thiol–ene coupling of renewable monomers: at the forefront of bio
TRANSCRIPT
Thiol–ene Coupling of Renewable Monomers: at the forefront of bio-based polymeric materials
Mauro Claudino
Licentiate Thesis
Kungliga Tekniska Högskolan, Stockholm 2011
AKADEMISK AVHANDLING
Akademisk avhandling som med tillstånd av Kungliga Tekniska högskolan i Stock-holm, framlägges till offentlig granskning för avläggande av teknologie licentiatexa-men fredagen den 30 september 2011, kl.14.00 i sal K1, Teknikringen 56, KTH.
Copyright © 2011 Mauro Claudino All rights reserved Paper I © 2010 European Polymer Journal Paper II accepted 2011, Journal of Polymer Science Part A: Polymer Chemistry TRITA-CHE Report 2011:49 ISSN 1654-1081 ISBN 978-91-7501-094-6
To My Family
Abstract Plant derived oils bear intrinsic double-bond functionality that can be utilized di-rectly for the thiol–ene reaction. Although terminal unsaturations are far more reac-tive than internal ones, studies on the reversible addition of thiyl radicals to 1,2-disubstituted alkenes show that this is an important reaction. To investigate the thiol–ene coupling reaction involving these enes, stoichiometric mixtures of a tri-functional propionate thiol with monounsaturated fatty acid methyl esters (methyl oleate or methyl elaidate) supplemented with 2.0 wt.% Irgacure 184 were subjected to 365-nm UV-irradiation and the chemical changes monitored. Continuous (RT–FTIR) and discontinuous (NMR and FT–Raman) techniques were used to follow the progress of the reaction and reveal details of the products formed. Experimental results supported by numerical kinetic simulations of the system confirm the reac-tion mechanism showing a very fast cis/trans-isomerization of the alkene monomers (<1.0 min) when compared to the total disappearance of double-bonds, indicating that the rate-limiting step controlling the overall reaction is the hydrogen transfer from the thiol involved in the formation of final product. The loss of total unsatura-tions equals thiol consumption throughout the entire reaction; although product formation is strongly favoured directly from the trans-ene. This indicates that initial cis/trans-isomer structures affect the kinetics. High thiol–ene conversions could be easily obtained at reasonable rates without major influence of side-reactions demon-strating the suitability of this reaction for network forming purposes from 1,2-disubstituted alkenes. To further illustrate the validity of this concept in the for-mation of cross-linked thiol–ene films a series of globalide/caprolactone based copolyesters differing in degree of unsaturations along the backbone were photopol-ymerized in the melt with the same trithiol giving amorphous elastomeric materials with different thermal and viscoelastic properties. High thiol–ene conversions (>80%) were easily attained for all cases at reasonable reaction rates, while maintain-ing the cure behaviour and independent of functionality. Parallel chain-growth ene-homopolymerization was considered negligible when compared with the main cou-pling route. However, the comonomer feed ratio had impact on the thermoset properties with high ene-density copolymers giving networks with higher glass tran-sition temperature values (Tg) and a narrower distribution of cross-links than films with lower ene composition. The thiol–ene systems evaluated in this study serve as model example for the sustainable use of naturally-occurring 1,2-disubstituted al-kenes at making semi-synthetic polymeric materials in high conversions with a range of properties in an environment-friendly way.
Sammanfattning
Vegetabiliska oljor som innehåller dubbelbindningar kan användas direkt för thiol-ene reaktioner. Trots att terminala dubbelbindningar är mycket mer reaktiva än interna visar dessa studier att den reversibla additionen av thiyl radikaler till 1,2-disubstituerade alkener är en viktig reaktion. För att undersöka tiol–ene reaktioner-na, som ivolverar dessa alkener förbereddes stökiometriska blandningar av en tri-funktionell propionat tiol och enkelomättade fettsyrametylestrar (metyloleat eller metyl elaidat) samt 2.0 vikt.% Irgacure 184. Dessa blandningar utsattes för 365-nm UV-strålning och de kemiska förändringarna studerades. De kemiska förändringar-na analyserades med olika kemiska analysmetoder; realtid RT–FTIR, NMR och FT–Raman. Dessa användes för att analysera de kemiska reaktionerna i realtid och följa bildandet av produkterna. Reaktionsmekanismen bekräftades med hjälp av experimentella data och beräkningar av numeriska och kinetiska simuleringar för systemet. Resultaten visar en mycket snabb cis/trans-isomerisering av alkenmonome-ren (<1.0 min) jämfört med den totala förbrukningen av dubbelbindningarna, vilket indikerar att det hastighetsbegränsande steget kontrolleras av väteförflyttningen från tiolen till slutprodukten. Förbrukningen av den totala omättade kolkedjan är lika med tiolförbrukningen under hela reaktionen, även om bildandet av produkten gynnas från trans-enen. Detta indikerar att den första cis/trans-isomerstrukturen påverkar kinetiken. Höga tiol-ene utbyten kan enkelt erhållas relativt snabbt utan inverkan av sidoreaktioner. Detta innebär att denna reaktion kan användas som nätverksbildande reaktion för flerfunktionella 1,2-disubstituted alkenmonomerer. Vidare användes fotopolymerisation i smälta på en serie globalid/kaprolakton-baserade sampolyestrar med varierad grad av omättnad med samma tritiol vilket resulterade i bildandet av amorfa elastomeriska material med olika termiska och viskoelastiska egenskaper. Hög omsättning (>80%) uppnåddes relativt enkelt för samtliga blandningar oberoende av den initiala funktionaliteten. Homopolymerisat-ion av alkenen var försumbar i jämförelse med den tiol–en-reaktionen. Mängden alkengrupper har inverkan på härdplastsegenskaperna där en hög andel alken ger en nätstruktur med högre glastransitionstemperatur (Tg). Tiol–ene reaktionen utvärde-rades i modellsystem baserade på naturlig förekommande 1,2-disubstituterade al-kener för att demonstrera konceptet med tiol-förnätade halvsyntetiska material.
List of Publications This thesis is an extensive summary of the following articles, which are referred to in the text by Roman numerals and appended at the end of the thesis:
Paper I (published)
Thiol–ene Coupling of 1,2-Disubstituted Alkenes: the kinetic effect of cis/trans-isomer structures. Claudino, M.; Johansson, M.; and Jonsson, M. Euro-pean Polymer Journal, 2010 (46), p. 2321–2332.*
Paper II (accepted manuscript)
Photoinduced Thiol–ene Crosslinking of Globalide/ε-caprolactone Copoly-mers: curing performance and resulting thermoset properties. Claudino, M.; van der Meulen, I.; Trey, S.; Jonsson, M.; Heise, A.; and Johansson, M. Journal of Pol-ymer Science Part A: Polymer Chemistry, 2011.
* For corrections to Paper I see page 49.
Contribution to the Papers The contribution of the author to the appended papers goes as follows:
Paper I. All of the experimental work and measurements, all numerical kinetic simulations (modelling) and chemical data analysis, and most of the preparation of the manuscript.
Paper II. Most of the experimental work and measurements, a large part of the data analysis and writing of the manuscript.
