P
I
som med ti offentlig gr maj 2010, k försvaras på de Ciencia y
P OLYM
IN T HI
illstånd av K ranskning fö kl 10.00 i sal å engelska. F
Tecnología de
MER N
IN F ILM
Linda
AKADEM
Kungliga Tek ör avläggand
D1, Lindste Fakultetsopp e Polímeros (I
N ANOC
M A PP
a Fogelst
MISK AVHAN kniska högsk de av teknisk dtsvägen 17, ponent: Profe ICTP‐CSIC), M
COMPO PLICAT
tröm
NDLING
kolan i Stock k doktorsexa , KTH, Stock essor José M Madrid, Span
OSITES
TIONS
kholm framl amen fredag kholm. Avha M. Kenny från
nien.
lägges till gen den 7 andlingen n Instituto
Copyright © 2010 Linda Fogelström All rights reserved
Paper I © 2006 Elsevier B.V.
Paper V © 2008 Elsevier B.V.
TRITA‐CHE‐Report 2010:12 ISSN 1654‐1081
ISBN 978‐91‐7415‐615‐7
A BSTRACT
The introduction of a nanoscopic reinforcing phase to a polymer matrix offers great possibilities of obtaining improved properties, enabling applications outside the boundaries of traditional composites.
The majority of the work in this thesis has been devoted to polymer/clay nanocomposites in coating applications, using the hydroxyl‐functional hyperbranched polyester Boltorn® as matrix and montmorillonite clay as nanofiller. Nanocomposites with a high degree of exfoliation were readily prepared using the straightforward solution‐intercalation method with water as solvent. Hard and scratch‐resistant coatings with preserved flexibility and transparency were obtained, and acrylate functionalization of Boltorn® rendered a UV‐curable system with similar property improvements. In order to elucidate the effect of the dendritic architecture on the exfoliation process, a comparative study on the hyperbranched polyester Boltorn® and a linear analogue of this polymer was performed. X‐ray diffraction and transmission electron microscopy confirmed the superior efficiency of the hyperbranched polymer in the preparation of this type of nanocomposites.
Additionally, an objective of this thesis was to investigate how cellulose nanofibers can be utilized in high performance polymer nanocomposites. A reactive cellulose “nanopaper” template was combined with a hydrophilic hyperbranched thermoset matrix, resulting in a unique nanocomposite with significantly enhanced properties. Moreover, in order to fully utilize the great potential of cellulose nanofibers as reinforcement in hydrophobic polymer matrices, the hydrophilic surface of cellulose needs to be modified in order to improve the compatibility. For this, a grafting‐from approach was explored, using ring‐opening polymerization of ε‐caprolactone (CL) from microfibrillated cellulose (MFC), resulting in PCL‐modified MFC. It was found that the hydrophobicity of the cellulose surfaces increased with longer graft lengths, and that polymer grafting rendered a smoother surface morphology. Subsequently, PCL‐grafted MFC film/PCL film bilayer laminates were prepared in order to investigate the interfacial adhesion. Peel tests demonstrated a gradual increase in the interfacial adhesion with increasing graft lengths.
Tillförandet av en förstärkande fas i nanostorlek till en polymermatris innebär stora möjligheter till egenskapsförbättringar, vilket möjliggör användande utanför ramarna för traditionella kompositer.
Majoriteten av arbetet i denna avhandling har tillägnats polymer/lera‐
nanokompositer i lackapplikationer, i vilka den hydroxyl‐funktionella hyper‐
förgrenade polyestern Boltorn® har använts som matris och montorillonit‐lera som nanofyllmedel. Nanokompositer med hög grad av exfoliering kunde relativt enkelt framställas genom att använda den okomplicerade lösningsmedels‐
interkaleringsmetoden, med vatten som lösningsmedel. Hårda och reptåliga lacker med bevarad flexibilitet och transparens erhölls och akrylat‐
funktionalisering av Boltorn® gav ett UV‐härdbart system med liknande egenskapsförbättringar. För att klarlägga vilken effekt den dendritiska strukturen har på exfolieringsprocessen, gjordes en jämförelsestudie mellan den hyper‐
förgrenade polyestern Boltorn® och en linjär analog till denna polymer.
Röntgendiffraktion och transmissionselektronmikroskopi bekräftade den överlägsna förmågan hos den hyperförgrenade polymeren i framställandet av denna typ av nanokompositer.
Vidare var målet med denna avhandling att undersöka hur nanofibrer av cellulosa kan användas i polymera nanokompositer med hög prestanda. Ett reaktivt ”nanopappers”‐templat bestående av cellulosananofibrer kombinerades med en hydrofil hyperförgrenad härdplastmatris, vilket resulterade i en unik nanokomposit med signifikant förbättrade egenskaper. För att till fullo kunna utnyttja den stora potentialen hos cellulosananofibrer som förstärkare i hydrofoba polymermatriser, måste cellulosans hydrofila yta modifieras för att öka kompatibiliteten. För detta ändamål undersöktes ringöppningspolymerisation som metod att ympa poly(ε‐kaprolakton) (PCL) från mikrofibrillerad cellulosa (MFC). Resultaten visade att hydrofobiciteten hos cellulosaytan ökade med längden av den ympade polymeren och att polymer‐ympningen gav en slätare ytmorfologi. För att undersöka vidhäftningen mellan gränsytorna framställdes laminat av PCL‐ympade MFC‐filmer och rena PCL‐filmer. Analys av dessa visade att vidhäftningen mellan gränsytorna successivt ökade med växande ymplängder.
L IST OF PAPERS
The thesis is a summary of the following papers:
I.
”UV‐curable Hyperbranched Nanocomposite Coatings” L. Fogelström, P.Antoni, E. Malmström, A. Hult, Progress in Organic Coatings 2006, 55, 284–
290.
II.
“Hard and Flexible Nanocomposite Coatings using Nanoclay‐filled Hyperbranched Polymers” L. Fogelström, E. Malmström, M. Johansson, A. Hult, ACS Applied Materials & Interfaces, submitted
III.
ʺLinear vs. Hyperbranched Polymers in the Preparation of Polymer/Clay Nanocomposites” L. Fogelström, S. Hansson, A. Carlmark, A. Hult, E.Malmström, Polymer, submitted
IV.
