Compounding MPS : Multi-phase polymeric systems

Publisher’s version / Version de l'éditeur:

Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.

Questions? Contact the NRC Publications Archive team at

PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.

https://publications-cnrc.canada.ca/fra/droits

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

2nd Conference and Workshop on Nanotechnology in Petroleum Industry [Proceedings], 2004-10-11

READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.

https://nrc-publications.canada.ca/eng/copyright

NRC Publications Archive Record / Notice des Archives des publications du CNRC :

https://nrc-publications.canada.ca/eng/view/object/?id=a1590084-5d57-4b9d-bbc4-83143078400d https://publications-cnrc.canada.ca/fra/voir/objet/?id=a1590084-5d57-4b9d-bbc4-83143078400d NRC Publications Archive

Archives des publications du CNRC

This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at

Compounding MPS : Multi-phase polymeric systems

L. A. Utracki

NRCC/IMI, 75 de Mortagne, Boucherville, QC, Canada J4B 6Y4

Compounding MPS

Multi-phase Polymeric Systems

2nd _{Conference on Nanotechnology in Petroleum Industry,}

Outline

Introduction to MPS

Mixing-blending-compounding

Equipment: LIM, SSE, & TSE

Dispersive and distributive mixing

Compounding in shear and extensional flow

fields

MPS compounding

Blends

Nanocomposites

The multiphase polymeric systems (MPS) accounts

for over 95% of plastics production

:

Molten polymer + solid (excluding semi-crystalline polymers)

[ca. 50%]:

Composites Filled systems Nanocomposites

Molten polymer + immiscible liquid [ca. 38%]:

Polymer alloys and blends

Suspension and emulsions (paints, adhesives)

Molten polymer + gas [ca. 11%]:

Foams, micro- and nano-foams

More complex systems, viz. foamed nanocomposites with polymer blend as the matrix are being developed.

Introduction 1

Processing consists of: (1) Compounding, (2) Forming, and (3) Finishing.

Compounding is the key to profit:

Mechanics: (1) disperse the minor phase, (2) distribute it homogenously in the matrix and (3) stabilize the morphology.

Used minimum energy, minimize degradation and/or attrition.

Common steps while compounding MPS:

Preparation of ingredients (stabilization, metal extraction, drying, etc.). Interphase control (sizing, compatibilization, or intercalation).

Incorporation into molten polymer and wetting (de-aeration).

Breaking up the large domains by progressively increasing stresses agglomerates (dispersive mixing).

Distribute the dispersed phase uniformly throughout the melt (distributive mixing).

Impose the desired morphology by process variables.

Introduction 2

Mixing is a general term = to combine ingredients into one mass, so that the constituent parts are indistinguishable, a homogenization.

Blending = the processes that lead to formation of polymer blends.

Compounding = formulation + compatibilization + engineering processing;

preparation of ―a compound‖ with fillers or reinforcements.

Compounding must yield a compound with desired morphology, that either will remain stable during the following processing steps, or it will be

modified in a predictable manner.

To toughen add 5-10 wt% elastomer dispersed to drop diameter d ~ 1m in (crazing & cracking) or to d < 100 nm (shear banding).

To control permeability, blending must generate lightly compatibilized drops with ca. 50 m diameter.

To obtain phase co-continuity, the desired composition must be compatibilized

then process under well controlled P, T, ₁₂ conditions

Polymer mixing occurs only within the laminar flow region where the Reynolds number

Re = VD << Re(critical) = 2100

.

Within the laminar region there are two mechanisms to be distinguished: the distributive (or extensive) and the dispersive (or intensive).

Introduction 3

Blending method depends on the system:

Compatibilization by addition or reactive

Shear mixing is adequate for blends with similar viscosity and low interfacial tension coefficient, ₁₂ < 10 (mNm), but extensional mixing

is useful in the full range of variables.

The thermo-mechanical degradation in extensional flow is smaller than that in shear.

Compounding of solids follows the steps:

1. Incorporation of a solid into molten polymer

2. Wetting the solid particles (de-aeration; compatibilization)

3. Using increasing stress, breaking up agglomerates and aggregates by rupture and erosion (dispersive mixing is rate-determining)

4. Homogenization and stabilization of the resulting dispersion (e.g., network formation by controlled flocculation; orientation).

Theory of dispersive mixing may be based on microrheology, while that of distributive mixing on laminar flow (extensional,

Introduction 4

Mixing

— Machines 1

Melt mixing machines comprise the single-screw (SSE) and

twin-screw (TSE) extruders. The latter are classified as co- and

counter-rotating, intermeshing or not.

Mixing

— Machines 2

Compounder N (rpm) Q (kg/h) # Manufact. Applications

SSE 20-250 5-30,000 _{130 Forming, devolatilization}

Co-Kneader 20-500 5-4,000 _{1 company Compounding, calender feeding}

of PVC or PP

TSE-CORI 100-2500 1-80,000 _{50 Compounding, powder}

processing at Q ? 1 ton/h TSE-ICRR 100-1200 5-2,000 _{12 Compounding, processing of}

PVC TSE-CRNI 100-500 100-15,000 7 Devolatilization Planetary roller 10-60

400-4,000 5 Compounding, calender feeding _{of PVC, masterbatching}

Multi-screw 20-200 20-5,000 _{1 company Devolatilization, reactions with}

Preponderance, performance, & application

of modern compounders

Internal mixers 1

Depending on the size, these machines are mainly used for:

Production of elastomers

Preparation of experimental compositions, viz. laboratory internal mixers (LIM)

There are several problems for analyzing and scaling up

these machines:

Slow heating to the desired temperature

Large temperature differences (measured: DT > 100°C) between

barrel and rotor surfaces (thermal energy drain) Large variation of stress fields (material slip)

Differences in the flow requirements between the different types of MPS, e.g., blends vs. filled systems and composites (blend

dispersion is strain-controlled; breaking and eroding filler aggregates requires progressive stress increase).

Scaling is always difficult due to reduction of the

surface-to-volume ratio as the machine size increases.

