University of Groningen

Chemical modifications and applications of alternating aliphatic polyketonesZhang, Youchun

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2008

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Zhang, Y. (2008). Chemical modifications and applications of alternating aliphatic polyketones. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 05-02-2020

Chemical Modifications and Applications of Alternating Aliphatic Polyketones

Copyright © 2008 by Youchun Zhang. All right reserved.

No part of this book may be reproduced or transmitted in any forms by any means

without permission from the author.

ISBN: 978-90-367-3573-5

ISBN: 978-90-367-3576-6 (electronic version)

Cover design: Youchun Zhang and GrafiMedia

Printed by: GrafiMedia, Facilitair Bedrijf RUG, Groningen

RIJKSUNIVERSITEIT GRONINGEN

Chemical Modifications and Applications

of Alternating Aliphatic Polyketones

Proefschrift

ter verkrijging van het doctoraat in de Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen

op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op

vrijdag 10 oktober 2008 om 16.15 uur

door

Youchun Zhang geboren op 17 februari 1980

te Jiangsu, China

Promotor : Prof. dr. A.A. Broekhuis Copromotor : Prof. dr. F. Picchioni Beoordelingscommissie : Prof. dr. F. Ciardelli Prof. dr. J.A. Loontjens Prof. dr. H.J. Heeres

ISBN: 978-90-367-3573-5

Table of contents

1 Introduction 1

1.1 Historical overview of aliphatic polyketones 2 1.2 Properties and applications of alternating aliphatic polyketones 3 1.3 Functional polymers from alternating polyketones 5 1.4 Chemical product engineering 6 1.5 Aim and scope of this thesis 7 1.6 References 8

2 Polymeric amines by chemical modifications of alternating aliphatic polyketones 13

2.1 Introduction 14 2.2 Experimental 15 2.3 Results and discussion 18

2.3.1 Chemical modifications of polyketones 18 2.3.2 Characterization of polymeric amines 24

2.4 Conclusions 32 2.5 References 33

3 Wood adhesive emulsions from thermosetting alternating polyketones 35

3.1 Introduction 36 3.2 Experimental 37 3.3 Results and discussion 39

3.3.1 Emulsification 39 3.3.2 Emulsion stability 45 3.3.3 Rheology 50 3.3.4 Wood adhesive testing 53 3.4 Conclusions 55 3.5 References 55

4 Thermally self-healing polymeric materials: the next step to recycling thermoset polymers? 57

4.1 Introduction 58 4.2 Experimental 59 4.3 Results and discussion 61

4.3.1 Synthesis of furan-functionalized polyketones 61 4.3.2 Diels-Alder and Retro-Diels-Alder reaction of PK-furan 66 4.3.3 Thermal self-healing 72 4.4 Conclusions 78 4.5 References 78

5 Cross-linking of Multi-walled carbon nanotubes with polymeric amines 81

5.1 Introduction 82 5.2 Experimental 83 5.3 Results and discussion 85

5.3.1 Grafting of polyamines onto MWNTs 85 5.3.2 Solubility 89 5.3.3 SEM and TEM 90 5.3.4 Blends with polyethylene 93 5.4 Conclusions 95 5.5 References 95

6 Cell behavior on polymeric amines derived from alternating polyketones 99

6.1 Introduction 100 6.2 Experimental 101 6.3 Results and discussion 102

6.3.1 Characterization of polyamine films 102 6.3.2 Cell culture studies 107

6.4 Conclusions 113 6.5 References 113

Summary 115

Samenvatting 119

Acknowledgements 123

Introduction

Chapter 1

Introduction

Abstract

A general overview on historical development, synthesis, catalysis, properties, applications, and chemical modifications of polyketones is provided. In particular, applications of polyketones as such and their deviratives obtained after chemical modifications are discussed. Chemical product engineering, an emerging paradigm within the chemical engineering sciences, is introduced and the framework of chemical product design is discussed. Finally, the aim and scope of this thesis is presented.

1

Chapter 1

1.1 Historical overview of aliphatic polyketones

Aliphatic polyketones, a new family of polymers, are produced by copolymerization of carbon monoxide (CO) and various unsaturated hydrocarbon monomers (e.g. olefins). When compared to the corresponding polyolefins, these polyketones have the following advantages: (i) CO as monomer is a particularly low-cost feedstock; (ii) the polymers can act as excellent precursors for the preparation of functional polymers by chemical modifications using the recurring highly reactive 1,4-di-carbonyl functionality; (iii) the reactive carbonyl groups in the backbone confer photo- and biodegradability; (iv) they can be used for a variety of applications due to their unique physical and chemical properties.

The first copolymerization of CO and ethylene, leading to random aliphatic polyketones, was discovered in 1941 by the researchers of Farbenfabriken Bayer using extreme conditions (230 °C, 2000 atm).1 One decade later, alternating aliphatic polyketones were reported by Reppe and Magin using a nickel-based catalyst in water under relatively mild conditions (200 °C, 200 atm).2 Since then, considerable efforts have been made by both academia and industry towards the development of new and industrially viable catalytic systems. However, only in the 1980s, a technological breakthrough was achieved by Shell researchers with the development of a highly efficient homogeneous palladium-based catalytic system.3-6 This catalyst, a palladium (II) complex bearing bidentate phosphine or nitrogen ligands with a Brønsted acid as co-catalyst, enabled the production of high molecular weight, perfectly alternating copolymers of carbon monoxide and olefins. This catalytic system can also be applied to mixtures of olefins, which in turn can be simple aliphatic or heteroatom functionalized ones. Production yields that satisfied the economic requirements for industrial production were achieved, i.e. 6 (kg of polymer) (g of Pd)-1h-1 under mild conditions (90 °C, 4-5 MPa).

Following the discovery of these catalyst systems, alternating polyketones have attracted considerable interest from both academia and industry.7-15 The polymerization mechanisms to achieve alternating polyketones have been determined by Sen et al.8 and Drent et al.9. In 1996, after nearly 20 years of intensive research, Shell introduced a number of CO/ethylene/propylene-based terpolymers, a class of engineering thermoplastics, in the plastics market under the trademark of Carilon.16 However, Shell withdrew the business of Carilon Polymers from the market in 2001 because of a strategy change in their chemical business. Nevertheless, research on catalyst improvements and polyketone application development continued in academia. Particularly interesting is the preparation of functionalized polyketone derivatives, which may open many new applications in different research fields (e.g. material and biomedical sciences) as demonstrated in the present thesis.

2

Introduction

1.2 Properties and applications of alternating aliphatic polyketones

Alternating polyketones (Figure 1.1) are produced by the copolymerization of CO and alkenes, dienes, styrene and its derivatives, using catalysts based on transition metal compounds. The obtained polymers range from amorphous liquids of low molecular weight (thermosets) to crystalline solids of high molecular weight (thermoplastics). The molecular weight depends on the reaction conditions, the type of olefins, and the type of catalyst.

Figure 1.1 Scheme of perfectly alternating polyketones (R=H, CH3, C6H5, etc.).

Most of the product and research & development is focused on the synthesis of aliphatic copolymers of CO with ethylene, propylene or mixtures of olefins.9 High molecular weight copolymers of ethylene and CO, in a perfectly 1:1 alternating order, represent the simplest members of the family of alternating aliphatic polyketones. Their alternating chemical structure has been thoroughly studied and confirmed by elemental analysis, nuclear magnetic resonance (NMR), and FTIR spectroscopy.17-19 These polymers are semi-crystalline thermoplastics with levels of crystallinity in the range of 35-50%.10 Two different crystalline structures (α and β form) in an all-trans configuration have been detected by X-Ray diffraction.20-23 Semicrystalline polyketones are insoluble in common organic solvents and only dissolve in highly polar and acidic solvents (e.g. hexafluoroisopropanol, m-cresol). The strong interaction between polymer chains and the high stereo-regularity of the ethylene-CO polymers lead to highly crystalline materials with a relatively high melting point (Tm, typically between 250 °C and 260 °C). However, this high melting range of CO-ethylene copolymers is close to the thermal decomposition temperature which makes their melt processing very difficult. In addition to thermal degradation, the high reactivity of the 1,4-di-carbonyl moiety often results in condensation reactions, leading to the formation of furan structures at elevated temperatures. As a result, ethylene/CO copolymers have limited practical use in the applications that require the melt-processing of the polymers.

Terpolymers of CO, ethylene, and propylene show a reduced Tm due to the incorporation of a certain amount of propylene/CO segments into the polymer backbone.19 The melting point of the terpolymers is dependent on the number of propylene units in the polymer

3

Chapter 1

backbone. For instance, the incorporation of 6% and 17% propylene leads to a melting point depression till 220 °C and 170 °C, respectively. The terpolymers possess excellent properties like fast crystallization rate, better mechanical properties in comparison to other thermoplastics (e.g. polyamide-6, polypropylene, low and high density polyethylene), high chemical resistance to a broad range of chemicals, and good barrier properties to gases (e.g. oxygen) and hydrocarbon fuels.24-25 These properties enable aliphatic polyketones to be used in a variety of applications: as fibers26-33, in polymer blends34-43 and polymer composites44-52, in packaging 53-60, as flame resistant materials61-65, etc.

Another class of alternating aliphatic polyketones is represented by thermosetting polyketone oligomers (the main focus of this thesis) produced by the co- and ter-polymerization of CO, ethylene, and proplyene.66-68 Like high molecular weight polyketones, thermosetting oligomers are synthesized in an organic solvent using a homogeneous palladium-based catalyst at a reaction temperature of 70–100 °C and a pressure of typically up to 8.5 MPa. The product composition is a function of the molar ratio between propylene and ethylene in the oligomers, ranging from viscous resins for the ethylene-free types to waxy or melting solids at ethylene contents of about 50% based on total olefin content. Because of the presence of highly active carbonyl groups, chemical conversion of thermosetting polyketones can be carried out in a variety of ways (e.g. condensation to polyfurans, reduction to poly-alcohol, and condensation with commercial formaldehyde resins).69 The Paal-Knorr reaction, involving pyrrole ring formation between an amino group and two adjacent carbonyl groups, has been used as a powerful cross-linking tool for these materials (Figure 1.2). In this curing reaction, aromatic pyrrole units are formed through the elimination of water, using a variety of di- or multi-functional amino-based curing agents familiar to the epoxy-resin technology. Product and application development of these thermosetting resins mainly focused on coating70-71, wood adhesives72-75, electronic adhesives76-78, and polymer composites, which are thermosetting areas dominated by commercial resins like epoxy, urea-formaldehyde or phenol-formaldehyde. All these applications require relatively fast cure chemistry, easy handling, and the absence or control over volatile and toxic components.

O

O

O+ NH2 R NH2

NO

R

NO

H2O+

Figure 1.2 Paal-Knorr curing chemistry for thermosetting polyketones.

4

Introduction

1.3 Functional polymers from alternating polyketones

Functional polymers are by definition those that bear specified chemical groups or have specified physical, chemical, biological, pharmacological applications which in turn depend on the presence of specific chemical groups.79 The functional groups could be part of the polymer backbone or be linked to the main chain as pendant groups (either directly or via a spacer group). Functional polymers have found numerous applications like organic electronics, chemical sensors, medical devices, artificial organs, fuel cells, nano devices, etc., in a great variety of areas.80-88 Direct polymerization of functional monomers (e.g. anionic, cationic, radical, and metal catalyzed polymerization) and chemical modification of synthetic polymers are considered as the two general routes for the preparation of functional polymers.89 Each of the two approaches has its own advantages and disadvantages. One approach may be suitable for the preparation of a particular functional polymer whereas the other would be totally impractical. In general, the polymerization route can lead to polymers with a homogeneous and uniform structure (e.g. distribution of the functional groups within a single chain and between different chains), while the use of chemical modification enables the creation of new classes of polymers which are difficult to synthesize by direct polymerization of the monomers. The design simplicity and the ease of synthesis are the accepted main features of the chemical modification route. The choice between the two different approaches is mainly dependent on the required chemical and physical properties of the polymers and the feasibility of the desired chemistry.

Alternating polyketones constitute a very interesting class of polymers as they can be applied as starting materials for the preparation of functional polymers by chemical modifications. As these polymers contain the highly reactive 1,4-di-carbonyl functionality, the systems can readily be converted into a great variety of polymers containing functional groups such as pyrroles90, furans91, thiophenes91, bisphenols92, alcohols93-94, ketals95-96, thiols97, oximes98-99, methylene100, cyanohydrins101, etc., via different reaction pathways (Figure 1.3). Among these modifications, the classic Paal-Knorr reaction, in which the 1,4-di-carbonyl moiety of the polyketones reacts with a primary amine function yielding a pyrrole unit, is one of the dominating reaction routes for the functionalization of alternating polyketones.74,102-106 The obtained poly-pyrrole derivatives exhibit good solubility in common organic solvents. The pyrrole function is also one of the most important heterocycles and the key unit in many biologically active compounds and synthetic pharmaceuticals.107 The Paal-Knorr reaction can be carried out under mild experimental conditions without the need for any catalysts and solvent. Particularly interesting is the preparation of polymeric amines via this type of reaction74, which has been main focus study in this thesis. Polymers containing amino functionality (e.g. primary, secondary, and

5

Chapter 1

tertiary amines) have found wide applications in various areas, such as chelating agents for metal ion108, polymeric surfactants109, polyelectrolytes110, ion exchange resin111, DNA carriers for gene delivery112, polymeric pharmaceuticals113, and antibacterial and bacterium adsorbing polymers114.

NR SO

OOH

HO OR

NOMe

HO CN

SH

HO P(OR)2O

NR2

HO OH

O O

R

O O

Figure 1.3 Scheme of chemical modifications of polyketones.

1.4 Chemical product engineering

Chemical Engineering is by definition the branch of engineering that deals with the application of natural sciences (e.g. chemistry and physics) with mathematics to the process of converting raw materials or chemicals into more useful or valuable forms.115 However, the activities of chemical industries during the last few decades have been undergoing dramatic changes: from commodity chemicals (large quantity, low value added) to specialty chemicals (small quantity, high value-added, e.g. pharmaceuticals, functional polymers) and to even more sophisticated structured chemical products (e.g. paints, personal care or household products, biomedical devices, and semiconductors).116-118 To address the current industrial demands, chemical engineering sciences have been reacting to this shift in industrial development.119,120 As one of the main results, Chemical Product Design and Engineering (CPDE) is becoming a well-established branch of Chemical Engineering.119-126 It can be defined as the discipline which combines science and technology for the development and production of structured chemical products able to

6

Introduction

meet the demands and requirements of society. This is usually achieved by improving existing products and designing new ones.124 This goal and the corresponding research methodology combine market pulls with technological pushes to arrive at unique products.119,120 A new conceptual model or structure for this new discipline (CPDE) has also been proposed by Moggridge et al.,126 in which three fundamental inter-related pillars were suggested to support the major objective of designing new chemical products: (i) the chemical product pyramid, i.e. the combination of usage functions, process, and property in the design of a chemical product; (ii) the integration of chemical product and process design; (iii) a multifaceted approach considering product design at different dimensional levels.

Chemical product design (CPD), considered as the core part of CPDE, is defined as a systemic procedure or framework of methodologies and tools, whose aim is to provide a more efficient and faster design of chemical products able to meet market needs.126 CPD has recently been introduced into the Chemical Engineering curriculum at a number of universities because of its well-accepted importance.121,126-129 CPD, according to Cussler et al., mainly comprises four essential steps: (i) identification of customers needs; (ii) generation of product ideas; (iii) selection of the most promising ideas; (iv) development of a process to manufacture the products.130-131 Other similar systematic frameworks about the design of chemical products have also been proposed through many detailed case studies of chemical-based products.131-140 A toolbox of molecular structure-property relations, based on the maturity of current scientific understanding, has been developed by Wei119,141 to design as well as manufacture new materials or products with desired properties. The available methods proposed by Wei include the use of additives and blending, addition and substitution of functional groups, isomerization and skeletal rearrangements, and the cross-linking of molecular chains. It should be pointed out that a multidisciplinary team is generally required for a successful product development due to the complexity of this concurrent process and product creation activity.

1.5 Aim and scope of this thesis

The aim of this thesis is to enhance the fundamental understanding regarding the development of new products and product applications by using aliphatic alternating polyketones based on low molecular weight ethylene/propylene/CO co- and terpolymers. The work has been carried out within several research disciplines and covers the full pathway from molecular design (i.e. chemical modifications of polyketones) to product development. New chemical products based on polyketones have been developed with good performance: (i) polymeric surfactants; (ii) water-borne formaldehyde-free wood adhesives; (iii) thermally self-healing polymeric materials; (iv) polymeric interlinkers for

7

Chapter 1

carbon nanotubes; (v) new biomaterials for controlling cell behavior. The layout of this thesis is given below.

Chapter 2 focuses on chemical modifications of alternating polyketones by using a series of di-amines via the Paal-Knorr reaction to prepare polymeric amines (polyamines). The chemical reactivity of the polyketones with the di-amines and characterization of the obtained polyamines in water as polymeric surfactants have been studied.

Chapter 3 describes the preparation of wood adhesive emulsions. The stability and structure of the emulsions were thoroughly studied at different experimental conditions with respect to the storage time (up to 2 years) at room temperature. The performance of the wood adhesives was evaluated according to the European Standard (EN-314) for wood adhesive testing.

Chapter 4 describes a thermally self-healing polymer system on the basis of furan functionalized alternating olefin-CO polyketones and bismaleimide by using the Diels-Alder and Retro-Diels-Alder reaction scheme. The self-healing ability of this system has been studied by using several analytical and mechanical testing techniques, which prove excellent self-healing performance.

Chapter 5 reports an approach for the preparation of carbon nanotube interconnects by using polymeric amines to functionalize and cross-link multi-walled carbon nanotubes (MWNTs) via the amidation reaction. The obtained systems have been characterized using several analytical and microscopy techniques.

Chapter 6 describes the use of polymeric amines for biomedical applications. The response and behavior of vascular smooth muscle cells and bovine arterial endothelial cells upon exposure to the polyamine films and polyamine solutions were investigated in vitro.

1.6 References

1. Ballauf, F.; Bayer, O.; Teichmann, L. DE 863711, 1953. 2. Reppe, W.; Magin, A. US 2577208, 1951. 3. Drent, E. EP 0121965, 1984. 4. Drent, E. EP 0181014, 1986. 5. Van Broekhoven, J. A. M.; Drent, E.; Klei, E. EP 0213671, 1987. 6. Van Broekhoven, J. A. M.; Drent, E. EP 0235865, 1987. 7. Sen, A. Adv. Polym. Sci. 1986, 73/74, 125. 8. Sen, A. Acc. Chem. Res. 1993, 26, 303. 9. Drent, E.; Budzelaar, P. H. M. Chem. Rev. 1996, 96, 663. 10. Sommazzi, A.; Garbassi, F. Prog. Polym. Sci. 1997, 22, 1547.

8

Introduction

11. Bianchini, C.; Meli, A. Coord. Chem. Rev. 2002, 225, 35. 12. Sen, A. Catalytic synthesis of alkene carbon monoxide copolymers and cooligomers,

Catalysis by metal complexes, Volume 27, 1st ed., Kluwer Academic Publishers, 2003. 13. Belov, G. P.; Novikova, E. V. Russ. Chem. Rev. 2004, 73, 267. 14. Nozaki, K.; Hiyama, T. J. Organomet. Chem. 1999, 576, 248. 15. Durand, J.; Milani, B. Coord. Chem. Rev. 2006, 250, 542. 16. Alperowicz, N. Chemical Week 1995, July 2, 22. 17. Lai, T. W.; Sen, A. Organometallics 1984, 3, 866. 18. Zhao, A. X.; Chien, J. C. W. J. Polym. Polym. Chem. 1992, 30, 2735. 19. Drent, E.; Van Broekhoven, J. A. M.; Doyle, M. J. J. Organom. Chem. 1991, 417, 235. 20. Lommerts, B. J.; Klop, E. A.; Aerts, J. J. Polym. Sci. Polym. Phys. 1993, 31, 1319. 21. Lommerts, B. J. Ph.D thesis, University of Groningen, 1994. 22. Lagaron, J. M.; Vickers, M. E.; Powell, A. K.; Davidson, N. S. Polymer 2000, 41, 3011. 23. Lagaron, J. M.; Vickers, M. E.; Powell, A. K.; Bonner, J. G. Polymer 2002, 43, 1877. 24. Wakker, A.; Kormelink, H. G.; Verbeke, P.; Jordaan, J. C. M. Kunststoffe 1995, 85,

1056. 25. SRI international, Carilon thermoplastic polymers, Information sheet, www.sri.com. 26. Van Breen, A. W.; Van Deursen, J. H.; Wetser, F. I. M. EP 0360358, 1990. 27. Brown, H. S. US 5597389, 1997. 28. Shin, H. WO 9733021, 1997. 29. Middeljans, H.; Veurink, J.; Juijn, J. A. WO 9941437, 1999. 30. Taniguchi, T.; Morita, T. JP 2001131825, 2001. 31. Taniguchi, R.; Morita, T. JP 2001207330, 2001. 32. Taniguchi, T.; Morita, T. JP 2002309442, 2002. 33. Kato, J.; Morita, T.; Fujieda, K. US 7015303, 2006. 34. Lutz, R. G. US 4814378, 1989. 35. Lutz, R. G.; Gergen, W. P. US 4880908, 1989. 36. Gergen, W. P.; Lutz, R. G. US 4904744, 1990. 37. Gergen, W. P.; Machado, J. M.; Waters, D. G.; Gingrich, R. P. US 5043389, 1991. 38. Machado, J. M. EP 0457374, 1991. 39. George, E. R.; Korcz, W. H. US 5084518, 1992. 40. Machado, J. M. US 5175210, 1992. 41. George, E. R. US H1404, 1995. 42. Marklund, E.; Gedde, U. W.; Hedenqvist, M. S.; Wiberg, G. Polymer 2001, 42, 3153. 43. Zuiderduin, W. C. J.; Vlasveld, D. P. N.; Huetink, J.; Gaymans, R. J. Polymer 2004, 45,

3765. 44. Van Breen, A. W.; Gillemans, J. C. M. US 4861428, 1989.

9

Chapter 1

45. George, E. R. US 4880865, 1989. 46. Kammula, S. C.; Wife, R. L. GB 2217720, 1989. 47. Hanley, S. J. US 5034431, 1991. 48. Proctor, M. G.; Jordaan, J. C. M. EP 0474309, 1992. 49. Hanley, S. J. US 5114992, 1992. 50. Hanley, S. J.; Dolce, T. J. EP 0474308, 1992. 51. Machado, J. M. EP 0486104, 1992. 52. George, E. R. US 5166252, 1992. 53. Machado, J. M. US H1169, 1993. 54. Matsuki, Y; Yoshimura, I. JP 6226925, 1994. 55. George, E. R. US H1404 1995. 56. Bonner, J. G.; Powell, A. K. WO 9606889, 1996. 57. Lofgren, L.; Oveby, C. WO 9833647, 1998. 58. Eggers, H.; Lund, K.; Brandt, R.; Sperlich, B. EP 0920991, 1999. 59. Eggers, H.; Lund, K.; Brandt, R.; Sperlich, B. WO 9928130, 1999. 60. Taniguchi, K. JP 2001200154, 2001. 61. Danforth, R. L.; Smith, D. G. US 4921897, 1990. 62. Notorgiacomo, V. J. EP 0448176, 1991. 63. Lutz, R. G. US 4761449, 1988. 64. Smaardijk, A. A.; De Wit, H. DE 19907533, 1999. 65. Gosens, J. C.; De Wit, G. WO 0012608, 2000. 66. Drent, E.; Keijsper, J. J. US 5225523, 1993. 67. Mul, W. P.; Dirkzwager, H.; Broekhuis, A. A.; Heeres, H. J.;Van der Linden, A. J.;

Orpen, A. G. Inorg. Chim. Acta 2002, 327, 147. 68. SRI international, Carilite thermoset resins, Information sheet, www.sri.com. 69. Smaardijk, A. A.; Kramer, A. H. EP 0372602, 1990. 70. Broekhuis, A. A.; Van Hek, J. R.; Kooijmans, P. G. US 6309527, 2001. 71. Broekhuis, A. A.; Kooijmans, P. G.; Van Ogtrop, J. US 6576688, 2003. 72. Van Der Heide, E.; Vietje, G.; Wang, P. C. US 5684080, 1997. 73. Wong, P. K.; Pace, A. R.; Weber, R. C. US 5955563, 1999. 74. Broekhuis, A. A.; Freriks, J. US 5952459, 1999. 75. Wong, P. K. US 6214941, 2001. 76. Shridhar, R.; Wong, P. K. US 5760337, 1998. 77. Shridhar, R.; Wong, P. K. US 5726391, 1998. 78. Shridhar, R.; Wong, P. K. US 5840215, 1998. 79. Horie, K. et al. IUPAC, Pure Appl. Chem. 2004, 76, 889. 80. Jagur-Grodzinski, J. React. Funct. Polym. 1999, 39, 99.

10

Introduction

81. McQuade, D. T.; Pullen, A. E.; Swager, T. M. Chem. Rev. 2000, 100, 2537. 82. Frechet, J. M. J. Prog. Polym. Sci. 2005, 30, 844. 83. Gil, E. S.; Hudson, S. M. Prog. Polym. Sci. 2004, 29, 1173. 84. Wegner. G. Polym. Adv. Technol. 2006, 17, 705. 85. Dhal, P. K.; Holmes-Farley, S. R.; Huval, C. C.; Jozefiak, T. H. Adv. Polym. Sci. 2006,

192, 9. 86. Liu, T.; Burger, C.; Chu, B. Prog. Polym. Sci. 2003, 28, 5. 87. Inoue, K. Prog. Polym. Sci. 2000, 25, 453. 88. Kumar, A.; Srivastava, A.; Galaev, I. Y.; Mattiasson, B. Prog. Polym. Sci. 2007, 32,

1205. 89. Akelah, A.; Moet, A. Functionalized polymers and their applications, 1st ed., Chapman

and Hall, 1990. 90. Kiovsky, T. E.; Kromer, R. C. US 3979374, 1976. 91. Jiang, Z.; Sanganeria, S.; Sen, A. J. Polym. Sci. Polym. Chem. 1994, 32, 841. 92. Fitko, C. W.; Ravve, A. US 3317472, 1967. 93. Drent, E. US 5071926, 1991. 94. Lommerts, B. L.; Sikkema, D. J. Macromolecules 2000, 33, 7950. 95. Green, M. J.; Lucy, A. R.; Lu, S.; Paton, R. M. J. Chem. Soc., Chem. Commun. 1994,

2063. 96. Lucy, A. R. US 5091486, 1992. 97. Scott, S. L.; Del, W. US 2495293, 1950. 98. Lu, S.; Paton, R. M.; Green, M. J.; Lucy, A. R. Eur. Polym. J. 1996, 32, 1285. 99. Khansawai, P.; Paton, R. M.; Reed, D. Chem. Commun. 1999, 1297. 100. Nozaki, N.; Kosaka, N.; Grabner, V. M.; Hiyama, T. Macromolecules 2001, 34, 6167. 101. Brubaker, M. M.; Coffman, D. D.; Hoehn, H. H. J. Am. Chem. Soc. 1952, 74, 1509. 102. Chen, J.; Yeh, Y.; Sen, A. J. Chem. Soc., Chem. Commun. 1989, 965. 103. Sen, A.; Jiang, Z.; Chen, J. Macromolecules 1989, 22, 2012. 104. Brown, S. L. US 5081207, 1992. 105. Sinai-Zingde, G. D. US 5605988, 1997. 106. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F. J. Appl. Polym. Sci. 2008,

107, 262. 107. Jones, R. A. Pyrroles, Wiley, 1990. 108. Rivas, B. L.; Pereira, E. D.; Moreno-Villoslada, I. Prog. Polym. Sci. 2003, 28, 173. 109. Liu, S.; Armes, S. P. Curr. Opin. Colloid Interface Sci. 2001, 6, 249. 110. Radeva, T. Physical chemistry of polyelectrolytes, CRC Press, 2001. 111. Zagorodni, A. A. Ion exchange materials: properties and applications, Elsevier, 2006. 112. De-Smedt, S. C.; Demeester, J.; Hennink, W. E. Pharmacol. Res. 2000, 17, 113.