List of Abbreviations and Terms UV Ultraviolet BP Benzophenone BDE Bond Dissociation Energy RSH Thiol Group PI Photoinitiator RSSR Dissulfide Product Tg Glass Transition Temperature PGI Polyglobalide ROP Ring-opening (co)Polymerization THF Tetrahydrofurane TMP 2-ethyl-(hydroxymethyl)-1,3-propanediol NMR Nuclear Magnetic Ressonance CDCl3/d-solvent Deuterated Chloroform TMS Tetramethylsilane RT–FTIR Real-Time Fourier Transform Infrared Spectroscopy MCT Mercury Cadmium Telluride (photoconductive detector) FTIR Fourier Transform Infrared Spectroscopy ATR Attenuated Total Reflectance TGS Triglycine Sulfate (piroelectric detector) FT Fourier Transform NIR Near Infrared MO Methyl Oleate ME Methyl Elaidate FAME Fatty Acid Methyl Ester P(GI-co-CL) Poly(globalide-caprolactone) copolymer PCL Poly(ε-caprolactone) LSODA “Livermore Solver for Ordinary Differential Equations” ODE Ordinary Differential Equation
DSC Dynamic Scanning Calorimetry Tm Melting Point Temperature DMTA Dynamic Mechanical Thermal Analysis RTIR Real-Time Infrared Spectroscopy TE Thiol–ene CL (or ε-CL) ε-caprolactone PAm Polyambrettolide DCP Dicumyl Peroxide GI Globalide Am Ambrettolide DXO 1,5-dioxepan-2-one 4MeCL 4-methyl caprolactone CALB Candida Antarctica Lipase B PS Polystyrene fene Ene Functionality fthiol Thiol Functionality α Critical Fractional Conversion (gel-point¶) r Thiol–ene Molar Ratio (based on functional groups) FA Fatty Acid SEC Size Exclusion Chromatography (analogous to GPC) In Initiator I• Primary Initiating Free-Radical / or Iodide Radical SD Standard Deviation TBD Triazabicyclodecene (catalyst) AIBN Azobisisobutyronitrile (thermal radical initiator) DMPA 2,2-dimethoxy-2-phenylacetophenone (Irgacure 651, photoinitiator)
([1, 2]). ¶ Defined as the point in the polymerization reaction at which an infinite network first ap-pears in the system (onset of gelation) accompanied with fluidity loss and the viscosity be-comes so large that an air bubble cannot rise through it [1, 2].
Nomenclature of the polymers and films
Composition Polymer P(GI-co-CL)
Homopolymerized film Thiol–ene film
G:C 100:0 P(G:C) 100:0 f-H(G:C) 100:0 f-TE(G:C) 100:0 G:C 50:50 P(G:C) 50:50 f-H(G:C) 50:50 f-TE(G:C) 50:50 G:C 40:60 P(G:C) 40:60 f-H(G:C) 40:60 f-TE(G:C) 40:60 G:C 30:70 P(G:C) 30:70 f-H(G:C) 30:70 f-TE(G:C) 30:70 G:C 20:80 P(G:C) 20:80 f-H(G:C) 20:80 f-TE(G:C) 20:80 G:C 10:90 P(G:C) 10:90 f-H(G:C) 10:90 f-TE(G:C) 10:90
Table of Contents 1. PURPOSE OF THE STUDY ..................................................................... 1
2. INTRODUCTION .................................................................................... 2
2.1. Free-radical Induced Thiol–ene Reaction ................................................ 2
2.1.1. Short Historical Overview .................................................................... 2
2.1.2. Basic Chemistry and Reaction Mechanisms ........................................... 3
2.1.3. Advantages of the Thiol–ene Reaction ................................................... 8
2.2. Renewable Resources for the Thiol–ene Reaction .................................... 9
3. EXPERIMENTAL ................................................................................... 12
3.1. Chemicals ............................................................................................. 12
3.2. Techniques and Instrumentation .......................................................... 12
3.2.1. NMR Spectroscopy ............................................................................. 12
3.2.2. FTIR Spectroscopy ............................................................................. 13
3.2.3. FT–Raman Spectroscopy .................................................................... 13
3.2.4. UV/Vis–Spectrometry ........................................................................ 13
3.2.5. UV–light Sources .............................................................................. 13
3.3. Procedures ............................................................................................ 14
3.3.1. Sample Preparation ........................................................................... 14
3.3.2. Initiator Photolysis ............................................................................ 15
3.3.3. UV–induced Reactions ...................................................................... 15
3.3.4. Discontinuous Conversion Studies ...................................................... 15
3.3.5. Product Identification ........................................................................ 17
3.3.6. Simulation Software and Modelling ................................................... 17
3.3.7. Film-formation and UV-curing.......................................................... 17
3.3.8. Dynamic Scanning Calorimetry (DSC) .............................................. 18
3.3.9. Dynamic Mechanical Thermal Analysis (DMTA) ............................... 18
3.3.10. Sol-content Determination ................................................................. 18
4. RESULTS AND DISCUSSION ............................................................... 20
4.1. Thiol–ene Addition of 1,2-disubstituted Alkenes: Z/E-isomerization mechanism and kinetics ........................................................................ 20
4.1.1. Thiol–ene Reaction Dynamics ............................................................ 20
4.1.2. Product Formation ............................................................................ 23
4.1.3. Kinetic Modelling .............................................................................. 23
4.2. Thiol–ene Photocuring of Globalide/ε-caprolactone Copolymers: kinetic behaviour and resulting film–properties from a series of poly-functional macrolactone precursors ........................................................................ 27
4.2.1. Thiol–ene Curing Kinetics and Conversion ......................................... 29
4.2.2. Thermoset Evaluations ...................................................................... 32
5. CONCLUSIONS ..................................................................................... 38
6. FUTURE WORK .................................................................................... 40
7. ACKNOWLEDGEMENTS ..................................................................... 42
8. REFERENCES......................................................................................... 43
0
1
1. PURPOSE OF THE STUDY
ree-radical induced photopolymerization of thiol–ene systems have obtained a renewed interest during the last decades thanks to several attractive key features
differing from conventional free-radical polymerization reactions such as fast reac-tion rates with reduced influence of oxygen inhibition, uniform cross-linking densi-ty, a step-wise growth chain mechanism leading to a late gel point, improved curing control, low shrinkage at high monomer conversions, and the ability to initiate polymerization without addition of photoinitiator [3]. Another area with increased attention concerns materials based on renewable resources to obtain a greener sus-tainable material production in the future. One group of bio-based alkenes is fatty acids and derived compounds containing internal main-chain double-bonds. De-spite the relatively slow rates of reaction when compared to olefins with external unsaturations, studies on the effect of thiol structure and different unsaturations within the aliphatic chain demonstrate potential applicability in the creation of cross-linked thiol–ene networks [4].
The main aim of this thesis is to provide a better understanding of the free-radical thiol–ene coupling of 1,2-disubstituted alkenes in bulk with special focus on the kinetic effect of initial cis/trans-configurations promoted by photogenerated thiyl radicals. The purpose is furthermore to determine the main reaction routes at high conversions relevant for end-use applications such as organic coatings and how ini-tial ene compositions of closely-related unsaturation systems, such as those based on polyglobalide copolymers, affects the cross-linking kinetics and final film properties as consequence of possible diffusion and/or mobility restriction(s) developed within the network upon the cure process.
F
2
2. INTRODUCTION
2.1. Free-radical Induced Thiol–ene Reaction
2.1.1. Short Historical Overview
The thermal cross-linking of natural rubber [poly(cis-isoprene)] with elemental sul-fur was discovered in 1839 by Charles Goodyear [5]. This process later became known as vulcanization, and is still to this date extensively used by the tire industry. The thiol–ene reaction was already observed in 1905 by Postner showing that enes and thiols could react spontaneously with each other or in the presence of an acid [6]. In 1926 it was presented for the first time as a polymer forming reaction where it was discovered that allyl mercaptan ‘spontaneously gelled’ upon heating [7]. However, its basic mechanistic formulation as a free-radical ‘mediated’ polymeriza-tion, including the individual reaction steps, was accomplished later by Kharasch et al. just before the Second World War [8]. The early work involving the coupling of thiols to olefins was concisely described in 1970 by Griesbaum [9] and in 1993 Jacobine reviewed extensively all the aspects of thiol–ene photopolymerizations [10]. Since then, this unique reaction has attracted significant attention especially in organic [11] and polymer syntheses [12, 13].