Scaling criteria (to larger size or different equipment):

Strain:

(N, L, and H is rotor speed, length, and clearance, while V_s and V_b are sheared and batch volumes, respectively; t is the mixing time). The desired strain level is 104 _{[Funt, 1977].}

Dissipated energy in the power-law region:

where

f

is local volume of sheared or elongated material. The local deformation rates must be calculated from the mixer

geometry and process variables. The shear and elongational viscosities,



and



_e, are taken at the deformation rate of 1 sec-1_.

Internal mixers 2

s b

( LV / HV )Nt





n 1 e 2 dis e

E





f



dV





f 



dV

11

No. Variable Relation Variable Relation

1 Screw speed N/No = (D/D o)n Residence time t/to = (D/D o)t 2 Channel depth H/H o = (D/D o)h Throughput Q/Q o = (D/D o)q 3 Screw length L/Lo = (D/D o)l Shear rate /o = (D/D o)g 4 Power P/Po = (D/D o)po Strain /o = ( D/Do)gs 5 Torque T/To = (D/D o)m Specific Surface S/So = (D/D o)s 6 Specific energy E/Eo = (D/D o)es

Internal mixers 3

Ingen Housz (1982) relations for continuous mixers

(symbol with subscript ―o‖ indicate the reference machine):

All the exponents can be written as functions of three principal ones: n, h, and l:

g

o o

/ = (D/D )

 

No. Variable Relation Variable Relation

1 Power po = 3 + n + l - h Throughput q = 2 + n + h 2 Torque m = 3 + 2n + l - h Shear rate g = n – h + 1 3 Specific energy es = 1 + n + l – 2h Strain gs = l - h 4 Residence time t = l – 1 - n Specific Surface s = l – n – h -1

The numerical values of

n

,

h

, and

l

depend on the scaling

method:

Scaling from laboratory internal mixer (LIM) to

production-scale TSE must be carried out in several steps:

From LIM to small TSE (ca. 25 mm diameter) virtually empirical, with some help of method 2.

From TSE 25 to 55 mm; also empirical by method 3 or 4. Only for the TSE sizes above 55 mm (e.g., 90 to 305 mm)

Internal mixers 4

No. Method n h l 1 Geometry 0 1 1 2 Laminar flow -1 ½ 1 3 TSE (L/D = constant) -2/3 2/3 1 4 TSE (H/D = constant) -½ 1 1½ -1.5 -1 -0.5 0 0.5 1 1.5 2 n h l #1 #2 #3 #4

Q = Q

_D

- Q

_P

- Q

_L

SSE

— Flow 1

Transport flows:

V - drag or longitudinal flow

in the channel direction

T - in transverse direction

Drag flow: Q_D; Pressure flow: Q_p; Leakage flow: Q_L

Q

_L

is the only flow element

that may induced mixing:

Q_DL/Q_D = 3.5 %

Q_PL/P_P = 0.5 %

Flow in SSE 2

Flow in a SSE channel

resembles flow through a pipe

with little mixing.

Leakage flow, Q

_L,

is the only

component that has some

extensional flow element.

The estimated percentage of

the leakage flow per turn is:

Q_DL/Q_D = 3.5 % Q_PL/P_P = 0.5 %

Blending in SSE 1

Lindt & Ghosh model [1992]:

Pellets of two polymers melt into layers Layers are stretched during flow into threads

Threads break into droplets by the capillarity forces

Blending action is provided by

progressive stretching of the

power-law fluids by the shear stress:





n z 12 dv k exp b T dy   _ _  D  

Blending in SSE 2

Experimental evidence for

mixtures of PS/EVAc(14%VAc) compatibilized by SB-star

copolymer confirmed the assumptions.

Dispersive Mixing

Screw modifications & high-shear add-

on’s

Distributive Mixing

Examples of

distributive

mixing with

screw pins

and add-ons

Others:

Barmag’s mixer,

RAPRA’s CTM,

U.-Twente mixer,

static mixers

,

etc.

Static mixers operate on the principle of repetitive division of a flow channel into at least two new channels, reorienting them by 90°, and dividing again. The flow is purely pressure driven,

laminar shear.

Mixing is related to the numbers of striations, generated by a

number of elements, and the number of divisions (new channels) engendered by each element.

Static mixers

Static Mixer L/D DP_rel DV_rel D_rel L_rel

Koch S MX 9  1.0 1.0 1.0 Koch S MX L 26  1.8 0.8 2.4 Koch S MV 18  4.6 1.3 2.7 Kenics 29 0.6 1.9 0.8 2.7 Etoflo HV 32  2.0 0.8 2.7 Komax 38  8.9 1.3 5.4 Lightnin 100  29.0 1.4 15.3 PMR 320  511.0 2.4 86.0 Toray 13  1.9 1.1 1.6 N -Form 29  4.5 1.1 3.6 Ros s IS G 10  9.6 2.1 2.3

CHARACTERISTIC

SSE

TSE

Construction simple & sturdy complex

Capital cost low to moderate high

Operational cost low moderate

Velocity profile can be calculated complex

Residence time wide narrow

Shear stress wide range as desired

Mixing poor good

Heat generation high low

TSE

screw arrangements

The independent

control of the shear & conveying is the

reason why TSE are used extensively to mix and reactively compound polymers. TSE are available with screw diameter

D  455 mm,

L/D  100 operating at N  2500 rpm, yield Q  80 ton/h and specific energy input  1.80 kWh/kg.

ICRR CORI

Screw geometry and cross-section through the

intermeshing region. Note a good seal between two

screws (top) bur relatively large openings between

channels

Counter-rotating TSE

— ICRR

Screw geometry and cross-section through the intermeshing

region. Close spacing between the screws and small openings

in the intermeshing region provide highly positive conveying

characteristics

Ranking third in popularity are the

counter-rotating, non-intermeshing

TSE’s, the CNRI, used as reactors,

with limited applicability as mixers.