11

Chapter 1

113. Dhal, P. K.; Huval, C. C.; Holmes-Farley, S. R. Ind. Eng. Chem. Res. 2005, 44, 8593. 114. Tashiro, T. Macromol. Mater. Eng. 2001, 286, 63. 115. http://en.wikipedia.org/wiki/Chemical_engineering. 116. Villadsen, J. Chem. Eng. Sci. 1997, 52, 2857. 117. Wintermantel, K. Chem. Eng. Sci. 1999, 54, 1601. 118. Kind, M. Chem. Eng. Process. 1999, 38, 405. 119. Wei, J. Ind. Eng. Chem. Res. 2002, 41, 1917. 120. Cussler, E. L.; Wei, J. AICHE J. 2003, 49, 1072. 121. Wesselingh, J. A. Powder Technol. 2001, 119, 2. 122. Favre, E.; Marchal-Heusler, L.; Kind, M. Trans IchemE-Part A. 2002, 80, 65. 123. Broekhuis, A. A. Chem. Eng. Res. Des. 2004, 82, 1409. 124. Voncken, R. M.; Broekhuis, A. A.; Heeres, H. J.; Jonker, G. H. Chem. Eng. Res. Des..

2004, 82, 1411. 125. Edwards, M. F. Chem. Eng. Res. Des. 2006, 84, 255. 126. Costa, R.; Moggridge, G. D.; Saraiva, P. M. AICHE J. 2006, 52, 1976. 127. Saraiva, P. M.; Costa, R. Trans IchemE-Part A 2004, 82, 1474. 128. Kavanagh, L.; Lant, P. Trans IchemE-Part D 2006, 1, 66. 129. Wei, J. Ind. Eng. Chem. Res. 2008, 47, 1. 130. Moggridge, G. D.; Cussler, E. L. Trans IchemE-Part A 2000, 78, 5. 131. Cussler, E. L.; Moggridge, G. D. Chemical product design, Cambrigde University

Press, 2001. 132. Charpentier, J. C. Chem. Eng. Sci. 2002, 57, 4667. 133. Rahse, W.; Hoffmann, S. Chem. Eng. Technol. 2003, 26, 931. 134. Wibowo, C.; Ng, K. M. AICHE J. 2002, 48, 1212. 135. Fung, K. Y.; Ng, K. M. AICHE J. 2003, 49, 1193. 136. Hill, M. AICHE J. 2004, 50, 1656. 137. Gani, R. Comput. Chem. Eng. 2004, 28, 2441. 138. Ng, K. M.; Gani, R.; Dam-Johansen, K. Chemical product design: toward a

perspective through case studies, Elsevier, 2007. 139. Wesselingh, J. A.; Kiil, S.; Vigild, M. E. Design and development of biological,

chemical, food and pharmaceutical product, Wiley, 2007. 140. Brockel, U.; Meier, W.; Wagner, G. Product design and engineering, Wiley, 2007. 141. Wei, J. Product engineering–molecular structure and properties, Oxford University

Press, 2006.

12

Polymeric amines

Chapter 2

Polymeric amines by chemical modifications of

alternating aliphatic polyketones

Abstract

Alternating aliphatic polyketones were chemically modified using di-amines to obtain polymeric products having pendant amino groups. The used Paal-Knorr reaction involves the formation of pyrrole rings along the polyketone backbone. The corresponding kinetics and final conversions are clearly dependent, among others, on statistical factors (two adjacent carbonyls must react in order to obtain ring formation) as well as on the steric hindrance (sterically hindered amino groups react very slowly). The corresponding reaction products (polymeric amines) display very interesting physical properties in aqueous solution such as formation of micelles at low protonation level, polyelectrolyte behaviour at high protonation level. These properties have been characterized by using dynamic light scattering, Cryo-electron microscopy, and drop tensiometry. Key word: Chemical modifications; Polymeric amines; Polymeric surfactants; Paal-Knorr

Based on: Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F. Polymeric amines

by chemical modifications of alternating aliphatic polyketones, J. Appl. Polym. Sci. 2008,

107, 262-271.

13

Chapter 2

2.1 Introduction

The synthesis of polymers containing amine groups has stimulated an enhanced interest in recent years due to their wide applications as emulsifiers, drug carriers, DNA-carriers for gene delivery, cation-exchange resins, and pharmaceutical uses.1-4 However, the high reactivity of the amine groups due to their inherent ability to act as bases, reducing agents, nucleophiles or as ligands for transition metals, make the preparation of this kind of polymers quite difficult and problematic.5 Alternating polyketones can act as precursors for the preparation of polymeric amines (polyamines) by chemical modifications via the classic Paal-Knorr reaction, in which the 1,4-di-carbonyl moiety of the polyketones reacts with a primary amine function yielding a pyrrole unit. In contrast to a variety of polymerization mechanisms such as anionic, cationic or radical ones, the route of chemical modifications of polyketones described here just consists of a two components/one-pot reaction without the need of any catalysts and organic solvent using mild conditions during the whole process. It represents a simple, low cost, and straightforward way to prepare the polyamines.

In the present work, we report the synthesis of a family of polyamines by chemical modifications of alternating aliphatic polyketones with a set of functional di-amines via the Paal-Knorr reaction. The resulting polyamines possess double functionalities: the N-substituted 2,5-pyrrole-diyl group incorporated in the polymer backbone and a substituent containing an amino functionality which can be either primary, secondary or tertiary and both aliphatic or aromatic. The factors affecting the chemical reactivity of the polyketones with the di-amines such as reaction conditions (time, temperature, and initial molar ratio between amino groups of the di-amines and 1,4 arrangement of di-carbonyl groups on the polyketones), ethylene content of the polyketones, and chemical structure of the di-amines were thoroughly studied. The polyamines are expected to transform into water-soluble cationic species by simple protonation with weak acid or by quaternization of the amino groups. Thus physical properties of the polyamines as water dispersion were characterized by a combination of dynamic light scattering (DLS), surface tension, and Cryo-transmission electron microscopy (Cryo-TEM). Furthermore, as aromatic rings (pyrroles) are present in the backbone of the polymer, a study of the fluorescence property of the polyamines was carried out in water solution.

14

Polymeric amines

2.2 Experimental

Materials. The alternating polyketones with 0% ethylene (PK0, Mw-1680), 30% ethylene (PK30, Mw-3970), and 50% ethylene (PK50, Mw-5350) based on the total olefin content were synthesized according to a reported procedure.6-7 1,2-diaminopropane (1,2-DAP, Acros, 99%), 1,3-diaminopentane (1,3-DAPe, Aldrich, 98%), 2-(Dimethylamino) ethylamine (2-DAEA, Aldrich, ≥98%), 2-diethylaminoethylamine (2-DEAEA, Aldrich, 95%), 4-Picolylamine (4-PcA, Aldrich, ≥97%), 2,5-hexanedione (Aldrich, ≥99%), acetic acid (Acros, ≥99.7%), methyl iodide (Aldrich, 99%), and tetrahydrofuran (THF, Acros, ≥99%) were purchased and used as received.

Model component reaction. 2,5-hexanedione (0.3 mmol) was added to 1,2-diaminopropane (0.3 mmol) in 0.7 ml of CDCl3. The resulting mixture was shaken for 2 min and transferred into an NMR tube (5 mm in diameter). The progress of the reaction was monitored with 1H NMR spectroscopy at 50 °C for 11.25 h. The spectra were recorded at regular time intervals of 25 min.

Chemical modifications of the polyketones. The chemical modifications were carried out in a sealed 250 ml round bottom glass reactor with a reflux condenser, a U-type anchor impeller, and an oil bath for heating. In order to calculate the stoichiometry of the Paal-Knorr reaction, the amount of 1,4-arrangement of di-carbonyl reactive species in 40 g polyketones were first calculated. Here, PK30 was taken as example. The repeating units of CO/propylene and CO/ethylene is C8H12O2 with a mol mass of 140 and C6H8O2 with a mol mass of 112, respectively. Thus, the average mol mass of the repeating unit for PK30 is (0.7×140)+ (0.3×112)=131.6, so that 40 g of PK30 contains 40/131.6=0.3 mol reactive species. The amount of di-amines (e.g. 1,2-DAP) when used for reaction in equimolar ratio between the di-amines and di-carbonyl groups of the polyketones is 0.3×74.12 (Mw of 1,2-DAP)=22.24 g. After the polyketones (40 g) were preheated to the liquid state at the employed reaction temperature (80-100 °C), the di-amines were added dropwise into the reactor in the first 20 min. The stirring speed was set at a constant value of 500 rpm. During the reaction, the mixture of the reactants changed from the slight yellowish, low viscous state into a highly viscous brown homogeneous paste. The reaction conversion was monitored using potentiometric titrations. After the reactions were completed and the polymers cooled down to room temperature, the products derived from the 1,2-DAP and 4-PcA changed into rigid solid material, while the other products appeared as waxy or soft materials at room temperature. The resulting polyamines were washed several times with de-ionized Milli-Q water to remove unreacted di-amines. After filtering and freeze-drying, light brown polymers were obtained as the final products.

15

Chapter 2

Preparation of polymer solutions. The polyamines as such are not soluble in water. However, they display good solubility in acidic water. Accordingly, 100 mg/ml stock solutions of PK30 modified with 1,2-DAP and 1,3-DAPe in de-ionized Milli-Q water were prepared by adding the appropriate amount of acetic acid solution to match a desired protonation level and stirring with a magnetic stir-bar at 80 °C for 1.5 h. The tertiary amino groups of PK50 modified with 2-DAEA, 2-DEAEA, and 4-PcA, and PK30 modified with 4-PcA were quaternized by treating the polymers with a ten molar excess of methyl iodide over the amino groups in THF at room temperature for 24 h. The solvent and excess methyl iodide were removed under reduced pressure. The 10 mg/ml stock solutions of the quaternized polymers were prepared in the de-ionized Milli-Q water by stirring with a magnetic stir-bar at 80 °C for one hour and then passing through 200 nm syringe filters after complete dissolution. All solutions were equilibrated overnight before any further measurements were performed.

Nuclear magnetic resonance (NMR) and and FTIR spectroscopy. 1H- and 13C-NMR spectra were recorded on a Varian Mercury Plus 400 MHz spectrometer using CDCl3 as a solvent. FTIR spectroscopy was performed on a Perkin-Elmer Spectrum 2000 and spectra of the polymers were recorded by using KBr pellets.

Potentiometric titrations. Potentiometric titrations were performed at room temperature for determination of the amine number of the modified polyketones using a titrator (702 SM Titrino, Metrohm) with an automatic burette of 10 ml capacity. A perchloric acid (HClO4, 0.1 M) solution in water, standardized by potassium hydrogen phthalate, was used as the titrating agent. A THF/water (9:1) mixture, which is able to solubilize the polyamines at all protonation levels (i.e. during all the titrations), was used as titration solvent (120 ml) for the samples (0.2-0.3 g). The inflection point was identified from the titration curve. Three or four measurements were performed for each reaction sample. The average amine number was taken and the obtained standard error of the average value was less than 1%. The conversion of the di-amines, Xamine, and carbonyl groups of the polyketones, XCO, were determined by the following formulas:

Xamine= (Ai – Af)/A×100% (1) XCO= Xamine × Iam/CO ×100% (2)

where Ai is the amine number before the modification, Af is the amine number after the modification, A is the standard amine number for 100% conversion value, Iam/CO is the initial reaction molar ratio between the di-amines and 1,4-arrangement of di-carbonyl groups of the polyketones.

16

Polymeric amines

Dynamic light scattering. The dynamic light scattering (DLS) measurements were performed on a Zetasizer 5000 instrument (Malvern Instruments, UK) at a wavelength of 633 nm and at a temperature of 25 °C. Scattered light was detected at a 90 degree angle. The viscosity (0.89 mPa s) and the refractive index (1.33) of water at 25°C were used for data analysis. The intensity autocorrelation functions obtained from the DLS were analyzed by using CONTIN algorithm for all measurements. The apparent hydrodynamic diameter, DH, was obtained using the Stokes-Einstein relation:

D0 = kT/ 3π ηDH (3)

where D0 is the diffusion coefficient, k is the Boltzmann constant, T is the absolute temperature, and η is the viscosity of the solution.

Drop tensiometry. Surface tension measurements on the polymer solutions were performed by using a drop tensiometer (Lauda, TVT-1), with a Lauda RM6 temperature controller. The temperature was controlled at 25±0.1 °C during the period of measurements. The inner radius of the steel capillary was 1.055 mm and the employed syringe had a volume of 500 µl. The surface tension value for de-ionized Milli-Q water (70 mN/m) was taken as an internal standard before starting measurements. The error given by the instrument was less than 0.1 mN/m.

Cryo-transmission electron microscopy. A drop of the polymer solution was deposited on a glow discharged carbon-coated grid. After blotting away the excess of solution with filter paper, the grids were rapidly plunged into liquid ethane. The frozen specimen were transferred to a Gatan (model 626) cryo-stage and examined in a Philips CM 120 cryo-electron microscope operating at 120 kV. Micrographs were recorded under low-dose conditions at about -170 °C.

Fluorescence spectroscopy. The fluorescence measurements were performed on a Fluorolog 3-22 spectrofluorimeter. The spectra were recorded between 365 nm and 650 nm with the excitation wavelength of 350 nm. The slit width of the emission was kept at 1 nm step size. All the measurements were performed in the de-ionized Milli-Q water at room temperature.

17

Chapter 2

2.3 Results and discussion 2.3.1 Chemical modifications of polyketones

The current study focused on a new class of the polyketones (Mw 1500-5500), i.e. alternating co- and ter-polymers of carbon monoxide, ethylene, and propylene (Figure 2.1), as starting materials for modifications. Depending on the molar ratio of ethylene and propylene, the consistency of this type of polyketones at room temperature varies from viscous flowing materials for ethylene free materials to waxy or melting solids at 50% ethylene based on the total olefin content.

O

O

O

O

OR

R

R

R

CO + CH2CH2 CHCH3CH2Pd based Catalyst

+

R= H or CH3

Figure 2.1 Synthesis of CO-ethylene-propylene based low molecular weight polyketones.

Since the use of polyketones resulted in low resolution spectra due to the polymeric nature of the substrate, we determined the exact chemical structure of the product by using model compounds. In particular, we carried out the reaction between 1,2-DAP and 2,5-hexanedione in an NMR tube in CDCl3. The NMR spectra of the starting reaction mixtures and after 11.25 h reaction time are reported in Figure 2.2a. The appearance of the pyrrole ring absorptions at about 5.7 ppm, the -CH2- groups attached to the pyrrole ring (3.6 ppm), and the >CH- in β-position with respect to the pyrrole rings (3.2 ppm) all testify the formation of the reported product. Moreover, from the ratio of peaks at 1 ppm (methyl group of the reacted 1,2-DAP) and at 0.95 ppm (methyl groups of unreacted 1,2-DAP), it is possible to calculate the conversion at a given time (Figure 2.2b). It is clear that the reaction rate at 50 °C is actually slow, reaching a di-amine conversion of 80% in 11.25 h. On the other hand, by comparison of the 1H-NMR spectra before and after reaction, it can also be seen that only the less hindered amino group can react with the carbonyl moieties and the other one remains as pendant from the pyrrole ring, which is in agreement with what is reported in previous studies.8

18

Polymeric amines

123456

NCH3 CH3

CH2

CH

NH2

CH3

aa

bb

c

de

ppm

ac

d b

e

(a)

0 100 200 300 400 500 600 700

0

10

20

30

40

50

60

70

80

90

X amin

e(%

)

Time (min)

(b)

Figure 2.2 Model compound reactions: (a) 1H-NMR spectra before reaction (upper spectrum) and after 11.25 h (lower spectrum); (b) reaction kinetics, determined by 1H-NMR.

The chemical reactivity of polyketones with different kinds of di-amines has been systematically investigated here (Figure 2.3) by using potentiometric titrations. The mechanism of the Paal-Knorr reaction between a 1,4-di-carbonyl compound and a primary amine involves three steps: addition to the first carbonyl, addition/elimination and rearrangement steps, and stable product formation.9 The chemical characterization of the final stable product has been well established.10-15 The presence of pyrrole rings in the backbone of the polyamines has been confirmed by FTIR and NMR analysis on all our reaction products. Typical is the appearance of 1H-NMR peaks (Figure 2.4) around 6 ppm (assigned to the protons attached to the pyrrole rings), 13C-NMR peaks at around 130 ppm

19

Chapter 2

(assigned to the C-atoms of the pyrrole rings), and FTIR absorption peaks (Figure 2.5) at around 3100 cm-1 and 1500-1680 cm-1 (assigned to the pyrrole rings).

O O

O

O

+

N

N O

=

NH2

NH2

NH2

NH2 NNH2

NNH2 N

NH21,2-DAP 1,3-DAPe2-DAEA

2-DEAEA 4-PcA

R

R

O

R NH2

R NH2

Figure 2.3 Scheme of the reaction between polyketones and a series of di-amines.

1234567

Polyamines

ppm

Polyketones

Figure 2.4 1H-NMR spectra of polyketones (PK30) and polyamines (PK30 with 1,2-DAP).

20

Polymeric amines

5001000150020002500300035004000

Abs

orpt

ion

Wavenummber (cm-1)

Polyketones

Polyamines

Figure 2.5 FTIR spectra of polyketones (PK30) and polyamines (PK30 with 1,2-DAP).

A more detailed study for the polymeric systems was set up by carrying out the reactions under different experimental conditions. We first investigated the kinetics of the reaction under mild conditions (50 °C) by determining and calculating the di-amine conversion as function of time (Figure 2.6). An initial molar ratio (Iam/CO) of 1 between 1,2-DAP and the 1,4-arrangement of di-carbonyl groups on the PK30 was employed. Due to the steric effect of the 1,2-DAP, only the non-sterically hindered amino group (position 1 in 1,2-DAP) could react with the 1,4-di-carbonyl unit and led to the desired regiospecific reaction. The other one (position 2 in 1,2-DAP) remained intact as functional group. This is consistent with the results of the model component reactions between 2,5-hexanedione and di-amines with steric hindrance8 and to the data reported above. It can be observed that the conversion increases with the reaction time and reaches a steady value of about 65 % after about 4 h reaction time. The fact that not all amino groups are actually consumed in the reaction (i.e. conversion never reaches 100% for an initial equimolar ratio), is easy to understand if one considers that every amino groups must react with two adjacent carbonyl groups to yield the final pyrrole. Statistically, however, there exists always the chance that two different amino groups will react and result, after ring closure, in two pyrroles separated by a single carbonyl group. In this case, further reaction of the isolated carbonyl-group with another amino group is not possible, thus preventing the amine conversion to reach 100% for an initial equimolar ratio.

21

Chapter 2

50 100 150 200 250 300 350 40025

30

35

40

45

50

55

60

65

X am

ine

(%)

Reaction time (min) Figure 2.6 Di-amine conversion as a function of time in the reaction of PK30 with 1,2-DAP.

The situation changes when using a different Iam/CO (Table 2.1) at 80°C after 240 min of reaction time. The di-amine conversion steadily decreases with the amount of the same reactant present originally in the reaction mixture. Moreover, di-amine conversions very close to 100% are observed when using an Iam/CO of about 0.6 or 0.7. The latter clearly confirms that the limitation in the conversion is not due to the amine reactivity but to the statistical restraint mentioned above. Moreover, from a practical point of view, the same fact indicates that a maximum of about 70% of the carbonyl groups can actually react with the amino groups due to the availability of the 1,4-di-carbonyl groups. The reaction temperature has very little impact on the conversion between 80°C and 100°C (Table 2.2), at least if compared to the influence of the composition. Between 80°C and 100°C the di-amine and carbonyl conversion remains basically unaffected at Iam/CO=0.6 while it increases slightly at Iam/CO=0.8. These data indicate once more that there is a maximum to the amine amount able to react with the PK30. At the same time the conversion might be improved by increasing the reaction temperature when working in excess of di-amines with respect to the 70 % of carbonyl groups as mentioned earlier.

Iam/CO

Xamine

(%) XCO (%)

0.6 99 59 0.7 92 64 0.8 83 67

Table 2.1 Conversion of di-amines and carbonyl groups in the reaction of PK30 with 1,2-DAP as a function of Iam/CO at 80 °C.

22

Polymeric amines

Iam/CO T (°C)

Xamine (%)

XCO (%)

0.6 80 99 59 0.6 100 99 59 0.8 80 83 67 0.8 100 90 72

Table 2.2 Conversion of 1,2-DAP and carbonyl groups as functions of temperature and Iam/CO for the reaction with PK30.

We finally investigated the effect of the chemical structure of the di-amines and the polyketones on the conversion (Table 2.3). All the reactions proceeded exceedingly well except for the unexpected cross-linking of PK50 with 1,2-DAP and 1,3-DAPe. No cross-linking can occur for the PK50 with 2-DAEA, 2-DEAEA, and 4-PcA due to the fact that only primary amino groups are reactive in the Paal-Knorr reaction.

PK0 PK30 PK50

Di-amines Xamine(%) XCO (%) Xamine(%) XCO (%) Xamine(%) XCO (%)

1,2-DAP 80 64 90 72 Cross-linking

1,3-DAPe 81 65 82 65 Cross-linking

2-DAEA 76 61 82 66 94 76

2-DEAEA 71 57 76 61 98 78

4-PcA 74 60 77 61 99 79

Table 2.3 Conversion of di-amines and carbonyl groups as functions of the kind of polyketones and di-amine (Iam/CO=0.8, T=100 °C, t=240 min).

Due to the steric hindrance, only one amino group of 1,3-DAPe could react with PK0 and PK30, a very similar effect to the one of the 1,2-DAP. The conversion of different di-amines are found to be quite comparable to each other. All the resulting polyamines can be readily dissolved into common organic solvents, such as methanol, ethanol, THF, and chloroform. In contrast, virgin polyketone oligomers (before modifications) are only slightly soluble in protic solvents such as methanol and ethanol. Conversion increases with increasing ethylene content of the polyketones, in particular this leads for 1,2-DAP and 1,3-DAPe to cross-linking when using PK50. It is interesting to note that the di-amine conversion with PK50 can even reach almost 100%, which corresponds to around 80% carbonyl group consumption. This trend could be due either to a different solubility of the different di-amines in the molten polyketones or to the higher reactivity at high ethylene content because of the lower steric hindrance of the carbonyl groups of PK50. However, a

23

Chapter 2

closer look at the value of the solubility parameters16 for every possible couple of polyketones and di-amines shows that there is no real correlation between solubility (estimated as absolute value of the solubility parameter difference) and di-amine conversion (Figure 2.7). As a consequence, like already stated in the literature10-11, we can conclude that the steric effect in the polyketone backbone plays here a major role.

0.0 0.4 0.8 1.2 1.6 2.0 2.4

72

76

80

84

88

92

96

100

X amin

e(%)

|δpolymer - δdiamine| ((cal/cm3)0.5)

PK0PK30PK50

Figure 2.7 Effect of the difference of solubility parameters between the polyketones and

di-amines on the conversion of the di-amines.

2.3.2 Characterization of the polymeric amines

The synthesized polyamines were protonated with the use of acetic acid (Figure 2.8) and successively dispersed in water. They exhibit many interesting physical properties in aqueous solution.

N

N ONH2

NH2

OH

ON

N ONH2

NH3+

Figure 2.8 Protonation of polyamines with acetic acid (only derivatives of 1,2-DAP is shown for brevity).

24

Polymeric amines

We first investigated the influence of the protonation level on the particle dimension of the resulting dispersion of PK30 modified with 1,2-DAP (Figure 2.9a). Below 35% protonation level, the polymer is insoluble in water and precipitates. In first instance, it must be noticed that the particle size decreases quite significantly with the protonation level. This is most probably related to the stabilization effect which prevents coagulation, due to the positive charges at the interface between the particles and water according to the well-established DLVO theory17. A confirmation of this hypothesis is given by the observed salt effect. The addition of NaCl (150 mM) to the dispersion results in a screening of the electrostatic repulsion between particles and therefore leads to an increased average particle size. The salt effect is rather pronounced at low protonation level. Within the protonation level range from 35% to 45%, the addition of the salt leads to the flocculation (salted out) of the polyamines in the solution.

The protonation level affects the particle size but has no remarkable influence on the width of the distribution (Figure 2.9b), which remains basically monomodal at all protonation levels. Moreover, the repulsive electrostatic interaction is strong enough to stabilize the corresponding dispersion even after 1 month storage (Figure 2.9a) at room temperature. The same lack of effect is observed when comparing the scattering intensity, which corresponds to the number of the particles in the solution, as functions of protonation level and time (Figure 2.9c). The increase of the protonation level makes the polymers more hydrophilic and keen to molecularly dissolve into water, thus leading to a decrease of the scattering intensity. A very weak scattering intensity can be observed with further increasing the protonation level above 60%, indicating that the polymers are able to behave like polyelectrolytes and become water soluble.

35 40 45 50 55 6010

20

30

40

50

60

70

80

90

Part

icle

siz

e (n

m)

Protonation level (%)

1 day 1 month Salt effect

(a)

25

Chapter 2

1 10 10002468

10121416182022

Inte

nsity

(%)

Particle size (nm)

35% 37.5% 40% 45% 50% 55% 60%

(b)

35 40 45 50 55 600

100

200

300

400

500

600

700

Scat

terin

g in

tens

ity (k

cps)

Protonation level (%)

1 day 1 month

(c)

Figure 2.9 Influence of protonation level on (a) particle size; (b) particle size distribution; (c) scattering intensity for PK30 modified with 1,2-DAP in water (5 mg/ml).

A deeper insight into the particle size and, in general, the morphology of the dispersions was obtained by using Cryo-TEM (Figure 2.10). It is clear how the average particle size decreases with the protonation level, in striking agreement, also from a quantitative point of view, with the light scattering measurements. In this respect, such dispersions could rightly be called “nano”, starting with average dimensions of about 50 nm for the spherical polymer micelles at 40% protonation level and ending up with an average of even less than 10 nm at 100% protonation level. Such outstanding uniformity of the dispersions as well as their stability with respect to time has clear advantages for many practical applications.

26

Polymeric amines

Figure 2.10 Cryo-TEM images of PK30 modified with 1,2-DAP dispersed in water at (a) 40% protonation level; (b) 60% protonation level; (c) 100% protonation level. Bar represents 50 nm.

27

Chapter 2

One factor that may remarkably affect both particle size and the light scattering intensity of the dispersions is the sudden pH drop caused by the addition of the strong acid (0.5 M HCl) into polyamine dispersions (Figure 2.11), which may provide useful information for further application as drug carriers18. For a derivative of PK30 with 1,2-DAP at 40% protonation level, it is found that once the micelles are formed, the micelle size does not strongly depend on the pH drop even after 1 day and 3 days time-period. However the considerable decrease of the scattering intensity indicates very fast kinetics for the conversion of micelles into unimers, corresponding to deeply protonation of amino groups, once the pH drop is made.

5.8 6.0 6.2 6.4 6.6 6.8 7.025

30

35

40

45

50

55

60

20

30

40

50

60

70

80

90

Part

icle

siz

e (n

m)

pH

1 day 3 days

Scat

terin

g in

tens

ity (k

cps)

Figure 2.11 Influence of pH on the particle size and scattering intensity of PK30 modified with 1,2-DAP (2 mg/ml).