“Novel Nanocomposite Concept based on Hyperbranched Polymers in Reactive Cellulose Nanopaper Templates” M. Henriksson, L. Fogelström, M. Johansson, A. Hult, L. Berglund, Composites Science and Technology, submitted
V.
ʺSurface Grafting of Microfibrillated Cellulose with Poly(ε‐caprolactone) – Synthesis and Characterization” H. Lönnberg, L. Fogelström, Q. Zhou, A. Hult, L. Berglund, E. Malmström, European Polymer Journal 2008, 44, 2991–2997
VI.
“Investigation of the Graft‐Length Impact on the Interfacial Toughness in a Cellulose/Poly(ε‐caprolactone) Bilayer Laminate” H. Lönnberg, L.Fogelström, Q. Zhou, A. Hult, L. Berglund, E. Malmström, Composites Science and Technology, submitted
I. The majority of the experimental work and most of the preparation of the manuscript.
II. All of the experimental work and the preparation of the manuscript.
III. The majority of the experimental work and most of the preparation of the manuscript.
IV. Half of the experimental work and the preparation of the manuscript.
Involved in all parts of the work.
V. Minor parts of the experimental work and the preparation of the manuscript.
VI. About half of the experimental work and minor parts of the preparation of the manuscript.
Papers not included in this thesis:
VII.
ʺEnzymatic One‐Pot Route to Telechelic Polypentadecalactone Epoxide:Synthesis, UV Curing, and Characterization” M. Eriksson, L. Fogelström, K. Hult, E. Malmström, M. Johansson, S. Trey, M. Martinelle,
Biomacromolecules 2009, 10(11), 3108‐3113
VIII.
“Honeycomb‐Patterned Membranes from Polymer‐Modified Silica Nanoparticles” D. Nyström, P. Antoni, E. Östmark, D. Nordqvist, J.Örtegren, L. Fogelström, E. Malmström, M. Lindgren, A. Hult, manuscript
A BBREVIATIONS
AFM atomic force microscopy
ATRP atom transfer radical polymerization Bis‐MPA 2,2‐bis(methylol)propionic acid
Boltorn® hydroxyl‐functional aliphatic hyperbranched polyester Boltorn H30 Boltorn® of the third pseudo‐generation
CA contact angle
CC conventional calibration in SEC CDCl3 deuterated chloroform
ε‐CL ε‐caprolactone
CRP controlled radical polymerization Cu(I)Cl copper(I)chloride
Cu(II)Br2 copper(II)bromide DCM dichloromethane
DMA dynamic mechanical analysis DMF dimethylformamide
DMSO‐d6 deuterated dimethyl sulfoxide DP degree of polymerization DSP differential speckle photography
EB electron beam
EBiB 2‐bromoisobutyrate
FE‐SEM field emission scanning electron mircroscopy FTIR fourier transform infrared spectroscopy GMA glycidyl methacrylate
HCl hydrochloric acid
HMMM hexamethoxymethyl melamine [I] concentration of initiator Irgacure 651 UV‐initiator
[M] concentration of monomer MCC microcystalline cellulose MEK methyl ethyl ketone
MMT montmorillonite
Mn number average molecular weight MW molecular weight
Na+MMT natural montmorillonite
NMP nitroxide mediated polymerization NMR nuclear magnetic resonance spectroscopy PCL poly(ε‐caprolactone)
PDI polydispersity index PGMA poly(glycidyl methacrylate) PMDETA pentamethyldiethyltriamine pTSA p‐toluenesulfonic acid
RAFT reversible addition fragmentation chain transfer polymerization RH relative humidity
ROP ring‐opening polymerization
RT room temperature
SEC size exclusion chromatography Sn(Oct)2 tin octoate
std standard deviation
tan δ storage modulus/loss modulus in DMA TBAB tetrabutyl ammonium bromide
TEM transmission electron microscopy
Texanol 2,2,4‐trimethyl‐1,3‐pentanediol monoisobutyrate Tg glass transission temperature
TGA thermogravimetric analysis THF tetrahydrofuran
TONE 0301 polyol spacer
TPGDA tripropyleneglycol diacrylate UC universal calibration in SEC UV ultra violet
Vf cellulose nanofiber volume fractions XRD x‐ray diffraction
Zcrit critical entanglement chain length
T ABLE OF C ONTENTS
1. PURPOSE OF THE STUDY ... 1
2. INTRODUCTION ... 2
2.1 POLYMER NANOCOMPOSITES ... 2
2.2 POLYMER/CLAY NANOCOMPOSITES... 3
2.2.1 Montmorillonite clay ... 3
2.2.2 Structure of polymer/clay nanocomposites ... 4
2.2.3 Preparation of polymer/clay nanocomposites ... 5
2.2.4 Characterization of polymer/clay nanocomposites ... 5
2.2.5 Properties of polymer/clay nanocomposites ... 6
2.3 DENDRITIC POLYMERS ... 7
2.3.1 Hyperbranched polymers ... 8
2.3.1.1 Properties of hyperbranched polymers ... 8
2.3.1.2 Derivatization of hyperbranched polymers ... 9
2.3.1.3 Boltorn® ... 9
2.4 DENDRITIC POLYMER/CLAY NANOCOMPOSITES ... 10
2.5 CELLULOSE ... 11
2.5.1 Cellulose nanofibers ... 12
2.5.2 Cellulose nanocomposites ... 12
2.5.3 Surface modification of cellulose ... 12
2.6 CONTROLLED POLYMERIZATION ... 13
2.6.1 Atom transfer radical polymerization, ATRP ... 14
2.6.2 Ring-opening polymerization, ROP ... 14
2.7 THIN FILM APPLICATIONS ... 15
2.7.1 Coatings ... 15
2.7.1.1 Coating types ... 16
2.7.1.2 Rheology of coatings ... 17
2.7.1.3 Dendritic polymers in coatings ... 17
2.7.1.4 Nanocomposite coatings ... 17
3. EXPERIMENTAL ... 18
3.1 MATERIALS ... 18
3.2 CHARACTERIZATION METHODS ... 18
3.3 EXPERIMENTAL PROCEDURES ... 21
3.3.1 Boltorn H30/Na+MMT nanocomposite coatings ... 21
3.3.1.1 Preparation of Boltorn H30/Na+MMT nanocomposites (Paper I and II) ... 21
3.3.2.1 Synthesis of acrylated Boltorn H30 ... 22
3.3.2.2 Preparation of acrylated Boltorn H30/Na+MMT nanocomposites ... 22
3.3.2.3 Preparation of UV-cured coatings ... 22
3.3.3 Linear vs. hyperbranched polymers in polymer/clay nanocomposites ... 23
3.3.3.1 Synthesis of PGMA using ATRP... 23
3.3.3.2 Synthesis of Bis-MPA-modified PGMA – a linear analogue to Boltorn H30 .... 23
3.3.3.3 Preparation of Boltorn H30/Na+MMT nanocomposites (Paper III) ... 24
3.3.3.4 Preparation of Bis-MPA-modified PGMA/Na+MMT nanocomposites ... 24
3.3.3.5 Preparation of thermally cured coatings (Paper III) ... 24
