Table of Contents

Related Titles

Title Page

Copyright

Dedication

Acknowledgments

Preface

Nomenclature

Acronyms

Symbols

Greek Letters

Superscripts and Subscripts

Chapter 1: Introduction to Polyolefins

1.1 Introduction

1.2 Polyethylene Resins

1.3 Polypropylene Resins

Further Reading

Chapter 2: Polyolefin Microstructural Characterization

2.1 Introduction

2.2 Molecular Weight Distribution

2.3 Chemical Composition Distribution

2.4 Cross-Fractionation Techniques

2.5 Long-Chain Branching

Further Reading

Chapter 3: Polymerization Catalysis and Mechanism

3.1 Introduction

3.2 Catalyst Types

3.3 Supporting Single-Site Catalysts

3.4 Polymerization Mechanism with Coordination Catalysts

Further Reading

Chapter 4: Polyolefin Reactors and Processes

4.1 Introduction

4.2 Reactor Configurations and Design

4.3 Olefin Polymerization Processes

4.4 Conclusion

References

Further Reading

Chapter 5: Polymerization Kinetics

5.1 Introduction

5.2 Fundamental Model for Polymerization Kinetics

5.3 Nonstandard Polymerization Kinetics Models

5.4 Vapor-Liquid-Solid Equilibrium Considerations

Further Reading

Chapter 6: Polyolefin Microstructural Modeling

6.1 Introduction

6.2 Instantaneous Distributions

6.3 Monte Carlo Simulation

Further Reading

Chapter 7: Particle Growth and Single Particle Modeling

7.1 Introduction

7.2 Particle Fragmentation and Growth

7.3 Single Particle Models

7.4 Limitations of the PFM/MGM Approach: Particle Morphology

References

Further Reading

Chapter 8: Developing Models for Industrial Reactors

8.1 Introduction

References

Further Reading

Index

Epilogue

Isayev, A. I. (ed.)

Encyclopedia of Polymer Blends

Volume 2: Processing

Series: Encyclopedia of Polymer Blends 2011

ISBN: 978-3-527-31930-5

Isayev, A. I. (ed.)

Encyclopedia of Polymer Blends

Volume 1: Fundamentals

Series: Encyclopedia of Polymer Blends 2010

ISBN: 978-3-527-31929-9

Xanthos, M. (ed.)

Functional Fillers for Plastics

Second, Updated and Enlarged Edition

2010

ISBN: 978-3-527-32361-6

Elias, H.-G.

Macromolecules

Series: Macromolecules (Volumes 1–4)

2009

ISBN: 978-3-527-31171-2

Matyjaszewski, K., Gnanou, Y., Leibler, L. (eds.)

Macromolecular Engineering

Precise Synthesis, Materials Properties, Applications

2007

ISBN: 978-3-527-31446-1

Meyer, T., Keurentjes, J. (eds.)

Handbook of Polymer Reaction Engineering

2005

ISBN: 978-3-527-31014-2

Severn, J. R., Chadwick, J. C. (eds.)

Tailor-Made Polymers

Via Immobilization of Alpha-Olefin Polymerization Catalysts

2008

ISBN: 978-527-31782-0

Asua, J. (ed.)

Polymer Reaction Engineering

2007

ISBN: 978-4051-4442-1

The Authors

Prof. Dr. João B. P. Soares

University of Waterloo

Department of Chemical Engineering

University Avenue West 200

Waterloo, ON N2L 3G1

Canada

Prof. Dr. Timothy F. L. McKenna

C2P2 UMR 5265

ESCPE Lyon, Bat 308F

43 Blvd du 11 Novembre 1918

69616 Villeurbanne Cedex

France

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Library of Congress Card No.: applied for

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.

© 2012 Wiley-VCH Verlag & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany

All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form — by photoprinting, microfilm, or any other means — nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.

Print ISBN: 978-3-527-31710-3

ePDF ISBN: 978-3-527-64697-5

ePub ISBN: 978-3-527-64696-8

mobi ISBN: 978-3-527-64695-1

oBook ISBN: 978-3-527-64694-4

To our wives, Maria Soares and SalimaBoutti-McKenna, for their love, dedication, and patience while we wrote this book, not to mention the interminably long hours we spent discussing polyolefins in their presence. This book belongs to both of you, but you don't need to read it – you have heard all about it already.

João Soares and Timothy McKenna

Acknowledgments

Personally I'm always ready to learn, although I do not always like being taught.

Sir Winston Churchill (1874–1965)

Several of the concepts covered in this book arose from our daily interactions with students and colleagues in academia and industry. They are too many to be named individually here, but we would like to express our sincere gratitude to their outstanding contributions that are summarized in this work. We did like being thought by all of you.

First, we would like to thank our former mentors, who trusted and guided us when we were starting our careers, and kept encouraging us throughout these years. Their mentoring, support, and friendship are greatly appreciated.

This book could not have been written without the dedication of our graduate students, post-doctoral fellows, and research assistants, who toiled day after day in our laboratories to propose and test hypotheses, challenge us with unexpected new results, and in the process advance our understanding of polyolefin reaction engineering. Several of their results are interspersed throughout this book and constitute main contributions to the field of olefin polymerization science and engineering. We are very thankful to their hard work, perseverance, and confidence in us as their supervisors.

We would also like to thank our academic and industrial collaborators who over the years helped us better understand olefin polymerization and polyolefin characterization, often kindly allowing us to use their laboratory facilities (for free!) to complement the work done in our institutions. We are indeed indebted to these extraordinary colleagues and look forward to continue working with them in the future.