Co- vs. Counter-rotating, fully

intermeshing TSE: CORI vs. ICRR

CHARACTERISTIC Co-rotating

Counter-rotating

Clearance small large

Screw speed, rpm • 600 Š 300

Throughput high moderate

Self-wiping good moderate

Gap pressure & abrasion negligible high

Shear history uniform extreme differences

Conveying less effective high ; length &

crosswise closed cells Residence time relatively broad narrow

Surface regeneration relatively low high Devolatilization inefficient efficient

Extensional flow low high

Dispersive vs. Distributive Mixing

                                                 

1

2

3

4

Melt mixing of polymeric systems is always within the laminar type of flow, i.e., the Reynolds number:

Dispersive mixing

Mixing is usually carried out in a SSE or TSE

Flow in SSE is Poiseuille-type pipe flow, with

skin-core structure and little mixing.

Owing to compression ratio of the machine and ―corkscrew‖ flow, during blending, polymer layers

may be stretched, fibrillated and then broken by capillarity forces.

Systems with the yield stress flows by the plug flow mechanism.

Flow in TSE is more complex, as there are different screw

arrangements and a plethora of screw elements.

The flow is mainly shear with few percent of energy converted into elongation.

Theoretical basis is provided by microrheology modified to account for viscoelasticity and concentration effects.

Microrheology

Consider



_d

/

_m

:

In shear there are 4 regions of drop deformation: for 0.1   tip spinning

for 0.1   1 drop splitting

for 1   3.8 fibrillation and break for 3.8   deformation but not break

Consider

d

_

:

In shear or elongation drops will:

for 0.1   not deform

for 0.1   1 deform but not break for 1   2 deform and break for 2   deform into filaments

Consider t



:

In Newtonian systems drops deform and break when t he deformation time exceeds the critical value:

for t = 25 equilibrium de formation, for t = 160 drop break

t_d* t b *

t* = 

10 -1 10 1 10 3 10 -5 10 -3 10 -1 10 1 10 3

C RITIC AL CAPILLARITY NUMB ER VS. VISC O S ITY RATIO

 cri t = d/ ij  shear fl ow ex tens iona l fl ow  = (dispersed) / (matrix) averag ed

Distributive mixing is related to strain:

where t

_r

is the residence time.

For good mixing   10

4

_{is required; achievable in a TSE.}

Laminar flow theories

are the basis for development and

control of distributive mixing.

Static mixers

with pressure driven laminar shear flow are

most commonly used distributive mixers.

Chaotic mixers

are being developed. They are based

either on the journal bearing or split-and recombine flow

principles.

Owing to the vorticity component, the shear flow is less

efficient than extensional flow for generation of high

strains, thus distributive mixing.

Distributive mixing

;

r r

Laminar flows

1 100 104 106 108 1 10 100 n= 1 n= 2 n= 4 exp In te r fa c ia l a r ea r at io , R i St rai n:    R_ =   n)n R i = e xp{/3}

On all three accounts

[Erwin, 1991]:

1.

the magnitude of the interface increase,

2.

the energy required for mixing, and

3.

the rate of spatial separation of two material points,

Better mixing (by several orders of magnitude) was

found for the extensional than for the shear flow.

10-1 103 107

100 102 104

Mixing energy vs. interface growth for four types of flow

Et o / R simple shear uniaxial extension biaxial extension planar extension

1

2

₃

10 0 10 10 10 20 10 30 10 40 0 40 80 L o ca l s tr a in ,  Time, t(s) uniaxial e longation simple shear

Advantages of extensional flow field over that of

shear are evident from analysis of laminar flows:

100

1000 Computed deformation rates for 90° staggered kneading blocks in 50 mm APV at 200 rpm.

R at e o f d e fo r m et io n ( 1/ s) Shear rate

Extens ional rate

va n der W aal, 199 8

In a co-rotating TSE the extensional flow field is present mainly in the staggered kneading block sections.

3D computations for CORI with narrow, transport screw elements with pitch angle 10.9°, and bi-lobal blocks, gave the results (N = 200 rpm shown in the Figs. below.

Even within the blocks the elongation rate is ca. 32% lower than that

in shear.

9 10

Computed deformation rates for transporting section with pitch angle 10.9° in 50 mm APV at 200 rpm . R at e o f d e fo r m et io n ( 1/ s) 20 van d er Waal, 1 99 8 Shear rate

Extens ional rate

Concept of EFM

The extensional flow mixer (EFM) is:

Unlimited by viscosity Less degradative

It is inexpensive and easy to use



_{Dispersion mechanism}

A Relative viscosity 4 1 Simple shear Elongation B

EFM with converging and diverging (C-D) plates

Radial flow of progressively increasing extensional

stress.

The intensity of the flow field is controlled by

adjusting the gap between C-D plates.

EFM uses plates of different geometries (number

and position of the converging zones).

EFM-

3

The microrheology indicates that the dispersion process is controlled by three dimensionless parameters:

The viscosity ratio: The capillarity ratio: The reduced time:

In these relations:



_d and



_m is viscosity of the dispersed phase and the matrix, respectively;



is the deforming stress, d is domain

diameter, and



₁₂ is the interfacial tension coefficient;



is the imposed strain.

Basing on the experimental data [Grace, 1982; Elemans, 1989] and theory

[Palierne, 1991], the reduced time can be expressed as:

The last relation provides a guide for mixing in extensional flow field, for imposing the same strain, independent on the blend components. The microrheology indicates that the dispersion process is controlled by three dimensionless parameters:

The viscosity ratio: The capillarity ratio: The reduced time:

In these relations:



_d and



_m is viscosity of the dispersed phase and the matrix, respectively;



is the deforming stress, d is domain

diameter, and



₁₂ is the interfacial tension coefficient;



is the imposed strain.

Basing on the experimental data [Grace, 1982; Elemans, 1989] and theory

[Palierne, 1991], the reduced time can be expressed as:

The last relation provides a guide for mixing in extensional flow field, for imposing the same strain, independent on the blend components.