3.3.4 Cellulose nanocomposite films ... 24
3.3.4.1 Preparation of MFC nanofiber films ... 24
3.3.4.2 Preparation of MFC nanofiber films for Boltorn H30/MFC nanocomposites .... 24
3.3.4.3 Preparation of Boltorn H30/MFC nanocomposites ... 24
3.3.5 Surface modification of cellulose nanofibers ... 25
3.3.5.1 Grafting of PCL from MFC using ROP ... 25
3.3.6 Cellulose nanocomposite laminates ... 26
3.3.6.1 Grafting of PCL from MFC nanofiber films using ROP ... 26
3.3.6.2 Preparation of PCL-film and PCL-MFC bilayer laminates ... 26
4. RESULTS AND DISCUSSION ... 27
4.1 POLYMER/CLAY NANOCOMPOSITES ... 27
4.1.1 Boltorn H30/Na+MMT nanocomposite coatings ... 27
4.1.1.1 Nanocomposite characterization ... 27
4.1.1.2 Coating characterization ... 31
4.1.2 Acrylated Boltorn H30/Na+MMT nanocomposite coatings ... 34
4.1.2.1 Synthesis of acrylated Boltorn H30 ... 34
4.1.2.2 Nanocomposite characterization ... 35
4.1.2.3 Coating characterization ... 37
4.1.3 Linear vs. hyperbranched polymers in polymer/clay nanocomposites ... 39
4.1.3.1 Synthesis of linear analogue to Boltorn H30 ... 39
4.1.3.2 Nanocomposite and film preparation ... 40
4.1.3.3 Nanocomposite characterization ... 40
4.2 CELLULOSE NANOCOMPOSITES ... 42
4.2.1 Cellulose nanocomposite films ... 42
4.2.1.1 Structure of cellulose nanocomposite films ... 43
4.2.1.2 Mechanical properties ... 45
4.2.2 Surface modification of cellulose nanofibers ... 46
4.2.2.1 Grafting of PCL from MFC using ROP ... 47
4.2.2.2 Thermal characterization ... 49
4.2.3 Cellulose nanocomposite laminates ... 49
4.2.3.1 Grafting of PCL from MFC nanofiber films using ROP ... 49
4.2.3.2 Adhesion in PCL-grafted MFC film/PCL film bilayer laminates ... 51
5. CONCLUSIONS ... 53
6. FUTURE WORK ... 56
7. ACKNOWLEDGEMENTS ... 57
8. REFERENCES ... 59
1. P URPOSE OF THE STUDY
The overall purpose of this study has been to investigate different approaches to enhance properties of polymer thin films through the introduction of a nanoscopic reinforcing phase.
The majority of the work in this thesis has been devoted to polymer/clay nanocomposites in coating applications, using a hyperbranched polymer as matrix and montmorillonite clay as nanofiller. Previous work by Månson et al.1 has demonstrated the successful use of the hydroxyl‐functional hyperbranched polyester Boltorn® in the preparation of polymer/clay nanocomposites, using the solution intercalation method with water as solvent, resulting in highly exfoliated materials. The present study aimed to explore the use of this system to improve hardness and scratch resistance in thermoset coatings, while maintaining flexibility and transparency. Furthermore, it was hypothesized that the hydroxyl end‐groups of the hyperbranched polymer would be available for further reaction after the exfoliation of clay nanoparticles, enabling functionalization with, e.g., acrylate groups, in order to obtain a UV‐curable coating system. Another aim was to elucidate the effect of the dendritic architecture on the exfoliation process by performing a comparative study on the hyperbranched polyester Boltorn® and a linear analogue of this polymer, in order to investigate their respective efficiency in the preparation of these nanocomposites.
Additionally, an objective of this thesis has been to study how cellulose nanofibers can be utilized in high performance polymer nanocomposites. First the combination of a reactive cellulose nanopaper template and a hydrophilic hyperbranched thermoset matrix was investigated. Moreover, in order to fully utilize the great potential of cellulose nanofibers as reinforcement in hydrophobic polymer matrices, the hydrophilic surface of cellulose needs to be modified in order to improve the compatibility. For this, a polymer‐grafting approach was explored, using ring‐opening polymerization of poly(ε‐caprolactone) from microfibrillated cellulose (MFC). Subsequently, the aim was to investigate the interfacial adhesion in modified MFC/PCL bilayer laminates.
2. I NTRODUCTION
2.1 POLYMER NANOCOMPOSITES
A composite consists of two or more physically distinct components. When mixed together the resulting material achieves superior properties compared with the individual constituents.2 In nanocomposites the size of at least one of the components is in the nanometer (10‐9 m) range. Apart from the obvious particle‐
size difference, nanocomposites differ from conventional composites in some other important ways. In traditional polymer composites the amount of filler usually has to be over 10 wt% to achieve the desired mechanical properties.3 This high degree of loading leads to deteriorating properties, such as increased density and a loss of tenacity. With nanocomposites, however, the desired properties can be achieved at a much lower filler loading, generally up to a few weight percent.
Figure 1. Schematic picture showing a) one‐dimensional nanoparticles, b) two‐dimensional nanoparticles, c) three‐dimensional nanoparticles.
Nanoparticles are divided into three groups, Figure 1, differentiated by how many of their dimensions that are in the nanometer range.4 When all three dimensions are in the nanometer range, the nanoparticles are called isodimensional, with spherical silica nanoparticles as a typical example. Particles with two dimensions in the nanometer range are rodlike, such as carbon nanotubes or cellulose whiskers. In the third type of nanocomposites, only one of the dimensions is in the nanometer range, which gives a platelet‐ or sheet‐like structure, exemplified by layered silicate clays. The sheets are typically one or a
a) b) c)
few nanometers thick with a surface of several thousand square nanometers, which gives a very high aspect ratio.