Finally, we would like to thank the polyolefin companies all over the world that have hired us as consultants and instructors of our industrial short course on Polyolefin Reaction Engineering. This book is a result, in large part, from the stimulating discussions we had with the scientists and engineers who took these courses. If it is true, as said by Scott Adams, the creator of the comic strip Dilbert, that “Give a man a fish, and you'll feed him for a day. Teach a man to fish, and he'll buy a funny hat. Talk to a hungry man about fish, and you're a consultant”, then we hope that talking to the course participants over these years has at least stimulated them to look deeper into the vast sea of polyolefin reaction engineering.

Preface

It is the mark of an instructed mind to rest satisfied with the degree of precision which the nature of the subject permits and not to seek exactness where only an approximation of the truth is possible.

Aristotle (384–322 BC)

The art of being wise is the art of knowing what to overlook.

William James (1842–1910)

The manufacture of polyolefins with coordination catalysts has been a leading force in the synthetic plastic industry since the early 1960s. Owing to the constant developments in catalysis, polymerization processes, and polyolefin characterization instruments, it continues to be a vibrant area of research and development today.

We have been working in this area for over 15 years, always feeling that there was a need for a book that summarized the most important aspects of polyolefin reaction engineering. This book reflects our views on this important industry. It grew out of interactions with the polyolefin industry through consulting activities and short courses, where we first detected a clear need to summarize, in one single source, the most generally accepted theories in olefin polymerization kinetics, catalysis, particle growth, and polyolefin characterization.

As quoted from Aristotle above, we will rest satisfied with the degree of precision which the nature of the subject permits and hope that our readers agree with us that this is indeed the mark of an instructed mind. It was not our intention to perform an extensive scholarly review of the literature for each of the topics covered in this book. We felt that this approach would lead to a long and tedious text that would become quickly outdated; several excellent reviews summarizing the most recent findings on polyolefin manufacturing and characterization are published regularly and are more adequate for this purpose. Instead, we present our interpretation of the field of polyolefin reaction engineering. Since any selection process is always subjective, we may have left out some approaches considered to be relevant by others, but we tried to be as encompassing as possible, considering the limitations of a book of this type. We have also sparsely used references in the main body of the chapters but added reference sections at their end where we discussed some alternative theories, presented exceptions to the general approach followed in the chapters, and suggested additional readings. The reference sections are not meant to be exhaustive compilations of the literature but sources of supplemental readings and a door to the vast literature in the area. We hope this approach will make this book a pleasant reading and also provide the reader with additional sources of reference.

Chapter 1 introduces the field of polyolefins, with an overview on polyolefin types, catalyst systems, and reactor configurations. We also introduce our general philosophy of using mathematical models to link polymerization kinetics, mass and heat transfer processes at several length scales, and polymer microstructure characterization for a complete understanding of olefin polymerization processes.

We discuss polyolefin microstructure, as defined by their distributions of molecular weight, chemical composition, stereo- and regioregularity, and long-chain branching, in Chapter 2. It is not an overstatement to say that among all synthetic polymers, polyolefins are the ones where microstructure control is the most important concern. Polyolefin microstructure is a constant theme in all chapters of this book and is our best guide to understanding catalysis, kinetics, mass and heat transfer resistances, and reactor behavior.

Chapter 3 is dedicated to polymerization catalysis and mechanisms. The field of coordination catalysis is huge and, undoubtedly, the main driving force behind innovation in the polyolefin manufacturing industry; to give it proper treatment, a separate book would be necessary. Rather, we decided to focus on the most salient aspects of the several classes of olefin catalysts, their general behavior patterns and mechanisms, and how they can be related to polymerization kinetics and polyolefin microstructural properties.

The subject of Chapter 4, polymerization reactors, is particularly dear to us, polymer reactor engineers. In fact, polyolefin manufacturing is a “dream come true” for polymer reactor engineers because practically all possible configurations of chemical reactors can be encountered. A great deal of creativity went into reactor design, heat removal strategies, series and parallel reactor arrangements, and cost reduction schemes of polyolefin reactors. We start the chapter by discussing reactor configurations used in olefin polymerization and then continue with a description of the leading processes for polyethylene and polypropylene production.

Chapter 5 is the first chapter dedicated to the mathematical modeling of olefin polymerization. We start our derivations with what we like to call the fundamental model for olefin polymerization kinetics and develop, from basic principles, its most general expressions for the rates of catalyst activation, polymerization, and catalyst deactivation. The fundamental model, albeit widely used, does not account for several phenomena encountered in olefin polymerization; therefore, some alternative polymerization kinetic schemes are discussed at the end of this chapter.

In Chapter 6, we develop mathematical models to describe the microstructure of polyolefins. This is one of the core chapters of the book and helps connect polymerization kinetics, catalysis, and mass and heat transfer resistances to final polymer performance. We opted to keep the mathematical treatment as simple as possible, without compromising the most relevant aspects of this important subject.

Particle fragmentation and growth are covered in Chapter 7. These models are collectively called single particle models and can be subdivided into polymer growth models and morphology development models. The two most well-established particle growth models are the polymeric flow model and the multigrain model. These models are used to describe heat and mass transfer in the polymeric particle after fragmentation takes place. The fragmentation of the catalyst particles themselves (described with morphology development models) is much harder to model, and there is still no well-accepted quantitative model to tackle this important subject. We review the main modeling alternatives in this field.

