ICSAM Course Pissis

1

National TechnicalUniversity of Athens

Seminar on Nanomaterials(Short Course in Polymer Nanocomposites)

Polymer nanocomposites

1. Introduction2. Synthesis-Processing3. Properties-Applications

(a) polymer/silica nanocomposites(b) polymer/clay nanocomposites(c) polymer/CNT nanocomposites

We do not have the ambition to cover with this seminar the whole area of nanomaterials, we focus from the beginning on polymer nanocomposites.After an extended introduction into the topic, we discuss in Part 2 synthesis and processing of polymer nanocomposites. Processing has been here included in the title to stress its significance for determining and tuning the final properties of polymer nanocomposites. In Part 3 we discuss three families of polymer nanocomposites selected on the basis of their significance for the field in terms of applications. It is interesting to note that in these three families the filler is nano in 3, 1 and 2 dimensions, respectively. With nano we mean here and in the following between 1 and 100 nm.

2


Polymer nanocompositesIntroduction

1. Introduction2. Properties improvement3. Effects of interfaces

In this extended introduction we will focus on two points: properties improvement (with respect to the pure matrix) and the significance of interfacial effects for that improvement.

3


Dependence of tensile modulus E at 120°C on clay content for organo-modified montmorillonite and saponite-based nanocompositesY. Kojima et al., J. Mater. Res. 6 (1993) 1185-9


First report on improvement of polymer properties by incorporation of nanofillersA. Okada, M. Kawasumi, T. Kurauchi, O. Kamigaito, Polym.Prepr. (Am. Chem. Soc. Div. Polym. Chem.) 28 (1987) 447(from the Toyota research group)

The figure shown here is not from the original report but from a later publication by the same research group.The striking result in that report and subsequent reports (and in the figure shown here) is the significant improvement of polymer properties at very low filler contents, much lower than those required to achieve the same improvement with conventional (macro- or microscale) composites.

4


Number of published journal articles and issued and pending patentson nanotubes and nanotube/polymer composites as a function of yearM. Moniruzzaman, K. I. Winey, Macromolecules 39 (2006) 5194-5205


Searching in SCOPUS for polym nanocompo OR polymer inorganic hybrids yields more than 8.000 hits from journals

As you know (and this is reflected also in the scientific programme of this conference – ICSAM 2007) research in the field of polymer nanocomposites is rapidly growing all over the world.

5


silica

rubber

strain

stre

ss

L. Bokobza, J. P. Chauvin, Polymer 46, 4144-4151, 2005.

Properties Improvement in Nanocomposites (1)Mechanical properties

Guth & Gold equation:

G = G0(1 + 2.5φ + 14.1φ2)

G = G0 . X . Y


What kind of properties we typically seek to improve?First, mechanical properties.The figure shows the results of stress-strain measurements in natural rubber/silica nanocomposites. The filled symbols show results predicted on the basis of the experimental data for natural rubber and the Guth and Gold equation, a well established equation for conventional equations (where Go is the modulus of the pure matrix and fi the filler factor). This equation fails here, it underestimates the modulus. It is generally accepted that the additional improvement comes from interfacial effects (to be discussed later). An alternative to the Guth and Gold equation is then to consider two factors of improvement, X being the expression in brackets in the “classical” Guth and Gold equation and Y representing the additional improvement coming from interfacial effects.

6

National TechnicalUniversity of AthensPolymer nanocomposites

Introduction

Water permeability in poly(ε-caprolactone)/clay nanocompositesP. B. Messersmith, E. P. Giannelis, J. Polym. Sci. Part A Polym. Chem. 33 (1995) 1047-57

Properties Improvement in Nanocomposites (2)Barrier properties

A second property we would like to improve is related to packaging applications, namely barrier to diffusion of small molecules. Please note the low filler factors, permeability is reduced by a factor of 5 on addition of 5 vol% clays.

7


Predicted thermal longevity in polyimide/silica nanocomposites calculated from thermogravimetric dataV. A. Bershtein et al., J. Polym. Sci. Part B Polym. Phys. 40, 1056-69 (2002)


Properties Improvement in Nanocomposites (3)Thermal stability

A third property we would like to improve is thermal stability. Here we show results for the predicted thermal longevity of polyimide/silica nanocomposites, calculated on the basis of thermogravimetric data and a model for thermal degradation. We observe improvement of thermal stability up to about 200 C with respect to the pure matrix.

8


MWCNTs in PCEvaluation of dc conductivity in terms of percolationσdc(p) ~ (p- pc)t

Properties Improvement in Nanocomposites (4)Electrical propertiesHere we make use of the good properties of the filler

P. Poetschke et al., Polymer 44 (2003) 5023-30


So far we discussed examples of improved properties of the polymer matrix. In this example we utilize the good properties of the filler.In the figure we observe a transition from an insulating material to a conductive one at a low filler content, when a conducting network of carbon nanotubes(CNTs) (here multiple walled carbon nanotubes – MWCNTs) is formed.In the equation coming from percolation theory p is the filler content (actually in vol%), pc the percolation threshold and t an exponent related with dimensionality. From percolation theory we know that the percolation threshold for a homogeneous distribution of spherical inclusions in three dimensions is about 16 vol%. The low percolation threshold of only 1.5 wt% (which means less than 1.5 vol%) here is related to the large aspect ratio of CNTs.Other properties of nanofillers which are often utilized in polymer nanocomposites include magnetic and optical ones.

9


Introduction

Properties Improvement in Nanocomposites (5)Applications

Timing belt covered with injection molded nylon-6/clay nanocomposites M. kawasumi, J. Polym. Sci. Part A Polym. Chem. 42 (2004) 819-24

There are already several commercial applications of polymer nanocomposites, two examples being given here and in the following slide.