In the past, the main large-scale applications of free-radical thiol–ene chemistry included the manufacture of relief printing plates (also known as the Latterflex pro-cess), wear layers for floor tiles (based on UV-curable resins) and coatings for elec-tronics [14]. However, the use of thiol–ene systems had been restricted to some extent due to issues of bad odor and difficulties in stabilizing the systems leading to short pot-life of formulated monomer mixtures. Moreover, the erroneous impres-sion that all thiol–ene coatings were subject to rapid yellowing (caused by residues of photoinitiator) and discoloration upon weathering, in part due to the large usage of benzophenone (BP) as photoinitiator, as well as the introduction of cheap, readi-ly-available acrylate monomers, made the popularity of thiol–ene photopolymeriza-tion decrease severely and gave way to acrylate-based photocurable systems. The revival of this chemistry was attributed mainly to the development of cleavage-type photoinitiators to initiate the thiol–ene photopolymerization (eliminating the prob-lem of yellowing) and the incorporation of thiols into acrylate formulations to de-crease oxygen sensitivity and improve the final network properties. In the last 10 years most of the research in thiol–ene chemistry has been focused on the develop-
3
ment of new materials and applications, inclusively substrate surface modifications, formation of networks with unique properties, polymer functionalization and pho-tocuring, high-impact energy absorbing materials, among very many others [14, 15]. Due to its high efficiency, the term (‘click’) was recently coined to this reaction (2008) [16-18] and there are already a vast number of excellent reviews on the sub-ject covering a broad range of scientific fields and applications [10, 12, 14, 15, 19].
2.1.2. Basic Chemistry and Reaction Mechanisms
The lability of thiol hydrogens differentiates thiol–ene polymerizations from con-ventional free-radical polymerizations. The thiol–ene reaction takes advantage of the easily abstractable hydrogen atom of the thiol group due to the relatively weak sul-phur-hydrogen bond (BDE, MeS–H=368.44 kJ⋅mol–1) [20]. This happens because the electron-poor hydrogen is bonded to the less electronegative sulphur atom (RSδ–
–Hδ+) if compared with a more electronegative oxygen from an alcohol group where this is more difficult (BDE, MeO–H=435.43 kJ⋅mol–1). The cleavage of S–H bonds can be promoted either by direct photolysis (or simply by thermolysis) or indirectly from heat- or light- generated nucleophilic alkyl radicals obtained from the cleavage of initiators. The resulting electrophilic thiyl radicals (RS•) are extremely reactive and can add to a wide variety of unsaturated compounds (both electron-rich/poor double-bonds) to form new carbon–carbon linkages. This addition reaction (also termed hydrothiolation of C=C bonds) is exothermic [15] and energetically fa-voured as a new strong σ C–C bond (∼370 kJ⋅mol–1) is formed at the expense of a weaker alkene π-bond (∼235 kJ⋅mol–1) [21]. However, the overall rate of addition is strongly dependent on the chemical structure of the thiol and ene (internal, isolated, conjugated, non-conjugated, and substituted), with thiyl radicals adding faster to electron-rich double-bonds (terminal and monosubstituted) than to electron-deficient ones (e.g. acrylates). An accurate summary of the general trends in reactivi-ty involving thiols and enes is given by Hoyle et al. [12, 14].
The thiol–ene reaction proceeds as a typical radical chain process with initiation, propagation and termination steps. The characteristic two-step mechanism for the hydrothiolation of an isolated unsaturation is represented in Scheme 1. At first, the reaction starts via initiation (often UV-induced) which promotes hydrogen transfer from the thiol to one of the initiating free-radicals generated, for instance, via the cleavage of a photoinitiator. The resulting thiyl radical then adds across the C=C double-bond (propagation step 1) yielding an intermediate β-thioether carbon-centered radical followed by chain transfer to a second thiol group (propagation step
4
2) to give the final thiol–ene addition product with anti-Markovnikov orientation. The mechanism regenerates the RS• radical, thus there is no net consumption of thiol groups, allowing the polymerization to be continued in a cyclic sequence. Termination reactions are frequently considered unimportant if compared with the rates of propagation and usually involve bimolecular combination of the intervening radical species (β-carbon or thiyl radicals), although these processes still remain obscure.
One of the most prominent competing reactions to thiol–ene additions is the pro-pensity of the ene to homopolymerize via a pure chain-wise radical growth mecha-nism. In this case the choice of the ene will affect the progress and outcome of the polymerization as one route is favoured over the other thus leading to different structure build-up patterns in multifunctional monomer systems. For example, enes such as acrylates, prone to rapid homopropagation, will to some extent homopoly-merize even in the presence of thiol whereas monomers such as allyl ethers to a sig-nificantly less extent will do this (Scheme 2) [15]. This feature has allowed the ad-justment of the overall reaction sequence in systems based on more than two mon-omers to create novel polymeric structures with exclusive properties as described by Bowman et al. [22-24]. A thiol–ene system is therefore to a larger extent affected by the reaction kinetics for the different reactions than conventional systems.
propagation cycle
R1 Sinitiator (if used)+
thiol–ene addition product
i.
ii.
Scheme 1. Step-wise growth mechanism of the free-radical thiol–ene coupling involving a terminal ene with alternating propagation (i.) and chain transfer (ii.). In ideal stoichiometry and absence of competing reactions, such as homopolymerization of the alkene monomer (i.e., chain growth), a single thiol group couples with an ene functionality to yield the final thioether (C–S linkage) product.
5
PI + h 2 Ikd
I +R2
I
R2
kadd
I
R2
+R2
kpr1I
R2 R2
n
R2
R1 S +R2
R1 SR2 R2
n
kpr2
(1)
(2)
(3)
(4)
AcrylateVinyl ether Allyl etherMethacrylate
no homo‐polymerization
very slow homo‐polymerization
slow homo‐polymerization
fast homo‐polymerization
< < <
Scheme 2. Chain-wise radical growth mechanism showing the degrees in homopolymeriza-tion of the ene-monomer.
Although internal olefins show generally much lower reactivity towards the hy-drothiolation of unsaturations than singly substituted ones, this is not uniquely due to an increased degree of substitution of the double-bonds (causing steric hindrance effects) but also because of isomerization. One exception to this is the double-bond of norbornene which exhibits an exceptionally high reactivity toward thiol addition attributed to bond angle distortion in association with ring strain relief [14]. When 1,2-disubstituted enes are involved, for instance those present in fatty acids, isomer-ization further reduces the rate at which this occurs due to interchangeability be-tween cis/trans configurations (cf. Scheme 3). Additionally, the intermediate alkyl radical formed between the two isomeric forms has a rather short lifetime and low resonance stability which further hampers its reaction with the thiol. This reduces the rate of reaction of the second hydrogen-transfer step due to the inherently low hydrogen-abstraction rate constant [9]. Two common ways used to speed-up this process is by increasing the concentration of thiol in the reaction system or by low-ering the reaction temperature; even though sometimes these approaches are not feasible in practical terms [9, 25]. When equimolar thiol–ene ratios are required, for example in network formation, thiol concentration cannot be increased and the reaction temperature should be elevated to prevent crystallization of the polymer.
6
Additionally, higher amounts of thiol groups promotes the occurrence of secondary reactions such as disulfide formation [3]. The contribution of all these factors has made internal main-chain alkenes less accountable for the thiol–ene reaction, in particular to what concerns polymer synthesis.
Initiation
Isomerization – propagation
Termination
Thiol–ene coupledproduct
Scheme 3. General mechanism proposed for the free-radical thiol–ene reaction involving an internal (isomerizable) ene. The species (A•) denotes the equilibrium radical structure of the intermediary fragmentation adduct.
From a chemical perspective, the cis-to-trans conversion of unsaturations is a ther-modynamically driven process, which can also be induced by reversible addition with thiyl radicals (RS•) using the thiol–ene reaction [26, 27]. The mechanism con-ceived involves the abstraction of hydrogen from the thiol group, the insertion of the generated thiyl radical to the cis unsaturation of an oleate moiety to form the
radical adduct ( A Z
• ) followed by half-rotation about the C9–C10 bond to give ( AE
• ) and subsequent ejection of the thiyl radical by β-fragmentation (regeneration) (cf. Scheme 4). Mechanistically, the reaction scheme shows essentially the same elemen-tary processes as for terminal enes except the existence of the isomerization step.