Finally, Chapter 8 is dedicated to macroscopic reactor modeling. This chapter is, in a way, the most conventional chapter from the chemical engineering point of view, since it involves well-known concepts of reactor residence time distribution, micromixing and macromixing, and reactor heat removal issues. The combination of macroscopic reactor models, single particle models, detailed polymerization kinetics, and polymer microstructural distributions, however, is very challenging and represents the ultimate goal of polyolefin reactor engineers.

Nomenclature

What's in a name? William Shakespeare (1564–1616)

Acronyms

CCD chemical composition distribution

CEF crystallization elution fractionation

CFC cross-fractionation

CGC constrained geometry catalyst

CLD chain length distribution

CRYSTAF crystallization analysis fractionation

CSLD comonomer sequence length distribution

CSTR continuous stirred tank reactor

CXRT computed X-ray tomography

DEAC diethyl aluminum chloride

DIBP di-iso-butylphthalate

DSC differential scanning calorimetry

EAO ethylaluminoxane

EB ethyl benzoate

EDX energy dispersive X-ray spectroscopy

EGMBE ethylene glycol monobutylether

ELSD evaporative light scattering detector

EPDM ethylene-propylene-diene monomer rubber

EPR ethylene–propylene rubber

FBR fluidized bed reactor

FFF field flow fractionation

FTIR Fourier-transform infrared

GPC gel permeation chromatography

HDPE high-density polyethylene

HMDS hexamethyldisilazine

HPLC high-performance liquid chromatography

HSBR horizontal stirred bed reactor

IR infrared

LALLS low-angle laser light scattering

LCB long-chain branch

LDPE low-density polyethylene

LLDPE linear low-density polyethylene

LS light scattering

MALLS multiangle laser light scattering

MAO methylaluminoxane

MDPE medium-density polyethylene

MFI melt flow index

MFR melt flow rate

MGM multigrain model

MI melt index

MWD molecular weight distribution

MZCR multizone circulating reactor

NMR nuclear magnetic resonance

NPTMS n-propyltrimethoxysilane

ODCB orthodichlorobenzene

PDI polydispersity index

PFM polymer flow model

PFR plug flow reactor

PP polypropylene

PSD particle size distribution

RND random number generated in the interval [0,1]

RTD residence time distribution

SCB short-chain branch

SEC size exclusion chromatography

SEM scanning electron microscopy

SLD sequence length distribution

SPM single particle model

tBAO t-butylaluminoxane

TCB tricholorobenzene

TEA triethyl aluminum

TEM transmission electron microscopy

TGIC temperature gradient interaction chromatography

TMA trimethyl aluminum

TOF turnover frequency

TREF temperature rising elution fractionation

UHMWPE ultrahigh-molecular weight polyethylene

ULDPE ultralow-density polyethylene

VLDPE very low-density polyethylene

VISC viscometer

VSBR vertical stirred bed reactor

Symbols

a = Mark–Houwink equation constant, Eq. 2.7

a_s = specific surface area of the support

A = monomer type A

A = total reactor heat transfer area

A_i = Arrhenius law preexponential factor for reaction of type i

A_S = support specific surface area

Al = cocatalysts

[AS^*] = concentration of active sites per unit surface area in the microparticle

B = monomer type B

B_n = average number of long-chain branches per polymer chain

C = catalyst precursor or active site

C^* = active site

[C₀] = initial concentration of active sites

C_d = deactivated catalytic site

C_p = heat capacity

D_b = bulk diffusivity

D_eff = effective diffusivity in the macroparticle

d_p = (or catalyst) particle diameter

D_p = diffusivity in the primary particle

D_r = dead polymer chain

D_{r, i} = dead polymer chain of length r having i long-chain branches

= dead polymer chain of length r having i long-chain branches and a terminal unsaturation (macromonomer)

E(t) = reactor residence time distribution

E_i = Arrhenius law activation energy for reaction of type i

f⁼ = molar fraction of macromonomers in the reactor

f_i = molar fraction of monomer type i in the polymerization medium

f_r = frequency Flory chain length distribution, Eq. 6.13

= overall frequency chain length distribution for chain having long-chain branches, Eq. 6.101

f_rk = frequency chain length distribution for chains with k long-chain branches per chain, Eq. 6.86

f_{log rk} = frequency chain length distribution for chains with k long-chain branches per chain, log scale, Eq. 6.88

F = monomer molar flow rate to the reactor

F_A = comonomer molar fraction in the copolymer

= average comonomer molar fraction in the copolymer

F_Br = molar fraction of comonomer B as a function of chain length

F_M,in = molar flow rate of the monomer feed to the reactor

F_M,out = molar flow rate of the monomer exiting the reactor

g = branching index, Eq. 2.18

g′ = viscosity branching index, Eq. 2.17

ΔG = Gibbs free energy change

h = average convective heat transfer coefficient between the macroparticle and surroundings

ΔH = enthalpy change

ΔH_p = average enthalpy of polymerization

ΔH_r = enthalpy of reaction

ΔH_u = enthalpy of melting for a crystallizable repeating unit, Eq. 2.26

ΔH_vap = enthalpy of vaporization

I₁ = Bessel function of the first kind and order 1

k_a = site activation rate constant

k_c, = forward and reverse rate constants, respectively, for the formation of dormant site with Ni-diimine catalysts, Table 5.8

k_d = first-order deactivation rate constant

= second-order deactivation rate constant

k_f = forward rate constant for reversible monomer coordination or β-agostic interaction; thermal conductivity

k_fL = effective thermal diffusivity in the macroparticle

k_fp = thermal conductivity of the polymer layer around the catalyst fragment in the microparticle

k_iH = rate constant for initiation of metal hydride active sites

k_p = propagation rate constant

k_P′ = apparent propagation rate constant, Eq. 5.115

= pseudo-propagation rate constant

= apparent propagation rate constant

= propagation rate constant for monomer type i (Bernoullian model)