0.35 0.35

o 12

t*



cons tant(

  

/

)



d

  

/



cons tant

The reduced time scale

d m 12

/

d /

t*

/

  

 



 







Extensional Flow Mixer, EFM

Commercial mixers:

1. Spiral feed for better feed T-control and added rigidity.

2. Redesigned C-D plates and mounting.

4. Redesigned attachment to the die.

35

DEFM may be used

as a torpedo

attached to SSE or

as a ―stand-alone‖

unit fed from SSE,

TSE, gear pump, etc.

Sliding barrel

Dynamic Extensional Flow Mixer:

(DEFM)

Compatibilization - General

Compatibilization must accomplish:

1. Reduce interfacial tension coefficient, ₁₂, and dispersion size.

2. Stabilize morphology against stress-destruction during forming.

3. Provide adhesion between the phases in the solid state.

Compatibilization is accomplished by:

1. Addition (Note: assumption of thermodynamic equilibrium!) of: A small amount of tailored copolymer (e.g., tapered block copolymer,

A-B or X-Y type, with entropic and/or enthalpic interactions)

Addition of large amount of a polymeric, core-shell or multi-layered type, compatibilizer-cum-impact modifier

Addition of co-solvent (e.g., Phenoxy).

Helfand’s theory — principal conclusions:

Chain ends of both polymers concentrate in the interphase.

Low M_W components thermodynamically forced to the interphase. Interfacial tension coefficient, _ increases with M_W to an

asymptotic value.

Reciprocity between interfacial tension and thickness, Dl _.

Noolandi et al.:

It is possible to design a universal compatibilizer, X-Y, utilizing principles of competitive repulsive interactions between the homopolymers and copolymer’s segments.

Experimentally:

It was shown that di-block copolymers have higher interfacial activity than tri-block, or graft copolymers.

How to De v e lop PAB

7

Det er mine stabilit y of mor phology Def ine Desir ed

Propert ies

START ₀

Select components with compensat ing

pr opert ies

1

Check miscibility or compat ibiliz at ion

met hods 2 Examine t he economy 3 Compare pr opert ies wit h

specif icat ions

9

Select pr ocessing met hods

8

Def ine desir ed mor phology 4 Select compounding met hod Select r heology of blend component s Pat enting Production y e s y e s y e s no _no no

Alloying strategy

S. Bruce Brown’s Classification

1 Redistribution or Trans-reactions  Block & Random Copolymers

1a reactive end-groups of polymer-1 (P-1) attack main chain of polymer-2 (P-2) 1b chain cleavage/recombination involving all polymers

2 Graft Copolymer Formation  Graft Copolymers

2a direct reaction of end-group of P-1 with pendent groups of P-2

2b reaction of end-group of P-1 with pendent group of P-2 in the presence of a condensing agent 2c reaction of end-group of P-1 with pendent group of P-2 in presence of a coupling agent (“c”) 2d reaction of pendent groups of P-1 with main chain of P-2 in a degradative process

3 Block Copolymer Formation  Block Copolymers

3a direct reaction of end-group of P-1 with end-group of P-2

3b reaction of end-group of P-1 with end-group of P-2 in the presence of a condensing agent 3c reaction of end-group of P-1 with end-group of P-2 in the presence of a coupling agent (“c”) 3d reaction of end-group of P-1 with main chain of the P-2 in a degradative process

4 Crosslinked Copolymer Formation  Crosslinked Structures

4a direct reaction of pendent functionality of P-1 with pendent functionality of P-2

4b reaction between pendent functionalities of P-1 and P-2 in the presence of a condensing agent 4c main chain of P-1 reacts with main chain of P-2 in the presence of a radical initiator

4d reaction between pendent functionalities of P-1 and P-2 in presence of a coupling agent ("c")

5 Ionic Bond Formation  Block, Graft or Crosslinked Structures

5a ion-ion association mediated by metal cations as linking agents ("c") 5b ion-neutral donor group association mediated by metal cations

Reactive blending in a TSE 1

Am ps m a x. 2 7 rp m m a x. 3 90 Am ps m a x. 2 7 rp m m a x. 3 90 Am ps m a x. 2 7 rp m m a x. 3 90 P P P P P P ag ent gon fl an t P P P P P P P P P P P P P P P PS

Co-rotating

intermeshing TSE

with in-line flow and

US monitoring

-2.0 0.0 2.0 4.0 6.0 8.0 0 50 100 150 200 250 300 350 400 configuration 5 100 rpm 4 kg/h a tt e n u a tio n ( d B /c m ) time (s) Résultats RTD 102.161 t début: 345.204 t fin: 172.018 tavg: 0.0573253 variance: 499.682 aire: 1.23292 95% confid.: 0.00236081 base noise:

The overall residence time distribution for TSE-35 mm at N = 100 rpm, T = 220°C The

residence

time distribution

inside a TSE is narrow, represented by a single peak without a shoulder (Q = 4 kg/h; t_r = 100 s). The shoulder

originates from the complex flow

between the screw ends and the die.

0 100 200 300 400 500 600 700 800 0 50 100 150 200 250 300 350 400 1 kg/h 2 4 8.75 14.3 20 t _{r ( s )}

The average residence time in a TSE, t

_r

, depends on

throughput (i.e., degree of fill), Q, and the screw speed, N.

The higher Q and N, the smaller t_r. For N = 100 rpm the residence time of t_r  40 sec was recorded. Modern TSE operate at N  2500 rpm, with the

Reactive blending in a TSE 4

Microstructure monitoring during reactive compatibilization involves

observation of melting, measurements of reaction, analysis of dispersion and flow monitoring [Sakai, 1995]:

The left schematic below shows a cross-section of a barrel element with a window for monitoring the melting process.

The right schematic below shows the barrel element with the melt sampling port. The element must be located at the high pressure zone. Its use is mainly for monitoring the reaction kinetics.

Several in- or on-line melt flow and ultrasonic monitors are available. The most difficult is the determination of blend morphology.

Grafting of EPR or EPDM with methylmethacrylate (MMA) was carried out in a tangential or intermeshing TSE at T = 175°C [Staas, 1981].