The fact that the modified PK30 has a surfactant-like structure or amphiphilic character is not only testified by its chemical structure but also by the measurement of surface tension of the resulting dispersions as functions of the concentration and protonation level of the polyamines (Figure 2.12). Surface tension, indicating the adsorption of the surfactants at the air-water interface, decreases as expected with the concentration of the polyamines and with the protonation level. The latter effect is probably related to the more hydrophobic nature of the modified PK30 with decreasing protonation level, which thus determines its surface activity. This phenomenon is similar to other synthetic block copolymers with a hydrophilic-hydrophobic balance.19 Surface tension measurements result in a relatively smooth curve due to molecular weight distributions of the polymers, which is a clear difference with that of low molecular weight surfactants. As a result, no well-defined critical micelle concentration (CMC) could be identified.20

28

Polymeric amines

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

45

50

55

60

65

70

Su

rfac

e te

nsio

n (m

N/m

)

C (mg/ml)

40% 60% 80% 100%

Figure 2.12 Surface tension of PK30 modified with 1,2-DAP as functions of protonation level and concentration.

Aqueous properties were also studied on the PK30 with 1,3-DAPe (Table 2.4 and Figure 2.13) at two different protonation levels. A very similar aqueous behavior to the one displayed by PK30 modified with 1,2-DAP can be observed. Partial protonation of the amine groups on the PK30 with 1,3-DAPe induces larger, loose aggregates with an average particle size around 300 nm, compared to an average particle size of less than 50 nm for PK30 with 1,2-DAP. This may be correlated with the difference in chemical structure of the used di-amines for modifications, which thus leads to the diversity of the aqueous properties. The polymer solutions display good stability during the 2 weeks of measurement time. The salt effect (150 mM NaCl) here is not as much pronounced as for PK30 with 1,2-DAP. In terms of surface activity, it can be seen in Figure 2.13 that the surface tension decreases with an increase in the concentration and, as expected, low protonation levels result in a higher surface activity.

Protonation level (%)

Particle size (nm)

Scattering intensity (kcps)

70 (1 day) 302 425 70 (2 weeks) 295 350

70 (salt effect) 264 410 80 (1 day) 296 163

80 (2 weeks) 265 190 80 (salt effect) 263 176

Table 2.4 Influence of the protonation level on the particle size and scattering intensity of

PK30 modified with 1,3-DAPe.

29

Chapter 2

0.0 0.2 0.4 0.6 0.8 1.0 1.2

48

50

52

54

56

58

60

62

64

66

68

Surf

ace

tens

ion

(mN

/m)

C (mg/ml)

80% 100%

Figure 2.13 Surface tension of PK30 modified with 1,3-DAPe as functions of protonation level and concentration.

The solution behavior of the modified polymers after methylation was also studied by measuring the surface tension (Figure 2.14). The surface tension of the modified PK50 with 2-DAEA, 2-DEAEA, and 4-PcA after methylation decreases in a less dramatic manner to a limited value around 65-67 mN/m with an increase in polymer concentration. This may be explained by the fact that the modified PK50 after the methylation has a high content of the ionized amino groups around the backbone, leading to the enhanced hydrophilicity. Thus the polymers become water-soluble cationic polyelectrolytes. This is also confirmed by the fact that there are no small aggregates formed even at the high polymer concentration (10 mg/ml) by the DLS studies. It is worth to note that the observed difference in surface activity for the modified PK30 and PK50 with 4-PcA might be due to the increasing of the di-amine conversion from the 77% to 99% and to the different molecular weight of the polyketones. Indeed, we already showed and discussed above (Figure 2.12) how the protonation level (i.e. the hydrophilicity of the polymers) negatively affects the surface activity. In the present case one might notice that higher conversion for 4-PcA (as in the case of the reaction with PK50 compared with PK30) results in a stronger hydrophilic character of the polyamines (simply more hydrophilic groups are grafted along the backbone) and thus a lower surface activity, in agreement with the concept expressed above.

30

Polymeric amines

0.0 0.2 0.4 0.6 0.8 1.040

45

50

55

60

65

70

Surf

ace

tens

ion

(mN

/m)

C (mg/ml)

2-DAEA(PK50) 2-DEAEA(PK50) 4-PcA(PK50) 4-PcA(PK30)

Figure 2.14 Surface activity measurements for water dispersion of PK50 modified with 2-DAEA, 2-DEAEA, 4-PcA and PK30 modified 4-PcA.

All synthesized polymers behaved like intrinsically photo-luminescent materials and displayed fluorescent properties, which can be attributed to the multiple aromatic pyrrole rings in the backbone. For example, the fluorescent spectra of PK30 modified with 1,2-DAP at 100% protonation level and with 4-PcA after methylation are shown in Figure 2.15. In aqueous solution, the fluorescence spectra show maxima at around 420-430 nm for the PK30 with 1,2-DAP and around 470 nm for the PK30 with 4-PcA. A small red shift of the fluorescence emission for PK30 with 1,2-DAP takes place with an increase of concentration. The different maximum emission wavelength between these two polyamines can be attributed to their different chemical structure. The fluorescence intensity increases with increasing the initial polymer concentrations. After reaching the maximum, the fluorescence intensity decreases with further increasing the concentration. This behavior is most probably due to a self-quenching effect.21

400 450 500 550 600 6500.0

0.2

0.4

0.6

0.8

1.0

Fluo

resc

ence

(nor

mal

ized

)

Wavelength (nm)

0.1 mg/ml

0.05 mg/ml

0.025 mg/ml

0.2 mg/ml

0.0125 mg/ml

0.3 mg/ml

(a)

31

Chapter 2

400 450 500 550 600 6500.0

0.2

0.4

0.6

0.8

1.0

Fluo

resc

ence

(nor

mal

ized

)

Wavelength (nm)

0.5 mg/ml

0.75mg/ml

0.3 mg/ml

0.2 mg/ml

0.1 mg/ml

(b)

Figure 2.15 Normalized fluorescence intensity of (a) PK30 modified with 1,2-DAP; (b) PK30 modified with 4-PcA in water at a function of concentration.

The aqueous properties described above are very interesting when considering the possible applications of the synthesized polyamines. Their surface activity strongly suggest the application as polymeric surfactants (e.g. the dispersion of virgin polyketones in water, see Chapter 3). Moreover, the simple presence of ammonium groups, whose amount is easily tunable by controlling the protonation reaction, suggests the use of these cationic polymers in the formation of complexes with polyanions (e.g. DNA, polyacrylic acid).

2.4 Conclusions

A series of new polymeric amines were prepared by chemical modifications of low molecular weight alternating polyketones with different kinds of di-amines. It was found that the reaction conversion strongly depends on the steric hindrance displayed by the reactants. During the reaction, around 70-80% of the polyketones carbonyl groups could be converted into pyrrole units along the backbone. It is demonstrated that the resulting polymers, after protonation or alkylation of the amino groups, display interesting aqueous solution behavior and they could act as polymeric surfactants or polyelectrolytes. Furthermore, they exhibit fluorescence in aqueous solution, which may be promising for the use as water-soluble fluorescence probes.

32

Polymeric amines

2.5 References

1. Butun, V.; Liu, S.; Weaver, J. V. M.; Bories-Azeau, X.; Cai, Y.; Armes, S. P. React. Funct. Polym. 2006, 66, 157.

2. Gil, E. S.; Hudson, S. M. Prog. Polym. Sci. 2004, 29, 1173. 3. Aoyama, Y.; Novak, B. M. Macromolecules 2001, 34, 6842. 4. De-Smedt, S. C.; Demeester, J.; Hennink, W. E. Pharmacol. Res. 2000, 17, 113. 5. March, J. Advanced organic chemistry, 4th ed., Wiley, 1992. 6. Drent, E.; Keijsper, J. J. US 5225523, 1993. 7. Mul, W. P; Dirkzwager, H.; Broekhuis, A. A.; Heeres, H. J.; Van Der Linden, A. J.;

Orpen, A. G. Inorg. Chim. Acta 2002, 327, 147. 8. Kostyanovsky, R. G.; Kadorkina, G. K.; Mkhitaryan, A. G.; Chervin, I. I.; Allev, A. E.

Mendeleev. Commun. 1993, 21. 9. Amarnath, V.; Anthony, D. G.; Amarnath, K.; Valentine, W. M.; Wetterau, L. A.;

Graham D. G. J. Org. Chem. 1991, 56, 6924. 10. Chen, J.; Yeh, Y.; Sen, A. J. Chem. Soc., Chem. Commun. 1989, 965. 11. Sen, A.; Jiang, Z.; Chen, J. Macromolecules 1989, 22, 2012. 12. Kiovsky, T. E.; Kromer R. C. US 3979374, 1976. 13. Brown, S. L. US 5081207, 1992. 14. Sinai-Zingde, G. D. US 5605988, 1997. 15. Broekhuis, A. A.; Freriks, J. US 5952459, 1999. 16. Coleman, M. M.; Graf, J. F.; Painter, P. C. Specific interactions and the miscibility of

polymer blends, Technomic Publishing, 1991. 17. Stokes, R. J.; Evans, D. F. Fundamentals of interfacial engineering, Wiley, 1997. 18. Gillies, E. R.; Frechet, J. M. J. Chem. Commun. 2003, 1640. 19. De Paz-Banez, M. V.; Robinson, K. L.; Vamvakaki, M.; Lascelles, S. F.; Armes, S. P.

Polymer 2000, 41, 8501. 20. Butun, V.; Armes, S. P.; Billingham, N. C.; Tuzar, Z.; Rankin, A.; Eastoe, J.; Heenan,

R. K. Macromolecules 2001, 34, 1503. 21. Lakowicz, J. R. Principles of fluorescence spectroscopy, 2nd ed, Kluwer

academic/Plenum Publishers, 1999.

33

Chapter 2

34

Wood adhesive emulsions

Chapter 3

Wood adhesive emulsions from thermosetting

alternating polyketones

Abstract

Aqueous polymer emulsions were prepared by chemical modifications of thermosetting alternating polyketones in a single one-pot reaction. Polymeric amines derived from the polyketones can act as polymeric surfactants for the self-emulsification of thermosetting polyketones. The stability and structure of the resulting emulsions with respect to the storage time at room temperature (20 °C) at different experimental conditions were thoroughly studied by dynamic light scattering, rheology, and Cryo-SEM. Emulsions with an average particle size smaller than 1 μm and a viscosity less than 1 Pa s could be achieved and remained stable for at least 2 years. The prepared emulsions were qualified as wood adhesives for the wood composite industry, according to the European Standard for wood adhesive testing.

Keywords: Polyketones; Wood adhesives; Polymeric surfactants; Water emulsions

Based on: Zhang, Y.; Broekhuis, A. A.; Picchioni, F. Aqueous polymer emulsions by chemical modifications of thermosetting alternating polyketones, J. Appl. Polym. Sci. 2007, 106, 3237-3247.

35

Chapter 3

3.1 Introduction

Thermosetting polyketones, produced by alternating co- and ter- polymerization of carbon monoxide, ethylene, and propylene, constitute a relatively new class of viscous low molecular weight resins.1-2 Due to the presence of highly active carbonyl groups, thermosetting polyketones can be cured in a variety of ways, e.g. conversion to polyfurans, reduction to poly-alcohol, co-condensation with commercial formaldehyde resins, etc.3 The Paal-Knorr reaction, involving pyrrole ring formation between an amino group and two adjacent carbonyl groups, has been used as basic cross-linking chemistry for these materials.4-6 In this curing reaction, water-resistant pyrrole units are formed with the elimination of water, using a variety of di- or multi-functional amino-based curing agents. Based on this curing chemistry, thermosetting polyketones can be used in various areas like coatings, foaming, adhesives, and polymer composites. However, thermosetting polyketones display a relatively high viscosity and therefore can be difficult to directly apply on a substrate. In this respect and also because of the environmental restrictions on organic solvent-based systems, the development of waterborne dispersions or emulsions is the preferred route for the application of thermosetting polyketones.6-9

Polymeric emulsions in water are generally produced by using surfactants or direct emulsion polymerization.10 In recent years, block and graft structured polymeric surfactants were used to replace low molecular weight ones in a variety of applications as emulsifiers, dispersion stabilizers, wetting agents, and compatibilizers.11-12 Copolymers consisting of hydrophilic and hydrophobic blocks, may exhibit interesting properties, such as self-assembly, micelle-formation, surface absorption, and molecular association.13 In the previous chapter, the synthesis of a new family of polymeric amines (polyamines) by chemical modifications of polyketones via the Paal-Knorr reaction has been described. Interesting properties in aqueous solution (e.g. micelle-formation, long-term stability of the formed micelles, high surface activity, and fluorescence) were observed for these polyamines.14 Thus, they may hold great promise for many applications, such as emulsifier, drug carriers, cation exchange resins, pharmaceuticals, and photoluminescent materials. In the present work we report a simple route to prepare water-borne polyketone emulsions, using the protonated polyamines as polymeric surfactants, previously derived by the modifications of polyketones. This approach not only allows to combine all different processing steps in a single one-pot reaction, but also represents a very economical way to disperse polyketones in water.

The use of water-based polyketone dispersions as adhesives in the wood composite industry such as plywood, particleboard, fiberboard, and wood panels is currently seen as the most promising application of polyketones. Dominant wood adhesive systems used in

36

Wood adhesive emulsions

the market are nowadays based on urea-formaldehyde or phenol-formaldehyde. However, both of these adhesives are based on formaldehyde, a known hazardous chemical and suspected carcinogen to the human health.15-17 Thus there is an increasing need to develop formaldehyde-free and environmental-friendly wood adhesives. Polyketone-based emulsions do not release any volatile toxic components during the processing steps and the byproduct of all emulsification steps (including the polyketone modifications) is only water.

In this work we systematically investigated the different factors that have an effect on the stability and structure of the emulsions (i.e. emulsification conditions, protonation level and amount of polyamines, the molecular weight and ethylene content of the polyketones to be emulsified) by determining the particle size and viscosity of the emulsions up to the storage time of 2 years at room temperature (20 °C). The morphology of the emulsions was studied by Cryo-SEM. According to the European Standard (EN-314) for wood adhesive testing, the quality of the emulsions applied as wood adhesives was evaluated by measuring the shear strength on hard wood (maple) substrates.

3.2 Experimental

Materials. The alternating polyketones with 0% ethylene (PK0, Mw-1680), 30% ethylene (PK30, Mw-3970), and 50% ethylene (PK50, Mw-5350) based on the total olefin content, were synthesized, according to a previously reported procedure.1-2 1,2-diaminopropane (Acros, 99%) and acetic acid (Acros, ≥99.7%) were purchased and used as received.

Emulsion preparation. The emulsions were prepared by using the following procedure (Figure 3.1). Initially, chemical modifications of the PK30 (we denote with mPK30, the intial weight of the PK30 undergoing the modification reaction) with 1,2-diaminopropane were carried out to prepare polyamines in a sealed 250 ml round bottom glass reactor with a reflux condenser, U-type anchor impeller in an oil bath. Details of the procedure and successive protonation were described in Chapter 2. After modifications, 70% of the carbonyl groups have been converted to pyrrole rings (determined by potentiometric titration), thus meaning that only 30% of the carbonyl groups remain unreacted. During the emulsification step, a second quantity of unmodified polyketones (PK0, PK30, and PK50) and de-ionized milli-Q water were added into the protonated polyamine solutions for a final fixed 50 wt% solids resin composition (the amount of water was corrected for the amount of water developed during the polyamine synthesis). All these different steps were carried out in a single one-pot reactor. The obtained emulsions were stored in sealed high

37

Chapter 3

density polyethylene (HDPE) containers at room temperature for a period of up to 2 years and analyzed at regular intervals.

O O

O

O NH2

NH2

O

Polyketones (mPK30)

Chemical

Polyamines

NN

O

NH3NH2

Protonation

Water emulsions

Acetic acid

Protonated polyamines

Emulsification

Second Polyketones(PK0,PK30,PK50)

+

NN

O

NH2NH2modifications

Figure 3.1 Scheme of the preparation of aqueous polymeric emulsions based on thermosetting polyketones.

Dynamic light scattering. The dynamic light scattering (DLS) measurements were performed on a Zetasizer 5000 instrument (Malvern Instruments, UK) at a wavelength of 633 nm and at a temperature of 25 °C. Scattered light was detected at a 90 degree angle. The viscosity of (0.89 mPa s) and the refractive index (1.33) of water at 25°C were used for data analysis. The intensity autocorrelation functions obtained from the DLS were analyzed using CONTIN algorithm in all the measurements. The apparent hydrodynamic diameter, DH, was obtained using the Stokes-Einstein relation:

D0 = kT/ 3π ηDH (1)

where D0 is the diffusion coefficient, k is the Boltzmann constant, T is the absolute temperature, and η is the viscosity of the solution. All samples were diluted with de-ionized milli-Q water prior to the measurements.

Rheological analysis. The viscosity of the emulsions was measured with an AR 1000 rheometer (TA Instruments, USA), using an aluminum cone-and-plate fixture of 2° and 40 mm in diameter. The apparent viscosity of the samples was measured at a constant shear rate of 15 s-1. The viscosity-shear rate relationship was measured as the shear rate was increased from 5 s-1 to 60 s-1. All experiments were carried out at 20 °C.

Cryo scanning electron microscopy (Cryo-SEM). The Cryo fixation technique was used to prepare the samples for SEM observation. A cold field emission scanning electron microscope (JEOL 6301 F) equipped with an Oxford CT 1500 HF Cryotransfer system

38

Wood adhesive emulsions

was used. A drop of diluted (four times) emulsions was placed on a piece of freshly cleaved mica surface and was Cryo fixated into a liquid nitrogen slush at -210 °C after blotting of the drop to a thin layer. After transport of the specimen-holder with the sample into the Cryotransfer system, surface water was removed by sublimation for 10 min and a 3 nm thick coating of Au/Pd was sputtered at –120 °C. SEM micrographs of the samples were recorded at -120 °C.

Wood adhesive testing. The hard wood maple veneers for adhesive testing were vacuum dried at 105 °C for 10 h to reduce moisture content. Salicylic acid (0.5 wt% based on the second quantity of polyketones, i.e. PK0, PK30 or PK50) was added into the emulsions as curing catalyst. According to the European Standard (EN-314) for wood adhesive testing, the emulsions were applied as 150 g/m2 single adhesive line onto one side of the hard wood maple veneer (a thickness of 4 mm, dimension of 50 mm×25 mm) and adhesive line for each piece veneer sample was 25 mm×25 mm (Figure 3.2). The lapped two pieces of glue-applied veneer was hot-pressed for 5 min at 200 °C under a 3 MPa constant pressure. After hot-pressing, the samples were subjected to immersion in boiling water for a period of 72 h. The shear strength of the samples (previously cooled down but on purpose not dried) was tested on an Instron 4301 machine using 5 kN power sensor with a crossing head speed 2 mm/min. The shear strength is defined as the load required to break the specimen, divided by the area of the adhesive bonds. At least 10–13 replicates were performed for each experiment series with standard deviation less than 0.3 MPa.

Figure 3.2 Lapped specimen for shear strength test.

3.3 Results and discussion

3.3.1 Emulsification

The influence of different factors on the emulsification process and the resulting average particle size (Table 3.1) was first investigated. As can be seen in Figure 3.3a, the particle size decreases initially with the mixing time and then it levels off after roughly 40 min. To obtain an emulsion, sufficient energy input is required to overcome the interfacial tension between the two fluids and mechanically break up large droplets into smaller ones. The

39

Chapter 3

observed trend clearly indicates that the equilibrium particle size is reached after 40 min mixing time. A descending trend is observed when plotting the particle size as a function of the rotor speed (Figure 3.3b). In this case we observe a clear decrease of particle size as function of the rotor speed, which we can roughly estimate as proportional to the mechanical energy input. This is actually not surprising and in agreement with the general principles of emulsion formation.18 It is remarkable that the emulsions can be prepared even with a very low energy input (10 min, 300 rpm). In general, an important aspect in the preparation of polymeric emulsions is to obtain an average droplet size lower than 1 µm

and a narrow size distribution in order to ensure high kinetic stability. Here, average particle size lower than 1 µm can be achieved by simply increasing the energy input either at higher rotor speed or at longer emulsification time. By comparing Figure 3.3a and Figure 3.3b, it can be seen that energy density, which is the mechanical energy input per unit volume of the zone where the droplets are disrupted, plays a more important role in the emulsification process than the total energy input. The influence of the emulsification temperature shows an interesting behavior (Figure 3.3c). The particle size remains almost constant below 50 °C and starts to increase up to several micrometers when the temperature rises to 70 °C. The latter effect is probably due to the increased thermal energy that favors eventually re-coalescence of the droplets.

Factors Unit Range

Rotor speed Rpm 300-1300

Emulsification time Min 10-90

Emulsification temperature °C 25-70

Protonation level of the polyamines

% 40-80

Polyamine content (PK0/mPK30, PK30/mPK30 or PK50/mPK50)

weight ratio 1-2

Polyketone grade or ethylene content

% 0-50

Table 3.1 Investigated factors for the emulsification step.

40

Wood adhesive emulsions

0 20 40 60 80

700

800

900

1000

1100

1200

1300

1400

1500

100

Part

icle

siz

e (n

m)

Time (min)

(a)

200 400 600 800 1000 1200 1400

400

500

600

700

800

900

1000

Part

icle

siz

e (n

m)

Speed (rpm)

(b)

20 30 40 50 60 70500

1000

1500

2000

2500

3000

3500

Part

icle

siz

e (n

m)

Temperature (°C)

(c)

Figure 3.3 Effect of (a) mixing time (500 rpm, 50 °C); (b) rotor speed (1 h, 50 °C); (c) emulsification temperature (500 rpm, 1 h) on the particle size of the emulsions (60% protonation level, PK30/mPK30=1.5).

41

Chapter 3

The particle size is also related to the kind of polyketones employed for the emulsification (Figure 3.4). According to the general theory of emulsion formation, the average diameter of particles dispersed in the matrix is proportional to the viscosity of the dispersed phase.18 Low ethylene content is usually related to a lower molecular weight of the polyketones and therefore a lower viscosity of the dispersed phase. On the basis of this, it is reasonable to assume that the observed effect (particle size increasing with ethylene content) might be simply related to the difference in viscosity. The particle size of the emulsions is also influenced by the protonation level and amount of the employed surfactants (polyamines) (Figure 3.5). It is found that a low protonation level results in a much smaller particle size. This effect is correlated with the nature and the aqueous properties of the polyamines. In general, surfactants play two main roles during the emulsification: (i) they lower the interfacial tension between the two different phases to facilitate droplet break-up; (ii) they prevent re-coalescence. It is already known that polyamines at low protonation level display higher surface activity.14 In addition, as the protonation level of the polyamine increases (higher than 60%) the polyamines start to dissolve into water molecularly and act more like polyelectrolytes instead of surfactants. The particle size increases also with the PK30/mPK30 ratio. This is an expected trend, since the amount of surfactants determines the total interfacial area and the decrease of the PK30/mPK30 ratio causes a better stabilization of the interface. The most striking result in the emulsification studies is that the use of the polyamines even at low protonation level (40%, 50%) and high PK30/mPK30 ratio of 2 can still result in emulsions with an average particle size in the order of 500 nm.

1.00 1.25 1.50 1.75200

400

600

800

1000

1200

1400

1600

1800

2000

Part

icle

siz

e (n

m)

Polyamine content

PK0 PK30 PK50

Figure 3.4 Effect of ethylene content of the second quantity polyketones and polyamine content on the particle size of the emulsions (500 rpm, 1 h, 50 °C, 60% protonation level).

42

Wood adhesive emulsions

1.00 1.25 1.50 1.75 2.000

500

1000

1500

2000

2500

3000

3500

Part

icle

siz

e (n

m)

PK30/mPK30 ratio

80% 60% 50% 40%

Figure 3.5 Effect of protonation level and polyamine content on the particle size of the emulsions (500 rpm, 1 h, 50 °C).

The particle size distributions of the emulsions can also be effectively varied by adjusting the protonation level of the polyamines and the PK30/mPK30 ratio (Figure 3.6). A narrow and mono-modal distribution can be achieved at low protonation level and low PK30/mPK30 ratio. The evolution of the particle size distribution is worth noticing here. The narrow distribution is gradually transformed into broad distribution and even to asymmetric bimodal with the increase of the PK30/mPK30 ratio (Figure 3.6a). The same trend of the particle size distribution can also be observed as a function of increasing protonation level (Figure 3.6b). In this case the width of the distribution increases with the protonation level, probably as consequence of the lower surface activity of the polyamines at high protonation levels.

Effect of (a) PK30/mPK30 ratio (60% protonation level); (b) protonation level

100 10000

10

20

30

40

50

Figure 3.6 (PK30/mPK30=1.5) on the particle size distribution of the emulsions (500 rpm, 1 h, 50°C).

Inte

nsity

(%)

Particle size (nm)

1 1.25 1.5 1.75

PK30/mPK30(a)

100 1000 100000

5

10

15

20

25

30

35

Inte

nsity

(%)

Particle size (nm)

50% 60% 80%

(b)

43

Chapter 3

In order to obtain a more quantitative correlation between the particle size and the different factors outlined above, we summarized all our results in a statistical model.19 In this case, we used a multivariable, not-linear regression technique to be able to express the average particle size (y) as function of the mentioned experimental parameters (x’s). The resulting model has the following form:

022

1 )**( fxfxfyi

iiii ++= ∑ (2)

where the summation is extended over all investigated factors namely: protonation level (x1), polyamine content (x2), rotor speed (x3), emulsification temperature (x4), ethylene content (x5), and emulsification time (x6). The fji (j=1,2; i=1-6) are the corresponding coefficients determined by the fitting procedure. The f0 represents the intercept of the model.

We first started to check the validity of the model by the analysis of variance (Table 3.2). The very low P-value indicates that the model is statistically significant (confirmed also by the random distribution of the residuals, not shown here for brevity) and that the observed differences are not due to a random noise of the experimental data. The R2 value of the model is 0.941 with an adjusted R2 value of 0.909, thus indicating that there is a reasonable agreement between the model itself and the experimental data. The latter is also confirmed by the plot of predicted vs. experimental values (Figure 3.7). Once established that the model is statistically significant and describes reasonably the experimental data, we calculated the value of the regression coefficients (Table 3.3). Combination of the f-values with the equation (2) reported above, yields the equation for the developed model immediately. The latter allows easy prediction of the particle size as functions of the employed experimental conditions.

Sum of squares

Degrees of Freedom

Mean Square

Point of F-distribution P-value

Model 1.741×107 12 1.450×106 27.716 3.661×10-10

Error 1.099×106 21 5.233×104

Total 1.850×107 33

Table 3.2 Analysis of variance (ANOVA) for the statistical model.

44

Wood adhesive emulsions

Factor Factor name f1i f2i

x1 Protonation level 1.174 -83.648

x2 Polyamine content -463.177 2.725×103

x3 Rotor speed 4.07×10-4 -1.212

x4 Emulsification

temperature 2.849 -226.44

x5 Ethylene content 0.417 -11.814

x6 Mixing time 0.147 -21.419

Intercept f0=3.975×103

Table 3.3 Values of the regression coefficients.

0 500 1000 1500 2000 2500 3000 3500

0

500

1000

1500

2000

2500

3000

3500

Pred

icte

d pa

rtic

le s

ize

(nm

)

Experimental particle size (nm) Figure 3.7 Plot of predicted vs. experimental values.

3.3.2 Emulsions stability

The stability or life-time of emulsions is considered a primary requirement for a wide variety of industrial applications of emulsions. It could be examined from the phase separation by visual inspection and the change of particle size with respect to storage time. After 2 month storage, it was observed that only the emulsions with the particle size higher than 2 µm (protonation level of 80% with PK30/mPK30≥1.75 and protonation level of 60% with PK30/mPK30 = 2) experienced phase separation by visual examination (left part in Figure 3.8). However no sigh of phase separation can be observed for the samples at low protonation level (40-50%) after a storage time of 2 years (right part in Figure 3.8).

45

Chapter 3

Figure 3.8 Pictures of the emulsions: left sample at 80% protonation after 2 month storage; right sample at 40% protonation after 2 years storage (PK30/mPK30=2, 500 rpm, 50 °C, 1 h).