= propagation rate constant for chain terminated in monomer type i coordinating with monomer type j (terminal model)

= propagation rate constant for chain terminated in monomer types i and j coordinating with monomer type k (penultimate model)

k_pm = propagation rate constant for meso insertion (propylene)

k_pr = propagation rate constant for racemic insertion (propylene)

k_r = reverse rate constant for reversible monomer coordination or β-agostic interaction

k_tAl = rate constant for transfer to cocatalyst

k_t = rate constant for β-hydride elimination

k_tH = rate constant for transfer to hydrogen

k_tM = rate constant for transfer to monomer

K = Mark–Houwink equation constant, Eq. 2.7

K_a = initiation frequency, k_a[Al]

K_eq = equilibrium constant for dormant sites, Eq. 5.67

K_g–l, = gas–liquid partition coefficients, Eq. 5.113

K_g–s, , = gas–solid partition coefficients, Eqs 5.111 and 5.114

K_l–s = liquid–solid partition coefficient, Eq. 5.112

K_H = Henry law constant

K_T = lumped chain-transfer constant, Eq. 5.74

= lumped chain-transfer constant, Eq. 5.70

= lumped chain-transfer constant, Eq. 5.71

m_i = mass fraction of polymer made on site type i

m_k = mass fraction of chains with k long-chain branches

m_p = mass of polymer, polymer yield

m_vap = vaporization rate

mw = molecular weight of repeating unit; in the case of copolymers, the average molecular weight of the repeating units

M = molecular weight

M = monomer

M_C = molar mass of catalyst

M_n = number average molecular weight

M_v = viscosity average molecular weight

M_w = weight average molecular weight

MW = polymer molecular weight

n = number of long-chain branches per chain; number of active site types

n_c(v) = polymer particle size distribution

= number of moles of catalyst

n_LCB = average number of long-chain branches in a polymer sample

n_M = number of moles of monomer

n_w = weight average number of long-chain branches per chain

N_A = Avogadro number

N_i = flux of species i

N_s = number of macroparticles per unit volume of the reactor

Nu = Nusselt number

P_A, P_B = probability of propagation of monomers A and B, respectively

P_H* = metal hydride active site

P_M = partial pressure of monomer

P_p = propagation probability

P_r = living chain with length r

P_rⁱ = living polymer chain with length r terminated in monomer type i(A or B for binary copolymers) or 1-2 or 2-1 insertions for polypropylene

P*_{r i} = living polymer chain of length r having i long-chain branches

P₁ = dormant site due to β-agostic interaction, Table 5.5

= dormant site for Ni-diimine catalysts, Table 5.8

P_t = termination probability

PDI = polydispersity index

= polydispersity index for chains containing long-chain branches

= polydispersity index as a function of copolymer composition

PDI_k = polydispersity index for chain with k long-chain branches, Eq. 6.106

Pr = Prandtl number

= heat generation rate

r = polymer chain length

r_i = comonomer reactivity ratio

r_L = radial position in macroparticle (multigrain model)

r_n = number average chain length

= number average chain length for chains containing long-chain branches, Eq. 6.107

= number average molecular weight that would result in the absence of long-chain branch formation reactions, see footnote 13 in Chapter 6

= number average chain length as a function of copolymer composition

r_nk = number average chain length for chains with k long-chain branches, Eq. 6.103

r_s = radial position in the microparticle (multigrain model)

r_w = weight average chain length

= weight average chain length for chains containing long-chain branches, Eq. 6.108

= weight average chain length as a function of copolymer composition

r_wk = weight average chain length for chains with k long-chain branches, Eq. 6.104

= z-average chain length for chains with k long-chain branches, Eq. 6.105

= root-mean-square end-to-end distance of a polymer chain

R = gas constant

R_c = catalyst fragment radius (multigrain model)

R_i = reaction rate of species i

= squared radius of gyration of branched chains

= squared radius of gyration of linear chains

R_L = macroparticle radius (multigrain model)

R_p = polymerization rate

= average polymerization rate per unit volume of the reactor

′ = average polymerization rate per polymer particle

R_S = microparticle radius (multigrain model)

R_t = chain-transfer rate

Re = Reynolds number

Δ S = entropy change

Sc = Schmidt number

Sh = Sherwood number

t = time

= average reactor residence time

t_R = reactor residence time

= catalyst half-time

T = temperature

T_c = crystallization temperature

T_i0 = reactor inlet temperature

T_m = melting temperature

= melting temperature of an infinitely long polyethylene chain

T_S = temperature in the microparticle

T_w = reactor coolant temperature

U = global heat-transfer coefficient

V = Monte Carlo control volume; reactor volume

V_e = elution volume

V_i = interstitial volume

V_p = pore volume

V_R = reactor volume

w_{log r} = weight Flory chain length distribution, log scale, Eq. 6.24

w_r = weight Flory chain length distribution, Eq. 6.17

= cumulative weight Flory chain length distribution

= overall weight chain length distribution for chain having long-chain branches, Eq. 6.102