10


Introduction

Properties Improvement in Nanocomposites (6)Applications

Thermoplastic polyolefine nanocomposites, applications for automotive partF. Hussain et al., J. Compos. Mater. 40 (2006) 1511-75

The second example, applications for automotive parts.

11


IntroductionProperties Improvement in Nanocomposites (7)Size effects (classical)

(a) Size dependence of the meltingpoints of Sn particles(b) Size dependence of the normalized heat of fusion of Sn particlesClassical size effect

Lai et al., Phys. Rev. Lett. 77 (1996) 99

How to calculate the properties of a polymer nanocomposite from those of the matrix and the filler? We often use classical theories we know from conventional composites. We have to take, however, some additional points into account. We will consider only two here: size effects and interfacial effects.An example of size effect is given here. The reduction in melting point and heat of fusion with decreasing particle size below about 50 nm is obviously due to the fact that an increasing number of atoms is on the surface, i.e. more free. This is a classical size effect.

12


Photoluminescence blue shift versus polymer dielectric constant at a frequency of 60 Hz. PSi-polymer nanocomposites were produced (74% porosity) by diffusion of the polymers for a period of 2-3 days. Initial peak PL energy before infiltration was 1.63 eV.H. A. Lopez et al., J. Luminescence 80 (1999) 115-9


Properties Improvement in Nanocomposites (8)Synergy

An example of quantum size effect is photoluminescence of Si nanocrystals: a shift to higher energy of luminescence (blue shift) with decreasing crystal size. In the figure, by filling the pores of nanoporous silica with different polymers we achieve a fine tuning of blue shift depending on the dielectric constant of the polymer (which modifies the applied electromagnetic field). This is a good example of synergy in a composite material.

13


Effects of interfaces (1)

Polymeric chains close (a few nm) to a solid surface: changes instructure (density, chain configuration)phase transitionsmolecular mobility

The properties of the interface affect or dominate over the bulk properties

new properties or combination of properties interesting for applications(fundamental physics) effects close to interfaces become bulk properties


G. C. Papanicolau et al., Coll. Polym. Sci., The concept of the boundary interface in composite mechanics, Coll. Polym. Sci. 256 (1978) 625-30G. C. Papanicolau et al., Compos. Interf. 14 (2007) 131-52

Now to interfacial effects, a point we have to consider if we would like to predict the properties of polymer nanocomposites (see slide 5).Interfacial effects, arising from the presence of an interfacial layer around the filler particles with different properties than the bulk (interphase, mesophase), are significant also for conventional composites. They become more significant for nanocomposites, due to the larger surface to volume area, as we discuss in the following.

14


Schematic drawings of microstructural appearance of typical-particulate vs. fine-particulate vs. nano-particulate composites based on electronic microscopic observations: (a) 3 vol% of particles with 10 μm diameter (2.86 particles within a volume of 50 000 cubic μm); (b) 3 vol% of particles with 1 μm diameter (2860 particles within a volume of 50 000 cubic μm); and (c) 3 vol% of particles with 0.1 μm (100 nm) diameter (2.86 million particles within a volume of 50 000 cubic μm)M. Z. Rong et al., Polymer 42 (2001) 3301-4



We consider a volume of a sample of 50.000 cubic microns with 3 vol% of spherical filler. We have then in that volume about 3 particles if their mean diameter is 10 microns, about 3.000 particles if their mean diameter is 1 micron, and about 3.000.000 particles if their mean diameter is 0.1 micron. Is that not impressiv?

15



M. R. Bockstaller et al., Adv. Mater. 17 (2005) 1331-49

Calculated interfacial area per volume of particles (in 1/nm), assuming a right –circular cylindrical particle shape, for different particle diameters and aspect ratios


Now we allow for the aspect ratio, which is 1 for spherical fillers, to change, i.e. we consider clays and nanotubes. The calculated interfacial area per volume of particles becomes very high for clays.

16


Papakonstantopoulos et al., Phys. Rev. E 72 (2005) 031801Simulation of local mechanical properties reveals that a glassy layer is formed in the vicinity of an attractive filler, contributing to the increased strength of the composite materialLocal shear modulus as a function of the distance from the center of the nanoparticlefor the three types of interaction in (a) and for the attractive particle in the melt and the glassy regime in (b)



No additional comments are necessary here.

17


CNTs in PVA, Polymer crystallinity (left) and rate of increase of modulus with volume fraction against rate of increase of crystallinity with volume fraction (right)Correlation between polymer ordering and reinforcement – the mayor role played by the CNTs in improving the mechanical properties of composites is to nucleatean ordered polymer coatingColeman et al., Polymer 47 (2006) 8556




18


Effects of interfaces (6)Computer simulations

Systems investigated: polymer nanocomposites, liquids confined in pores,liquids close to a surface -> similar results

Reduction or increase of mobility depending onstrength of interaction with the surfaceroughness of the surface

Gradual variation of mobility, effect of temperature

rough surface smooth surface


Computer simulations are an attractive tool to calculate molecular mobility of a polymer or a liquid near a solid surface.Here calculations of the relaxation time as a function of the distance from a rough and a smooth solid surface. Please note the gradual variation of tau and the effect of temperature.

19


Polymer nanocompositesSynthesis-processing

1. Introduction2. Mixing with preformed particles3. Sol-gel techniques4. Using nanobuilding blocks5. Nanoengineering of composite materials

How to synthesize polymer nanocomposites?Typically polymer and inorganic filler do not like each other – they are incompatible. So, the main task is how to render them compatible, i.e. how to distribute the nanoparticles in the matrix. Typically (however, there are exceptions!) we would like to have a homogeneous distribution of nanoparticles. A big effort is made in the field of polymer nanocomposites to optimize and control the dispersion of nanoparticles – processing is very significant in that respect.