7
The formation of either cis- or trans-isomers depends on the conformational state of the intermediary fragmentation adduct at the instant of thiyl radical loss [28]. This event changes the double-bond geometry leading to the thermodynamically more stable trans-isomer. The energy difference between the two geometrical isoforms based on the catalysis of 2-butene by HOCH2CH2S• is 1.0 kcal⋅mol–1 at 20°C [26]. Thiyl radicals are amongst the most effective agents known to catalyze the cis/trans-isomerization because even a small amount of radical species is capable of making the reaction to proceed [27, 29]. Constitutional isomers cannot be obtained as reac-tion products since the mechanism does not allow positional shift of the double-bond [29]. Also, the location of the unsaturation along the aliphatic chain, together with tail length, has proven to have no relevant effect on the isomerization itself [28] although it affects the reactive character of the C=C bond as a result of steric hindrance effects. It was shown, for instance, that 1-hexene is 8-times more reactive than trans-2-hexene and 18-times more reactive than trans-3-hexene based on equal C=C/RSH mole ratios [14]. Many other free-radicals (e.g., RSO2
•, R3Sn•, RSe•, NO2
•, or (Me3Si)3Si•) and elementary radicals (such as Br• or I•) are known to in-duce cis/trans-isomerization through an insertion–elimination sequence, although with different efficiency than with thiyl radicals [27, 29-31]. There is also support-ing evidence that oxygen (<0.3 mM) does not seem to play a strong influential role in the effectiveness of cis/trans-isomerization if internal cis-alkenes are employed [32, 33].
aZk a
EkfEkf
Zk
A•EA•
Z
Scheme 4. Proposed reversible addition mechanism of the thiyl radical to a 1,2-
disubstituted ene at position C9–C10 (adapted partially from [32]). The conformers ( A Z
• and
AE
• ) denote intermediary fragmentation states of the carbon-centered radical adduct.
8
2.1.3. Advantages of the Thiol–ene Reaction
Some distinct benefits of the thiol–ene coupling reaction have already been men-tioned in the above sections. One of the key advantages offered by this chemistry over classical free-radical polymerizations is the high ability to overcome oxygen inhibition. In the case of free-radical polymerization using (meth)acrylates, if oxy-gen adds to a polymeric propagating chain terminus, that chain will end immediate-ly because the formed alkyl peroxy radical will have insufficient reactivity to add to a new ene monomer. This results in short chain fragments, a loose network structure and reduced cross-link density [34]. For the thiol–ene polymerization this does not happen since the peroxy radical is still able to abstract hydrogen from a thiol mon-omer producing a new thiyl radical that propagates the polymerization with only minimal impact on the main reaction route (Scheme 5). Since thiols act as strong hydrogen donors, they can be added (even in small amounts) to acrylic formulations to suppress the effect of oxygen through trapping peroxy radicals [35]. Importantly, the thiol–ene reaction is also considered an environmentally benign tool that pro-ceeds without the need of solvents and under clean and mild reaction conditions (if induced photochemically), and that rejects potentially toxic metal catalysts so com-monly employed in other ‘click’ reactions [12, 19].
final thiol–ene product
Scheme 5. Hydrogen abstraction vs. oxygen-scavenging routes for the free-radical-mediated thiol–ene reaction involving terminal enes.
Concerning thin-film formation as thermosets for organic coating applications, individual attributes of the thiol–ene reaction are summarized as follows:
9
• Self-initiation without the need of photoinitiator. This allows the polymeriza-tion of very thick geometries since UV-light is able to penetrate throughout the thickness of the film without being absorbed by the photoinitiator result-ing in homogeneous cure. Therefore, the photopolymers formed are less prone to degradation/yellowing, without generation of any coloured or vola-tile by-products, and fast aging during prolonged exposure to sunlight [36]. If a photoinitiator is to be used, then it must be ensured that the main emis-sion peak of the UV-light source overlaps with the absorption peaks of the initiator for maximum cure efficiency.
• High uniformity in the network cross-link density. This results in materials with homogeneous properties across all dimensions as a consequence of free-radical two step-growth mechanism which ensures optimal mechanical per-formances and tendency for extensive sub-Tg relaxation processes related with physical aging (enthalpy and volume relaxation) [15, 37].
2.2. Renewable Resources for the Thiol–ene Reaction
One remarkable feature inherent to the thiol–ene reaction is that virtually any al-kene functional group can participate [9]. As mentioned already, the chemical na-ture of the double-bond affects, to a large or small extent, the efficiency of the reac-tion and this much often dictates which ene monomers can be selected for a particu-lar application. So far external enes and norbornenes have had preferential choice due to their highly reactive character. They are, however, derived mostly from pet-rochemical feedstocks and a continual search for sustainable ‘green’ alternatives is of paramount importance in order to curtail their dependence on evermore depleting fossil oil reserves. Moreover, by using bio-based alkenes the nature’s synthetic po-tential is fully exploited directly in a very efficient way. Many good examples of natural ene-compounds are reported in the literature as building-blocks in polymer science using a wide variety of chemistries [38-41], but only ‘a handful’ are men-tioned in particular for the thiol–ene reaction.
For instance, Johansson et al. (2003) functionalized metallic aluminium sheets with mercapto silanes and then reacted the pendant thiol groups with linseed oil under photochemical conditions to yield thin-film vegetable coatings (∼25Å) that exhibit-ed reduced surface friction when subjected to heavy loads [42]. Bantchev and co-workers (2009) investigated the formation of sulphide-modified vegetable and cano-la oils bearing internal cis-unsaturations by butanethiol under UV-irradiation, with
10
the aim of producing new lubricants from natural renewables to improve wear and friction properties. The results showed that, working under optimized reaction con-ditions, it was possible to attain double-bond conversions up to 97% with an isolat-ed yield of 61% [43]. In another study, Türünç et al. (2010) reported the addition of mono- and di-functional thiols to methyl 10-undecenoate, a castor oil derived ene monomer, and successfully polymerized the resulting monomers using TBD as a catalyst, to linear as well as hyperbranched aliphatic polyesters bearing thioether linkages that exhibited good thermal properties [44]. An indirect approach involves the grafting of the double-bonds via an intermediate step with thiol-terminated precursors to make the polymers amenable for further chemical modifications or even post-curing purposes. In this case, Ates et al. (2011) successfully functionalized PGI (a linear C=C unsaturated polyester produced by enzymatic ROP from the macrocyclic lactone globalide) with 6-mercapto-1-hexanol (MH), butyl-3-mercapto propionate (BMP) and N-acetylcysteamine (nACA) as pendant side-chain linkers [45]. Following the same approach, Desroches et al. (2011) successfully synthesized bio-based oligo-polyols by free-radical photo-addition of 2-mercaptoethanol onto rapeseed oil and then the functionalized triglyceride was used in the synthesis of polyurethanes [3]. The combination of all these different methods has culminated recently (2011) in the development of a new vegetable-oil based polyamine issued from grapeseed oil and cysteamine chloride by use of the thiol–ene photoreaction which was then employed in a second step to thermally create a cross-linked materi-al from epoxidized linseed-oil [46]. Limonene, a readily available monoterpenic fragrance, has also been studied in the synthesis of monomers and polymers using methyl thioglycolate as model compound [47]. According to the authors, optimal reaction conditions were developed for the selective functionalization of the termi-nal vinyl-bond allowing the synthesis of a diversity of different monomers. Direct polymerization of limonene with dithiols was also reported in this proceeding. A few more examples exist in the literature [48-50] but concerning the formation of cross-linked thiol–ene networks the references on bio-based enes are even more scarce. For example, several allyl-, acrylate- and vinyl- ether derivatives of ricinolein (the chief unsaturated triglyceride constituent of castor oil) were synthesized and used together with multifunctional thiols in the preparation of UV-curable systems. Testing on the films immediately after UV-exposure and one week later indicated increased cross-linking and superior physical properties upon aging [41].