The effect of storage time on the particle size was studied at different protonation levels by using several polyketone grades. The obtained emulsions are kinetically stable at least for one year and their average particle size slightly increases to a small extent over that period of time (Figure 3.9). The monomodal particle size distribution becomes slightly broader over the same time period. By comparison of the graphs at different protonation levels, we did not find any immediate correlation between the experimental conditions upon which an emulsion is formed and its stability. In general, the instability of emulsions arises from the processes of creaming (or sedimentation), Ostwald ripening, flocculation, and coalescence. They may take place concurrently at different rates leading eventually to complete phase separation if equilibrium is achieved. Creaming or sedimentation rate depends on droplet size, the density difference between the dispersed and continuous phases, continuous phase viscosity, and interdroplet interactions. The phase separation of emulsions with large particles here may result from creaming or sedimentation. Ostwald ripening and coalescence is the process whereby larger droplets grow at the expense of smaller ones depending on the solubility and diffusion rate of the dispersed phase in the continuous phase. Small particle size is against sedimentation (or creaming) because of the Brownian motion. Consequently, the diffusion rate is higher than the sedimentation (or creaming) rate induced by the gravity force. Thus, it can be believed that the slight increase of particle size with time can be attributed to Ostwald ripening which is also the main instability mechanism for emulsions of submicron particle size.20

46

Wood adhesive emulsions

Figure 3.9 Effect of storage time on the particle size of the emulsions at (a) 40 % protonation level; (b) 50% protonation level; (c) 60% protonation level; (d) 80% protonation level (500 rpm, 1 h, 50 °C).

It must be stressed here that the two general mechanisms of emulsion stabilization21 (i.e. electrostatic and steric repulsion) are probably both active in our system and responsible for the observed stability. The polyamines contain both positive charges (due to the protonation) as well as hydrophilic polymer chains which can significantly extend in water. A quantitative estimation of these effects is difficult to achieve due to the extreme complexity of the emulsions in terms of chemical composition and molecular weight distribution of the starting polymers. In terms of the effect of ethylene content of the second polyketones, it can be seen that particle size also grows bigger after one year, but on a very small scale (Figure 3.10). We must notice again substantial stability of the particle size as function of time with no remarkable difference as a function of the ethylene content. It is worth noticing the lack of any significant change in particle size of the emulsions for the samples at low protonation level (i.e. 40%, 50%) even after 2 year storage (Figure 3.11).

0 40 80 120 160 200 240 280 320 360300

400

500

600

700

800

Part

icle

siz

e (n

m)

Storage time (day)

1.5 1.75 2

(b)

PK30/mPK30

0 50 100 150 200 250

420

440

460

480

500

520

540

560

Part

icle

siz

e (n

m)

Storage time (day)

1.75 2

(a)

PK30/mPK30

0 40 80 120 160 200 240 280 320 360 4001200

1400

1600

1800

2000

2200

2400

2600

Part

icle

siz

e (n

m)

Storage time (day)

1 1.25 1.5

(d)

PK30/mPK30

0 40 80 120 160 200 240 280 320 360

300400500600700800900

10001100120013001400

Part

icle

siz

e (n

m)

Storage time (day)

1 1.25 1.5 1.75

(c)

PK30/mPK30

47

Chapter 3

0 40 80 120 160 200 240 280 320 360

500

600

700

800

900

1000

1100

1200

1300

1400

Part

icle

siz

e (n

m)

Storage time (day)

1 1.25 1.5

(a)

PK50/mPK30

0 40 80 120 160 200 240 280 320 360

300

400

500

600

700

800

900

1000

1100

Part

icle

siz

e (n

m)

Storage time (day)

1 1.25 1.5 1.75

(b)

PK0/mPK30

Figure 3.10 Effect of storage time on the particle size of emulsions (a) use of PK50 as second polyketones; (b) use of PK0 as second polyketones (60 % protonation, 500 rpm, 1 h, 50 °C).

0 100 200 300 400 500 600 700200

250

300

350

400

450

500

Part

icle

siz

e

Storage time (day) Figure 3.11 Effect of storage time on the particle size of emulsions for a period of 2 years (1300 rpm, 1 h, 50 °C, PK30/mPK30=1.5).

48

Wood adhesive emulsions

Cryo-SEM studies of the emulsion morphology (Figure 3.12) after 6 months’ storage time indicate uniform spherical shape and narrow particle size distribution of the particles at both low and high magnification. The average particle size of the emulsions at two different protonation levels (50% and 60%) is around 400 nm and 800 nm, respectively. It can also be seen that the size distribution of the particles in Figure 3.12b is also much broader than that in Figure 3.12a. These observations from the Cryo-SEM are strikingly agreement with the results from the DLS measurements.

Figure 3.12 Cryo-SEM images of the emulsions at (a) 50% protonation level at low and high magnification; (b) 60% protonation level at low and high magnification (500 rpm, 1 h, 50 °C, PK30/mPK30=1.5).

49

Chapter 3

3.3.3 Rheology

The rheology of the emulsions was also studied at different protonation levels, polyamine content, and ethylene content of the polyketones. The rheological properties of emulsion-based materials play a major role in determining their suitability for particular applications. In terms of wood adhesives, it may be desirable to control the viscosity in view of the applying methods on the wood surface, such as the use as a paste or spraying. Furthermore, rheological studies can provide useful information on the stability and internal microstructure change of emulsions. We first report in Figure 3.13 the viscosity as a function of the storage time for two different protonation levels, namely 40% and 50%. It can be seen that there is a dramatic decrease in viscosity after 1 week and then no significant change within the storage time up to one year. The long term stability (i.e. after the first week) is in agreement with the particle size measurements discussed above. The sharp decrease of viscosity during the first week is not explainable in terms of average particle size or particle size distribution, since the latter do not change significantly.

0 50 100 150 200 250 300 350

0

10

20

30

4090

100

110

120

Figure 3.13 Effect of storage time on the viscosity of the emulsions at (a) 40% protonation level; (b) 50% protonation level (500 rpm, 1 h, 50 °C).

As a result one might speculate that a rearrangement of the polymers within each particle is taking place with the surfactant migrating at the interface and possibly retreating a part of his chains from the water solution in the first week (Figure 3.14). Due to the solubility of (highly) protonated chains in water only the gray particles in the picture are actually seen by the dynamic light scattering technique used for particle size measurements. This hypothesis has in first instance a chemical basis. Potentiometric titrations show that a fully protonated polyamine in de-ionized water gives a pH of about 4.5, thus indicating that the amonium groups easily deprotonate in water. This would mean that as soon as the surfactant is in de-ionized water, a slight deprotonation of the former at the surface should take place. As a result the chains become more hydrophobic and thus less keen to be solubilized in water. One might speculate that in the beginning all the highly protonated

Visc

osity

(Pa.

s)

Storage time (day)

1.5 1.75 2

(a)

PK30/mPK30

0 50 100 150 200 250 300 350

0

10

20

30

40

70

80

90

Visc

osity

(Pa.

s)

Storage time (day)

1.5 1.75 2

(b)

PK30/mPK30

50

Wood adhesive emulsions

chains will stretch in water while later, due to deprotonation or even proton exchange with the amino groups close to the surface, some of them will retract and come closer to the hydrophobic surface of the polyketone particle. Thus, we could see none or little difference in the particle size as measured by the light scattering, but observe a difference in viscosity.

Time

Figure 3.14 Representation of chain retraction for the emulsions.

The mechanism of arms-retraction is indirectly but substantially confirmed also by another experimental evidence. The fresh samples reveal a non-Newtonian or pronounced shear thinning flow behavior (Figure 3.15). Slow transition from a non-Newtonian to a Newtonian behavior, with viscosity independent of the shear rate, is observed with respect to storage time. Such transition might be explained by the retraction of the polymeric chains (previously stretched into the water solution) on the surface of the polymer particle. Indeed, shear thinning behavior is typical of polymeric systems in the melt and related to chain entanglements. In the beginning the system (left part of Figure 3.14) is composed of particles at high concentration (50 wt% of total polymers in water) which will form entanglements between the chains by overlapping. On the other hand, as the systems rearranges in time (right part of Figure 3.14), the overlap between the chains and the number of entanglements diminishes, resulting in a Newtonian-like behavior. This is very similar to what is reported in the literature for polystyrene-end-capped polyelectrolytes22, where a transition from non-Newtonian to Newtonian behavior has also been observed as a function of concentration. Thus most probably the arm retraction causes dis-entanglement of the polymer chains between different particles, leading to the observed rheological properties. Apart from the exact mechanism of the emulsion formation and stability, it is worth noticing that a low viscosity less than 1 Pa s could be achieved at high solid content (50 wt%) even after 1 year storage time. An increase in the PK30/mPK30 ratio is accompanied by a pronounced reduction in the viscosity at all protonation levels. This may be attributed to an increase in the effective volume fraction with increasing particle size, caused by the increase of the PK30/mPK30 ratio.

51

Chapter 3

0 10 20 30 40 50 60

01234

12

16

20

24

28

32

Visc

osity

(Pa.

s)

Shear rate (s-1)

1 day 8 days 13 days 35 days 85 days 235 days 310 days

Figure 3.15 Effect of shear rate on the viscosity of the emulsions (50% protonation level, 500 rpm, 1 h, 50 °C, PK30/mPK30=1.5).

At 60% and 80% protonation level, viscosity decreases initially after one week storage and then increases up to the point when emulsions become gel-like materials (Figure 3.16). These effects may be related to the associative behavior of the used polymeric surfactants and polyketone complex aggregates. However, up to now, the microstructure and the association mechanism of these polymers in aqueous solutions are still unclear. Water-soluble amphiphilic polymers can also possess bulk-thickening properties when compared with conventional surfactants. The polyamines may act as thickening agents to increase the viscosity with time. Another possible explanation is that the amino groups of the polyamines may slowly react with the second quantity of the polyketones at room temperature and therefore lead to low cross-linking density, which cause the viscosity to increase with time. The latter hypothesis was confirmed by studying the viscosity of the emulsions in the presence of the curing catalysts (0.5 wt% salicylic acid based on second polyketones) as functions of time at room temperature and different protonation levels (40%, 50%, and 60%). It was found that the presence of the curing catalysts could result in a much higher viscosity compared with the emulsions without catalysts. This strongly indicates that the increase in viscosity can be explained by a chemical reason and namely the cross-linking mechanism proposed above, which could be both intra- and inter-particle. In case of the effect of ethylene content of the second polyketones, similar rheological behavior of the PK0 and PK50 can be observed as that of the PK30. This is in agreement with the particle size measurements discussed above.

52

Wood adhesive emulsions

Figure 3.16 Effect of storage time on the viscosity of the emulsions at (a) 60% protonation level; (b) 80% protonation level (1 h, 500 rpm, 50 °C).

3.3.4 Wood adhesive testing

The polyketone based emulsions can be easily applied and evenly spread at maple wood surface (left part of Figure 3.17). During the curing reactions in the hot press, water starts to evaporate and the polyamines at this stage act as curing agents to react with the second polyketones in presence of curing catalysts (salicylic acid) leading to water-resistant and stable bis-pyrrole units. The shear strength of the adhesive bonds has been studied on the maple substrates, according to the European Standard (EN-314), a 72-hours boiling water test. In this test, wood swells into all the directions in water and the adhesive must be able to withstand all these forces. Furthermore, the adhesive itself is tested on the hydrolysis, water adsorption, and thermal stability. The samples must fail by giving wood failure and contain fibers from the opposite veneer, indicating that wood adhesives are stronger than wood (right part of Figure 3.17). The shear strength should be higher than 1 MPa, which is the standard value of the EN-314. It has been proven that all tested emulsions pass the EN-314 with significantly higher shear strength values (almost 3 times of the standard value of the EN-314) (Figure 3.18). It is also found that protonation level and PK30/mPK30 ratio have no significant effect on the shear strength of the emulsions within the employed range. The adhesion is most probably related to the reaction at the surface between the carbonyl groups (mostly of the second virgin polyketones) and the hydroxyl groups present on the wood. In this respect it is reasonable to assume that the polyamines mainly act as surfactants and curing agents in this system and do not play an important role in term of the interaction with wood. Indeed only 30% of the carbonyl groups remain unreacted after reaction with the di-amines, and those remaining ones have obviously a high steric hindrance with respect to the “virgin” carbonyl groups of the second polyketones. When tested again after two months’ storage time, similar shear strength data as compared to freshly prepared adhesives was obtained. This illustrates the long shelf-life of the

0 40 80 120 160 200 240 280 320 360

0102030405060708090

100110120

Visc

ostiy

(Pa.

s)

Storage time (day)

1.25 1.5 1.75 2

PK30/mPK30(a)

0 40 80 120 160 200 240 280 320

0

20

40

60

80

100

120

140

Visc

ostiy

(Pa.

s)

Storage time (day)

1.25 1.5 1.75 2

PK30/mPK30(b)

53

Chapter 3

polyketones based adhesives, prepared according to the procedures described above. The wood adhesive test was also performed for a sample after the storage time of 2 years (Figure 3.18). It is suprising to observe that the performance of the polyketone-based emulsions remains the same as the fresh ones when applied as wood adhesives.

Figure 3.17 Spread of the polyketone based emulsions on the wood surface before hot-pressing (left) and the ruptured sample after shear strength test (right).

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.52 months

1 day2 years2 months

1 day2 months1 day

60605050504040

PK30/mPK30

Shea

r str

engt

h (M

Pa)

Protonation level (%)

1.5 1.75 2

Figure 3.18 Effect of protonation level, storage time, and PK30/mPK30 ratio on the shear strength of wood adhesives (1 h, 500 rpm, 50 °C).

54

Wood adhesive emulsions

3.4 Conclusions

mulsions are produced by chemical modifications of thermosetting

.5 References

er, J. J. US 5225523, 1993. A.; Heeres, H. J.; Van der Linden, A. J.;

90.

A.; Van Der Heide, E.; Vietje, G. US

; Freriks, J. US 5952459, 1999. 84080, 1997.

006, 31, 443.

Aqueous polymer epolyketones in a one-pot laboratory procedure. The properties of the resulting emulsions are dependent on the emulsification process, the protonation level and the quantity of the in-situ polymeric surfactants (polyamines). A multivariable, non-linear regression model has been developed in order to determine the effect of each experimental parameter on the resulting average particle size of the emulsions. The same model allows easy prediction of the particle size as functions of the experimental conditions. The prepared emulsions exhibit extremely long shelf life at room temperature (20 °C) and remain stable for at least 2 years. The average particle size, as measured by dynamic light scattering, remains basically the same while the viscosity of the system decreases in the first week and then levels off to a almost constant value. A hypothetical mechanism, which is able to explain this behavior, has been proposed. It involves a kind of “arms” (free polymer chains) retraction of the polyketone particles with respect to time. A direct proof of such mechanism is very difficult to obtain due to the complexity of the system. However, the hypothesis is substantiated both by the considerations on a deprotonation mechanism and viscosity measurements (the latter showing a clear transition from non-Newtonian to Newtonian behavior). When the emulsions are applied as a wood adhesive, a high shear strength can be achieved, which far exceeds the requirement of 1 MPa as demanded by the European Standard (EN-314).

3

1. Drent, E.; Keijsp2. Mul, W. P.; Dirkzwager, H.; Broekhuis, A.

Orpen, A. G. Inorg. Chim. Acta 2002, 327, 147. 3. Smaardijk, A. A.; Kramer, A. H. EP 0372602, 194. Van Der Heide, E.; Vietje, G. GB 2277520, 1994. 5. Van Druten, M. L.; Ruyter, H. P.; Ten Hoeve, A.

5633299, 1997. 6. Broekhuis, A. A.7. Van Der Heide, E.; Vietje, G.; Wang, P. C. US 568. Wong, P. K.; Pace, A. R.; Weber, R. C. US 5955563, 1999. 9. Wong, P. K. US 6214941, 2001. 10. Chern, C. S. Prog. Polym. Sci. 2

55

Chapter 3

11. Riess, G.; Labbe, C. Macromol. Rapid. Commun. 2004, 25, 401. 249.

.; Picchioni. F. J. Appl. Polym. Sci. 2008,

, K. Marcomol. Rapid. Commun. 2002, 23, 739. 004, 24, 327.

//www.iarc.fr/.

experiments, 5th ed., Wiley, 2000. r. Opin.

engineering, Wiley, 1997.

12. Liu, S.; Armes, S. P. Curr. Opin. Colloid. Interface. Sci. 2001, 6, 13. Riess, G. Prog. Polym. Sci. 2003, 28, 1107. 14. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A

107, 262. 15. Liu, Y.; Li16. Li, K.; Geng, X.; Simonsen, J.; Karchesy, J. Int. J. Adhes. 217. Press Release No.153 International Agency for Research on Cancer. http:18. Walstra, P. Chem. Eng. Sci. 1993, 48, 333. 19. Montgomery, D. C. Design and analysis of 20. Solans, C.; Izquierdo, P.; Nolla, J.; Azemar, N.; Garcia-Celma, M. J. Cur

Colloid. Interface. Sci. 2005, 10, 102. 21. Stokes, R. J.; Evans, D. F. Fundamentals of interfacial 22. Tsitsilianis, C.; Iliopulos, I.; Ducouret, G. Macromolecules 2000, 33, 2936.

56

Self-healing materials

Chapter 4

Thermally self-healing polymeric materials:

the next step to recycling thermoset polymers?

Abstract

Thermally self-healing polymeric materials were developed on the basis of furan-functionalized, alternating thermosetting polyketones (PK-furan) and bis-maleimide by using the Diels-Alder and Retro-Diels-Alder reaction sequence. The highly cross-linked PK-furan can be thermally remended through several repetitive cycles without any loss in fracture loading performance. It is found that the achieved self-healing ability of this easily accessible system provides recyclability and reworkability, which often is perceived as difficult or impossible for thermosetting polymers. The simplicity of the synthesis, the broad range of available polyketone precursors, and the striking healing ability of this system could expand the scientific understanding of self-healing materials and introduce the cradle-to-cradle concept for thermoset-based plastics and composites.

Key word: Self-healing materials; Diels-Alder; Polyfuran; Themoset recycling

57

Chapter 4

4.1 Introduction

Inspired by the phenomenon of self-healing in biological systems, the synthesis of man-made self-healing polymeric materials has become a newly emerging paradigm and a fascinating area of research.1-3 Self-healing materials have the capability to repair or recover themselves when suffering mechanically and/or thermally induced damage, which can occur autonomously4-5 or be activated by external stimuli6-8 (e.g. heat) for once or multiple times. This valuable characteristic could extend the life-time use of various polymeric products. However, costly and complex synthetic pathways, and the loss of mechanical properties1-8 after self-healing or remending so far limited their further development.

The Diels-Alder (DA) reaction and its Retro-Diels-Alder (RDA) analogue, represent a highly promising route to introduce self healing properties into polymeric systems.1-3 The earliest work dealing with thermally reversible polymeric network that contain DA functionalities was reported by Craven et al. in 1969.9 During the last few decades, two different strategic applications of this reaction sequence have been studied: (i) the polymerization of multifunctional monomers,6,7,10-15 e.g. a di- or tri-functional furan derivative and a bis-maleimide; (ii) the formation of cross-linked polymer networks16-25 from linear thermoplastics bearing pendant furan and/or maleimide groups. However, limited thermal reversibility and costly synthesis pathways in both of the strategic approaches prohibit their practical applications as self-healing materials.6-7,10-25 In contrast, we report a simple polymer system that could resolve these problems. The starting materials for this system consist of alternating polyketones, a new class of thermosetting polymers, obtained by alternating co- or ter-polymerization of carbon monoxide, ethylene, and propylene using homogeneous palladium-based catalysis.26-28 These polyketones can be easily converted into furan derivatives (PK-furan) via the Paal-Knorr reaction with furfurylamine. The PK-furan is successively converted by the DA reaction with the indicated bis-maleimide, resulting in a highly cross-linked polymeric network. Subsequently, the dissociation of the network can be accomplished at elevated temperature by the RDA reaction. Thus the sequence of cross-linking (DA), de-cross-linking (RDA), and re-cross-linking (DA) processes makes this polymer system not only remendable but also reworkable.

Nowadays themosetting resins are used in a wide range of applications including adhesives, coatings, polymer composites, electrical insulation, printed circuit boards, etc. However, the recycling at the end of their life cycle is a very difficult challenge due to the cross-linked nature.29-33 The cross-linked thermosetting resins can not be remelted or reshaped like thermoplastics, since heating leads to decomposition and degradation of the

58

Self-healing materials

materials. Furthermore, thermoset-based products, in particular thermoset composites, are usually formulated with various additives such as fillers (e.g. mineral powders, fire retardants), fibrous reinforcement (e.g. glass or carbon fiber), which result in complex mixtures that are difficult or impossible to separate. Thus, most of the thermoset-based products end-up in landfills or are incinerated after their functional use. Owing to the serious concern regarding the environment and the finite natural resources, new recycling technologies for thermosetting resins are urgently needed. In the present work, the before-mentioned thermally self-healing concept could well address these issues.

Here we first investigated the chemical reactivity of the polyketones with furfurylamine at different reaction conditions. A model component reaction was used for this study in order to facilitate the structure characterization for the prepared polymeric materials. The DA and RDA reactions between PK-furan and bis-maleimide were thoroughly studied by measuring gel time and gel content, and by performing NMR and FTIR spectroscopy. Thermal analysis of the cross-linked PK-furan was also performed by using differential scanning calorimetry (DSC), while the self-healing efficiency was evaluated by dynamic mechanical analysis (DMA) and 3-point bending tests.

4.2 Experimental

Materials. The alternating polyketones, co- and ter-polymers of carbon monoxide, ethylene, and propylene, with 0% ethylene (PK0, Mw 1680), 30% ethylene (PK30, Mw 3970), and 50% ethylene (PK50, Mw 5350) based on the total olefin content were synthesized according to a reported procedure.27,28 Furfurylamine (Aldrich, ≥99%), 2,5-hexanedione (Aldrich, ≥99%), 1,1′-(methylenedi-4,1-phenylene)bismaleimide (bis-maleimide, Aldrich, 95%), choloform (Lab-Scan, 99.5%), dichloromethane (Lab-Scan, 99.8%), and dimethyl sulfoxide (DMSO, Acros, 99.7%) were purchased and used as received.

Model component reaction. The reaction between 2,5-hexanedione (80 mg, 0.7 mmol) and furfurylamine (68 mg, 0.7 mmol) was first carried out in CDCl3 at 50 °C for 12 h directly in an NMR tube. The progress of the reaction was monitored with 1H-NMR spectroscopy and the spectra were recorded at regular time intervals of 30 min. A relatively large scale reaction between 2,5-hexanedione (5 g, 44 mmol) and furfurylamine (4.26 g, 44 mmol) without using any solvent was performed in a 50 ml round bottom flask equipped with a magnetic stirrer. The reaction mixture was heated at 70 °C for 20 h with vigorous stirring. After the reaction, the mixture was diluted with dichloromethane (150 ml) and dried over sodium sulfate. The solvent was removed under vacuum to yield pyrrolic furan (7.06 g, 94% yield).

59

Chapter 4

Furan functionalization of alternating polyketones. The reaction between polyketones (80 mg) and furfurylamine (in equimolar ratio between the 1,4-di-carbonyl groups of the polyketones and the amino groups) was first carried out in CDCl3 at 50 °C for 12 h in an NMR tube. The progress of the reaction was monitored with 1H-NMR spectroscopy and the spectra were recorded at regular time intervals of 30 min. The chemical modifications of the polyketones with furfurylamine were also carried out in the bulk in a sealed 250 ml round bottom glass reactor with a reflux condenser, a U-type anchor impeller, and an oil bath for heating. After the polyketones (40 g) were preheated to the liquid state at the employed reaction temperature (100 °C), furfurylamine was added dropwise into the reactor in the first 10 min. The stirring speed was set at a constant value of 500 rpm and the employed reaction time was 4 h. The conversion of furfurylamine was determined by 1H- NMR and reported as average of four different values (different spectra). The resulting experimental error (standard deviation) on the average value was less than 1%. Moreover, the conversion was found to be in full agreement with the results determined by potentiometric titration.34 The resulting polymers (after modifications) were washed several times with de-ionized Milli-Q water to remove unreacted furfurylamine. After filtering and freeze-drying, light brown polymers were obtained as the final products (PK-furan).

Model Diels-Alder and Retro-Diels-Alder reaction. The Diels-Alder (DA) reaction of pyrrolic furan (0.5 g, 2.85 mmol) and bis-maleimide (0.54 g, 1.43 mmol) was carried out in chloroform (5 ml) in a 50 ml round bottom flask equipped with a magnetic stirrer. The reaction mixture was heated at 50 °C for 24 h with vigorous stirring. After the reaction, the solvent was removed under vacuum to afford the model adduct as a yellowish solid. The Retro-Diels-Alder (RDA) reaction of the model adduct was performed in an NMR tube. The model adduct (50 mg) was dissolved in DMSO-d6 (0.6 ml) in a 4 ml glass vial. After transferring the solution in the NMR tube, the latter was immersed in a 150 °C oil bath for 5 min, followed by quenching in an ice-water bath. After that, 1H-NMR spectra were recorded immediately.

Polymer Diels-Alder and Retro-Diels-Alder reaction. The DA reaction of PK-furan (15 g) and bis-maleimide using chloroform (150 g) as solvent (10 wt% polymer based on solvent) was carried out in a 250 ml round bottom flask equipped with a magnetic stirrer. The gelation time was determined as the point at which the magnetic stirrer stopped to rotate. The cross-linked polymers were obtained by drying the polymer gel to constant weight at 50 °C under vacuum. The gel content was determined by Soxhlet extraction with boiling dichloromethane for 20 h. The DA and RDA reaction of the polymer were also performed in an NMR tube. PK-furan (25 mg) and bis-maleimide (21.5 mg) in equimolar

60

Self-healing materials

ratio (between furan and maleimide groups) were dissolved in DMSO-d6 (0.7 ml) and then transferred into an NMR tube. The reaction mixture was heated at 50 °C for 24 h to form the polymer adduct. After the reaction, the NMR tube was immersed in a 150 °C oil bath for 5 min, followed by quenching in an ice-water bath. 1H-NMR spectra were recorded instantly after quenching. For the cross-linked PK-furan after Soxhlet extraction and the removal of the solvent, the sample (50 mg) was added in DMSO-d6 (0.7 ml) in a small glass vial and then heated at 150 °C in an oil bath for 5 min. After quenching in an ice-water bath, the completely dissolved polymer solution was transferred in an NMR tube and then 1H-NMR spectra were instantly recorded.

Characterization. 1H-NMR spectra were recorded on a Varian Mercury Plus 400 MHz using CDCl3 or DMSO-d6 as a solvent. FTIR spectra were recorded using a Perkin-Elmer Spectrum 2000, equipped with a heating stage and temperature controller. Differential scanning calorimetry (DSC) was performed on a Perkin Elmer differential scanning calorimeter Pyris 1 under N2 atmosphere. The sample for DSC was weighted (ca. 8 mg) in an aluminum pan, which was then sealed. The sample was first heated from 25 °C to 180 °C, and then kept at 180°C for 1 min, followed by cooling to 25 °C. Four cycles were performed and the heating and cooling rates were set at 10 °C/min throughout the DSC measurements.

Quantifying healing efficiency. Dynamic mechanical analyses were conducted on a Rheometrics scientific solid analyzer (RSA II) under air environment using dual cantilever mode at an oscillation frequency of 1 Hz at a heating rate of 5 °C/min with the specimen size of 6 mm in width, 1.4 mm in thickness, and 54 mm in length. 3-point bending test was performed on a 4301 Instron machine using a 1 kN power sensor at a crossing head speed of 1 mm/min with the specimen size of 12.7 mm in width, 4 mm in thickness, and 64 mm in length. At least 8 specimens of every formulation for 3-point bending were tested with the standard deviation of fracture load less than 0.2 kN. The fracture surface of the samples after 3-point bending was examined by scanning electron microscopy (SEM) (JSM-6320 instrument). The samples were sputtered with Pt/Pd prior to SEM observation. The specimens for 3-point bending were prepared by compression molding of the cross-linked PK-furan into rectangular bars at 120 °C for 20 min under a pressure of about 4 MPa, followed by the thermal treatment at 50 °C for 24 h in an oven.