MMA (or another vinyl monomer) was fed through injection port 15, toluene solution of peroxide was fed through injection port 13. After reaction in the zone 12, the graft polymer was devolatilized through port 14; 40 wt% mass gain was recorded.

Grafting of PP with glycidylmethacrylate (GMA) was carried out in a CORI TSE at T = 165-220°C. Importance of feeding peroxide & GMA into fully filled zone (mixing) was stressed [Hu & Cartier, 1998].

Redistribution 1

(trans-esterification)

Redistribution is a slow statistical process in the melt between: (1) ester groups, (2) amide groups, or (3) amide-ester groups. Redistribution may lead to destruction of crystallinity and/or lowering of T_g — as observed for PET/PC or PET/PA-6.

Transesterification has been observed in PEST/PC, PEST/LCP,

LCP/PC, PAr/LCP, etc.

It can be catalyzed by SnO₂ or Ti(OBu) ₄, and hindered by tri-phenyl phosphite (TPPite).

The reaction occurs between ester groups of the main chain as well as these of side groups, viz. of PMMA, EVAc, acrylic

impact modifiers, etc.

Controlled redistribution can be used for compatibilization (PET/LCP) and/or impact modification (PET/MABS).

Redistribution between PA and PA is a slow statistical

process resulting from the reversibility of trans-reaction

between amine and amide groups.

Trans-amidation products of PA-6 and PA-66 are

commercially available. They show improved

processability, enhanced mechanical and barrier

properties.

Blends of aliphatic and aromatic PA’s are usually

immiscible, compatibilized by trans-reaction, that may

reduce crystallinity.

The reaction is controlled by T and intensity of mixing.

Commercial examples:

PA-6I has been co-reacted with PA-46 for compatibilization.

Redistribution 2

Redistribution between PA and PEST or PC is slow and it may reduce crystallinity, thus progress of these reactions must be

strictly controlled, e.g., by concentration, process variables (e.g., screw speed, T, P, residence time, etc.) and/or a catalyst. PET has been co-reacted mainly with PA-66 and PA-6, PC with PA-6. The reaction can be catalyzed by, e.g., p-tolueno-sulfonic acid (PTSA), dimethylol propionic acid (DMPA), etc.

Since the PEST and PA resins usually are crystalline and brittle, the reaction product should be impact-modified.

Commercial examples:

PET/PA-66 = 9:1 with 0.2 wt% PTSA was blended at T = 290°C in a TSE and spun into monofilaments. Improved processability, zero-rejects, zero shrinkage, and enhanced mechanical properties were reported.

PET (grafted with p-tolueno sulfonic acid groups) was blended with PA-46 for improved impact and HDT.

Redistribution 3

0.1 0.2 0.3 0.4

PET/PA-6 = 1:1 with 0.2 wt% of p-tolueno sulfonic acid (PTSA)

C o n ver sio n , c( -) c = 1/(1 + 320 exp((297-T)/7.7))

Redistribution 4

The reaction between PET and PA-6 is mainly controlled by temperature (see the plot below) [Pillon & Utracki, 1984].

At T = 280 - 330o_{C the concentration of PTSA is not critical.}

PTSA initiates the reaction, decomposes & devolatilizes.

The reaction results in excellent adhesion of PET matrix to PA. PET-co-PA formation was demonstrated (see micrograph).

The compatibilized crystalline/crystalline blends still required toughening.

Most often used reactions for compatibilization that lead to

graft or block copolymers [Brown, 1992].

Polymer Modification Aids

Compatibilization of addition polymer blends involves

generation of free radicals via:

Thermal decomposition of initiators, e.g., azo compounds, peroxides, etc.

Photochemical decomposition of initiators Gamma-rays and electron beam irradiation Redox sources (at low temperatures)

High stress, mechano-chemical initiation.

A vector fluid, viz.

a co-solvent, monomers, oligomers, Supercritical Fluids (SCF, e.g., CO₂), or hydro-fluorocarbons,

may be used to transports the reactants to the

interphase.

51

The first-stage polymer (e.g., PO, PS, PPE) may require:

Grafting acidic groups, viz. AA, MAH, GMA or oxazoline to the main chain.

Note: owing to reversibility of the reaction with the equilibrium shifted toward MAH, the extend of grafting is small

Presence of ester side group, as in PMMA, MABS, EVAc, etc., Blending with halogenated PO or PS.

The second-stage involves reacting the first-stage polymers with PA or PEST chain-end groups (-NH₂, -OH or -COOH).

For toughening PA or PEST the following styrenics are often used: SEBS-g-MAH, SB-SMA, ABS-g-GMA, SB-MMA-g-GMA, etc.

The third-stage may involve compounding a polymer blend with the stage-two graft copolymer as a compatibilizer.

Commercial examples: PA-6/EEA; PA/LDPE/LDPE-g-BA; PET/HDPE/SEBS-g-MAH; PET/LCP/PP/EEA-g-GMA;

PET/PC/SEBS-g-GMA/PPE/HIPS, etc.

74 wt% of PA-66 (46NH₂ / Mg) was fed to a TSE with 26 wt% of PP-g-MAH (130 MAH / Mg). The copolymer formation was conducted at the amine / anhydride ratio = 1.

Alternatively, 10 wt% of PA-66 was blended with 0-4 wt% of PP-g-MAH, and 86-90 wt% PP.

The blends were characterized using DSC, FTIR, morphology.

1 2 3 P P -c o -P A -6 6 (w t% ) 2 4 10 wt% PA-66/PP-co-Pa-66/PP A va r ag e d ia m e te r , d 50 ( m) Copolymer added Reactive compatibilization Helmert et al., 1995

2-step compatibilization

Graft copolymers of PA with engineering resins are limited to PA alloys with PEST, PPE and more recently with PC.

Two strategies for PA/PPE compatibilization are:

grafting PPE with AA, fumaric acid (FA), itaconic acid (IA), citraconic acid (CA), MAH, GMA, oxazoline, etc.

grafting styrenics with AA, MAH, GMA, oxazoline, etc.

