCOTEX 72.5a)
0.110
13.56g
METHOCEL F50b)
0.020
5.53g
VINOL 540c)
0.020
2.35g
Diisobutyl peroxydicarbonate (SBP)d) ALCOTEX 72.5
0.045
0.24g
0.010
1.23g mix with SBP
Diethylhydroxylamine (DEHA) Sodium hydroxide
0.005
0.03g shortstop
0.009
0.94g with shortstop
Vinyl Chloride
a): Polyvinyl alcohol (degree of hydrolysis: 71.5-73.5 mole%), commercially available from Synthomer. b): A cellulose ether, commercially available from Dow. c): A partial hydrolyzed polyvinyl alcohol, commercially available from Air Products & Chemicals, Inc. d): Acronym SBP, commercially available from Akzo Nobel Polymer Chemicals.
§2.2.5 GPC Characterization of PVC Samples GPC measurements were performed using a Waters GPC system equipped with a Model 510 Pump, Model 410 Refractive Index Detector, Model 996 Photodiode Array Detector and a Column Heater Module with a Styragel HR 5E column and a Styragel HR 4E column (5μm, 100Å, 7.8x300mm) in series. The GPC system was controlled by a stand-alone workstation through a Waters Bus LAC/E Card. HPLC grade tetrahydrofuran - 38 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ was used as the mobile phase and the flow rate was set to 1.0 mL/min. The Styragel columns were kept at 35 oC in the column heater module. Poly(vinyl chloride) narrow molecular weight distribution standards (purchased from Polymer Standards ServiceUSA) with molecular weights of 327,000, 213,000, 146,000, 119,000, 94,000, 64,000, 48,000, 36,000, and 14,200 were used to make a GPC calibration curve. All the dried PVC samples, including the standards, were dissolved in cyclohexanone (CHX) solvent overnight and heated at 90-120 oC for 60 minutes prior to injection. The hot solution was injected to avoid aggregation. The concentration of injected PVC solution was 1~2 mg/mL and a volume of 100~200 μL solution was injected into the GPC system. The data acquisition and processing was handled by Millennium32 Software provided by Waters. The number-average molecular weight (Mn), weight-average molecular weight (Mw), z-average molecular weight (Mz) and the polydispersity index (PDI) of the samples were calculated from the chromatograms.
§2.2.6 Dynamic Thermal Stability Test of PVC Samples Some suspension polymerization PVC samples were chosen to be studied using the Dynamic Thermal Stability (DTS) test in a Brabender Plasticorder equipped with a mixing chamber, operating at 190 oC with a rotor speed of 60 rpm. About 63 g of virgin PVC resin were mixed with specified plasticizers and lubricants and the melted mixture was extruded strip by strip at 2-minute intervals; the strips were cooled and pasted onto a record pad. The test was stopped when the extruded PVC strip had turned black. The DTS time was defined as the time required, taken at 2-min intervals, for the melted mass in Brabender chamber to turn a relatively dark color. - 39 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.2.7 Static Thermal Stability Test of PVC Samples 12 PVC samples were dissolved in dichlorobenzene in 12 5-mm NMR tubes. The concentration was 5wt% for each sample and the tubes were held at 90±0.5 oC for 14 hours; then all tubes were removed and the color changes for the polymer solution were recorded and compared by visual inspection.
§2.3 Results and Discussion §2.3.1 Molecular Weight and Molecular Weight Distribution of PVCs §2.3.1.1 Polymer aggregation in solution Gel
Permeation
Chromatography
(GPC),
also
called
Size
Exclusion
Chromatography (SEC), is a very rapid analytical technique for measuring the molecular weight (MW) and the molecular weight distribution (MWD) of polymers. GPC separation is based on the molecular size of the dissolved polymer; thus the polymer must be dispersed at a molecular level in the solution. However, the molecular weight and size information derived from PVC solutions is often distorted because the polymer trends to form aggregates even in dilute solution.
Macromolecular aggregation of PVC was first described by Doty and coworkers17 back in 1947. The association of PVC molecules into densely packed aggregates in dilute dioxane was demonstrated by osmotic-pressure, light-scattering, and ultracentrifuge measurements17. Since then, various procedures have been used to disintegrate these aggregates into single PVC molecules in solution to prevent anomalous
- 40 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ molecular weight measurements18-24. Lyngaae-Jφrgensen18, 19 found that heating the PVC solution in THF at 120 oC for 3 hours could dissociate the aggregates. Rudin and Benschop-Hendrychova20 reported that the ultrasonic treatment of the PVC solution in THF at room temperature for 15 minutes destroyed the aggregates; the simultaneous degradation of PVC was prevented by adding a small amount of a nonionic surfactant to the THF solution. Abdel-Alim and Hamielec21 reported that heating the solution at 90 oC for 10 min was adequate for PVC prepared at polymerization temperatures between 30 and 70 oC. However, for PVC prepared at -50 oC, a temperature of 200 oC was needed to dissociate the polymer into single PVC molecules22. Pang and Rudin23 reported the preparation of aggregate-free PVC solution in 1,2,4-trichlorobenzene (TCB) by heating at 120 oC for 12 hrs, followed by GPC measurements at 110 oC, with TCB as the mobile phase. Recently, Manabe and coworkers24 reported a new method. PVC was dissolved in TCB, held at 130-140 oC for 6 hours, precipitated in methanol, and dried. Aggregate-free PVC solution could be prepared by dissolving the pre-treated PVC in THF at room temperature. They also pointed out that the existence of aggregates sometimes could not be observed by a refractive index detector, but a light-scattering detector showed them clearly.
Generally, the molecular aggregation in PVC solution is affected by many factors, such as the temperature, the molecular weight and the tacticity of PVC (polymerized at different temperatures), the concentration of the polymer, the solvent used, the method of dissolution and so on. The most important factor is the molecular weight of the PVC. In this dissertation, the number average molecular weights of the PVC samples were mostly
- 41 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ in the range of 40,000-60,000 (Pn=640~960), and cyclohexanone (CHX) was used to dissolve PVC samples. A 0.1-0.2% CHX solution of PVC was prepared and heated at 110 oC for various periods of time prior to injection. A volume of 100~200 μL solution was injected into the GPC system. Since PVC does not have UV sensitive groups, only the RI defector gave positive signal peaks. The RI chromatogram for PVC/control is shown in Figure 2.1. (Chromatograms for other samples are shown in Figures 2.3, 2.4 and 2.5. They are discussed later.)
10
11
12
13
60 min
15.5
13.5
0 min
14 15 16 Retention time (minute)
17
18
19
20
Figure 2.1. GPC traces of PVC/control prepared by bulk polymerization at 55 oC, initiated by AIBN (⎯) before heating treatment; (⎯) after heating at 110 oC for 60 minutes.
Figure 2.1 shows two peaks for the untreated PVC/control, a small, higher molecular weight one at ~13.5 min, and a large, lower molecular weight one at ~15.5 min. After the solution was heated at 110 oC for 1 hour, the peak at 13.5 min disappeared; the lower molecular weight peak was unchanged and showed unimodal distribution. The high molecular weight shoulder peak can be ascribed to molecular aggregation of PVC in - 42 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ cyclohexanone (CHX) solution. By comparison with the calibration curve, it was found that the molecular weight of the ‘aggregate peak’ in Figure 2.1 was about 6-8 times that of the corresponding disaggregated PVC. Since aggregated polymers occupy a smaller volume than the unaggregated polymers do, a molecular weight increase of 6-8 times probably means 15-20 molecules in the aggregate. After disaggregation treatment, the molecular weights of the PVC polymerized at 55 oC in the absence of additives were calculated as Mn=43,700, Mw=80,100 (Mw/Mn=1.83). For most of the PVC samples studied, the aggregates could be effectively disaggregated by heating at 110 oC for 60 minutes; this pretreatment method was applied to all of the PVC samples polymerized in bulk or in suspension, in the presence or absence of organic additives, to determine the weight average and number average molecular weights.
§2.3.1.2 GPC data for bulk polymerized PVCs
Representative GPC data for PVCs prepared by bulk polymerization initiated by AIBN in the presence of various additives are listed in Table 2.2. The number average molecular weight of PVC prepared in the absence of additive is about 43,600 and the molecular weight distribution is about 1.83. For PVC prepared in the presence of various additives, the molecular weight varies. PVCs prepared in the presence of 1 mole% of 1,3dimethyl-2-imidazolidinone (DMI), 2-methyl benzothiazole (MBTZ), 2-methylpyrazine (MPZ), pyridine (PYR), and 2,4,6-trimethylpyridine (TMPY) have lower molecular weights (20~40% loss compared to normal polymer); these additives are relatively strongly basic nitrogen compounds. Figure 2.2 shows that the dichlorobenzene solutions of PVCs prepared in the presence of DMI, MBTZ, MPZ, PYR, and TMPY darkened after - 43 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ heating in dichlorobenzene at 90 oC for 14 hours. However, the other samples remained colorless after the same treatment. It is very interesting that, for PVC prepared in the presence of a less basic compound, 2,6-dichloropyridine (DCPY), the molecular weight did not decrease. 14 hours at 90 oC static thermal stability test also shows that the resulting polymer is reasonably stable. Further studies on the effect of DCPY will be discussed in section 2.3.7. The molecular weights slightly decreased or stayed the same for PVCs prepared in the presence of 2-coumaranone (CMN), 1,4-dimethoxybenzene (DMB), and phosphine oxide compounds. However, the number average molecular weights increased about 8~15% for PVCs prepared in the presence of 1.0% of dimethyl terephthalate (DMT), ethylene carbonate (EC) or trimethyl-1,3,5-benzene-tricarboxylate (TMB). A detailed discussion of the influence of the individual additive on the molecular weights of resulting PVCs will be given in sections 2.3.7, 2.3.8, and 2.3.9.
- 44 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.2. Molecular weights and molecular weight distributions for PVCs prepared by bulk polymerization at 55 oC in the presence of organic additives at given AIBN concentration. Mw Mn Mz Mw/Mn Mz/Mw Sample designationa) AIBN319b)
80,100
43,700
128,600
1.83
1.61
AIBN320 b)
82,800
43,600
139,200
1.90
1.68
AIBN326 b)
80,600
43,500
132,600
1.85
1.64
AIBN416 b)
74,000
40,900
116,500
1.80
1.57
GBL01d
78,700
44,200
119,700
1.78
1.52
GBL02b
78,500
44,200
117,100
1.77
1.49
GBL04b
74,600
42,400
113,000
1.76
1.52
CMN001a
75,500
41,400
118,300
1.82
1.57
CMN001b
68,000
37,200
103,500
1.83
1.52
DCPY005d
80,300
45,000
120,200
1.78
1.49
DCPY01h
84,900
46,900
129,000
1.83
1.54
DMB001a
73,400
41,500
112,600
1.77
1.53
DMB001b
80,100
45,200
120,900
1.77
1.51
DMI01a
50,500
24,700
78,200
2.04
1.55
DMI05d
22,600
13,000
27,100
1.73
1.20
DMT01a
89,500
50,200
135,200
1.78
1.51
DMT01b
99,500
53,200
152,200
1.87
1.53
DMT01d
85,500
46,200
128,300
1.85
1.50
EC01c
85,100
48,100
127,700
1.77
1.50
EC04b
82,500
45,600
122,100
1.81
1.48
(continued on next page) - 45 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.2 (continued) EC04d
80,700
45,600
121,00
1.77
1.50
MBTZ01b
63,700
31,200
107,900
2.04
1.69
MPZ01a
63,700
31,200
100,900
2.04
1.58
PYR01a
75,300
37,000
132,700
2..04
1.76
TBPO01b
74,000
44,400
105,800
1.66
1.43
TBPO04a
84,900
39,700
155,100
2.13
1.82
TBPO05c
94,400
41,700
204,900
2.26
2.17
TEPO005a
79,400
45,600
110,900
1.74
1.40
TEPO01b
80,100
38,700
150,400
2.06
1.87
TMB005d
82,000
45,500
125,400
1.80
1.53
TMB01a
85,000
47,300
128,600
1.80
1.51
TMPO01a
70,400
36,400
100,800
1.93
1.43
TMPO01c
67,100
39,200
93,800
1.71
1.40
TMPY01a
72,500
35,800
139,400
2.02
1.92
TMPY05c
41,000
21,500
55,500
1.91
1.35
TPPO01a
66,000
40,800
92,000
1.61
1.39
TPPO01b
85,500
38,400
145,100
2.23
1.70
a): Samples prepared in the presence of additives are named by additive abbreviation plus a 2 or 3 digit number followed by a lower-case letter (the number represents the mole ratio of additive to monomer(decimal point omitted), and the lower-case letter identifies different runs); GBL: γ-butyrolactone; CMN: 2-coumaranone; DCPY: 2,6bichloropyridine; DMB: 1,4-dimethoxybenzene; DMI: 1,3-dimethyl-2-imidazolidinone; DMPY: 2,6-dimethyl-pyridine; DMT: dimethyl terephthalate; EC: ethylene carbonate; MBTZ: Methyl benzothiazole; MPZ: 2-methyl-pyrazine; PYR: Pyridine; TEPO: triethylphosphine oxide; TMB: trimethyl-1,3,5-benzenetricarboxylate; TMPO: trimethylphosphine oxide; TMPY: 2,4,6-trimethyl-pyridine; TPPO: triphenylphosphine oxide. b): Control polymerization was initiated by AIBN in the absence of additives; samples were named as AIBN plus a random 3-digit number to distinguish different runs. - 46 -
- 47 Test conditions: 5% solution in dichlorobenzene, 90 oC, 14 hours
Figure 2.2. Thermal stability test for some PVC samples.
PYR01a TMPY05c TMPY01a DMI05d DMI01a MBTZ01b TPPO01b DCPY01a TBPO03d EC01c EC04d AIBN319
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.1.3 PVC/DMT01b and PVC86k aggregation in solution
10
11
12
13
60 min
15.3
13.2
0 min
14 15 16 Retention time (minute)
17
18
19
20
Figure 2.3. GPC traces of PVC/DMT01b prepared by bulk polymerization at 55 oC, initiated by AIBN in the presence of 1.0 mole% of DMT: (⎯) before heating treatment; (⎯) after heating at 110 oC for 60 minutes.
(A) PVC/DMT01b The GPC chromatograph of PVC/DMT01b, obtained by polymerization at 55 oC in the presence of 1.0 mole% of dimethyl terephthalate (DMT), is shown in Figure 2.3. It can be seen that PVC/DMT01b also has a bimodal distribution for untreated sample as did the PVC/control. There is a small peak at 13.2 min, and a large peak at 15.3 min. After the solution was heated for one hour at 110 oC, the peak at 13.2 min disappeared and the peak at 15.3 min remained but with a very small shoulder on the high molecular weight side. After heating for 3 more hours at 110 oC, the shape for the 15.3 min peak did not change, which means that high molecular weight shoulder is not aggregated polymer peak. The reason for having that shoulder is not clear. The molecular weights were calculated as Mn=50,200, Mw=89,500, with a molecular weight distribution of 1.78. The - 48 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ weight average molecular weight of PVC prepared in the presence of 1.0 mole% of DMT is about 12% higher than that of PVC/control, prepared without additive (Mn=43,700, Mw=80,100, Mw/Mn=1.83). The GPC curves for disaggregated PVC/DMT01b and PVC/control are shown in Figure 2.4. It can be seen that the main eluent peak for PVC/DMT01b came out 0.2 min earlier than that for the PVC/control.
Control
15.3 15.5
DMT01b
10
11
12
13
14 15 16 Retention time (minute)
17
18
19
20
Figure 2.4. GPC traces of PVC/DMT01b and PVC/control. (both dissolved samples were heated in cyclohexanone at 110 oC for 60 minutes).
(B) PVC86k In our polymerization runs, we found that after degassing, vinyl chloride monomer in Ace-8648 pressure tube spontaneously polymerized at room temperature under visible light. The ‘spontaneous’ polymerization was extremely slow. After 7 days a very thin white film formed inside the tube wall and the colorless monomer liquid became opaque. After venting the unreacted monomer, the white polymer film was collected and the conversion was less than 0.1%. This type of polymer was purified by - 49 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ the same procedure as other PVCs and named PVC86k (86k is its number-average molecular weight).
10
11
12
13
60 min
14.7
13.2
0 min
14 15 16 Retention time (minute)
17
18
19
20
Figure 2.5. GPC traces of PVC86k prepared by bulk polymerization at room temperature, initiated by visible light: (⎯) before heating treatment; (⎯) after heating at 110 oC for 60 minutes.
The GPC chromatograph of PVC86k is shown in Figure 2.5. Since this polymer is polymerized at room temperature, the molecular weight is higher than that of those prepared at 55 oC. When Figures 2.5 and 2.1 are compared, the so-called ‘aggregate peak’ moves from 13.5 min to 13.2 min and the main peak moves from 15.5 min to14.7 min. After the solution was heated at 110 oC for 60 minutes the ‘aggregate peak’ at 13.2 min disappeared; the 14.7 min peak remained almost unchanged. After the solution was heated for 5 more hours at 110 oC, the chromatograph shape was still unchanged; the sample was slightly degraded. The molecular weights of PVC86k were calculated as Mn=86,000, Mw=157,400 (Mw/Mn=1.83).
- 50 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.1.4 GPC data for suspension polymerized PVCs Representative data on the suspension polymerization of vinyl chloride in the presence of some additives are listed in Table 2.3. By comparing Tables 2.3 and 2.2, we can see that the number-average molecular weight (Mn) of suspension polymerized PVC is about 30% higher than that of the bulk polymerized PVC. Mn for PVC prepared in the absence of additive is about 55,400, the molecular weight distribution is 1.85. The additives used in suspension polymerization were γ-butyrolactone (GBL), dimethyl phthalate (DMP) and dimethyl terephthalate (DMT). The concentration of GBL used was 0.2 mole% relatively to VC and the polymerization conversion was about 56%. The molecular weight of the PVC prepared in the presence of GBL is almost the same as that of the polymers without additive. Since GBL is water soluble, most of GBL goes into the water layer with very little in the vinyl chloride micelles. Thus GBL did not influence the suspension polymerization of vinyl chloride. As shown in Table 2.3, the addition of DMP at the 0.1 mole% level increases Mn of the resulting PVC by about 12% (Mn=61,800). DMT at 0.5 mole% level also increases the PVC molecular weight about 12%. For polymerizations stopped at low conversion ( 1 . Substituting equation (2.5) into equation (2.4), we get
dX = k p (1 − X − AX + AQX )[ M • ]I dt 1
⎛ fk ⎞ 2 1 = [1 + ( AQ − A − 1) X ]k p ⎜⎜ d ⎟⎟ [ I ] 2 ⎝ ktI ⎠ = (1 + qX ) k
' p
1 2 0
[I ]
− kdt ) ……………......................(2.6) 2
exp( 1 2
⎛ fk ⎞ Where q = AQ − A − 1, k p' = k p ⎜⎜ d ⎟⎟ , and [I ]0 is the concentration of initiator at t = 0 . ⎝ ktI ⎠ By integrating equation (2.6) between 0 and X for
X
and between 0 and t for t , one
obtains:
X=
1 ⎧⎪ ⎡⎛ 2q ⎞ ' 1 2 ⎛ ⎛ − k d t ⎞ ⎞⎤ ⎫⎪ ⎟ ⎟⎟⎥ − 1⎬ ………………...(2.7) ⎨exp ⎢⎜⎜ ⎟⎟k p [I ]0 ⎜⎜1 − exp⎜ q ⎪⎩ ⎣⎝ k d ⎠ ⎝ 2 ⎠ ⎠⎦ ⎪⎭ ⎝
⎛ ⎛ − k t ⎞⎞ k t If k d t is less than 0.2, ⎜⎜1 − exp⎜ d ⎟ ⎟⎟ ≈ d , then equation (2.7) can be ⎝ 2 ⎠⎠ 2 ⎝ simplified to
- 57 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ X =
[ (
) ]
1 12 exp qk p' [I ]0 t − 1 ………………………………...(2.8) q
and equation (2.8) can be rewritten as
k p' =
ln (qX + 1) …………………………………………(2.9) q[ I ]10/ 2 t
Thus equations (2.8) and (2.9) can describe the kinetics of bulk polymerization of vinyl chloride at low conversion. Here q , which could be called ‘gel factor’, can be determined by fitting experimental X ~ t curves with equation (2.8). Many papers35-39 reported those X versus t or X versus t • [I ]0.5 curves in the literature. The initiators included 2,2’-azobisisobutyronitrile (AIBN); benzoyl peroxide (BPO), LPO: lauroyl peroxide (LPO), as well as γ-rays. The experimental q was found to be in the range of 3.3-6.4. Talamini et al.35 recently summarized the literature data for q and found q was not very sensitive to the type of initiation or to the nature of the chemical initiators. In this dissertation, q = 6 was used to calculate the experimental rate constant k p' for bulk polymerization of vinyl chloride at 55 oC, since it fits very well with the re-plots of many published x~ t curves.
From equation (2.6), we define the initial polymerization rate, R p0 , as:
⎛ dX ⎞ R p0 = M 0 ⎜ ⎟ ⎝ dt ⎠ X →0
= k p [ M • ]I M 0 = k p' [ I ]10/ 2 M 0 …………………………(2.10)
- 58 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The initial polymerization rate R p0 can be determined by the initial slope of X ~ t curve multiplied by M 0 . From equation (2.10) we see that R p0 is proportional to the
square root of [I ]0 . Different polymerization rate expressions were derived using other models28, 31, 34, but the same initial polymerization rate, equation (2.10), was obtained from those models. Here, we need this initial polymerization rate to evaluate the mechanism of the vinyl chloride polymerization.
In bulk polymerization of vinyl chloride, if we ignore chain transfer to the initiator, the initial degree of polymerization ( Pn0 ) can be described as follows:
⎛ fk d k t 1 ⎜ C = + M ⎜ k2 Pn0 ⎝ p
1 2
1
⎞ [I ]2 ⎟ ⎟ [ M ] ……………………………..(2.11) ⎠
Where k t is the rate constant for propagating radical-radical termination and CM is the chain transfer constant to monomer. Here we assume the chain transfer constants in both monomer-rich and polymer-rich phases are the same. By using equations (2.9), (2.10) and (2.11), we can get the experimental rate constant
k 'p , the initial polymerization rate R 0p
and the chain transfer constant to monomer C M for the bulk polymerization of vinyl chloride.
- 59 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.3 Bulk Polymerization of Vinyl Chloride at 55 oC initiated by AIBN The free radical bulk polymerization of vinyl chloride was conducted in an ACE8648 Pressure Tube with Ace-Thred and Plunger Valve. The initiator used was 2,2’azobisisobutyronitrile (AIBN), and the concentration of AIBN varied from 2 to16 x10-3 mole/L, listed in Table 2.4. The polymerization was stopped at low conversion. As shown in Table 2.4, the polymerization had a short induction time, in the range of 1~9 minutes, but mostly around 3-5 minutes. The induction period is probably due to impurities or residual O2 in the polymerization system. The conversion was 3-7%. From the literature34, the half-life of initiator AIBN was calculated as 38.5 hours at 55 oC. The polymerization was stopped in less than 2 hours so the concentration of the initiator could be treated as constant. Using equation (2.9), the experimental rate constant k 'p was calculated and is listed in Table 2.4. It can be seen that within the experimental error k 'p is constant in the range of [I]=2-16x10-3 mole/L. The average k 'p is estimated to be 11x10-5 L0.5•mole-0.5•s-1.
- 60 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.4. Bulk polymerization of vinyl chloride initiated by AIBN at 55oC. AIBN VC Ind* Time PVC Conv k 'p x105 Run (**) (%) (mg) (g) (min) (min) (g) AIBN8a 8.8 25.20 4 115 0.69 2.74 8.7
R 0p x105 (***) 4.7
AIBN8b
8.3 23.76
5
133
0.90
3.79
10.1
5.4
AIBN16a
16.1 25.90
9
100
1.03
3.98
10.5
7.4
AIBN16b
15.8 24.53
5
94
0.86
3.51
9.8
7.1
AIBN319
32.5 27.64
1
67
1.16
4.20
12.0
11.5
AIBN320
33.0 26.65
4
75
1.31
4.92
12.0
11.6
AIBN326
33.0 25.07
3
76
1.43
5.70
13.0
12.8
AIBN416
32.3 21.92
2
87
1.61
7.34
13.4
13.5
AIBN430
31.6 26.15
5
77
1.59
6.08
14.2
13.3
AIBN507
33.1 25.23
2
78
1.21
4.80
11.0
11.0
AIBN64a
64.4 24.61
3
45
0.88
3.58
10.3
14.9
AIBN64b
64.5 25.68
1
58
1.57
6.11
13.1
17.6
*: induction time; **: L0.5•mole-0.5•s-1; ***: mole•L-1•s-1.
- 61 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.5. Molecular weights and molecular weight distributions for PVCs prepared by bulk polymerization at 55 oC initiated by AIBN. Sample 1 / Pn0 designation Mw Mn Mz Mw/Mn Mz/Mw x103 AIBN8a 91,000 49,300 141,800 1.85 1.56 1.268 AIBN8b
89,800
49,100
138,700
1.83
1.55
1.273
AIBN16a
86,300
47,500
135,500
1.82
1.57
1.316
AIBN16b
89,800
49,000
140,600
1.83
1.57
1.276
AIBN319
80,100
43,700
128,600
1.83
1.61
1.430
AIBN320
82,800
43,600
139,200
1.90
1.68
1.433
AIBN326
80,600
43,500
132,600
1.85
1.64
1.437
AIBN416
76,000
42,200
119,300
1.80
1.57
1.481
AIBN430
70,500
41,200
102,200
1.71
1.45
1.517
AIBN507
81,000
44,500
128,000
1.82
1.58
1.404
AIBN64a
78,000
41,000
126,900
1.90
1.63
1.524
AIBN64b
78,700
41,300
133,500
1.91
1.69
1.513
The molecular weights and molecular weight distributions (Mw/Mn) of the resulting PVCs were determined by GPC. Narrow distribution PVC samples (14,000~323,000), purchased from Polymer Standards Service-USA, were used as standards. The polymers were dissolved in cyclohexanone and heated at 90-110 oC for 60 minutes before injecting into the GPC system. The results are listed in Table 2.5. The average number molecular weights of the PVCs were found to be 41,000-49,100 and the molecular weight distributions were 1.71-1.91. The degree of polymerization ranged from 660 to 790, depending on the initiator concentration. - 62 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ In this dissertation, since all the polymerizations were conducted in ACE 8647 pressure tubes, not in dilatometers, no X ~ t curve was obtained for the individual polymerization run, so it is not possible to get the initial polymerization rate R p0 from the initial slope of X ~ t curve. However, since the polymerization was stopped at very low conversion (3-7%), R p0 can be approximately determined as X M 0 divided by (1 + qX ) to t
correct for the gel effect:
R p0 ≈
X M 0 …………………………………………(2.12) t (1 + qX )
Values of R p0 calculated by using equation (2.12) were also listed in Table 2.4. It can be seen in the range of [AIBN]=2-16x10-3mole/L R p0 ranged from 5-18x10-5 mole•L1
•s-1. At 6.4x10-3mole/L initiator concentration, the average R 0p was about 10.6x10-5
mole•L-1•s-1. In the following study of vinyl chloride polymerization in the presence of additives, the initiator concentration was kept constant at 6.4 x 10-3 mole/L, and R p0 = 10.6x10-5 mole•L-1•s-1 was used as a reference to evaluate the effect of additives on the polymerization rate.
The plot of natural logarithm of R p0 against natural logarithm of [AIBN] is shown in Figure 2.7. The reaction order with respect to the initiator obtained from the slope of Ln(R0p) vs Ln([AIBN]) is 0.62±0.04. This is significantly greater than 0.5, the theoretical order indicated from equation (2.10). We think the conventional assumption of a propagating radical-radical termination, i.e. Rn• − Rn• termination, is in doubt. Actually, - 63 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ there is disagreement in the literature on the initiator order in bulk polymerization of vinyl chloride. The value of 0.5 was reported by Bengough44 for bulk polymerization at 40 oC. Talamini46 reported a value of 0.56 for vinyl chloride bulk polymerization at -50 o
C, and a value of 0.67 was reported by Olaj30 for bulk polymerization of vinyl chloride
at 50 oC. Danusso45 found the initiator order was 0.7 for vinyl chloride polymerization in methanol (in which vinyl chloride is soluble and the polymer is insoluble). Moreover, Cotman et al.47 claimed that the initiator order was 0.5 at 5% conversion but it became 0.56 at 10% conversion for bulk polymerization of vinyl chloride.
-Ln( R0p ) ( moleL -1s-1 )
12 y = 0.616x + 5.972 R2 = 0.9494
11 10 9 8 7 6 3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
-1
-Ln( [AIBN] ) ( moleL )
Figure 2.7. The plot of Ln(R0p) as a function of Ln([AIBN]) for bulk polymerization of vinyl chloride at 55 oC.
- 64 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.4 New Kinetic Model for Vinyl Chloride bulk Polymerization initiated by AIBN In bulk polymerization, since PVC is not soluble in its own monomer, the produced polymer precipitates from vinyl chloride monomer at the very beginning of the polymerization, and the polymerization becomes heterogeneous. Also, since the propagating radical is so active, it can easily chain transfer to monomer to terminate the chain and generate “monomer radicals”. These monomer radicals could be chlorine atoms, chlorine atoms π-bonded with monomers or dichloroethane radicals. These radicals could add monomer to reinitiate polymerization or terminate with propagating radicals to form polymers. It is generally believed53,
54, 56-58
that the termination rate
constant for vinyl chloride polymerization is about 109 L•mole-1•s-1, which implies a very high diffusion rate for the radicals involved in the termination. It is extremely difficult for two propagating radicals to find each other at this high speed in the bulk and gel phase. Instead, small radicals such as Cl● or Cl● π-bonded with a monomer are more likely to diffuse rapidly and terminate propagating radicals. The dichloroethane radical is expected to be highly reactive and add to monomer. So, we believe, the propagating radicals either chain transfer to monomer, or terminate with these small radicals. Propagating radicalradical termination is very rare and can be ignored. The mechanism of vinyl chloride polymerization under bulk condition can then be described as following.
(1) Initiation reactions: kd I ⎯⎯→ 2I • ki I • + M ⎯⎯→ R•
- 65 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Here I is the initiator, k d the rate constant for the initiator decomposition, ki , the rate constant for initiator radical adding monomer M . Since the first step is the ratedetermining step, the initiation rate ( Ri ) can be described as Ri = 2 fk d [ I ] ………………………………………………………(2.13) Here f is the initiator efficiency, [ I ] , initiator concentration.
(2) Propagation reactions: p Rn• + M ⎯⎯→ Rn•+1
k
Here, k p is the rate constant for the propagation reaction; we assume k p is chain-length independent. Rn• is the propagating radical and M , the monomer, vinyl chloride. The initial rate of propagation, R0p , can be described as
R p0 = k p [ Rn• ][M ] ………………………………………...(2.14) (3) Chain-transfer reaction: k tr Rn• + M ⎯⎯→ Pn + M •
Here,
ktr
is the chain-transfer rate constant and M • is conveniently called the the
“monomer radical”. M • could be Cl●, CH2=CH●, CH2=CCl●, ClCH2-ClCH●, or Cl● πbonded with a monomer. We do not need to know what it is at this point (detailed end - 66 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ group analysis will be given in Chapter 4 using NMR technique), and only assume M • is active enough to reinitiate relatively slowly as well as terminate. Since the initiator concentration is about 10-3mole/L and the conversion is less than 10% for almost all of the polymerization runs in this thesis, we ignore chain transfer to initiator and to polymer and only consider chain-transfer to monomer. Thus the rate of chain-transfer reactions, Rtr , is given by:
Rtr = k tr [ Rn• ][ M ] ……………………………………....(2.15)
(4) Reinitiation reaction: Radicals M • may add vinyl chloride monomer (M) to reinitiate: k ri M • + M ⎯⎯→ R•
Here k ri is the rate constant for reinitiation, and the reinitiation rate Rri , is given as
Rri = k ri [ M • ][ M ] ……………………………………….(2.16) Since [ M ] is constant in each phase, the reinitiation reaction is essentially first order for [ M • ] .