Based on the works developed by Samuelsson et al. (2004) [4] and van der Meulen et al. (2008, 2011) [51, 52], herein it is highlighted the use of monounsaturated oils (Paper I) and globalide based (co)polymers (Paper II) together with a propionate
11
ester trithiol to evaluate the reactivity of these 1,2-disubstituted alkenes in bulk and aiming at the production of cross-linked thiol–ene networks.
12
3. EXPERIMENTAL
Bellow follows a comprehensive summary of the methods and experimental proce-dures for the presented results. Complete details can be found in Paper I and Paper II.
3.1. Chemicals
The tri-functional thiol cross-linker trimethylolpropane tris(3-mercaptopropionate) (TMP-trimercapto propionate, 398.56 g⋅mol–1) was kindly supplied by Bruno Bock (Marschacht, Germany). Methyl oleate (99%), methyl elaidate (99%), ε-caprolactone and THF were purchased from Sigma-Aldrich (Stockholm, Sweden) and the photoinitiator 1-hydroxy-cyclohexyl-phenyl-ketone (Irgacure 184®) was obtained from Ciba Specialty Chemicals. Globalide was a kindly gift of Symrise. All chemicals were used as received without further purification (Figure 3.1).
TMP–trimercapto propionate
Irgacure 184
Methyl oleate Methyl elaidate
Globalide ε-caprolactone
Figure 3.1 List of chemical compounds used in this study.
3.2. Techniques and Instrumentation
3.2.1. NMR Spectroscopy
1H and 13C NMR spectra of the samples were recorded on a 400 MHz Bruker As-pect NMR spectrophotometer (Karlsruhe, Germany). CDCl3 containing 0.05 vol.% of tetramethylsilane (TMS) was used as d-solvent. Chemical shifts (δ) were
13
reported in parts per million (ppm) relative to the tetramethylsilane internal stand-ard (TMS, δ=0.00 ppm). Spectral analysis was made using the Mestrec® Software.
3.2.2. FTIR Spectroscopy
RT–FTIR spectra were recorded in the mid-region with a Perkin-Elmer Spectrum 2000 (Norwalk, CT) using an MCT detector cooled with liquid nitrogen. The FTIR instrument was equipped with a heat-controlled Golden Gate single reflection ATR-accessory from Graseby Specac Ltd. (Kent, England). The horizontal ATR-sampling unit was modified in order to accommodate a vertical UV-light cable. All FTIR measurements were performed in the reflection mode via the single-bounce diamond ATR crystal. Conventional ATR-FTIR measurements were performed on a Perkin-Elmer Spectrum 2000 equipped with a TGS detector using the Golden Gate setup. Each spectrum collected was based on 32 scans averaged at 4.0 cm–1 resolution in the range of 600–4000 cm–1. Data were acquired and processed using the software Spectrum from Perkin-Elmer.
3.2.3. FT–Raman Spectroscopy
FT–Raman measurements were performed with a Perkin-Elmer Spectrum 2000 NIR–Raman equipment with Spectrum software to determine the residual unsatu-ration and pendant thiol groups remaining in the thiol–ene mixtures and cross-linked films. Each spectrum collected was based on 16-scans using a laser power of 800–1000 mW.
3.2.4. UV/Vis–Spectrometry
Ultraviolet–visible spectroscopy was conducted on a double-bean Cary E1 UV–Vis spectrophotometer. The scan resolution of the UV–Vis spectrometer was 1.0 nm.
3.2.5. UV–light Sources
A Hamamatsu L5662 equipped with a standard medium-pressure 200W L6722-01 Hg–Xe lamp and provided with optical fibers was used as the UV-source for the photo-RTIR measurements. A condenser lens adapter, model A4093 from Hama-matsu, was employed to focus the UV-beam. Neutral density filters with optical densities of 0.2, 0.4, 0.6, 1.0 and 2.0 from CVI Laser Corp. LCC were used to obtain a constant irradiance of 5.0 mW⋅cm–2. The UV-intensity was measured using
14
a Hamamatsu UV-light power meter (model C6080-03) calibrated for the main emission line centred at 365 nm. An UV Fusion Conveyor MC6R equipped with Fusion electrodless bulbs standard type BF9 (UV-fusion lamp) was employed for network formation purposes. The UV-light intensity was determined with a UVICURE®Plus from ET, Sterling, VA.
3.3. Procedures
3.3.1. Sample Preparation
Two different reaction mixtures containing TMP-trimercapto propionate (89.6 mg) and 1,2-disubstituted alkene (MO or ME, 200 mg) in a molar ratio of 1:1 with respect to thiol–ene functionalities were prepared. Irgacure 184 (2.0 wt.%, ∼2.8 mg) was added to each mixture. The approximate molarities of trithiol, FAMEs and PI in the samples were 0.74, 2.21 and 0.045 M, respectively. The initial bulk mix-tures had to be heated slightly prior to the analysis in order to mix all the compo-nents completely.
Thiol–ene mixtures of the different unsaturated P(GI-co-CL) copolymers were pre-pared by dissolving the copolymer in 5.0 ml of THF solvent and then mixing equimolar amounts of trithiol with respect to thiol and ene functionalities so that all groups could react theoretically with each other at the same stoichiometry. The samples were supplemented with a small amount of the UV-initiator Irgacure 184 (∼2.0 wt.%) and then kept at 4°C protected from light until further use. A typical formulation is as follows: 525 mg of P(GI-co-CL) 47/53, 190 mg of TMP-trimercapto propionate and 14.3 mg of Irgacure 184. The same recipe and amounts were used, with omission of the thiol reactant, for the homopolymerized samples. For the thiol–ene films, the mass ratio necessary for equimolar reaction of thiol–ene functional groups was calculated using the expression:
⋅ ⋅=
⋅ trithiol
polym. n
t
3m f DP Mm M
(1)
where f is the mole fraction of ene in the copolymer, DP is the degree of polymeri-zation, Mt is the molecular weight of the trithiol and Mn the number-average mo-lecular weight of the copolymer. The number of ene functionalities per copolymer chain is given by ⋅f DP .
15
3.3.2. Initiator Photolysis
To obtain a rough estimate for the production rate of primary free-radicals, a dilut-ed solution of Irgacure 184 in methyl oleate was prepared (0.02 wt.%, ∼0.86 mM). The sample was sealed inside a 1-mm thick quartz cell and then irradiated intermit-tently in the presence of air at 20°C with an UV-intensity of 5.0 mW⋅cm−2. The absorbance was measured in the range 190–600 nm. The decomposition of pho-toinitiator with time was monitored by the decrease of the main absorption peak at 247 nm until there was no significant change in the absorption spectrum. All spec-tra were subtracted from a blank sample containing only methyl oleate in order to eliminate spectral contribution of the solvent (background spectra). Concentrations over time were calculated from the Lambert–Beer law.
3.3.3. UV–induced Reactions
For RT–FTIR measurements of photoinitiated reactions, the heat-controller was set to 60°C and the system left to equilibrate for 10 min. Two drops of the mono-mers/photoinitiator mixture (∼75 μl) were then applied onto the surface of the ATR diamond probe connected in-situ to a circular heater plate and the measurements were started immediately. The photoreaction and photopolymerizations were initi-ated by vertically irradiating UV-light from the Hamamatsu lamp. Duplicate IR-runs were conducted in the presence and absence of air (by covering the liquid sam-ples with a thin quartz lamella) over a period of 30 min using independent samples. RTIR continuously recorded the chemical changes over the range of 4000–600 cm–
1. Spectroscopic data were collected at an optimized scanning rate of 1 scan per 1.67 seconds with a spectral resolution of 4.0 cm–1 using the TimeBase® software from Perkin-Elmer. The course of the reaction was followed by monitoring the peaks corresponding to the cis- and trans-unsaturation carbon-carbon double-bonds oc-curring at ∼3010 and 968 cm–1, respectively. Figure 3.2 illustrates the custom-made RTIR setup utilized in the experiments. Aliquot samples for 1H NMR analysis were taken before and after the reaction in order to estimate the final conversion of dou-ble-bonds (photoreaction system only).