4.3 Results and discussion

4.3.1 Synthesis of furan-functionalized polyketones

The classic Paal-Knorr reaction, in which the 1,4-di-carbonyl moiety of the polyketones reacts with a primary amine function yielding a pyrrole unit, is one of the dominating

61

Chapter 4

reaction routes for the functionalization of alternating polyketones.34 In the present work, a class of low molecular weight polyketones (Mw 1500–5500), alternating co- and terpolymers of carbon monoxide, ethylene, and propylene were used as starting materials for chemical modifications. The furan-functionalized polyketones (PK-furan), which have N-substituted 2,5-pyrrolediyl groups incorporated into the backbone with a furan group pendant from the main chain, were synthesized by reacting the polyketones with furfurylamine (Figure 4.1). The mechanism of the Paal-Knorr reaction between a 1,4-di-carbonyl compound and a primary amine has already been described in Chapter 1. In contrast to various polymerization techniques,25 chemical modifications of polyketones provide a feasible, simple, and highly efficient route to prepare furan-functionalized polymers.

O

O

O

O

O

OH2N+

N

N

O

O

O

Figure 4.1 Scheme of the reaction between polyketones and furfurylamine.

In order to facilitate structural identification of the obtained PK-furan, a model compound reaction between 2,5-hexanedione and furfurylamine was first studied directly in NMR tubes using CDCl3 as solvent at 50°C. The NMR spectra of the starting reaction mixture (I) and that after 12 h reaction time (II) are reported in Figure 4.2a. After the Paal-Knorr reaction, the appearance of new signals for the pyrrole ring (about 5.8 ppm) and the –CH2- group (attached to the pyrrole at 4.9 ppm) as well as the down-field shifting of the furan ring signal all demonstrate the formation of the reported product. The conversion of furfurylamine (Figure 4.2b) could be calculated from the intensity ratio of peaks at 4.9 ppm (-CH2- group of the converted furfurylamine) and 3.8 ppm (-CH2- of unreacted furfurylamine). The conversion increased slowly with reaction time and reached 64% after the reaction time of 12 h. With respect to the bulk reaction of model compounds, the model reaction proceeded to almost 100% furfurylamine conversion at elevated temperature (70 °C) at the reaction time of 20 h.

62

Self-healing materials

234567

78

1

3 2

45

6

N

O

II6 5 4

2,31

ppm

7,8(a)

I

0 2 4 6 8 10 120

10

20

30

40

50

60

70

Con

vers

ion

(%)

Reaction time (h)

(b)

Figure 4.2 NMR tube reaction between 2,5-hexanedione and furfurylamine at 50 °C (a) 1H-NMR spectra before reaction (I) and after 12 h (II); (b) furfurylamine conversion, determined by 1H-NMR.

With the information given by the model compound reaction, NMR tube reactions between polyketones (PK0, PK30, and PK50) and furfurylamine were investigated at equimolar ratio between furfurylamine and the 1,4-di-carbonyl groups of the polyketones. NMR spectra of the starting reaction mixture (PK50 and furfurylamine) (I) and that after 12 h reaction time at 50 °C (II) are shown in Figure 4.3a. Consistent with the model compound analysis, the broad absorbance peaks at 5.8 ppm and 4.9 ppm are assigned to the pyrrole ring in the backbone and the –CH2- group attached to the pyrrole ring, respectively. The broad absorbance peaks at 7.3, 6.2, and 5.9 ppm can be ascribed to the hydrogens on the furan ring attached to the polymer backbone. Using the intensity ratio of the -CH2- peak of furfurylamine before and after reaction, the reaction conversion of furfurylamine are shown as a function of reaction time in Figure 4.3b. It is interesting to observe that the

63

Chapter 4

Paal-Knorr reaction for the polyketones with furfurylamine can easily take place even under mild condition (50 °C). The reaction kinetics of the polymers exhibits a similar profile as that of the model compounds after 12 h reaction time, reaching 17.5% (PK0), 35% (PK30), and 68% (PK50), respectively. The reaction rates and conversion values were found to increase with the ethylene content of the polyketones, which is due to lower steric hindrance of the carbonyl groups and a reduced C-C rotation barrier for the polymers with the higher ethylene content.34 Furthermore, FTIR spectra of polyketones (PK50) before and after chemical modifications (Figure 4.4) also demonstrated the presence of pyrrole rings (3112 cm-1 and 1596 cm-1) in the backbone and the bearing of furan rings (734 cm-1) at the side chain for the furan-functionalized polyketones.

1234567ppm

N

N

O

O

O1

1

3 2

3

5

2

5

4

46

6

II6 5

4 2,3

(a)

1

I

0 2 4 6 8 10 120

10

20

30

40

50

60

70

Con

vers

ion

(%)

Reaction time (h)

PK0 PK30 PK50

(b)

Figure 4.3 NMR tube reactions between polyketones (PK0, PK30, and PK50) and furfurylamine at 50 °C (a) 1H-NMR spectra (PK50 and furfurylamine) before reaction (I) and after 12 h (II); (b) furfurylamine conversion, determined by 1H-NMR.

64

Self-healing materials

500100015002000250030003500

Abs

orpt

ion

I

II

Wavenummber (cm-1)

Figure 4.4 FTIR spectra of polyketones (PK50) (I) and furan-functionalized PK50 (II).

The bulk reactions between polyketones and furfurylamine (without the use of any catalysts and/or additives) were studied at 100 °C using a reaction time of 4 h while varying the amine/di-ketone ratio (INH2/CO) and the ethylene content of the polyketones. The conversion data for furfurylamine (XNH2) and carbonyl groups (XCO) are shown in Table 4.1. All amine conversions with PK50 proceeded exceedingly well to 98% for INH2/CO=0.8 and 100% for INH2/CO of 0.6, 0.4, and 0.2, respectively. We can see that the degree of furan functionality can simply be tuned by varying the INH2/CO. With respect to the effect of ethylene content of the polyketones, amine conversions of 75% and 62% for PK30 and PK0 were obtained at the applied reaction conditions, respectively, which also indicates that a higher ethylene content resulted in higher conversion values. This is in line with the results obtained for the reactions carried out in the NMR tubes.

PK-furan Polyketones INH2/CO XNH2 (%) XCO (%)

PK50f-1 PK50 0.8 98 78

PK50f-2 PK50 0.6 100 60

PK50f-3 PK50 0.4 100 40

PK50f-4 PK50 0.2 100 20

PK30f PK30 0.8 75 60

PK0f PK0 0.8 62 50

65

Chapter 4

Table 4.1 Conversion data of furfurylamine (XNH2) and carbonyl groups of polyketones (XCO) as a function of INH2/CO for different types of polyketones ( reaction temperature 100 °C, reaction time 4 h).

4.3.2 Diels-Alder and Retro-Diels-Alder reaction of PK-furan

The cross-linking of PK-furan with bis-maleimide via the DA reaction (Figure 4.5) was studied as a function of the initial molar ratio between maleimide and furan groups (Ima/fur), the degree of furan functionality on the polymer backbone, and the ethylene content of the polyketones (Table 4.2). The PK-furan derived from PK50 can easily be cross-linked with bis-maleimide to form a gel, irrespective of Ima/fur and the degree of furan functionality. The fastest gelation time was about 2 h for PK50f-1 at Ima/fur=1. It is worth noticing that the gelation time increases with decreasing the Ima/fur and the degree of furan functionality. This is related to the fact that the reaction kinetics at a fixed weight content of the polymers strongly depend on initial concentration of furan and maleimide groups. Regarding the effect of the ethylene content in the polyketones, the gelation time of PK50f is much faster that of PK30f and no gel formation was observed for PK0f even after the reaction time of 4 days. This discrepancy is believed to be due to the difference in molecular weight of the starting polyketones. With respect to gel content (corresponding to the number of cross-linking points) for all the studied samples, it is found that high gel contents (in the range of 92-95%) were obtained at high maleimide/furan ratios and a high degree of furan functionality, clearly indicating the high conversion level for the DA reaction.

N

N

O

O

O

RDA reaction

N

O

O

N

O

O

DA reaction

N

N

OO

O

NN

O

O

N

O

O

NO

O

O

Figure 4.5 Scheme of the Diels-Alder and Retro-Diels-Alder reaction between PK-furan and bis-maleimide.

66

Self-healing materials

PK-furan Ima/fur gelation time (h)

gel content (%)

PK50f-1 0.25 9 59

PK50f-1 0.5 3 92

PK50f-1 0.75 2.5 95

PK50f-1 1 2.2 93

PK50f-2 1 2.5 95

PK50f-3 1 3 87

PK50f-4 1 8.5 68

PK30f 1 4.5 87

Table 4.2 Cross-linking of PK-furan with bis-maleimide via the DA reaction at the effect of maleimide/furan ratio (Ima/fur), the degree of furan functionalization, and the ethylene content of the polyketones (reaction temperature 50 °C, reaction time 24 h, polymer coding as in Table 4.1).

The cycles of the DA and RDA reaction of PK-furan with bis-maleimide were first investigated in solvent. Figure 4.6a shows the photographs of the DA and RDA reaction in DMSO solvent (10 wt% of PK-furan based on solvent) for PK50f-1 at Ima/fur=1. The forward reaction occurred at low temperature and led to gel formation (2 of Figure 4.6a). The resulting polymer gels were completely reversed to clear and fluid solutions (3 of Fig 4.6a) upon heating, which shows the same appearance as the starting mixture (1 of Fig. 4.6a). Contrary to the results16-25 described in the literature for the thermoplastics bearing pendant furan and/or maleimide groups, the DA and RDA reactions here are characterized by ultra-fast kinetics (gel formation in 2 h at 50 °C and its reversal in 5 min at 150 °C or in 10 min at 120 °C) which in turn can also be tuned by adjusting the maleimide to furan molar ratio. The cycle of gelation and its reversal was repeated 4 times without noticing any relevant changes in appearance, thus giving a clear preliminary indication that the PK-furan/bis-maleimide system is fully thermally reversible. Moreover, the fast kinetics of the RDA reaction were further confirmed in Figure 4.6b where the highly cross-linked polymers (PK50f-1 at Ima/fur=1) after Soxhlet extraction and the removal of the solvent were insoluble in DMSO (upper part in Figure 4.6b) but converted back to a clear solution (bottom part in Figure 4.6b) in less than 5 min upon heating.

67

Chapter 4

(a) (b)

Figure 4.6 (a) Gelation and its reversal for PK50f-1 with bis-maleimide at Ima/fur=1 in DMSO solvent: 1. initial reaction mixture; 2. gel formation at 50 °C after 2 h; 3. back to soluble solution after 5 min at 150°C; (b) Insoluble cross-linked PK50f-1 at Ima/fur=1 in DMSO (upper) at 50 °C and fully soluble solution (bottom) after 5 min at 150 °C.

1H-NMR spectroscopy was used to study the forward and reverse DA reaction of PK-furan with bis-maleimide. To facilitate structure identification, the model DA adduct of pyrrolic furan with bis-maleimide was first used to study the thermal behavior. It is observed that the reaction product after DA reaction (I of Figure 4.7a) contains the endo and exo forms of the DA adduct and some of the unreacted starting materials. According to the relative intensity ratio of the reacted and unreacted pyrrolic furan, the yield of the DA (including exo and endo foams) adduct was found to be 96% with an exo/endo ratio of around 70/30. Upon heating at 150 °C for 5 min, the RDA reaction on the model compound proceeded almost completely, as demonstrated by the corresponding NMR spectrum (II of Figure 4.7a), where only peaks from the pyrrolic furan and the bis-maleimide could be identified. For the polymeric systems, the features of the DA adduct that are in line with the model study, namely H1, H2 at 3.2 ppm, H3 at 5.1 ppm, and H4, H5 at 6.5 ppm, appeared after the reaction (II of Figure 4.7b). The RDA reaction of the polymer adduct was performed using the same conditions as for the model adduct. No features of the DA adduct could be detected in III of Figure 4.7b. Another NMR study for the RDA reaction was carried out for the cross-linked PK50f-1 at Ima/fur=1 after Soxhlet extraction and removal of the solvent. Upon heating in less than 5 min at 150 °C or 10 min at 120 °C, a clear polymer solution in DMSO-d6 was obtained. The NMR spectrum of this polymer solution was found to be practically the same as I of Figure 4.7b that only consists

68

Self-healing materials

of absorptions assigned to PK50f-1 and bis-maleimide. All these results provide the strong evidence of 100% or full thermal reversibility of the cross-linked polymer systems.

234567

exoendo

exoendo

exo

endo

7,8 6 9

ppm

111,122,3

endo+exo

II

s

N

O

O

N

O

O

O O

N N

1234

5

67

8910

1112

13

Iexo

endo

exo

endo

exoendos

4,5

(a)

10,13

12345678 ppm

(b)

1,24,5

I

II

ON

O

O1

23

4

5

3

III

Figure 4.7 (a) NMR spectra of model adduct (I) and reversal of model adduct (II); (b) NMR spectra of initial reaction mixture of PK50f-1 with bis-maleimide at Ima/fur=1 (I), polymer adduct after 24 h at 50 °C (II), and reversal of polymer adduct after 5 min at 150 °C (III).

69

Chapter 4

In addition, FTIR spectroscopy was used to study the cycle of the DA and RDA reactions of PK-furan with bis-maleimide in the solid state. The following steps were used: DA-RDA-DA (Figure 4.8). The striking phenomenon during the heating and cooling cycle is the up and down switch of the intensity of the absorption peak at 1182 cm-1. The spectrum of the cross-linked PK-furan (PK50f-1 at Ifur/Ma =1) after solvent extraction and removal of the solvent is shown in I of Figure 4.8. Upon heating at 150 °C for 5 min, a significant intensity decrease of the absorption peak at 1182 cm-1 (II of Figure 4.8) was observed, which could be ascribed to the debonding of the DA adduct (C-O-C) via the retro-reaction. No change of the spectra could be detected after a further increase of heating time to 20 min at 150 °C, indicating the complete debonding of the DA adduct in the short time frame of 5 min, which is in line with our study in the liquid state. After slow cooling from 150 °C to room temperature in 20 min, III in Figure 4.8 shows the recovery of intensity at 1182 cm-1 with further the same characteristics as given in I of Figure 4.8. This can be explained by the regeneration of the DA adduct (C-O-C) during the cooling step. Four cycles of RDA and DA reactions were performed and similar results (i.e. up and down switch of intensity at 1182 cm-1) were obtained. Therefore, it can be concluded that the PK-furan can be repeatedly de-cross-linked and re-cross-linked with bis-maleimide in the solid state by simple heating and cooling cycles.

60080010001200140016001800

III

II

Wavenumber (cm-1)

I

Abs

orba

nce

C-O-C1182

Figure 4.8 FTIR spectra of the cross-linked PK-furan (PK50f-1 at Ifur/Ma=1) before heat treatment at room temperature (I), upon heating at 150 °C for 5 min (II), and after cooling down to room temperature in 20 min (III).

70

Self-healing materials

The high efficiency of the DA and RDA process is further confirmed by the thermal behavior of the cross-linked PK-furan (Figure 4.9), as measured by the repeated differential scanning calorimetry (DSC) cycles. DSC thermograms display an overlapping of the endothermic (ascribed to the RDA reaction) and exothermic transitions (ascribed to the DA reaction) during the successive thermal cycles. The slight deviations of the absorption profile especially in the heating cycles can be attributed to the relatively short time scale of the DSC experiment, which does not allow a full reconstitution of the DA network. With respect to thermal behavior of previously reported DA polymer products10,13-15,17, none of them was able to show any endothermic transition at a second heating cycle, which suggests limited reversibility or very slow kinetics of the DA adduct formation during the first cooling scan. This highlights the capability of the current system as a self-healing material.

20 40 60 80 100 120 140 160 18016

18

20

22

24

1 2 3 4

Hea

t flo

w (m

W) e

ndo

up

Tempertature (°C)

Figure 4.9 Thermal behavior of the cross-linked PK-furan (PK50f-1 at Ima/fur=1).

71

Chapter 4

4.3.3 Thermal self-healing

The fact that the cross-linked PK-furan is thermally remendable is further demonstrated by compression molding of small granules of the cross-linked PK-furan into uniform bars at elevated temperature (typically 110-150 °C, 10-30 min processing time) (Figure 4.10). When exposed to heat, the polymers display the relevant properties of linear thermoplastics, such as remeltability, reprocessability, and recyclability, because of the opening of the DA adduct. After slowly cooling to room temperature (30-40 min), a rigid structural polymer network can be achieved due to the regeneration of the DA adduct. In this process, polymer chains are able to reorganize and therefore reconstruct or remodel themselves into any desired physical shape. Besides the remendability, a breakthrough has been made here in terms of full recyclability and reworkability of this system. Nowadays, recycling of thermosetting materials at the end of the life-cycle is often considered as difficult and remains an unresolved issue.29-33 The current recycling techniques involve mainly mechanical (the use of grinding techniques to comminute thermoset recyclates as reinforcing fillers in new composites) and thermal processing steps (the use of heat to break the scarp composites down to recover the energy and feedstock), but they do not represent convenient choices from an economic and environmental point of view.29-33 In contrast, our system can provide a feasible way to resolve the recycling issues of thermosetting polymers.

Figure 4.10 Photographs of the cross-linked polymers before self-healing (upper) in form of granules and after remending in form of uniform bars (bottom).

72

Self-healing materials

The thermal self-healing ability of the cross-linked PK-furan was further studied by using dynamic mechanical analysis (DMA). Dynamic mechanical properties of the cross-linked PK-furan (storage modulus G΄, loss modulus G΄΄, tanδ versus temperature) are shown in Figure 4.11. We observed that dynamic mechanical properties can be easily modulated by adjusting Ima/fur. It can be seen in Figure 4.11a that there is a gradual increase of G΄ and a decrease of G΄΄ with the increase of Ima/fur from 0.5 to 1 at the glassy state, indicating that the stiffness of the polymers increases with Ima/fur. The shift of tanδ to a higher level with Ima/fur also gave an indication of less damping characteristic or less flexibility at higher Ima/fur values. The observed dynamic mechanical properties can be ascribed to the fact that the change of Ima/fur leads to the difference in cross-linking density in terms of the number of the formed DA adducts. The glass transition (Tg) or softening temperature, as determined from the inflection point of the G΄ curves, for the cross-linked PK50f-1 at Ima/fur of 1, 0.75, and 0.5 are around 100 °C, 96 °C, and 87 °C, respectively, thus indirectly confirming the change in cross-linking density. The obtained Tg is found to be comparable or even higher than other DA related polymer systems reported in the literature7,12. Nevertheless, the relatively low Tg in comparison to some commercial thermosetting resins (e.g. epoxy) may limit the use of the proposed materials for applications that require softening points higher than 100 °C. Similar results are obtained when DMA is performed for samples with a relatively low degree of furan functionality (PK50f-2, PK30f) compared to PK50f-1 at Ima/fur =1 (Figure 4.11b), where Tg values of 93 °C and 91 °C are obtained for PK50f-2 and PK30f, respectively.

50 60 70 80 90 100 110

1E7

1E8

1E9

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

tan δ

G''

G'

tan

δ

G',

G''

(Pa)

Temperature (°C)

1 0.75 0.5

(a)

73

Chapter 4

50 60 70 80 90 100 110

1E7

1E8

1E9

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

tan

δ

tan δ

(b)

G''

G'

Temperature (°C)

G',

G''

(Pa)

PK50f-1 PK50f-2 PK30f

Figure 4.11 DMA analysis of (a) the cross-linked PK50f-1 as a function of Ima/fur; (b) the cross-linked PK50f-1, PK50f-2, and PK30f at Ima/fur=1.

DMA was also performed in a cyclic manner to check reworkability of the DA and RDA sequence on the sample PK50f-1 at Ima/fur=1 (Figure 4.12). When reaching its glass transition temperature, the sample starts to become soft due to the occurrence of the RDA reaction but can still retain its original shape. Upon cooling down in 10 min, the G΄ and G΄΄ of the samples recover due to the reformation of the DA adduct. The cycle was repeated sequentially for 6 times. Even without using antidegradants in the formulation, it is shown that the dynamic mechanical properties of the tested sample (Figure 4.12a) remain almost unchanged after the repeated 6 cycles, although a small drop in Tg value of 3-4 degrees is observed after the first cycle. This effect could be attributed to the discrepancy in time scale between the measurement time of the DMA and the kinetics of the DA adduct formation, in analogy to the results found in the DSC study. The confirmation for this hypothesis was obtained by thermal treatment of the same 6-cycle sample at 50 °C for 24 h in order to fully recover the DA adduct qualities. When testing again, the behavior of cycle 7 (Figure 4.12b) almost matches that of cycle 1 in terms of the G΄, G΄΄, and tanδ, which proves that this polymer system is 100% self-repairable under heat damage.

74

Self-healing materials

50 60 70 80 90 100 110

1E7

1E8

1E9

0.00

0.05

0.10

0.15

0.20

0.25

0.30

tan δ

G''

G'

tan

δ

G',

G''

(Pa)

Temperature (°C)

1 2 3 4 5 6

(a)

50 60 70 80 90 100 110

1E7

1E8

1E9

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

tan

δ

tan δ

G''

G'

G',

G''

(Pa)

Temperature (°C)

1 7

(b)

Figure 4.12 (a) dynamic mechanical properties in response to continuous heating cycle scanning for the cross-linked PK50f-1 at Ima/fur = 1; (b) dynamic mechanical properties of the sample of (a) after heat treatment at 50 °C for 24 h.

Self-healing efficiency4 (e.g. remended to original in fracture load) of the repaired polymers was further studied by using a 3-point bending test (Figure 4.13) at room temperature (23 °C). Representative “load to displacement” curves show typical thermosetting mechanical behavior (linear load to displacement relationship) of the prepared samples (PK50f-1 at Ima/fur of 1, 0.75, and 0.5) (Figure 4.14). It can be seen that a lower amount of bis-maleimide can lead to a higher fracture load, which could be due to a more efficient formation of the cross-linked network and higher toughness at relatively low cross-linking densities. After the bending test, the fractured samples were shredded into

75

Chapter 4

small granulates using a table-top hammer mill and then remended into rectangular bars by using compression molding at 120 °C for 20 min. When testing the remended samples once again, a full recovery of the fracture load with a healing efficiency of 100% was achieved (Figure 4.14a). For other reported self-healing systems in the literature4-8, the recovery efficiency never reaches a full completion and a relevant deterioration of the mechanical behavior was found. SEM examination of the fracture surface of the original and healed samples after testing (Figure 4.15) indicates that the polymers have the ability to remend themselves to a similar microstructure. Both fracture surfaces gave similar appearance (sharp, clear, and ligament shapes), which are a typical for a brittle or thermosetting material.35-37 The healing performance was also evaluated for multiple cycles, indicating that the polymers, after 3 cycles, do display full recovery to the original fracture load (Figure 4.14b).

Figure 4.13 Photograph representation of 3 point-bending test.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.20.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

Displacement (mm)

1 0.75 0.5

(a)

Original

Healed

Load

(kN

)

76

Self-healing materials

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.60.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

Load

(kN

)

Displacement (mm)

Original Cycle 1 Cycle 2 Cycle 3

(b)

Figure 4.14 (a) Representative load to displacement behavior of the original and healed samples (PK50f-1 at Ima/fur of 1, 0.75, and 0.5); (b) Representative load to displacement behavior of the samples (PK50f-1 at Ima/fur =0.75) upon multiple healing cycles.

Figure 4.15 SEM images of fracture surface of (a) original and (b) healed polymers.

77

Chapter 4

4.4 Conclusions

We have developed a thermally self-healing polymeric material using a simple and efficient processing method. To the best of our knowledge, this system introduces a new and practically applicable concept to recycle thermosetting polymers, through a Diels-Alder and Retro-Diels-Alder reaction sequence. The polyketones can easily be converted into furan derivatives in bulk without the need for a catalyst or a solvent just using mild conditions, while the degree of furan functionality can be well tuned by changing the initial reaction conditions. The NMR and FTIR spectroscopy as well as thermal analysis indicated that the furan-functionalized polyketones could be repeatedly cross-linked and de-cross-linked with bis-maleimide by using only one external stimulus (heat). DMA analysis and 3-point bending tests demonstrated that this polymeric system can repeatedly heal or repair itself to the extent of 100% for multiple times. Thus, we hope that our polymer system will move the research on self-healing materials into a new stage.

4.5 References

1. Bergman, S. D.; Wudl, F. J. Mater. Chem. 2008, 18, 41. 2. Wu, D. Y.; Meure, S.; Solomon, D. Prog. Polym. Sci. 2008, 33, 479. 3. Wool, R. P. Soft. Matter. 2008, 4, 400. 4. White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore. J. S.; Kessler, M. R.; Sriram, S. R.;

Brown, E. N.; Viswanathan, S. Nature 2001, 409, 794. 5. Toohey, K. S.; Sottos, N. R.; Lewis, J. A.; Moore, J. S.; White, S. R. Nat. Mater. 2007, 6,

581. 6. Chen, X.; Dam, M. A.; Ono, K.; Mal, A.; Shen, H.; Nutt, S. R.; Sheran, K.; Wudl, F.

Science 2002, 295, 1698. 7. Chen, X.; Wudl, F.; Mal, A. K.; Shen, H.; Nutt, S. R. Macromolecules 2003, 36, 1802. 8. Cordier, P.; Tournilhac, F.; Soulié-Ziakovic, C.; Leibler, L. Nature 2008, 451, 977. 9. Craven, J. M. US 3435003, 1969. 10. Goussé, C.; Gandini, A. Polym. Intern. 1999, 48, 723. 11. Kamahori, K.; Tada, S.; Ito, K.; Itsuno, S. Macromolecules 1999, 32, 541. 12. Mcelhanon, J. R.; Russick, E. M.; Wheeler, D. R.; Loy, D. A.; Aubert, J. H. J. Appl.

Polym. Sci. 2002, 85, 1496. 13. Teramoto, N.; Arai, Y.; Shibata, M. Carbohydr. Polym. 2006, 64, 78. 14. Liu, Y.; Hsieh, C. J. Polym. Sci. Polym. Chem. 2006, 44, 905. 15. Watanabe, M.; Yoshie, N. Polymer 2006, 47, 4946.

78

Self-healing materials

16. Chujo, Y.; Sada, K.; Saegusa, T. Macromolecules 1990, 23, 2636. 17. Canary, S. A.; Stevens, M. P. J. Polym. Sci. Polym. Chem. 1992, 30, 1755. 18. Laita, H.; Boufi, S.; Gandini, A. Eur. Polym. J. 1997, 33, 1203. 19. Goussé, C.; Gandini, A.; Hodge, P. Macromolecules 1998, 31, 314. 20. Jones, J. R.; Liotta, C. L.; Collard, D. M.; Schiraldi, D. A. Macromolecules 1999, 32,

5786. 21. Imai, Y.; Itoh, H.; Naka, K.; Chujo, Y. Macromolecules 2000, 33, 4343. 22. Gheneim, R.; Perez-Berumen, C.; Gandini, A. Macromolecules 2000, 35, 7246. 23. Liu. Y.; Hsieh, C.; Chen, Y. Polymer 2006, 47, 2581. 24. Liu, Y.; Chen, Y. Macromol. Chem. Phys. 2007, 208, 224. 25. Gandini, A.; Belgacem, M. N. Prog. Polym. Sci. 1997, 22, 1203. 26. Drent, E.; Budzelaar, P. H. M. Chem. Rev. 1996, 96, 663. 27. Drent, E.; Keijsper, J. J. US 5225523, 1993. 28. Mul, W. P; Dirkzwager, H.; Broekhuis, A. A.; Heeres, H. J.;Van Der Linden, A. J.;

Orpen, A. G. Inorg. Chim. Acta 2002, 327, 147. 29. Pickering, S. J. Compos. Pt. A-Appl. Sci. Manuf. 2006, 37, 1206. 30. Derosa, R.; Telfeyan, E.; Mayes, J. S. J. Thermoplast. Compos. Mater. 2005, 18, 219. 31. Raghavan, J.; Wool, R. P. J. Appl. Polym. Sci. 1999, 71, 775. 32. Buchwalter, S. L.; Kosbar, L. L. J. Polym. Sci. Polym. Chem. 1996, 34, 249. 33. Chen, J.; Ober, C. K.; Poliks, M. D. Polymer 2002, 43, 131. 34. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F. J. Appl. Polym. Sci. 2008,

107, 262. 35. Lekakou, C.; Kontodimopoulos, I.; Murugesh, A. K.; Chen, Y. L.; Jesson, D. A.; Watts,

J. F.; Smith, P. A. Polym. Eng. Sci. 2008, 48, 216. 36. Jayakumari, L. S.; Thulasiraman, V.; Sarojadevi, M. Polym. Compos. 2008, 29, 709. 37. Li, Y.; Feng, L.; Zhang, L. J. Appl. Polym. Sci. 2006, 100, 593.