= Stockmayer bivariate distribution, Eq. 6.60

w_rk = weight chain length distribution for chains with k long-chain branches per chain, Eq. 6.87

w_{log rk} = weight chain length distribution for chains with k long-chain branches per chain, log scale, Eq. 6.89

= Stockmayer bivariate distribution, log scale, Eq. 6.61

= trivatiate distribution of chain length, chemical composition, and long-chain branching, Eq. 6.117

w_ry = Stockmayer bivariate distribution, Eq. 6.56

w_{log MW} = weight Flory molecular weight distribution, log scale, Eq. 6.32

w_MW = weight Flory molecular weight distribution, Eq. 6.30

W_C = mass of catalyst

x_c = mass fraction of catalyst in a supported catalyst

x_i = molar fraction of comonomer i in the copolymer

y = deviation from average comonomer molar fraction in the copolymer, Eq. 6.57

y_k = molar fraction of chains with k long-chain branches

Y_i = ith moment of living polymer

[Y₀] = total concentration of active sites or living polymer chains

Z = compressibility factor

Greek Letters

α = polymer chain hydrodynamic volume constant, Eq. 2.4; long-chain branching parameter, Eq. 6.93

β, = Stockmayer bivariate distribution parameters, Eqs. 6.58 and 6.62, respectively

ε = exponent relating g′ to g, Eq. 2.19; macroparticle void fraction

η = catalyst site efficiency, Eq. 5.45

[η] = intrinsic viscosity

ϕ = fraction of actives sites with growing polymer chains, Eq. 5.55

ϕ_i = fraction of living chains terminated in monomer type i

ϕ_k = Catalan numbers, Eq. 6.95

Φ = polymer chain hydrodynamic volume constant defined in Eq. 2.4

κ = size exclusion partition coefficient

λ, λ_n = number of long-chain branches per 1000 C atoms

μ = long-chain branching parameter, Eq. 6.94; viscosity

ρ_C = support (catalyst) density

= average value of the heat capacity per unit volume of the macroparticle

τ = Flory most probable chain length distribution parameter; ratio of all chain-transfer rates to the propagation rate, Eq. 6.29; g macroparticle tortuosity

= Flory most probable molecular weight distribution parameter, Eq. 6.31

τ_B = chain length distribution parameter for polymers containing long-chain branches, Eq. 6.90

τ_d = characteristic diffusion time in the macroparticle

Superscripts and Subscripts

∧ = pseudokinetic constant

— = average

12, 21 = 1-2 or 2-1 propylene insertions

A, B = monomer types

bulk = bulk conditions

C = catalyst, monomer type C in the case of terpolymerization

l = liquid phase

M = monomer

MC = Monte Carlo simulation rates and constants

P = polymer

s = solid polymer phase

Chapter 1

Introduction to Polyolefins

It is a near perfect molecule […].

Jim Pritchard, Phillips Petroleum Company

1.1 Introduction

Polyolefins are used in a wide variety of applications, including grocery bags, containers, toys, adhesives, home appliances, engineering plastics, automotive parts, medical applications, and prosthetic implants. They can be either amorphous or highly crystalline, and they behave as thermoplastics, thermoplastic elastomers, or thermosets.

Despite their usefulness, polyolefins are made of monomers composed of only carbon and hydrogen atoms. We are so used to these remarkable polymers that we do not stop and ask how materials made out of such simple units achieve this extraordinary range of properties and applications. The answer to this question lies in how the monomer molecules are connected in the polymer chain to define the molecular architecture of polyolefins. By simply manipulating how ethylene, propylene, and higher α-olefins are bound in the polymer chain, polyolefins with entirely new properties can be produced.

Polyolefins can be divided into two main types, polyethylene and polypropylene, which are subdivided into several grades for different applications, as discussed later in this chapter.¹ Taking a somewhat simplistic view, three components are needed to make a polyolefin: monomer/comonomer, catalyst/initiator system, and polymerization reactor. We will start our discussion by taking a brief look at each of these three components.

Commercial polyethylene resins, despite their name, are most often copolymers of ethylene, with varying fractions of an α-olefin comonomer. The most commonly used α-olefins are 1-butene, 1-hexene, and 1-octene. They are used to decrease the density and crystallinity of the polyolefin, changing its physical properties and applications. Industrial polypropylene resins are mostly isotactic materials, but a few syndiotactic grades are also available. There are two main types of propylene copolymers: random propylene/ethylene copolymers² and impact propylene/ethylene copolymers.

At the heart of all polyolefin manufacturing processes is the system used to promote polymer chain growth. For industrial applications, polyethylene is made with either free radical initiators or coordination catalysts, while polypropylene is produced only with coordination catalysts. Low-density polyethylene (LDPE) is made using free radical processes and contains short chain branches (SCB) and long chain branches (LCB). Its microstructure is very different from that of polyethylenes made with coordination catalysts. Coordination catalysts can control polymer microstructure much more efficiently than free radical initiators and are used to make polyolefins with a range of properties unimaginable before their discovery. The quotation at the beginning of this chapter was motivated by the revolutionary synthesis of unbranched, linear polyethylene, the “near perfect molecule,” using a Phillips catalyst.