3.3.4. Discontinuous Conversion Studies
Both 1H NMR and FT–Raman spectroscopies were used for the discrete evaluation of thiol–ene conversions (photoreaction system).The procedure involved the pour-
16
ing of a small volume of the thiol–ene sample on top of a glass slide set in contact with the ATR crystal and thermostatted to 60°C. The liquid films were irradiated with a certain dose (mJ⋅cm–2) and the main portion analysed by FT–Raman to as-sess the level of intensity of the –SH peak at ∼2576 cm–1 after spectral normalization with respect to the ester carbonyl peak at 1735 cm–1. The residual fraction attached to the glass was immersed in cold CDCl3 and then analysed by NMR. Proton inte-gration signals of the C=C bonds (vinylic: 5.05–5.54 ppm and allylic: 2.0 ppm) were used to estimate the extent of the reaction using the integral areas of two con-served signals: (i) the methyl ester protons at 3.66 ppm, and (ii) the aliphatic pro-tons at 2.29 ppm next to the ester group. The degree of conversion into C–S bonds was expressed as the cross-average of the values obtained from both unchanged sig-nals.
Figure 3.2 Experimental RT–FTIR setup equipped with an ATR-accessory for single reflec-tion with a fixed incidence angle of 45°.
For the UV-cured films based on PGI/PCL the degree of conversion was estimated via FT–Raman spectroscopy by first normalizing all spectra with respect to the ester carbonyl peak and then taking the ratio of the band between 2600–2560 cm–1 re-sulting from the thiol functional group and/or the band at 1660–1680 cm–1 result-ing from the trans-unsaturation in the unreacted and reacted spectra. Conversions for the homopolymerized ene films and thiol–ene films were calculated by using equation (2), where A denotes the band area before and after cure.
17
final
start
Conversion (%) 1 100AA
= − ×⎛ ⎞⎜ ⎟⎝ ⎠
(2)
3.3.5. Product Identification
Identification of reaction products and characterization of the chemical structures was accomplished via proton and carbon NMR.
3.3.6. Simulation Software and Modelling
Dynamic kinetic simulations of the reaction system (Scheme 3) were performed using the application software GEPASI version 3.30 [53]. A deterministic routine algorithm called LSODA was used by GEPASI to compute the numerical solution of a set of ODEs. LSODA is a very robust adaptive step-size solver that calculates the stiffness of equations and dynamically switches the method of integration ac-cording to this measure [54-57]. The mechanistic model was entered in the software according to the elements listed in Table 1. The kinetic curves generated were plot-ted against the experimental profiles obtained from RT–FTIR data. To check the agreement between the generated output profiles and experimental conversion data, the simulated concentrations for the Z- and E-isomers were combined and plotted against time conversions of total unsaturations (cis + trans) estimated by 1H NMR. A similar procedure was used to verify the simulated-experimental agreement for thiol conversion evaluated by FT–Raman.
3.3.7. Film-formation and UV-curing
For the photo cross-linking of the different (co)polymers, thiol–ene mixtures were spread on glass substrates that were previously cleaned and rinsed with acetone. Liquid films were applied sequentially up to two layers by letting the solvent evapo-rate between applications and then melted in the oven pre-heated to 85°C for 2–3 min until transparent films were observed. The coated slides were immediately placed in the conveyor belt while still in the molten-state (by use of a 0.4 cm thick glass plate to help the sample retain the molten temperature) and passed four times under the UV-fusion lamp with a line speed of 6.52 m⋅min–1 to give an overall ex-posure dose of 0.14 J⋅cm–2. The samples were then left at room temperature to cool. Smooth, non-tacky films of 30–40 μm thickness resulted.
18
3.3.8. Dynamic Scanning Calorimetry (DSC)
The thermal properties of the cross-linked films were analyzed by DSC. The exper-iments were conducted on a DSC 820 equipped with a sample robot and a cry-ocooler (Mettler Toledo). The DSC runs were carried out in closed sample pans sealed in air, using the following temperature program: heating from 25 to 120°C (50°C⋅min–1), cooling from 100°C to –65°C (50°C⋅min–1), then heating up to 120°C (5°C⋅min–1). Isothermal segments of 5 min were performed at the conclusion of each dynamic segment. The melt enthalpy was determined from the integration of the Tm peak of the second heat.
3.3.9. Dynamic Mechanical Thermal Analysis (DMTA)
To examine the physical properties of the thiol–ene networks, DMTA was per-formed on a Q800 DMTA (TA-instruments), equipped with a film fixture for ten-sile testing. Film tension DMTA measurements were performed on rectangular dried film samples (5 × 0.05 mm, width × thickness) were performed between –30 and 140°C, with a heating rate of 3°C⋅min–1. The tests were performed in controlled strain mode with a frequency of 1 Hz, oscillating amplitude of 0.12 μm, and force-track of 125%.
3.3.10. Sol-content Determination
All cross-linked thiol–ene films were cut into small rectangular sections with ap-proximate dimensions of 1.0 cm × 2.0 cm, dried in the vacuum oven at 50°C for 1 hour, weighed and then soaked overnight in 5.0 ml of THF under gentle stirring conditions. The films were subsequently washed (2×) with THF solvent and placed again in the vacuum oven until all the residual solvent was evaporated. Sol-fractions for the thin-film specimens were determined from mass losses relative to initial dry mass according to the following equation:
f
s
Sol content (%) 1 100WW
= − ×⎛ ⎞⎜ ⎟⎝ ⎠
(3)
where Ws is the initial dry weight of the film samples and Wf is the dry weight of the same film specimens after sol-extraction. The analyses were performed in triplicate from two independent UV-cured films and the results averaged.
19
Tab
le 1
. Mod
el de
finiti
on u
sed
to g
ener
ate
the
simul
atio
ns w
ith G
EPAS
I.
Sequ
ence
R
eact
ion
Che
mic
al e
quat
ions
a
Rat
e co
nsta
nts
Ref
.
Initi
atio
n 1a
d
In2
Ik
•⎯⎯→
3
1
d1
410
sk
.−
−×
th
is w
ork
1b
R
SHI
RSH
IHR
Sk
••
++
⎯⎯
⎯→
RSH(1
)=1.
010
Ms
k−
−×
⋅7
11
[25]
Isom
eriza
tion
2a
a f
RS
AZ Zk k
Z•
•+
aM
sZ k
.−
−=
×⋅
51
116
10
[26]
71
f2
010
sZ k
.−
=×
[2
5]
2b
f a
AR
S
E Ek kE
••
+
51
1
a2
910
Ms
E k.
−−
=×
⋅
[2
6]
81
f1
610
sE k
.−
=×
[2
5, 2
6]
Prop
agat
ion
3 R
SHA
RSH
PR
Sk
••
++
⎯⎯⎯→
RS
H(2
)=1.
010
Ms
k−
−×
⋅7
11
[25,
26]
Ter
min
atio
n 4
t
2R
SR
SSR
k•⎯⎯→
91
1
t3
010
Ms
k.
−−
=×
⋅
[26]
a Ini
tial c
once
ntra
tions
: [In
] 0=4.
48×1
0–2 a
nd [R
SH] 0=
[Z] 0=
[E] 0=
2.21
M.