The method may be conducted in 3, 2 or 1 step, i.e.:

3: pre-functionalizing PPE, preparation of PPE-co-PA, and blending PPE/PA/copolymer. PO-impact modifier may also be added.

2: pre-functionalizing PPE, then blending with PA

1: feeding all polymers and functionalizing agents to the extruder.

Commercial examples:

PPE-g-MAH/SEBS-g-MAH /PA-66 in TSE at 300o_{C [Nakazima & Izawa, 1989].}

PPE grafted with MAH in a TSE at 300o_{C with N-bromo succin-imide, then}

blended with 40 wt% PA-6 [Akkapeddi et al., 1992].

Toughening PA

To toughen amorphous, semi-aromatic PA non-reactive and maleated EP was used [Scott & Macosko, 1995].

The reaction between EP-MAH and the terminal -NH₂ groups was very rapid.

The droplet distribution in reactive and non-reactive system was log-normal, with significant scatter.

Maleation reduced the EP drop diameter by a factor of ca. 5.

-1 -0.5 0 0.5 1

Dispersion of 20 wt% of EP or EP-MAH in amorphous PA

EP EP-MAH y = -0.628 + 0.392norm(x) R= 0.978 y = -0.162 + 0.443norm(x) R= 0.976 lo g d ( m)

data: Scott & Macosko, 1995

0.4 0.6 0.8 1 1.2 d /d o EP-MAH/EP dn (m) d/d0 0.0 1.13 1 0.01 0.55 0.49 0.03 0.61 0.54 0.10 0.72 0.64 0.30 0.32 0.28 1.0 0.22 0.19

Block Copolymers 1

Block copolymers are formed in end-group reactions of two polymers, e.g., PA and oxidized PO or carboxyl-terminated NBR, e.g., PA with PP-MAH, PP-AA, oxidized PE, maleated-PPE.

Multifunctional coupling agents can also be used, viz. diglycidyls, oxazolines, carbodiimides, isocyanates, etc. (epoxy group generates alcohol that may re-enter reaction.)

Phosphite accelerates the reaction between the end groups of immiscible PA blend components, forming block copolymer (and secondary

phosphites), e.g., PA-6 with either PA-11, PA-12, PA-66, PA-6T, etc. The method can also be used for PA/PEST.

Block copolymers with PS are formed using trimellitic chloride and hydroxyl-terminated PS, then reacting with PA.

Mixtures of PA’s or PA with PEST can be compatibilized by addition of multifunctional epoxide coupling resin.

Mixtures of PA’s or PA with PEST can be compatibilized by addition of functionalized PO or styrenic, viz. PE-MAH,

SEBS-GMA, etc.

The preferred method of compatibilization of PA/PO is by reaction between end-group amine of PA and COOH end-group of PO (viz.

carboxylterminated NBR, oxidized PP or LDPE). The most common process: PA6 -12, or -66 with PP and PP-MAH or PP-AA.

Block Copolymer 2

PA/TPU compatibilized by bis-isocyanate coupler.

Block copolymers of hydroxyl-terminated PS with PA were formed using trimellitic chloride.

Similarly, block copolymers can be formed

between MAH end groups of PPE and PA-66, ...

PS (70 wt%) with Polyisoprene (PI) blends

mixed at 180°C [Orr et al., 1997]. Left: the

micrographs (after OsO₄ staining):

A: Non-reactive PS & PI

B: PS terminated with aromatic amine with maleic anhydride-terminated PI

C: PS terminated with aliphatic amine with maleic anhydride-terminated PI  lamellar morphology on 30 nm scale.

Crosslinking

Several methods of compatibilization by crosslinking

have been developed. They are based on:

Activating agent,

Coupling agent, or

Use of ion-ion, ion-dipol or dipol-dipol interactions .

The

activating

agent causes crosslinking without

entering into the product, for example 0-0.5 % DCP or

DBP can be used to generate free radical coupling.

Popular, commercially available

coupling

agents are:

Bis-maleimide, di-methylol phenol.

Tri-allyl isocyanurate (TAIC), tri-methylol propane triacrylate Tetra-glycidyl di-amino phenyl methane (TGDDM), etc.

To predict blend morphology the following

assumptions were made

:

1. Mixing occurs in isothermal shear field.

2. Two power-law fluids with the interfacial tension.

3. Low concentration, f < f_perc —the blend morphology determined by

drop deformation, break-up and coalescence (microrheology!). 4. Drop deformation takes place only within the pressurized screw

zones.

5. Droplet deformation and break-up, as well as fibrillation and disintegration, contribute to dispersion.

6. Fibrillated drops disintegrate when either _cror when the fiber diameter becomes smaller than = 1 m.

7. The mechanism of drop coalescence is the same in an internal

mixer as in twin screw extruders, thus its characteristic constant can

Microrheology

makes it possible to predict evolution of drop size during mixing in a TSE The model

assumed shear flow with drop

breakup, fibrillation and coalescence. Neither normal forces nor extensional flow was considered.

Flow chart

Coalescence

1

Microrheology has been developed for infinitely diluted emulsions of Newtonian liquids, thus:

The non-Newtonian flow is accommodated by using constant stress viscosity. The concentration effect is accounted for by considering the coalescence. At dynamic equilibrium the rate of drop breakup (as described by

microrheology) must be equal the rate of coalescence, given by theory of dynamic coalescence [Utracki, 1973]:

Validity of the relation was experimentally verified.

Coagulation is related to the projected area of the drop.

* * 8 / 3 8 / 3

/

/ 3

/

/

cr b break eq Taylor cr b coalescence

d

t

d

t

d

d

C

t

d

t

C

d

0 4 8 12

0 0.02 0.06 0.08

Concentration effect on PE-drops' diameter in PS at 200°C; line — theory

30 rpm 40 rpm 50 rpm 60 rpm d eq ( m) f   Y = M0 + M1*X 0.7032 M0 139.31 M1 0.99061 R

The single coalescence parameter, C, was determined

experimentally in an internal mixer, then used for computation of the coalescence effects in a TSE.