Catalyst design is behind the success of modern industrial olefin polymerization processes because the catalyst determines how the monomers will be linked in the polymer chain, effectively defining the polymer microstructure and properties. Industrial and academic research on olefin polymerization catalysis have been very dynamic since the original discoveries of Ziegler and Natta (Ziegler–Natta catalysts) and Hogan and Banks (Phillips catalysts), with many catalyst families being developed and optimized at a rapid pace. There are basically four main types of olefin polymerization catalysts: (i) Ziegler–Natta catalysts, (ii) Phillips catalysts, (iii) metallocene catalysts, and (iv) late transition metal catalysts. Ziegler–Natta and Phillips catalyst were discovered in the early 1950s, initiating a paradigm shift in olefin polymerization processes, while metallocene and late transition metal catalysts (sometimes called post-metallocenes) were developed in the 1980s and 1990s, respectively. In Chapter 3, olefin polymerization catalysts and mechanisms are discussed in detail, while polymerization kinetic models are developed in Chapter 5. For our purposes in this chapter, it suffices to say that polymerization with all coordination catalyst types involves monomer coordination to the transition metal active site before insertion in the metal–carbon bond at one end of the polymer chain. The coordination step is responsible for the versatility of these catalysts: since the incoming monomer needs to coordinate to the active site before propagation may occur, the electronic and steric environment around it can be changed to alter polymerization parameters that control the chain microstructure, such as propagation and chain transfer rates, comonomer reactivity ratios, stereoselectivity, and regioselectivity. This concept is illustrated in Figure 1.1.

Figure 1.1 Coordination step before ethylene insertion on a polyethylene chain (active site control).

This mechanism controls the polymer microstructure better than free radical polymerization, where the chemical nature of the free radical initiator stops being important after a few propagation steps since the polymerization locus moves away from the initiator molecule, as illustrated in Figure 1.2. Free radical processes for LDPE production are not the main topic of this book and are only discussed briefly as comparative examples.

Figure 1.2 Generic free radical polymerization (chain-end control). R^*, free radical initiator; M, monomer.

Even though a few processes for ethylene polymerization use homogeneous catalysts in solution reactors, most olefin polymerization processes operate with heterogeneous catalysts in two-phase or three-phase reactors. This adds an additional level of complexity to these systems since inter- and intraparticle mass and heat transfer resistances during polymerization may affect the polymerization rate and polymer microstructure. If significant, mass and energy transport limitations create nonuniform polymerization conditions within the catalyst particles that lead to nonuniform polymer microstructures. Many other challenging problems are associated with the use of heterogeneous catalysts for olefin polymerization, such as catalyst particle breakup, agglomeration, growth, and morphological development, all of which are discussed in Chapter 7.

All the phenomena mentioned above take place in the polymerization reactor. The variety of polyolefin reactors can be surprising for someone new to the field: polyolefins are made in autoclave reactors, single- and double-loop reactors, tubular reactors, and fluidized-bed reactors; these processes may be run in solution, slurry, bulk, or gas phase. Each reactor configuration brings with it certain advantages, but it also has some disadvantages; the ability to select the proper process for a given application is an important requirement for a polyolefin reaction engineer. Chapter 4 discusses these different reactor configurations and highlights some important polyolefin manufacturing processes. Reactor models that take into account micromixing and macromixing effects, residence time distributions, and mass and heat transfer phenomena at the reactor level are needed to simulate these processes, as explained in Chapter 8.

Catalyst type, polymerization mechanism and kinetics, inter- and intraparticle mass and heat transfer phenomena, and macroscale reactor modeling are essential for the design, operation, optimization, and control of polyolefin reactors. Figure 1.3 shows how these different length scales are needed to help us understand olefin polymerization processes.

Figure 1.3 Modeling scales in polyolefin reaction engineering.

In summary, much science and technology is hidden behind the apparent simplicity of everyday polyethylene and polypropylene consumer goods. No other synthetic polymer is made with such a variety of catalyst types, reactor configurations, and microstructural complexity. In this book, we will explain how, from such simple monomers, polyolefins have become the dominant commodity plastic in the twenty-first century.

1.2 Polyethylene Resins

Polyethylene resins are classified into three main types: LDPE, linear low-density polyethylene (LLDPE), and high-density polyethylene (HDPE). This traditional classification distinguishes each polyethylene type according to its density range: 0.915–0.940 g cm⁻³ for LDPE, 0.915–0.94 g cm⁻³ for LLDPE,³ and 0.945–0.97 g cm⁻³ for HDPE, although these limits may vary slightly among different sources. Lower-density polyethylene resins ( < 0.915 g cm⁻³) are sometimes called ultra low-density polyethylene (ULDPE) or very low-density polyethylene (VLDPE). HDPE with molecular weight averages of several millions is called ultrahigh molecular weight polyethylene (UHMWPE). To reduce the number of acronyms in this book, we have often grouped MDPE, LLDPE, ULDPE, and VLDPE under the generic term LLDPE and have made no distinction between HDPE and UHMWPE; these resins are very similar from a structural point of view, as explained below.

This division according to polyethylene density or molecular weight, although standard for commercial resins, tells us little about their microstructures. A more descriptive classification, based on their microstructural characteristics, is presented in Figure 1.4.

Figure 1.4 Classification of polyethylene types according to branching structure and density.

HDPE and LLDPE are made with coordination catalysts, while LDPE is made with free radical initiators. LDPE has both SCB and LCB, while polyethylenes made by coordination polymerization generally have only SCBs. Some polyethylene resins made with specific metallocenes or Phillips catalysts may also have some LCBs, but their LCB topology is distinct from that of LDPE resins. Most commercial HDPE and LLDPE grades are made with Ziegler–Natta or Phillips catalyst. Phillips catalysts are very important for the production of HDPE but are not used for LLDPE manufacture. Metallocenes can be used in making both HDPE and LLDPE, but metallocene resins are very different from the ones made with either Ziegler–Natta or Phillips catalyst, as explained below. The market share of metallocene resins is still relatively small but has been increasing steadily since the 1990s. Resins made with late transition metal catalysts have had no significant commercial applications to date.