(5) Termination reaction: We assume the propagating radicals Rn• mainly terminate with small radicals such as M • , all other terminations such as primary radical termination and mutual propagating radical termination are insignificant. Thus we have:
- 67 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ kt 1 Rn• + M • ⎯⎯→ Pn
Here kt1 is the termination rate constant*. The rate of termination, Rt , is given by
Rt = kt1[ Rn• ][ M • ] ………………………………………..………..(2.17)
Under steady-state assumption, the concentrations of
Rn• and M •
are constant,
so we have the following two equations:
⎛ dRn• ⎞ ⎜⎜ ⎟⎟ = 2 fk d [ I ] + k ri [ M • ][ M ] ⎝ dt ⎠ − ktr [ Rn• ][ M ] − kt1[ Rn• ][ M • ] ≈ 0 ………..….……………(2.18)
⎛ dM • ⎞ ⎟⎟ = k tr [ Rn• ][ M ] − k ri [ M • ][ M ] ⎜⎜ ⎝ dt ⎠ − kt1[ Rn• ][M • ] ≈ 0 ……………………………………...(2.19) Adding equation (2.18) and equation (2.19), we get:
fk d [ I ] = kt1[ Rn• ][ M • ] ………………………………………………(2.20)
Equation (2.20) simply tells us that under the steady-state condition the rate of radical generation is equal to the rate of radical disappearance. __________________________________ * Here we use kt1 as the rate constant for termination of propagating radicals with small radicals to distinguish with it from kt , the rate constant for mutual propagating radical termination, generally used in the literature. - 68 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ From equation (2.19), one gets
[M • ] =
ktr [ M ] [ Rn• ] …………………………..…………..(2.21) • k ri [ M ] + kt1[ Rn ]
Substituting equation (2.21) into equation (2.20) gives equation (2.22):
[ Rn• ]2 −
fk d [ I ] • fk k [ Rn ] − d ri [ I ] = 0 ……………………..(2.22) ktr [ M ] ktr k t1
Equation (2.22) is a quadratic equation, which has two roots, and the meaningful root is:
[ Rn• ] =
fk d [ I ] ⎛ fk d k ri [ I ] ⎞ ⎟ +⎜ 2ktr [ M ] ⎜⎝ ktr kt1 ⎟⎠
1 2
1+
fk d kt1[ I ] …………(2.23a) 4ktr k ri [ M ]2
fk d kt1[ I ] is estimated to be ~3[I]. In this thesis [I] is in the range of 10-3 to 10-2 mole/L, 4 ktr k ri [ M ]2
so fk d k t1[ I ] is much smaller than 1, and can be ignored, i.e., 4 k tr k ri [ M ]2
1+
fk d k t1 [ I ] ≈ 1. 4k tr k ri [ M ] 2
Then equation (2.23a) can be approximated as: 1
1
⎛ fk k ⎞ 2 fk d [ I ] …...............................................(2.23b) [ Rn• ] ≈ ⎜⎜ d ri ⎟⎟ [ I ] 2 + 2k tr [ M ] ⎝ k tr k t1 ⎠ Substituting equation (2.23b) into equation (2.14), one gets: 1
1 fk d k p [ I ] ........................................(2.24a) ⎛ fk d k ri ⎞ 2 2 0 ⎜ ⎟ Rp ≈ k p ⎜ [ ] [ ] + I M ⎟ 2k tr ⎝ k tr k t1 ⎠
- 69 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ or 1
1 ⎛ fk d k ri k p ⎞ 2 2 fk 0 ⎟ ⎜ Rp ≈ ⎜ [ I ] [ M ] + d [ I ] …………………………...(2.24b) ⎟ 2C M ⎝ C M kt1 ⎠
Here C M = k tr / k p is defined as the chain transfer constant to monomer. From equation (2.24a), it can be seen that the initial polymerization rate equations, R p0 , has two terms: the first term is proportional to the one-half power of initiator concentration, the second term to the first power of initiator concentration. In other words, there is no single initiator order which can correlate experimental results at all initiator concentrations. At a narrow range of initiator concentrations, one may be able to fit the initial rates with a single initiator order, but the value of the initiator order may vary with the initiator concentration range. At higher initiation concentrations, the second term becomes more and more important.
For the bulk polymerization of vinyl chloride, the initial degree of polymerization, Pn0 , can be defined as
Pn0 =
R p0 Rtr + Rt
……………………………………………………...(2.25)
Equation (2.25) can be rewritten as
R +R 1 = tr 0 t ……………………………………………………..(2.26) 0 Pn Rp Substituting equations (2.14), (2.15) and (2.17) into equation (2.26), we get:
- 70 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ ktr kt1 [ M • ] 1 ……………………………………………….(2.27) = + Pn0 k p k p [ M ]
From equation (2.21), if [ Rn• ] is very small or k ri is very large, k ri [ M ] >> k t1 [ Rn• ], the contribution for termination reactions to the radical concentration can be ignored and we get
[M • ] ≈
k tr • [ Rn ] ………………………………..………….………(2.28) k ri
Substituting equation (2.28) into equation (2.27) gives equation (2.29): k tr k t1 k tr [ Rn• ] 1 …………………………………………….(2.29) ≈ + Pn0 k p k p k ri [ M ]
Substituting equation (2.23b) into equation (2.29), we have
1 ⎞2
1
ktr ⎛⎜ fk d kt1ktr ⎟ [ I ] 2 fk d k t1 [ I ] ……………………….(2.30a) 1 + ≈ + Pn0 k p ⎜⎝ k ri k p2 ⎟⎠ [ M ] 2k p k ri [ M ]2 or 1
1
⎛ C M fk d kt1 ⎞ 2 [ I ] 2 fk d kt1 [ I ] .....................................(2.30b) 1 ⎟ ⎜ + C ≈ + M ⎟ [ M ] 2 k k [ M ]2 ⎜ k k Pn0 ri p p ri ⎠ ⎝
When equations (2.30a) and (2.11) are compared, we find equation (2.30a) has one extra term. At very low initiator concentrations this extra term is very small compared with the first two terms and can be ignored. Then equation (2.30a) has the same form as equation (2.11). So at low initiator concentration we can - 71 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ see k t ≈ k t1 k tr / k ri . If k ri ≈ k tr , k t 1 will be of the same magnitude as k t . If k ri ≈ 10 k tr ,
k t1 will be 10 times as k t .
§2.3.5 Kinetic Parameters for Bulk Polymerization of Vinyl Chloride at 55 oC Equations (2.24b) and (2.30b) describe the kinetics of bulk polymerization of vinyl chloride. From Tables 2.4 and 2.5, we found that for [AIBN] in the range of 216x10-3 mole/L R 0p was in the range of 5-18x10-5 mole•L-1•s-1 and 1 / Pn0 was in the range of 1.27-1.51 x 10-3. Plots of 1 / Pn0 and R p0 vs [AIBN]0.5 are shown in Figure 2.8. The ○ and □ points show the experimental data and the solid curves are the calculated results from applying least-squares regression for equations (2.24b) and (2.30b) to the experimental data. Here we get CM=(11.3±0.1)x10-4;
fkd=(3.9±0.3)x10-6s-1 and
krikp/kt1=(2.5±0.2)x10-6 L•mole-1•s-1. It should be pointed out that the two solid curves in Figure 2.8 are not straight lines. It can be seen that the curves rise with higher initiator concentrations, which means that the second term in equation (2.24b) and third term in equation (2.30b) become more and more important as initiator concentration rises.
- 72 -
48
1.80
42
1.60
36
1.40
30
1.20
24
1.00
18
0.80
12
0.60
6
1000/Pn
2.00
0.40 0.00
0.02
0.04
0.06 0.08 0.10 [AIBN]0.5 (mole 0.5L-0.5)
0.12
0.14
R0p x 10 5 (moleL -1s-1)
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
0 0.16
Figure 2.8. Plots of 1 / Pn0 and R p0 vs [AIBN]0.5 for bulk polymerization of vinyl chloride at 55 oC. ○ and □ are the experimental data and the solid curves are the calculated results of equation (2.24b) and equation (2.30b) by least-squares regression. Table 2.6. Chain transfer constant to monomer (CM) for vinyl chloride polymerization. Initiatora CMx104 Polym. Temp(oC) Medium Reference AIBN
11.3±0.1
55
bulk
this work
AIBN
12.5b
55
bulk
[29]
LPO
11.0
50
1,2-dichlorethane
[39]
γ-rays
12.7c
55
bulk
[40]
AIBN
10.4
50
bulk
[42]
AIBN
17.1d
55
bulk
[43]
a: AIBN: 2,2’-azobisisobutyronitrile; LPO: lauroyl peroxide. b: calculated using the equation CM=5.78exp(-2768.1/T). c: calculated using equation CM=108.5exp(-3726.3/T). d: calculated using equation CM=125exp(-3676.0/T).
The calculated CM together with the results reported by other authors is listed in Table 2.6. It can be seen that the value of CM determined in this dissertation is very close - 73 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ to the literature results. CM from reference 39 was determined from solution polymerization, and the agreement of the values of CM between bulk and solution polymerization indicates that heterogeneous polymerization does not affect the value of ratio ktr/kp. The decomposition rate constant (kd) of AIBN has been studied by many researchers48-52 for different solvents over a wide temperature range. It was found that kd for AIBN is independent of solvent type and level34. The temperature dependence was given as kd=2.88x1015exp(-13042/RT)34. kd at 55 oC was calculated to be 5.0x10-6 s-1. The AIBN efficiency f is determined as 0.78±0.05 for the bulk polymerization of vinyl chloride from the experimental value of fkd=(3.9±0.3)x10-6 s-1. A value of 0.77 for AIBN was reported by Peterson et al.41 for vinyl chloride polymerization in solution at 50 oC. We can see that our experimental result of f is in very good agreement with the literature value.
To get kt1, we need to know the values of kp and kri. Vinyl chloride polymerization has been extensively studied in the literature, but only few papers reported the kp value for vinyl chloride. Burnett and Wright53 in 1954 first reported kp=6,200 L•mole-1•s-1 at 25 oC and 11,000 L•mole-1•s-1 at 55 oC for polymerization of vinyl chloride in tetrahydrofuran using the rotating sector method. Bengough and Thomson54 in 1965 reported a value of 3,130 L•mole-1•s-1 at 25 oC for vinyl chloride polymerization in chloroform using a modified rotating sector method. Recently Kajiwara and Kamachi55 reported a value of 5,400±1,500 L•mole-1•s-1 at 25 oC for vinyl chloride polymerization in benzene at 25 oC using electron spin resonance spectroscopy. In 1990 Kiparissides et al.56 used kp=5x107exp(-27600/RT) to fit their comprehensive mathematical modeling of vinyl chloride suspension polymerization. Their model - 74 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ predictions were in very good agreement with the experimental results over a wide range of polymerization conditions and were further validated by dynamic simulation of industrial PVC batch suspension polymerization reactors57. From Kiparissides’ result, the value of kp was calculated as 2,020 L•mole-1•s-1 at 55 oC. Up to now, the kp of vinyl chloride has not yet been estimated by the pulse laser polymerization method.
We have determined CM=ktr/kp=11.3x10-4 for polymerization at 55
o
C. If
kp=11,000 L•mole-1•s-1, ktr is calculated as 12.4 L•mole-1•s-1. It was proposed28 that kri was about 1~10ktr. From krikp/kt1=2.5x10-6 L•mole-1•s-1, if kri=ktr we get kt1= 5.5x1010 L•mole-1•s-1; if kri=10ktr we get kt1= 5.5x1011 L•mole-1•s-1. These values of kt1 seem too high, as most models53, 54, 56-58 calculated a termination rate constant of the order of 109. If kp=2,020 L•mole-1•s-1, we get ktr= 2.3 L•mole-1•s-1 and kt1= 1.8x109 L•mole-1•s-1 if we assume that kri=ktr; kt1= 1.8x1010 L•mole-1•s-1 if we assume that kri=10ktr. In 1981, Litt and Chang59 found that for bulk and emulsion polymerization of vinyl acetate, the termination reaction was mainly that of propagating radicals with monomer radicals (generated by chain transfer to monomer), and that the termination rate constant of the propagating radicals with monomer radicals (called k80 in their designation) was determined to be 1.1x109 L•mole-1•s-1 for vinyl acetate at 60 oC. It can be seen that the value of kt1 for vinyl chloride found in this thesis is comparable to that of k80 for vinyl acetate. This value of kt1 is also close to that of kt=2.3x109 L•mole-1•s-1, proposed as propagating radical-radical termination rate constant for vinyl chloride53. Earlier in this section we pointed out that under low initiator concentration k t ≈ k t1k tr / k ri . If we assume that k ri ≈ 1 ~ 10k tr , we get k t ≈ 1 ~ 10k t1 . k t was defined as the rate constant of - 75 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ termination with the assumption that the termination involves two propagating radicals. However, the propagating radical-radical termination assumption can not generate a polymerization rate order with respect to initiator concentration that is larger than 0.5. So we propose here that the termination of propagating radicals with monomer or chlorine radicals is the dominant reaction under heterogeneous condition. We can see that the values of the rate constants match very well with our experimental results and explain why the polymerization rate is proportional to an initiator concentration greater than 0.5 power.
In the above discussion, we simplified equation (2.21) into equation (2.28) by assuming k ri [ M ] >> kt1[ Rn• ] . This is true only when the initiator concentration is very low. As initiator concentration rises, the kt1[ Rn• ] part can not be ignored. Thus we directly substitute equation (2.21) into equation (2.27), to get equation (2.31):
k tr k tr k t1[ Rn• ] 1 = + ……………….…………..(2.31) Pn0 k p k p ( k ri [ M ] + k t1[ Rn• ])
• n
And using [ R ] =
R p0 k p [M ]
to substitute into equation (2.31), we have:
kt1 R p0 1 ktr ktr = + ……………..………..(2.32) Pn0 k p k p (k ri k p [ M ]2 + kt1 R p0 ) Equation (2.32) can be arranged as:
- 76 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
k t1 ( Pn0' − Pn0 ) = R p0 ……………………..………….(2.34) 0 0' 2 (2 Pn − Pn ) k ri k p [ M ] Here Pn0 ' = k p / k tr is defined as the theoretical degree of polymerization under the assumption that the polymer is generated only by chain transfer to monomer. It can be seen that equation (2.34) required no approximation. It describes the molecular weight variation over a wide initiator concentration range. The plot of ( Pn0 ' − Pn0 ) /( 2 Pn0 − Pn0 ' ) vs
R p0 for vinyl chloride polymerization at 55 oC is shown in Figure 2.9, where Pn0 ' was taken as 885 and the straight line is the best fit for the experimental points. From the slope we get krikp/kt1=(1.7±0.2)x10-6 L•mole-1•s-1. This value of krikp/kt1 is close to the value, (2.5±0.2)x10-6 L•mole-1•s-1 we calculated earlier with the assumption of k ri [ M ] >> kt1[ Rn• ] . All
the kinetic parameters for the bulk polymerization of vinyl chloride are
summarized in Table 2.7.
0.8 y = 0.0317x R2 = 0.8948
(P0'n -P0n )/(2P0n -P0'n )
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
2
4
6
8 10 12 14 R0p x 105 (moleL-1s-1)
16
18
20
22
Figure 2.9. Plot of ( Pn0 ' − Pn0 ) /( 2 Pn0 − Pn0 ' ) vs Rp0 for bulk polymerization of vinyl chloride at 55 oC ( krikp/kt1=(1.7±0.2) x 10-6 L•mole-1•s-1, Pn0 ' = 885 ).
- 77 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.7. Summary of kinetic parameters for bulk polymerization of vinyl chloride initiated by 2,2’-azobisisobutyronitrile (AIBN) at 55 oC. Varaible Value Units Reference f for AIBN
0.78±0.05
kd for AIBN
a)
CM
this work
5.0x10-6
s-1
11.3±0.1x10-4
[34] this work
2020
L•mole-1•s-1
[56]
ktr
2.3
L•mole-1•s-1
this work
krikp/kt1
2.5±0.2x10-6 or1.7±0.2x10-6
L•mole-1•s-1 L•mole-1•s-1
this work
kri
1~10ktr
L•mole-1•s-1
[28]
kt1
(1.8~18)x109 or (2.7~27)x109
L•mole-1•s-1 L•mole-1•s-1
this work
kt
2.3x109
L•mole-1•s-1
[53]
kp
b)
a): calculated using equation kd=2.88x1015exp(-130400/RT) s-1. b): calculated using equation kp=5x107exp(-27600/RT) L•mole-1•s-1.
The bulk polymerization of vinyl chloride initiated by AIBN has been studied extensively in the literature. The purpose of this duplication was to set a reference to study the influence of selected additives on the polymerization of vinyl chloride under similar conditions. The kinetic parameters determined in this section will be used to evaluate other parameters for vinyl chloride polymerization in the presence of additives. In the next section we will first derive the kinetic equations for bulk polymerization of vinyl chloride in the presence of additives.
- 78 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.6 Kinetics and Mechanism of Vinyl Chloride Polymerization in the presence of Additives It is well-known that PVC can form compatible blends with poly(methyl methacrylate) (PMMA) and poly(carbonate) (PC). The miscibility of PVC with poly(methyl methacrylate) (PMMA) or polycarbonate (PC) was proposed due to hydrogen bonding between the carbonyl groups (>C=O) of PMMA or PC and hydrogens from chloromethylene (~CHCl~) groups of PVC15, 16. Our data imply that the hydrogen on the propagating chloromethylene radical is more acidic than that of the chloromethylene group, and is acidic enough to compete successfully with the internal hydrogens, in hydrogen bonding with compounds that have weak proton accepting groups. These compounds could be esters, carbonates, phosphine oxides etc. This interaction can be schemed as following:
~CH
C
H
+ O
2
Rn
Cl
OR'
+
C
~CH
R
C
H
..
Cl
2
K
A
(R nA)
Here Rn• is the propagating radical ~ CH 2 C HC l , A, the additive in the polymerization system, Rn• A is the hydrogen bonded propagating radical, and K the equilibrium constant.
Since [ A] >> [ Rn• ] + [ Rn• A] , we have
- 79 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ [ Rn• A] ≈ K [ Rn• ][ A] …………………………………………………(2.35)
and [ Rn• A] + [ Rn• ] = [ RT• ] ……………………………………………(2.36) Here [ RT• ] is the total concentration of propagating radicals in the polymerization system.
From equations (2.35) and (2.36), we have: [ Rn• ] =
1 [ RT• ] ……………………………………..………..(2.37) 1 + K [ A]
and [ Rn• A] =
K [ A] [ RT• ] ……………………………………..............(2.38) 1 + K [ A]
Vinyl chloride (M) polymerization in the presence of additive (A) can then be described as following:
(1) Initiation reactions: this is the same as in absence of additives. kd I ⎯⎯→ 2I • ki I • + M ⎯⎯→ R•
The rate of initiation ( Ri ) can be described as Ri = 2 fk d [ I ]
(2) Propagation reactions: in the presence of additives, the propagation reactions become p1 Rn• + M ⎯⎯→ Rn•+1
k
p2 Rn• A + M ⎯⎯→ Rn•+1 A
k
- 80 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
The initial rate of propagation R p0 can then be described as:
R p0 = k p1[ Rn• ][M ] + k p 2 [ Rn• A][M ] = k p [ RT• ][M ] …………...(2.39) Here kp1, kp2 are propagation rate constants. We assume k p 2
≈ k p1 = k p , since any
difference is absorbed by the equilibrium constant K. Equation (2.39) describes the overall initial rate of propagation for the bulk polymerization of vinyl chloride in the presence of additives.
(3) Chain-transfer reactions: the chain-transfer reactions include chain transfer to monomer M and to additive A by both the free radicals and the complexed radicals: ktr1 Rn• + M ⎯⎯→ Pn + M • tr 2 Rn• A + M ⎯k⎯→ Pn + A + M •
tr 3 Rn• + A ⎯k⎯→ Pn + A•
tr 4 Rn• A + A ⎯k⎯→ Pn + A + A•
Here M • is the monomer radical. It could be chlorine radical, or chlorine radical associated with monomers. A• is the additive radical. We assume ktr 4 ≈ ktr 3 , and ignore other chain transfer reactions such as chain transfer to the initiator and to polymer. It will be shown that when chain transfer to additive becomes unimportant, essentially all the radicals exist as
Rn• A . Thus the rate of chain-transfer reactions, Rtr, is given by
- 81 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Rtr = k tr 1[ Rn• ][ M ] + k tr 2 [ Rn• A][ M ] + k tr 3{[ Rn• ] + [ Rn• A]}[ A]
[ RT• ][M ] K[ A][RT• ][M ] = ktr1 + ktr 2 + ktr 3 [ RT• ][ A] 1 + K [ A] 1 + K[ A]
= ktr1Φ[ RT• ][ M ] + ktr 3 [ RT• ][ A] ………………………………….(2.40) ktr 2 K [ A] ktr1 . 1 + K [ A]
1+
Here Φ =
(4) Reinitiation reactions: radicals M • and A• may reinitiate vinyl chloride monomer (M): kri 1 M • + M ⎯⎯→ R• ri 2 A• + M ⎯k⎯→ R•
(5) Termination reactions: same as shown in sections 2.3.4 and 2.3.5, the propagating radicals mainly terminate with small radicals such as M • or A• and ignore other termination reactions: kt 1 Rn• + M • ⎯⎯→ Pn kt 1 Rn• A + M • ⎯⎯→ Pn + A
kt 2 Rn• + A• ⎯⎯→ Pn kt 2 Rn• A + A• ⎯⎯→ Pn + A
The rate of termination, Rt, is given by
- 82 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Rt = kt1[ Rn• ][ M • ] + kt1[ Rn• A][ M • ] + kt 2 [ Rn• ][ A• ] + kt 2 [ Rn• A][ A• ] = kt1[ RT• ][ M • ] + kt 2 [ RT• ][ A• ] …………….…………………….(2.41)
Under steady-state assumption, the concentrations of RT• , M • and A• change very slowly, so we have the following equations:
⎛ dRT• ⎞ ⎜⎜ ⎟⎟ = 2 fk d [ I ] + k ri1[ M • ][ M ] + k ri 2 [ A• ][ M ] ⎝ dt ⎠ − ktr1
[ RT• ][ M ] K [ A][ RT• ][ M ] − ktr 2 − ktr 3 [ RT• ][ A] 1 + K [ A] 1 + K [ A]
− kt1[ RT• ][ M • ] − kt 2 [ RT• ][ A• ] ≈ 0 ……………………...(2.42)
⎛ dM • ⎞ [ R • ][ M ] K [ A][ RT• ][M ] ⎜⎜ ⎟⎟ = ktr1 T + ktr 2 1 + K [ A] 1 + K [ A] ⎝ dt ⎠ − k ri1[ M • ][ M ] − kt1[ M • ][ RT• ] ≈ 0 …………………........(2.43)
⎛ dA• ⎞ ⎜⎜ ⎟⎟ = ktr 3[ RT• ][ A] − kri 2 [ A• ][M ] − kt 2 [ A• ][ RT• ] ≈ 0 ……….....(2.44) ⎝ dt ⎠ Adding Equations (2.42), (2.43) and (2.44), we get
fk d [ I ] = kt1[ RT• ][ M • ] + kt 2 [ RT• ][ A• ] ……………………………….(2.45)
To solve for [ RT• ] , we need to know [ M • ] and [ A• ] . Here we will solve equation (2.45) with two different, extreme approximations: (I) A• is a stable radical and only terminates, and (II) A• is an active radical and reinitiate at least as rapidly as M • . - 83 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
(I) Stable Additive Radical Approximation If the additive radical A• is very stable, it is mainly consumed by termination with propagating radicals and its reinitiation reaction can be ignored. Thus, equation (2.44) becomes:
⎛ dA• ⎞ ⎜⎜ ⎟⎟ = ktr 3 [ RT• ][ A] − kt 2 [ A• ][ RT• ] ≈ 0 ⎝ dt ⎠ • • • So ktr 3 [ RT ][ A] ≈ k t 2 [ A ][ RT ] ………………………….……….(2.46)
Substituting equation (2.46) into equation (2.45), we have fk d [ I ] = kt1[ RT• ][ M • ] + ktr 3 [ RT• ][ A] …………………......................(2.47)
Equation (2.43) can be rearranged to
[M • ] =
ktr1[ M ]Φ [ RT• ] ……………………………...(2.48) • k ri1[ M ] + kt1[ RT ]
Substituting equation (2.48) into equation (2.47), we get: k t1 ⎧ [ A] ⎫ • 2 ⎨ktr1Φ + ktr 3 ⎬[ RT ] k ri1 ⎩ [M ] ⎭
⎧ k [I ] ⎫ • + ⎨ktr 3 [ A] − fk d t1 ⎬[ RT ] − fk d [ I ] = 0 ………….…..(2.49) k ri1 [ M ] ⎭ ⎩
- 84 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Equation (2.49) is a quadratic equation, which has two roots. The meaningful root can be approximated as: 1
⎛ fk k ⎞ 2 1 fkd [ I ] k C [ RT• ] = ⎜⎜ d ri1 ⎟⎟ [ I ] 2 + − ri1 A [ A] .........................(2.50) 2ktr1Γ [ M ] 2kt1CM 1Γ ⎝ ktr1kt1Γ ⎠ Here Γ = Φ +
k k C A [ A] , CM 1 = tr1 , and C A = tr 3 . kp kp C M 1[ M ]
Substituting equation (2.50) into equation (2.39), we get: 1 2
1 C k k ⎛ fkd kri1k p ⎞ fkd ⎟⎟ [ I ] 2 [ M ] + R = ⎜⎜ [ I ] − A ri1 p [ A][ M ] ……….(2.51) 2CM 1Γ 2CM 1kt1Γ ⎝ CM 1kt1Γ ⎠ 0 p
We can see that if [A]=0, Γ will be 1 and equation (2.51) becomes equation (2.24).
Substituting equations (2.39), (2.40) and (2.41) into equation (2.26), we have ktr1 ktr 3 [ A] kt1 [ M • ] 1 ………………………………...(2.52) = Φ + + Pn0 k p k p [M ] k p [M ]
From equation (2.48), we get [ M • ] ≈
ktr1 Φ[ RT• ] . If kri [M ] >> kt1[ Rn• ], the k ri1
contribution of termination reactions can be ignored. Thus equation (2.52) can be rewritten as - 85 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ ktr1 ktr 3 [ A] kt1ktr1 [ RT• ] 1 = Φ + + Φ Pn0 k p k p [ M ] k ri1k p [ M ]
k t1 [ RT• ] = C M 1Γ + C M 1Φ …………………………………...(2.53) k ri1 [ M ]
Substituting equation (2.50) into equation (2.53), we get: 1
1
⎛ C M 1 fk d k t1Φ 2 ⎞ 2 [ I ] 2 1 ⎟ = C M 1Γ + ⎜ ⎜ k k Γ ⎟ [M ] Pn0 ri1 p ⎝ ⎠ +
fk d kt1Φ [ I ] C AΦ [ A] − ………………………………(2.54) 2k ri1k p Γ [ M ]2 2Γ [ M ]
Again, we can see that if [A]=0, Γ and Φ will be 1, and equation (2.54) becomes equation (2.30).
(II) Active Additive Radical Approximation
If the additive radical
A• is
active enough, it mainly reinitiates. Then the
termination reaction can be ignored. So equation (2.45) becomes: fk d [ I ] = kt1[ RT• ][ M • ] ……………………………………………...(2.55)
Substituting equation (2.48) into equation (2.55), one gets:
Φ[ RT• ]2 −
fk d [ I ] • fk k [ RT ] − d ri1 [ I ] = 0 ………………………...(2.56) ktr1 [ M ] ktr1 kt1
- 86 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ This is a quadratic equation and the one meaningful root can be approximated as follows: 1
⎛ fk k ⎞ 2 1 fk d [ I ] [ RT• ] = ⎜⎜ d ri1 ⎟⎟ [ I ] 2 + …………………………………….(2.57) 2ktr1Φ [ M ] ⎝ kt1ktr1Φ ⎠
Similarly, substituting equation (2.57) into equation (2.39) and equation (2.53), we get: 1
⎛ fk k k ⎞ 2 1 fk d R p0 = ⎜⎜ d ri1 p ⎟⎟ [ I ] 2 [ M ] + [ I ] …………..………………...(2.58) 2CM 1Φ ⎝ CM 1kt1Φ ⎠
1 [ A] = C M 1Φ + C A 0 Pn [M ] 1
1
⎛ C fk k Φ ⎞ 2 [ I ] 2 fk k [ I ] ……………….(2.59) + ⎜ M 1 d t1 ⎟ + d t1 ⎜ k k ⎟ [ M ] 2 k k [ M ]2 1 ri p p ri 1 ⎝ ⎠
Equations (2.51), (2.54), (2.58) and (2.59) can describe the kinetics of bulk polymerization of vinyl chloride in the presence of additives. In the following three sections, three groups of the additives were studied to find out which were the best additives to give higher molecular weight and higher thermal stability. The first group of additives has medium basicity: pyridine, its derivatives, pyrazine, benzothiazol and imidazolidinone. The second group of additives contains high dipole moment phosphine oxide compounds. The third group of additives has weakly basic moieties: lactones, carbonates and aromatic carboxylates, ether compounds like dimethoxybenzene etc. - 87 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
§2.3.7 Bulk Polymerization of Vinyl Chloride at 55 oC in the presence of Pyridine, its Derivatives, Pyrazine, Benzothiazol, or Imidazolidinone
S
CH3
CH3
CH2-N N
C=O CH2-N
2-methylbenzothiazole (MBTZ)
CH3 1,3-dimethyl-2-imidazolidinone (DMI) N N CH3 2-methylpyrazine(MPZ) pKa=1.45
Cl
N
N pyridine(PYR) pKa=5.25 CH3
Cl
N CH3 CH3 2, 4, 6-trimethylpyridine (TMPY)
2,6-dichloropyridine (DCPY)
Scheme 2.3 Chemical structures for 1,3-dimethyl-2-imidazolidinone (DMI), 2-methylbenzothiazole (MBTZ), 2-methylpyrazine (MPZ), pyridine (PYR), 2,6-dichloro-pyridine (DCPY), and 2,4,6-trimethylpyridine (TMPY).
The bulk polymerization of vinyl chloride was initiated by AIBN. The concentration of the initiator was kept constant at about 6.4x10-3 mole/L for all the reactions. The additives were 1,3-dimethyl-2-imidazolidinone (DMI), 2-methylbenzothiazole (MBTZ), 2-methylpyrazine (MPZ), pyridine (PYR), 2,6-dichloro-pyridine (DCPY), and 2,4,6-trimethylpyridine (TMPY). The structures of the additives are shown in Scheme 2.3. The concentration of the individual additives varied, but was set at 1.0 mole% for most reactions. The reactions were stopped at low conversion as shown in - 88 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.8. The initial polymerization rate, R p0 , calculated according to equation (2.12) is also listed in Table 2.8. The molecular weights and the molecular weight distributions of the PVC samples, determined by GPC, are listed in Table 2.9.
It can be seen from Table 2.8 that for most polymerizations in the presence of these additives, the extrapolated values of R p0 were lower than that without additives (control polymerizations). There was an induction time for most reactions. The average induction time for vinyl chloride polymerization without additive was about 3-5 minutes, Table 2.4. The induction time was longer than 10 minutes for most reactions with these additives, and it sometimes reached 50 minutes. The molecular weights of the resulting polymers decreased compared to the control, which implies that those additives are active chain transfer agents. The chain transfer constants to DMI, MBTZ, MPZ, PYR and TMPY were calculated as 689±95x10-4, 1531±?x10-4, 1491±?x10-4, 246±?x10-4, and 272±4x10-4, respectively. We can see they are much higher than the chain transfer constant to monomer (11.3±0.1x10-4). At higher concentrations, these additives can decrease the resulting polymer molecular weight drastically. Figure 2.4 showed that PVCs prepared in the presence of DMI, MBTZ, MPZ, PYR and TMPY degraded rapidly in dichlorobenzene solution and darkened within 14 hours at 90 oC, but other samples remained colorless under the same condition. It is probably because they are all too basic and could attack Cl groups of the polymer to form internal double bonds during or after polymerization, which cause PVC to decompose faster at elevated temperatures.