79

Chapter 4

80

Crosslinking of MWNTs

Chapter 5

Cross-linking of multi-walled carbon nanotubes

with polymeric amines

Abstract

Functionalization of carbon nanotubes is considered as an essential step to enable their manipulation and application in potential end-use products. In this chapter, we introduce a new approach to functionalize multi-walled carbon nanotubes (MWNTs) by applying an amidation-type grafting reaction with amino-functionalized alternating polyketones. The functionalized MWNTs were characterized by using thermogravimetric analysis (TGA), X-ray photoemission spectroscopy (XPS), element analysis, Raman spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). Around 40 wt% polyamines based on the total weight of the MWNTs can be covalently attached to the surface of the MWNTs. It is found that polyamines act as cross-linking agents to interconnect or cross-link the MWNTs within and between bundles, as demonstrated by SEM and TEM analysis. After cross-linking, the functionalized MWNTs are insoluble in any solvent. The cross-linked MWNTs can be melt-blended into polyethylene and the resulting composites show comparable mechanical properties to those obtained by simple blending of “un-cross-linked” nanotubes with polyethylene. Key word: Carbon nanotubes; Cross-linking; Polyamines; Carbon nanotube junctions

Based on: Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Fernandez-Landaluce, T.; Fausti, D.; Rudolf, P.; Picchioni, F. Cross-linking of multi-walled carbon nanotubes with polymeric amines, Macromolecules 2008, 41, 6141-6146.

81

Chapter 5

5.1 Introduction

Carbon nanotubes have attracted great research interest because of their unique structural, mechanical and electrical properties.1 Functionalization of carbon nanotubes via non-covalent adsorption, wrapping of various functional molecules or covalent attachment of chemical groups is essential to facilitate their manipulation (e.g. dispersion of carbon nanotubes in various solvents) for many further applications.2-4 It has been shown that carbon nanotubes can be dispersed in aqueous media or organic solvents via non-covalent adsorption of low molecular weight surfactants5 and polymers6-8 (such as diblock copolymers and amphiphilic polymers). The covalent functionalization with organic molecules is mostly achieved by exploiting the reactivity of carbon nanotube-bound carboxylic acids via amidation or esterification reactions.9-11 The carboxylic acid groups can be obtained on the surface of the nanotubes by oxidation (usually from intrinsic or induced surface defects). In terms of polymer systems, the so-called “grafting onto” (attaching polymers with reactive or functional groups onto carbon nanotubes) and “grafting from” (growing polymers from carbon nanotube surfaces by in-situ polymerization) approaches are applied to functionalize carbon nanotubes.12-15

Some chemical functionalizations using bi-functional molecules enable the interlinking or cross-linking of individual carbon nanotubes to form carbon nanotube junctions or complex networks for the application as nano-scale electronic circuits.16,17 The first interconnection (e.g. end-to-end and end-to-side heterojunctions) of single wall carbon nanotubes (SWNTs) has been achieved by using aliphatic diamines as linkers via amide linkages.17-19 Using a cycloaddition reaction, SWNTs can also be cross-linked within a bundle as well as between bundles by using bifunctional nitrenes.20 With respect to multi-walled carbon nanotubes (MWNTs), it has been reported that interconnects of MWNTs can be formed through amide linkage with an inorganic metal complex.21 Alternative ways for cross-linking of carbon nanotubes include the use of electron or ion-beam irradiation.22-

25 Introducing cross-links between the carbon nanotube bundles could lead to dramatic improvement in mechanical strength,26,27 since the mechanical properties of nanotube bundles after production are limited by the sliding of individual carbon nanotubes along each other because of the relatively weak van der Waals interaction between them. Here we report the use of a polymeric component to create cross-linking points within or between the bundles of MWNTs via an amidation reaction. Based on the “grafting onto” approach, chemical cross-linking of MWNTs was carried out by using amino-functionalized alternating polyketones as the polymeric component.

82

Crosslinking of MWNTs

The polymeric amines (polyamines) used for this study were first prepared by chemical modifications of alternating aliphatic polyketones via the Paal-Knorr reaction.28 XPS, element analysis, TGA, and Raman spectroscopy were employed to characterize the functionalized MWNTs. The presence of polymer linkage between carbon nanotubes was examined with TEM and SEM. Furthermore, LDPE/cross-linked MWNTs composites were prepared using a twin-screw microextruder and the morphology and mechanical properties of the composites were investigated. The use of polymers for the cross-linking of carbon nanotubes has not been reported before. We believe this research can open up new approaches for the preparation of carbon nanotube interconnects.

5.2 Experimental

Materials. MWNTs (>90%) produced by CVD (Chemical Vapour Deposition), were purchased from Aldrich with outer-diameter 10-15 nm, inner-diameter 2-6 nm and length 0.1-10 μm. The alternating polyketones (Mw 3970), ter-polymers of carbon monoxide, ethylene, and propylene were synthesized according to a reported procedure.29 1,2-diaminopropane (1,2-DAP, Acros, 99%), thionyl chloride (Fluka, ≥99%), nitric acid (Merck, 65%), THF (Acros, >99%), toluene (Lab-Scan, 99.5%), chloroform (Lab-Scan, 99.5%), and DMF (Acros, >99% ) were purchased and used as received. Anhydrous THF, toluene (Aldrich, anhydrous, 99.8%) for the functionalization of MWNTs were dried over Al2O3 (Fluka). Anhydrous solvents were degassed prior to use and stored under nitrogen. The low density polyethylene (LDPE) in this study was purchased from Aldrich with a melt index 25 g/10min (190 °C/2.16 kg) and density 0.925 g/ml at 25 °C.

Preparation of polyamines. Polyamines were prepared in bulk by reacting two components (polyketones and 1,2-DAP) in a one-pot synthesis at 100 °C for 4 hours. After reaction, the resulting mixtures were washed several times with de-ionized Milli-Q water. After filtering and freeze-drying, light brown polymers were obtained as final products. Details of the procedure and characterization of the prepared polyamines were described in Chapter 2.

Chemical functionalization of MWNTs. MWNTs (3 g) were added to 65% HNO3 aqueous solution (100 ml). The mixture was first treated in an ultrasonic bath (Bransonic 2510) for 30 min and then stirred for 24 h under reflux. After cooling down to room temperature, the resulting mixture was diluted with 200 ml of deionized water and vacuum-filtered through 0.2 μm polycarbonate membrane. The obtained solid was washed with deionized water until the pH of the filtrate was 7. After drying overnight under vacuum at 60 °C, a solid (MWNTs-COOH, 1.65 g) was obtained. MWNTs-COOH (0.5 g)

83

Chapter 5

was suspended in thionyl chloride (40 ml). The mixture was stirred for 24 h under reflux, followed by the removal of excess thionyl chloride under vacuum, giving acyl chloride-functionalized MWNTs (MWNT-COCl). The MWNT-COCl was washed with anhydrous THF (20 ml) and the excess THF was removed under vacuum. The polyamines (6.5 g) were first dissolved in anhydrous toluene (40 ml) at room temperature. Then the dissolved polyamines in toluene were added into MWNT-COCl (0.45 g). The reaction mixture was first sonicated for 30 min at room temperature and then vigorously stirred at 90 °C under nitrogen atmosphere for the reaction time of 30 h. After the reaction was finished, the entire mixture was centrifuged at 3000 rpm for 10 min. The functionlized MWNTs in the mixture after centrifuging were all precipitated at the bottom of the glass vial and the upper solvent layer containing the excess polyamines was carefully decanted. The resulting solid (MWNTs-PA) were repeatedly washed with toluene and centrifuged for 4-5 cycles to remove the unreacted polyamines from the surface of the carbon nanotubes and then again washed with THF for 3-4 cycles to ensure complete removal of excess polyamines. The final product (MWNTs-PA) (0.74 g) was obtained by the removal of the solvent under vacuum.

Blends with polyethylene. The LDPE was melt-blended with pristine MWNTs and MWNTs-PA (1-6 wt%) using a 5 cm3 microextruder (DSM research products B.V., The Netherlands) with barrel temperature of 150 °C and a screw speed of 160 rpm for 10 min. The extruder was operated with two corotating conical screws (self-wiping type). The mixed samples were then compressed into the specimens (length 22 mm, width 4.7 mm, thickness 0.75 mm) with dog-bone shape under a pressure of about 8 MPa at 150 °C for 10 min using a hot press. Thin films of the composites were also prepared using hot press in order to study the dispersion of carbon nanotubes in the LDPE with a Zeiss Axiophot microscope equipped with a Plan-Neofluar 20x/0.45 objective.

Characterization. Thermal gravimetric analysis (TGA) was conducted in a nitrogen environment on a PerkinElmer TGA 7 instrument from 20 °C to 800 °C at a heating rate of 10 °C/min. Element analysis of C, H, N were performed with an Euro EA elemental analyzer. X-ray Photoemission Spectroscopy (XPS) measurements were performed with an SSX-100 (Surface Science Instruments) photoemission spectrometer with a monochromatic Al Kα X-ray source (hv=1486.6 eV). A suspension of the MWNTs in ethanol was drop cast on a polycrystalline gold substrate. The base pressure in the spectrometer was 1×10-10 Torr. The energy resolution was set to 1.3 eV to minimize the data acquisition time and the photoelectron take-off angle was 37 °C. Raman spectra, excited by the second harmonic light of a Nd:YVO4 laser (532 nm) were recorded on a micro-Raman spectrometer (T64000 Jobin Yvon) equipped with a liquid nitrogen-cooled

84

Crosslinking of MWNTs

charged coupled device (CCD) detector. Scanning electron microscopy (SEM) was carried out on a JSM-6320 instrument with an accelerating voltage of 2 kV and the samples were sputtered with Pt/Pd prior to SEM observation. Transmission electron microscopy (TEM) micrographs were recorded on a Philips CM 120 electron microscope operating at 120 kV. After 30 min sonication, one drop of the MWNTs suspension in water was placed on a glow discharged carbon-coated grid and followed by the evaporation of the solvent for the TEM examination. Tensile properties of the LDPE/MWNTs composites were tested on an Instron 5565 machine using 0.1 kN power sensor at a crosshead rate of 10 mm/min. The tensile fracture surface of the composites after mechanical testing was examined by SEM.

5.3 Results and discussion

5.3.1 Grafting of polyamines onto MWNTs

The approach to the cross-linking of MWNTs with polyamines is illustrated in Figure 5.1. The used polyamines, which have N-substituted 2,5-pyrrolediyl groups incorporated in the backbone with an amino functional group pendant from the main chain, were synthesized by reacting the polyketones with 1,2-DAP. Due to the steric effect of the 1,2-DAP, only the non-sterically hindered amino group (position 1 in 1,2-DAP) could react with 1,4-di-carbonyl unit of the polyketones and the other one (position 2 in 1,2-DAP) remains intact as functional group. After modification, around 70% of the carbonyl groups of the polyketones have been converted into pyrrole units bearing in a primary amino group in β-position with respect to the nitrogen atom on the pyrrole ring. Carboxylic acid functionalized MWNTs (MWNTs-COOH) were obtained by the oxidation reaction of MWNTs with the concentrated HNO3. It has been reported that the carbon bonds at the open ends and at defect sides of the side walls of the carbon nanotubes could be terminated with carboxylic acid groups after the strong acid treatment.30-32 These functional groups attached to the nanotubes can be converted into the corresponding acyl chlorides (MWNTs-COCl) by treatment with thionyl chloride. The acyl chloride-functions are then susceptible to react with the amino groups of the polyamines (yielding amide linkages) to produce the interlinked MWNTs with polyamines (MWNTs-PA). This is similar, in terms of the used chemistry, to earlier work on the chemical functionalization of carbon nanotubes with aliphatic diamines.17

85

Chapter 5

Figure 5.1 Schematic illustration of the functionalization of MWNTs with the polyamines.

The TGA spectrum of MWNTs-PA (Figure 5.2) exhibited a main weight loss between 200 °C and 450 °C, which can be attributed to thermal degradation of the attached polyamines since it mirrors the weight loss behavior of pure polyamines. Pristine MWNTs show no weight loss up to 800 °C and MWNTs-COOH has a weight loss of approximately 13 %. The noticeable slight increase in weigh loss of MWNTs-PA as compared to MWNTs-COOH and polyamines in the temperature range 100-200 °C is due to the small amount of residual solvent (2-3%) after functionalization. Thus according to this TGA analysis, the amount of the polyamines covalently bonded to the MWNTs was estimated, based on the total weight of MWNTs-PA, to be about 40 wt%.

0 100 200 300 400 500 600 700 8000

102030405060708090

100

Wei

ght l

oss

(%)

Temperature (°C)

MWNTs MWNTs-COOH MWNTs-PA Polyamines

Figure 5.2 TGA analysis of the pristine MWNTs, MWNTs-COOH, MWNTs-PA, and polyamines.

The attachment of the polyamines at the surface of MWNTs-PA was further confirmed by elemental analysis (Table 5.1). No nitrogen could be detected from the pristine MWNTs and small amount of nitrogen (0.3%) was observed for MWNTs-COOH due to the

86

Crosslinking of MWNTs

remaining impurities after the HNO3 oxidation step. Based on the Table 5.1, the nitrogen content (4.6%) of MWNTs-PA can also provide an estimation of the amount of attached polyamine (around 40 wt%), which was comparable to the results of TGA analysis. Furthermore, the increase of hydrogen content of MWNTs-PA (4.91%) in comparison to that of MWNT-COOH (0.48%) may also provide the direct evidence for the grafting of the polyamines. In addition, based on the assumption that all amino groups of the attached polyamines will react with carboxylic acid groups of MWNTs-COOH, the cross-linking density of the functionalized MWNTs according to element analysis can be obtained from the ratio of amine number of polyamine and carboxylic acid number of MWNTs-COOH. As a result, around 60% carboxylic acid groups of MWNTs-COOH have been reacted into amide groups.

Sample C (%) H (%) N (%)

MWNTs 92.33 0.09 0

MWNTs-COOH 81.28 0.48 0.3

MWNTs-PA 75.69 4.91 4.60

Polyamines 70.77 9.27 12.30

Table 5.1 Element analysis of MWNTs, MWNTs-COOH, MWNTs-PA, and polyamines.

XPS was employed to determine the surface composition (in terms of elements present) of the MWNTs before and after functionalization (Figure 5.3). After oxidation with strong acid, it is clear the increase of the intensity for the oxygen at 532 eV in the spectrum of MWNT-COOH compared to that of pristine MWNTs. No nitrogen could be found for pristine MWNTs and MWNT-COOH. However, a signal of nitrogen is observed at 400 eV in the spectrum of MWNTs-PA due to the presence of the polyamines at the surface of MWNTs-PA. From the relative intensities of the photoemission lines, the ratio of carbon, oxygen, and nitrogen for MWNTs-PA is 81%, 12%, and 7%, respectively. The nitrogen content in MWNTs-PA from XPS is consistent with the results from element analysis.

Raman spectroscopy is a widely used tool for studying structure, diameter, and electronic properties of carbon nanotubes.33 As shown in Figure 5.4, the disorder transition mode around 1345 cm-1 (D band) and characteristic tangential stretch mode peak at around 1580 cm-1 (G band) were observed for both pristine MWNTs, MWNTs-COOH, and MWNTs-PA. The D- to G- band intensity ratios (ID/IG) is typically taken as a measure standard of surface defects in carbon nanotubes.34,35 We found that this ID/IG ratio increases from 0.60 for the pristine MWNTs to 0.96 for MWNTs-COOH, which is an indication of the increment in the defects in the nanotube lattice after the strong acid treatment. It is worth noting that the D΄ band at 1617 cm-1 is hardly observed for pristine

87

Chapter 5

MWNTs but becomes more distinguishable for MWNTs-COOH, directly indicating an increase in the number of defects along the carbon nanotube wall after the oxidation.36 The Raman spectrum of MWNTs-PA showed the similar characteristic peaks compared to that of MWNTs-COOH. The ID/IG ratio of MWNTs-PA (0.93) remains virtually unchanged with respect to that of MWNT-COOH (0.96), which indicates that the electronic structure of the MWNTs-COOH are not perturbed by the covalent attachment of the polyamines.37,38

200250300350400450500550600

C

Inte

nsity

(Arb

. Uni

ts)

Binding Energy (eV)

MWNTs

C

C

O MWNT-COOH

MWNTs-PANO

Figure 5.3 X-ray photoemission survey spectra of the pristine MWNTs, MWNTs-COOH, and MWNTs-PA.

1300 1400 1500 1600

MWNTs

Inte

nsity

(Arb

. Uni

t)

Wavenummber (cm-1)

MWNTs-PA

MWNTs-COOH

D',1617

G,1580D,1345

Figure 5.4 Raman spectra of the pristine MWNTs, MWNTs-COOH, and MWNTs-PA.

88

Crosslinking of MWNTs

5.3.2 Solubility

Due to a high density of amino functionalities along the backbone of the polyamines, the latter act as linkers for interconnecting or cross-linking the MWNTs as could be verified by the study of the dispersibility properties of the MWNTs in a variety of solvents (Figure 5.5).

Figure 5.5 (a) Suspensions of MWNTs (1 mg/ml) in various solvent after 30 min sonication. Contents of glass vials, from left to right: 1. pristine MWNTs in water; 2. MWNT-COOH in water; From 3. to 8. MWNT-PA in deionized water, 0.5 M HCl water solution, DMF, THF, toluene, and CHCl3, respectively. The images were taken after overnight storage. (b) Suspensions of MWNTs-PA (1 mg/ml) in water before hydrolysis (left) and after hydrolysis (right). The images were taken after 3 days’ storage.

The pristine MWNTs (Figure 5.5a(1)) are not dispersable in water because of the tendency to assemble into aggregates or ropes due to the van der Waals interaction.5 After strong acid treatment and sonication, the dispersability of MWNTs-COOH (Figure 5.5a(2)) was greatly improved in water and no precipitation was observed at the bottom of the glass vial as reported in the literature39-41. However after reaction with the polyamines, the obtained MWNTs-PA (Figure 5.5a(3-8)) were not dispersable in any of the tested solvents

89

Chapter 5

after 30 min sonication, namely, deionized water, acid water (0.5 M HCl), DMF, THF, toluene, and CHCl3, while the pure polyamines are readily soluble in acidic water and the organic solvents mentioned here. Thus the decrease in dispersability after reaction clearly supports the fact that the polyamines were responsible for the cross-linking within or between the bundles of MWNTs. This is also in analogy with what was reported in previous studies of the use of bi-functional nitrenes as linkers.20 The de-cross-linking of MWNTs-PA by acid-catalyzed hydrolysis reaction was used to confirm the cross-linking of MWNTs-PA via the formation of amide bonding.42 MWNTs-PA were treated in HCl water solution (2 M) under reflux for 3 days and then washed with deionized water until the pH was 7. After sonication for 30 min, MWNTs-PA could partially be redispersed in deionized water as shown in Figure 5.5b. These results provide additional evidence for the amide linkage in the cross-linking process of the MWNTs.

5.3.3 SEM and TEM

The morphology and structure of pristine MWNTs, MWNTs-COOH, and MWNTs-PA was investigated by SEM. MWNTs-PA (micrograph in Figure 5.6c) presents a major change in the morphology and structure in comparison to pristine MWNTs (micrograph in Figure 5.6a) and MWNTs-COOH (micrograph in Figure 5.6b). While pristine MWNTs present a smooth surface and a loosely packed arrangement, the image of MWNTs-COOH shows not only that the length of nanotubes was greatly reduced by the strong acid treatment but also that the oxidized carbon nanotubes are highly tangled with each other and form mat-like structure, called “bucky paper” 30. By contrast, it is clear that the image of MWNTs-PA shows irregular blobs of material wrapped around the surface of the carbon nanotubes, resulting from the covalent attachment of the polyamines. It is quite difficult to identify the single carbon nanotubes and some parts of the tubes are fully coated with polymers. Both randomly entangled structures and compact bundles of carbon nanotubes can be observed for MWNTs-PA. We assign this morphology to the covalent linkage of the bundles and the cross-linking of the MWNTs. Confirmation for this interpretation comes from TEM studies. The image of MWNTs-COOH (Figure 5.7a) shows that the acid-treated nanotubes have a rather smooth and clean surface without any extra phase adhering to and between them. In contrast, TEM pictures of MWNT-PA (Figure 5.7b-e) clearly show an irregular and discontinuous coating of an amorphous polymer layer around and between the carbon nanotubes and the interlinking of carbon nanotubes by the polymers. The polyamines can also act as bridge to interconnect or assemble carbon nanotubes into new configurations, mainly end-to-side (Figure 5.7c), side-to-side (Figure 5.7d) junctions and carbon naotubes bundles (Figure 5.7e), which are potentially useful for the assembly of carbon-nanotube-based electronic devices.17-19

90

Crosslinking of MWNTs

Figure 5.6 SEM images of (a) pristine MWNTs, (b) MWNTs-COOH, and (c) MWNTs-PA. The white bar represents 100 nm.

91

Chapter 5

Figure 5.7 TEM images of (a) MWNTs-COOH and (b)-(e) MWNTs-PA.

92

Crosslinking of MWNTs

5.3.4 Blends with polyethylene

Carbon nanotubes are considered promising reinforcing materials for polymer composites which improve the mechanical, electrical, and thermal properties of polymers.43,44 However, cross-linked nanotubes might in principle display a lower processability as compared to the virgin ones. LDPE/MWNTs-PA composites were prepared by melt- blending of MWNTs-PA into LDPE using a micro-extruder to check if the cross-linked MWNTs which could be further processed into polymers. As shown in Figure 5.8, optical microscopy revealed a similar dispersion behavior of pristine MWNTs and MWNTs-PA in the LDPE matrix. However, big clusters or clumps with the diameter upto 50 μm (black spot in Figure 5.8) of carbon nanotubes can be found in both polymer composites. Nevertheless, a more homogenous distribution of the polyamine-cross-linked nanotubes clusters can be observed (Figure 5.8b). It seems that the mechanical energy in the melt blending process is not enough to overcome the van den Waals interaction between carbon nanotubes, which is also main disadvantage of the use of melt blending process as dispersion techniques.43,44 It is believed that the reinforcing ability of nanotubes will be lower when carbon nanotubes are dispersed poorly into polymer matrix.

Figure 5.8 Optical micrograph of thin films of (a) LDPE/MWNTs composites (1 wt% MWNTs) and (b) LDPE/MWNTs-PA composites (1 wt% MWNTs-PA).

The presence of big clusters of the MWNTs-PA embedded in the LDPE matrix are also confirmed by the SEM micrographs of the fracture surface of the LDPE/MWNTs-PA composites after mechanical testing (left part of Figure 5.9). Failure only occurred within the PE matrix and the carbon nanotubes are difficult to distinguish due to the coating or wrapping of the carbon nanotubes by the LDPE at the surface of the cluster (right part of Figure 5.9). Tensile measurements of the LDPE/MWNTs and LDPE/MWNTs-PA composites presented in Figure 5.10 reveal similar mechanical properties for different loading of carbon nanotubes in the range from 1 wt% to 6 wt%. Compared with the neat

93

Chapter 5

LDPE, tensile strength of both types of composites remained almost unchanged with the increase of MWNTs content and the presence of the carbon nanotubes leads to a slightly increase in Young’s modulus. These results are similar to what is reported in the literature45 for PE/MWNTs composites prepared by the melt blending and indicate that both the processability and reinforcing ability of the carbon nanotubes as filler for the LDPE are retained after cross-linking with the polyamines. However, the enhancement on mechanical property of the LDPE matrix when using the cross-linked MWNTs as fillers is not realized due to poor dispersion of the fillers when using the melt-blending. Other methods, e.g. solution blending or use of surfactants, may be required in order to efficiently disperse MWNTs in a polymer matrix.43-44

Figure 5.9 SEM image of the cluster surface of the LDPE/MWNTs-PA composites (2 wt% MWNTs-PA).

0 1 2 3 4 5 66

7

8

9

10

11

12

320

330

340

350

360

370

380

390

400

410

420

Tens

ile s

tren

gth

(MPa

)

Weight percent (wt %)

MWNTs MWNT-PA

M

odul

us (M

Pa)

Figure 5.10 Mechanical properties of LDPE/MWNTs and LDPE/MWNTs-PA composites at a function of carbon nanotube loading.

94

Crosslinking of MWNTs

5.4 Conclusions

We report a simple and novel approach to cross-link multi-walled carbon nanotubes (MWNTs) by using polyamines synthesized from alternating polyketones. The covalent attachment of the polyamines on the MWNTs is confirmed by TGA, XPS, Raman spectra, and element analysis. It is found that around 40 wt% polyamines are grafted into the surface of the MWNTs. The cross-linked MWNTs display poor solubility in water and several other organic solvents. The polymer bridges within, as well as between the bundles of the MWNTs after cross-linking were clearly observed by SEM and TEM analysis. No detrimental effects on the processability of the nanotubes as well as on the mechanical properties of the composites was found when melt blending the cross-linked MWNTs into a LDPE matrix. The cross-linked MWNTs may find applications in electronic circuits or in reinforcing materials for polymer nanocomposites. In contrast to low molecular weight cross-linking agents (such as aliphatic diamines), the use of a polymeric material might open new processing possibilities for interconnected MWNTs.

5.5 References

1. Baughman, R. H.; Zakhidov, A. A.; de-Heer, W.A. Science 2002, 297, 787. 2. Hirsch, A. Angew. Chem. Int. Ed. 2002, 41, 1853. 3. Banerjee, S.; Kahn, M. G. C.; Wong, S. S. Chem. Eur. J. 2003, 9, 1898. 4. Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chem. Rev. 2006, 106, 1105. 5. Vaisman, L.; Wagner, H. D.; Marom, G. Adv. Colloid. Interface. Sci. 2006, 128-130, 37. 6. Shin, H.; Min, B. G.; Jeong, W.; Park, C. Macromol. Rapid. Commun. 2005, 26, 1451. 7. Sinani, V. A.; Gheith, M. K.; Yaroslavov, A. A.; Rakhnyanskaya, A. A.; Sun, K.;

Mamedov, A. A.; Wicksted, J. P.; Kotov, N. A. J. Am. Chem. Soc. 2005, 127, 3463. 8. Cotiuga, I.; Picchioni, F.; Agarwal, U. S.; Wouters, D.; Loos, J.; Lemstra, P. J.;

Macromol. Rapid. Commun. 2006, 27, 1073. 9. Chen, J.; Hamon, M. A.; Hu, H.; Chen, Y.; Rao, A. M.; Eklund, P. C.; Haddon, R. C.