20


Synthesis – processingIntroduction

Two approaches for device fabrication where nanoparticle properties can be exploited

1. “top-down” approachphotolithography, electron-beam lithography(microelectronics)

2. “bottom-up” approachself-assembly processescontrol of structural arrangement of nanoparticles and of morphology

Going a step back, there are in general the two above approaches.The second one is going to be our approach.

21


Different kinds of inorganic-organic composite materialsa) Embedding of the inorganic moiety into the organic polymerb) Interpenetrating networks (IPNs)c) Incorporation of inorganic groups by bonding to the polymer backboned) Dual inorganic-organic hybrid polymerG. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingIntroductionFor a given composition, properties are determined to a large extent by the final morphology

Here a rough classification of morphologies in terms of interpenetrating networks.We use the term hybrid in the case of chemical bonds between the filler and the polymer. So, C is a hybrid nanocomposite and d a hybrid dual inorganic-organic polymer.

22

National TechnicalUniversity of AthensSynthesis-processing

Introduction

Three routes for preparing polymer nanocomposites

1. Mixing with preformed particles2. Sol-gel techniques to generate the particles in-situ3. Incorporation of nanobuilding blocks (clusters) into the polymer matrix

In route 1 the main difficulty is to control dispersion of nanoparticles and morphology. Sol-gel techniques are very good in that respect, however reactions are often difficult to control and there are problems also with scaling up in industry. Route 3 is then a compromise.

23

National TechnicalUniversity of AthensSynthesis-processing

Mixing with preformed particles

Mixing with preformed particles

Mixing in solution (dispersion), i.e. use of a common solvent, which will be then evaporated, and melt-mixing

Mixing with the monomer followed by in-situ polymerization

Typically monomer and filler do not like each other, so there is a need for Functionalization (chemical treatment) prior to mixing

Conditions of mixing and processing very significant for the final properties


24


Mixing with inorganic layered 2D materials (natural or synthetic)

G. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingMixing with preformed particles

This is an example of mixing with preformed particles (here natural or synthetic). Depending on the type of the layered inorganic material and the polymer and on the conditions of mixing, different morphologies are obtained. We will come back to these morphologies later with respect to applications, here it is just an example.

25


Formation of metal oxide frameworks by sol-gel techniquesG. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingSol-gel techniques

Sol-gel techniques were known before for preparing ceramic particles and powders of high purity and with less pollution to the environment than solid state chemistry. The two steps, hydrolysis of metal alkoxides and polycondensation, are schematically shown here.

26


Silicon alkoxides possessing polymerizable alkoxy groups were used and the alkoxy groups were liberated during the sol-gel processing as the corresponding alcohol. Aqueous ring-opening metathesis polymerization (ROMP) as well as free radical polymerization were used for the organic polymerization reaction. G. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114


Simultaneous formation of interpenetrating networksThree processes in competition: (a) the kinetics of formation of the inorganic phase, (b) the kinetics of polymerization of the organic phase, (c) the thermodynamics of phase separation between the two phases

It is possible to generate the inorganic particles and the polymer simultaneously. For that we combine the two precursors in one, two examples are given here.

27




Dual network precursorsThe network forming functionalities can also be covalently connected to each other. One way to realize that is with the functional groups for the organic and inorganic polymerization/polycondensation process being incorporated in one molecule (commercially available molecules GLYMO or MEMO, Scheme below). Another possibility is that functional groups for the second polymer are incorporated in a preformed polymer of the other type

Here we would like to develop the inorganic and the organic network with chemical bonds between them.

28


Functionalization of a preformed network or polymer with pending functional groups capable of forming the second network and use of end-capped polymersG. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114


Dual network precursors

Examples for that are given here.

29


Different paths for obtaining hybrid materials. Conventional sol-gel route (path A), use of templates capable of self-assembly (paths B, D), assembly of nanobuilding blocks (paths C, D)C. Sanchez et al., Chem. Mater. 13 (2001) 3061-83

Synthesis-processingUsing nanobuilding blocks

Here different paths for obtaining hybrid materials.A sol-gel techniques.B, D using of organic templates (structure directing agent), like block copolymers, to control morphology, not further discussed here (we will come back to that later).C using nanobuilding blocks.In the following we focus on nanobuilding blocks.

30


Examples include modification of commonly used inorganic fillers or particles but also commercially available polyhedral oligomericsilsequioxanes (POSS) nanoclustersG. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingUsing nanobuilding blocksSurface modification with polymerizable groups

Here an example of a nanobuilding block: commercially available POSS, which are compact structures of nm size, typically with organic substituents. We may replace a part of the organic substituents R by reactive ones, R’, and do chemistry.

31


Depending on the number of reactive functional groups available on the surface linear or crosslikned systems can be preparedG. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingUsing nanobuilding blocksSurface modification with polymerizable groups

Other examples of surface functionalization are schematically given here.

32


Initiating groups attached to the inorganic core allow the grafting of polymers from the surface of the inorganic moiety. This approach has some advantages over the more often used “grafting to” technique, where end-functionalized polymers are grafted to the surface of inorganic particles, e.g. leading to restrictions in the surface coverage.G. Kickelbick, Prog. Polym. Sci. 28 (2003) 83-114

Synthesis-processingUsing nanobuilding blocksSurface modification with initiating groups

Here another possibility of surface modification with functional groups.The concept of grafting from the surface (by using initiating groups) against the concept of grafting to the surface. We have a better arrangement of polymer around the particle in grafting from the surface.

33


Schematic representation of the various possible strategies that can be used to assemble functionalized tin-12 clustersC. Sanchez et al., Chem. Mater. 13 (2001) 3061-83

Synthesis-processingUsing nanobuilding blocks

Here a summary of different functionalization strategies at the example of tin clusters.