- 89 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.8. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of pyridine, its derivatives, pyrazine, benzothiazol, and imidazolidinone. AIBN Addit R0px105 PVC Conv VC Indb) Time Runa) (mg) (min) (min) (%) (mole•L-1•s-1) (g) (g) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 DCPY00025a
31.7
0.015
26.42
2
68
1.20
4.54
12.0
DCPY0005b
32.0
0.030
25.67
3
70
1.32
5.14
12.9
DCPY001b
32.7
0.060
25.83
14
75
1.38
5.34
12.4
DCPY005a
16.0
0.153
15.04
8
108
1.04
6.91
10.4
DCPY005b
16.0
0.148
15.12
10
65
0.36
2.38
7.3
DCPY005c
16.0
0.154
14.63
8
90
0.72
4.92
9.7
DCPY005d
32.0
0.293
24.25
10
95
0.98
4.04
7.9
DCPY0075a
32.4
0.446
25.20
4
108
1.22
4.84
8.0
DCPY01h
32.7
0.592
26.71
4
110
1.44
5.39
8.5
DMI01a
16.0
0.235
14.30
50
97
0.93
6.51
11.1
DMI05d
16.0
1.177
13.85
n/a
197
2.02
14.57
9.2
MBTZ01a
26.0
0.238
27.60
45
180
1.50
5.43
5.3
MPZ01a
26.0
0.151
27.04
25
240
1.23
4.55
3.5
PRY01a
16.0
0.158
12.85
n/a
180
0.42
3.27
3.5
TMPY01a
16.0
0.246
12.72
42
120
1.61
12.64
13.8
TMPY05c
16.0
1.210
11.73
10
162
1.12
9.55
8.7
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 90 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.9. Molecular weights and molecular weight distributions for PVCs prepared at 55oC in the presence of pyridine, its derivatives, pyrazine, benzothiazol, and imidazolidinone. Sample Mw Mn Mz Mw/Mn Mz/Mw Designationa) Control 78,500 43,100 125,000 1.82 1.59 DCPY00025a
80,500
44,000
131,700
1.83
1.64
DCPY0005b
81,400
43,800
131,800
1.86
1.62
DCPY001b
75,000
42,400
115,300
1.77
1.54
DCPY005a
83,900
45,700
126,700
1.83
1.51
DCPY005b
76,800
43,400
114,500
1.77
1.49
DCPY005c
71,300
43,400
104,800
1.64
1.46
DCPY005d
80,300
45,000
120,200
1.78
1.49
DCPY0075a
87,800
48,100
134,900
1.83
1.54
DCPY01h
84,900
46,900
129,000
1.83
1.54
DMI01a
50,500
24,700
78,200
2.04
1.55
DMI05d
22,600
13,000
27,100
1.73
1.20
MBTZ01b
63,700
31,200
107,900
2.04
1.69
MPZ01a
63,700
31,200
100,900
2.04
1.58
PRY01a
75,300
37,000
132,700
2..04
1.76
TMPY01a
72,500
35,800
139,400
2.02
1.92
TMPY05c
41,000
21,500
55,500
1.91
1.35
a): Sample named as in Table 2.8.
- 91 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ From Tables 2.8 and 2.9, we can see that as 2,6-dichloropyridine (DCPY) concentration increased, R p0 decreased, but the molecular weights of the polymers remained constant within the experimental error. As will be shown in Chapter 4, the resulting PVC rate of HCl loss (dehydrochlorination) increased rapidly as DCPY concentration increased. Since DCPY, like PYR and TMPY, is a basic compound, it may attack a Cl group of the polymer to form double bonds or conjugated double bonds in the polymer backbone. These double bonds promote allylic dehydrochlorination and HCl is lost more rapidly than found in the control. Both PYR and TMPY, and other basic compounds such as DMI, MBTZ, and MPZ, reduced the vinyl chloride polymerization rate and also decreased the molecular weights of the resulting polymers.
§2.3.8 Bulk Polymerization of Vinyl Chloride at 55 oC in the presence of Phosphine Oxides The bulk polymerization of vinyl chloride in the presence of phosphine oxide compounds was initiated by AIBN. The initiator concentration was kept constant at around 6.4x10-3 mole/L for all reactions. The four compounds used were trimethylphosphine oxide (TMPO), triethylphosphine oxide (TEPO), tributylphosphine oxide (TBPO), and triphenylphosphine oxide (TPPO). The structures of those additives are shown in Scheme 2.4. The concentration of the individual additive used was usually 1.0 mole%, as shown in Table 2.10. The reactions were stopped at low conversion. The initial polymerization rate R p0 was calculated according to equation (2.12). The molecular weight and the molecular weight distribution of the PVC samples, determined by GPC, are listed in Table 2.11. It can be seen that the vinyl chloride polymerizations in the - 92 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ presence of phosphine oxides have an average induction time of 15 minutes. The longer induction time is probably due to impurities in the additives. The initial polymerization rate R p0 is higher than that of the control for TBPO and TEPO at 1mole% concentration or less. R p0 is lower than that of the control for TMPO and TPPO additives.
CH3
CH2CH3
CH3 P=O
CH3CH2
CH3
P=O CH2CH3
triethylphosphine oxide (TEPO)
trimethylphosphine oxide (TMPO) n-Bu n-Bu
P=O
P=O n-Bu
tributylphosphine oxide triphenylphosphine oxide (TBPO) (TPPO) Scheme 2.4. Chemical structures for trimethylphosphine oxide (TMPO), triethylphosphine oxide (TEPO), tributylphosphine oxide (TBPO), and triphenylphosphine oxide (TPPO).
- 93 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.10. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of phosphine oxides. PVC Conv R0px105 AIBN Addit VC Indb) Time Runa) (min) (min) (g) (%) (mole•L-1•s-1) (mg) (g) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 TBPO01b
16.0
0.439
16.26
15
134
1.85
11.38
11.7
TBPO04a
16.0
2.163
16.99
59
122
1.40
8.22
10.4
TBPO04b
16.0
2.183
13.64
12
140
1.05
7.68
8.7
TEPO005a
16.0
0.202
14.03
15
105
1.50
10.69
14.3
TEPO01b
16.0
0.321
13.72
10
160
2.16
15.74
11.7
TMPO01a
16.0
0.168
13.00
15
130
0.77
5.92
7.8
TMPO01c
16.0
0.180
20.92
15
170
0.66
3.15
3.6
TPPO01a
10.0
0.556
14.29
n/a
420
0.96
6.72
2.8
TPPO01b
33.0
1.113
25.57
25
290
0.89
3.48
2.4
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 94 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.11. Molecular weight and molecular weight distributions for PVCs prepared at 55oC in the presence of phosphine oxides. Sample Mw Mn Mz Mw/Mn Mz/Mw Designationa) Control 78,500 43,100 125,000 1.82 1.59 TBPO01b
74,000
44,400
105,800
1.66
1.43
TBPO04a
84,900
39,700
155,100
2.13
1.82
TBPO04b
94,400
41,700
204,900
2.26
2.17
TEPO005a
79,400
45,600
110,900
1.74
1.40
TEPO01b
80,100
38,700
150,400
2.06
1.87
TMPO01a
70,400
36,400
100,800
1.93
1.43
TMPO01c
67,100
39,200
93,800
1.71
1.40
TPPO01a
66,000
40,800
92,000
1.61
1.39
TPPO01b
85,500
38,400
145,100
2.23
1.70
a): Sample named as in Table 2.10.
As shown in Table 2.11, the average number molecular weight at 1 mole% TBPO concentration was about 3% higher than that of the control. The polymer molecular weight decreased at higher TBPO concentration. When the TEPO concentration was 0.5 mole%, the resulting polymer molecular weight was about 6% higher than that of the control, but the molecular weight decreased at higher concentrations. For PVC polymers prepared in the presence of 1 mole% TEPO, TMPO or TPPO, the molecular weights decreased about 10%, compared to that of the control. The molecular weight distributions of the resulting PVCs were mostly about 1.61-1.74, but some had a broader distribution, over 2.00, which indicates much lower molecular weight polymers were in these samples. - 95 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The chain transfer constants to TBPO, TEPO, TMPO, and TPPO were calculated as 24±16x10-4, 126±188x10-4, 296±43x10-4, and 147±75x10-4, respectively. Due to the lack of experimental data, these calculated results have a relatively large error. But, it’s still obvious that the chain transfer constants for TBPO and TEPO are much smaller than those for TMPO and TPPO. It should be pointed out that both TBPO and TEPO increased the initial polymerization rate and the resulting polymer molecular weights at low concentration.
§2.3.9 Bulk Polymerization of Vinyl Chloride at 55
o
C in the presence of
Dimethoxybenzene, Carbonates, Carboxylates or Lactones
Section 2.3.7 showed that polymerization of vinyl chloride in the presence of pyridine, its derivatives, or other basic nitrogen compounds such as pyrazine, benzothiazol, or imidazolidinone, had a long induction period and only low molecular weight PVC was obtained. Section 2.3.8 showed that polymerization of vinyl chloride in the presence of weakly basic phosphine oxides also had long induction periods and lower molecular weight in most cases. However, the molecular weight of 1 mole% TBPO polymer was 3% higher than that of the control and 0.5 mole% TEPO polymer was 6% higher than that of the control. More moderate additives, with much weaker basicity are considered in this section.
The additives discussed in this section are 1,4-dimethoxybenzene (DMB), ethylene carbonate (EC), 2-coumaranone (CMN), γ-butyrolactone (GBL), dimethyl terephthalate (DMT), and trimethyl-1,3,5-benzene tricarboxylate (TMB). The structures - 96 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ of those additives are shown in Scheme 2.5. The initiator was AIBN at a concentration of about 6.4x10-3 mole/L for all reactions. The concentration of the individual additive was varied and the reactions were stopped at low conversion. The initial polymerization rates, R p0 , calculated according to equation (2.12), are listed in Tables 2.12, 14, 16, 18, 20 and
22. The molecular weights and the molecular weight distributions of the PVC samples, determined by GPC, are listed in Tables 2.13, 15, 17, 19, 21 and 23.
CH2-O
CH3 O
O
C=O CH3
CH2-O
1,4-dimethoxybenzene (DMB)
ethylene carbonate (EC)
C O
CH2 O
O γ-butyrolactone (GBL)
C=O
2-coumaranone (CMN)
O C
OCH3
OCH3
O C
C
O
O
CH3O dimethyl terephthalate (DMT)
O
C
C
OCH3
OCH3
trimethyl-1,3,5-benzene tricarboxylate (TMB)
Scheme 2.5. Chemical structures for 1,4-dimethoxybenzene (DMB), ethylene carbonate (EC), 2-coumaranone (CMN), γ-butyrolactone (GBL), dimethyl terephthalate (DMT) and trimethyl-1,3,5-benzenetricarboxylate (TMB).
- 97 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.9.1 Polymerization in the presence of 1,4-Dimethoxybenzene The bulk polymerizations of vinyl chloride in the presence of 1,4dimethoxybenzene (DMB) had 1-9 minute induction times, with an average induction time of 3-4 minutes, Table 2.12. The presence of DMB, at least up to 0.1 mole% level, does not retard the vinyl chloride polymerization. The initial polymerization rate, R p0 , remained at the same level as that of the control within the experimental error, Figure 2.10. Conversion was around 4-7%. The molecular weights and the molecular weight distributions of the resulting PVCs are listed in Table 2.13. They are about the same as that of the control PVC. The molecular weight distribution was around 1.77-1.84. Figure 2.10 shows that plots of both 1 / Pn0 and R p0 v.s. [DMB]/[VC] are flat. Any effect of DMB on the polymerization rate or on the resulting polymer molecular weights is within the
2.0
32
1.8
28
1.6
24
1.4
20
1.2
16
1.0
12
0.8
8
0.6
4
0.4
0 0.0
0.1
0.2
0.3
0.4 0.5 0.6 0.7 1000x[DMB]/[VC]
0.8
0.9
1.0
R0p x 10 5 (moleL -1s-1)
1000/P0n
experimental error at these concentrations.
1.1
Figure 2.10. DMB effect on polymerization of vinyl chloride (VC) at 55 oC.
- 98 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.12. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of 1,4-dimethoxybenzene (DMB). Time PVC Conv AIBN DMB VC Indb) R0px105 Runa) (%) (mole•L-1•s-1) (mg) (mg) (g) (min) (g) (min) Control 32.6 0.0 25.44 3 76 1.43 5.70 10.6 DMB00005a
32.4
2.5
22.96
2
87
1.70
7.40
13.5
DMB00005b
31.8
2.6
27.10
2
72
1.23
4.54
11.4
DMB00005c
32.5
3.1
25.18
1
72
1.18
4.69
11.6
DMB0001a
32.0
5.3
24.22
2
68
1.03
4.25
11.4
DMB0001b
32.1
5.5
23.69
3
67
1.13
4.76
12.7
DMB0002a
31.7
11.1
23.99
2
96
1.24
5.17
9.5
DMB0002b
31.7
11.9
22.43
4
85
1.38
6.15
12.1
DMB0002c
33.2
12.4
24.33
4
71
1.16
4.77
12.0
DMB0005a
32.1
27.3
25.22
4
79
1.07
4.24
9.8
DMB0005b
32.4
27.3
26.84
7
88
1.07
3.99
8.4
DMB0008a
33.0
41.4
26.07
4
83
1.64
6.29
12.6
DMB001a
32.8
56.4
25.30
9
90
1.38
5.45
10.5
DMB001b
32.2
55.6
26.07
5
84
1.47
5.63
11.5
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 99 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.13. Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of DMB. Sample Mw Mn Mz/Mw Mz Mw/Mn Designationa) Control 78,500 43,100 125,000 1.82 1.59 DMB00005a
78,600
43,200
123,400
1.82
1.57
DMB00005b
85,500
47,900
131,600
1.79
1.54
DMB00005c
85,400
46,500
136,200
1.84
1.59
DMB0001a
79,700
44,200
120,100
1.80
1.51
DMB0001b
84,500
45,800
128,400
1.84
1.52
DMB0002a
80,000
44,100
123,000
1.81
1.54
DMB0002b
82,100
45,600
123,800
1.80
1.51
DMB0002c
82,600
46,800
124,100
1.77
1.50
DMB0005a
83,200
46,400
126,200
1.79
1.52
DMB0005b
77,100
43,000
118,400
1.79
1.53
DMB0008a
75,900
42,400
114,900
1.79
1.51
DMB001a
73,400
41,500
112,600
1.77
1.53
DMB001b
80,100
45,200
120,900
1.77
1.51
a): Sample named as in Table 2.12.
§2.3.9.2 Polymerization in the presence of Ethylene Carbonate Bulk polymerization of vinyl chloride in the presence of ethylene carbonate (EC) had an 8-15 minute induction time, with initial polymerization rates, R p0 , about 5-14x10-5 mole•L-1•s-1, Table 2.14. The molecular weight of PVC prepared in the presence of 1
- 100 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ mole% EC was about 10% higher than that of control PVC. When the concentration of EC was raised to 4 mole%, the molecular weight of the resulting PVC dropped, but was still about 6% higher than that of the control polymer, Table 2.15. The chain transfer constant to EC was calculated as 21±2x10-4. This value is much smaller compared to those for other compounds.
Table 2.14. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of ethylene carbonate. Time PVC Conv R0p x105 AIBN Addit. VC Indb) Runa) (min) (min) (g) (%) (mole•L-1•s-1) (mg) (g) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 EC01c
26.0
0.352 36.80
15
210
2.01
5.46
4.6
EC04b
16.0
0.886 14.70
15
120
0.85
5.76
8.2
EC04d
16.0
0.881 16.21
8
125
2.20
13.55
13.8
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time.
Table 2.15. Molecular weights and molecular weight distributions for PVCs prepared at 55oC in the presence of ethylene carbonate. Sample Mw Mn Mz Mz/Mw Mw/Mn Designationa) Control 78,500 43,100 125,000 1.82 1.59 EC01c
85,100
48,100
127,700
1.77
1.50
EC04b
82,500
45,600
122,100
1.81
1.48
EC04d
80,700
45,600
121,000
1.77
1.50
a): Sample named as in Table 2.14.
- 101 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Ethylene carbonate is a polar compound with high dipole moment and could Hbond with the propagating group to form a reversible complex. Such a complex could reduce chain transfer to the monomer and thus increase the molecular weight. The static thermal stability test, Figure 2.4, showed that the polymer solution remained colorless after heating at 90 oC for 14 hours. More analyses and characterization of the resulting PVCs will be discussed in the following chapters.
§2.3.9.3 Polymerization in the presence of 2-Coumaranone The bulk polymerization of vinyl chloride in the presence of 2-coumaranone (CMN) had a 1-7 minute induction time, Table 2.16, with an average induction time of 34 minutes, the same level as in the control. The presence of CMN reduced the polymerization rate even at very low concentration. As can be seen in Table 2.16 and Figure 2.11, R p0 dropped as CMN concentration increased. The conversion was about 36%. The molecular weights and the molecular weight distributions of the resulting PVCs, determined by GPC, are listed in Table 2.17. As can be seen, the molecular weights of the resulting PVCs, obtained by polymerization at very low concentrations of CMN were higher than that of the control. The molecular weight increased with the addition of CMN and reached the highest value (~13% higher than that of control) at a CMN concentration of about 0.01 mole% (relative to VC). The molecular weight then decreased with increasing CMN. CMN acts as a strong chain transfer agent and tends to decrease the PVC molecular weight. The reason is that hydrogen atoms of the CH2 group in CMN
- 102 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ next to the carboxyl group and benzene ring are very vulnerable to attack by propagating radicals to form the following radical60: H H
H
H
O
O
O
O
O
O
This radical is very stable since it has many conjugated structures. We assume these radicals disappear mainly by terminating with propagating radicals. Their ability to add monomer to generate new propagating radicals is very low, so reinitiation can be ignored. The kinetic behavior can be evaluated using equation (2.51) and equation (2.54), developed in section 2.3.5:
1
C k k ⎛ fk k k ⎞ 2 1 fkd R p0 = ⎜⎜ d ri1 p ⎟⎟ [ I ] 2 [ M ] + [ I ] − A ri1 p [ A][ M ] …..…...(2.51) 2CM 1Γ 2CM 1kt1Γ ⎝ CM 1kt1Γ ⎠
And 1
1
⎛ CM 1 fkd kt1Φ 2 ⎞ 2 [ I ] 2 1 ⎜ ⎟ = Γ + C M1 0 ⎜ ⎟ Pn ⎝ kri1k p Γ ⎠ [ M ] +
Here Γ = Φ +
C A [ A] ,Φ= C M 1[ M ]
fk d kt1Φ [ I ] C AΦ [ A] − …………………(2.54) 2k ri1k p Γ [ M ]2 2Γ [ M ]
ktr 2 K [ A] ktr1 ktr1 , and C A , CM 1 = kp 1 + K [ A]
1+
- 103 -
=
ktr 3 . kp
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 35
1.6
30
1.4
25
1000/P0n
R0p x 10 5 ( moleL -1s-1 )
1.8
1.2
20
1.0
15
0.8
10
0.6
5
0.4
0 0.0
0.1
0.2
0.3
0.4 0.5 0.6 0.7 1000x[CMN]/[VC]
Figure 2.11. Plots of 1 / Pn0 and
R 0p
0.8
0.9
1.0
1.1
vs [CMN]/[VC] for bulk polymerization of vinyl
chloride in the presence of CMN at 55 oC. ○ and ● are the experimental data and the solid curves are the calculated results of equation (2.51) and equation (2.54) by leastsquares regression (K=60000, ktr2/ktr1=0.89±0.01, CA=0.53±0.04).
Using least-squares regression, we fit the experimental data of Pn0 and R p0 to equations (2.51) and (2.54). At an equilibrium constant, K=60,000, the regression best fit gave parameter values for k tr 2 k tr1 =0.89±0.01, and for CA=0.53±0.04. The calculated curves for 1 / Pn0 and
R p0
vs [CMN]/[VC] are shown in Figure 2.11. We can see the curves
match very well with the experimental results. The value of k tr 2 k tr1 for CMN is less than 1, which means that the chain transfer rate constant to monomer decreased for the propagating radicals complexed with CMN and thus the propagating chain could add more monomers before chain transferring to monomer. This is the main reason why low concentrations of CMN increased the polymer molecular weights. Since k tr 2 k tr1 is about 0.9, the maximum increase of the molecular weight is about 10% when [CMN]/[VC] ≈ 10-4, Figure 2.11. The chain transfer constant for CMN is very high. It - 104 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ was determined as CA=0.53±0.04. At higher CMN concentrations, chain transfer to CMN became the dormant chain terminating process and 1 / Pn0 increased almost linearly with the increase of [CMN], Figure 2.11. Since CMN generates a stable radical that mainly terminates, Figure 2.11 shows that
R p0
decreases with the increase of CMN concentration.
Table 2.16. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of 2-coumaranone (CMN). Indb) Time PVC Conv AIBN CMN VC R0px105 Runa) (mg) (%) (min) (min) (mg) (g) (g) (mole•L-1•s-1) Control 32.6 0.0 25.44 3 76 1.43 5.70 10.6 CMN00005a
32.2
3.0
28.72
1
80
1.25
4.35
9.9
CMN00005b
34.0
2.7
25.89
3
84
1.45
5.60
11.5
CMN0001a
32.0
5.4
19.12
7
78
0.96
5.02
11.3
CMN0001b
32.0
5.4
26.55
2
86
1.11
4.18
8.9
CMN00025a
32.0
14.2
21.63
7
91
0.93
4.30
8.6
CMN00025b
32.0
14.0
27.91
2
90
1.21
4.33
8.8
CMN0005b
32.0
28.0
31.15
5
97
1.00
3.21
6.4
CMN0005c
31.8
27.1
22.51
7
85
0.99
4.40
9.4
CMN0005d
32.0
26.9
26.89
3
94
0.89
3.31
6.8
CMN001a
32.1
52.4
29.07
3
97
0.91
3.13
6.2
CMN001b
32.2
52.7
24.72
4
105
0.98
3.96
7.0
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 105 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.17. Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of CMN. Sample Mw Mn Mz Mz/Mw Mw/Mn Designationa) Control 78,500 43,100 125,000 1.82 1.59 CMN00005a
81,100
47,800
117,700
1.70
1.45
CMN00005b
81,700
46,900
121,000
1.74
1.48
CMN0001a
84,500
48,500
125,600
1.74
1.49
CMN0001b
88,400
49,100
136,600
1.80
1.55
CMN00025a
75,900
43,500
113,500
1.75
1.50
CMN00025b
85,300
48,500
127,300
1.76
1.49
CMN0005b
77,800
44,200
120,100
1.76
1.54
CMN0005c
74,200
41,500
112,400
1.79
1.52
CMN0005d
80,100
45,600
121,300
1.75
1.52
CMN001a
66,800
39,400
97,300
1.69
1.46
CMN001b
68,900
39,400
102,200
1.75
1.48
a): Sample named as in Table 2.15.
§2.3.9.4 Polymerization in the presence of γ-Butyrolactone Bulk polymerization of vinyl chloride in the presence of γ-butyrolactone (GBL) had a 3-7 minute induction time for most runs, as shown in Table 2.18. The initial polymerization rate R p0 increased about 10% over the control with added GBL, and stayed that level until GBL reached 4 mole%. The molecular weights and the molecular weight distributions of the resulting PVCs, determined by GPC, are listed in Table 2.19. - 106 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The molecular weights of the resulting PVCs increased with added GBL, reached a maximum at about 0.5 mole% GBL concentration (Mn was about 10% higher than that of the control), and decreased at higher GBL concentration. The molecular weight distribution was about 1.76-1.87. We believe GBL forms a complex with the propagating radical, retarding chain transfer to monomer. Thus the growing radical adds more monomers before chain transferring (i.e. increasing the molecular weight). Chain transfer on GBL generates a very reactive radical, butyrolactonyl radical that reinitiates very quickly. Its termination with a propagating radical is negligible. Therefore the polymerization kinetics can be evaluated using equations (2.58) and (2.59), developed in section 2.3.5:
1
⎛ fk k k ⎞ 2 1 fk d R p0 = ⎜⎜ d ri1 p ⎟⎟ [ I ] 2 [ M ] + [ I ] …………..………………..(2.58) 2CM 1Φ ⎝ CM 1kt1Φ ⎠
And 1 [ A] = C M 1Φ + C A 0 Pn [M ] 1
1
⎛ C fk k Φ ⎞ 2 [ I ] 2 fk k [ I ] ……………..(2.59) + ⎜ M 1 d t1 ⎟ + d t1 ⎜ k k ⎟ [ M ] 2 k k [ M ]2 ri1 p p ri1 ⎝ ⎠ ktr 2 K [ A] ktr1 ktr1 , and , CM 1 = kp 1 + K [ A]
1+
Here Φ =
CA =
ktr 3 . kp
- 107 -
32
1.6
28
1.4
24
1.2
20
1.0
16
0.8
12
0.6
8
0.4
4
0.2
0
1000/P0n
1.8
0.0
0.5
1.0
1.5
Figure 2.12. Plots of 1 / Pn0 and
2.0 2.5 3.0 100x[GBL]/[VC] R p0
3.5
4.0
R0p x 10 5 ( moleL -1s-1 )
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
4.5
vs [GBL]/[VC] for bulk polymerization of vinyl
o
chloride in the presence of GBL at 55 C. ○ and ● are the experimental data and the solid curves are the calculated results of equation (2.58) and equation (2.59) by least-squares regression (K=1000, ktr2/ktr1=0.87±0.05, CA=(73.4±19.5)x10-4).
Using least-squares regression, we fit the experimental data of Pn0 and R p0 with equations (2.58) and (2.59). The regression returned a best fit for parameters k tr 2 k tr 1 =0.87±0.05
The plots of
1 / Pn0
and CA=(73.4±19.5)x10-4 using an equilibrium constant of K=1000. and
R 0p
vs [GBL]/[VC] are shown in Figure 2.12. We can see the
calculated curves match very well with the experimental results. The value of k tr 2 k tr1 is 0.87±0.05, which means that the chain transfer rate constant to monomer decreased for the propagating radicals complexed with GBL. This increased the molecular weight about 10% at the optimal GBL concentration of 0.3 mole%. Since k tr 2 k tr1 is less than 1, Φ will always less than 1 except when [A]=0. From equation (2.58), we can see that
R p0
also
increases, when Φ is lower than 1. When K[A] becomes very large, Φ will be close to
- 108 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ k tr 2 k tr 1 . R 0p will
reach a maximum value and remain constant as GBL increases. From
Figure 2.12, it can be seen that the experimental
R p0 did
increase about 10% with added
GBL. When GBL concentration was over its optimum value, the experimental R 0p seemed to drop slightly. We believe this is because chain transferring on GBL is generating more radicals at higher high GBL concentrations. These GBL radicals can terminate as well as reinitiate and thus slow the polymerization rate.
Table 2.18. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of γ-butyrolactone (GBL). Conv. R0px105 AIBN GBL VC Indb) Time PVC Runa) (%) (moleL-1 s-1) (mg) (g) (g) (min) (min) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 GBL0025b
31.7
0.086
23.42
5
87
1.40
5.98
11.6
GBL005b
31.8
0.170
27.07
5
85
1.29
4.77
10.0
GBL0075a
32.1
0.263
25.43
7
82
1.50
5.90
12.2
GBL01a
16.0
0.187
17.94
5
130
1.27
7.08
8.8
GBL01d
31.6
0.351
24.10
7
79
1.53
6.35
13.4
GBL02a
16.0
0.352
15.35
5
78
0.64
4.17
9.8
GBL02b
31.8
0.823
24.72
3
89
1.78
7.20
13.0
GBL04a
16.0 0.6904
16.92
7
76
0.60
3.55
8.8
GBL04b
32.0 1.3766
24.62
3
92
1.42
5.77
10.7
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 109 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.19. Molecular weights and molecular weight distributions for PVCs prepared at 55oC initiated by AIBN in the presence of GBL. Sample Mw Mn Mz Mz/Mw Mw/Mn Designationa) Control 78,500 43,100 125,000 1.82 1.59 GBL0025b
84,700
47,100
133,000
1.80
1.57
GBL005b
84,700
47,400
132,500
1.78
1.56
GBL0075a
82,400
46,000
126,200
1.79
1.53
GBL01a
84,400
45,100
125,700
1.87
1.49
GBL01d
78,700
44,200
120,000
1.78
1.52
GBL02a
80,000
43,800
124,800
1.85
1.54
GBL02b
78,500
44,200
117,100
1.77
1.49
GBL04a
76,000
42,700
114,800
1.78
1.51
GBL04b
74,600
42,400
113,000
1.76
1.52
a): Sample named as in Table 2.18.
§2.3.9.5 Polymerization in the presence of Dimethyl terephthalate The bulk polymerization of vinyl chloride in the presence of dimethyl terephthalate (DMT) had a relatively short induction time, only 2 minutes for most runs, Table 2.20. Two runs had induction times of 10 minutes, probably because the larger apparatus needed a longer time to reach the equilibrium temperature. DMT up to a 2% level did not retard the polymerization but actually have increased the polymerization rate. The initial polymerization rate R p0 increased as DMT increased and reached a maximum at about 1 mole% of DMT, then dropped a little bit with higher DMT - 110 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ concentration, but still remained higher than that of the control, up to 2mole% level. The polymerization conversion was 6-8% in most cases. The molecular weights and the molecular weight distributions of the resulting PVCs are listed in Table 2.21. As can be seen, the molecular weights of most samples are higher than that of the control. The molecular weight distributions were about 1.73-1.87 for most cases.
Again, we believe DMT formed a complex with the propagating radical that slowed chain transfer to monomer and thus increased the molecular weight. Since the propagating radical is active it can also chain transfer to DMT to terminate the growing chain and generate an additive radical by attacking the aromatic ring of DMT to form a σ-complex radical. This complex radical is unstable and returns to aromaticity by
releasing a proton radical. This process is probably very fast and the proton radical can easily react with monomer and form a new propagating radical. Then, its termination with propagating radicals can be ignored. Thus the kinetic behavior of the polymerization of vinyl chloride in the presence of DMT can be evaluated using equation (2.58) and equation (2.59) as were developed in section 2.3.5.
By least-squares regression we analyze our experimental data of Pn0 and R p0 using equations (2.58) and (2.59). The regression returns a best fit for parameters k tr 2 k tr 1
=0.89±0.02 and CA=(60.9±19.3)x10-4 at an equilibrium constant K=1370. The
plots of 1 / Pn0 and
R 0p
curves of 1 / Pn0 and
vs [DMT]/[VC] are shown in Figure 2.13. We can see the calculated
R p0
match the experimental results within the experimental error. The
value of k tr 2 k tr1 is 0.89±0.02. Thus both
Pn0 and R 0p
- 111 -
increased about 10% at an optimal
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ DMT concentration of 0.3 mole%, relative to VC, Figure 2.13. Due to the experimental variation, the points seemed scatter, but still follow the trend that equation (2.58) and
36
1.6
32
1.4
28
1.2
24
1.0
20
0.8
16
0.6
12
0.4
8
1000/P0n
1.8
R0p x 10 5 ( moleL -1s-1 )
equation (2.59) describe.
4
0.2 0.0
0.2
0.4
0.6
Figure 2.13. Plots of 1 / Pn0 and
0.8 1.0 1.2 100x[DMT]/[VC] R 0p
1.4
1.6
1.8
2.0
vs [DMT]/[VC] for bulk polymerization of vinyl
chloride in the presence of DMT at 55 oC. ○ and ● are the experimental data and the solid curves are the simulation results of equation (2.58) and equation (2.59) by leastsquares regression (K=1370, ktr2/ktr1=0.89±0.02, CA=(60.9±19.3)x10-4).