The mechanism of SCB and LCB formation in LDPE is different from that in coordination polymerization; in LDPE, LCBs are formed by transfer-to-polymer reactions, while SCBs result from backbiting reactions. Contrarily, SCBs in HDPE and LLDPE are produced by the copolymerization of α-olefins added to the reactor as comonomers. LCBs, when present, are also formed by copolymerization reactions with a polymer chain having a terminal group containing a reactive double bond (macromonomer).

Figure 1.5 illustrates the SCB formation mechanism during the copolymerization of ethylene and α-olefins. The SCB behaves as a defect in the polymer chain, decreasing polymer density, crystallite size, and melting temperature. Therefore, the higher the molar fraction of α-olefin in the polymer chain, the lower its crystallinity. HDPE resins have very low α-olefin comonomer fractions (typically below a few mole percentage), while the comonomer content increases from LLDPE to ULDPE to VLDPE.

Figure 1.5 Short-chain branch formation mechanism in coordination polymerization. The chains are shown growing on a titanium active site.

Density is, therefore, a reflection of the α-olefin molar fraction in the polyolefin chain, and it also depends, to a lower degree, on its molecular weight; all other factors being the same, polyolefins with higher molecular weight averages tend to have a slightly lower density than those with lower molecular weight averages.

Density has been used for decades to classify polyethylene resins, but it is a poor descriptor for these materials because the microstructure of commercial polyethylenes is too complex to be captured with a single density value. Let us first focus on the chemical composition distribution (CCD) of LLDPEs, that is, the distribution of α-olefin fraction in the polymer chains. Most commercial LLDPEs are made with heterogeneous Ziegler–Natta catalysts. These catalysts have more than one type of active site, each one producing polymer chains with different average comonomer fractions and molecular weights. In addition, active sites that favor α-olefin incorporation also result in polymers with lower average molecular weights. As a consequence, the CCDs of Ziegler–Natta LLDPE resins are very broad, generally bimodal, and the average α-olefin content is correlated to the polymer molecular weight, as illustrated in Figure 1.6.

Figure 1.6 Generic chemical composition distribution of an LLDPE resin made with a heterogeneous Ziegler–Natta catalyst.

Two distinct regions can be identified in Figure 1.6: a sharp high-crystallinity peak (low α-olefin fraction) and a broad low-crystallinity peak (high α-olefin fraction). These two regions are associated with at least two types of active sites, one with much lower reactivity ratio toward α-olefin incorporation than the other. As the relative amounts of polymer made under these two modes vary, polyethylene resins vary from HDPE, with a unimodal, high-crystallinity peak and sometimes a small, lower-crystallinity tail, to MDPE, LLDPE, ULDPE, and VLDPE, with increasingly pronounced lower-crystallinity peak. The shape of the CCD is a strong function of the catalyst type, but it also depends on ethylene/α-olefin ratio, α-olefin type, and polymerization temperature.

From the discussion of Figure 1.6, it is apparent that classifying these complex microstructures according to a single density value is inadequate. The picture becomes even more complex when we take a look at the joint distribution of molecular weight and chemical composition (MWD × CCD) for Ziegler–Natta LLDPE, such as the one depicted in Figure 1.7. This tridimensional plot summarizes the complexity inherent to most commercial polyolefin resins. It also demonstrates that microstructural characterization techniques are indispensable tools to understand these polymers, as discussed in Chapter 2.

Figure 1.7 Joint distribution of molecular weight and chemical composition (MWD × CCD) for an LLDPE made with heterogeneous Ziegler–Natta catalysts.

The MWD × CCD correlation exemplified in Figures 1.6 and 1.7 is not desirable for certain polyolefin applications. A notable example are bimodal pipe resins, where better mechanical properties are achieved if the higher molecular weight chains also have a higher α-olefin fraction than the lower molecular weight component. The reason for this improved performance has been linked to the presence of tie molecules, a subject that is, unfortunately, beyond the scope of this book. The reader is directed to the references at the end of the chapter for more information on this subject. The usual MWD × CCD correlation observed in polyethylene resins made with heterogeneous Ziegler–Natta catalysts can be partially reversed using at least two reactors in series, as depicted in Figure 1.8. The polymers are called bimodal polyethylenes because they have broad, and sometimes bimodal, molecular weight distribution (MWD). The first reactor makes low molecular weight HDPE in the absence, or under very low concentration, of α-olefin. Hydrogen, the standard chain transfer agent in olefin polymerization, is used in the first reactor to lower the polymer molecular weight. The polymer made in the first reactor is then transferred continuously to the second reactor, which is operated under higher α-olefin concentration in the absence, or under a much lower concentration, of hydrogen, thus producing an LLDPE component with higher average molecular weight than the HDPE component made in the first reactor. Many polyethylene industrial processes include two reactors in series to broaden the range of product properties, as described in Chapter 4.

Figure 1.8 Polyethylene with regular and reverse comonomer incorporation made in single- and dual-reactor systems, respectively.