34


Aggregate size control using recognition-functionalized block copolymers. a) Schematic representation and TEM images demonstrating the control of aggregate core and corona size using different diblock lengths. b) Comparison between aggregate core size (from TEM) and hydrodynamic radius (from DLS).R. Shenhar et al., Adv. Mater. 17 (2005) 657-69

Synthesis-processingNanoengineering of composite materials3D control of aggregate shape and size

Two slides to nanoengineering. The term refers to tailored-made synthesis using block copolymers as templates. The particles prefer one of the blocks. The size of the blocks determines the size of particle aggregates.

35


Schematic representation of the formation of structured nano- or mesoporous materials by a self-assembly process of block copolymers


Synthesis-processingNanoengineering of composite materials

Here the second example, starting again with diblock copolymers.The template can be removed in the last step by heating or by using a solvent.

36


Polymer nanocompositesProperties-Applications(a) polymer/silica nanocomposites1. rubber/silica – mechanical properties2. polyimide/silica – low-k materials3. PHEA/silica hydrogels - biomaterials

Now to the first of the three families of polymer nanocomposites which will be discussed in some detail.In some of these and in other applications silica can be replaced by other oxides, such as alumina, titania etc. However, silica is much more used than any other oxide.

37


PDMS/silica nanocomposites

Polydimethylsiloxane (PDMS)

OH OH| |

– O – Si – O – Si – O -| |

Silica nanoparticles prepared by sol-gel techniques in the presence of crosslinked PDMS,diameter ~10 nmVery good dispersionHydrogen bonding between polymer and filler

L. Dewimille, B. Bresson, L. Bokobza, Polymer 46 (2005) 4135-43

Silica particles are here generated by sol-gel techniques in the presence of crosslinked PDMS. The dispersion is by far much better than in the case of mixing preformed silica with the monomer and then polymerization and crosslinking or mixing preformed silica with the polymer and then crosslinking.

38


Natural rubber = cis-1,4-polyisopreneVulcanization with sulfur

No strong interaction between matrix and filler

Two series of samplesSeries A Series Bafter crosslinking before crosslinkingGood dispersion Aggregation

Natural rubber (NR) / silica nanocomposites

L. Bokobza & J. P. Chauvin, Polymer 46, 4144-4151, 2005.

Here a second example, where PDMS has been replaced by natural rubber.We compare two series of samples with each other. Dispersion of silica nanoparticles in the polymer matrix is less good if we generate the nanoparticles before crosslinking. The reason is that in this case there is more space for the nanoparticles to grow and aggregate.

39


Natural rubber (NR) / silica nanocomposites

Stress–strain curves for the unfilled NR and NR films filled withsilica after the cross-linking process.L. Bokobza & J. P. Chauvin, Polymer 46, 4144-4151, 2005

Other mechanical properties are also improved.An example is shown here, these materials being investigated with respect to applications in the tires industry, where carbon black is being replaced by silica.

40


Polyimide/silica nanocomposites

• Increasing demand in microelectronics for low-k materials, i.e. materials with low ε′, to replace SiO2 as inter-metal and inter-layer dielectrics (IMD, ILD).

• Candidate materials should combine low ε′ values (below 3.0, possible even below 2.5) with several other good properties.

• Aromatic polyimides (PIs) are interesting materials for such applications.• However ε′ of the starting material (in the range 3.2-3.5) should be further

reduced (by keeping the other good properties).• Two possible ways to achieve that is by introducing porosity and by

reinforcing with inorganic nanoparticles.• Here we combined these two ways: we prepared hybrids of polyimide (PI)

and porous organosilicon nanophase (ON) and investigated their dielectric properties

D. Fragiadakis et al., J. Phys. Conf. Series 10 (2005) 139-42V. Yu. Kramarenko et al., Polym. Adv. Techn. 15 (2004) 144-8

No comments are necessary here.

41


Polyimide/silica nanocomposites

Preparation – structure Anticipated structure

• From polyamic acid of molar mass 5.000 (series 5) or 10.000 (series 10) or 15.000 (series 15) with ethoxysilane end groups (PAAS) and methyl triethoxysilane (MTS).

• Sol-gel techniques• The PAAS/MTS mass ratio was

systematically varied from 100/0 to 100/120 corresponding to PI/ON mass fractions varying from 100/0 to 64.4/35.6

V. Yu. Kramarenko et al., Polym. Adv. Techn. 15 (2004) 144-8

The structure has been confirmed by IR spectroscopy.

42

National TechnicalUniversity of AthensPolyimide/silica nanocomposites

Dielectric relaxation spectroscopy (DRS)

10-1 101 103 105 107

3,0

3,2

3,4

100/80 - 10

100/70 - 10

100/60 - 10

ε'

frequency (Hz)

100/8 - 10

Frequency dependencies of the r.t. dielectric permittivities for the samples indicated onthe plot for series 10.

Frequency dependencies of the r.t.dielectric permittivities for the PNs with PAAS/MTS ratios from 100/0 (uppermost curve) to 100/120 (lowermost curve) for series 5.

D. Fragiadakis et al., J. Phys. Conf. Series 10 (2005) 139-42

Measurements were performed at room temperature.At high frequencies we observe a drop of the dielectric constant, it is due to the secondary gamma relaxation , arising from the imide cycles.

43


Table. Experimental and calculated dielectricpermittivities of nanocomposites

εtot= εtot(εi, εm,...)