- 112 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.20. Bulk polymerization of vinyl chloride at 55oC initiated by AIBN in the presence of dimethyl terephthalate (DMT). Conv R0px105 AIBN DMT VC Indb) Time PVC Runa) (%) (mole•L-1•s-1) (mg) (g) (g) (min) (min) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 DMT0005a
32.5
0.039
26.43
3
74
1.34
5.07
12.0
DMT001c
33.3
0.077
26.83
3
71
1.60
5.96
14.2
DMT004a
16.0
0.154
18.93
2
98
1.25
6.60
11.1
DMT005a
64.0
0.773
86.75
10
90
3.68
4.24
8.6
DMT005d
31.6
0.382
25.29
3
76
1.57
6.21
13.6
DMT01a
16.0
0.389
16.84
2
98
1.65
9.80
14.5
DMT01b
64.0
1.558
85.71
10
90
5.12
5.97
11.2
DMT01c
32.0
0.776
33.89
2
85
3.06
9.03
15.8
DMT01d
32.7
0.773
29.59
2
85
1.84
6.22
12.2
DMT015a
32.3
1.165
25.58
2
93
2.29
8.95
14.4
DMT015b
16.3
0.586
11.34
2
84
0.82
7.23
13.8
DMT02b
32.0
1.554
34.18
5
90
2.21
6.47
11.9
DMT02c
16.5
0.778
13.28
2
84
0.90
6.78
13.2
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time
- 113 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.21. Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of DMT. Sample Mw Mn Mz Mz/Mw Mw/Mn Designationa) Control 78,500 43,100 125,000 1.82 1.59 DMT0005a
83,200
45,700
133,200
1.82
1.60
DMT001c
87,100
48,300
136,900
1.80
1.57
DMT004a
79,200
46,300
116,100
1.71
1.47
DMT005a
85,100
48,000
130,200
1.77
1.53
DMT005d
82,400
45,500
127,100
1.81
1.54
DMT01a
89,500
50,200
135,200
1.78
1.51
DMT01b
99,500
53,200
152,200
1.87
1.53
DMT01c
85,500
47,100
130,000
1.82
1.52
DMT01d
85,500
46,200
128,300
1.85
1.50
DMT015a
72,300
47,000
107,400
1.73
1.49
DMT015b
77,800
44,100
119,900
1.76
1.54
DMT02b
73,900
41,900
108,900
1.76
1.47
DMT02c
80,300
45,300
122,600
1.77
1.53
a): Sample named as in Table 2.20.
- 114 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.3.9.6 Polymerization in the presence of Trimethyl 1,3,5-benzenetricarboxylate The bulk polymerization of vinyl chloride in the presence of trimethyl 1,3,5benzenetricarboxylate (TMB) had 3-9 minute induction times, with an average of 4-5 minutes, Table 2.22. The conversion was kept at 4-5%. The initial polymerization rate, R p0 ,
increased slightly with the addition of a small amount of TMB, and then remained
constant from 0.2 to 1.0 mole%. The molecular weights and molecular weight distributions of the resulting PVCs are listed in Table 2.23. They are all higher than that of PVC prepared in absence of additive (control). The molecular weight distribution is around 1.79-1.86. Again, we postulate that the added TMB formed a complex with the propagating radical that slowed chain transfer to monomer. Also, the propagating radical could chain transfer to TMB and finally generate a proton radical, which mainly added monomer very quickly to regenerate a propagating radical. So the kinetics behavior of the polymerization can also be evaluated using equation (2.58) and equation (2.59) as we developed in section 2.3.5.
The parameters were determined by applying equations (2.58) and (2.59) to the experimental data for Pn0 and R p0 using least-squares regression. The regression returns a best fit for parameters k tr 2 k tr1 =0.89±0.07 and CA=(124.5±92.1)x10-4 with an equilibrium constant K=2370. The plots of 1 / Pn0 and
R 0p
vs [TMB]/[VC] are shown in Figure 2.14. The
theoretical curves for 1 / Pn0 and
R p0
match very well with the experimental results within
the experimental error. Both
Pn0
and
R p0
increased about 10% at the optimal TMB
concentration with k tr 2 k tr1 = 0.89 . The optimal concentration is 0.17 mole% relative to VC - 115 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ as shown in Figure 2.14. The experimental
R 0p
decreased at higher TMB concentration;
this is most likely due termination with TMB radicals.
1.8
28
1.6
24
1.4
20
1.2
16
1.0
12
0.8
8
0.6
4
0.4
0 0.0
0.1
0.2
0.3
0.4 0.5 0.6 0.7 100x[TMB]/[VC]
Figure 2.14. Plots of 1 / Pn0 and
R p0
0.8
0.9
1.0
R0p x 10 5 ( moleL -1s-1 )
32
1000/P0n
2.0
1.1
vs [TMB]/[VC] for bulk polymerization of vinyl
chloride in the presence of TMB at 55 oC. ○ and ● are the experimental data and the solid curves are the simulation results of equation (2.58) and equation (2.59) by leastsquares regression (K=2370, ktr2/ktr1=0.89±0.07, CA=(124.5±92.1)x10-4).
- 116 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 2.22. Bulk polymerization of vinyl chloride at 55 oC initiated by AIBN in the presence of trimethyl 1,3,5-benzenetricarboxylate (TMB). R0px105 AIBN TMB VC Indb) Time PVC Conv Runa) (%) (mole•L-1•s-1) (mg) (g) (g) (min) (min) (g) Control 32.6 0.000 25.44 3 76 1.43 5.70 10.6 TMB00025a
32.5
0.025
25.03
3
69
1.11
4.43
11.6
TMB0005b
32.5
0.051
26.41
5
70
1.23
4.65
11.9
TMB001c
31.7
0.101
25.75
3
74
1.18
4.58
11.1
TMB0025c
33.0
0.254
24.10
9
86
1.03
4.27
9.1
TMB005d
33.2
0.504
25.42
4
80
1.34
5.27
11.5
TMB0075a
31.7
0.756
23.24
8
83
0.99
4.25
9.4
TMB01a
32.7
1.021
24.84
5
90
1.29
5.19
10.1
a): The experimental run is named by combining the additive abbreviation with a number followed by a lower-case letter. The number represents the mole ratio of additive to vinyl chloride (omitting the decimal point), and the lower-case letter is used to identify the different run. b): Induction time. Table 2.23. Molecular weights and molecular weight distributions for PVCs prepared at 55 oC initiated by AIBN in the presence of TMB. Mw Mn Mz Mw/Mn Mz/Mw Samplea) Control 78,500 43,100 125,000 1.82 1.59 TMB00025a
85,700
46,000
138,900
1.86
1.62
TMB0005b
89,500
48,200
142,800
1.86
1.60
TMB001c
86,400
46,700
136,500
1.85
1.58
TMB0025c
78,600
44,000
120,000
1.79
1.53
TMB005d
82,000
45,500
125,400
1.80
1.53
TMB0075a
79,400
44,200
119,700
1.80
1.51
TMB01a
85,000
47,300
128,600
1.80
1.51
a): Sample named as in Table 2.22. - 117 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The values of K , k tr 2 / k tr1 and CA for CMN, DMT, GBL and TMB are summarized Table 2.24. As can be seen, the values of k tr 2 / k tr 1 for these four additives are close to each other. They are all esters or lactones. These additives can hydrogen-bond with the propagating radicals and lower the chain transfer rate to monomer, increasing the PVC molecular weight. Of these four additives, CMN has the largest equilibrium constant, and complexes with most of the propagating radicals at a very low concentration. CMN is also an active chain transfer agent (CA=0.53±0.04) and chaintransfer to CMN generates a stable radical that terminates with Rn• or M • . Therefore the PVC molecular weight increased only at very low concentrations of CMN. The optimal concentration was found to be [CMN]/[VC]=10-4, as listed in Table 2.24. On the other hand, much higher optimal concentrations were found for GBL, DMT and TMB, Table 2.24, since they all have much smaller K’s and CA’s compared to those for CMN.
EC, TBPO and TEPO also increased the initial polymerization rate and resulting polymer molecular weights at low additive concentrations, but there are too few experimental data points to calculate ktr 2 ktr1, CA and K values for these additives. However, we believe they formed a similar complex with the propagating radical as CMN, GBL, DMT and TMB did, and reduced the chain transfer rate constant to monomer. The propagating radical could grow a longer chain before chain transfer. EC and TBPO are particularly interesting since they have much smaller chain transfer constants, 21±2x10-4 for EC and 24±16x10-4 for TBPO. Both additives increased the resulting polymer molecular weights at very low concentrations, and the molecular weights remained higher than that of the control even at higher concentrations. This - 118 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ implies that they easily complex with propagating radicals and have smaller chain transfer constants.
Table 2.24. Values of K ,
k tr 2 / k tr1 and C A for vinyl chloride polymerization at 55 oC in
the presence of 2-Coumaranone (CMN), dimethyl terephthalate (DMT), γ-butyro-lactone (GBL) or trimethyl 1,3,5-benzenetricarboxylate (TMB). ktr 2 Optimal 4 [A]/[VC]x104 Additive K C A x 10 ktr1 CMN
60000
0.89±0.01
5300±400
1.0
GBL
1000
0.87±0.05
73.4±19.5
30
DMT
1370
0.89±0.02
60.9±19.3
30
TMB
2370
0.89±0.07
124.5±92.1
17
§2.4 Conclusions The polymerization rate for bulk polymerization of vinyl chloride was found to be proportional to the 0.62 power to the initiator concentration. A proposed mechanism suggested
that,
under
the
heterogeneous
condition,
mutual
propagating
radical/propagating radical termination was relatively rare and the dominant termination reaction was the reaction of propagating radicals with small radicals, formed by chain transferring to monomer or additive. A kinetic model was developed based on the proposed mechanism. The rate constant for the cross-termination reaction was determined to be about 109-1010 L•mole-1•s-1.
- 119 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Various organic materials were added to vinyl chloride for bulk as well as suspension polymerization. The initial polymerization rates and the molecular weights of the resulting polymers increased in the presence of weakly ‘basic’ compounds such as dimethyl terephthalate (DMT), ethylene carbonate (EC), γ-butyrolactone (GBL), tributylphosphine oxide (TBPO) and trimethyl-1,3,5-benzene tricarboxylate (TMB). A modified kinetic model was developed for the bulk polymerization of vinyl chloride in the presence of these weakly basic additives, assuming a hydrogen-bond complex formed between an additive and the terminal hydrogen of the propagating radical. The H-bonded propagating radicals could grow, on the average, longer chains, and the molecular weights of the resulting polymers increased. The kinetic model was tested on four sets of experimental data and the calculated results were in good agreement with the experimental findings.
Differential Scanning Calorimetry (DSC), Nuclear Magnetic Resonance (NMR) analysis and dehydrochlorination of the PVC samples, prepared in presence or absence of organic additives, will be discussed in Chapter 3 and Chapter 4.
- 120 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §2.5 References 1. Burleigh, P. H. J. Am. Chem. Soc. 1960, 82(3), 749. 2. Rosen, I.; Burleigh, P. H.; Gillespie, J. F. J. Polym. Sci. 1961, 54, 31-44. 3. Imoto, M.; Takemoto, K.; Nakai, Y. Makromol. Chem. 1961, 48, 80-8. 4. Sumi, M.; Imoto, M. Makromol. Chem. 1961, 50, 161-5. 5. Zegel'man, V. I.; Shlykova, M. N.; Svetozarskii, S. V.; Zil'berman, E. N. Vysokomolekulyarnye Soedineniya, Seriya A 1968, 10(1), 114-18. 6. Razuvaev, G. A.; Zil'berman, E. N.; Zegel'man, V. I.; Svetozarskii, S. V.; Pomerantsva, E. G. Doklady Akademii Nauk SSSR 1966, 170, 1092-5. 7. Boeckman, O. C. J. Polym. Sci. A 1965, 3, 3399-404. 8. Pham, Q. T.; Taieb, M. J. Polym. Sci. Polym. Chem. Ed. 1972, 10, 2925-34. 9. Zil'berman, E. N. J. Macromol. Sci.-Rev. Macromol. Chem. Phys. 1995, C35(1), 47-62. 10. Turska, E.; Obloj-Muzaj, M. Acta Polymerica 1981, 32(6), 295-9. 11. Grishin D. F.; Semenycheva, L. L.; Pavlovskaya, M. V.; Sokolov, K. V. Russian Journal of Applied Chemistry 2001, 74(9), 1591-9.
12. Semenycheva, L. L.; Grishin D. F. Russian J. Appl. Chem. 2003, 76(6), 851-8. 13. Percec, V.; Popov, A. V.; Ramirez-Castillo, E.; Weichold, O. J. Polym. Sci. Part A Polym. Chem. 2003, 41, 3283-99.
14. Percec, V.; Ramirez-Castillo, E.; Hinojosa-Falcon, L. A.; Popov, A.V. J. Polym. Sci. Part A Polym. Chem. 2005, 43, 2185-7.
15. Vorenkamp, E. J.; Challa, G. Polymer 1988, 29(1), 86-92. 16. Belhaneche-Bensemra, N.; Belaabed, B.; Bedda, A. Macromol. Symp. 2002, 180, 203-15. - 121 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 17. Doty, P.; Wagner, H.; and Singer, S. J. Phys. Chem. 1947, 51, 32-57. 18. Lyngaae-Jorgensen, J. H. J. Chromatogr. Sci. 1971, 9, 331-40. 19. Lyngaae-Jorgensen, J. H. Makromol. Chem. 1973, 167, 311-9. 20. Rudin, A.; Benschop-Hendrychova, I. J. Appl. Polym. Sci. 1971, 15(12), 2881-904. 21. Abdel, A. H; Hamielec, A. E. J. Appl. Polym. Sci. 1972, 16(5), 1093-101. 22. Abdel, A. H; Hamielec, A. E. J. Appl. Polym. Sci. 1973, 17(10), 3033-47. 23. Pang, S.; Rudin, A. J. Appl. Polym. Sci. 1993, 49(7), 1189-96. 24. Manabe, N.; Kawamura, K.; Toyoda, T.; Minami, H.; Ishikawa, M.; Mori, S. J. Appl. Polym. Sci. 1998, 68(11), 1801-9.
25. Talamini, G. J. Polym. Sci. Chem. Ed. 1966, 4, 535-7. 26. Crosato-Amaldi, A.; Gasparini, P.; Talamini, G. Makromol. Chem. 1968, 117, 14052. 27. Ugelstad, J.; Flogstad, H.; Hertzberg, T.; Sund, E. Makromol. Chem. 1973, 164, 17181. 28. Ugelstad, J. J. Macromol. Sci.-Chem. 1977, A11(7), 1281-305. 29. Abdel-Alim, A. H.; Hamielec, A. E. J. Appl. Polym. Sci. 1972, 16, 783-99. 30. Olaj, O. F. J. Macromol. Sci.-Chem. 1977, A11(7), 1307-17. 31. Suresh, A. K.; Chanda, M. Eur. Polym. J. 1982, 18, 607-16. 32. Xie, T. Y.; Yu, Z. Z.; Cai, A. Z.; Pan, Z. R. J. Chem. Ind. And Eng. (China), 1984, 2, 93-100. 33. Weickert, G.; Henschel, G.; Weiβenborn, K. Angew. Makromol. Chem. 1987, 147, 133. 34. Xie, T. Y.; Hamielec, A. E.; Wood, P. E.; Woods, D. R. Polymer 1991, 32(3), 537-57.
- 122 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 35. Xie, T. Y.; Hamielec, A. E.; Wood, P. E.; Woods, D. R. J. Vinyl Tech. 1991, 13(1), 225. 36. Tamamini, G.; Visentini, A.; Kerr, J. Polymer 1998, 39(10), 1879-91. 37. Tamamini, G.; Kerr, J.; Visentini, A. Polymer 1998, 39(18), 4379-84. 38. Danusso, F.; Pajaro, G.; Sianesi, D. Chimica e l'Industria (Milan) 1955, 37, 695-701. 39. Vidotto, G.; Crosato-Arnaldi, A.; Talamini, G. Makromol. Chem. 1968, 114, 217-25. 40. Russo, S.; Stannett, V. Makromol. Chem. 1971, 143, 47-56. 41. Arnett, L. M.; Peterson, J. H. J. Am. Chem. Soc. 1952, 74, 2031-3. 42. Dausso, F; Pajaro, G.; and Sianese, D. Chimica e l'Industria (Milan) 1959, 41, 117075. 43. Breitenbach, J. W. Makromolekulare Chemie 1952, 8, 147-55. 44. Bengough, W. I.; Norrish, R. G. W. Proc.Roy. Soc. London Series A 1950, 200, 30120. 45. Danusso, F.; Sabbioni, F. Chimica e l'Industria (Milan, Italy) 1955, 37, 1032-4. 46. Talamini, G.; Vidotto, G. Makromol. Chem. 1962, 53, 21-7. 47. Cotman, J. D.; Gonzalez, M. F.; Claver, G. C. J. Polym. Sci. Part-A 1967, 5 1137-64. 48. Bawn, C. E. H.; Mellish, S. F. Trans. Faraday Soc. 1951, 47, 1216-27. 49. Talat-Erben, M.; Bywater, S. J. Am. Chem. Soc. 1955, 77(14), 3712-4. 50. Van Hook, J. P. Tobolsky, A. V. J. Am. Chem. Soc. 1958, 80(4), 779-82. 51. Bawn, C. E. H.; Verdin, D. Trans. Faraday Soc. 1960, 56, 815-22. 52. Barrett, K. E. J. Appl. Polym. Sci. 1967, 11(9), 1617-26. 53. Burnett, G. M.; Wright, W. W. Proc. Roy. Soc. Ser. A, 1954, 211, 41-53. 54. Bengough, W. I.; Thomson, R. A. M. Trans. Faraday Soc. 1965, 61, 1735-44.
- 123 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 55. Kajiwara, A.; Kamachi, M. Macromol. Chem. Phys. 2000, 201(16), 2165-9. 56. Sidiropoulou, E.; Kiparissides, C. J. Macromol. Sci.-Chem. 1990, A27(3), 257-88. 57. Kiparissides, C.; Daskalakis, G.; Achilias, D. S.; Sidiropoulou, E. Ind. Eng. Chem. Res. 1997, 36, 1253-67.
58. Krallis, A.; Kotoulas, C.; Papadopoulos, S.; Kiparissides, C.; Bousquet, J.; Bonardi, C. Ind. Eng. Chem. Res. 2004, 43, 6328-99. 59. Chang, K. H. S.; Litt, M. H.; Nomura, M. Emulsion Polymerzation of Vinyl Acetate, edited by El-Aasser, M. S. and Vanderhoff, J. W. 1981, Chapter 6, p89-136. 60. Bejan, E. V.; Font-Sanchis, E.; and Scaiano, J. C. Organic Letters 2001, 3(25), 405962.
- 124 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Chapter 3 Differential Scanning Calorimetry and Thermogravimetric Analysis of Poly(vinyl chloride)s
- 125 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.1 Introduction It is well-known that commercial poly(vinyl chloride) (PVC) is a low crystallinity polymer. PVC prepared at 50-60 oC contains approximately 10% crystallinity. But this relatively low level of crystallinity has a surprisingly significant effect on its processing and properties. Many techniques have been used to investigate the crystalline nature of the polymer. These include electron and x-ray diffraction1-14, vibrational spectroscopy such as infrared and Raman6,
9-10, 15-18
, thermal analysis such as differential scanning
calorimetry (DSC)19-25, density determination10,
26-28
and nuclear magnetic resonance
spectroscopy29. Among these, wide angle X-ray scattering and infrared spectroscopy were the most extensively used techniques. DSC was the most widely used thermal analysis method. It should be pointed out that each method examines somewhat different aspects of the crystalline nature of the polymer. Because the methods look at different aspects of the crystallinity, each method tends to call different levels of order crystalline, and draws the demarcation line between crystalline and non-crystalline regions differently, according to its own limit of detection. It is therefore very important to remember what each technique is actually measuring. Also, the level of order in any given sample is a very strong function of thermal history. So, it is not surprising that reported PVC crystallinity varies widely from polymer to polymer, as well as between measurement methods.
Natta and Corrandini1 first studied the unit cell structure of PVC using X-ray diffraction in 1956. They reported that PVC crystalline unit cell is orthorhombic and its lattice parameters are a=1.04nm, b=0.53nm, c=0.51nm, respectively. Each cell contains - 126 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ two chains arranged in a planar syndiotactic conformation. In 1973, Wilkes, Folt and Krimm4 prepared low molecular weight PVC by polymerization in butryraldehyde and, after purification, produced a single crystal mat and studied its X-ray diffraction patterns. They found a more compact orthorhombic crystal structure with lattice constants of a=1.024nm, b=0.524nm, and c=0.508nm. The nature of larger scale order is still far from clear. Both nodular and lamellar crystalline structures have been proposed in the literature7, 8. Blundell8 favored the former, but accumulating evidence appears to point to lamellar-type structures. Wenig7 carried out a detailed study of commercial PVC samples, using small-angle x-ray scattering (SAXS), wide-angle x-ray diffraction (WAXD), and small-angle light scattering (SALS). Wenig concluded that most crystallites were lamellar, though some rodlike entities were also present. Biais et al.9 suggested that PVC crystallites are ribbonlike and narrow in both b- and c-directions, which would be consistent with Wenig’s rodlike crystallites.
Both X-ray diffraction and thermal analysis provide evidence for the atypical crystalline nature of PVC. The X-ray diffraction trace for so-called amorphous PVC is bimodal6, 30, rather than having a broad unimodal diffraction peak as is found for most amorphous polymers. The crystalline diffraction peaks for non-oriented PVC are unusually broad, suggesting the presence of small, imperfect crystallites. A DSC trace of PVC powder shows a very broad endotherm, covering the range from just above the glass transition temperature to above 200 oC. The rapid crystallization of PVC has been demonstrated by thermal analysis19-25 and is one of the reasons why it is difficult to get a 100% amorphous sample, which is sometimes required for measurement of crystallinity.
- 127 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ In this chapter, Differential Scanning Calorimetry (DSC) was used to study the crystallization behavior of PVC samples (preparation discussed in the previous chapter). The annealing effect on crystallinity was extensively studied for PVC samples prepared in the presence of various additives, as well as samples prepared without additives. A subtraction method was used to calculate the crystallinity of PVC samples. Crystallization kinetics of PVC is discussed based on the crystallinity obtained as a function of annealing time. The additive effect was evaluated by comparing the crystallinity of different PVCs after the same thermal cycling. The degradation behavior of a few representative samples was studied using dynamic thermogravimetric analysis (TGA). The overall degradation activation energy was determined by several different methods and the result is in agreement with the literature results.
§3.2 Experimental §3.2.1 Materials PVCs prepared by free radical bulk and suspension polymerization methods as described in Chapter 2 were studied. The polymers were precipitated twice from a 5% THF solution into a large volume of methanol, and dried in a vacuum oven at 60 oC for 48 hours. The purified PVC samples were white, cotton-like fibers.
- 128 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.2.2 DSC Measurement of PVC Samples Differential Scanning Calorimetry (DSC) measurements were carried out using a TA 2910 modulated differential scanning calorimeter calibrated using Indium (melting point: 156.60 oC, heat of fusion: 28.71 J/g). The data was collected using Thermal Advantage version 1.1A instrument control software and analyzed by Universal Analysis 2000 version 3.1E software. A heating rate of 20 oC/min and a nitrogen flow rate of 50ml/min were used for the calorimetric measurements. The PVC sample used weighed about 10~15mg. For the annealing studies, PVC samples were heated under a nitrogen atmosphere in the DSC cell to 230 oC, rapidly cooled to the annealing temperature, held there for a specified time, and then cooled rapidly to -60oC. The samples were then scanned at 20 oC/min to about 250 oC to observe the melting behavior. For the quenching studies, the samples were heated to 230 oC, kept there for one minute, then rapidly cooled to -60oC, using a dry-ice filled cold cylinder. Then the samples were scanned to 250 oC. Quenched PVC DSC curves were used as a reference to measure the increase of crystallinity of annealed samples.
§3.2.3 TGA Measurement of PVC Samples The thermogravimetric experiments were carried out using a TA 2950 thermogravimetric analysis instrument with platinum crucibles. The weights of the samples were about 4-8 mg. Small PVC samples were used to reduce the temperature difference between the sample and the crucible, and minimize any heat transfer barriers between the organic polymer and metal crucibles. The tests were performed in a dynamic
- 129 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ mode, going from room temperature to 600 oC. Experiments were carried out under nitrogen, with a flow rate of 100 ml/min in order to remove the evolved corrosive gases rapidly. Heating rates of 2, 5, 10, 15 and 20 oC/min were used for selected samples.
§3.3 Results and Discussion §3.3.1 Crystallinity of PVCs prepared in the presence of Additives This section reports on the determination of crystallinity of PVCs using Differential Scanning Calorimetry (DSC). The PVCs studied here were prepared in bulk at 55 oC in the presence of 2-coumaranone (CMN), 2,6-dichloropyridine (DCPY), 1,4dimethoxybenzene (DMB), dimethyl terephthalate (DMT), ethylene carbonate (EC), γbutyrolactone (GBL) and trimethyl-1,3,5-benzenetricarboxylate (TMB). The sample designations are described in Chapter 2. All samples were purified using the procedure described in Chapter 2. All samples were annealed at 100 oC for 30 minutes after heating to 230 oC, to eliminate differences in thermal history, so samples could be reliably compared. The effect of additive concentration on the crystallinity of the resulting PVCs is investigated in this section. The effects of annealing time and annealing temperature on the crystallinity of PVC samples will be discussed in section 3.3.2 and 3.3.3.
- 130 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Figure 3.1. DSC curves for PVC/AIBN416 recorded at a heating rate of 20 oC/min. A: sample without any pretreatment; B: sample heated to 230 oC followed by quenching; C: sample annealed at 100 oC for 30 minutes after heating to 230 oC and then quenched; D: curve C subtracted by curve B.
Figure 3.1 shows typical DSC traces for PVC samples. Curve A was recorded for the sample without any pretreatment, directly heating from -60 oC to 230 oC at a heating rate of 20 oC/min under nitrogen atmosphere. Section 3.3.4 will show that at this heating rate, PVC decomposition during the short time at 230 oC can be ignored (refer to Figure 3.68). Curve B is the 2nd run of the specimen used in curve A, after quenching from 230 o
C to -60 oC. As seen in curve A, there is a small endotherm, beginning at 85 oC to about
100 oC with a maximum at 93 oC; then a second broad endotherm is observed from 100 to 230 oC, with a maximum at 174 oC. In curve B the small endotherm at 93 oC disappears, but the broad endotherm remains almost the same. The small endotherm at or below the glass transition temperature (sub-Tg) is attributed to volume relaxation of
- 131 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ PVCs during drying and storage at room temperature. Many researchers19, 20, 31, 32 have studied this sub-Tg relaxation behavior, and they believe this relaxation is not involved with crystallization, but is due to the decrease of free volume in PVC.
It can also be seen that after quenching the shape and the scale of the broad endotherm of curve B changes only slightly. This broad endotherm is attributed to the melting of imperfect crystallites, which means it is very difficult to get 100% amorphous PVC by quenching. Due to the baseline shift, it is very hard to evaluate the fraction of such crystallites that are left after quenching. It is possible that the crystallites melted after first heating polymer undergo fast recrystallization during the short quenching time (from 230 oC down to 90 oC in 1~2 minutes). The curve B endotherm is similar to that seen in curve A. There is a maximum at 176 oC for curve B.
Curve C is the DSC curve for a quenched PVC sample, subsequently annealed at 100 oC for 30 minutes, followed by quenching to -60 oC. Compared with curve B, one can see that a new small endotherm, starting at 105 oC and going to 130 oC, with a maximum at 118 oC, is formed during annealing. Curve D is the difference curve, curve C minus curve B. One can see that, after subtraction, only the endothermic peak around 118 oC is left. The baseline was corrected to be almost parallel to the x axis. Integrating this 118 oC endotherm, we find that the heat of fusion, ΔH, of this sample to be 2.5J/g. This is attributed to the crystallites formed during the 30-min annealing time at 100 oC. These crystallites have a characteristic melting temperature, approximately 20 degrees higher than the annealing temperature (i.e. crystallization temperature).
- 132 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ As we mentioned earlier, due to the rapid crystallization, it is very difficult to produce a 100% amorphous PVC sample. Here we subtract the quenched sample curve from the annealed sample curve (we use the quenched sample curve as the baseline), so the calculated ΔH for the annealed sample is a relative one, but it is convenient to evaluate the annealing effect on crystallization and make comparisons between samples under the same pretreatment. This “subtraction method” was also used by other authors42. Other techniques such as X-ray diffraction also use the similar method to get the polymer crystallinity*. All the following DSC curves were analyzed by this method. A difference curve is defined as the annealed sample curve minus the quenched sample curve. The quenched-sample-curve, called the reference curve, is displayed in some Figures in the following sections.
Figure 3.2. DSC curves for control PVCs prepared in the absence of additives (samples annealed at 100 oC for 30 minutes after heating to 230 oC, heating rate: 20 oC/min). ___________________________________ *In X-ray diffraction, the scattering patterns from the quenched sample was used to determine the amorphous areas in the annealed sample patterns and the crystallinity was calculated as crystalline_areas/(crystalline_areas+amorphous_areas)6, 12. - 133 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ A typical series of thermograms for PVCs prepared in bulk, using 0.05% 2,2’azobis-isobutyronitrile (AIBN) as initiator, is shown in Figure 3.2. The 8 curves are from six PVC samples designated as AIBN319, AIBN320, AIBN326, AIBN416, AIBN430, AIBN507 (two of them run twice). These six samples are from multiple runs using the same polymerization recipe; the number average molecular weights are between 41,200 and 44,500, and their molecular weight distributions are between 1.80 and 1.90. All the samples were annealed at 100 oC for 30 minutes. It can be seen from Figure 3.2 that the samples have the same glass transition temperature (Tg) and similar melting behavior. There is a Tg at 91 oC, a small endotherm at about 119 oC, and a second broad endotherm at about 175 oC.
Figure 3.3. DSC Difference curves for control PVCs prepared in the absence of additives. A reference curve (i.e. a quenched-sample-curve) was subtracted from the curves in Figure 3.2; the resulting curves are shown in Figure 3.3. The remaining endotherms in the difference spectra all show similar melting temperatures (Tm) at about 119 oC, about - 134 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 20 oC higher than the annealing temperature. The corresponding heats of fusion, ΔH, of the annealed samples (by integrating the endotherm from 100-130 oC) are listed in Table 3.1. The calculated values of ΔH are very close to each other; the average ΔH is 2.42±0.11 J/g, as shown in Table 3.1.
The degree of crystallinity can be calculated from equation (3.1):
Xc =
ΔH ΔH u
…………………………………………………………(3.1)
where Xc is the degree of crystallinity, ΔH is the observed heat of fusion, and ΔHu is the crystalline PVC heat of fusion.
Values of ΔHu given in the literature33-39 range from 2.74 to 11.29 kJ/mole. Anagnostopoulos et al33 (1960) obtained the value of 2.74 kJ/mole from the melting point depression of PVC in diluents. A value of 11.29 kJ/mole was estimated by Kockott34 for a hypothetical completely syndiotactic PVC in 1964. A value of 3.28 kJ/mole was obtained by Nakajima et al35 in 1966. Juijin et al36 got a similar value of 3.14 kJ/mole in 1969. Colborne37 reported a value of 3.92 kJ/mole in 1970, obtained from a study of solubility of high-boiling diluents (i.e. plasticizers) in PVC. Guinlock38 estimated ΔHu as 4.93±0.38 kJ/mole by studying the highly crystalline PVC prepared in the presence of the chain-transfer agents (butyraldehyde and butyl mercaptan). Finally, a value of 2.74±0.26 kJ/mole was obtained by Patterson et al39 in 1982, which compares favorably to Anagnostopoules’ original value of ΔHu. Since the calculated crystallinity using the value
- 135 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ of 2.74 kJ/mole was consistent with results determined by other methods such as X-ray and density measurement, and widely cited by many other researchers19-25, 40-43, in this dissertation ΔHu=2.74 kJ/mole (or 43.89J/g) was chosen for calculation of the crystallinity for annealed PVCs. The calculated crystallinities, (Xc), using equation (3.1), are listed in Table 3.1; the average Xc due to annealing for all six PVC control samples is 5.5±0.3%. This value will be used as a reference to compare with the crystallinity data for other PVC samples, using the same annealing condition, discussed in the following sections.