2.2 POLYMER/CLAY NANOCOMPOSITES
The great interest in using clay nanoparticles as reinforcement in polymers emerged in the late 1980s when a research group from Toyota reported their findings regarding the possibility of building a nanostructure from a polymer and an organophilic clay, with dramatic property improvements.5, 6
2.2.1 Montmorillonite clay
Expandable layered silicate clays, with montmorillonite (MMT) as a typical example,7 are among the most commonly used nanofillers for polymer nanocomposites. MMT is a naturally abundant, easily accessible soil mineral with a layered structure, Figure 2.
Figure 2. Structure of the layered silicate clay montmorillonite.
The layers consist of two tetrahedral sheets sandwiching an edge‐shared octahedral sheet of either aluminum or magnesium hydroxide.4 When the clay is
stacked, each layer is separated by regular van der Waals gaps, called interlayers or galleries. Located in the interlayers are cations such as Ca2+ and Na+, counterbalancing an excess of negative charge, caused by isomorphous substitutions of Si4+ for Al3+ within the tetrahedral lattice, and of Al3+ for Mg2+ in the octahedral lattice. There are also often water molecules present between the layers, due to the high hydrophilicity of the clay. The repeating unit, d‐spacing, is represented by the sum of the single layer thickness and the interlayer. In natural MMT (Na+MMT), the d‐spacing varies from 9.6 Å, for the stacked dry clay, to 19 Å for the clay in 99% relative humidity.8 The microstructure of the clay consists of several layers, lamellas (1 nm), lying parallel forming a primary particle (8‐10 nm). The primary particles are, in turn, aggregated into larger irregular particles (0.1‐10 μm).9 The layered structure together with the suitable layer charge density enable a rich intercalating chemistry, thus allowing for MMT to be intercalated directly with hydrophilic polymers or chemically modified to enhance the compatibility with hydrophobic polymers.10
2.2.2 Structure of polymer/clay nanocomposites
Polymer/layered silicate nanocomposites can either be referred to as intercalated if the polymer is present between the clay layers, but the regularity of the layered structure is maintained, or as exfoliated if the clay layers are completely dispersed in the matrix,11 Figure 3.
Figure 3. Schematic structure of polymer/clay nanocomposites
layered silicate polymer
conventional composite
intercalated nanocomposite
exfoliated nanocomposite
In order to fully utilize the property‐improving potential of the nanofiller, complete dispersion is necessary, so that the entire surface area (ca 760 m2/g) of the nanosized clay particles is available to interact with the polymer; the coupling between the phases facilitates stress transfer to the reinforcing phase, allowing for tensile strength and toughening improvements.10
2.2.3 Preparation of polymer/clay nanocomposites
Various methods have been employed to prepare polymer/clay nanocomposites.
The first method, used by the Toyota group, was the in‐situ synthesis.12‐16 In this method, the polymerization is initiated in between the clay layers, and the polymer build‐up causes the interlayer gallery to expand. Another commonly used method is the melt processing technique, in which the clay is dispersed in a polymer in the molten state, using high shear forces and high temperatures.17‐20 In the solution intercalation method the layered silicate is exfoliated into single layers using a solvent in which the polymer is soluble.4 The weak forces stacking the layers together make the dispersion of the inorganic material easy, if the solvent is adequate. Subsequently, the polymer adsorbs onto the delaminated sheets and the nanocomposite is formed upon removal of the solvent. The above described nanocomposite preparation methods generally require modification of the naturally hydrophilic clay to improve the compatibility with commonly used hydrophobic polymers. Depending on the desired properties, several modifying methods are applicable. The simplest method is the reaction with alkylammonium ions, due to the ease with which the alkyl chains can change their conformation by transition into kink and gauche conformers.21 In addition, many different properties can be obtained depending on the choice of carbon chain in the alkylammonium ion. Therefore, this type of modifier has shown to be particularly versatile in polymer/clay nanocomposites.22, 23 The purpose of the alkylammonium ion is not just to compatibilize the clay and the polymer but it also expands the clay galleries, making intercalation easier.10 However, modification of the clay could be avoided using the solution intercalation method with water as solvent, thereby making use of native MMT’s affinity for water and its ability to exfoliate in aqueous solutions.24, 25
2.2.4 Characterization of polymer/clay nanocomposites
There are two main methods in the characterization of polymer/clay nanocomposites:7, 26 x‐ray diffraction (XRD) and transmission electron microscopy (TEM). XRD is a straightforward method for determining the spacing between the clay layers; the sample preparation is relatively easy and the analysis can be
performed within a few hours. The potential of XRD, however, is limited since all its data is averaged over the whole sample. Furthermore, XRD does not give any information concerning the spatial distribution of the clay particles. These limitations may lead to incorrect conclusions regarding the nanocomposite structure, and need to be taken into careful consideration when interpreting the results. TEM is thus an important complement to XRD. TEM provides direct visual information of the morphology, atom arrangement, spatial distribution of the phases and structural defects of an elected area of the sample. In this method, the limitations are rather operational than instrumental, since TEM requires substantial skills in specimen preparation and analysis in order to ensure that a representative section is examined.
2.2.5 Properties of polymer/clay nanocomposites
Improvement of mechanical properties is the most common application for polymer/clay nanocomposites.10, 17, 27‐32 A proposed theory suggests that the polymer which is physisorbed onto the clay is stiffened by its affinity for, and adhesion to, the silicate surface, thus having a higher modulus than the surrounding bulk polymer.33 Polymer‐clay nanocomposites have also shown to substantially decrease the permeability compared with the pristine polymer.28, 34‐38 The impermeable, platelet‐like nanolayers provide a tortuous pathway for the diffusing molecules as they transverse the nanocomposite, see Figure 4. Hence the high aspect ratio clay layers effectively decrease the diffusion‐rate through the material.7
Figure 4. Schematic picture showing the tortuous diffusion pathway in an exfoliated polymer/clay nanocomposite.
penetrating molecule
Closely related to the improvement of barrier properties is the improvement of thermal properties, benefitted by in thermal‐stability and fire‐resistance applications.4, 39‐44 The thermal stability is improved since the clay layers act as superior insulator and mass transport barrier to the volatile products generated during decomposition, as well as by assisting the char formation after thermal decomposition, especially on the outer layer.33 The clay‐rich char layer in turn protects the polymer from hot flames during fire and slows the oxygen uptake and the escape of volatile gases produced by the polymer degradation. Important for many applications,12, 45, 46 for example coating applications,47 are the optical properties of nanocomposites. The nanometric dimension of the filler is smaller than the wavelength of visible light, i.e., the light is not refracted, thus rendering the nanocomposites highly transparent.48 Furthermore, the introduction of nanoparticles to crystalline materials affects the crystallinity, in turn affecting the optical clarity.49 The transparency of nanocomposites makes them applicable outside the boundaries of traditional composites.