•Series 5: three EM models to calculate the dielectric permittivity of the ON

•Looyenga model ε′1/3= εm′ 1/3 (1-φ) + εi′ 1/3φ , (1a)

•Lichtenecker model lnε′= lnεm′(1-φ) + lnεi′ 1/3φ , (1b)

•Stoltze et al.′s model ε′a = εm′a (1-φ) + εi′a φ (1c)

polyimide/silica nanocomposites

effective medium theory (EMT) analysis

V. Yu. Kramarenko et al., Polym. Adv. Techn. 15 (2004) 144-8

EMT has been used to calculate the dielectric constant of the inclusion (silica) from the measured dielectric constant of the composite (effective dielectric constant) and the measured dielectric constant of the matrix. This is a way to calculate, under some assumptions, the dielectric constant of a dispersed phase. We have chosen a middle frequency (1 kHz) for that calculation, i.e. we used values measured for the composite and the matrix at 1 kHz. The result is that we get values for the dielectric constant of ON around 2, different for the different models (this is obvious) but with the same trend. This is much less than the dielectric constant of bulk silica, which is around 4. The conclusion is that our ON generated by sol-gel in the nanocomposites is porous, i.e. a mixture of compact silica (dielectric constant 4) and air (dielectric constant 1). This result was confirmed by density measurements not shown here.

44

National TechnicalUniversity of AthensPHEA/silica hydrogels

Introduction: PHEA matrix

► PHEA Hydrogels

PHEA – poly(hydroxyl ethyl) acrylate .........

Bulk polymerization of 2-hydroxyethyl acrylate and ethylenglycol dimethacrylate (as crosslinking agent)

Good water sorption/diffusion properties PHEA accommodates about 1/4 to 1/3 of its dry mass

as bound (non-freezable) water, it becomes however very soft

Kyritsis et al. J Polym Sci Part B Polym Phys 33 (1995) 1737Kyritsis et al. Polym Gels Networks 3 (1995) 445

The next example is a hydrogel.PHEA is an interesting material for applications in medicine and biotechnology: materials for implantation, scaffolds for tissue engineering, drug delivery systems etc. For some of these applications (e.g. the first two here) mechanical stability has to be improved.

45


Techniques: SEM, TGA, DMA, water sorption-desorption, DRS, TSDC

PHEA/silica hydrogels

Reinforcement: Silica nanoparticles

The mechanical properties of polymer nanocomposites may be significantly improved with respect to those of the pure matrix. It is assumed that the large surface to volume ratio of the nanoparticles plays a significant role

The properties are to a large extent determine by the morphology (degree of dispersion of the nanoparticles)

► Inclusion of silica nanoparticles

J. A. Gomez Tejedor et al., J. Polym. Sci. Part B Polym. Phys. 46 (2007) 43-54J. G. Rodriguez Hernandez et al., Eur. Polym. J. 43 (2007) 2775-83

PHEA/silica hydrogels were studied by several techniques listed here. Some of the results obtained are also listed here.

46


PHEA/silica hydrogels - Preparation

OHHCOHSiOHHCOSi 52252 +−→+−−≡

1. Polymerisation of PHEA

2. Hydrolysis of TEOS

3. Condensation

and/or

f= 0% to 30% in silica

SEM: excellent dispersion

TGA: Nanoparticles form a continuous network

=> An organic and an inorganic network combined

OHSiOSiSiHOOHSi 2+≡−−≡→≡−+−≡

OHHCSiOSiSiOHCOHSi 5252 +≡−−≡→≡−−+−≡

OH

OE t

OHHO

3nm

domains 30-50 nmdomains < 400 nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

OH

OE t

OHHO

3nm



domains 30-50 nmdomains 30-50 nmdomains < 400 nm

organic polymer chains

silica tunneled nanoparticles

OH

OE t

OHHO

3nm


OH

OE t

OHHO

3nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

OH

OE t

OHHO

3nm




OH

OE t

OHHO

3nm


OH

OE t

OHHO

3nm


OH

OE t

OHHO

3nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

OH

OE t

OHHO

3nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

3nm

OH

OE t

OHHO

OH

OE t

OHHO

3nm






organic polymer chains

silica tunneled nanoparticles


Polymerization of PHEA and hydrolysis/condensation occur simultaneously. TEOS is the precursor of silica.From the results of scanning electron microscopy (SEM) and atomic force microscopy (AFM) measurements we conclude on the structure shown here.

47

National TechnicalUniversity of AthensPHEA/silica hydrogels

Dynamic Mechanical Analysis

-150 -100 -50 0 50 100 150 200

106

107

108

109

1010

E´

at 1

Hz

T (C)

Silica Content 0%15%30%

At 1 Hz: glass-rubbery transition at 10-20 C

• mechanical enhancement by 3 o.o.m. compared to pure PHEA!


Dynamic mechanical analysis (DMA) measurements.In the rubbery phase we observe a significant increase of E’ (the rubbery plateau). This result confirms the formation of a network of silica nanoparticles, observed on the same samples by thermogravimetric analysis (TGA).

48

0.00.10.2-0.050.000.050.100.150.200.250.300.350.400.45Hydr Silic 0.050.100.150.200.30

The results here suggest that the hydrophilicity of the matrix is retained, which can be explained randomly distributed in the nanocomposite. The steep increase of water content at higher values

49


Polymer nanocompositesProperties - Applications(b) polymer/clay nanocomposites

1. Morphology2. Synthesis3. Mechanical properties4. Barrier properties

We move now to the second family of polymer nanocomposites, polymer/clay nanocomposites.

50


Structure of 2:1 phyllosilicates

S. Sinha Ray, M. Okamoto, Prog. Polym. Sci. 28 (2003) 1539-1641

polymer/claymorphology

A typical structure of a layered silicate is shown here: platelets of a thickness of about 1 nm separated by galleries, where cations are located. They can be exchanged, e.g. with organic cations, so that the clay, which is originally hydrophilic, becomes organophilic.

51


M. Alexandre, P. Dubois, Mater. Sci. Eng. 28 (2000) 1-63


Depending on the method and the conditions of preparation we may have one of the three distinct morphologies shown here. In practice we have often mixtures of these morphologies.