Table 3.1. DSC difference curve data for PVCs prepared in bulk at 55 oC in absence of additives. Sample* Tg Tm Xc ΔH o o ( C) ( C) (J/g) (%) AIBN319 92 118 2.54 5.8 AIBN320
93
118
2.24
5.1
AIBN320
92
118
2.30
5.2
AIBN326
93
119
2.28
5.2
AIBN416
91
119
2.40
5.5
AIBN416
91
119
2.47
5.6
AIBN430
91
119
2.72
6.2
AIBN507
91
120
2.39
5.4
(average)
92±1
119±1
2.42±0.11
5.5±0.3
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
- 136 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.1.1 Crystallinity of PVCs prepared in the presence of 2-Coumaranone
Figure 3.4. DSC curves of PVCs prepared at 55 oC in the presence of the indicated mole fraction (in ppm) of 2-coumaranone (samples annealed at 100 oC for 30 minutes after heating to 230 oC).
Typical DSC curves of 10 PVC samples prepared in the presence of different amounts of 2-coumaranone (CMN) are shown in Figure 3.4. The number-average molecular weights of these 10 samples ranged from 39,400-49,100, with molecular weight distributions of 1.70-1.82. All 10 samples were annealed at 100 oC for 30 minutes, after heating to 230 oC. They show a glass transition temperature at approximately 92 oC, a small endotherm from 100 to 130 oC, and a broad endotherm from 130 to 230 oC.
The difference curves are shown in Figure 3.5. The broad, high temperature endotherm disappears; only the small endotherms with maxima at 118 oC are left. The calculated heats of fusion, ΔH, and the corresponding crystallinity, Xc, for these 10 samples are listed in Table 3.2. By comparing the values of Xc for the 10 samples in - 137 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.2 to those for the 8 controls in Table 3.1, using Wilcoxon’s sum of ranks test44, we find the probability of no significant difference between these two sets of measurements is bigger than 10%, which means they have no significant difference. The average Xc for the 10 samples is 5.5±0.6%, almost the same as that for the average control (5.5±0.3%). The effect of CMN concentration on the crystallinity for the resulting PVCs is shown in Figure 3.6. We can see that CMN has no or little effect on the resulting polymer crystallinity within the experimental error. As mentioned in Chapter 2, the molecular weights of the resulting polymers reached a maximum at about 0.01mole% of CMN and then dropped as CMN increasing. Since DSC technique does not provide any direct information about the microstructure of the resulting polymers, other techniques are required to investigate the additive effect on the PVC microstructure.
Figure 3.5. DSC difference curves for PVCs prepared in the presence of indicated mole fraction (in ppm) of 2-coumaranone.
- 138 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.2. DSC difference curve data for PVCs prepared at 55 oC in presence of CMN. Sample* [CMN] Tg Tm Xc ΔH o o (ppm) ( C) ( C) (J/g) (%) Control PVC 0 92±1 119±1 2.42±0.11 5.5±0.3 (avg value)
CMN00005a
49
92
119
2.50
5.7
CMN00005b
50
93
118
2.25
5.1
CMN0001b
95
92
118
2.66
6.1
CMN0001a
132
92
119
2.63
6.0
CMN00025b
234
92
118
2.20
5.0
CMN00025a
306
92
118
2.51
5.7
CMN0005d
466
92
118
2.63
6.0
CMN0005c
561
91
118
1.94
4.4
CMN001a
840
92
119
2.12
4.8
CMN001b
993
92
118
2.66
6.1
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
10.0
Crystallinity (%)
8.0 6.0 4.0
control
2.0 0.0 0.00
0.02
0.04
0.06
0.08
CMN concentration ( mole% )
0.10
Figure 3.6. Effect of CMN on the crystallinity for the resulting PVCs.
- 139 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.1.2 Crystallinity of PVCs prepared in the presence of 2,6-Dichloropyridine
Figure 3.7. DSC curves for PVCs prepared in the presence of indicated mole percent of 2,6-dichloropyridine (samples annealed at 100 oC for 30 minutes after heating to 230 oC).
DSC curves of seven PVC samples prepared in the presence of different amounts of 2,6-dichloropyridine (DCPY) are shown in Figure 3.7. The number-average molecular weights of these seven samples ranged from 42,400-46,900, with molecular weight distributions of 1.77-1.86. All 7 DSC curves show the same glass transition from 85 to 105 oC, followed by a small endotherm from 105 to 135 oC and a broader endotherm from 135 to 230 oC. The difference curves obtained by subtracting a quenched sample curve are shown in Figure 3.8. Only the small endotherm with a maximum at 119 oC is left. The corresponding heats of fusion, ΔH, and the crystallinity, Xc, calculated using equation (3.1), are listed in Table 3.3. By comparing the values of Xc for the 7 samples in Table 3.3 to those for the 8 controls in Table 3.1, using Wilcoxon’s sum of ranks test44, we find the probability of no-significant-difference between these two sets of - 140 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ measurements is less than 1%. Since no-significant-difference probability is very small, these two sets of measurements are considered different. The average Xc for the 7 samples is 6.1±0.5%, about 10% higher than that of the average control (5.5±0.3%). The effect of DCPY concentration on PVC crystallinity is shown in Figure 3.9. It seems DCPY increased the crystallinity for the resulting polymers. Due to the shortage of data, it is hard to tell whether or not this increase reached a plateau at higher DCPY concentrations. As we mentioned in Chapter 2, DCPY decreased the polymerization rate, but had no influence on the resulting PVC molecular weights. Because the DSC technique does not provide any direct information about the microstructure of the resulting polymers, other techniques are required to investigate the DCPY effect on the PVC microstructure.
Figure 3.8. DSC difference curves for PVCs prepared in presence of indicated mole percent of 2,6-dichloropyridine.
- 141 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.3. DSC data for PVCs prepared at 55 oC in the presence of DCPY. Sample* [DCPY] Tg Tm ΔH o (mole%) ( C) (oC) (J/g) Control PVC 0 92±1 119±1 2.42±0.11
Xc (%) 5.5±0.3
(avg value)
DCPY00025a
0.03
91
120
2.68
6.1
DCPY0005b
0.05
93
119
2.30
5.2
DCPY001b
0.10
91
119
2.75
6.3
DCPY0025a
0.25
91
119
2.72
6.2
DCPY005d
0.51
91
119
2.55
5.8
DCPY0075a
0.75
92
118
2.87
6.6
DCPY01h
1.00
91
119
2.95
6.7
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
12.0
Crystallinity (%)
10.0 8.0 6.0 4.0 control
2.0 0.0 0.0
0.2
0.4
0.6
0.8
1.0
DCPY concentrati on ( m ole% )
Figure 3.9. Effect of DCPY on the crystallinity for the resulting PVCs.
- 142 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.1.3 Crystallinity of PVCs prepared in the presence of 1,4-Dimethoxybenzene
DSC thermograms of 10 PVCs prepared in the presence of different amounts of 1,4-dimethoxybenzene (DMB) are shown in Figure 3.10. The number-average molecular weights of these 10 samples ranged from 43,000-47,900, with a molecular weight distribution of 1.77-1.84. All 10 samples were annealed at 100 oC for 30 minutes, after heating to 230 oC. All 10 curves show similar baseline changes between 85 and 100 oC due to the glass transition, followed by a small endotherm from 100 to 135 oC, and a broad endotherm from 135 to 230oC. The difference thermograms, after subtraction of the quenched sample curve, are shown in Figure 3.11. Only the small endotherms, with maxima at about 119 oC, are left. The calculated heats of fusion, ΔH, and the corresponding crystallinity, Xc, calculated using equation (3.1), are listed in Table 3.4. Again, by comparing the values of Xc for the 10 samples in Table 3.4 to those for the 8 controls in Table 3.1, using Wilcoxon’s sum of ranks test, we find the probability of no significant difference between these two sets of measurements is bigger than 10%, which means they have no significant difference. The average Xc for the 10 samples is 5.8±0.6%, very close to that for the average control (5.5±0.3%). The effect of DMB concentration on the crystallinity for the resulting PVCs is shown in Figure 3.12. We can see that, within the experimental error, DMB has no obvious effect on the resulting polymer crystallinity. As mentioned in Chapter 2, we found that DMB, under our investigation level, had no effect on the polymerization rate and the resulting polymer molecular weights either.
- 143 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Figure 3.10. DSC curves forPVCs prepared in the presence of indicated mole fraction (in ppm) of 1,4-dimethoxybenzene.
Figure 3.11. DSC difference curves for PVCs prepared in the presence of indicated mole fraction (in ppm) of 1,4-dimethoxybenzene.
- 144 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Table 3.4. DSC difference curve data for PVCs prepared at 55 oC in presence of DMB. Sample* [DMB] Tg Tm Xc ΔH o o (ppm) ( C) ( C) (J/g) (%) Control PVC 0 92±1 119±1 2.42±0.11 5.5±0.3 (avg value)
DMB00005b
43
94
120
2.88
6.6
DMB00005a
49
92
119
2.34
5.3
DMB0001a
99
92
118
2.12
4.8
DMB0001b
105
92
119
2.48
5.7
DMB0002c
231
92
119
2.90
6.6
DMB0002b
240
92
119
2.74
6.2
DMB0005b
460
92
118
2.32
5.3
DMB0005a
489
93
119
2.73
6.2
DMB001b
965
91
119
2.26
5.2
DMB001a
1008
91
118
2.59
5.9
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
12.0
Crystallinity (%)
10.0 8.0 6.0 4.0 control
2.0 0.0 0.00
0.02
0.04
0.06
0.08
0.10
DMB concentration ( mole% )
Figure 3.12. Effect of DMB on the crystallinity for the resulting PVCs. - 145 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.1.4 Crystallinity of PVCs prepared in the presence of Dimethyl Terephthalate The melting endotherms of the annealed PVCs prepared in the presence of different mole percents of dimethyl terephthalate (DMT) are shown in Figure 3.13. The number-average molecular weight of these samples ranges from 41,900-53,200, with a molecular weight distribution of 1.71-1.87. Figure 3.13 shows that all DSC curves have the same profile, a glass transition at ~92 oC, followed by a small endotherm from 105 to 135 oC, and a large, broad endotherm from 135 to 230 oC. The difference thermograms, after subtraction of the quenched sample curve, are shown in Figure 3.14. Only the small endotherms with a maximum at 119 oC are left. The corresponding heats of fusion, ΔH, and their crystallinity, Xc, calculated using equation (3.1), are listed in Table 3.5*. It can be seen that the crystallinities for most samples are higher than that of the control. By comparing the values of Xc for the 12 samples in Table 3.5 to those for the 8 controls in Table 3.1, using Wilcoxon’s sum of ranks test44, we find the probability of no significant difference between these two sets of measurements is less than 1%. Since this probability is very small, the no-significant-difference assumption is not proven, which means these two sets of measurements are most probably different.
____________________________________ *Two curves were omitted in Figures 3.13 and 3.14 to simplify the figure, but their calculated Xc’s were listed in Table 3.5. - 146 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Figure 3.13. DSC curves for PVCs prepared in the presence of indicated mole percent of dimethyl terephthalate. (samples annealed at 100 oC for 30 minutes after heating to 230 oC).
Figure 3.14. DSC difference curves for PVCs prepared in the presence of indicated mole percent of dimethyl terephthalate.
- 147 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.5. DSC data for PVCs prepared at 55 oC in the presence of DMT. Sample* [DMT] Tg Tm ΔH o (mole%) ( C) (oC) (J/g) Control PVC 0 92±1 119±1 2.42±0.11
Xc (%) 5.5±0.3
(avg value)
DMT0005a
0.05
93
120
2.77
6.3
DMT001c
0.09
92
119
2.22
5.1
DMT004a
0.26
90
119
2.92
6.7
DMT005a
0.29
91
120
2.83
6.5
DMT005d
0.49
91
119
2.84
6.5
DMT01b
0.59
92
119
2.86
6.5
DMT01a
0.74
92
118
2.75
6.3
DMT01d
0.84
94
119
2.71
6.2
DMT02b
1.46
90
119
2.59
5.9
DMT015a
1.47
92
118
2.73
6.2
DMT015b
1.66
92
119
2.73
6.2
DMT02c
1.89
94
119
2.75
6.3
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
- 148 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 12.0
Crystallinity (%)
10.0 8.0 6.0 4.0 control
2.0 0.0 0.0
0.4
0.8
1.2
1.6
2.0
DM T co ncentration ( mole% )
Figure 3.15. Effect of DMT on the crystallinity for the resulting PVCs.
The effect of DMT concentration on the resulting PVC crystallinity is shown in Figure 3.15. The resulting PVC crystallinity increased with the addition of DMT. The average value of Xc is 6.2±0.4%, about 13% higher than that of the control PVCs (Xc=5.5±0.3%). This increase is real, according to Wilcoxon’s test evaluation. As mentioned in Chapter 2, both the polymerization rate and the molecular weights of the resulting polymers were increased with the addition of DMT. It seems that DMT influenced the crystallinity of the resulting polymers. Since DSC technique does not provide any direct information about the microstructure of the resulting polymers, other techniques are required to investigate the additive effect on the PVC microstructure.
- 149 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.1.5 Crystallinity of PVCs prepared in the presence of γ-Butyrolactone
Figure 3.16. DSC curves for PVCs prepared in the presence of indicated mole percent of γ-butyrolactone (samples annealed at 100 oC for 30 minutes after heating to 230 oC).
Figure 3.16 shows the DSC curves of six PVC samples prepared in the presence of different amounts of γ-butyrolactone (GBL). The number-average molecular weights of these six samples range from 42,400-47,100, with molecular weight distributions of 1.76-1.87. The glass transition temperatures of these six annealed samples ranged from 92 to 94 oC, averaging 1 oC higher than that of the control. Figure 3.17 shows the difference thermograms. The endotherm peak is around 119 oC. The corresponding heats of fusion, ΔH, and the crystallinity, Xc, calculated using equation (3.1), are listed in Table 3.6. The crystallinity data in Tables 3.6 and 3.1 were evaluated by Wilcoxon’s sum of ranks test and the probability of no significant difference between these two Tables is only 0.2%. The probability is so small that a significant difference is ‘almost certain’ between these two set of measurements. - 150 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Figure 3.17. DSC difference curves ofor PVCs prepared in the presence of indicated mole percent of γ-butyrolactone.
Table 3.6. DSC data of PVCs prepared at 55 oC in the presence of GBL. [GBL] Tg Tm Sample* ΔH o (mole%) ( C) (oC) (J/g) Control PVC 0 92±1 119±1 2.42±0.11
Xc (%) 5.5±0.3
(avg value)
GBL0025b
0.27
94
121
2.69
6.1
GBL005b
0.46
93
120
3.14
7.2
GBL0075a
0.75
93
119
2.98
6.8
GBL01d
1.06
92
118
2.93
6.7
GBL02b
2.42
92
119
3.08
7.0
GBL04b
4.05
93
120
3.05
6.9
*All samples annealed at 100 oC for 30 minutes after heating to 230 oC.
- 151 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 12.0
Crystallinity (%)
10.0 8.0 6.0 4.0
control
2.0 0.0 0.0
1.0
2.0
3.0
4.0
GBL concentration ( mole% )
Figure 3.18. Effect of GBL on the crystallinity for the resulting PVCs.
The effect of GBL concentration on the crystallinity for the resulting PVCs is shown in Figure 3.18. The crystallinity for the resulting PVCs is higher than that of ‘control PVCs’ even at low concentrations of GBL (60 minutes are omitted. It can be seen in Table 3.28 that the values of the crystallization rate constant, K/(n+1), are relatively large, since they are calculated based on the relative crystallinity Xt. At t=60 minutes, Xt is defined to be greater than 0.97 for most cases. Assuming all seven PVC samples have the same crystallization mechanism, we got an overall Avrami plot shown in Figure 3.66 and overall Tg & Tm vs Ta plots - 206 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ shown in Figure 3.67. From Figure 3.66, exponent n+1= 1.0, crystallization rate constant K/(n+1)=1.1x10-3s-1, and half-time t1/2 =11 min were obtained.
3 Overall Crystallization Kinetics of PVC
Ln ( -Ln (1-Xt) )
2 1 0 -1
y = 1.0039x - 6.8102 R2 = 0.9193
-2 -3 4.5
5.5
6.5 Ln( t ) ( second )
8.5
7.5
Figure 3.66. Modified Avrami overall plot of Ln(-Ln(1-Xt)) against Ln( t ) for PVC/AIBN416, DMB0001a, DMB0001b, DMT01d, EC01c, SDMT01a, and SBP124 isothermally crystallized at 125 oC for up to 3600 seconds.
210 Tm
Tg
Tg & Tm ( o C )
180 y = 1.0692x + 14.121
150
R2 = 0.9938
120
90
60 80
90
100
110 120 130 140 150 Annealing Temperature, Ta ( o C )
160
170
180
Figure 3.67. The overall trend of Tg and Tm with annealing temperature (Ta) for PVC/AIBN416, DMB0001a, DMB0001b, DMT01d, EC01c, SDMT01a, and SBP124. - 207 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.4 Thermal Degradation Behavior of PVC Samples 1.2
127.16°C 220.49°C 100.0% 99.81%
302.71°C 77.58%
100
332.85°C 53.10% 465.22°C 22.40%
80 Weight (%)
1.0
60
558.54°C 0.2 7.63%
20 0
0.6 0.4
250.96°C 99.14%
40
0.8
Deriv. Weight (%/°C)
120
394.75°C 35.59% 0
100
200
300 400 Temperature (°C)
500
0.0 -0.2 600
Figure 3.68. TGA and negative derivative TGA curves for purified PVC/EC01c sample at the heating rate of 20 oC/min under nitrogen atmosphere.
In this section, the kinetics of thermal degradation of some PVC samples under non-isothermal conditions is investigated using thermogravimetric analysis (TGA). Figure 3.68 shows the exemplary thermogravimetric (TG) and differential (negative first derivative) thermogravimetric (DTG) curves (heating rate 20 oC/min) of approximately 10mg of PVC/EC01c using nitrogen as purge gas. The thermograms show two distinct degradation stages, with two characteristic DTG peaks, between ambient temperature and 600 oC. Stage one is between 230 oC and 400 oC and is divided into two steps with a degradation peak at 303 oC (77.6 wt%) and a distinguishable shoulder peak at 333 oC (53.1 wt%) as shown in DTG thermogram. About 40% of the initial weight is lost in the first step, and the remaining weight drops from 60 wt% to approximately 35.6 wt% in the second step. During stage two, from 400 oC to 600 oC, the weight of the sample is further reduced from 35.6 to 7.6w%. Degradation in stage one is due mainly the evolution of - 208 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ hydrogen chloride (H35Cl, H37Cl) (i.e. dehydrochlorination) with a small amount of benzene49. At the end of stage one, the PVC sample has lost most of its chlorine as HCl plus small amounts of aromatic compounds, mainly benzene, and the weight remaining in the thermobalance is 35.6w%. In stage two, the remaining 35.6w% residue is further decomposed into a mixture of hydrocarbons containing mostly aromatic compounds49; the char residue is about 7.6wt% at 600 oC. In this section, only ‘stage one’ is analyzed in detail, since we are focusing on how the PVC microstructure affects the degradation behavior. Eight samples were selected for TGA measurements at a single heating rate and two samples were chosen for measurements at multiple heating rates. Several kinetic methods were used to evaluate the thermogravimetric data with respect to kinetic parameters such as reaction order, activation energy, pre-exponential factor, etc.
At first, a brief introduction to the kinetic methods is given here. It is convenient to define the degree of degradation (α) as
α=
w0 − w w0 − w f
where w0, w, and wf are initial, instantaneous, and final weight of the sample during the degradation process. For non-isothermal degradation, the degradation rate (dα/dt) of a decomposing polymer can be generally described52-57 dα = k (T ) f (α ) …….……………………………………………..(3.7) dt
where f(α) is a general differential function of degradation. Depending on the degradation reaction mechanism f(α) presents different forms. Since α is a dimensionless - 209 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ quantity, dα/dt simply has unit of inverse time. k(T) is the temperature-dependent rate constant which often has an Arrhenius-type dependence:
⎛ − Ea ⎞ k (T ) = A exp⎜ ⎟ …………………………...………………….(3.8) ⎝ RT ⎠
where A is pre-exponential factor (having a unit of time-1), Ea, apparent activation energy, R, gas constant and T, absolute temperature. Using equation (3.8) and introducing the linear heating rate β = dT dt , equation (3.7) can be rewritten as:
dα dα ⎛ − Ea ⎞ ≡β = A exp⎜ ⎟ f (α ) ………………...…......................(3.9) dt dT ⎝ RT ⎠
The use of equation (3.9) assumes that the three parameters (Ea, A, f(α)) describe a chemical or physical change, irrespective of its complexity. Starting from equation (3.9) various kinetic evaluation methods have been developed50-58. These methods can be classified as differential, integral, and special methods. In this section, one differential and three integral methods are used. Some of these require recording the α=α(T) curves for several heating rates, so-called ‘isoconversional’ methods. The isoconversional methods allow evaluation of the activation energy without the knowledge of the explicit form of f(α). However, if the activation energy depends on α, the use of various isoconversional methods could lead to various activation energies for a given degree of degradation.
- 210 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The following methods are used to evaluate the degradation kinetic parameters of some selected PVC samples.
(1) Friedman method (FR method): this differential isoconversional method suggested by Friedman53 is based on the general degradation rate equation (3.9), written in its logarithmic form:
Ea ⎛ dα ⎞ ⎛ dα ⎞ Ln⎜ ⎟ ≡ Ln⎜ β ⎟ = Ln ( Af (α )) − …………………..(3.10) RT ⎝ dt ⎠ ⎝ dT ⎠
From TGA thermograms recorded at several heating rates, plotting Ln(dα/dt) against (1/T) at several given degrees of degradation (α) should generate a set of straight lines with slopes (-Ea/R). Thus, it is possible to obtain the activation energy over a range of degradations without knowing the explicit form of f(α).
(2) Flynn-Wall-Ozawa method54, 55 (FWO method): this isoconversional linear integral method is based on the following equation ⎛ AEa ⎞ Ea ⎟⎟ − 5.331 − 1.052 Lnβ = Ln⎜⎜ ……………………………(3.11) RT ⎝ Rg (α ) ⎠ α
here, g (α ) = ∫ 0
dα , β is the heating rate ( β = dT dt ). Equation (3.11) is obtained by f (α )
integrating equation (3.9) using a linear empirical approximation given by Doyle59, 60.
- 211 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The details are in references 53 and 54. For α=a specific value for each heating rate, the plot of Lnβ vs. 1/T, obtained from thermograms recorded at several heating rates, should be a straight line. The activation energy can be calculated from its slope.
(3). Kissinger-Akahira-Sunose method56, 57 (KAS method): this isoconversional integral method is based on the equation
⎛ AR ⎞ Ea ⎛ β ⎞ ⎟⎟ − Ln⎜ 2 ⎟ = Ln⎜⎜ …………………….………..........(3.12) Eg α ( ) ⎝T ⎠ ⎠ RT ⎝ α again, g (α ) = ∫ dα and β is the heating rate ( β = dT dt ). Equation (3.12) is obtained by f (α ) 0
integrating equation (3.9) using an approximation based on successive integration by parts and then retaining the first term in a rapidly converging series. Thus, for α=a specific value, a plot of Ln(β/T2) vs. 1/T, obtained from thermograms recorded at several heating rates, should be a straight line whose slope can be used to evaluate the activation energy. This expression is also given by the American Society for Testing and materials (ASTM)61. All the above three methods do not need an explicit form of f(α) to calculate the activation energy. If one needs to calculate the pre-exponential factor A and reaction order n, an analytical form of f(α), has to be provided.
(4) Coats-Redfern method52 (CR method): this is also an integral method and it involves the thermal degradation mechanism. In non-isothermal degradation of polymers, the mechanism is generally very complicated and various expressions of function f(α) - 212 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ have been proposed in the literature. The most used one is f (α ) = (1 − α )n , assuming that a simple nth-order kinetic relationship holds for the degradation function term. It should be pointed out that function f (α ) = (1 − α )n is just an overall description of the degradation. n is just an overall reaction order, and the real degradation mechanism may involve several elementary reactions. By introducing f (α ) = (1 − α )n equation (3.9) can be rearranged and integrated as following:
α
dα A ⎛ − Ea ⎞ ∫0 (1 − α )n = β ∫0 exp⎜⎝ RT ⎟⎠dT …………………………………….(3.13) T
The right-hand integral in equation (3.13) has no analytical solution. Therefore, approximations have to be applied to solve this equation. By making the substitution u=Ea/RT, Coats and Redfern52 used the following series expansion approximation:
∞
∫e
−u
−b
1−b −u
u du ≅ u e
∞
∑ n =0
u
(− 1)n (b)n u n+1
………………………….(3.14)
The integral of equation (3.13) can then be approximated as
1 − (1 − α ) 1− n
1− n
=
ART 2 β Ea
⎡ 2 RT ⎤ ⎛ − Ea ⎞ ⎢⎣1 − Ea ⎥⎦ exp⎜⎝ RT ⎟⎠ …………….(for n≠1; 3.15)
ART 2 or, − Ln(1 − α ) = βEa
⎡ 2 RT ⎤ ⎛ − Ea ⎞ ⎢⎣1 − Ea ⎥⎦ exp⎝⎜ RT ⎟⎠ ………....(for n=1; 3.16)
- 213 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ and the logarithmic expression:
⎡ AR ⎛ 2 RT ⎞⎤ Ea ⎛ g (α ) ⎞ ………………………….(3.17) Ln⎜ 2 ⎟ = Ln ⎢ ⎜1 − ⎟ − Ea ⎠⎥⎦ RT ⎝ T ⎠ ⎣ βEa ⎝ 1 − (1 − α ) where g (α ) = 1− n
1− n
, or, − Ln(1 − α ) .
Thus a plot of Ln(g(α)/T2) vs. 1/T should result in a straight line of slope –Ea/R for the correct value of n. CR method does not need to record the TGA thermograms at several heating rates, but picking the correct value of n is not straightforward. It is possible to use a computational approach to select a value of n which gives the best fit, but it doesn’t necessarily mean the reaction order is constant throughout the temperature range. It is just an overall, apparent reaction order.
- 214 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §3.3.4.1 Evaluation of Degradation Activation Energy of PVC samples by singleheating-rate TGA curve
100 Weight ( % )
-3.0
AIBN319 AIBN416 PVC40k PVC86k
80
-2.5 -2.0
60
-1.5
40
-1.0
20
-0.5
0
0.0 250
275
300
325
350
Temperature ( o C )
375
400
Deriv. Weight ( %/ o C )
120
425
Figure 3.69. TG and DTG curves for PVC/AIBN319, AIBN416, PVC40k, PVC86K at the heating rate of 20 oC/min under nitrogen atmosphere.
Figures 3.69 and 3.70 show the TG and DTG curves of 8 PVC samples at a heating rate of 20 oC/min using nitrogen as purging gas (the curves only show the region ranging from 250 to 425 oC). It can be seen from Figures 3.69 and 3.70 that all 8 samples start to degrade at around 262 oC, have a maximum degradation rate at approximately 330 oC, and reach a plateau at around 412 oC with a weight fraction of 35.6% remaining. In the region from 262 to 330 oC, the degradation of PVC samples mainly evolves hydrogen chloride49 and only this region is considered in this dissertation.
- 215 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 120
80
-2.5 Deriv. Weight ( %/ o C )
100 Weight ( % )
-3.0
CMN00025b DMB00005b DMB0001a DMT01b
-2.0
60
-1.5
40
-1.0
20
-0.5
0
0.0 250
275
300
325
350
Temperature ( o C )
375
400
425
Figure 3.70. TG and DTG curves for PVC/CMN00025b, DMB00005b, DMB0001a, DMT01b at the heating rate of 20 oC/min under nitrogen atmosphere.
From the curves, the degradation temperature at certain percentage of weight loss can be obtained, and the data are listed in Table 3.29. Here the degradation temperatures at 0.5%, 1%, 2%, 5% weight loss are defined as T0.5%, T1%, T2%, T5%, respectively. T0.5% can be treated as the effective starting point for sample degradation; the higher the T0.5% the more stable the sample is. From Table 3.29, it can be seen that the values of T0.5% for samples CMN00025b, AIBN319, and AIBN416 are very close to each other at 273-275 o
C, and the values of T0.5% for DMT01b, DMB00005b, DMB0001a samples are about 2-3
o
C higher; this difference may indicate that there is a microstructural difference between
them and others. It is believed that defect structures initiate the degradation of PVC and more defects in a sample may be associated with a lower T0.5% of decomposition. The dynamic TGA may not be a good method to investigate the effect of PVC defect differences on the thermal degradation of samples prepared in this thesis, since that - 216 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ difference is small and the reaction is drastic under the dynamic TGA conditions. Not surprisingly, the T0.5% of PVC86k is much higher than those of all other samples, because PVC86k, prepared at lower temperature, has higher molecular weight, higher tacticity than all other polymers and probably lower structural defects.
Table 3.29. The activation Energies (Ea) and the degradation temperatures of some PVC samples at different percentage of weight loss at a heating rate of 20 oC/min. Sample Mw T0.5% ( oC) T1% ( oC) T2% ( oC) T5% ( oC) *Ea (kJ/mole) PVC40k
68,700
271
280
291
306
158±5
CMN00025b
85,300
273
282
292
305
179±3
AIBN416
76,000
274
283
291
308
168±12
AIBN319
80,000
275
284
293
306
188±3
DMT01b
99,500
277
285
295
307
196±4
DMB00005b
85,500
278
287
298
312
171±4
DMB0001a
79,700
279
286
295
308
198±7
PVC86k
157,400
281
290
299
311
194±5
*calculated by applying equation (3.17), assuming n=0.
By applying equation (3.17), as shown in Figures 3.71 and 3.72, we can calculate the initial degradation activation energy (Ea) for all eight samples. The results are listed in Table 3.29. The activation energy values for these 8 PVC samples are in the range of 158-198 kJ/mole. Within the experimental error, they are the same. Even though different samples may have different defect concentrations that may cause the sample to decompose with different T0.5%, the decomposition mechanism remains the same under the same condition and the activation energies are also the same for these samples. Other - 217 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ techniques are needed to further investigate the effect of microstructure on the degradation behavior of PVC samples.
18.0 AIBN319
AIBN416
PVC40k
PVC86k
-Ln( α /T2 )
17.5 17.0 16.5 16.0 15.5 15.0 1.70
1.72
1.74
1.76 1.78 1000/T ( K-1 )
1.80
1.82
1.84
Figure 3.71. Application of Coats-Redfern method (eq. 3.17) for degradation of PVC/AIBN319, AIBN416, PVC40k, and PVC86k.
18.0 CMN00025b
DMB00005b
DMB0001a
DMT01b
-Ln( α /T2 )
17.5 17.0 16.5 16.0 15.5 15.0 1.70
1.72
1.74
1.76 1.78 1000/T ( K-1 )
1.80
1.82
1.84
Figure 3.72. Application of Coats-Redfern method (eq. 3.17) for degradation of PVC/CMN00025b, DMB00005b, DMB0001a, and DMT01b.
- 218 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The values of activation energy determined by TGA can be affected by many factors, such as sample weights, purge gas, heating rate, degree of degradation, as well as the kinetic model used62-69. It will be shown later in this section that the activation energy (Ea) values calculated using CR method are relatively higher than the values calculated using other methods, and these Ea values are also higher than those determined by isothermal degradation (please refer to Chapter 4 for dehydrochlorination measurements). Actually, when α≤0.1, g(α) is insensitive to reaction order, n. In Table 3.29, Ea is calculated by conveniently assuming n=0 (n=1, 2 give the same results).
§3.3.4.2 Evaluation of Degradation Activation Energy of PVC samples by FR, WFO and KAS methods Thermal degradations of PVC/CMN001a and EC01c were studied in this section. The thermogravimetric curves were recorded at multiple heating rates and the data were evaluated applying FR, WFO and KAS methods.
(A) PVC/CMN001a The thermogravimetric (TG) and negative derivative thermogravimetric (DTG) curves for PVC/CMN001a at heating rates of 5, 10 and 15 oC/min, respectively, are shown in Figure 3.73. It can be seen that with the increase of heating rate, the degradation starts at higher temperature. The reason for this shift is that as the heating rate increases, the time needed to achieve a certain temperature decreases. Since the degradation is kinetically controlled, at a higher heating rate, there is less degradation at a specific temperature. This heating-rate dependence is also indicated in the DTG thermograms. - 219 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ That is the peaks of the DTG thermograms shift towards higher temperatures as the heating rate increases. The peak temperatures are 272, 285 and 293 oC for heating rates of 5, 10, and 15 oC/min, respectively. The difference in heights (rates of weight loss, dw/dt) in the DTG thermograms shown in Figure 3.73 is due to the different heating rates. A plot of the weight change with respect to temperature (dw/dT) against temperature (T) would show DTG peaks of approximately the same height and shape (please refer DTG curves as shown in Figure 3.79).