The advent of metallocene catalysts added a new dimension to commercial polyolefin resins. Metallocenes are single-site catalysts that are used to make polyethylenes with completely different microstructures from those made with Ziegler–Natta and Phillips catalysts, but are still classified loosely as HDPE and LLDPE. Polyethylenes made with metallocene catalysts have uniform microstructures, with narrow MWDs and CCDs. Figure 1.9 shows the CCDs for a series of ethylene/1-hexene copolymers made with a metallocene catalyst. All distributions are narrow and unimodal in sharp contrast to the behavior observed for Ziegler–Natta LLDPEs. Notice the uniform incorporation of 1-hexene and the absence of the high-crystallinity peak. In addition, their average copolymer composition is independent of their MWD.

Figure 1.9 Chemical composition distributions of ethylene/1-hexene copolymers made with a metallocene catalyst.

From a polymerization reaction engineering point of view, polyolefins made with metallocene catalysts provide an excellent opportunity for model development because they have “well-behaved” microstructures. Chapter 6 shows that models developed for single-site catalysts can also be extended to describe the more complex microstructures of polyolefins made with multiple-site complexes such as Ziegler–Natta and Phillips catalysts.

Substantially linear polyethylenes are an important new class of polyolefins. These polymers are also made with single-site catalysts, but their main characteristic is the presence of LCBs formed by terminal branching. They are called substantially linear because their LCB frequencies are typically below a few LCBs per 1000 carbon atoms. Terminal branching is illustrated in Figure 1.10, showing a polymer chain with a reactive terminal bond being copolymerized with ethylene to form an LCB. The branching topology resulting from this mechanism is very different from that of LDPE resins. Figure 1.11 illustrates a model prediction for the chain length distribution and LCB frequency for a polyolefin formed via this mechanism. More details on this LCB formation mechanism are provided in Chapter 6.

Figure 1.10 Long-chain branch formation through terminal branching promoted by a single-site catalyst.

Figure 1.11 Model predictions for the long-chain branch and chain length (r) distributions of substantially linear polyethylene.

1.3 Polypropylene Resins

Because propylene is an asymmetrical monomer, polypropylene can be produced with different stereochemical configurations. The most common types of polypropylene, shown in Figure 1.12, are isotactic, syndiotactic, and atactic. In isotactic polypropylene, the methyl groups are placed on the same side of the backbone; in syndiotactic polypropylene, on alternating sides; and in atactic polypropylene, the methyl groups are arranged randomly along the chain. Atactic polypropylene is amorphous and has little commercial value. Both isotactic and syndiotactic polypropylene are semicrystalline polymers with high melting temperatures. Isotactic polypropylene dominates the market, likely because it is easily produced with heterogeneous Ziegler–Natta and metallocene catalysts; syndiotactic polypropylene can be produced only with some metallocene catalysts and has much less widespread commercial use.

Figure 1.12 Main polypropylene types: (a) isotactic, (b) syndiotactic, and (c) atactic.

Modern Ziegler–Natta catalysts used for propylene polymerization make isotactic polypropylene with a very small fraction of atactic polypropylene. Non-specific sites are responsible for the formation of atactic chains in Ziegler–Natta catalysts. Many years of catalyst development were required to minimize the fraction of these catalyst sites, as discussed in Chapter 3.

Ziegler–Natta catalysts also are used to make chains with very high regioregularity, favoring 1-2 insertions and head-to-tail enchainment (Figure 1.13). Defects such as a 2-1 insertion following a 1-2 insertion create irregularities along the polymer chain, decreasing its crystallinity and melting temperature. Several metallocene catalysts produce polypropylene with very high isotacticity but lower regioregularity, which causes their melting temperatures to be lower than of those made with Ziegler–Natta catalysts. Metallocene catalysts can also make polypropylenes with other stereostructures such as atactic–isotactic block chains, but these products have not found commercial applications yet.

Figure 1.13 Regioregularity in polypropylene polymerization.

Impact propylene/ethylene copolymers⁴ are produced using at least two reactors in series with heterogeneous Ziegler–Natta catalysts or supported metallocenes. The first reactor makes isotactic polypropylene, and the second produces propylene–ethylene random copolymer. The copolymer component is amorphous or has very low crystallinity, and is intimately dispersed in the homopolymer phase, even though the two phases are immiscible. The copolymer phase dissipates energy during impact, greatly increasing the impact resistance of these resins. Several processes have been designed to produce impact polypropylene of high quality, as discussed in Chapter 4. Figure 1.14 schematically illustrates a process for the production of impact polypropylene.

Figure 1.14 Process for the production of impact polypropylene.

Finally, the same comments made for Ziegler–Natta versus metallocene polyethylene apply to polypropylene resins. Metallocene catalysts make polypropylene with narrower MWD and, in the case of copolymers, narrower CCD.

Notes

¹Ethylene-propylene-diene(EPDM) terpolymers, another important polyolefin type, are elastomers with a wide range of applications. They are made using catalysts and processes similar to those used to produce polyethylene and polypropylene. Even though they are not discussed explicitly in this book, several of the methods explained herein can be easily adopted to EPDM processes.

²Propylene/1-butene copolymers and propylene/ethylene/α-olefin terpolymers are also manufactured, but in a smaller scale.

³Sometimes polyethylene resins in the range 0.926–0.940 g cm⁻³ are called medium-density polyethylene (MDPE).

⁴These materials are sometimes (erroneously) called block copolymers. In fact, they are heterophasic materials composed of isotactic polypropylene and random propylene/ethylene copolymer chains.

Further Reading