2.3 DENDRITIC POLYMERS
The origin of the word dendritic is the Greek word for tree, dendro. Accordingly, dendritic polymers are characterized by densely branched structures and a large number of end‐groups ‐ a tree‐like structure.50‐52 The monomers used to obtain this type of polymers are ABx‐monomers, typically AB2. These can be either homo‐ or copolymerized, with or without a multifunctional core, to achieve a number of different property profiles.
Figure 5. Schematic picture showing a) a dendrimer, b) a hyperbanched polymer with core, and c) a hyperbranched polymer without core.
If a core is used and the product is monodisperse ‐ a dendrimer ‐ each layer of propagation is represented by a generation, Figure 5a, and if the product is polydisperse ‐ a hyperbranched polymer ‐ the term pseudo‐generation is used, Figure 5b. If no core is used in the synthesis, the product is a hyperbranched polymer with a theoretical infinitely broad molecular‐weight distribution,50 Figure 5c. Also belonging to the family of dendritic polymers are linear‐dendritic‐
hybrid polymers, such as dendrigrafts and dendronized polymers.52
2.3.1 Hyperbranched polymers
Hyperbranched macromolecules, theoretically described in the 1950’s,53 were first reported synthesized in the early 1990’s54 and have been studied in detail since then.55, 56 The synthesis of dendrimers,57‐60 is often tedious since it generally requires protection and deprotection steps as well as purification between successive generations. Hyperbranched polymers on the other hand, are readily synthesized through a one‐pot61 or pseudo‐one‐pot62 synthesis. One‐pot‐
synthesis, with or without a core, is performed by adding all ABx monomers simultaneously, resulting in an uncontrolled branching and a very broad molecular‐weight distribution, typically 2 or higher.52 In the pseudo‐one‐pot synthesis a core is used, and the ABx monomers are added in a stoichiometric ratio, one pseudo‐generation at a time,62 enabling a better controlled polymerization with a higher degree of branching and a more narrow molecular‐
weight distribution. While having the same amount of end‐groups, the main distinguishing properties for hyperbranched polymers compared with dendrimers, are higher molecular weight distribution and irregular branching.
These, in some cases undesired, properties are a result of the one‐pot, or pseudo‐
one‐pot synthesis, applied on hyperbranched polymers. However, the simple synthesis renders the preparation of hyperbranched polymer less expensive,50 which is a highly desired feature, facilitating commercial applications.61, 63‐65
2.3.1.1 Properties of hyperbranched polymers
The structure of hyperbranched polymers, densely branched with a large number of end‐groups, differs from the structure of linear or conventionally branched polymers; hence, the properties differ as well. The globular structure of hyperbranched polymers precludes the entanglements that conventional polymers undergo, which leads to lower viscosity at a given molecular weight.66 Also, the hyperbranched polymer exhibits a Newtonian behavior, which is unusual for polymers.67 As can be anticipated from their highly branched architecture, hyperbranched polymers are most often amorphous; however, some
exceptions are reported in the literature.68, 69 The large number of end‐groups has a great effect on the properties of the hyperbranched polymer since they strongly affect the interactions between the hyperbranched polymer and its neighboring molecules, thus affecting the solubility and viscosity. Furthermore, the nature of the end‐groups governs the reactivity of the molecule. Due to the large number of end‐groups, hyperbranched polymers have a great capacity for rapid crosslinking even at moderate molecular weights, if the end‐groups are reactive. Hence, the possibilities for designing thermosetting networks are numerous.56
2.3.1.2 Derivatization of hyperbranched polymers
The large number of end‐groups of hyperbranched polymers strongly governs the final unique properties of the material, as mentioned above. Often only a small amount of end‐groups has to be derivatized to achieve, for instance, a desired reactivity and therefore the rest can be used to control other properties, such as viscosity, solubility etc. For example, acrylate and methacrylate derivatives of hyperbranched polymers can be synthesized using standard procedures, resulting in UV‐curable systems.70, 71
2.3.1.3 Boltorn®
Boltorn® is one of few commercially available hyperbranched polymers. It is a hydroxyl‐functional hyperbranched polymer, manufactured and sold by Perstorp AB, Sweden.72 The core is a polyalcohol, ethoxylated pentaerythritol, and the backbone, built up by the monomer 2,2‐bis(methylol)propionic acid, Bis‐MPA, is an aliphatic polyester,62 Figure 6.
Figure 6. Structure of the hyperbranched polyester Boltorn H30.
Since Boltorn® is produced through a pseudo‐one‐pot synthesis, offering some degree of control over the polymerization, the polydispersity is fairly low, with a PDI value ranging from 1.4 to 2.4, as reported in the literature.62, 72, 73 The use of a core also allows for some control over the molecular weight. The Boltorn® used in this thesis is Boltorn H30, which is a hyperbranched polymer of the third pseudo‐
generation, having a molecular weight (Mn) of about 3600 g/mol and an average of 32 hydroxyl end‐groups per molecule.
2.4 DENDRITIC POLYMER/CLAY NANOCOMPOSITES
As described previously, polymer/clay nanocomposites can be prepared using the solution intercalation method, preferably with water as solvent, thus avoiding prior clay modification. Linear water‐soluble polymers, processed with this method, have in some cases resulted in reaggregated intercalated structures, rather than exfoliated structures, after drying.74, 75 The use of dendritic polymers such as hyperbranched polyesters, on the other hand, has been shown to promote exfoliation and stabilize the exfoliated structure after drying,1, 76 Figure 7. This stabilization is thought to be enabled by the compact globular structure of the hyperbranched polymer, with a large number of polar end‐groups and a relatively hydrophobic core, preventing the molecule from collapsing onto the clay layers.77
Figure 7. Schematic picture showing the structure of hyperbranched polymer/clay nanocomposites.