30
120 10C/min
15C/min 25
80
20
60
15
40
10
20
5
0
0
Weight (% )
100
200
240
280
320
Temperature ( o C )
360
Deriv. Weight (%/min)
5C/min
400
Figure 3.73. Thermogravimetric (TG) and negative first derivative thermogravimetric (DTG) curves for PVC/CMN001a at heating rates of 5, 10 and 15 oC/min, respectively.
To apply the FR, WFO and KAS methods, two sets of data are needed from TG and DTG curves recorded at different heating rates. First, a set of temperatures at given degree of degradation (α) is needed, and second, a set of degradation rates (dα/dt) at a given α is needed from the DTG curves. These two sets of data are listed in Tables 3.30 and 3.31, with degrees of degradation ranging from 0.16% to 47.27% (i.e. loss of - 220 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ approximately half of HCl content of PVC). Within this region, the degradation of PVC is mainly by evolution of HCl (i.e. dehydrochlorination) and the evolution of aromatic compounds can be ignored.
2.8 2.5
-Ln( β )
2.2 1.9 1.6 1.3 1.0 30%
20%
15%
10%
5.0%
2.0%
1.0%
0.5%
0.2%
0.1%
0.7 1.70
1.75
1.80
1.85
1.90
1.95
1000/T ( K-1 )
2.00
2.05
2.10
Figure 3.74. The plots of natural logarithm of heating rate(β) versus the reciprocal absolute temperature for PVC/CMN001a degradation at different weight loss(%).
By applying equations (3.10), (3.11), and (3.12), the values of activation energy of PVC/CMN001a at different degrees of degradation (α) were obtained and the results are listed in Table 3.32. Figure 3.74 shows the plots obtained applying FWO method. The plots for the other methods are omitted. It can be seen from Table 3.32 that the calculated degradation activation energies (Ea) for CMN001a are very close to each other when α > 5%. The plot of calculated Ea as a function of α is shown in Figure 3.75. It can be seen that the Ea for CMN001c remains constant between degradation 7.88 to 47.27%, but drops at lower degrees of degradation (α). It is possible that the degradation mechanism is different in the very early degradation stage. Or the so-called error-propagation effect - 221 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ might be important when the logarithmic method is used to calculate activation energies. The uncertainty is generally larger in the early degradation stage than in the middle degradation stage. As can be seen from Figure 3.75, the Ea for PVC/EC01c (described in the next section) is higher in the early stage of degradation and significantly lower above 15% degradation than that of CMN001a.
Table 3.30. The degradation temperature of PVC/CMN001a under different degrees of degradation (α) at heating rates of 5 , 10 and 15 oC/min, respectively. Degree of Degrad. Degradation temperature at the heating rate β ( oC) α (%)
β=5 oC/min
β=10 oC/min
β=15 oC/min
0.16
210
221
230
0.32
215
227
234
0.79
225
237
243
1.58
231
244
250
3.15
239
251
257
7.88
250
262
268
15.76
259
271
277
23.64
264
276
283
31.51
268
280
288
47.27
275
288
295
- 222 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.31. The derivative weight (dα/dt) of PVC/CMN001a at different degrees of degradation (α) at heating rates of 5 , 10 and 15 oC/min, respectively. Degree of Degrad. Derivative weight(dα/dt) at the heating rate β (%/min) α (%)
β=5 oC/min
β=10 oC/min
β=15 oC/min
0.16
0.13
0.25
0.35
0.32
0.10
0.23
0.32
0.79
0.22
0.49
0.70
1.58
0.42
0.94
1.53
3.15
1.04
1.91
2.71
7.88
1.98
4.21
5.79
15.76
4.08
7.82
11.01
23.64
5.51
10.77
15.06
31.51
6.42
12.56
17.89
47.27
6.78
13.40
19.31
- 223 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.32. Apparent Activation Energies (Ea) for CMN001a calculated by three isoconversional methods*. Activation Energy, Ea (kJ/mole) Degradation, α (%)
FR method
FWO method
KAS method
0.16
102±7
107±1
105±6
0.32
125±9
115±1
113±1
0.79
131±2
118±5
116±5
1.58
153±9
124±7
122±7
3.15
118±6
128±7
127±7
7.88
142±7
136±4
134±4
15.76
132±1
138±3
136±3
23.64
133±4
137±2
135±2
31.51
134±4
135±1
133±1
47.27
135±4
134±1
132±1
*FR method: Friedman method; FWO method: Flynn-Wall-Ozawa method; KAS method: Kissinger-Akahira-Sunose method.
240 CMN001a
EC01c
Ea (kJ/mole)
200 160 120 80 40 0 0
10
20 30 Degree of Conversion ( % )
40
50
Figure3.75. Apparent activation energies (Ea) for the degradation of PVC/CMN001a and EC01c as a function of degree of degradation (α). - 224 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ (B) PVC/EC01c Thermogravimetric (TG) and negative derivative thermogravimetric (DTG) curves for PVC/EC01c sample at heating rates of 2, 5, 10 and 15 oC/min, respectively, are shown in Figure 3.76. It can be seen that with the increase of heating rate, the degradation starts at higher temperature. This heating-rate dependence is also indicated in the DTG thermograms. The DTG thermogram peaks shift towards higher temperatures as the heating rate increases. The peak temperatures corresponding to the maximum degradation rates((dα/dt)max) are 256, 276, 289 and 301 oC for heating rates of 2, 5, 15 and 20 oC/min, respectively. The difference in heights (rates of weight loss, dw/dt) in the DTG thermograms, shown in Figure 3.76, is due to the different heating rates. A plot of the weight change with respect to temperature (dw/dT) against T would show DTG peaks of approximately the same height and shape (refers DTG curves shown in Figure 3.80).
120
30 ––––––– 15C/min –––– 10C/min ––––– – 5C/min 25 ––– – – 2C/min
100
Weight (%)
15
275.72°C 75.56%
40
0 200
20
288.73°C 76.28%
60
20
Deriv. Weight (%/min)
300.79°C 77.95%
80
10
256.04°C 76.09%
240
5
280 320 Temperature (°C)
360
0 400
Figure 3.76. Thermogravimetric (TG) and negative first derivative thermogravimetric (DTG) curves of PVC/EC01c at heating rates of 2, 5, 10, 15 oC/min.
- 225 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The degradation temperatures (T) and degradation rates (dα/dt) of PVC/EC01c at a given degree of degradation (α), at heating rates (β) of 2, 5, 10 and 15 oC/min, respectively, are listed in Tables 3.33 and 3.34. The two sets of data were obtained over a degradation range from 0.16 to 47.27%, in which the degradation of PVC is mainly dehydrochlorination. The evolution of aromatic compounds is small and can be ignored.
By applying equations (3.10), (3.11), and (3.12), the activation energy values of EC01c sample at different degrees of degradation (α) are obtained and the results are listed in Table 3.35. Plots applying FWO method are shown in Figure 3.77. The plots for the other methods are omitted. It can be seen from Table 3.35 that the calculated values of activation energy (Ea) of EC01c degradation are very high at lower degradation but drop rapidly as degradation increases. The plot of calculated Ea as a function of degree of degradation (α) is shown in Figure 3.75. It is likely due to experimental uncertainty, rather than to mechanism changes from the very early degradation stage. The uncertainty is generally larger in the early degradation stage than in the middle degradation stage. It can also be seen from Table 3.35 that the values of Ea calculated by the FWO & KAS methods are close to each other, but the values of Ea calculated by FR method do not agree with the others.
- 226 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.33. The degradation temperature of PVC/EC01c with different degrees of degradation (α) at heating rates of 2, 5 , 10, 15 oC/min. Degr. of degrad. Degradation temperature at the heating rate β ( oC) α (%)
β=2 oC/min
β=5 oC/min
β=10 oC/min
β=15 oC/min
0.16
205
215
230
227
0.32
211
223
228
234
0.79
222
233
240
248
1.58
229
241
248
258
3.15
236
249
256
267
7.88
244
259
268
280
15.76
249
266
277
289
23.64
252
270
282
295
31.51
255
273
286
299
47.27
258
279
293
308
3.8 3.4
-Ln( β )
3.0 2.6 2.2 1.8 1.4 1.0 0.6 1.70
30%
20%
1.75
15%
1.80
10%
5.0%
1.85
2.0%
1.90
1.0%
1.95
1000/T ( K-1 )
0.5%
2.00
0.2%
0.1%
2.05
2.10
Figure 3.77. The plots of natural logarithm of heating rate(β) versus the reciprocal absolute temperature for PVC/EC01c degradation at selected weight losses (%). - 227 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.34. The derivative weight (dα/dt) of PVC/EC01c at different degrees of degradation (α), at heating rates of 2, 5 , 10, 15 oC/min. Degr. of degrad. Derivative weight(dα/dt) at the heating rate β ( %/min) α (%)
β=2 oC/min
β=5 oC/min
β=10 oC/min
β=15 oC/min
0.16
0.03
0.06
0.09
0.15
0.32
0.04
0.11
0.19
0.30
0.79
0.12
0.25
0.44
0.68
1.58
0.22
0.49
0.81
1.25
3.15
0.42
0.84
1.33
1.93
7.88
1.25
2.47
3.86
5.51
15.76
2.80
5.19
8.18
11.42
23.64
4.09
7.30
11.15
15.89
31.51
4.92
8.64
13.25
18.17
47.27
4.79
8.45
12.45
15.60
- 228 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.35. Apparent Activation Energies (Ea) for PVC/EC01c calculated by three isoconversional methods*. Activation Energy, Ea (kJ/mole) Degradation, α (%)
FR method
FWO method
KAS method
0.16
151±3
179±14
177±7
0.32
169±5
169±11
170±11
0.79
143±9
161±15
162±16
1.58
137±8
152±14
151±15
3.15
118±8
148±15
148±16
7.88
100±6
131±11
128±12
15.76
88±4
120±8
119±9
23.64
80±2
115±7
112±8
31.51
75±3
110±7
107±7
47.27
63±4
102±5
98±6
*FR method: Friedman method; FWO method: Flynn-Wall-Ozawa method; KAS method: Kissinger-Akahira-Sunose method.
§3.3.4.3 Evaluation of Degradation Kinetic Parameters (Ea, A, n) of PVC samples by Coats-Redfern method
In the previous section, the activation energies (Ea) for CMN001a and EC01c degradation were evaluated by FR, FWO, and KAS methods. By applying FR, FWO, and KAS methods, we do not need to know the explicit form of f(α) or g(α) to evaluate the activation energy Ea. If we assume that a simple n-th order kinetic model holds for the degradation region in which we are interested, the form f(α) can be simply defined as
- 229 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ f (α ) = (1 − α ) . Here n is the overall reaction order. From equation (3.10) (Friedman n
method), we plot Ln(dα/dt) against 1/T to get a straight line, the slope is –Ea/R and the intercept is Ln(Af(α)); by substituting f (α ) = (1 − α ) , the intercept Ln(Af(α)) can be n
written as Ln ( Af (α )) = Ln ( A) + nLn (1 − α ) ………………………………….(3.18)
Thus, the plotting of Ln(Af(α)) against Ln(1-α) should be a straight line; the reaction order n and the pre-exponential factor A can be determined from its slope and intercept, respectively. The plots of Ln(Af(α)) as a function of Ln(1-α) for PVC CMN001a and EC01c are shown in Figure 3.78.
50 CMN001a
EC01c
Ln(Af (α ))
40 30 20 10 0 -1.0
-0.8
-0.6
Ln(1-α )
-0.4
-0.2
0.0
Figure 3.78. Plots of Ln(Af(α)) as a function of Ln(1-α) for PVC/CMN001a and EC01c.
We can see from Figure 3.78 that the data points for both samples are very scattered when α ≤0.1. The data for both samples have a linear trend only when α>0.1, which gives A=1.2x1012 min-1, n=1.68 for CMN001a sample, and A=5.9x107 min-1, - 230 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ n=9.27 for EC01c. It seems that due to the error-propagation effect, the linear Friedman method not appropriate to evaluate the pre-exponential factor A and reaction order n. FWO and KAS methods also have the similar error- propagation effects, due to the linear regression, and fail to get meaningful (A, n) pairs. So other methods are needed to evaluate the (A, n) pairs.
10K/min
0.80 Norm arized Weight
-0.04
5K/min
-0.03
15K/min simulation
0.60
-0.02
0.40
-0.02
0.20
-0.01
0.00
0.00 475
525
575
625
Derivati ve Wei ght (1/ K)
1.00
675
Temperature (K)
Figure 3.79. Experimental and simulated TG and DTG curves for PVC/CMN001a thermal degradation at heating rates of 5, 10, 15 K/min. (simulated curves were obtained by fitting experimental data using equation (3.15) by least-squares regression: A=1.9x1011min-1, Ea=126kJ/mole, n=1.71)
If we assume the activation energy Ea and pre-exponential factor A are independent of the degree of degradation (α) and the heating rate β, we can use the Coats-Redfern method, i.e. equation (3.15), to evaluate the (Ea, A, n) triplet by non-linear least-squares regression. Here values of Ea, A, and n determined by Friedman method can be used as initial values for the least-squares regression. Thus, the least-squares regressions apply equation (3.15) to α=α(T) curves recorded at multiple heating rates for - 231 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ both CMN001a and EC01c samples. The results are shown in Figures 3.79 and 3.80. A triplet of A=1.9x1011min-1, Ea=126kJ/mole, n=1.71 was obtained for the CMN001a degradation, and A=1.8x1011min-1, Ea=127kJ/mole, n=1.64 for EC01c degradation. It can be seen that the results of CMN001a sample and Ec01c sample are very close to each other, but the simulation results are only close to the experimental data at very low degradations. Our results as well as some other authors’ are listed in Table 3.36. As can been seen, our results are in good agreement with the literature ones, but neither can fit the experimental data.
1.2
Normalized weight
1.0 0.8 0.6
2K/min
0.4
5K/min
0.2
10K/min 15K/min simulation
0.0 450
500
550 Temperature (K)
600
650
Figure 3.80. Experimental and simulated TG curves for PVC/EC01c thermal degradation at heating rates of 2, 5, 10, 15 K/min. (simulated curves were obtained by fitting experimental data using equation (3.15) by least-squares regression: A=1.8x1011min-1, Ea=127 kJ/mole, n=1.64)
- 232 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 3.36. Comparison of the kinetic parameters of PVC degradation obtained by different authors. Author Ea(kJ/mole) A(min-1) Method* f(α) This work
126--127
1.8--1.9x1011
(1-α)1.64--1.71
CR
Bockhorn62
140
1.0x1013
(1-α)1.50
CR
Oh64
159
--
(1-α)1.25
NM
Budrugeac66
103
2.1x109
(1-α)1.80
IKP
Budrugeac69
104
6.4x109
α0.63(1-α)1.55
IKP
Hirschler70
164
5.8x1014
(1-α)1.54
FR
Kunmann71
143
3.2x1012
(1-α)1..54
CR
Jimenez72
179
--
α0.2(1-α)2.0
FR
Jimenez73
145
1.6x1011
(1-α)1.73
FR
Slapak74
125
2x1012
(1-α)2.80
NM
*CR: Coats-Redfern Method; FR: Friedman method; IKP: Invariant Kinetic Parameters69 method; NM: Numerical Model using 4th Runge-Kutta algorithm74.
The simulated curves in Figures 3.79 & 3.80 were obtained by applying equation (3.15) to the experimental data using least-square regression. It can be seen that the simulated curves are only close to the real thermogravimetric curves at very low degradation ( K a , and K a t ≈ 0 , which happens if K a is very small or the dehydro-chlorination is in its initial stage (i.e. t very small), HCl s will reach a steady-state (st) value:
(HCl s )st
≈
K a …………………………………………..…….……..(4.9) Kb
Then,
⎛ dHCle ⎞ ⎜ ⎟ ≈ K a ………………………………………….…...........(4.10) dt ⎝ ⎠ st
How fast dHCl e dt reaches its steady-state ( K a ) depends on K b . From equation (4.7), we see that when t = 1 K b ,
dHCl e dt ≈ 0.63 K a . When t = 3 K b ,
dHCl e dt ≈ 0.95 K a and when t = 5 K b , dHCl e dt ≈ 0.99 K a . This assumes that K b >> K a and K a t ≈ 0 holds in this phase of dehydrochlorination. After t ≥ 5 K b , the HCle vs. t curve is linear and the slope, i.e., the dehydrochlorination rate, equals K a .
In the literature22,
63-65
, people often treated the dehydrochlorination curve, i.e.
HCle vs. t as a 2-part curve. The first part was non-linear and the second part was “linear”. The initial non-linear part was often called as induction period22. The slope of the “linear” part was reported as the rate of dehydrochlorination for that PVC and the reaction was said to be zero order. Actually setting the starting point for the “linear” - 255 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ section is arbitrary, especially for dehydrochlorination at low temperatures. Usually the “linear” portion included part of gradient curve and gave a dehydrochlorination rate lower than that we determined. From the two-parameter model, as mentioned earlier, the dehydrochlorination rate reached a constant value when t ≥ 5 K b . Actually, from equation (4.8) it is not necessary to run the dehydrochlorination reaction to its linear period. The fitting kinetic parameter, Ka, from the non-linear region can always be considered the dehydrochlorination rate. Also from the 2-parameter model, it is seen that the dehydrochlorination is a complex of first or zero order HCl releasing reaction plus a first order HCl diffusion reaction. The zero order reaction is only true for the initial degradation stage, since other factors may become important as degradation proceeds.
In the following sections, the 2-parameter model is used to evaluate the dehydrochlorination behavior of some representative PVCs prepared in the presence or absence of organic additives, in bulk and in suspension.
- 256 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.3.2 Dehydrochlorination of PVCs prepared in Bulk without Additives 0.9
AIBN319 AIBN416
0.8
AIBN320 AIBN430
AIBN326 AIBN507
HCle ( mole% )
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
10
20
30
40 50 60 Time ( minute )
70
80
90
100
Figure 4.3 Kinetic curves for the dehydrochlorination of control PVCs at 190 oC under fast nitrogen flow.
Typical kinetic curves for the dehydrochlorination of six PVC samples at 190 oC under fast nitrogen flow are shown in Figure 4.3. These six samples were prepared in bulk initiated by about 6.4x10-3 mole/L of 2,2’-Azobis-isobutyronitrile (AIBN) and named as PVC/AIBN319, AIBN320, AIBN 326, AIBN416, AIBN430 and AIBN507. These were called control samples. The number average molecular weights of these six samples were 41,200 to 44,500, and the molecular weight distributions were 1.71 to 1.90. Using least-squares regression, we fit our experimental data with equation (4.8) to get values of Ka and Kb, Table 4.1. It can be seen that the calculated Ka’s are very close to each other, but the calculated Kb’s are somewhat scattered. Since Kb was defined as the rate constant for HCls diffusing through the polymer surface, it is affected by factors such as sample size, particle morphology, etc. Since our purified samples did not have uniform - 257 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ shapes and morphology, the calculated values of Kb are expected to vary. On the other hand, Ka was defined the dehydrochlorination rate constant, which was proportional to the structural defect concentration in the PVC samples. The dehydrochlorination reaction became apparently zero order after a certain time period, the “induction” time, and the dehydrochlorination rate is then equal to Ka. Here, we mainly focus on Ka values. Since these six samples are bulk polymerization runs using the same recipe, called bulk controls, these Ka’s are very close to each other. The averaged Ka is 1.42±0.06x10-6s-1. In the following study of dehydrochlorination of PVCs prepared in the presence of additives, Ka=1.42±0.06x10-6s-1 is used as a reference to evaluate the additive effect on the dehydrochlorination of resulting PVCs.
Table 4.1. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in bulk using 2,2’-azobisisobutyronitrile (AIBN) as initiatora). Sample [AIBN]x103 Kb x 103 Induction time (ti)b) Ka x 106 (mole/L) (s-1) Calc (min) Exp (min) (s-1) AIBN319 6.1 1.29 13 14 1.35 AIBN320
6.4
1.52
11
13
1.32
AIBN326
6.8
1.03
16
15
1.46
AIBN416
7.6
2.00
8
10
1.51
AIBN430
6.2
1.17
14
12
1.45
AIBN507
6.8
0.77
22
16
1.41
(average)
6.6±0.4
1.30±0.31
14±3
13±2
1.42±0.06
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle.
- 258 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.3.3 Dehydrochlorination of PVCs prepared in Bulk in the presence of Additives In this section, a series of dehydrochlorination measurements that were carried out for PVCs prepared in bulk polymerization in the presence of additives are discussed. The additives were 2-coumaranone (CMN), 2,6-dichloropyridine (DCPY), 1,4-dimethoxybenzene (DMB), dimethyl terephthalate (DMT), γ-butyrolactone (GBL) and trimethyl1,3,5-benzenetricarboxylate (TMB). The polymerization was described in detail in Chapter 2 and the PVC samples are named as designated in Chapter 2. The dehydrochlorination temperature was set as 190 oC with rapid removal of HCl for all samples. The dehydrochlorination kinetic parameters, Ka and Kb, were determined using equation (4.8) by least-squares regression.
§4.3.3.1 Dehydrochlorination of PVCs prepared in the presence of 2-Coumaranone The dehydrochlorination of eleven PVC samples prepared in the presence of 2coumaranone (CMN) was carried out in a 5-unit DEHYDRO apparatus described in section 4.2.2. The dehydrochlorination temperature was set as 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes; the HCl loss was less than 1mole% (relative to the theoretical HCl content in the polymer) for all samples. The dehydrochlorination curves are omitted but the calculated kinetic parameters, Ka and Kb, using equation (4.8), are listed in Table 4.2. The plot of kinetic parameter Ka as a function of CMN concentration is shown in Figure 4.4. We can see that the values of Ka for 8 out of 11 samples were lower than that of the control PVC, Table 4.2 and Figure 4.4. By comparing the values of Ka in Table 4.2 to those in Table - 259 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 4.1 using Sudent’s’ t test62, we find the probability of no-significant-difference between these two sets of Ka’s is less than 5%, which means the difference is ‘probably’ significant. The average decrease for Ka is about 10%, which indicates that defect concentration decreased about 10%. The thermal stability increased about 10% over that of the control for those PVCs.
Table 4.2. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 2-coumaranone (CMN)a). Sample [CMN] Kb x 103 Induction time (tin)b) Ka x 106 (mole%) (s-1) Calc (min) Exp (min) (s-1) Controlc) 0.000 1.30 14 13 1.42 CMN00005a
0.005
1.30
13
14
1.50
CMN00005b
0.005
1.71
10
9
1.28
CMN0001b
0.010
1.02
16
13
1.30
CMN0001a
0.013
1.75
10
8
1.36
CMN00025b
0.023
1.17
14
12
1.22
CMN00025a
0.031
1.20
14
13
1.33
CMN0005b
0.042
2.34
7
7
1.27
CMN0005d
0.047
1.08
15
13
1.26
CMN0005c
0.056
1.50
11
10
1.51
CMN001a
0.084
1.36
12
10
1.51
CMN001b
0.099
1.70
10
9
1.35
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results.
- 260 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
2.80
Ka x 10 6 ( s-1 )
2.40 2.00 1.60 1.20 0.80
control
0.40 0.00 0.00
0.02
0.04 0.06 [CMN] (mole%)
0.08
0.10
Figure 4.4. Effect of CMN on Ka for the dehydrochlorination of the resulting PVCs.
As mentioned in Chapter 2, the molecular weights for PVCs prepared in the presence of CMN increased as CMN increased and reached a maximum (10% higher than that of the control) when the CMN concentration was about 0.01 mole%. They decreased at higher CMN concentrations. In Chapter 3, we found that the crystallinity for the resulting PVCs was about that for the control within the experimental error. Here the thermal stability increased for PVCs prepared in the presence of CMN. The increased stability is most likely due to decreased structural defect concentration, which initiates the dehydrochlorination reaction in the polymer chains. The structural defects of the resulting PVCs are discussed in detail in section 4.3.6.
- 261 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.3.3.2 Dehydrochlorination of PVCs prepared in the presence of 2,6-Dichloropyridine
The dehydrochlorination of six PVC samples prepared in the presence of 2,6dichloropyridine (DCPY) was carried out in a 5-unit DEHYDRO apparatus described in section 4.2.2. The dehydrochlorination temperature was set as 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes; the HCl loss was less than 1mole% (relative to the theoretical HCl content in the polymer) for most samples. The dehydrochlorination curves are shown in Figure 4.5. The kinetic parameters, Ka and Kb, were determined using equation (4.8) by least-squares regression, and the results are listed in Table 4.3. The plot of kinetic parameter Ka as a function of DCPY concentration is shown in Figure 4.6. As can be seen, the values of Ka increased drastically with the increase of DCPY. As we mentioned in chapter 2, DCPY decreased the initial polymerization rate, but, had no influence on the molecular weights of the resulting PVCs within the experimental error. An explanation for the effect of DCPY on the resulting polymers is following. Since DCPY is a basic compound, it may attack resulting polymer Cl groups to form double bonds or conjugated double bonds in the polymer backbone. They promote allylic dehydrochlorination at elevated temperatures as HCl is lost rapidly than found in the control. Other techniques such as Nuclear Magnetic Resonance (NMR) are needed to identify and quantify these double bonds, so-called structural defects, in the polymer.
- 262 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 1.4 1.2
DCPY0005b
DCPY001b
DCPY005b
DCPY005d
DCPY0075a
DCPY01h
HCle (mole%)
1.0 0.8 0.6 0.4 0.2 dehydrochlorination temperature: 190 oC 0.0 0
10
20
30
40 50 60 Time ( minute )
70
80
90
100
Figure 4.5 Kinetic dehydrochlorination curves for PVCs prepared in the presence of 2,6-dichloropyridine (DCPY).
Table 4.3. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 2,6-dichloropyridine (DCPY)a). Sample [DCPY] Kb x 103 Induction time (tin)b) Ka x 106 (mole%) (s-1) Calc (min) Exp (min) (s-1) c) Control 0.00 1.30 14 13 1.42 DCPY0005b
0.05
0.92
18
14
1.41
DCPY001b
0.10
1.40
12
11
1.53
DCPY005b
0.41
1.28
13
11
1.60
DCPY005d
0.51
0.94
18
14
1.96
DCPY0075a
0.75
1.00
17
13
2.02
DCPY01h
0.94
0.63
26
16
2.94
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results. - 263 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 3.60
Ka x 10 6 ( s-1 )
3.00 2.40 1.80 1.20 0.60 0.00 0.00
control
0.20
0.40 0.60 [DCPY] ( mole% )
0.80
1.00
Figure 4.6. Effect of DCPY on Ka for the dehydrochlorination of the resulting PVCs.
§4.3.3.3 Dehydrochlorination of PVCs prepared in the presence of 1,4-Dimethoxybenzene The dehydrochlorination of ten PVC samples prepared in the presence of 1,4dimethoxy-benzene (DMB) was carried out in a 5-unit DEHYDRO apparatus described in section 4.2.2. The dehydrochlorination temperature was set as 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes and HCl loss was less than 1mole% (relative to the theoretical HCl content in the polymer) for all samples. The dehydrochlorination curves are omitted. The kinetic parameters, Ka and Kb, were determined using equation (4.8) by least-squares regression, and the results are listed in Table 4.4. The effect of DMB concentration on Ka for the dehydrochlorination of the resulting PVCs is shown in Figure 4.7. We can see that the values of Ka remain at the same level as that of the control within the experimental error. As we mentioned in Chapter 2, DMB, at the concentration level we investigated, had no - 264 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ or little effect on the initial polymerization rate and the resulting polymer molecular weights. In Chapter 3, we also found that DMB had no effect on the crystallinity for the resulting PVCs.
Table 4.4. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of 1,4-dimethoxybenzene (DMB)a). Sample [DMB] Kb x 103 Induction time (tin)b) Ka x 106 -1 (mole%) (s ) Calc (min) Exp (min) (s-1) Controlc) 0.000 1.30 14 13 1.42 DMB00005b
0.004
1.16
14
12
1.38
DMB00005a
0.005
0.86
19
17
1.53
DMB0001a
0.010
0.81
21
15
1.33
DMB0001b
0.011
1.39
12
11
1.26
DMB0002c
0.023
0.50
33
20
1.47
DMB0002b
0.024
1.13
15
13
1.34
DMB0005b
0.046
1.14
15
12
1.37
DMB0005a
0.049
1.22
14
11
1.35
DMB0008a
0.072
0.81
21
15
1.68
DMB001b
0.097
0.64
26
17
1.56
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results.
- 265 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2.80
Ka x 10 6 ( s-1 )
2.40 2.00 1.60 1.20 0.80
control
0.40 0.00 0.00
0.02
0.04 0.06 [DMB] (mole%)
0.08
0.10
Figure 4.7. Effect of DMB on Ka for the dehydrochlorination of the resulting PVCs.
§4.3.3.4 Dehydrochlorination of PVCs prepared in the presence of Dimethyl terephthalate The dehydrochlorination of eleven PVC samples prepared in the presence of dimethyl terephthalate (DMT) was carried out in a 5-unit DEHYDRO apparatus described in section 4.2.2. The dehydrochlorination temperature was set as 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes and the HCl loss was less than 1 mole% (relative to the theoretical HCl contents in the polymer) for all samples. The dehydrochlorination curves are omitted. The kinetic parameters, Ka and Kb, were determined by least-squares regression using equation (4.8), and the results are listed in Table 4.5. The effect of DMT on Ka for the dehydrochlorination of the resulting PVCs is shown in Figure 4.8. It can be seen that for samples prepared with DMT concentration higher than 0.5 mole%, 5 out of 7 samples have a decreased Ka (about 10% lower than that of the control). By comparing the values - 266 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ of Ka of these 7 samples to those of the control samples in Table 4.1 using Sudent’s’ t test62, we find that the probability of no-significant-difference between these two sets of Ka’s is less than 5%, which means the difference is ‘probably’ significant. As we mentioned in Chapter 2, DMT increased both the polymerization rate and the resulting polymer molecular weights.
Table 4.5. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of dimethyl terephthalate (DMT)a). Sample [DMT] Kb x 103 Induction time (ti)b) Ka x 106 (mole%) (s-1) Calc (min) Exp (min) (s-1) Controlc) 0.00 1.30 13 14 1.42 DMT0005A
0.05
1.25
13
14
1.47
DMT001c
0.09
0.87
19
15
1.43
DMT004a
0.26
1.05
16
14
1.50
DMT005a
0.29
1.15
14
15
1.31
DMT01b
0.59
1.06
16
16
1.10
DMT01a
0.74
1.11
15
16
1.20
DMT01c
0.74
1.21
14
12
1.28
DMT01d
0.84
1.33
13
9
1.30
DMT02b
1.46
0.98
17
15
1.43
DMT015b
1.66
0.99
17
14
1.46
DMT02c
1.89
1.41
12
10
1.36
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results.
- 267 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2.80
Ka x 10 6 ( s-1 )
2.40 2.00 1.60 1.20 0.80
control
0.40 0.00 0.00
0.40
0.80 1.20 [DMT] (mole%)
1.60
2.00
Figure 4.8. Effect of DMT on Ka for the dehydrochlorination of the resulting PVCs.
§4.3.3.5 Dehydrochlorination of PVCs prepared in the presence of γ-Butyrolactone The dehydrochlorination curves for six PVC samples prepared in the presence of γ-butyrolactone (GBL) are shown in Figure 4.9. The dehydrochlorination temperature was 190
o
C and the nitrogen flow was 333 mL/min for all samples. The
dehydrochlorination was stopped after 100 minutes and the HCl loss was less than 1mole% (relative to the theoretical HCl content in the polymer) for all samples. The kinetic parameters, Ka and Kb, were determined using equation (4.8) by least-squares regression, and the results are listed in Table 4.6. The effect of GBL concentration on Ka for the dehydrochlorination of the resulting PVCs is shown in Figure 4.10. It can be seen that the values of Ka were the same as that of the control within the experimental error. As we mentioned in Chapters 2, GBL did increase the initial polymerization rate and the molecular weights for the resulting polymers.