19
20
4. RESULTS AND DISCUSSION
4.1. Thiol–ene Addition of 1,2-disubstituted Alkenes: Z/E-isomerization mechanism and kinetics
Studies on the effect of thiols on the isomerization of fatty acids demonstrate that the reversible addition of thiyl radicals to 1,2-disubstituted alkenes is an important reaction [4, 25, 26]. It is however still unclear which main reaction routes governs the overall reaction system and how the initial ene structure (cis- vs. trans-) affects the final sulphide product at high conversions. It is vital in polymerization reactions that a high isolated yield of the desired product is achieved at high conversions in order to obtain high molecular weight polymers, i.e., the influence of side reactions should be minimal. In this part of the study the influence of the isomerization mechanism and kinetics on the efficiency in end-product formation is evaluated by reacting in bulk equally balanced thiol–ene mixtures of MO vs. ME based on TMP-trimercapto propionate under photochemical conditions. The purpose is further-more to investigate if this reaction has enough potential to yield the final coupled C–S product in high conversions and in a timely fashion relevant for the develop-ment of cross-linked polymeric materials for coating applications by exploring such natural class of alkene monomers.
4.1.1. Thiol–ene Reaction Dynamics
To assess the production rate of initiating free-radicals obtained from the photolysis of Irgacure 184 in MO, a series of irradiation experiments were performed at differ-ent timed intervals and the resulting solution analysed by UV/Vis-spectro-photometry (Figure 4.1). Irgacure 184 is a Norrish type I photoinitiator (α-cleavage) frequently used in the polymerization of thin-films and coatings given its high efficiency. Under UV-light it decomposes from its lowest triplet state into benzoyl and cyclohexanol radicals which, to a lesser or larger extent, have the ability to abstract hydrogens from thiol groups and produce reactive thiyl radicals [58-63]. An estimate from the initial slope shown in the graph indicates a rate of photolysis of ∼1.2×10–6 M–1⋅s–1 up to 110 seconds of reaction with a first-order rate constant (kd) of ∼1.4×10–3 s–1. The rate coefficient for the decomposition of photoinitiator was then used for model simulations of the overall reaction system (Table 1).
21
RT–FTIR spectroscopy has demonstrated to be a versatile tool for monitoring the course of photoinduced polymerization reactions [4, 36, 64-67]. Figure 4.2 depicts the 3D-spectra evolution of ME during the thiol–ene reaction between MO and the trithiol. At the beginning, the reaction system consists of stoichiometric functional group amounts of thiol and ene supplemented with a small percentage of photoin-itiator. Immediately upon exposure to UV-light a sharp development of trans-unsaturations from the cis-alkene is observed reaching a maximum at ∼40 seconds which then fades out slowly as the intermediary alkyl radical (A ) reacts with the thiol monomer to give the coupled thiol–ene product. The observed behaviour goes well along with the conceived isomerization–propagation mechanism.
Figure 4.1 Spectral evolution of the main UV/Vis-absorption peak at 247 nm and photoin-itiator concentration for a diluted solution of Irgacure 184 in MO. Dashed line indicates the linear decay of PI extrapolated from the first 110 seconds of reaction.
To better evaluate the combined action of the isomerization step with the product formation route in the equilibrium, complementary FT–Raman measurements were conducted discontinuously throughout the photoinduced reaction (Figure 4.3). Very early in the reaction the cis-peaks appearing at 3010 cm–1 (C=C–H symmetric stretching) and 1652 cm–1 (C=C symmetric stretching, non-conjugated) immediate-ly decreased their intensities followed by a shift towards the formation of a trans-ene peak at ∼1670 cm–1 (isolated C=C trans-stretching) which reduced in intensity as the reaction proceeded. Simultaneously, depletion of the thiol peak at ∼1576 cm–1 was observed over time which further supports the results obtained from RT–FTIR.
22
Figure 4.2 Stacked waterfall plot for the RT–FTIR spectra recorded continuously over time showing the 3D-evolution of the trans-isomer band from the reaction of an equimolar tri-thiol/MO mixture in the presence of air and supplemented with 2.0 wt.% Irgacure 184.
Figure 4.3 Changes observed in the FT–Raman spectra over selected times for the photore-action of a stoichiometric thiol–ene mixture containing MO.
23
4.1.2. Product Formation
To further confirm the existence of the final thiol–ene addition product and quanti-fy the loss of total unsaturations over time, NMR spectroscopy was employed (Fig-ure 4.4). After 30 min of reaction about 80% of double-bonds initially present in the reaction mixture were converted into sulphide groups as shown by 1H NMR. This proves the high efficiency and robustness of the thiol–ene reaction undertaken even when using these cis-enes. Indeed, the signals for the allylic (CH2–CH, c) and vinylic (CH=CH, d) protons centered at 2.00 and 5.34 ppm almost completely disappeared, respectively. Also, initial methylene protons (k) and (l) next to the thiol group significantly decreased their intensity and were shifted slightly up-field as a new signal appeared at the vicinity which strongly suggests the formation of a cova-lent C–S bond. No important side-reactions were detected such as homopropaga-tion of the 1,2-disubstituted ene nor termination reactions such as the homocou-pling of carbon-centered radicals, disulfide formation and the heterocoupling with thiyl radicals thus indicating the viability of this reaction for network forming pur-poses. Complementary 13C NMR spectra further support these results (results not shown) together with analogous studies reported in the literature [43, 45]. A recent-ly published study (2011) based on the photo-addition of 2-mercaptoethanol onto oleic acid is also supportive of the NMR results obtained in this work [3].
4.1.3. Kinetic Modelling
In order to verify the reaction scheme proposed in Scheme 3 and give a better view of the dynamic behaviour of the system, a numerical simulation of the kinetics was performed (Table 1.). An exhaustive analytical treatment of the reaction kinetics has been reported previously for solutions of MO with β-mercaptoethanol in tert-butyl alcohol under photochemical or γ-radiolysis conditions using the steady-state ap-proximation [25, 26]. Dynamic kinetic modelling with GEPASI provides the same kind of mechanistic information, however in a simplified manner, with the ad-vantage of avoiding the steady-state condition many times assumed to facilitate the mathematics. Simulation of both reaction systems (MO vs. ME) confirms the reac-tion mechanism as demonstrated by the excellent trend-shape agreement with the experimental profiles, despite the observed deviations (Figure 4.5). The differences could be ascribed to an expected increase in the medium viscosity with reaction time slowing down the diffusion of reactants in the system, the existence of parallel competing reactions unaccounted in the model [43], in addition to the evident inaccuracy of the rate coefficients given that measurements were performed at 60°C
24
and not at room-temperature as stated by the literature values. Despite the visible differences, these are not extremely pronounced to question the mechanism suggest-ing that the ‘real’ rate coefficients should exhibit comparable orders of magnitude to the values used in the simulations.
Figure 4.4 1H NMR spectra of a thiol–ene mixture starting from pure cis-isomer (MO). (A) before reaction; (B) after 30 min of reaction at 60°C.
According to the analysis, initial configuration of the 1,2-disubstituted double-bonds significantly affects the progress and outcome of the reaction. Aliphatic trans-enes are consumed directly into product while showing insignificant conversion into the cis-isomer (graph (a)), whereas cis-enes are first converted to the corresponding trans-form and then consumed into product as described by the progressive decay of
25
trans-unsaturations (graphs (c–d)). In this case, before the maximum trans-peak is reached the governing route is the conversion of cis- into trans- unsaturations with only a minor fraction of cis-isomer being re-formed from the trans-monomer given the reversible condition of equilibrium, even though this was not observed experi-mentally. The preference in formation of trans-unsaturations is essentially due to a stabilization of the E-transition state which occupies a lower enthalpy level with respect to the Z-transition state. By starting with the trans-isomer, the higher energy barrier encountered makes difficult the conversion into cis-unsaturations and the product formation route is strongly favoured [26].