2.5 CELLULOSE
Cellulose, which is one of the world’s most abundant renewable resources, exists as a load‐bearing component in plant cell walls, in algae and tunicate sea animals and it can also be produced by certain bacteria.78 Cellulose is a linear polysaccharide, made up of β‐1‐4 D‐glucose units, Figure 8.
Figure 8. Molecular structure of cellulose.
The many hydroxyl groups form strong inter‐ and intramolecular hydrogen bonds, resulting in a crystalline, sheet‐like structure with a very high modulus, 134 GPa.79 In plants, the crystalline sheet stack together to form microfibrils, in turn bundling up to form fibrils, which, together with hemicellulose and lignin, build up the different layers of the wood cell (i.e., wood fiber) wall, Figure 9.
Figure 9. Schematic picture showing the hierarchical structure of cellulose.
2.5.1 Cellulose nanofibers
Nanosized cellulose fibers can be isolated from the wood cell wall in different ways, i.e., by chemical and mechanical treatments.80 Microcrystalline cellulose (MCC) is obtained through acid hydrolysis of wood fibers, thereby removing the amorphous and unordered cellulose parts, yielding highly crystalline cellulose nanofibers with low aspect ratio; the typical size of MCC is 200‐400 nm in length with an aspect ratio of about 10. Microfibrillated cellulose (MFC), developed in the 1980’s,81, 82 is produced through a combination of chemical and mechanical treatment, i.e., homogenization, of wood fibers, causing disintegration of microfibrils and microfibril aggregates from the cell wall. In order to further promote this disintegration, different pretreatments can be employed.83‐85 Compared with MCC, MFC has a much higher aspect ratio with a typical diameter of 10‐100 nm and a length up to several micrometers, depending on the preparation method. Furthermore, MFC contains both amorphous and crystalline cellulose, thus rendering MFC more flexible than MCC. When dried from a water suspension, MFC irreversibly forms a strong network, stabilized by the large amount of interfibrillar hydrogen bonds present in the material.86
2.5.2 Cellulose nanocomposites
The high modulus of cellulose, which is important for the mechanical properties,87, 88 together with other possible property improvements, including increased thermal stability,89 decreased thermal expansion,90 and increased thermal conductivity,91 makes cellulose in general, and cellulose nanofibers in particular, highly interesting as reinforcement in composite, or nanocomposite, materials. Additionally, if cellulose nanofibers are used to reinforce a transparent matrix, it is possible to maintain most of the transparency, even at fiber contents as high as 70 %, owing to the small scale (i.e., nanoscale) of the filler.90, 92
2.5.3 Surface modification of cellulose
Owing to its many hydroxyl groups, cellulose is very hydrophilic and adsorbs water easily, thus complicating the use of cellulose as reinforcement in polymer composites where adhesion to the often hydrophobic polymer is essential.
However, the hydroxyl groups can also act as reactive sites, endowing cellulose with great potential for modification and tailoring of properties. The adhesion between cellulose and hydrophobic polymers can be improved through the introduction of a low molecular weight coupling agent, such as acetyl‐, silanyl‐ or carboxyl‐functional compounds,93‐95 through physical modification using
secondary forces,84, 96 or through graft‐copolymerization of high molecular weight polymers, thereby making use of the possibility of obtaining chain entanglements that drastically improve the mechanical properties, as well as the possibility for co‐crystallization to occur with the polymer matrix. Co‐crystallization would further improve the interfacial adhesion between matrix and fiber, and thereby enhance the mechanical properties of the final composite. The surface grafting from cellulose has traditionally been performed through radiation‐initiated free radical polymerization of vinyl monomers, resulting in high molecular weight grafts with poor definition.97, 98 Surface modification of cellulose using controlled polymerization techniques, such as ATRP,99‐101 RAFT,102‐104 and ROP,105‐107 on the other hand, offer the possibility of obtaining surface‐confined polymers with controlled molecular weight and molecular weight distribution and chain structure, i.e., block copolymers, graft‐on‐graft copolymers etc.
2.6 CONTROLLED POLYMERIZATION
In order to control the molecular weight, the molecular weight distribution and the end‐group functionality of a polymer, the synthesis has to be performed using a living or controlled polymerization technique. Anionic and cationic polymerization are true living techniques, offering complete control over the polymerization. However, they are synthetically demanding in their requirement of ultrapure reactants and specific reaction conditions, thus limiting their commercial use. The most commonly used method for commercial applications is free radical polymerization, but since it lacks control over the polymerization regarding termination and chain transfer reactions, molecular weight and molecular weight distribution, and end‐group functionality, it cannot be used to synthesize well‐defined polymers, such as star, comb or block copolymers, requiring architectural control. In the effort to combine the control of the living polymerization techniques with the less‐demanding synthesis of the free radical polymerization, controlled radical polymerization (CRP) techniques have been developed; the three major types are nitroxide mediated polymerization (NMP),108 atom transfer radical polymerization (ATRP),109 and reversible addition fragmentation chain transfer polymerization (RAFT)110. By maintaining a low concentration of active propagating species, through a dynamic equilibrium as shown in Scheme 1, termination and unwanted side‐reactions are suppressed, thus offering control over the polymerization and the end‐group functionality.
Scheme 1. Fundamental equilibrium in controlled radical polymerization.
2.6.1 Atom transfer radical polymerization, ATRP
ATRP111‐113 is the most studied of the CRP techniques, owing to its versatility and compatibility with a range of monomers and solvents. The most elementary ATRP system consists of initiator (commonly an alkyl‐halide), monomer, and catalyst (a transition metal halide complexed with a ligand) and accurate matching of these components is essential in order to achieve a successful result.109 ATRP is based on a reversible redox process with equilibrium between active propagating chains and dormant – reversibly end‐capped – chains, unable to propagate. In order to achieve a well‐controlled polymerization, it is important that the rate of initiation is faster than the rate of propagation, since a fast initiation ensures that all the chains start growing at the same time. The main drawback of ATRP is the relatively large amount of catalyst used in the system, usually in the form of Cu+. Without purification of the final product, the remaining metal‐containing catalyst could have a detrimental effect in many applications, and therefore it needs to be removed after synthesis.