- 268 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 0.9 0.8
GBL0025b
GBL005b
GBL0075a
GBL01d
GBL02b
GBL04b
HCle ( mole% )
0.7 0.6 0.5 0.4 0.3 0.2 0.1 dehydrochlorination temperature: 190 oC 0.0 0
10
20
30
40 50 60 Time ( minute )
70
80
90
100
Figure 4.9 Kinetic dehydrochlorination curves for PVCs prepared in the presence of γ-butyrolactone (GBL).
Table 4.6. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of γ-butyrolactone (GBL)a). Sample [BL] Kb x 103 Induction time (ti)b) Ka x 106 -1 (%) (s ) Calc (min) Exp (min) (s-1) Controlc) 0.00 1.30 14 13 1.42 GBL0025b
0.27
0.98
17
14
1.40
GBL005b
0.46
0.80
21
17
1.43
GBL0075a
0.75
0.86
19
16
1.35
GBL01d
1.06
1.41
12
10
1.40
GBL02b
2.42
0.90
19
15
1.64
GBL04b
4.06
0.85
20
15
1.43
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results. - 269 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2.80
Ka x 10 6 ( s-1 )
2.40 2.00 1.60 1.20 0.80 control 0.40 0.00 0.00
1.00
2.00 [GBL] (mole%)
3.00
4.00
Figure 4.10. Effect of GBL on Ka for the dehydrochlorination of the resulting PVCs.
§4.3.3.6 Dehydrochlorination of PVCs prepared in the presence of Trimehtyl-1,3,5benzenetricarboxylate The dehydrochlorination curves for 6 PVC samples prepared in the presence of trimehtyl-1,3,5-benzenetricarboxylate
(TMB)
are
shown
in
Figure
4.11.
The
dehydrochlorination temperature was 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes. The kinetic parameters, Ka and Kb, were determined using equation (4.8) by least-squares regression, and the results are listed in Table 4.7. The effect of TMB concentration on Ka for the dehydrochlorination of the resulting PVCs is shown in Figure 4.12. Within the experimental error the values of Ka were the same as those of the control, at the TMB concentration levels we investigated. As mentioned in Chapter 2, TMB increased the initial polymerization rate and the resulting polymer molecular weights about 10% at the optimal TMB concentration.
- 270 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 1.0 0.9
TMB00025a
TMB001c
TMB0025c
TMB005d
TMB0075a
TMB01a
0.8 HCle ( mole% )
0.7 0.6 0.5 0.4 0.3 0.2 0.1
dehydrochlorination temperature: 190 oC
0.0 0
10
20
30
40 50 60 Time ( minute )
70
80
90
100
Figure 4.11 Kinetic dehydrochlorination curves for PVCs prepared in the presence of trimehtyl-1,3,5-benzenetricarboxylate (TMB).
Table 4.7. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in the presence of trimehtyl-1,3,5-benzenetricarboxylate (TMB)a). Sample [TMB] Kb x 103 Induction time (tin)b) Ka x 106 -1 (mole%) (s ) Calc (min) Exp (min) (s-1) Controlc) 0.00 1.30 13 14 1.42 TMB00025a
0.03
1.20
14
12
1.26
TMB001c
0.10
0.83
20
15
1.46
TMB0025c
0.26
1.55
11
10
1.40
TMB005d
0.49
1.16
14
14
1.55
TMB0075a
0.81
0.92
18
14
1.46
TMB01a
1.02
1.12
15
12
1.45
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle. c): PVC prepared in absence of additives; values of Ka and Kb are averaged results. - 271 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 2.80
Ka x 10 6 ( s-1 )
2.40 2.00 1.60 1.20 0.80 control 0.40 0.00 0.00
0.20
0.40 0.60 [TMB] (mole%)
0.80
1.00
Figure 4.12. Effect of TMB on Ka for the dehydrochlorination of the resulting PVCs.
§4.3.4 Dehydrochlorination of PVCs prepared in suspension In this section, the dehydrochlorination behavior of six PVC samples prepared in suspension polymerization is discussed. Three control samples were prepared in suspension polymerizations initiated by 0.05 mole% of diisobutyl peroxydicarbonate (SBP) without additives. They were named as PVC/SBP006, SBP124 and SBP138 with number-average molecular weights (Mn) of 54,700, 55,400, and 58,200, respectively. The other three were also prepared in suspension polymerization using 0.05 mole% of SBP as initiator, but with 0.2-1 mole% of dimethyl terephthalate (DMT). These samples are named PVC/SDMT002a, SDMT01a, SDMT01b, with number-average molecular weights of 60,300, 59,200, and 49,700, respectively. The molecular weights for most suspension samples are about 25-35% higher than those for the bulk controls (Mn≈44,000).
- 272 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 0.9 0.8
SBP006
SBP124
SBP138
SDMT002a
SDMT01a
SDMT01b
HCle ( mole% )
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
10
20
30
40 50 60 Time ( minute )
70
80
90
100
Figure 4.13. Kinetic dehydrochlorination curves for some suspension PVCs at 190 oC under fast nitrogen flow.
The kinetic dehydrochlorination curves for these six PVC samples are shown in Figure 4.13. The dehydrochlorination temperature was 190 oC and the nitrogen flow was 333 mL/min for all samples. The dehydrochlorination was stopped after 100 minutes and the total HCl loss was less than 1 mole%, Figure 4.13. The kinetic parameters, Ka and Kb, were determined by least-squares regression using equation (4.8), and the results are listed in Table 4.8. The average value of Ka for SBP006, SBP124 and SBP138 is 1.45±0.06x10-6s-1. This value is very close to the average Ka for the bulk controls (1.42±0.06x10-6s-1), even though the suspension sample molecular weights are 25-30% higher than those for the bulk polymers. This implies that Ka is not directly related to the polymer molecular weight. Later, in section §4.3.8, we will show that Ka is directly related to the labile structures in the polymer. If the labile structure concentration does not decrease with increased molecular weight, Ka does not decrease. - 273 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.8. Values of Ka and Kb for the dehydrochlorination of PVCs prepared in suspension with and without dimethyl terephthalate (DMT) using diisobutyl peroxydicarbonate (SBP) as initiator a). Sample [DMT] Kb x 103 Induction time (ti)b) Ka x 106 (mole%) (s-1) Calc (min) Exp (min) (s-1) SBP006 0.0 1.39 12 10 1.50 SBP124
0.0
1.46
11
11
1.46
SBP138
0.0
1.38
12
10
1.38
SDMT002a
0.2
1.79
9
8
1.34
SDMT01a
1.0
1.69
10
10
1.25
SDMT01b
1.0
1.39
12
11
1.28
a): Dehydrochlorination temperature was 190 oC. b): The calculated induction time (ti) is defined as ti=1/Kb and the experimental ti was found by extrapolating the “linear” region of the dehydrochlorination curve to zero HCle.
Table 4.8 shows that the values of Ka for SDMT002a, SDMT01a and SDMT01b are all lower than those for the suspension control samples. After evaluation using Sudent’s’ t test62, we find the probability of no-significant-difference between the values of Ka for these 3 samples and those for 3 suspension controls is less than 1%, which means the difference between them is ‘almost certainly’ significant. The average Ka for these 3 samples is 1.29±0.5x10-6s-1, about 10% lower than that for the suspension controls. This means SDMT002a, SDMT01a and SDMT01b samples are about 10% more thermally stable than the suspension controls. This finding is consistent with what we found in Chapter 2, where the dynamic thermal stability tests also showed that the PVCs prepared in suspension with added DMT were more thermally stable than the suspension controls. Earlier we also found that the Ka decreased about 10% for PVCs prepared in bulk when DMT concentration was 0.7 mole% or higher, section §4.3.3.4. It - 274 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ seems DMT decreases the resulting polymer structural defects in both bulk and suspension polymerization. The structural defects of the resulting PVCs will be reported in detail in sections §4.3.6 & §4.3.7 using 600MHz Nuclear Magnetic Resonance (NMR) technique.
§4.3.5 Activation energies for PVC Dehydrochlorination at early stage In the previous sections, dehydrochlorination measurements were carried out at only one temperature. The two-parameter model was successfully applied to various PVC samples prepared in bulk and in suspension. The kinetic parameters Ka and Kb were determined by least-squares regression using equation (4.8), and the additive effects on Ka for different samples were evaluated. In this section, the dehydrochlorination measurements for some selected samples were carried out at four different temperatures, ranging from 170 to 200 oC with intervals of 10 oC. Again the two-parameter model was used to analyze the PVC dehydrochlorination. The activation energies for both kinetic parameters, Ka and Kb, were determined and compared with the literature results.
- 275 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Exp_HCle @200C Exp_HCle @180C Calc_HCle Calc_HCls @190C Calc_HCls @170C
1.40
HCle ( mole% )
1.20
0.32
Exp_HCle @190C Exp_HCle @170C Calc_HCls @200C Calc_HCls @180C
0.28 0.24
1.00
0.20
0.80
0.16
0.60
0.12
0.40
0.08
0.20
0.04
0.00
0.00 100
0
10
20
30
40 50 60 Time ( minute )
70
80
90
HCls ( mole% )
1.60
Figure 4.14. Kinetic curves for the dehydrochlorination of PVC/AIBN319 in the temperature range of 170-200 oC (calculated HCls and HCle were obtained by fitting the experimental data with equations (4.6) and (4.8) by least-squares regression)
Nine PVC samples were chosen for dehydrochlorination measurements in the temperature range of 170-190 oC. The measurements were carried out in a 5-unit DEHYDRO apparatus. The temperatures were calibrated every time when the set temperature reached equilibrium. All dehydrochlorination curves are omitted, except for one sample, PVC/AIBN319, shown in Figure 4.14. The calculated HCle’s, evolved HCl, in Figure 4.14 were obtained by fitting the experimental data with equation (4.8) using least-squares regression. It can be seen that the calculated HCle’s fitted very well with the experimental results. The HCls’, HCl inside the polymer, were calculated using equation (4.6). As can be seen in Figure 4.14, at each temperature HCls initially increased and reached a constant value after a short period of time. As the temperature dropped, the time needed to reach an equilibrium value of HCls increased. - 276 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ The values of Ka and Kb for AIBN319 and eight other samples at four different temperatures were determined by least-squares regression using equation (4.8); the results are listed in Tables 4.9 and 4.10. The values of Ka for all samples were very close to each other at all four temperatures, except for one, DMT01b, whose Ka values were about 20% lower than the averaged Ka values for the eight other samples at all four temperatures. Also the value of Ka for all samples doubled for each 10 degree’s temperature increase. However, the values of Kb did not behavior similarly. For most samples Kb increased with temperature, but for CMN00025b and DMB0001a, the values at 190 oC were smaller than those at 180 oC, Table 4.10. This is most likely due to the lack of uniform sample morphology, since Kb is the rate constant for HCl diffusing through the polymer surface, which is more sensitive to the sample shape, size, etc.
If Ka and Kb are Arrhenius-type rate constants, the activation energies for both Ka and Kb can be obtained by plotting Ln(Ka) and Ln(Kb) versus reciprocal temperature. Arrhenius plots of Ka and Kb for four PVCs are shown in Figures 4.15 and 4.16. The activation energies were calculated from the slopes and the values are listed in Tables 4.9 and 4.11. The activation energies for the other five samples were calculated using the same method, but their Arrhenius plots are omitted. Table 4.9 shows that the activation energies (Ea) of Ka are very close to each other; the average Ea is 125±6 kJ/mole. This implies that the dehydrochlorination mechanism is the same for all samples in the temperature range of 170-200 oC. Table 4.10 lists the activation energies for PVC dehydrochlorination obtained by different authors. As can be seen, our result is consistent with the others.
- 277 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 17.5 AIBN319
CMN00025b
DMB00005b
DMT01b
16.5
-Ln(Ka)
15.5 14.5 13.5 12.5 11.5 2.10
2.14
2.18 2.22 1000/T (1/K)
2.26
2.30
Figure 4.15. Arrhenius plots of Ln(Ka) as a function of reciprocal temperature for some representative PVCs Table 4.9. Values of Ka, and activation energies (Ea) of Ka for the dehydrochlorination of some PVCs in the temperature range of 170-200 oC. Sample Ea Ka x 106 (s-1) o o o o @200 C @190 C @180 C @170 C (kJ/mole) AIBN319 2.70 1.35 0.66 0.28 132±3 AIBN416
2.86
1.51
0.75
0.35
124±2
CMN00025b
2.62
1.22
0.67
0.27
131±6
DMB00005a
3.08
1.53
0.70
0.38
125±5
DMB00005b
2.82
1.38
0.73
0.30
127±4
DMB0001a
2.56
1.33
0.62
0.31
119±4
DMB0001b
2.48
1.26
0.63
0.31
122±3
DMT01b
2.16
1.10
0.57
0.24
125±3
DMT02c
2.78
1.36
0.64
0.32
125±1
(average)
(125±6)
- 278 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.10. Comparison of the activation energies (Ea) for PVC dehydrochlorination obtained by different authors. Author Ea (kJ/mole) Medium Temperature (oC) Reference present work
125±6
N2
170-200
Abbas
117
N2
160-190
[63]
Guyot
113±2
140-170
[64]
Sharpiro
125.4±2
170-200
[65]
Mitani
121
150-190
[66]
Danforth
117-121
He
180-190
[67]
Klaric
N2 air in various solvents & in solid state
170-200
[68]
He
128 107 136
160-190
[69]
Yoshioka
116.5
80-280
[70]
N2
10.0 AIBN319
CMN00025b
DMB00005b
DMT01b
9.0
-Ln(Kb)
8.0 7.0 6.0 5.0 4.0 2.10
2.14
2.18 2.22 1000/T (1/K)
2.26
2.30
Figure 4.16. Arrhenius plots of Ln(Kb) as a function of reciprocal temperature for some representative PVCs.
- 279 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.11. Values of Kb and the activation energies (Ea) of Kb for the dehydrochlorination of PVC samples in the temperature range 170-200 oC. Sample Ea Kb x 103 (s-1) o o o o @200 C @190 C @180 C @170 C (kJ/mole) AIBN319 1.65 1.47 0.82 0.76 50±13 AIBN416
2.88
2.00
0.89
0.51
104±8
CMN00025b
2.59
1.17
1.25
0.92
53±21
DMB00005a
1.56
0.97
0.75
0.31
90±14
DMB00005b
1.28
1.06
0.68
0.56
51±6
DMB0001a
1.45
0.82
0.91
0.43
60±18
DMB0001b
1.22
1.31
0.85
0.49
54±16
DMT01b
2.35
1.29
1.14
0.93
48±13
DMT02c
2.70
1.40
1.01
0.96
60±16
As shown in Table 4.11, the calculated activation energies for Kb had large standard errors for most samples; the values varied from 48 to 104 kJ/mole, but six out of nine values were in 50-60kJ/mole range. As mentioned earlier, Kb was defined as the rate constant for HCl diffusing through the polymer surface. To our best knowledge, no one has yet reported an activation energy data for HCl diffusion in PVC in the temperature range of 170-200 oC. Tikhomirov et al71 reported activation energies for a number of gases and H2O diffusion in unplasticized PVC in the temperature range of 25-90 oC (the PVC glass transition temperature was 75 oC). They reported an activation energy (ED) for N2 diffusivity of 61.8kJ/mole, 51.4 kJ/mole for Ar and 54.5kJ/mole for O2. ED was proportional to the square of penetrant diameter. El Nadi72 reported an HCl diameter of 0.27nm. The O2 diameter was 0.27nm, 0.29nm for Ar, 0.30nm for N2. Geritsen et al73 - 280 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ estimated an optical collision diameter to be 0.24nm for N2 and 0.29nm for HCl. The HCl diameter is close to the N2 diameter, and our activation energy for HCl diffusion, 5060kJ/mole, is close to the value of 61.8kJ/mole for N2 found by Tikhomirov et al71. However there are large differences between these two studies. Our experiments were run in a much higher temperature range, about 100 oC higher than the polymer glass transition temperature (Tg), and HCl is a polar gas but N2 is not. Above Tg, diffusion activation energies generally decrease as temperature rises.
§4.3.6 Identification of PVC Structural Defects using 1D & 2D NMR Previous sections discussed the dehydrochlorination measurements of PVCs prepared in the presence of various additives. The kinetic parameters Ka and Kb for PVC dehydrochlorination were determined using an equation derived from a 2-parameter model. The kinetic parameter Ka is related to the polymer structural defects. We found that Ka was smaller for PVCs prepared in the presence of some additives. This could be due to a decrease in active structural defect concentration. In this section we report the characterization and identification of the structural defects in our experimental polymers using 1D 1H & 13C and 2D homonuclear scalar coupling correlation (COSY), pulsed-field gradient heteronuclear multiple quantum coherence (gHMQC) and heteronuclear multiple-bond correlation (gHMBC) NMR spectroscopy. A comprehensive H and C chemical shift assignment of the structural defects is given at the end of this section. In the next section the structural defects of some selected PVCs will be quantified using 1D 600MHz 1H NMR spectroscopy.
- 281 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ (A) PVC Structural Defect Analysis by 1D 1H & 13C and 2D 1H-1H COSY NMR
125.0
55.0
52.5
35,54
35,87
37,44
40,00
40,29
40,75
41,01
41,22
41,46
40.0
51,57
53,03
54,03
54,65
55,05
56,00
41,74
42,46
42,81
42.5
124,19
127,27
127,92 128,67
129,30
129,37
129,75
57.5
43,94
44,64 45.0
127.5
56,89 57,92
58,33 58,85 59,38
13
129,88
130,07 130,11 130,17
130.0
60.0
45,85
47.5
59,75
60,82
61,38
62.5
46,93
48,35
48,64
48,91
ppm (f1)
130,32
132.5
62,16
63,97
64,40
ppm (f1)
130,40
131,50
134,33 ppm (f1) 135.0
37.5
Figure 4.17. 150.9 MHz C NMR spectra of PVC/SBP160 in THF-d8 at 45 oC (relaxation delay 3.00 sec, pulse 45.0 degree, acquisition time 1.30 sec, width 2500.0 Hz, 20480 repetitions, total time 24.5 hr).
Figure 4.17 shows 1D 150.9 MHz 13C NMR spectra of PVC/SBP160 in THF-d8 at 45 oC. SBP160 was a low molecular weight sample (Mn=29.500, Mw=52,500), extracted - 282 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ from a commercial suspension PVC. From Figure 4.17 we can see three main peaks at 56.0, 56.9 and 57.9ppm, respectively, and three main peaks at 44.6, 45.8 and 46.9ppm, respectively. These peaks correspond to the carbon resonance of PVC backbone methine and methylene, respectively. Besides these main peaks, there are dozens of small peaks between 35 and 63ppm, which correspond to the carbon resonance of various saturated short branches or end groups of PVC. Due to lack of 13C references, it is very difficult to identify these peaks. There are also some peaks between 124 and 135ppm, which correspond to the unsaturated double bond structures in the polymer. Due to the low concentrations, these double bonds can hardly be quantified. In the following, we report the analysis of the PVC structures by using 600MHz 1D 1H and 2D COSY, gHMQC & gHMBC NMR.
A 1D 600 MHz 1H NMR spectrum (A) of PVC/DMT01b and the corresponding 2D 1H-1H COSY spectrum (B) of the same sample is shown in Figure 4.18. As can be seen from the 1D spectrum, there are two main peaks marked as e and m, respectively, in the regions of 4.3-4.7ppm and 2.0-2.5ppm. These are the hydrogen resonances corresponding to the methine moiety (CHCl) and methylene moiety (CH2) for the PVC backbone. Besides these main peaks, there are a dozen relatively small peaks between 2.0-6.0ppm, marked as a to l from downfield to upfield. These peaks correspond to the various structural defects and end structures. The peaks marked by double stars are from THF-d8, and those with single star are from solvent impurities.
- 283 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
(A)
**
** e d * a
b
*
m h f g
i
*
* * l
* * *
f'
k
c
* *
*
j
Figure 4.18. 600 MHz 1D 1H (A) and 2D 1H-1H COSY (B) NMR spectra of PVC/DMT01b in THF-d8 at 50 oC with the corresponding correlation assignments. From the 2D COSY spectrum, we can see an (a, f) proton-proton correlation and an (a, l) proton-proton correlation, Figure 4.18. Two sets of multiplets, a, with chemical - 284 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ shifts from 5.8-5.9ppm, correspond to a double bond moiety. Since f is a doublet, and is close to the methine absorption, e, it can be assigned as a CH2Cl group adjacent to a double bond. The l peak at ~2.5ppm can be assigned as a CH2 group adjacent to the double bond. Therefore a, f and l represent a ~CH2-CH=CH-CH2Cl end group. This end vinylene group has two possible structures. One is trans and the other is cis. From spectrum (B), an (a,f’) correlation can be found, so f’ can be assigned as cis and f as
trans, since the trans end group is thermodynamically preferred and thus has a higher intensity than the cis end group in the NMR spectrum. Following the same procedure, b was assigned to ~CH2-CH=CH~ units, since we found a (b, k) correlation in the COSY spectrum. And c was assigned as ~ClCH-CH=CH2, a vinyl end group. There is no obvious correlation in the 2D NMR due to its low concentration in the polymer.
Multiplet d is located at the left shoulder of the main e peaks. We first thought it might correspond to head-to-head (H-H) structures. But the COSY spectrum showed no obvious correlation patterns. There is disagreement in the literature regarding the numbers of H-H units in the polymer chain. Mitani et al.65 found that there were about 37 H-Hs per 1000 monomer units in the ordinary polymer. Hjertberg et al.74 repeated Mitani’s experiment but found only about 0-0.2 H-H per 1000 monomer units. Both Abbas75 and Starnes76 claimed that H-H concentration was below their
13
C NMR
detecting limit. The peak difference between d and e was 0.129ppm at 600MHz, but shrank to 0.102ppm at 750MHz. It is possible d is generated by
13
C-H splitting in
~CHCl~ backbone. Using the above data, the one-bond 1JC-H coupling constant was calculated to be 152-154Hz. This is very close to the value of one-bond 1JC-H coupling
- 285 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ constant for CH3Cl (150.0Hz)77. We used this one-bond coupling constant for PVC to optimize our gHMQC experiments.
Peaks g, h, i can be assigned as various types of CH2Cl units since they are close to f. There is a clear (e, g) correlation and (i, m) correlation in the COSY spectrum. So g can be assigned as ~CHCl-CH2Cl unit, and i as ~CH2CH2Cl unit. It is very difficult to assign h unambiguously from the COSY spectrum, and gHMQC & gHMBC are needed to get complete assignments.
(B) PVC Structural Defect Analysis by 2D gHMQC and gHMBC NMR In Heteronuclear Multiple Quantum Coherence (HMQC) experiments the signal is detected by observing protons, rather than carbons (i.e. inverse detection), which is inherently much more sensitive and the relaxation time is shorter, theoretically having a 64-fold sensitivity gain (γ3H/ γ3C, where γ is the magntogyric ratio of a nucleus). HMQC is used to correlate 1H and
13
C peaks for directly bonded C-H pairs (i.e. one-bond
correlation). The coordinates of each peak seen in the contour plots are the 1H and
13
C
chemical shifts. Heteronuclear multiple-bond correlation (HMBC) provides correlations between two- and three-bond J-couplings. HMBC is also an ‘inverse detection’ experiment. The coordinates of each peak seen in the contour plots are the chemical shifts of a carbon and protons separated by two or three bonds. HMBC experiment is designed to suppressed one-bond correlations, but a few are observed in some spectra; in concentrated samples of conjugated systems, four-bond correlations can be observed38.
- 286 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Pulsed-field gradient (PFG)34,35 HMQC and HMBC (gHMQC, gHMBC) can greatly increase the signal-to-noise ratio, lower the concentration thresholds and reduce dynamic-range problems36. PFG version of HMQC and HMBC experiments can provide artifact-free spectra which enable structural characterization of trace components in the presence of much larger signals from the major structural components of the polymer. It is a very powerful tool to study structural defects of polymers. In this section gHMQC and gHMBC are used to investigate PVC structural defects. We first thought we might need low molecular PVC to get reasonably good spectra. It turned out it was unnecessary.
Figure 4.19 Contour plots of 1H-13C gHMQC NMR spectrum of PVC/DMT01b in THFd8 at 45 oC in the regions of (1H) 3.1-4.1 ppm and (13C) 34.0-50.0 ppm with one-bond correlation assignments (arrows point from H to C). - 287 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Figure 4.19 presents the gHMQC NMR spectrum of PVC/DMT01b in the regions of (1H) 3.1-4.1 ppm and (13C) 34.0-50.0 ppm. Figure 4.20 presents the gHMBC NMR spectrum of the same sample in the regions of (1H) 3.5-4.0 ppm and (13C) 25.0-75.0 ppm. By correlating these two spectra, all peaks could be assigned to the corresponding structural fragments; the assignments are illustrated on the spectra. Since HMQC only detected one-bond 1H-13C correlation, HMBC was needed to make the final assignments for the HMQC peaks.
Figure 4.20 Contour plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b in THFd8 at 45 oC in the regions of (1H) 3.5-4.0 ppm and (13C) 25.0-75.0 ppm with long-range correlation assignments (arrows point from H to C).
- 288 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Here g was assigned as the 1,2-dichloroethyl end group and h was assigned as a chloromethyl branch. i was assigned as a 2,4-dichlorobutyl branch and i’ was assigned as a 1,2-dichloroethyl branch, which was believed to present only in high temperature polymerized samples (>100 oC)78. We can see in Figure 4.20 that the intensity of this group was very low; it could not be detected in HMQC, Figure 4.19. Since the two protons in C2 of 1,2-dichloroethyl end groups are diastereotopic, they showed two peaks in HMQC and four peaks in HMBC. The two protons from the chloromethyl branches are also diastereotopic and showed a similar type of correlation pattern.
Figure 4.21 Contour plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b in THFd8 at 45 oC in the regions of (1H) 2.5-4.2 ppm and (13C) 20.0-134.0 ppm with long-range correlation assignments (arrows point from H to C).
- 289 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Figure 4.21 shows the gHMBC NMR spectrum of PVC/DMT01b in the regions of (1H) 2.5-4.2 ppm and (13C) 20.0-134.0 ppm. Peaks marked as f 42 and f 43 correspond to the (H4, C2) and (H4, C3) correlations in the ~CH2-CH=CH-CH2Cl end group. Similarly, peaks marked as f12 and f13 correspond to the (H1, C2) and (H1, C3) correlations in the same end group. Since the 13C chemical shifts of C2 and C3 in this ~CH2-CH=CH-CH2Cl end group are very close to each other, the corresponding peaks in Figure 4.21 can hardly be distinguished. The ~CH2-CH=CH-CH2Cl end group was identified unambiguously by the COSY spectrum as described earlier.
Figure 4.22 Contour plots of 1H-13C gHMBC NMR spectrum of PVC/DMT01b in THFd8 at 45 oC in the regions of (1H) 3.4-4.9 ppm and (13C) 15.0-103.0 ppm with long-range correlation assignments (arrows point from H to C). - 290 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Figure 4.22 shows the gHMBC NMR spectrum of PVC/DMT01b in the regions of (1H) 3.4-4.9 ppm and (13C) 15.0-103.0 ppm. In this broad region, we found two correlated peaks marked as e2' and e2' ' ’ with coordinates (69.4, 4.6) and (35.9, 4.6). This correlation pattern probably corresponds to a head-to-head structure, since only this structure provides such 13C chemical shifts. As can be seen, their intensities are very low, and some correlation patterns are undetectable. There is no way it can be measured using 1D 1H NMR since it is buried in ~CHCl~ backbone peaks.
Finally, a comprehensive assignment of the hydrogen and carbon chemical shifts for all identified PVC structural defects was carried out with the help of COSY, gHMQC and gHMBC spectra. The results are listed in Table 4.14. These include three end groups, three short branches and one head-to-head addition segment. There is no way to make this comprehensive assignment using only 1D 1H and/or
13
13
C NMR, since many 1H and
C peaks are buried in the PVC backbone peaks. In the next section, we quantify these
structural defects using 1D 600MHz 1H NMR spectroscopy.
- 291 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.12 Hydrogen and carbon chemical shift assignments for PVC end groups, headto-head addition segments and short branchesa). Structure Chemical shift (ppm) 1 13 H C δ δ b) H1 C1 (59.8) Cl Ha Cl b) H2 C2 (134.3) 5.92 2 1 3 Hb 1' b) H3a C3 (117.3) 5.36 H
1-chloro-2-propenyl end group (c)
H3b
5.22
H1
C1
c)
H2
C2
c)
Cl
H3 1
2
3
5 4
Cl
Cl
Cl head-to-head addition fragment (e’)
Cl
1
3 1' 4 2 Cl 4-chloro-2-butenyl end group (f)
Cl
Cl
2'
1
4.60
C3
6
3' 2 1' 1,2-dichloroethyl end group (g)
(69) 69.4 c) (72) c)
H4
C4
H5
C5
H6
C6
H1’
C1’
59.0
H1a, d)H1b
(32) 35.9 c) (35) c)
(60)
d)
2.55, 2.51
C1
47.5
H2
5.87
C2
130.2
H3
5.78
C3
129.9
H4
4.05
C4
44.1
H1
4.43
C1
58.6
d)
3.86, 3.78
C2
48.1
H1’
2.46
C1’
43.2
H2’
4.46
C2’
56.0
d)
2.73, 2.66
C3’
46.3
H2a, d)H2b
Cl
(44)
H3’a, d)H3’b
(continued on next page) - 292 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.12 (continued) Structure
1
H d) H1a, d)H1b Cl
Cl 3'
1
H1’
1'
2'' Cl Chloromethyl branch (h)
Cl
2'
1
2
3
4
Cl
Cl 2,4-dichlorobutyl branch (i)
Chemical shift (ppm) 13 C δ 3.82, 3.78 C1 2.45
C1’
43.1
H2’
C2’
37.8
H2’’
C2’’
61.2
H1
2.58
C1
44.3
H2
4.56
C2
55.6
d)
2.24, 2.20
C3
42.0
H4
3.72
C4
41.3
C1
61.0
C2
44.3
H3a, d)H3b
H1a Cl 2'
1'
1
Cl
2 2'' Cl 1,2-dichloroethyl branch (i’)
δ 45.2
d)
H2a, d)H2b
3.71, 3.69
H1’
C1’
H2’
C2’
a): The assignments were based on 1D 1H NMR, 2D COSY, gHMQC, and gHMBC NMR. For those atoms whose peaks were not identified or indistinguishable with the ~CH2-CHCl~ backbone peaks, their chemical shifts were left as empty. b): Based on the low molecular PVC 1D 13C NMR. c): Estimated values by Keller and Mugge79. d): Diastereotopic protons were arbitrarily assigned as Ha at downfield, Hb at upfield.
- 293 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.3.7 Determination of PVC Structural Defects using 1D 1H NMR In this section, we report the determination of structural defect concentration for some PVCs using high-resolution one-dimensional 600 MHz 1H NMR. The structural defect concentration is reported as the number of structural defects per 1000 monomer units, and also the number of structural defects per macromolecule. Seventeen PVC samples prepared under different conditions were chosen for these measurements. The measurements were carried out using a Varian Inova 600 MHz spectrometer with tertahydrofuran-d8 as solvent for all PVC samples. The polymer concentration was ~35mg/mL for all samples. The running parameters were given in section 4.2.3. The temperature was set as 45 oC for all experiments and the run time was 1.2 hr for each sample. The structural identification is given based on the chemical shift data listed in Table 4.12. The integration of the individual peaks was performed after carefully making corrections for the baseline. Integrations were averaged from multiple determinations and the results are listed in Tables 4.13-4.16.