Science 1998, 282, 95. 10. Hamon, M. A.; Chen, J.; Hu, H.; Chen, Y.; Itkis, M. E.; Rao, A. M.; Eklund, P. C.;

Haddon, R. C. Adv. Mater. 1999, 11, 834. 11. Gabriel, G.; Sauthier, G.; Fraxedas, J.; Moreno-Manas, M.; Martinez, M. T.;

Miravitlles, C.; Casabo, J. Carbon 2006, 44, 1891. 12. Sun, Y. -P.; Fu, K.; Lin, Y.; Huang, W. Acc. Chem. Res. 2002, 35, 1096. 13. Lin, Y.; Rao, A. M.; Sadanadan, B.; Kenik, E. A.; Sun, Y. -P. J. Phys. Chem. B. 2002,

106, 1294.

95

Chapter 5

14. Kong, H.; Gao, C.; Yan, D. Macromolecules 2004, 37, 4022. 15. Baskaran, D.; Dunlap, J. R.; Mays J. W.; Bratcher, M. S. Macromol. Rapid. Commun.

2005, 26, 481. 16. Yao, Z.; Postma, H. W. C.; Balents, L.; Dekker, C. Nature 1999, 402, 273. 17. Chiu, P. W.; Duesberg, G. S.; Dettlaff-Weglikowska, U.; Roth, S. Appl. Phys. Lett.

2002, 80, 3811. 18. Dettlaff-Weglikowska, U.; Benoit, J.; Chiu, P. W.; Graupner, R.; Lebdekin, S.; Roth, S.

Curr. Appl. Phys. 2002, 2, 497. 19. Chiu, P. W.; Kaempgen, M.; Roth, S. Phys. Rev. Lett. 2004, 92, 246802-1. 20. Holzinger, M.; Steinmetz, J.; Samaille, D.; Glerup, M.; Paillet, M.; Bernier, P.; Ley, L.;

Graupner, R. Carbon 2004, 42, 941. 21. Frehill, F.; Vos, J. G.; Benrezzak, S.; Koos, A. A.; Konya, Z.; Ruther, M. G.; Blau, W.

J.; Fonseca, A.; Nagy, J. B.; Biro, L. P.; Minett, A. I.; in het Panhuis, M. J. Am. Chem. Soc. 2002, 124, 13694.

22. Terrones, M.; Terrones, H.; Banhart, F.; Charlier, J. C.; Ajayan, P. M. Science 2000, 288, 1226.

23. Ni, B.; Andrews, R.; Jacques, D.; Qian, D.; Wijesundara, M. B. J.; Choi, Y.; Hanley, L.; Sinnott, S. B. J. Phys. Chem. B. 2001, 105, 12719.

24. Terrones, M.; Banhart, F.; Grobert, N.; Charlier, J. C.; Terrones, H.; Ajayan, P. M. Phys. Rev. Lett. 2002, 89, 075505.

25. Federizzi, R. L.; Moura, C. S.; Amaral, L. J. Phys. Chem. B. 2006, 110, 23215. 26. Kis, A.; Csanyi, G.; Salvetat, J. P.; Lee, T. N.; Couteau, E.; Kulik, A. J.; Benoit, W.;

Brugger, J.; Forro, L. Nat. Mater. 2004, 3, 153. 27. Koos, A. A.; Horvath, Z. E.; Osvath, Z.; Tapaszto, L.; Niesz, K.; Konya, Z.; Kirisci, I.;

Grobert, N.; Ruhle, M.; Biro, L. P. Mater. Sci. Eng. C. 2003, 23, 1007. 28. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F. J. Appl. Polym. Sci. 2008,

107, 262. 29. Drent, E.; Keijsper, J. J. US 5225523, 1993. 30. Liu, J.; Rinzler, A. G.; Dai, H.; Hafner, J. H.; Bradley, R. K.; Boul, P. J.; Lu, A.;

Iverson, T.; Shelimov, K.; Huffman, C. B.; Macias, F. R.; Shon, Y. S.; Lee, T. R.; Colbert, D. T.; Smalley, R. E. Science 1998, 280, 1253.

31. Hu, H.; Bhowmik, P.; Zhao, B.; Hamon, M. A.; Itkis, M. E.; Haddon, R. C.; Chem. Phys. Lett. 2001, 345, 25.

32. Hu, H.; Zhao, B.; Itkis, M. E.; Haddon, R. C. J. Phys. Chem. B. 2003, 107, 13838. 33. Dresselhaus, M. S.; Dresselhaus, G.; Saito, R.; Jorio, A. Phys. Rep. 2005, 409, 47. 34. Jorio, A.; Pimenta, M. A.; Filho, A. G. S.; Saito, R.; Dresselhaus, G.; Dresselhaus, M.

S. New. J. Phys. 2003, 5, 139.

96

Crosslinking of MWNTs

35. Dyke, C. A.; Tour, J. M. J. Phys. Chem. A. 2004, 108, 11151. 36. Gao, C.; Jin, Y. Z.; Kong, H.; Whitby, R. L. D.; Acquah, S. F. A.; Chen, G. Y.; Qian,

H.; Hartschuh, A.; Silva, S. R. P.; Henley, S.; Fearon, P.; Kroto, H. W.; Walton, D. R. M. J. Phys. Chem. B. 2005, 109, 11925.

37. Zhao, B.; Hu, H.; Haddon, R. C. Adv. Funct. Mater. 2004, 14, 71. 38. Yang, Y.; Xie, X.; Wu, J.; Yang, Z.; Wang, X.; Mai, Y. Macromol. Rapid. Commun.

2006, 27, 1695. 39. Esumi, K.; Ishigami, M.; Nakajima, A.; Sawada, K.; Honda, H. Carbon 1996, 34, 279. 40. Jeong J. S.; Jeon, S. Y.; Lee, T. Y.; Park, J. H.; Shin, J. H.; Alegaonkar, P. S.;

Berdinsky, A. S.; Yoo, J. B. Diam. Relat. Mat. 2006, 15, 1839. 41. Tchoul, M. N.; Ford, W. T.; Lolli, G.; Resasco, D. E.; Arepalli, S. Chem. Mater. 2007,

19, 5765. 42. Fu, K.; Huang, W.; Lin, Y.; Riddle, L. A.; Carroll, D. L.; Sun, Y. -P. Nano. Lett. 2001,

1, 439. 43. Moniruzzaman, M.; Winey, K. I. Macromolecules 2006, 39, 5194. 44. Grossiord, N.; Loos, J.; Regev, O.; Koning, C. E. Chem. Mater. 2006, 18, 1089. 45. McNally, T.; Potschke, P.; Halley, P.; Murphy, M.; Martin, D.; Bell, S. E. J.; Brennan,

G. P.; Bein, D.; Lemoine, P.; Quinn, J. P. Polymer 2005, 46, 8222.

97

Chapter 5

98

Biomedical application of polyamines

Chapter 6

Cell behavior on polymeric amines derived from

alternating polyketones

Abstract

Polymeric amines (polyamines) derived from chemical modifications of alternating polyketones were used as thin films to study the response of vascular smooth muscle cells (VSMC) and bovine arterial endothelial cells (BAEC). The physical and chemical properties of the polyamines were characterized at different cross-linking levels by FTIR spectroscopy, elemental analysis, gel content (cross-linking degree), water contact angle, and atomic force microscope (AFM). In this study, we find that polyamines without cross-linking or at low cross-linking levels may induce cell death by apoptosis, this being confirmed by activation of Caspase-3/7 assay and direct visual observation at the microscope. However, polyamines at high cross-linking levels display good biocompability with both VSMC and BAEC (i.e. supporting proper attachment and proliferation). Key word: Polyamines; Polyketones; Apoptosis; Biomaterials

99

Chapter 6

6.1 Introduction

In the past two decades, there has been a growing interest in the use of polymers as biomaterials for application as tissue scaffolds, polymer carriers for drug delivery or medical devices.1-3 Currently, many studies in tissue engineering are driven by a few classes of polymers such as polylactic acid, polyglycolic acid (PGA), chitosan, polyurethane, and polycaprolactone. These polymers are generally required to be non-toxic, biocompatible with living cells, and biodegradable. In more recent years, a promising class of novel biocompatible materials, alternating aliphatic polyketones prepared by using homogeneous palladium catalyst systems, have been reported for potential applications in pharmacological and biological fields.4,5 The remarkable advantage of this class of polymers is that their average molecular weight (and therefore chain length), polarity, crystallinity, mechanical and surface properties can be easily tuned to meet the specific requirements of many diverse biomedical applications. Nontoxic behavior of the polyketones was demonstrated by in vitro tests over a period of 60 days.4 Polyketones have been even utilized as biocompatible scaffolds for primary urothelial cells in vitro and vivo studies.6 In addition, bioactive moieties such as monosaccharide fragments and protected tyrosine groups could be linked to the backbone of the polyketones by palladium catalyzed insertion polymerization.7,8

Polymers containing amino functionality (e.g. polyethylenimine, chitosan, and polylysine) have been used and studied for many biomedical applications, such as drug- and DNA-carriers for gene delivery.9,10 Polyketones can potentially act as excellent precursors for the preparation of polymeric amines (polyamines) by chemical modifications via the Paal-Knorr reaction due to the presence of the highly reactive 1,4-di-carbonyl groups along the backbone.11 Different kinds of polyamines, containing different amino functionality (which can be either primary, secondary, tertiary or aromatic) as side chains, can be prepared via this kind of modifications. The easiness of the chemical modification makes the synthesis of different polyamines (different cross-linking degrees, amount of cationic species, and chemical structure of the backbone) a straightforward and fast task. In the present work, we demonstrate that the polyamines derived from alternating aliphatic polyketones have the ability to induce and moulate apoptosis of cells (i.e. a programmed cell death12-14 that is essential for tissue and organ development, physiologic adaptation, and disease).

In this study, we have investigated the behavior of rat vascular smooth muscle cells (VSMC) and bovine arterial endothelial cells (BAEC) on exposure to the polyamines in vitro. Apoptosis assays were used to analyze the ability of the polyamines to induce apoptosis. A detailed structure-effect relationship was investigated in order to fully

100

Biomedical application of polyamines

understand which characteristics of the polyamines are actually responsible for apoptosis. Since cell adhesion and consequent states are dependent not only on the cell type but also on the physical and chemical properties of the polymers, a detailed characterization of the physical and chemical properties of the polyamine (e.g. wettability, surface roughness, and cross-linking level) was also carried out.

6.2 Experimental

Materials. The alternating polyketones (Mw 3970), ter-polymers of carbon monoxide, ethylene, and propylene were synthesized according to a reported procedure15. 1,2-diaminopropane (Acros, ≥99%) and tetrahydrofuran (THF, Acros, ≥99%) were purchased and used as received.

Preparation of polymeric amines. Polymeric amines (polyamines) were synthesized in bulk by reacting two components (polyketones and 1,2-diaminopropane) in a single one-pot synthesis.11 After reaction, the resulting mixtures were washed several times with deionized Milli-Q water. After filtering and freeze-drying, light brown polymers were obtained as the final products. The polyamines with low (2.7 mmol/g) and high (4.5 mmol/g) degree of amino functionality were prepared here for study, corresponding to 40% and 70% conversion value of carbonyl groups of the polyketones, respectively.

Preparation of polyamine films. Polyamine films were prepared by a solvent-casting method on glass petri dishes (Diameter 40 mm). Polyamines were first dissolved in THF at a concentration of 50 mg/ml. After passing through a 200 nm syringe filter twice, 0.5 ml of polyamine solution was added into the petri dishes, which was then covered with a lid and placed in a fume hood at room temperature overnight for slow evaporation. Cross-linking of the casted films was carried out at 140 ˚C in vacuum oven at the different time intervals.

Characterization of polyamine films. The FTIR spectroscopy was performed using a Perkin-Elmer Spectrum 2000. Thermogravimetric analysis (TGA) was conducted in a nitrogen environment on a Perkin-Elmer TGA 7 instrument from 20 °C to 600 °C at a heating rate of 10 °C/min. Elemental analysis of C, H, N were performed with an Euro EA elemental analyzer. The gel content of the cross-linked polyamines was determined by solvent extraction with tetrahydrofuran (THF). Water contact-angle measurements were carried out at room temperature (20 °C) by the sessile drop method, using a custom-built microscope–goniometer system. A 1.5 μl drop of ultrapure water was placed on a freshly prepared sample using a Hamilton micro-syringe and the contact angle was measured after 30 s. At least five different places on the film surface were measured and all quoted angles are subject to an error of ± 2°. The atomic force microscope (AFM) measurements of morphology of polyamine films were performed in tapping mode using a NanoScope IV

101

Chapter 6

multimode scanning probe microscope from Digital Instruments. The leaching of polyamines was studied on a HP 8453 Spectrophotometer. A phosphate buffer solution of pH 7.4 was added to the coated petri dish with polyamines. The coated petri dish was placed in an oven at 37 °C. After one day, 2 ml solution was withdrawn from the petri dish for UV-VIS spectrophotometer analysis.

Cell culture Studies. Rat vascular smooth muscle cells (VSMC) and bovine arterial endothelial cells (BAEC) were cultured in Dulbecco’s Modified Eagles Medium (DMEM, Life Technologies) supplemented with 10% Fetal calf serum, 2% Penstrep and incubated under 5% CO2 at 37 °C in a humidified incubator in all of the experiments described herein. Cell morphology was evaluated using a phase contrast, confocal laser microscope at 633 nm wavelength (Zeiss Microsystems LSM, Axiovert 135M) at different time intervals.

Cell behavior on polyamine films. Before being seeded with BAEC and VSMC, the polyamine films were first washed with 70% ethanol and three times with culture medium to remove contaminants. The cells were seeded at a density of 800 cells/cm2 (i.e. 10000 cells per dish) for BAEC cells and 1600 cells/cm2 for VSMC cells (i.e. 20000 cells per dish). To assess time-dependent cell viability and cell adhesion, the cells were cultured for various durations as indicated in the result section.

Cell response to polyamine solutions. The stock solution of polyamines (50 mg/ml) after protonation with acetic acid in de-ionized milli-Q water was diluted with culture medium to obtain different concentrations of polyamines, as indicated in the result section. BAEC were seeded into 96-well polystyrene plates (10000 cells per well) and incubated for 48 h at 37 ˚C with a 5 % CO2 atmosphere. The medium was replaced after 48 h with the medium containing the various concentrations of polyamines. After 24 h, the apoptosis was determined by Caspase-3/7 assay (Promega). The positive control for apoptosis was performed by incubation with menadione (45 µM for 1 h). The medium was removed and a 50/50 solution of Caspase 3/7 assay and DMEM was added to the cells. The cells were incubated in the dark for 1 h and the fluorescence was recorded on a Wallac Victor 2 1420 multilabel counter luminometer.

6.3 Results and discussion 6.3.1 Characterization of polyamine films

Polymeric amines (polyamines), which have N-substituted 2,5-pyrrolediyl groups incorporated in the backbone bearing an amino functional group pendant from the main chain, were synthesized by reacting a class of low molecular weight polyketones with 1,2-diaminopropane. The route of the chemical modifications of the polyketones described here just consists of a two component/one-pot reaction without the need of any catalysts,

102

Biomedical application of polyamines

organic solvent or any additives. Furthermore, the reaction can be easily carried out by using mild conditions during the whole process. The TGA analysis performed in nitrogen shows a degradation temperature of the polyamines at around 200 °C. The degree of amino functionality was finely adjusted by varying the initial molar ratio between the diamine and the 1,4-di-carbonyl functions of the polyketones. A maximum of around 70% of carbonyl groups on the polyketone backbone can be actually converted into pyrrole units. This is due to statistical factors (two adjacent carbonyls must react in order to obtain ring formation) and steric hindrance. Here polyamines with low (2.7 mmol/g) and high (4.5 mmol/g) degree of amino functionality were used for study: one with 40% carbonyl conversion (denoted as PA40) and the other with 70% carbonyl conversion (denoted as PA70). Upon heating at high temperature (140 ˚C), the polyamines can be cross-linked by formation of either instable imine bonds16 or stable bis-pyrrole units17 (Figure 6.1). The exact pathway is truly dependent on the degree of amino functionality and in turn on the availability of the 1,4-di-carbonyl functions at the main chain of the polyamines.

N

NH2

O

N

N N

ONH2

N

N

N O

N

NO

N

High amine degree(PA70)

Low amine degree(PA40)

H+

Figure 6.1 Cross-link reactions of the polyamines at high and low degree of amino functionality.

FTIR spectra for PA70 and PA40 are presented in Figure 6.2 before and after cross-linking. The peaks at 1707 cm-1, corresponding to carbonyl group stretching vibration18, as well as the range of 1500-1680 cm-1 due to the skeletal stretching of the pyrrole ring19,20, represent characteristic absorptions for the given system in all the spectra. At high degree of amino functionality (Figure 6.2a), the intensity of the absorption peak at 1707 cm-1 (carbonyl groups) decreases significantly after cross-linking. This, together with the invariance of the pyrrole rings absorption, constitute an indirect evidence of imine formation. With respect to low degree of amino functionality (Figure 6.2b), the increased

103

Chapter 6

intensity of the absorption peak at 1657 cm-1 (stretching vibration of the pyrrole rings) clearly indicates the bis-pyrrole formation upon cross-linking.

1500155016001650170017501800

After crosslinking

Before crosslinking

Abs

orpt

ion

Wavenumber (cm-1)

(a)

1500155016001650170017501800

After crosslinking

Before crosslinking

Wavenumber (cm-1)

(b)

Abs

orpt

ion

Figure 6.2 FTIR spectra of polyamines at (a) high (PA70) and (b) low (PA40) degree of amino functionality before and after cross-linking.

The cross-linking level can be easily tuned as a function of the cross-linking time. Such dependence makes it possible from PA70 to prepare low-cross-linked samples after 2 h (PA70-IX-l), medium-cross-linked ones after 4 h (PA70-IX-m), and highly-cross-linked ones after 8 h (PA70-IX-h). The same holds for PA40 despite the differences in chemical structures of the cross-linked points. For the latter medium-cross-linked samples (PA40-PX-m) and highly-cross-linked ones (PA40-PX-h) were prepared after 4 h and 8 h cross-linking time. The increased cross-linking level of the polyamines with the cross-linking time was further verified by the oxygen content values deduced from the elemental analysis (Figure 6.3). An increase in the cross-linking time leads to a gradual decrease of

104

Biomedical application of polyamines

the oxygen content because of the water formation and release during the cross-linking process. The same trend is confirmed by the gel content measurements (Table 6.1). The gel content of the crosslinked PA70 slightly increases with the cross-linking time and no weight loss is found for the crosslinked PA40 for 4 h and 8 h cross-linking time, indicating a fully-cross-linked structure and stable bis-pyrrole formation.

4

6

8

10

12

14

8 h4 h

0 h

Degree of amino functionality

Oxy

gen

cont

ent (

%)

High (PA70) Low (PA40)

0 h

2 h 4 h8 h

Figure 6.3 Oxygen content of the polyamines at high (PA70) and low (PA40) degree of amino functionality as a function of cross-linking time.

Cross-linked polyamines

Cross-linking Time (h)

Gel content (%)

PA70-IX-l 2 81

PA70-IX-m 4 90

PA70-IX-h 8 96

PA40-PX-m 4 100

PA40-PX-h 8 100

Table 6.1 Gel content of the polyamines at high (PA70) and low (PA40) degree of amino functionality as a function of cross-linking time.

Water contact angle measurements have been commonly used to characterize the relative hydrophilicity or hydrophobicity of polymer surfaces.21 The surface of the unmodified polyketones is relatively hydrophilic with an average contact angle of 56°. After chemical modifications, the contact angle of the polyamine film remains virtually unchanged: 52° for PA70 and 54° for PA40 before cross-linking. However, it is interesting to observe that the surface of the polyamines becomes more hydrophobic after cross-linking (Figure 6.4). The water contact angles of the polyamine surface gradually increase

105

Chapter 6

with cross-linking time from 52° to 100° for PA70 and from 54° to 90° for PA40. Such dramatic change can be explained by the fact that the cross-linking reaction between carbonyl groups and amino groups leads to a reduction in the number of the polar groups (amino and carbonyl) at the polymer surface. This also confirms the results of the elemental analysis and gel content with respect to the number of polar groups as a function of cross-linking time. All these observations indicated that the cross-linking process could effectively modify the surface wettability of the polyamines.

PA70 PA40

0 h 52°

83°

92°

100°

54°

86°

90°

2 h

4 h

8 h

0 h

4 h

8 h

Figure 6.4 Contact angle of the polyamine films at high (PA70) and low (PA40) degree of amino functionality as a function of cross-linking time.

Surface morphology of the polyamines before and after cross-linking was examined with atomic force microscope (AFM) (Figure 6.5). The AFM images indicate that the morphology of the surface of PA70 is rather smooth and flat. Roughness analysis was performed over the entire image and the surface roughness is expressed as the root mean square (RMS) roughness. The surface roughness of the PA70 thin films before cross-

106

Biomedical application of polyamines

linking is around 0.44 nm, which remains almost unchanged with respect to the cross-linking time: 2 h (0.44 nm), 4 h (0.42 nm), and 8 h (0.52 nm). A very similar surface morphology is also observed for the surface of the PA40 thin films before and after cross-linking. Thus it may conclude that the crosslinking process has no effect on the surface roughness of the polyamine films.

(a) (b)

(c) (d)

15 nm

Figure 6.5 AFM images of the polyamine films (PA70) as a function of cross-linking time: (a) 0 h; (b) 2 h; (c) 4 h; (d) 8 h.

6.3.2 Cell culture studies

It has been reported that the use or application of a given biomaterial is closely related to cell behavior upon contact with them and particularly to cell adhesion to the material solid surface.22 Cell adhesion and proliferation on the polyamine films of PA70 with different cross-linking levels have been qualitatively examined by using phase contrast microscopy here. Since physical and chemical properties of the polyamines vary with the cross-linking level, the latter is also found to play an important role in determining the cell behavior. The cell behavior and response to solid films of PA70 at different cross-linking level are summarized in Table 6.2.

107

Chapter 6

PA70-IX-l PA70-IX-m PA70-IX-h Time Event BAEC VSMC BAEC VSMC BAEC VSMC

Attachment + + + + + + Detachment – – – – – –

1 h

Cell death – – – – – – Attachment – +/– +/– + ++ ++ Detachment ++ +/– +/– – – –

1 day

Cell death ++ +/– +/– – – – Attachment – – – – ++ ++ Detachment ++ ++ ++ ++ – –

4

days Cell death ++ ++ ++ ++ – –

Table 6.2 Behavior of BAEC and VSMC on the films of PA70-IX-l, PA70-IX-m, and PA70-IX-h at the culture time of 1 h, 1 day, and 4 days. Qualitative scoring from ++ (= high / very well) to – (= absent / poor) is applied.

In general, BAEC and VSMC should attach and stretch out as to maintain viability. Poor attachment and rounding up of these cells are indicatives for improper cell-material interaction, which may be followed by cell death due to apoptosis.23,24 In 1 h, both BAEC and VSMC started to attach at the surface of PA70 at different cross-linking levels. On day 1, it was observed that BAEC died upon contact with the poorly cross-linked PA70-IX-l. On the medium cross-linked PA70-IX-m, they attached, but did not stretch out over the surface. On the highly cross-linked PA70-IX-h, attachment and subsequent morphological change, i.e. stretching out over the surface and developing the typical shape of BAEC, was good (Figure 6.6). With respect to VSMC, better tolerance to apoptosis effect of the polyamines can be observed than for BAEC: some cells were still attached at the film surface and assumed normal VSMC morphology whilst stretching over the surface on day 1. On day 4, no survived BAEC and VSMC were detected at the surface of PA70-IX-l and PA70-IX-m. However, PA70-IX-h supports both BAEC and VSMC after 4 days. Both BAEC and VSMC were well-attached and stretched out over the surface of PA70-IX-h to obtain proper morphology (Figure 6.7). Colonies of VSMC cells were even observed at the surface of PA70-IX-h (Figure 6.7b) on day 4. As a general observation, we must make note here that VSMC attached better and proliferated faster than BAEC. The observed behavior might in any case be explained by assuming that the released polyamines (from the surface of the films) is responsible for the apoptotic effect due to the non-stable imine crosslinking bond. By using UV-Vis spectroscopy, the absorbance at the wavelength of 350 nm, ascribed to pyrrole rings of the polyamines was detected in phosphate buffer at pH of 7.4 at 37 ˚C, supporting the probability of the release of the polyamines into the cell culture

108

Biomedical application of polyamines

medium at a function of culture time. As a consequence, we might conclude that cross-linking density (a structural factor) of the films is essential in determining cell survival and a high cross-linking level can lead to the disappearance of the apoptotic effect.

(a1) (a2)

(b3) (b1) (b2)

Figure 6.6 On day 1, Morphology of BAEC on the films of (a1) PA70-IX-M, (a2) PA70-IX-h and VSMC on the films of (b1) PA70-IX-l, (b2) PA70-IX-m, and (b3) PA70-IX-h at the magnification (10×).

(a) (b)

Figure 6.7 On day 4, morphology of (a) BAEC at the magnification (10×) and (b) VSMC at the magnification (10×) on the films of PA70-IX-h.

To further confirm that polyamines can induce cell death by apoptosis, the effect of a wide concentration range of the PA70 solutions on BAEC was studied to determine the cellular response by microscopical examination. Depending on the dose of PA70, substantial changes in cell morphology were detected upon exposure to polyamines (Figure 6.8). Upon treatment with increasing doses of polyamines, BAEC turned into bubbled spheres and partially detached, which is suggestive for an apoptotic death.23,24 At the

109

Chapter 6

highest dose of 250 μg/ml of PA70, cells perished. Cell death was either observed as a loss of cell morphology followed by detachment, or an appearance of cytoplasm-free cell remainders at the bottom of the wells, presumably representing bare cytoskeletons. Thus, it seems that at this very high dose necrosis (uncontrolled cell death) becomes the dominant death mechanism.

(a) (b)

(c)

(d)

(e)

Figure 6.8 Morphology of BAEC after treatment with PA70 after 4 h at different concentration: (a) 2.5 μg /ml; (b) 7.5 μg /ml; (c) 25 μg /ml; (d) 75 μg /ml; (e) 250 μg /ml.

110

Biomedical application of polyamines

The activation of the Caspase-3/7 at different doses of PA70 was measured (Figure 6.9) in order to further pinpoint the concentration at which apoptosis occurs. The results show that the cells died from apoptosis already at a concentration of 2.5 μg/ml. At 25 μg/ml of PA70, activation of Caspase-3/7 was substantially less than at lower doses (2.5 to 12.5 μg/ml), reaching the level of the negative control experiment. However, microscopical examination of these cultures showed that cell death occurred rapidly at 25 μg/ml of PA70. Thus, it can be concluded that at or below doses of 12.5 μg/ml of PA70, cell death involves apoptosis, while at 25 μg/ml of PA70 causes necrosis.

0

10000

20000

30000

40000

50000

60000

2.5 5pos. 7.5 12.5

Lum

. Val

ue

Concentration (µg/ml)25neg.

Figure 6.9 Caspase activation of BAEC at a effect of dose concentration of PA70.

Normally, apoptosis can be induced in various ways, such as a loss of intracellular water, chromatin condensation, internucleosomal DNA fragmentation, mitochondrial swelling, interaction with cell membrane receptors.25,26 The cause of apoptosis in this study may be correlated with the primary amino groups and in turn by the presence of positive charges on the polyamines. It has been well established that natural low-molecular aliphatic amines (e.g. putrescine, spermidine, and spermine) and many of their structural analogues are involved in the apoptosis process of cells and thus utilized as tools for apoptosis-based cancer therapies.27-29 However, the exact mechanism of cytotoxicity here must be further investigated to check whether cytotoxic effects are mainly mediated by interactions of the polycations with cell membranes or by cellular uptake and subsequent activation of the intracellular signal transduction pathway here.30 The observation for apoptosis effects of the polyamines may can find its application in the design of drugs or medical implants that require cytostatic properties.31 A most illustrative example is the drug-eluting stent (an implant designed to revascularize obstructed arteries), which is currently covered with cytostatic drugs to prevent hyperplastic renarrowing of the vessel lumen.32,33

111

Chapter 6

VSMC BAEC

day 9

day 3

day 1 day 1

day 3

day 9

Figure 6.10 Morphology of BAEC and VSMC on the films of PA40-PX-h on day 1, day 3, and day 9 at the magnification (10×).