2.6.2 Ring‐opening polymerization, ROP
Ring‐opening polymerization (ROP) is a controlled polymerization technique that was developed in the 1930s,114 and it is a versatile method for polymerizing various cyclic monomers, including lactones, lactides, lactams, cyclic carbonates, siloxanes and ethers. The high efficiency of ROP, and lack of condensation byproducts, makes it superior to traditional polycondensation reactions in the synthesis of high molecular weight polymers. Furthermore, ROP can be performed both through a controlled and a living mechanism, depending on the monomer and initiator/catalyst system, thus enabling synthesis of complex macromolecular architectures with well‐defined end‐groups.115 Depending on the catalyst system, ROP proceeds through a cationic, anionic or coordination‐
insertion mechanism. The most widely used catalyst for ROP of lactones and lactides is tin octoate (Sn(Oct)2), owing mainly to its availability, high efficiency and relatively low toxicity, while being inexpensive and approved for food and
drug applications.116 A general mechanism of ROP with this type of catalyst is shown in Scheme 2. Most commonly, ROP is initiated by hydroxyl‐functional compounds (or water), but amino‐functional is also possible. Drawbacks of Sn(Oct)2‐catalytic systems include requirement of water‐free reagents and glass‐
ware, occurrence of various transesterification reactions during synthesis, affecting the molecular weight distribution, and difficult catalyst‐removal after synthesis.
Scheme 2. Schematic description of ROP of lactones.
2.7 THIN FILM APPLICATIONS
The concept of polymer thin films entails various applications such as coatings (e.g., protective and aesthetic), free‐standing films (e.g., membranes and filters), and laminates (e.g., for packaging applications). This thesis contains work on free‐
standing polymer‐impregnated cellulose nanopapers and laminates of nanocellulose films. The main focus, however, regarding the thin film applications, has been on coatings.
2.7.1 Coatings
The potential functions of coatings are numerous, e.g., to protect, to decorate, or to provide information or special physical effects.117 A coating consists of four main components:118 The resin is the essence of the coating and contributes to most of its chemical and physical properties. The resin can be everything from a natural material, like egg‐white or blood, to a highly complex organic polymer obtained by modern chemistry. The main function of the resin is to be a film former, holding the coating cohesively together and adhesively to the substrate;
Solvents are primarily used to enable application of the coating. In chemically curing systems it also affects the reactivity. Solvent systems are often blends of different solvents, designed to improve flow and leveling of the film and to minimize volatile organic content and cost; Pigments are in general inorganic compounds such as oxides and silicates. Prime pigments add color and opacity
and they can protect the coating from deteriorating UV‐light. Functional pigments can provide for example fire resistance or corrosion protection.
Extender pigments increases volume without excessive cost. They add neither color nor opacity, but they can for instance provide mechanical strength and control gloss; In the coating industry numerous additives are used to improve various properties. Plasticizers can provide flexibility to brittle polymers, catalysts and driers can accelerate cure rates, etc. An important group of additives are the rheological additives, used to control consistency and pigment settling and improve flow, leveling and sag resistance of the film.
2.7.1.1 Coating types
There are four main classifications of coatings, categorized by the resin and how the coating is formed:118, 119 The simplest type of coating is the thermoplastic solvent‐
borne resins, in which a continuous film is formed when the solvent is evaporated and an amorphous solid is obtained. No cross‐linking occurs, but usually considerable physical entanglements are formed. Examples of this type are acrylics, vinyls, chlorinated rubbers, and cellulosic esters; In the thermoplastic dispersion resins, a high molecular weight polymer is segregated into discrete particles dispersed in a carrier phase. A continuous film is formed after application when the carrier phase is evaporated or displaced and the polymer particles are fused together. The dispersion enables a low viscosity despite the high molecular weight. Examples of this type are latex systems, plastisols, and thermoplastic powder coating resins; The chemically curing resins are crosslinked by intra‐ or intermolecular addition or condensation polymerization. The curing can proceed at either ambient or elevated temperature, with ultra‐violet (UV) or electron‐beam (EB) radiation, depending on the functional groups involved.
Examples of this type are epoxy systems, urethane systems, unsaturated polyesters, UV and EB curing acrylics, phenolics and amino cured systems; A very old film forming technology, still in use, is the oxidatively curing system, with a crosslinking chemistry yet to be fully understood. The oxidizing systems, however, involve an oxygen attack on conjugated double bonds on unsaturated fatty acids. This attack creates radicals ready to react with other fatty acid chains, and so the drying process proceeds. Examples of this type are oxidizing systems like unmodified oil systems, alkyds and epoxy esters, and water curing systems like ureas and alkyl silicates.
2.7.1.2 Rheology of coatings
A very important property of coatings is the viscosity; a sufficiently low viscosity is required in order to control the flow and leveling during film formation.120, 121 When designing a coating system there generally has to be a compromise between molecular weight and viscosity. While being essential in order obtain the right properties of the final film and to reduce cure shrinkage, a high molecular weight of the resin normally also renders a high viscosity. In order to lower the viscosity, an increased amount of solvent is required, which is undesirable also from an environmental point of view. Dendritic polymers, with their special features, offer an interesting alternative in the solution to this problem.
2.7.1.3 Dendritic polymers in coatings
One of the most prominent applications for dendritic polymers is as reactive multifunctional components in coating and resin formulations.56, 122 As discussed in section 2.3.1.1, the globular structure of hyperbranched polymers precludes the entanglement that conventional polymers undergo, which leads to lower viscosity at a given molecular weight and a Newtonian behavior. Additionally, the large number of end‐groups renders versatile cross‐linking possibilities and a great capacity for rapid curing.
2.7.1.4 Nanocomposite coatings
Many of the properties of nanocomposites described in section 2.2.5 are very useful in coating applications: improvements in mechanical properties would render a harder and more scratch resistant coating;47 the increased thermal stability and fire resistant properties are applicable in intumescent coatings123 and other fire retardant applications; the improved barrier properties are also pertinent, e.g., to packaging applications; the possibility of maintaining the transparency of a nanoparticle‐filled coating enables their use as clear‐coats, e.g., in wood or automotive applications.