The concentrations of 4-chloro-2-butenyl end groups, 1-chloro-2-propenyl end groups, 1,2-dichloroethyl end groups, chloromethyl branches, internal allylic segments and 2,4-dichlorobutyl branches were measured. Resonance peaks for head-to-head structural segments were buried in PVC backbone region and the resonance peaks for 1,2-dichloroethyl branches were buried in the 2,4-dichlorobutyl branch region. These two structural defects could not be quantified. However, the concentrations for these two structures were very low according to gHMBC spectra, so the two structures are ignored in the following discussion. - 294 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.13. Number of 4-chloro-2-butenyl end groups, 1,2-dichloroethyl end groups and chloromethyl branches per 1000 monomer units determined by 1D 1H-NMR. 4-chloro-2-butenyl 1,2-dichloroethyl chloromethyl Sample -CH2-CH=CH-CH2Cl -CHCl-CH2Cl >CH-CH2Cl (chem.shift) (5.7-5.9) (4.0-4.2) (3.9-4.0) (3.8-3.9) AIBN319 1.11 1.11 1.57 2.56 AIBN416
1.15
1.10
1.51
2.46
CMN0005c
0.93
0.91
1.70
2.51
CMN001a
0.95
0.95
1.45
2.48
DMB00005b
1.05
1.04
1.64
2.40
DMB0001b
1.17
1.09
1.46
2.39
DMT001c
0.96
0.96
1.82
2.65
DMT005a
0.79
0.80
1.47
2.43
DMT01a
0.87
0.87
1.52
2.32
DMT01b
0.66
0.64
1.16
2.21
DMT02b
1.24
1.13
1.91
2.48
SGBL002a
0.79
0.80
1.15
2.38
SDMP004a
0.98
0.97
1.51
2.46
SDMT01a
0.85
0.81
1.42
2.40
SBP006
0.90
0.77
1.52
2.40
SBP160a)
1.71
1.70
2.03
2.61
PVC86kb)
0.49
0.50
0.93
1.84
a): Low molecular weight sample (Mn=29,500, Mw=52,500) extracted from a normal suspension PVC. b): High molecular weight sample (Mn=86,000, Mw=157,400) prepared in bulk at room temperature.
- 295 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.14. Number of internal allylic chlorines (I), external_1 allylic chlorines (E) and tertiary chlorinesa) (T) per 1000 monomer units determined by 1D 1H-NMR plus 190 oC dehydrochlorination rate. -CHCl>CClCH2CHCl Labile dHCle/dt Sample -CH=CHb) CH=CH2 (E) -CH2-CH2Cl (T) chlorines CHCl- (I) x106 (chem. shift) (5.6-5.8) (5.3-5.4) (3.6-3.8) ( 1/s ) AIBN319 0.25 0.15 1.91 2.31 1.35 AIBN416
0.24
0.17
1.86
2.27
1.51
CMN0005c
0.23
0.22
2.04
2.49
1.51
CMN001a
0.20
0.27
2.16
2.63
1.51
DMB00005b
0.21
0.13
1.87
2.21
1.38
DMB0001b
0.22
0.21
1.73
2.15
1.26
DMT001c
0.17
0.18
2.07
2.42
1.43
DMT005a
0.17
0.17
1.85
2.18
1.31
DMT01a
0.18
0.17
1.69
2.04
1.20
DMT01b
0.12
0.10
1.52
1.73
1.10
DMT02b
0.26
0.24
2.18
2.68
1.43
SGBL002a
0.19
0.25
1.91
2.35
1.55
SDMP004a
0.20
0.18
1.98
2.36
1.42
SDMT01a
0.19
0.12
1.65
1.95
1.25
SBP006
0.20
0.22
1.96
2.37
1.50
SBP160c)
0.40
0.28
3.49
4.17
2.12
PVC86kd)
0.14
0.00
1.15
1.29
0.86
a): Here tertiary chlorines are mainly from 2,4-dichlorobutyl. Tertiary chlorines from long-chain branches are negligible at low conversion. b): Labile chlorines are defined as the sum of I , E and T. c): Low molecular weight sample (Mn=29,500, Mw=52,500) extracted from a normal suspension PVC. d): High molecular weight sample (Mn=86,000, Mw=157,400) prepared in bulk at room temperature. - 296 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.15. Number of 4-chloro-2-butenyl end groups, 1,2-dichloroethyl end groups and chloromethyl branches per macromolecule determined by 1D 1H-NMR. 4-chloro-2-butenyl 1,2-dichloroethyl chloromethyl Sample -CH2-CH=CH-CH2Cl -CHCl-CH2Cl >CH-CH2Cl (chem.shift) (5.7-5.9) (4.0-4.2) (3.9-4.0) (3.8-3.9) AIBN319 0.78 0.78 1.10 1.79 AIBN416
0.78
0.74
1.02
1.66
CMN0005c
0.62
0.60
1.13
1.67
CMN001a
0.60
0.60
0.91
1.56
DMB00005b
0.80
0.80
1.26
1.84
DMB0001b
0.86
0.80
1.07
1.75
DMT001c
0.73
0.73
1.39
2.02
DMT005a
0.61
0.61
1.13
1.87
DMT01a
0.70
0.70
1.22
1.86
DMT01b
0.56
0.54
0.99
1.88
DMT02b
0.83
0.76
1.28
1.66
SGBL002a
0.72
0.72
1.04
2.16
SDMP004a
0.92
0.91
1.42
2.31
SDMT01a
0.81
0.77
1.35
2.27
SBP006
0.79
0.67
1.33
2.10
SBP160a)
0.81
0.80
0.96
1.23
PVC86kb)
0.67
0.69
1.28
2.53
a): Low molecular weight sample (Mn=29,500, Mw=52,500) extracted from a normal suspension PVC. b): High molecular weight sample (Mn=86,000, Mw=157,400) prepared in bulk at room temperature.
- 297 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ Table 4.16. Number of internal allylic chlorines (I), external_1 allylic chlorines (E) and tertiary-carbon chlorinesa) (T) per macromolecule determined by 1D 1H-NMR. b) c) Sample -CH=CH-CHCl>CClCH2CHCl Total Total CHCl- (I) CH=CH2 (E) -CH2-CH2Cl (T) end group >C=C< (chem. shift) (5.6-5.8) (5.3-5.4) (3.6-3.8) AIBN319 0.17 0.10 1.34 1.98 1.05 AIBN416
0.16
0.11
1.26
1.89
1.03
CMN0005c
0.15
0.15
1.35
1.89
0.91
CMN001a
0.13
0.17
1.36
1.68
0.90
DMB00005b
0.16
0.10
1.43
2.16
1.06
DMB0001b
0.16
0.15
1.27
2.05
1.14
DMT001c
0.13
0.14
1.58
2.26
1.00
DMT005a
0.13
0.13
1.42
1.87
0.87
DMT01a
0.14
0.14
1.36
2.06
0.98
DMT01b
0.10
0.09
1.29
1.63
0.74
DMT02b
0.17
0.16
1.46
2.24
1.13
SGBL002a
0.17
0.23
1.73
1.99
1.12
SDMP004a
0.19
0.17
1.86
2.51
1.28
SDMT01a
0.18
0.11
1.56
2.25
1.08
SBP006
0.18
0.19
1.72
2.25
1.10
SBP160d)
0.19
0.13
1.65
1.90
1.13
PVC86ke)
0.19
0.0
1.58
1.96
0.87
a): Here tertiary chlorines are mainly from 2,4-dichlorobutyl. Tertiary chlorines from long-chain branches are negligible at low conversion. b): sum of the 1,2-dichloroethyl, 4chloro-2-butenyl and vinyl end group. c): sum of 4-chloro-2-butenyl, internal double bond and vinyl end group. d): Low molecular weight sample (Mn=29,500, Mw=52,500) extracted from a normal suspension PVC. e): High molecular weight sample (Mn=86,000, Mw=157,400) prepared in bulk at room temperature. - 298 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ From Table 4.13, we can see the number of chloromethyl branches per 1000 monomer units is almost constant for all samples except one. That sample is PVC86k, prepared at a lower temperature. All other samples were prepared at about the same temperature. This implies that the rate of formation of this branch was not influenced by additives or by the polymerization methods. The numbers of 1,2-dichloroethyl end groups per 1000VC are close to each other for all bulk polymerized samples, except for DMT01b, which has a lower value. However both DMT01a and DMT005a have a reasonably normal number. It is hard to say whether DMT had any influence on the formation of this kind of end group. The number of 4-chloro-2-butenyl end groups and vinyl end groups per 1000VC seems to relate to the molecular weights, since higher molecular weight samples have lower numbers. The number of 4-chloro-2-butenyl end groups and vinyl end groups per macromolecule were constant for almost all samples, except for DMT01b, Table 4.15. The total number of end groups per macromolecule for all samples was very close to 2, except for DMT10b, which was only 1.62. This situation for DMT01b seems strange.
From Table 4.14, we can see the numbers of 2,4-dichlorobutyl branches per 1000VC are also close to each other for most bulk samples. But DMT01a and DMT01b have relatively fewer 2,4-dichlorobutyl branches. SDMT01a also has fewer 2,4dichlorobutyl branches than do SBP006, SBL002 and SDMP004a. This implies that DMT has an important role in the formation of 2,4-dichlorobutyl branches in both bulk and suspension polymerization. Since the 2,4-dichlorobutyl branch connects to the backbone with a tertiary chlorine, which is believed to be a labile structure25, it could
- 299 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ initiate PVC dehydrochlorination at elevated temperatures. This finding rationalizes the decrease in dehydrochlorination rate for PVCs prepared in the presence of DMT.
Table 4.14 also shows that the numbers of internal allylic chlorines per 1000VC for DMT001c, DMT005a, DMT01a and DMT01b were significantly lower than for all other bulk polymerized samples. DMT02b has higher number of internal allylic chlorines, possibly because in the preparation of this polymer some impurities were introduced and the number-average molecular weight was 13-20% lower than other DMT samples. As can be seen in Table 4.14, SBP006, SDMT01a, SDMP004a and SGBL002a have essentially the same number of internal allylic chlorines per 1000 monomer units, but SDMT sample has fewer vinyl end groups than other suspension polymers (and fewer 2,4-dichlorobutyl branches as mentioned earlier). Here we have only one SDMT sample. More samples are needed to draw any conclusions for suspension polymerization. We can say that DMT addition reduces the internal double bond formation during vinyl chloride polymerization in bulk, and possibly during suspension condition. This finding is significant since this internal double bond has an allylic chlorine which is thought to be responsible for initiating PVC dehydrochlorination at elevated temperatures17, 22, 25.
§4.3.8 Correlation between the Dehydrochlorination Rate and the Structural Defects of PVCs Earlier in this Chapter we reported many dehydrochlorination measurements and the kinetic parameter Ka was determined for PVCs prepared in various conditions. That included 17 samples for which their structural defects were determined in detail in - 300 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ section 4.3.7. Ka is defined as the dehydrochlorination rate constant. Since the dehydrochlorination becomes zero-order after a short time and remains zero order at low levels of dehydrochlorination, the dehydrochlorination rate (dHCle/dt) is about equal to Ka in this region. The earlier Ka data is listed again in Table 4.14 as dHCle/dt, since we need
to
find
the
relationship
between
the
defect
concentration
and
the
dehydrochlorination rate. If a linear relationship holds, we can define the relationship between them.
The concentrations (in number per 1000 monomer units) for all structural defects (as stated earlier we treat some end groups as structural defects, too) are listed in Tables 4.13 and 4.14. Certain structural defects can initiate dehydrochlorination. Since the 2,4dichlorobutyl group is attached to a carbon with a tertiary chlorine (actually a tertiary carbon) in the polymer chain, it is believed this structure can initiate dehydrochlorination easily at elevated temperatures26. A double bond segment can also act as a dehydrochlorination initiator, if it has allylic chlorines, which lose HCl easily to form conjugated double bonds. There are three types of double bonds we identified. 4-Chloro2-butenyl end group, ~CH2-CH=CH-CH2Cl, has an isolated chlorine and is part of a chain end. There is no agreement as to whether or not this type of allylic structure can initiate dehydrochlorination. According to Starnes20, such allylic structures might undergo 1,4-dehydrochlorination. But Hjertberg and Sorvik21 found that these structures, and also 1,2-dichloroethyl end groups, had no major influence on the PVC thermal stability. Through NMR analysis, van den Heuvel and Weber80 reported that both ~CH2CH=CH-CH2Cl and ~CHClCH2Cl groups remained at the same concentration level after
- 301 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 5% HCl loss. We accept Hjertberg and van den Heuvel’s conclusion here. We plotted our data against various group combinations, but failed to get reasonably good correlation coefficients if these two groups were included. There is no dispute that chloromethyl branches do not initiate dehydrochlorination. Therefore, three groups are considered as labile groups. They are the internal allylic group, vinyl end group and 2,4-dichlorobutyl group. They either have an allylic chlorine or a tertiary chlorine, called the ‘labile chlorine’. All labile chlorines were added together and are listed in Table 4.14. A plot of dehydrochlorination rate (dHCle/dt) as a function of labile chlorine concentration is shown in Figure 4.23.
3.0
dHCle/dt x 10 6 ( s-1 )
2.5 2.0 1.5 1.0 0.5 0.0 0.0
1.0
2.0 3.0 4.0 Labile chlorines per 1000 monomer units
5.0
Figure 4.23. Correlation between the dehydrochlorination rate (dHCle/dt) and the labile chlorine concentrations in PVC.
Linear regression gave a straight line with slope of 4.3±0.3x10-4s-1 and an intercept of 4.0±0.8x10-7s-1. The correlation coefficient was 0.96.
We can see the
intercept is not zero, which indicates PVC may undergo dehydrochlorination by random - 302 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ initiation within the backbone at elevated temperatures even in the absence of labile structures81. In other words, low thermal stability may be an inherent property of PVC, due to this unavoidable process.
§4.3.9 PVC tacticity Finally the tacticity of seventeen PVC samples was determined by 1D 600 MHz 1H NMR; the results are listed in Table 4.17. The calculated number-average molecular weights for these 17 samples are also listed in Table 4.17, together with the GPC determined data. It can be seen that the calculated molecular weight is close the GPC determined one.
The syndiotacticity is about 55% for all samples except for PVC86k, prepared at room temperature, with a syndiotacticity of 57.5%.
All of the bulk samples were
prepared at 55 oC and the suspension samples at 53 oC. Table 4.17 shows that the syndiotacticity for DMT001a, DMT01a and DMT01b samples is about 0.5-1.0% higher than that for AIBN319 and AIBN416 samples, but DMT005a and DMT02b samples have essentially the same syndiotacticiy as do AIBN319 and AIBN416 samples. Suspension samples SDMT01a and SBP006 also have essentially the same syndiotacticiy, Table 4.17. It seems that the additives had little or no influence on the tacticity even though some additives influenced the molecular weights. Some additives also affected the structural defect concentration. A possibility in the constant tacticity is that our additives may be still not bulky enough to direct monomer addition. Even though they can interact with the hydrogen on the propagating chloromethylene radical, they have little or no - 303 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ effect on the penultimate position. So they have little or no effect on how the next monomer adds into the propagating chain.
Table 4.17. Tacticity and the number-average molecular weight (Mn) for some PVCs. Tacticity Mn Sample Triad (%) Diad (%) by GPC by NMR* rr mr mm r m AIBN319 29.9 49.3 20.8 54.4 45.4 43700 40717 AIBN416
29.9
49.6
20.6
54.7
45.3
42200
41391
CMN0005c
30.3
48.9
20.8
54.5
45.1
41500
45620
CMN001a
30.0
49.2
20.7
54.5
45.3
39400
44014
DMB00005b
30.2
49.4
20.5
54.8
45.1
47900
45126
DMB0001b
30.4
49.4
20.2
55.1
44.9
45800
40323
DMT001c
30.7
49.4
19.9
55.4
44.6
47700
47710
DMT005a
30.1
49.4
20.5
54.7
45.2
48000
55556
DMT01a
30.9
49.3
19.7
55.6
44.4
50200
51230
DMT01b
30.7
49.3
20.0
55.3
44.6
53200
72674
DMT02b
30.3
49.2
20.5
54.8
45.0
41900
37092
SGBL002a
30.5
49.3
20.2
55.1
44.8
56600
50607
SDMP004a
30.2
49.4
20.5
54.8
45.1
58800
46125
SDMT01a
30.7
49.0
20.3
55.0
44.7
59200
55310
SBP006
30.6
49.0
20.4
54.9
44.8
54700
50201
SBP160
29.8
49.2
21.0
54.2
45.5
29500
26205
PVC86k
33.3
48.6
18.0
57.5
42.3
86000
98425
*calculated by assuming every macromolecule has one double bond.
- 304 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.4 Conclusions A 2-parameter model was developed to describe the dehydrochlorination of PVCs in the bulk state with rapid removal of HCl. The kinetic equations derived were applied to determine the parameters Ka and Kb for dehydrochlorination of PVCs prepared in the absence and presence of various additives. Some additives influenced on the dehydrochlorination rate while others had no effect. The activation energies of dehydrochlorination for all samples were identical within the experimental error. This indicates that all the PVCs underwent dehydrochlorination through the same mechanism under our experimental conditions.
Structural defects in PVC were characterized by 1D 1H NMR and 2D correlation via homonuclear scalar coupling (COSY), pulsed-field gradient heteronuclear multiple quantum coherence (gHMQC) and heteronuclear multiple-bond correlation (gHMBC) NMR spectroscopy. Eight types of structural defects were identified using a combination of 1D and 2D NMR and six of them were quantified by 1D 1H NMR. Some additives stabilized the PVC, the dehydrochlorination rate was proportional to the defect concentration. Dimethyl terephthalate addition suppressed the formation of internal double bond structures and 2,4-dichlorobutyl branches. These two groups together with the vinyl end groups were found to be catalytic structures responsible for the initiation of PVC dehydrochlorination at elevated temperatures. A correlation was found between the dehydrochlorination rates and presence of some additives.
- 305 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ §4.5 References 1. Weiberg, E. L. In Encyclopedia of PVC, 2nd Edition, Vol. 1, Ed. Nass, L. I., Heiberger, C. A., Marcel Dekker Inc., New York and Basel, 1986, p1-35. 2. Tester, D. A. In Degradation and Stabilisation of PVC, Owen, E. D., Editor; Elsevier Applied Science, New York, 1984, p1-19. 3. Palma, G.; Carenza, M. J. Appl. Polym. Sci. 1970, 14, 1737. 4. Ivan, B.; Kelen, T.; Tudo, F. In Degradation and Stabilization of Polymers; Jellinek, H. H. G.; Kachi, H. Eds.; Elsevier Science, Amsterdam, Netherlands, 1989, Vol. 2, p483714. 5. Simon, P. Angew. Makromol. Chem. 1994, 216, 187-203. 6. Ivan, B. Advances in Chemistry Series 1996, 249, 19-32. 7. Hjertberg, T.; Sorvik, E. M. In Degradation and Stabilisation of PVC, Owen, E. D., Editor; Elsevier Applied Science, New York, 1984, p21-79. 8. Yassin, A. A.; Sabaa, M. W. J. Macromol. Sci. 1990, C30, 491-558. 9. Rogestedt, M.; Hjertberg, T. Macromolecules 1992, 25, 6332-40. 10. Rogestedt, M.; Hjertberg, T. Macromolecules 1993, 26, 60-4. 11. Minster, K. S. Polymer Yearbook 1994, 11, 229-41. 12. Starnes, W. H. Jr.; Girois, S. Polymer Yearbook 1995, 12, 105-31. 13. Milan, J.; Martinez, G. Revista de Plasticos Modernos 1997, 73, 570-9. 14. Troistskii, B. B.; Troitskaya, L. S. Int. J. Polym. Mater. 1998, 41, 285-324. 15. Troistskii, B. B.; Troitskaya, L. S. Polymer Yearbook 1999, 16, 237-66. 16. Starnes, W. H. Jr. Prog. Polym. Sci. 2002, 27, 2133-70. 17. Starnes, W. H. Jr. J. Polym. Sci., Polym. Chem. 2005, 43, 2451-67. - 306 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 18. Minster, K. S. Int. J. Polym. Mater. 1994, 24, 235-51. 19. Martinez, G.; Milan, J. L. J. Polym. Sci., Polym. Chem. 2002, 40, 3944-3949. 20. Cotman, J. D. Jr. J. Am. Chem. Soc. 1955, 77, 2790-93. 21. Starnes, W. H., Jr.; Hartless, R. L.; Schilling, F. C.; Bovey, F. A. Advances in
Chemistry Series 1978, 169, 324-32. 22. Hjertberg, T.; Wendel, A. Polymer 1982, 23, 1641-5. 23. Starnes, W. H. Jr.; Schilling, F. C.; Plitz, I. M.; Cais, R. E.; Bovey, F. A. Polym. Bull.
(Berlin) 1981, 4(9), 555-62. 24. Hjertberg, T.; Sorvik, E. M. Polymer 1983, 24, 673-84. 25. Abbas, K. B.; Sorvik, E. M. J. Appl. Polym. Sci. 1975, 19, 2991-3006. 26. Starnes, W. H. Jr.; Villacorta, G. M.; Schilling, F. C.; Plitz, I. M. Macromolecules 1985, 18, 1780-6. 27. Michel, A.; Schmidt, G.; Guyot, A. Polymer Preprints (Am. Chem. Soc., Divi. Polym.
Chem.) 1973, 14(2), 665-70. 28. Braun, D.; Bezdadea, E.; Holzer, G.; Sonderhof, D. Angew. Makromol. Chem. 1982, 104, 145-61. 29. Hjertberg, T.; Sorvik, E. M. Polymer 1983, 24, 685-92. 30. Wang, G.; Zhang, M. Shiyou Huagong 2001, 30(12), 933-4. 31. Muller, L. J. Am. Chem. Soc. 1979, 101, 4481-4. 32. Bax, A.; Griffey, R. H.; Hawkins, B. L. J. Am. Chem. Soc. 1983, 105, 7188-90. 33. Bax. A.; Summers, M. F. . J. Am. Chem. Soc. 1986, 108, 2093-4. 34. Hurd, R. E. J. Magn. Reson. 1990, 87, 422-8. 35. Hurd, R. E.; John, B. K. J. Magn. Reson. 1991, 91, 648-53.
- 307 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 36. Rinaldi, P. L.; Keifer, P. A. J. Magn. Reson. Ser A 1994, 108, 259-62. 37. Rinaldi, P. L.; Ray, D. G.; Litman, V. E. Polym. Inter. 1995, 36, 177-85. 38. Tokles, M.; Keifer, P. A.; Rinaldi, P. L. Macromolecules 1995, 28, 3944-52. 39. Wollenberg, K. F.; Kolp, C. J. Polymer Preprints (Am. Chem. Soc., Divi. Polym.
Chem.) 1997, 38(1), 845-6. 40. Liu, W.; Ray, D. G., Rinaldi, P. L.; Zens, T. J. Magn. Reson. 1999, 140, 482-6. 41. Liu, W.; Rinaldi. P. L.; McIntosh, L. M.; Quirk, R. P. Macromolecules 2001, 34, 4757-67. 42. Sahoo, S. K.; Zhang, T.; Reddy, D. V.; Rinaldi, P. L.; McIntosh, L. M.; Quirk, R. P.
Macromolecules 2003, 36, 4017-28. 43. Danforth, J. D., Spiegel, J.; Bloom, J. J. Macromol. Sci., -Chem. 1982, A17, 11071127. 44. Danforth, J. D. J. Macromol. Sci. –Chem. 1983, A19(6), 897-917. 45. Danforth, J. D. Polym. Stab. Deg., ACS Symposium Series 1985, 280, 285-298. 46. Georgiev, G.; L. Christov, L. J. Macromol. Sci. –Chem. 1990, A27(8), 987-997. 47. Knumann, R.; Bockhorn, H. Combust. Sci. Technol. 1994, 101, 285. 48. Troitskii, B. B.; Troitskaya, L. S. J. Polym. Sci., Polym. Chem. 1990, 28, 2695-2709. 49. Troitskii, B. B.; Troitskaya, L. S. Int. J. Polym. Mater. 1990, 13, 169-78. 50. Troitskii, B. B.; Troitskaya, L. S. J. Polym. Sci., Polym. Chem. 1993, 31(1), 75-81. 51. Troitskii, B. B.; Troitskaya, L. S. Eur. Polym. J. 1995, 31(6), 533-539. 52. Troitskii, B. B.; Troitskaya, L. S. Eur. Polym. J. 1997, 33(8), 1289-1294. 53. Simon, P. Polym. Degrad. Stab. 1990, 29, 155-163. 54. Simon, P. Polym. Degrad. Stab. 1990, 29, 253-270.
- 308 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 55. Simon, P. Polym. Degrad. Stab. 1992, 35, 45-48. 56. Simon, P. Polym. Degrad. Stab. 1992, 35, 157-161. 57. Simon, P. Polym. Degrad. Stab. 1992, 35, 249-253. 58. Simon, P. Polym. Degrad. Stab. 1995, 47, 265-273. 59. Bacaloglu, R.; Fisch, M. Polym. Degrad. Stab. 1994, 45, 301-338. 60. Radios, T.; Brown, G. R. J. Macromol. Sci., - Chem. 1997, A34, 743. 61. Beltran, M. Marcilla, A. Polym. Degrad. Stab. 1997, 73, 55. 62. Langley, R. in Practical Statistics Simply Explained, Revised Edition, Dover Publications, Inc, New York, 1970, p160-5. 63. Abbas, K. B.; Sorvik, E. M. J. Appl. Polym. Sci. 1973, 17(12), 3567-76. 64. Guyot, A.; Bert, M.; Burille, P.; Llauro, M.; Michel, A. Pure Appl. Chem. 1981, 53, 401-10. 65. Shapiro, J. S.; Starnes, W. H., Jr.; Plitz, I. M.; Hische, D. C. Macromolecules 1986, 19(1), 230-5. 66. Mitani, K.; Ogata, T.; Awaya, H.; Tomari, Y. J. Polym. Sci., Polym. Chem. Ed. 1975, 13(12), 2813-26. 67. Danforth, J. D.; Indiveri, J. J. Phys. Chem. 1983, 87(26), 5376-80. 68. Klaric, I.; Roje, U.; Kovacic, T. Hemijska Industrija 1982, 36(5), 145-7. 69. He, Z.; Hu, X.; Sun, G. Polymer Degradation and Stability 1989, 24(2), 127-35. 70. Yoshioka, T.; Saitoh, N.; Okuwaki, A. Chemistry Letters 2005, 34(1), 70-1. 71. Tikhomirov, B. P.; Hopfenberg, H. B.; Stannett, Vivian; Williams, Joel L.
Makromol. Chem. 1968, 118, 177-88. 72. El Nadi, M. Journal of Chemical Physics 1951, 19, 503-4.
- 309 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ 73. Gerritsen, H. J.; Ahmed, S. A. Physics Letters 1964, 13(1), 41-2. 74. Hjertberg, T.; Sorvik, E.; Wendel, A. Makromol. Chem., Rapid Commun. 1983, 4, 175-80. 75. Abbas, K. B. and Sorvik, E. M. J. Appl. Polym. Sci. 1976, 20, 2395-406. 76. Starnes, W. H.; Schilling, F. C.; Abbas, K. B.; Cais, R. E.; Bovey, F. A.
Macromolecules 1979, 12, 556-62. 77. Friebolin, H. Basic One- and Two-Dimensional NMR Spectroscopy, 2nd edition, VCH Publishers, New York, NY, 1993, p95. 78. Starnes, W. H., Jr.; Wojciechowski, B. J. Makromol. Chem. Macromol. Symp. 1993, 70/71, 1-11. 79. Keller, F.; Mugge, C. Faserforsch. Textiltech. 1976, 27, 347. 80. Van den Heuvel, C. J. M.; Weber, A. J. M. Macromol. Chem. 1983, 184, 2261-73. 81. Ivan, B.; Kelen, T.; Tudos, F. Makromol. Chem., Macromol. Symp. 1989, 29, 59-72.
- 310 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Chapter 5 Conclusions
- 311 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ In this dissertation, a comprehensive study was performed on the vinyl chloride polymerization in the presence of small amount of organic additives. These additives were weakly basic compounds, high dipole carbonyl compounds, ether compounds and some heteroraromatics. The initial polymerization rates and the molecular weights of the resulting polymers increased in the presence of weakly ‘basic’ compounds such as dimethyl terephthalate (DMT), ethylene carbonate (EC), γ-butyrolactone (GBL), tributylphosphine oxide (TBPO) and trimethyl-1,3,5-benzene tricarboxylate (TMB). A kinetic model was developed for vinyl chloride polymerization in the presence of these weakly basic additives, using the hypothesis that a hydrogen-bond complex formed between an additive and the terminal hydrogen of the propagating radical. The kinetic model successfully explained the increase of the polymerization rate and the resulting polymer molecular weights. An optimal additive concentration exists to get a maximum molecular weight increase and reduction of structural defects in the resulting polymers.
Various methods were applied to evaluate the additive effects on the polymerization and the resulting PVC polymer structures. Differential Scanning Calorimetry (DSC) was used to study the crystallization behavior of the resulting polymers. It was found that some additives had no effect on the crystallinity of the resulting polymers. Interestingly PVCs prepared in the presence of 2,6-dichloropyridine (DCPY), dimethyl terephthalate (DMT) and γ-butyrolactone (GBL) had a relatively higher fraction of crystallinity than that of the control. Some additives may decrease the structural defect concentration in the resulting polymers; DSC may be not a good technique to evaluate this decrease.
- 312 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ A series of dehydrochlorination measurements were carried out for PVCs prepared in the presence of various additives and a 2-parameter model was developed to describe the dehydrochlorination of PVCs in the bulk state with rapid removal of HCl. The additive effect on the thermal stability of the resulting polymers was evaluated using the equations derived from the 2-parameter model. It was found that some additives had no influence on the dehydrochlorination rate for the resulting polymers, but two additives, 2-courmaranone and dimethyl terephthalate had significant influence on the stability of the resulting polymers prepared at certain additive concentration.
This
finding was supported by NMR analysis of the resulting polymers.
It was found that the increase of the thermal stability is due to the decrease of the labile structures in the polymers. Additives like dimethyl terephthalate decreased two labile structures, internal allylic structures and 2,4-dichlorobutyl branches in the resulting polymers. These two groups were believed to be formed through chain-transfer to polymer and through radical ‘backbiting’. This proved our hypothesis that additivepropagating interaction might reduce the chance for propagating radicals to chain transfer or backbite to form labile structures.
- 313 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Appendix: Thesis Defense Presentation
- 314 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Kinetics and Mechanism of Vinyl Chloride Polymerization: Effects of Additives on Polymerization Rate, Molecular Weight and Defect Concentration in the Polymer Kun Si Advisors: Prof. Morton Litt and Prof. Jerome Lando Department of Macromolecular Science and Engineering Case Western Reserve University December 05, 2006
Kinetics and Mechanism of Vinyl Chloride Polymerization Outline: •Introduction •Background and Hypothesis •Experimental •Results and Discussion •Kinetics •Dehydrochlorination •1D & 2D NMR •Conclusions
- 315 -
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Kinetics and Mechanism of Vinyl Chloride Polymerization Background
• 1. 2. 3. 4.
Polymerization of vinyl chloride Radical polymerization High chain transfer constant Limited Molecular weights Various defects formed during polymerization
• Polyvinyl chloride 1. Low thermal stability 2. Release HCl gas just above its glass transition temperature (Tg) 3. Low crystallinity 4. Low stereoregularity (tacticity)
Kinetics and Mechanism of Vinyl Chloride Polymerization Background Possible reasons causing low thermal stability: 1. 2. 3. 4.
Molecular weights Tacticity Labile structural defects All of above
Possible structural defects: 1. 2. 3. 4. 5.
- 316 -
Short chain branches Long chain branches Chloroallylic groups End groups Head-to-head structures
Kinetics and Mechanism of Vinyl Chloride Polymerization ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯
Kinetics and Mechanism of Vinyl Chloride Polymerization Background Many materials can change the microstructure of PVC •Aldehydes do this while lowering the molecular weights Stability increased for polymers prepared in the presence of small quantities of bimuthiol, epoxidized soybean oil, etc PVC forms miserable blends with Poly(methyl methacrylate) (PMMA) •H on ~CHCl~ group may interact with carbonyl group in PMMA
JASC, 1960, 82(3), 749
Burleigh
Acta Polymerica, 1981, 23(6), 295
Turska, et al
Polym. Commun., 1983, 24, 290
Jager, et al
Kinetics and Mechanism of Vinyl Chloride Polymerization
Structures shown in order of increasing acidity CH 2 CH 2 C C H Cl H Cl
H Cl < C C H H