Coalescence

2

* 8 / 3 1/ 2 o cr b

D



D



( 6 C



t

f

)

10 0 10 2 10 4 0 400 800

PE-drop diameter v s. screw length; lines — computed

Q5, N150

Q5, N200 Q5, N250



m)

Position along the screw, L (mm)

PE/PS = 5/95 T = 200°C

PE E ( GPa)  (%) NIRT ( J/m)

( wt%) TSE SSE+EFM TSE SSE+EFM TSE SSE+EFM

A morphous PET 10 1.98±0.02 2.12±0.05 82±21 53±14 39.3 50.4 20 1.91±0.46 1.68±0.10 68±21 46±26 55.1 54.4 20,Comp 1.47±0.02 1.85±0.52 38±6 37±3 76.8 81.5 Crystalline PET 10 2.44±0.09 2.36±0.11 9.5±0.7 13.3±2.7 29.0 35.3 20 2.09±0.36 1.85±0.13 17.5±3.6 14.5±2.6 43.3 42.3 20,Comp 2.07±0.28 2.04±0.20 13.1±2.7 12.7±2.0 61.5 58.5

PET-A: amorphous; C: crystalline; EFM: 2A2M; 2000 psi

Mechanical properties of PET blends with metallocene

catalyzed polyethylene, mPE (10 and 20 wt%) as well as

with 15 wt% mPE + 5 wt% PE-GMA were tested.

Blends compounded in two TSE’s or in a SSE + EFM.

Transverse Longitudinal 20%w mPE

15%w mPE +5%w. 36A

SSE+EFM: 2A2M, 14 MPa; Fractured surfaces in  and // direction to flow TSE; Fractured surfaces in  and // direction to flow

Relative impact strength (RIS) of PP with 5, 9 or 15 wt% EPR. SSE + EFM data are shown as points, the TSE data as

horizontal, broken lines.

100 (

_blend

/

_PP

1)

RIS





NIRT

NIRT



EFM #7 EFM #8

Homogenization of reactor PP/EPR metallocene resin Comparing effects of TSE with SSE+EFM

EFM #7; P =2900 PSI

EFM #8 (TSE-1 + EFM Plates #4) P = 2200 PSI

EFM #9 (EFM #6 + Plates #4) P = 2300 PSI Relative number of particles original > 20 EFM #4  5.0 EFM #6 = 1.1 TSE  1.0 EFM #7 = 0.7 EFM #8 = 0.5 EFM #9 = 0.2 EFM #9

Film resin homogenization

MPS

=

matrix + dispersed phase + interphase

Matrix: Any polymeric substance

Dispersed phase:

Solids: clay, mica, metal, glass graphite, boron, carbon or metal oxide fibers or nanotubes, etc. Particles size in the Figure are (from top) d = 10, 80, 3300 and

58,000 nm; reinforcing limit is 50 nm

[Pukanszky, 1990].

Liquid: other (immiscible) polymer

Gas: a chemical or physical foaming agent

Interphase:

Solids: de-aggregation, delamination of clay stacks, adhesion between solid

particles and matrix in the molten and solid state – generated by wetting, chemical bonding, and mechanical interlocking.

Liquid: dispersion, stability of morphology, stress transfer

Compounding

10 nm 80 nm 3.3 m 58 m

no adhesion :

f

Agglomerate dispersion by fracture & erosion

(Manas-Zloczower et al., 1982):

Z  2 in shear, Z  1 in pure or biaxial elongation, and Z  0.5 in uniaxial elongation, k - geometry of the breakup process; (shear)/(elongation) = 4

Dry-blending of well-dried powders

Sufficient devolatilization/degassing

Dispersion progresses in stages: primary aggregates, secondary aggregates (by rupture), then individual particles (by erosion)

Provide low, increasing stresses,

& long residence time (strain)

Maximize extensional flow,

over that of shear

Distribution of particle size confirms the mechanism

0.01 1 0.1 10 F ra c ti o n o f p ar ti cl e s Particle size (m) #2 _#3 #1 #1 - individual particles, 150 nm #2 - secondary aggregates, 400 nm #3 - primary aggregates, 5 m 8 vol% of CaCO 3 in PP, batch mixing Suetsugu, 1994 ( _m ) / _ij / k Z k Z





 





Filler dispersion

Results of the C15A dispersion in 3-type of TSE & 8 screw

configurations correlate with the residence time and its distribution.

DD vs. residence time (t) correlated with r2 _{= 0.86}

When distribution of residence time

() was included: r2 _{= 0.98}

Thus, exfoliation depended not on stress, but on the residence time

For constant throughput the residence time is a measure of strain

Stress field is necessary

Exfoliation in extensional flow field should be at least 4x faster.





DD  4.37 exp 0.00972 t ; r  0.86

2 2

DD  4.2 0.21t 564   0.0017 t 4100 ; r  0.98

CORI = co-rotating, intermesh ICRR = intermesh, counter-CRNI = counter, non-intermesh

7 8 9 10 20 DD = 4.37 exp{0.00972t} D e gre e of dis pe rsion, D D CRNI ICRR ICRR CRNI CRNI ICRR CORI * * *   in elongation 4x faster???

data: Dennis et al., 2000

Melt exfoliation mechanism

Variation of the XRD with the mixing time for PS/Cloisite10A at 210°C under a shear rate of 65 (1/s) [Yoon et al., 2001].

Dennis et al., 2001

Thermal degradation of organoclay complicates the melt exfoliation

The advantage of a commercial EFM-3 for melt exfoliation of organoclay in a polymeric matrix was examined using three setups:

TSE (high shear and low shear screw configurations)

TSE + gear pump

TSE + gear pump + EFM-3

The system consisted of PET with 5 wt% of two organoclays: MMT-1 and MMT-2 [A. Garcia-Rejon & Y. Simard, 2001].

TSE-CORI with high-shear screw configuration was used, with a side-feeder for organoclay; the residence time varied from 50 to 240 sec. The extruded PNC was injection molded into ASTM #4 specimens for tensile tests.