52


XRD patterns of: (a) phase separated microcomposite (organo-modifiedfluorohectorite in a HDPE matrix); (b) intercalated nanocomposite (sameorganomodified fluorohectorite in a PS matrix) and (c) exfoliatednanocomposite (the same organo-modified fluorohectorite in a siliconerubber matrix) M. Alexandre, P. Dubois, Mater. Sci. Eng. 28 (2000) 1-63


How to find out which is our morphology?Typically with X-ray measurements taking advantage of the crystallinity of the clays. In the case of immiscibility (phase separation) no change in scattering angles (distance between the platelets) is observed. In case of intercalation the peaks shift to lower angles as the distance between the platelets increases. Finally, in the case of exfoliation the peaks disappear, as no crystal order is present. Scanning electron microscopy (SEM) and, in particular transmission electron microscopy (TEM), provide also information on morphology.

53


Schematic drawing of the preparation of polyurethane/clay nanocomposites. Theblack circles represent solvent molecules; the lines, PUU chains; and the blocksa single silicate layer. The relative size of the blocks are for illustrative purposesonly and the relative dimensions do not represent those of a single silicate layer( ~1 nm thick and having an aspect ratio of ~200-500). R. Xu et al., J. Biomed. Mater. Res. 64A (2003) 114-9

polymer/claySynthesis (solution intercalation)

This is the method of preparation used in the first years after the first report on polymer/clay nanocomposites appeared (1987).

54


Schematic illustration of nanocomposite synthesis (PDMS/MMT)

S. D. Burnside, E. P. Giannelis, Chem. Mater. 7 (1995) 1597-1600

polymer/claySynthesis (melt intercalation/processing)

The next milestone was the observation in 1993 that it is possible to melt-mix polymers and layered silicates.Here an example of synthesis of crosslinkedpolydimethylsiloxane/montmorillonite (PDMS/MMT) nanocomposites by melt intercalation.

55


Nylon-6 nanocomposite formed through in situ polymerizationwith 12-aminododecanoic acid (ADA)–MMT

http://www.nanocor.com

polymer/claySynthesis (in situ polymerization)

An alternative method is polymerization in situ. The first step consists of rendering the clay organophilic, i. e. the preparation of the organically modified clay. Surfactants are used to that aim.The second step consists of intercalation of the monomer, assisted by the presence of the surfactant. During the third step the monomer is polymerized. A problem with this method is that the optimal conditions of polymerization are usually different than for bulk polymerization and have to be found out before.

56


Effect of clay content on tensile modulus, measured at roomtemperature, of organomodified montmorillonite/nylon-6-based nanocomposite obtained by melt intercalation

L. M. Liu et al., J. Appl. Polym. Sci. 71 (1999) 1133-38

polymer/clayMechanical properties

In clay nanocomposites we have a large surface to volume ratio, so we expect a significant improvement of mechanical properties already at low filler factors (exfoliated morphology). The results confirm that expectation. At higher filler factors a saturation is observed, probably due to a change of morphology (no more exfoliation possible).

57


Trend of the storage modulus E′ at 25°C for SBS-basednanocomposites (□□) and microcomposites (▪▪) as a function of thefiller level(SBS = symmetric styrene-butadiene-styrene block copolymer)M. Laus et al., J. Mater. Res. 12 (1997) 3134-9


Improvement of mechanical properties depends strongly on on morphology.Nanocomposite here means that we have intercalation and/or exfoliation. Microcomposite means that we have a phase separated morphology, see slide 51.

58


Dependence of tensile modulus E at 120°C on clay content fororgano-modified montmorillonite and saponite-basednanocomposites

Y. Kojima et al., J. Mater. Res. 6 (1993) 1185-9


Improvement depends also on the surface to volume ratio of the clay.

59


Variation of elastic modulus as a function of particle content for extended graphite (EG) and for clay

A. Yasmin et al., Comp. Sci. Technol. 66 (2006) 1182-9


In addition to clays, other layered materials, natural or synthetic, can be used as fillers with similar or even better results, e.g. extended graphite.

60

National TechnicalUniversity of Athenspolymer/clay

Barrier properties

Water permeability in poly(ε-caprolactone)/clay nanocomposites

P. B. Messersmith, E. P. Giannelis, J. Polym. Sci. Part A Polym. Chem. 33 (1995) 1047-57

Good barrier properties to small molecules are typical for clay nanocomposites and form the basis for applications in food packaging.

61


Oxygen gas permeability of neat PLA and various nanocomposites as a function of OMLS content measured at 20 °C and 90% relative humidity. The filled circles represent the experimental data. Theoretical fits basedon Nielsen tortuousity model (J. Macromol. Sci. Chem. A1 (1967) 929-42).S. Sinha Ray, M. Okamoto, Prog. Polym. Sci. 28 (2003) 1539-1641

polymer/clayBarrier properties

OMLS means organically modified layer silicate. The fit is based on a model by Nielsen and the fitting parameter is the aspect ratio of the platelet.

62


Formation of tortuous path in PLS nanocomposites

S. Sinha Ray, M. Okamoto, Prog. Polym. Sci. 28 (2003) 1539-1641


PLS means polymer/layered silicate.The model s schematically presented here. The diffusing molecules can not go through the platelets, so their path through the nanocomposite becomes longer and diffusivity and permeability decrease.

63


Relative H2O vapor permeability for PU/clay nanocomposites. The solid lines represent the theoretical value for aspectratios =300 and 1000 (α is the aspect ratio, φ the volume fraction and μ a geometrical factor.

R. Hu et al., Macromolecules 34 (2001) 337-9


The experimental data indicate a decrease of the aspect ratio with increasing filler content, which is compatible with a gradual change of morphology from exfoliated at low filler content to intercalated and phase separated at higher filler content. Such a change can be confirmed by morphological characterization (X-rays and/or TEM).

64


Clay length dependence of the relative permeability coefficient forwater in poly(imide)/clay nanocomposites

K. Yano et al., J. Polym. Sci. Part A Polym. Chem. 35 (1997) 2289-94


In agreement with the model we get an improvement of barrier properties with increasing length of clay.