Day 9

On the other hand, the study of cell behavior on the films of PA40 after cross-linking demonstrated good biocompatibility with both BAEC and VSMC. The apoptotic effect completely disappeared for both PA40-PX-m and PA40-PX-h due to a non-reversible and stable cross-linking bond (Bispyrrole). Similar cell behavior and morphology were observed for PA40-PX-m and PA40-PX-h. The morphology of BEAC and VSMC on the surface of PA40-PX-h (Figure 6.10) with respect to different culture time was shown as

112

Biomedical application of polyamines

example here. Both types of cells were well-attached at the surface of PA40-PX-h. The increasing number of both type of cells over time demonstrated the occurrence of proper proliferation at the surface of polymer films. After 9 days of culture time, both types of cells had grown into confluent layers with a considerably higher cell population density than that of samples on day 3. The morphology of cells grown onto the surface of PA40-PX-h was found to be comparable with that of cells on polystyrene (cells on polyamine films and on polystyrene platforms both reached confluency at day 9). The results for good biocompatibility of PA40 after cross-linking may be utilized to enhance cell adhesion and tissue integration for tissue scaffold application.

6.4 Conclusions In the present study, a new type of polyamines, synthesized from the chemical modifications of alternating aliphatic polyketones, was used as biomaterial to study the behavior of BAEC and VSMC. Depending on the degree of the attached amino functionality, the polyamines can be cross-linked either by imine or bis-pyrrole formation. The water contact angle gradually increases with the cross-linking level, thus indicating that the material becomes more hydrophobic during the cross-linking process. In addition, a flat, smooth nature of the polyamine films can be obtained upon solvent casting before and after cross-linking. Based on Caspase-3/7 assay and the study of cell morphology at the surface of the polyamine films, it was found that cell behavior (e.g. apoptosis, growth, and proliferation) can be finely tuned by adjusting the cross-linking degree/time of the polyamines. The observed cell behavior with respect to the polyamines (i.e. low cross-linking level induce cell apoptosis and high crosslinking level support cell growth and proliferation) may open many special biomedical applications, e.g. using them as platforms to local delivery of drugs (such as drug eluting stent) or medical implants that require good biocompability. The present study warrants extensive studies with respect to biocompatibility of the polyamines that are specifically modified to serve these respective goals.

6.5 References

1. Cheung, H.; Lau, K.; Lu, T.; Hui, D. Compos. Pt. B-Eng. 2007, 38, 291. 2. Sokolsky-Papkov, M.; Agashi, K.; Olaye, A.; Shakesheff, K.; Domb, A. J. Adv. Drug

Deliv. Rev. 2007, 59, 187. 3. Jagur-Grodzinski, J. Polym. Adv. Technol. 2006, 17, 395. 4. Reuter, P.; Fuhrmann, R.; Mucke, A.; Voegele, J.; Rieger, B.; Franke, R. P. Macromol.

Biosci. 2003, 3, 123.

113

Chapter 6

5. Rohlke, W.; Fuhrmann, R.; Franke, R. P.; Mucke, A.; Voegele, J.; Rieger, B. Macromol. Biosci. 2003, 3, 131.

6. Bartsch, G. C.; Malinova, V.; Volkmer, B. E.; Hautmann, R. E.; Rieger, B. BJU Int. 2006, 99, 447.

7. Malinova, V.; Rieger, B. Macromol. Rapid. Commun. 2005, 26, 945. 8. Malinova, V.; Rieger, B. Biomacromolecules 2006, 7, 2931. 9. Schmaljohann, D. Adv. Drug Deliv. Rev. 2006, 58, 1655. 10. De-Smedt, S. C.; Demeester, J.; Hennink, W. E. Pharmacol. Res. 2000, 17, 113. 11. Zhang, Y.; Broekhuis, A. A.; Stuart, M. C. A.; Picchioni, F. J. Appl. Polym. Sci. 2008,

107, 262. 12. Thompson, C. B. Science 1995, 267, 1456. 13. Nagata, S. Cell 1997, 88, 355. 14. Elmore, S. Toxicol. Pathol. 2007, 35, 495. 15. Drent, E.; Keijsper, J. J. US 5225523, 1993. 16. Lu, S.; Paton, R. M.; Green, M. J.; Lucy, A. R. Eur. Polym. J. 1996, 32, 1285. 17. Zhang, Y.; Broekhuis, A. A.; Picchioni, F. J. Appl. Polym. Sci. 2007, 106, 3237. 18. Lai, T. W.; Sen, A. Organometallics, 1984, 3, 866. 19. Smela, E.; Zuccarello, G.; Kariis, H.; Liedberg, B. Langmuir 1998, 14, 2970. 20. Mahmud, H. N. M. E.; Kassim, A.; Zainal, Z.; Yunus, W. M. M. J. Appl. Polym. Sci.

2006, 100, 4107. 21. Goddard, J. M.; Hotchkiss, J. H. Prog. Polym. Sci. 2007, 32, 698. 22. Anselme, K. Biomaterials 2000, 21, 667. 23. Burridge, K.; Fath, K.; Kelly, T.; Nuckolls, G.; Turner, C. Annu. Rev. Cell Biol. 1988,

4, 487. 24. Hacker, G. Cell Tissue Res. 2000, 301, 5. 25. Robertson, J. D.; Orrenius, S.; Zhivotovsky, B. J. Struct. Biol. 2000, 129, 346. 26. Huerta, S.; Goulet, E. J.; Huerta-Yepez, S.; Livingston, E. H. J. Surg. Res. 2007, 139,

143. 27. Wallace, H. M.; Fraser, A. V.; Hughes, A. Biochem. J. 2003, 376, 1. 28. Wallace, H. M.; Fraser A. V. Biochem. Soc. Trans. 2003, 31, 393. 29. Schmitt, C. T.; Lowe, S. W. J. Pathol. 1999, 187, 127. 30. Fischer, D.; Li, Y.; Ahlemeyer, B.; Krieglstein, J.; Kissel, T. Biomaterials 2003, 24,

1121. 31. Kabanov, A. V. Adv. Drug Deliv. Rev. 2006, 58, 1597. 32. Burt. H. M.; Hunter W. L. Adv. Drug Deliv. Rev. 2006, 58, 350. 33. Coolong, A.; Kuntz, R. E. Am. J. Cardiol. 2007, 100, 17k.

114

Summary

Summary Alternating aliphatic polyketones, produced by co- and terpolymerization of carbon monoxide and olefins (mixtures of ethylene and propylene) using palladium-based homogeneous catalysis represent a very promising class of polymers for a wide range of applications. This research focused on chemical product development of low molecular weight, alternating aliphatic polyketones (Mw 1500-5500) (Figure 1).

O

O

O

O

OR

R

R

R

CO + CH2CH2 CHCH3CH2Pd based Catalyst

+

R= H or CH3

Figure 1 Synthesis of CO-ethylene-propylene based low molecular weight polyketones.

Besides many interesting chemical and physical properties, alternating polyketones can act as excellent precursors for the preparation of functional polymers by chemical modifications owing to the presence of highly reactive carbonyl groups along the backbone. The most interesting reaction route is the Paal-Knorr reaction (i.e. the 1,4-di-carbonyl moiety of the polyketones reacts with a primary amine function yielding a pyrrole unit), since this reaction can be accomplished at mild conditions without the need of any catalysts and organic solvent (Chapter 2 and Chapter 4). By using this route, a great variety of functional groups can be attached to the backbone of the polyketones (Figure 2). This could open up pathways to new chemical products and lead to the discovery and development of new applications for alternating polyketones. In this thesis, the work focused primarily on amine- and furan-based polyketone derivatives.

O O

O

O

+

N

N O

=

NH2

NH2

NH2

NH2

NNH2

NNH2

N

NH2

R

R

O

R NH2

O

NH2

R NH2

Figure 2 Preparation of functional polymers from polyketones via Paal-Knorr reaction.

115

Summary

The current market of wood adhesives is dominated by urea-formaldehyde or phenol-formaldehyde based thermosetting resins, which are known to be hazardous substances as formaldehyde is classified as a suspected carcinogenic chemical. In Chapter 3, water-based polyketone emulsions have been developed for the application as formaldehyde-free and environmental-friendly wood adhesives using a simple and cost-effective route. Polymeric amines derived from the polyketones were applied as polymeric surfactants for the self-emulsification of polyketones. All processing steps, including surfactant preparation, surfactant protonation, and resin emulsification can be performed in single one-pot reaction. Polyketone-based wood adhesives do not release any volatile toxic components during the manufacture nor during the application and end-use cycle. The byproduct obtained during the manufacturing and curing steps is only water. The prepared wood emulsions exhibit extremely long shelf-life at room temperature and were found to be stable and effective as wood adhesive even after a period of 2 years. The achieved submicron particle size (less than 500 nm) of the emulsions remained basically unchanged in time, thus demonstrating high kinetic stability of the emulsions. Cryo-SEM studies revealed uniform and spherical microstructures of the emulsion particles. The viscosity of the system decreases significantly in the first week which may be due to a rearrangement of the polymers at the surface of each particle (i.e. a kind of “arms”-free polymer chains retraction onto the surface of the polymer particle, causing the viscosity of the adhesive to level off quickly to a value less than 1 Pa s) and then remains constant over the storage time of 2 years. According to the European Standard (EN-314) for wood glue testing, the quality of the emulsions as adhesive was evaluated by measuring the shear strengths on the applied maple substrates (hard wood). Average shear strengths of 2.7 MPa could be achieved for both the fresh emulsions and the ones that were stored for 2 years. This by far exceeds the strength requirement of 1 MPa according to the EN-314 and the minimum shelf-life of some days for the currently applied commercial adhesives.

Self-healing polymeric materials have the capability to repair or recover themselves when suffering mechanical and/or thermal induced damage, which can occur autonomously or be activated by external stimuli (e.g. heat) for once or multiple times. An easy-accessible and highly-efficient self-healing system was reported in Chapter 4. The system is based on the Diels-Alder (DA) and Retro-Diels-Alder (RDA) reaction applied to furan-functionalized polyketones (PK-furan) and bis-maleimide. PK-furan can easily be obtained under mild conditions by the Paal-Knorr reaction of alternating polyketones with furfurylamine. The degree of furan functionality can be well adjusted by varying the molar ratio of the 1,4-di-carbonyl functionality of the polyketones and furfurylamine. A highly cross-linked polymeric network can be achieved by the DA reaction of PK-furan with bis-maleimide. The final properties of this network can be tailored by tuning the structure of

116

Summary

the polyketones, the degree of furan functionalization, and the level of cross-linking (i.e. the ratio of furan-functionality to bis-maleimide). NMR, FTIR spectroscopy, and thermal analysis demonstrated ultra-fast kinetics of the present system (gel formation in 2 h at 50 °C and its reversal in 5 min at 150 °C) and 100% or complete thermal reversibility. Self-healing ability was evaluated by using dynamic mechanical analysis (DMA) and 3-point-bending mechanical testing. DMA shows that dynamic mechanical properties of the cross-linked PK-furan remain almost unchanged after 6 repetitive heating-cooling cycles. The thermal remendability of the cross-linked polymers was demonstrated by 3-point-bending testing which showed complete recovery in fracture loading, while the remending process could be repeated multiple times without any loss in mechanical properties. The simplicity of the synthesis and the striking healing ability of this system open the pathway to 100% recyclability and reworkabilty of thermoset materials.

Functionalization of carbon nanotubes via non-covalent or covalent attachment of chemical groups is essential to facilitate their manipulation for many further applications. Chapter 5 reported a new and simple approach to functionalize multi-walled carbon nanotubes (MWNTs) by applying an amidation-type grafting reaction with amino-functionalized alternating polyketones (polyamines). It is found that polyamines can act as cross-linking agents to interlink or cross-link individual carbon nanotubes to form carbon nanotube junctions or complex networks, which may be applicable in nano-scale electronic circuits. The covalent attachment of polyamines (around 40 wt% based on the total weight of the MWNTs) was testified by using thermogravimetric analysis (TGA), X-ray photoemission spectroscopy (XPS), elemental analysis, and Raman spectroscopy. In contrast to good solubility of carboxylic acid-functionalized MWNTs in water, the functionalized MWNTs after cross-linking are insoluble in any solvent (e.g. water, toluene, and chloroform). After functionalization, the polymer bridges within, as well as between the bundles of the MWNTs and the interconnects of the MWNTs (end-to-side or end-to-side) were observed by using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The processability of the cross-linked MWNTs was also studied by melt blending of cross-linked MWNTs into low density polyethylenes (LDPE). The resulting composites show similar dispersibility and comparable mechanical properties to those obtained by simple blending of “un-cross-linked” carbon nanotubes with LDPE.

Polymers containing amino functionality (e.g. polyethylenimine, chitosan, and polylysine) have been used as tissue scaffolds, polymeric carriers for drug and DNA delivery, medical devices, etc. Polyamines, derived from the chemical modifications of alternating polyketones, can be considered as a very promising class of polymers for biomedical applications (Chapter 6). Polyamines can be cross-linked at a high temperature

117

Summary

(140 ˚C) by the formation of either imine bonds or bis-pyrrole units, depending on the amount of amino groups grafted onto the polymer backbone. The cross-linking degree can be fine-tuned as a function of the cross-linking time. The surface wettability of polyamine films was found to become more hydrophobic with increasing cross-linking degree. Atomic force microscope analysis indicated that smooth and flat surfaces with the roughness less than 0.5 nm can be achieved for polyamine films before and after cross-linking. Interesting cell behavior of rat vascular smooth muscle cells (VSMC) and bovine arterial endothelial cells (BAEC) were observed on exposure to polyamine films and polyamine solution in vitro study. It was found that polyamines without cross-linking or at low cross-linking levels may induce apoptosis, i.e. a programmed cell death that is essential for tissue and organ development, physiologic adaptation, and disease. This was confirmed by activation of Caspase-3/7 assay and direct visual observation at the microscope. This valuable property of the polyamines can find application in the design of drugs or medical implants that require cytostatic properties. On the other hand, polyamines at high cross-linking levels display good biocompability with both VSMC and BAEC, which can be utilized to enhance cell adhesion and tissue integration for tissue scaffolds.

118

Samenvatting

Samenvatting Alternerende alifatische polyketonen, geproduceerd door de co- en ter-polymerisatie van koolstof monoxide en olefinen (bijvoorbeeld mengsels van ethyleen en propyleen) met behulp van op palladium gebaseerde homogene katalyse, vertegenwoordigen een zeer veelbelovende klasse van polymeren voor een scala aan toepassingen. In dit onderzoek lag de nadruk op de ontwikkeling van nieuwe producten op basis van laag molecuul gewicht polyketonen (Mw 1500-5500) (Figuur 1).

O

O

O

O

OR

R

R

R

CO + CH2CH2 CHCH3CH2Pd based Catalyst

+

R= H or CH3

Figuur 1 Synthese van laag molecuul gewicht CO-ethyleen-propyleen polyketonen.

Naast vele interessante chemische en fysische eigenschappen van niet-gemodificeerde polyketonen, kunnen deze ook worden omgezet in functionele polymeren door chemische modificaties. Dit met name door de aanwezigheid van de zeer reactieve carbonyl groepen in de polymeerketens. De meest interessante modificatie route is de Paal-Knorr reactie, d.w.z. de reactie van de 1,4-di-carbonyl groep van de polyketonen met een primaire amino groep tot een pyrole ring. Deze reactie kan worden uitgevoerd onder milde condities, zonder de aanwezigheid van katalysator en organisch oplosmiddel (Hoofdstuk 2 en Hoofdstuk 4). Door deze syntheseroute te gebruiken, kunnen allerlei verschillende functionele groepen aan de polyketonen worden toegevoegd (Figuur 2).

O O

O

O

+

N

N O

=

NH2

NH2

NH2

NH2

NNH2

NNH2

N

NH2

R

R

O

R NH2

O

NH2

R NH2

Figuur 2 Synthese van functionele polymeren uit polyketonen via de Paal-Knorr reactie.

119

Samenvatting

Door de reactie uit te voeren met gefunctionaliseerde primaire amines kunnen nieuwe routes ontwikkeld worden voor de synthese van chemische producten die op een andere manier moeilijk toegankelijk zijn. Dit biedt tevens de mogelijkheid om nieuwe applicaties voor alternerende polyketonen te ontwikkelen. In dit proefschrift lag de nadruk op de synthese en het gebruik van primaire amino en furan gefunctionaliseerde polyketon derivaten.

De huidige markt voor houtlijmen wordt gedomineerd door ureum-formaldehyde en fenol-formaldehyde thermoset harsen, die bekend staan als gevaarlijke stoffen, omdat formaldehyde wordt verondersteld carcinogene eigenschappen te hebben. In Hoofdstuk 3 zijn op water gebaseerde polyketon emulsies ontwikkeld voor het gebruik als formaldehyde-vrije en milieuvriendelijke houtlijmen via een simpele en kosteneffectieve route. Amino-gefunctionaliseerde polymeren afgeleid van polyketonen zijn gebruikt als surfactanten voor de zelf emulsificatie van polyketon. Alle stappen in deze synthese, zoals de bereiding van het surfactant, de protonering van het surfactant en de hars emulsificatie kunnen in een één-pot reactie worden uitgevoerd. De productie en eindtoepassing van deze polyketon gebaseerde houtlijmen is een milieu-vriendelijk proces. Het enige bijproduct dat bij de synthese en de uitharding vrijkomt is water. De houtlijm emulsies tonen een extreem lange houdbaarheid bij kamertemperatuur en zijn zelfs na twee jaar nog steeds stabiel en effectief voor het gebruik als houtlijm. De deeltjesgroottes (minder dan 500 nm) van de emulsies bleven onveranderd over deze periode, wat de hoge kinetische stabiliteit van de emulsies aantoont. Cryo-SEM studies toonden gelijkvormige en bolvormige microstructuren van de emulsiedeeltjes. De viscositeit van de emulsie naam significant af in de eerste week na productie. Dit waarschijnlijk door herschikking van de polymeren aan het oppervlak van elk deeltje, d.w.z. de uitgestrekte ketens op een polymeer deeltje zullen na een bepaalde tijd heroriënteren op het oppervlak van het polymeerdeeltje. Dit zorgt voor een snelle reductie van de viscositeit van de hars naar een waarde van minder dan 1 Pa·s, welke gedurende de 2 jaar opslagtijd constant bleef. De kwaliteit van de emulsies voor houtlijm toepassingen werd getest volgens de Europese Standaard (EN-314) door de schuifspanning te meten op de toegepaste esdoorn substraten (hardhout). De gemiddelde schuifspanning van 2.7 MPa kon met zowel de verse emulsies als met de emulsies die 2 jaar werden opgeslagen bereikt worden. Dit overschrijdt veruit de sterktevereiste van 1 MPa volgens EN-314 en de minimum opslagtijd van een paar dagen voor de momenteel toegepaste commerciële harsen.

Zelfherstellende polymeren hebben de capaciteit om zich te herstellen bij mechanische en/of thermisch veroorzaakte schade. Dit zelfherstellende mechanisme kan zelfstandig of door externe stimuli (b.v. door verhitting) voor een keer of meerdere keren geactiveerd

120

Samenvatting

worden. Een zeer toegankelijk en zeer efficiënt zelfherstellend mechanisme wordt in Hoofdstuk 4 beschreven. Dit mechanisme is gebaseerd op Diels-Alder (DA) en Retro-Diels-Alder (RDA) reacties toegepast op polyketonen met furan functionele groepen (PK-furan) en bis-maleimide. PK-furan kan simpel worden gesynthetiseerd onder milde condities via de Paal-Knorr reactie van alternerende polyketonen met furfurylamine. De furan conversie kan goed worden gestuurd door de molaire verhouding 1,4-di-carbonyl functionaliteit van de polyketonen en furfurylamine te variëren. Door middel van de DA reactie van PK-furan met bis-maleimide kan een polymeer netwerk worden verkregen met een hoge vernettingsgraad. . De definitieve eigenschappen van dit netwerk kunnen worden aangepast door de structuur van de polyketonen, de graad van furan functionalisering, en het vernettingsniveau (d.w.z. de verhouding furan : bis-maleimide) te veranderen. NMR, FTIR spectroscopie en thermische analyse tonen een ultrasnelle kinetiek van het huidige systeem aan, d.w.z. gel vorming in 2 uur bij 50 °C en reversibiliteit in 5 minuten bij 150 °C en volledige thermische reversibiliteit. De zelfherstellende capaciteit werd geëvalueerd doormiddel van dynamisch mechanische analyse (DMA) en de evaluatie van mechanische eigenschappen in een 3-punts buigtest. DMA toont verder aan dat de dynamische mechanische eigenschappen van het vernette PK-furan na 6 repeterende verwarm- en koelcycli bijna onveranderd blijven. Ook de thermische herstelbaarheid van de vernette polymeren werd aangetoond middels de 3-punts buigtest, waarbij compleet herstel van de belasting bij breuk werd waargenomen. Dit herstelproces kon meerdere keren worden herhaald zonder enig verlies van mechanische eigenschappen. De eenvoud van de synthese en de opvallende herstellende capaciteit van dit systeem opent de weg voor 100% recycleerbaarheid en herbewerking van thermoset gebaseerde materialen.

Het functionaliseren van koolstof nanotubes via niet-covalente of covalente binding van chemische groepen is essentieel om hun manipulatie voor vele verdere toepassingen te vergemakkelijken. In Hoofdstuk 5 wordt een nieuwe en eenvoudige benadering beschreven om multi-walled koolstof nanotubes (MWNTs) te functionaliseren door een amidatie grafting reactie uit te voeren met amino gefunctionaliseerde polyketonen (polyaminen). Er werd aangetoond dat polyamines gebruikt kan worden voor het vernetten van individuele koolstof nanotubes in nanotube formuleringen die op nano-schaal in elektronische schakelingen toepasbaar kunnen zijn. De covalente binding van polyamines (rond 40 gew.% op het totale gewicht van de MWNTs) werd bevestigd met behulp van thermografiemetrische analyse (TGA), röntgenstraal foto-emissie spectroscopie (XPS), elementaire analyse en Raman spectroscopie. De vernette MWNTs zijn in geen enkel oplosmiddel (b.v. water, tolueen, chloroform) oplosbaar, dit in tegenstelling tot MWNTs met carboxyzuren als functionele groepen. Na het aanbrengen van de functionele groepen zijn de polymeerbruggen tussen de MWNT bundels en de individuele koolstof nanotubes

121

Samenvatting

(eind-aan-zij of zij-aan-zij) aangetoond met behulp van scanning elektronenmicroscopie (SEM) en transmissie elektronenmicroscopie (TEM). De bewerkbaarheid van vernette MWNTs werd ook bestudeerd door deze met laag molecuul gewicht polyethyleen (LDPE) via een smeltproces te mengen. De resulterende composieten tonen gelijke dispergeerbaarheid en vergelijkbare mechanische eigenschappen als die van LDPE mengsels geformuleerd met onbehandelde (niet-vernette) koolstof nanotubes.

Polymeren met amino functionaliteit (bijv. polyethyleenimine, chitosan, polylysine) zijn toegepast als dragermateriaal voor weefselkweek, als polymeerdragers voor medicijn en voor DNA toediening. Analoog kunnen polyamines gesynthetiseerd uit polyketonen als een zeer veelbelovende klasse van polymeren voor biomedische toepassingen (Hoofdstuk 6) worden beschouwd. Polyamines kunnen worden vernet bij hoge temperatuur (140 ˚C) onder vorming van imine bindingen of bis-pyrole groepen. De mate van vernetting kan als functie van de reactietijd worden afgestemd. De bevochtiging van het oppervlak van de polyamine films werd steeds hoger naarmate de vernettingstijd hoger werd. Atomic force microscopie wees uit dat polyamine films gemaakt kunnen worden met gladde en vlakke oppervlakken met een verschil in ruwheid van minder dan 0.5 nm voor en na de vernettingsreactie Blootstelling van vasculaire spier cellen (VSMC) van ratten en runder slagaderlijke endothelial cellen (BAEC) aan polyamine oplossingen en polyamine films in een in vitro studie tonen interessant celgedrag. Polyamines zonder vernetting of met een lage vernettingsgraad kunnen apoptosis veroorzaken. Apoptosis is een geprogrammeerde celdood die essentieel is voor weefsel en orgaan ontwikkeling, fysiologische aanpassing, en ziektes. Dit werd bevestigd door activering van Caspase-3/7 assay en directe visuele observatie met de microscoop. Deze belangrijke eigenschap van polyamines kan gebruikt worden voor het ontwikkelen van medicijnen of medische implantaten die cytostatische eigenschappen vereisen. Polyamines met een hoge mate van vernetting tonen goede bio-comptabiliteit met zowel VSMC als BAEC en kunnen mogelijk worden gebruikt als dragers voor celadhesie en weefselintegratie.

122

Acknowledgements

Acknowledgements

Finally, I have completed my thesis. I would like to thank all those people who made this possible and a thoroughly enjoyable experience for me during the last four years.

First of all, I would like to express my deep gratitude to Prof. A.A. Broekhuis, my promotor, for his continuous encouragement, guideline, and valuable support. He was always accessible, whenever I was in need. I appreciate very much his scientific insight and creative thinking to lead me to push the research works forward.

I am deeply grateful to Prof. F. Picchioni, my copromotor, who has been always there to listen and give valuable advices. His inspiration, enthusiasm, and great efforts contributed towards the completion of this thesis. Working with him was a great pleasure to me.

I would also like to thank the members of my reading committee, Prof. F. Ciardelli, Prof. J.A. Loontjens, and Prof. H.J. Heeres for their precious time, valuable comments and suggestions on my thesis.

My sincere thanks go to those whose have helped me during my lab and experimental works. I thank Anne Appeldoorn, Marcel de Vries, Erwin Wilbers, Laurens Bosgra, and Jan Henk Marsman for their various technical supports and helps. I am very thankful to Dr. Anton J.M. Roks for cell culture studies. I thank Dr. Marc C.A. Stuart for the TEM analysis. I thank Harry Nijland for the SEM analysis. I also thank Geert Alberda for many helps in the polymer analysis.

I would like to thank all my wonderful colleagues in the Chemical Engineering Department: Theo Jurriens, Sameer Nalawade, Leon Janssen, Francesca Fallani, Ignacio Melian Cabrera, Francesca Gambardella, Marya van der Duin-de Jonge, Nidal Hammound Hassan, Henky Muljana, Arjan Kloekhorst, Mochamad Chalid, Claudio Toncelli, Asal Hamarneh, Gerard Kraai, Asaf Sugih, Jaap Bosma, Fesia Lestari Laksmana, Boelo Schuur, Jelle Wildschut, Hans Heeres, Marcel Wiegman, C.B. Rasrendra, Buana Girisuta, Judy Retti Witono, Henk van de Bovenkamp, Camiel Janssen, Oscar Rojas, Teddy, Anant Samdani, Laura Justinia, Louis Daniel, Agnes Ardiyanti, Erna Subroto, Yongki Arianto, Poppy Sutanto for their helps and friendships. Special thanks to Nidal Hammound Hassan for the translation of my English Summary into Dutch. All of you made my life and research works easier and more pleasurable.

123

Acknowledgements

I would like to thank all my friends in Groningen: Yu Wu, Ji Liu, Shaozhong Ge, Jie Yao, Jianghua Ou, Xiaomei Wang, Junjun Shan, Dagang Gao, Yan Zhao, Huanjun Yu, Yuan Zhang, Xiaonan Sun, Fei Xiang…Thanks for all the happy moments, funs, and helps.

Finally, I would like to thank my dear parents for their endless love, support, encouragement, understanding, and sharing with all of my ups and downs through all those years.

Youchun Zhang Groningen, September 2008

124