Selected compositions were extrusion-blow molded into

500 ml bottles – significant improvement of hot-fill stability were

Hopper

Side feeder

Arrangement for ultrasound testing with silt die

Computer

Nanocomposite is extruded

Extensional Flow Mixer

Twin screw extruder

Gear pump

Hopper

Side feeder

Arrangement for ultrasound testing with silt die

Computer

Nanocomposite is extruded

Extensional Flow Mixer

Twin screw extruder

Gear pump

Standard compounder: Leistriz CORI-TSE,

f

= 34mm, L/D = 40.

Feed: matrix polymer + compatibilizer from hopper; organoclay from side feeder

Residence time controlled by the feed rate.

Gear pump (GP) provided constant pressure for EFM.

The results were repeated

at EFM Inc. (following

slide) for PET + MMT-1.

Generality of this

observation should be

validated using other

PNC systems and diverse

processing parameters.

Work is in progress.

Preliminary results indicate that to achieve

exfoliation: (1)

suitable organoclay must be used

,

and (2)

EFM leads to ca. 70% higher modulus

.

1 2 3 4

PET with organocly MMT-1 and MMT-2

MMT1: TSE+GP+EFM MMT1: TSE+GP MMT1: TSE MMT2: TSE+GP+EFM MMT2: TSE+GP MMT2: TSE R e lat ive Y o ung m o dul u s, E r error bar: ± 10%

EFM compounding 3

Organoclay MMT w/PET 0 1 2 3 4 5 6 7 8 50 100 150 200 250

Residence time, sec

R e la ti v e Y o u n g M o d u lu s EFM TSE+GP TSE

PET was compounded with 4 wt% organoclay.

TSE, TSE with gear pump (GP) and TSE + GP + EFM were used.

Relative performance is shown (data from EFM Inc.).

0 0.5 1 1.5 2 2.5 3

Flex strength Impact strength

R e la ti v e Str e n g th PET-MMT PET

EFM compounding 4

EFM increased the PNC

properties over that from TSE:

modulus by ca. 67% strength by ca. 60% Impact strength 80%.

PP compounded with proprietary organoclay & MA-g-PP in:

#1: TSE without EFM; x 50k – large stacks encircled. #2: TSE + EFM; x 200k – many short stacks.

#3: SSE + EFM; x 200k – few short stacks and exfoliation.

Better dispersion in SSE + EFM than in TSE + EFM is likely

due to higher thermal degradation in TSE

[UCL of ITRI, Taiwan]

.

#1

#2

#3

400 nm 400 nm

1.6 m

Effect of EFM gap

(2-wt% C15)

PA + 2-wt% organoclay

Medium shear TSE + GP + EFM

2 (degree) 2 4 6 8 10 In ten sity ( Cp s) 0 5000 10000 15000 20000 25000 30000 C 15A 5 30 60 250 1000 UBE1015C2 Interlayer spacing (nm) EFM gap (µm) Main Secondary 3.21 1.95/1.22 4.39 1.94 5.29 2.00 4.36 2.06 4.30 2.04 4.26 2.02

When dispersing 2-wt% C15A in PA-6, the effect of EFM is significant. For the smallest gap the main peak is very small, comparable to the behavior of the fully exfoliated UBE PA1015C2.

Summary of XRD data

Interlayer spacing (nm) Platelets/stack C15 wt% Specimen

(EFM gap) Main peak Secondary peak Main peak Secondary peak % Exfoliation 100 Cloisite 15A 3.21 1.95/1.22 3.34 3.23 - 4 Master-batch 4% 3.87 1.91 3.35 4.75 - 4 Stabilized master-batch 4% 3.93 1.92 3.40 5.31 10.41 4 TSE 3.97 1.94 3.16 4.87 20.57 4 TSE+GP 3.90 1.91 3.28 4.86 17.71 2 TSE 4.44 1.98 2.79 4.21 52.70 2 TSE+GP 4.42 2.02 2.89 3.80 64.74 4 (1000) 3.96 1.95 3.08 4.57 19.35 4 (250) 3.91 1.92 3.35 4.96 8.86 4 (60) 4.11 1.98 3.16 4.88 10.45 4 (30) 3.97 1.93 3.29 5.30 10.14 4 (5) 4.25 1.995 2.82 4.50 27.01 2 (1000) 4.26 2.02 2.92 4.20 56.10 2 (250) 4.30 2.04 2.86 4.16 60.84 2 (60) 4.36 2.06 2.83 4.02 65.35 2 (30) 5.29 2.00 2.83 4.84 76.03 2 (5) 4.39 1.94 2.78 2.53 83.33

d₀₀₁ decreases from 3.2 to 4.0 in 4% PNC, and to 4.3 - 5.3 nm in 2% PNC.

Common strategy for compounding MPS:

Compatibilization, sizing or intercalation

DISPERSIVE mixing, based on microrheology

DISTRIBUTIVE mixing, based on the theories of laminar flows

Extensional mixing in is superior to that in shear:

Local strains, increase of the interface, drop deformability, specific energy of mixing

Degradability of polymers, attrition of solid particles,

Inducing orientation in anisometric particle systems, etc.

MPS morphology depends on thermodynamics (miscibility),

flow and kinetics (non-equilibrium structures).

General purpose, static and dynamic extensional flow mixers (EFM and DEFM, respectively) have been developed for chemical, plastics, food, adhesives and other industries.

EFM has been successful for preparation of blends, elimination of ―fish eyes‖ in PO film, as well as for incorporation of organoclays.

The clay platelet must be treated as a large molecule, which must be chemically modified to induce miscibility, and then stress-dispersed. For successful melt exfoliation organoclay must be:

Miscible or suitably compatibilized with the matrix polymer. Thermally stable at the processing temperature

Preliminary data indicate that for melt exfoliation the use of EFM + SSE results in a better performance than that of TSE alone or TSE with EFM.

The improvement is probably due to:

Better mixing in EFM, viz. dispersion of aggregates & distribution Less degradation in EFM

Research on melt intercalation of PNC just started – there is a lack of

Conclusions 2