65


Polymer nanocompositesProperties - Applications(c) polymer/CNT nanocomposites

1. Introduction2. Preparation/processing3. Electrical properties (percolation)

Now to the last of the three families of polymer nanocomposites. The filler here is carbon nanotubes (CNT), which is nano in two dimensions (compare slide 1).

66


Number of published journal articles and issued and pendingpatents on nanotubes and nanotube/polymer composites as a function of year

M. Moniruzzaman, K. I. Winey, Macromolecules 39 (2006) 5194-5205

Introduction

Please note that CNT are intensively investigated also in microelectronics and other fields of research.

67


Schematic of a: (a) single-wall and (b) multiwall nanotube

F. Hussain et al., J. Comp. Mater. 40 (2006) 1511-75

Introduction

No comments.

68


Schematic diagram showing how a hexagonal sheet ofgraphene is “rolled” to form a carbon nanotube. The rolling shown inthe diagram will form a (3,2) nanotube.

M. Moniruzzaman, K. I. Winey, Macromolecules 39 (2006) 5194-5205

Introduction

No comments.

69


Tensile strength comparison of common engineering materials

F. Hussain et al., J. Comp. Mater. 40 (2006) 1511-75

Introduction

CNT possess excellent mechanical properties, the question is how to make use of these good properties in composite materials and in devices.

70


Preparation/processing methods for the production of polymer/CNTnanocomposites:Suspensions of nanotubes in polymer solutions, preparation as thin filmsIn-situ polymerization in presence of nanotubesMelt mixing of nanotubes with polymers (advantages: speed, simplicity, availabilityin the plastic industry, free of solvents and contaminants)

Main interesting property of CNT for composite application:→ Electrical conductivity In order to transfer this property in an insulating matrix (polymer)formation of a nanotube network required

CNT network should be:→ free of CNT agglomerates (agglomerates reduce amount

of CNT which can contribute to the network formation, act as defect in mechanical behavior)

→ good dispersion of single tubes (or small bundles in caseof MWCNT)

→ 3D network with contacts between CNT (within hoppingor tunneling distance of ~ 3-10 nm)

Preparation/processing

By courtesy of Dr. Petra Poetschke, IPF Dresden (Germany)

From the three methods for preparation of polymer/CNT nanocomposites given here melt mixing is that most widely used.

71


Starting from a masterbatch

Highly concentrated batch of polymer with 15-20 wt% CNTs (MWCNTs)Commercially available , i.e. Hyperion Catalysis Intern. Cambridge, USA

Direct incorporation

Starting from premixtures of polymer and nanotube powders,functionalization

Wetting of CNT by polymer• surface characteristics • interfacial tension polymer -CNT• melt viscosity of polymer

Distribution Dispersion

pure polymer

CNT

Tasks

pure polymer

masterbatch

nanocomposite

Preparation/processing (Incorporation of CNTs into a polymer by melt mixing)


Two different routes are mainly used for preparing polymer/CNT nanocomposites by melt mixing: (1) starting from a masterbatch and (2) by direct incorporation. Masterbatches (with typically 15-20 wt% CNT) are commercially available. They can be diluted with the same or a different polymer.

72


10-2 10-1 100 101 102 103101

102

103

104

105

106

PP OREVAC PPC PP OREVAC PPC, 1,2 wt% MWNT PP OREVAC PPC, 2,7 wt% MWNT PP OREVAC PPC, 4 wt% MWNT PP OREVAC PPC, 8 wt% MWNT

stor

age

mod

ulus

G´ [

Pa]

frequency [s-1]

Preparation/processingRheological measurements – MA PP/MWCNT series

Storage modulus (Fig.) and viscosity increase with CNT addition. At 190°C first clear rheological indication of percolation is seen starting at 8 wt% MWNT. At 4 wt% already a small increase in η* and G’ at low frequencies can be observed. This concentration is much higher than detected in electrical measurements (already percolated at 2.7 wt%)

Rheological measurements in the melt allow to follow in-situ the morphology, in particular the dispersion of CNT (and nanoparticles in general). The percolation here refers to the formation of a continuous network of CNT, which affects significantly the mechanical properties. In addition it affects also the electrical properties (see slide 73), there are however differences between the two percolations. The in-situ monitoring of the changing morphology is essential for optimizing the processing conditions, see also slide 78.

73


σ′(ω) = σ(0) + σac(ω) = σdc + Aωs (Universal Dynamic Response)A. K. Jonscher, Dielectric Relaxation in Solids, Chelsea Dielectric Press, London, 1983

σac(f)=2πfε0ε′′(f)

Electrical propertiesPA6/MWCNT nanocomposites

In contrast to the previous slide, the measurements here refer to the final nanocomposite, e.g. prepared by compression molding. Ac conductivity is obtained from dielectric measurements, see the equation on the slide. We observe a transition from non-conductive (insulating) to conductive state. This transition is related with connectivity of the conducting filler (percolation). The transition is manifested by the appearance of dc conductivity, i.e. ac conductivity becomes independent of frequency. The plateau values give the dc conductivity. The transition occurs here between 2.5 and 5.0 wt% CNTs. With increasing CNT content dc conductivity increases and the plateau is extended to higher frequencies. At each composition, the frequency dependence of ac conductivity is described by the equation given on the slide.

74

National TechnicalUniversity of AthensElectrical properties

Percolationin 3D

R. Zallen, The Physics of Amorphous Solids, Wiley, New York, 1983

This is an experiment to follow percolation in three dimensions. We mix conducting (metal, Al) and insulating (glass) spheres in a glass cylinder, apply electrodes (metal, Al) and a voltage and measure the current through the system. If we do that by systematically increasing the volume fraction of conducting spheres and mixing well each time, we will get the result that the system becomes conducting at about 16 vol% conducting spheres. 16 vol% is also the result we get by applying probability theory. This means that (statistically) the first conducting way is formed at about this composition. This is the percolation threshold.