3. E XPERIMENTAL
3.1 MATERIALS
The hydroxyl‐functional hyperbranched polymer Boltorn® H30 and 2,2‐
bis(methylol)propionic acid (Bis‐MPA) was kindly supplied by Perstorp AB Sweden. The clay used is a natural montmorillonite (Na+MMT) from the Cloisite® nanoclays series by Southern Clay Products, with a d‐spacing of 11.7 Å according to SCP’s X‐ray results. Hexamethoxymethyl melamine (HMMM) was supplied by Becker Industrial Coatings AB, and epoxy‐blocked p‐toluenesulfonic acid (epoxy‐
blocked pTSA) was supplied by Akzo Nobel Nippon Paint AB. Microfibrillated cellulose (MFC) was prepared according to the procedure described by Henriksson et al.85. Glycidyl methacrylate (GMA) from Fluka was passed through a column of neutral aluminum oxide from Sigma‐Aldrich prior to use. All solvents and the following chemicals were used as received: TONE polyol 0301 from Union Carbide, 2,2,4‐trimethyl‐1,3‐pentanediol monoisobutyrate (Texanol), benzyl alcohol, 2‐bromoisobutyrate (EBiB), ε‐caprolactone (ε‐CL), poly(ε‐
caprolactone) (PCL, MW 80 000 g/mol), copper(I)chloride (Cu(I)Cl), copper(II)bromide (Cu(II)Br2) hydroquinone, pentamethyldiethyltriamine (PMDETA) and tin octoate (Sn(Oct2)) from Sigma‐Aldrich, acrylic acid and methanesulfonic acid from Merck, nitrobenzene from Kebo Lab, tripropyleneglycol diacrylate (TPGDA) from UCB Radcure, Irgacure 651 from Ciba Specialty Chemicals, hydrochloric acid (HCl, 37 % solution in water) from Acros Organics, and tetrabutyl ammonium bromide (TBAB) from Nobel Chemicals.
3.2 CHARACTERIZATION METHODS
Atomic force microscopy (AFM) was performed using a Nanoscope III‐a system (Digital Instruments) equipped with an EV‐type vertically engaged piezoelectric scanner operating in tapping mode in air. Silicon AFM probes from Veeco (Nanosensors) were used with a resonance frequency of 275‐348 kHz.
Conventional coating characterization methods: pendulum hardness was studied with a 299 Koenig pendulum hardness tester from Erichsen; pencil hardness tests were principally performed according to SIS 18 41 87 standard, using pencils manufactured by KOH‐I‐NOOR; scratch tests were principally performed according to ISO 2409 standard, using a multi‐edge cutting tool and a brush, both manufactured by Erichsen; chemical resistance was analyzed through the MEK‐
rub test, using methyl ethyl ketone as solvent.
Density and porosity of the matrix and the cellulose nanofiber film in Paper IV was determined using an Archimedes scale. The density was calculated from sample displacement when immersed in mercury and the porosity was calculated assuming a cellulose density of 1500 kg/m3.
Dynamic mechanical analysis (DMA) was performed on freestanding films using a TA Instruments Q800 to determine the tensile properties of the coatings in Paper II and the films in Paper IV. The peel tests in Paper VI were also conducted with the DMA instrument. These measurements were performed on rectangular laminate samples (20x4 mm) at room temperature in a controlled stress‐strain mode with a preload force of 0.0010 N and a force ramp of 0.100 N/min.
Field emission scanning electron mircroscopy (FE‐SEM) was performed using a Hitachi S‐4800 scanning electron microscope. The specimens were coated by gold using an Agar HR sputter coater. The surfaces of the composites and the fracture surfaces were analyzed at 2 kV acceleration voltage.
Fourier transform infrared spectroscopy (FTIR) was conducted on a Perkin‐Elmer Spectrum 2000 FTIR equipped with a MKII Golden Gate, Single Reflection ATR system from Specac Ltd., London, UK. All spectra were normalized against a specific ATR crystal adsorption, thus enabling comparison between the polymer grafted cellulose substrates.99
Nuclear magnetic resonance (1H‐NMR and 13C‐NMR) spectra were recorded on a 400 MHz Bruker Aspect NMR, using DMSO‐d6 and CDCl3 as solvents.
Size exclusion chromatography (SEC), using THF (1.0 mL min‐1) as the mobile phase was performed at 35 °C using a Viscotek TDA model 301 equipped with two GMHHR‐M columns with TSK‐gel (Tosoh Biosep), a VE 5200 GPC autosampler, a VE 1121 GPC solvent pump, and a VE 5710 GPC degasser (all from Viscotek
corporation). A calibration method was created using narrow linear polystyrenes standards. Corrections for the flow rate fluctuations were made using toluene as an internal standard.
Static contact angle (CA) measurements were conducted on a KSV instruments CAM 200 equipped with a Basler A602f camera, using 5 μL droplets of MilliQ water and a relative humidity of 50%. The contact angles were determined using the CAM software.
Tensile tests of the films in Paper IV were performed with a Universal Materials Testing Machine from Instron, equipped with a 100 N or a 500 N load cell.
Specimens were 5 mm wide and thicknesses were uniform and 50 ‐ 230 μm depending on composition. Grip distance was 20 mm and testing was performed at a cross‐head speed of 2 mm/min. Specimens were conditioned and tested at 50% relative humidity and 23 °C. Displacement was measured by differential speckle photography (DSP). A pattern was prepared for the DSP by printer toner.
The toughness is defined as work‐to‐fracture calculated from the area under the stress‐strain curve.
Thermogravimetric analysis (TGA) was performed on a Mettler Toledo TGA/SDTA851e instrument. The samples in Paper I were heated from 40 °C to 800 °C with a heating rate of 10 °C/min, and a N2 or O2 flow of 50 mL/min. The samples in Paper V were heated from 30 °C to 600 °C with a heating rate of 10
°C/min, and a N2 flow of 80 mL/min.
Transmission electron microscopy (TEM) analysis was performed on a Philips Tecnai 10, using an acceleration voltage of 80 kV. The TEM samples were prepared from cured coatings, cast in epoxy and cut with an ultramicrotome.
UV‐curing of the coatings was performed with a Fusion System UV‐oven, equipped with an H‐bulb. The intensity of the UV‐light was 0.1 J/cm2, measured in the wavelength interval 320–390 nm.
X‐ray diffraction (XRD) analysis in Paper I was performed with a single crystal diffractometer STOE IPDS in transmission mode, using molybdenum radiation MoKα (λ≈0.71073 Å), and in Paper II with a PANalytical XPert Pro powder diffractometer in reflection mode, using copper radiation, CuKα1 (λ≈1.5406 Å).