75

National TechnicalUniversity of AthensElectrical properties

Percolationin 2D

R. Zallen, The Physics of Amorphous Solids, Wiley, New York, 1983

This is an experiment in two dimensions. The result we get is that the percolation threshold is about 50%. We could get the same result with the experiment of the previous slide by replacing the glass cylinder with a table surface (two-dimensional distribution of conducting and insulating spheres). We can easily understand why the percolation threshold is higher in two dimensions: less chance for the charge carriers to escape if they are confined in two dimensions. We can also easily predict what is going to be the percolation threshold in one dimension (a line of spheres), obviously 100%.

76


D. Stauffer and A. Aharony, Introduction to Percolation Theory, Taylor & Francis, London, 1992

Electrical properties

PercolationCalculations (sitepercolation)

We can do calculations on a two-dimensional lattice. We have occupied and empty places (we use then the term site percolation, there is also the alternative of bond percolation). P is the fraction of occupied sites and s is the number of neighbor occupied sites connected in a cluster. We have percolation if we get an infinite cluster.

77


σdc(p) ~ (p- pc)t


Evaluation of dc conductivity in terms of percolation (MWCNTs in PC)

P. Poetschke et al., Polymer 44 (2003) 5023-30

Percolation theory results for a nanocomposite with conducting inclusions (like here CNT) in an insulating matrix (polycarbonate here) in the equation given here for dc conductivity as a function of the volume fraction p of conducting inclusions. Pc is the percolation threshold and t a fractional power law exponent depending on the dimensionality. The figure shows experimental results (for convenience the weight fraction of CNT is used here to describe the composition). The line is the fit of the percolation equation to the data. The inset shows the same results in a different presentation. The percolation threshold is here much lower than the prediction for a three-dimensional statistical distribution of conducting spheres (about 16 vol%, slide 74), obviously because of the shape of CNT (needles).

78


10-2 10-1 100 101 102 103 104 105 106

10-8

10-7

10-6

10-5

10-4

σ(S/

cm)

f(Hz)

PP+ 5.3 wt% CNT

140rpm-1900C-15min

140rpm-1900C-5min

140rpm-2100C-15min

70rpm-1900C-15min

70rpm-1900C-5min

70rpm-2100C-15min

70rpm-2100C-5min


DRS Results – PP/CNT seriesEffects (optimization) of processing

Here results for ac conductivity for a fixed composition (5.3 wt% CNT in a polypropylene matrix) and different mixing conditions: speed, temperature and time of mixing. We use here conductivity as a measure of the quality of dispersion of CNT to optimize the processing (mixing) conditions.

79


• Percolation between 0.35 wt% (SWNT) and 2.0 wt% (MWNT)

Electrical conductivity – effects of CNT dimensions and purity (incorporation of different CNTs into PC)

PC Iupilon E2000, DACA Microcompounder, mixing under comparable conditions

DACA Microcompounder

(Pötschke et al. Fullerenes, Nanotubes, and Carbon Nanostructures 13 (2005), Suppl. 1, 211-224, Pötschke et al. GAK 58 (2005) 1, 45-51) By courtesy of Dr. Petra Poetschke, IPF Dresden (Germany)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0100

103

106

109

1012

1015

1018 MWNT2 (Nanocyl, very thin crude) MWNT3 (Nanocyl,very thin purified) MWNT4 (TsNaMWCnt1, C-nano, China) Nanocyl N 7000

(industrial product) DWNT (Nanocyl, purified) SWNT (CNI, HiPCo)

Vol

ume

resi

stiv

ity (O

hm c

m)

CNT content in PC (wt%)

This slide and the next two, as well as slides 70 and 71, were kindly given to me by Dr. Poetschke, an expert in the field. We follow here the effects of CNT dimensions and purity on percolation behavior.

80


• PC = PC Iupilon E 2000• PP = Moplen 520H• TPU = Elastollan 1185A

Electrical conductivity – effects of CNT/polymer interactions (incorporation of the same CNT into different polymers)

Nanocyl thin (10-15 nm) crude MWNT, DACA Microcompounder, 50 rpm, 15 min, T adapted to polymers

• Percolation concentration in PC < TPU < PP


(Pötschke et al. VDI Berichte 1920 (2005) 209)

0 1 2 3 4 5 6 7 8 9 10 11100

103

106

109

1012

1015

1018

TPU

PP

PCVol

ume

resi

stiv

ity (O

hm c

m)

MWNT content (wt%)

Here we follow the effects of CNT/polymer interactions on percolation behavior. We use three different matrices: polypropylene (PP), thermoplastic polyurethane (TPU) and polycarbonate (PC). The differences in percolation behavior reflect differences in the degree (quality) of dispersion of CNT.

81


• Percolation of MWNT in PA between 3 und 7 wt%,lower viscous matrix → lower percolation concentration

Electrical conductivity – effects of polymer viscosity (direct incorporation of Nanocyl N 7000 (10 nm) in different polyamide 6)


(Pegel, Pötschke et al., AP18, CD-ROM,Technomer 2005, ISBN 3-00-017458-3)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16101

104

107

1010

1013

1016

Capron 8202 low viscous/Nanocyl 7000 Capron B135WP high viscous/Nanocyl 7000 Capron 8202/Nanocyl 3150 (short CNT)

Vol

ume

resi

stiv

ity (O

hm c

m)

MWNT content (wt%)

Here effects of polymer viscosity on percolation behavior. The results confirm the expectation that the quality of dispersion increases with decreasing viscosity of the matrix.

Here we are at the end of our excursion in the world of polymer nanocomposites, many thanks for your attention!

Documents

ICSAM Course Pissis