Figure 4.5 Experimental results (symbols) vs. simulated data (lines) for the photoreaction of a stoichiometric thiol–ene mixture†. (a) concentration–time profiles for a mixture starting with pure trans-ene isomer (ME); (b) evolution of conversion (based on absolute moles) for a mixture starting from pure cis-alkene (MO); (c) cis/trans-isomerization of MO with con-current product formation; (d) zoom-in of graph (c) showing the first 400 seconds of reac-tion. Simulated data for product (P), cis-isomer (Z), trans-isomer (E) and thiol (RSH) are given on the inset in each graph. Predicted product line shown in graph (b) was obtained from equation (4). [†] Initial bulk concentrations for the thiol–ene monomers are given with respect to group functionalities and not to individual species.
26
Within the experimental errors, loss of total unsaturations (cis + trans) is to large extent matched by the disappearance of thiol groups and reach high conversions (∼80%) after 30 min of reaction as determined experimentally from graph (b). To a first approximation, the thiol–ene coupling reaction into product could be modelled as an ‘apparent’ second-order reaction (equation 4), even though the deterministic mechanism does not allow this.
[ ] [ ] app P obs RSH C=Ckr = ⋅⋅ (4)
For this case, an ‘observed’ second-order rate constant (kobs) for the S–C bond-forming reaction indicated a value of ∼1.5×10–3 M–1⋅s–1 obtained from the slope of the linearized (integrated) plot‡ of equation (4). The results also point out that cis/trans-isomerization is very rapid if compared with the coupling reaction that leads to product. This re-confirms that the rate-determining step controlling the overall reaction is the hydrogen-abstraction from the thiol by the intermediary ad-duct radical given that it is the slowest step. All these results are fully consistent with previous findings well documented in the literature [25-27, 66].
The work detailed in this first part illustrates the usefulness of the thiol–ene reaction involving naturally occurring cis-1,2-disubstituted alkenes (and their trans-congeners) in the synthesis of sulphide-modified vegetable oils in bulk conditions. The kinetic and mechanistic information collected provides a basis for the design of real photocurable systems involving multifunctional thiol–ene monomers intended for new materials and coating applications.
[‡]
[ ] [ ]= + ⋅
obs
0
1 1
TE TEk t , with the plot of
[ ]1
timeTE
vs. used to determine kobs from
linear regression.
27
4.2. Thiol–ene Photocuring of Globalide/ε-caprolactone Co-polymers: kinetic behaviour and resulting film–properties from a series of poly-functional macrolactone precursors
In this second part a different approach is utilized to investigate if the thiol–ene coupling reaction involving 1,2-disubstituted alkenes is still feasible when a mobility restriction factor is introduced into the system, as consequence of cross-linking, given the inherently slow reactivity of these enes [68-71]. The purpose is to evaluate the relationship between initial multifunctional ene-density, kinetic behaviour upon cure and the resulting conversion, and how the final cross-linked network structure affects the thermal and viscoelastic properties of the photocured materials.
As a role model pre-polymer for these studies a series of linear aliphatic polyesters derived from the macrolactone globalide were employed. Globalide (11/12-pentadecen-15-olide) is a synthetic unsaturated cousin of the naturally occurring fragrance ambrettolide (Am), widely used by the perfume industry due to its strong musky odor and ability to lose scent slowly [51]. These unsaturated polyesters have been successfully synthesized by lipase catalysed polymerization (Figure 4.6) using ε-caprolactone (CL) as intercalating monomer to provide random alkene copoly-mers with different degrees in ene-functionality within the backbone [51, 52]. The amount of globalide was varied between 10–50 mol.%, depending on how much caprolactone was introduced by the enzyme (Table 2). The 1,2-disubstituted enes of polyglobalide (PGI) and related polyesters such as polyambrettolide (PAm), has successfully allowed the thermal cross-linking in the molten-state with dicumyl peroxide (DCP) affording fully transparent network structures that contrasted sig-nificantly with the semi-crystalline morphology of the initial polymers [51]. The same approach was employed by using copolyesters derived from GI (or Am) com-bined with DXO (or ε-CL or 4MeCL) giving essentially the same results as with PGI and PAm alone [52]. In a recent study based solely on PGI was demonstrated the effectiveness of the thiol–ene reaction as a new straightforward methodology for the side-chain functionalization of these polymeric enes that could be aimed at a myriad of purposes [45]. It would however be of more interest if the alkene poly-mers could be reacted directly without any intermediate modifications.
Here is reported the utility of the thiol–ene reaction as an efficient means for the direct cross-linking of these unsaturated groups to high conversions and at reasona-ble reaction rates that could be used for practical applications.
28
15 + 6Novozym 435 (CALB)
(ROP)
globalide (GI) ε-caprolactone(CL)
unsaturated poly(globalide–caprolactone) copolymer: P(GI-co-CL)
+ R–SH UV
(i)
(ii)
thiol–ene addition productTMP-trimercapto propionate
R–SH:
Irgacure 184
Figure 4.6 Enzymatic ring-opening copolymerization of globalide/ε-caprolactone mono-mers (i) followed by thiol–ene UV-curing of the poly(globalide–caprolactone) copolymers with a tri-functional thiol in the presence of photoinitiator (ii). Globalide (GI) is a mixture of two constitutional isomers with the double-bond (mainly in a trans-configuration) located either at the 11 or 12 position (denoted by the dashed lines). Thioether C–S linkages are randomly located at positions 11, 12 or 13 of the aliphatic main-chain depending on which isomer is present at the instant of thiol–ene addition. Copolymer synthesis and characteriza-tion are reported elsewhere [51, 52].
Polymer Molar composition(G:C)
aMn b
(g×mol–1
)×103
PDI b DP b 1,2-disubstituted ene functionality
c
P(G:C) 100:0 100:0 15.0 2.8 65 65P(G:C) 50:50 47:53 20.0 3.3 118 55P(G:C) 40:60 40:60 21.0 3.0 125 50P(G:C) 30:70 32:68 23.0 2.4 152 49P(G:C) 20:80 22:78 18.0 2.8 125 28P(G:C) 10:90 10:90 16.0 2.4 125 13
Table 2. Properties of the synthesized polymers. General enzymatic polymerization proce-dure and characterization of the (co)polymers can be found elsewhere [51, 52].
a Determined by 1H NMR. b Determined by SEC in THF at 40°C using PS standards [52]. c Number of ene functionalities per copolymer chain.
29
4.2.1. Thiol–ene Curing Kinetics and Conversion
To investigate the direct influence of functional ene-density on the thiol–ene curing kinetics the various polymers differing in amount of unsaturations were exposed to UV-light and the progress of the reaction tracked online via RT–FTIR.
It was postulated that an increased dilution of double-bonds would make the system exhibit low reactivity early in the reaction due to thiol inaccessibility caused by a wider average spacing between unsaturated groups and then as the level of cross-linking increased higher conversions would be reached because of proximity effects. Conversely, a system based on higher density of unsaturations would cure faster initially but then reach a higher plateau with less conversion as a consequence of mobility restriction and seclusion of functional groups caused by the locked net-work structure (early gel-point) (Scheme 4.1.). The assumption is fairly consistent with results obtained from previous studies [52] using DCP as curing agent (cf. Table 3). In addition, it is well known from traditional multifunctional thiol–ene systems that the onset of gelation can be easily predicted using equation (5), and so if ene thiolf f as one would anticipate for linear polyfunctional polymers bearing internal double-bonds, then gelation would occur at fairly low thiol–ene conver-sions [14, 72]. In fact, according to the expression, using PGI as the highest ene-density polymer, a critical fractional conversion (gel-point) of ∼0.09 is attained which represents only 8.84% of the gel-point obtained (α=0.5) if standard tri-functional thiol–ene comonomers were reacted at the same functional group stoi-chiometry.
( ) ( )[ ]α−
= ⋅ − ⋅ −1
ene thiol1 1r f f (5)
Surprisingly, by using these polymeric main-chain alkenes via the thiol–ene reac-tion, no dramatic differences were observed experimentally as all kinetic profiles showed essentially the same range in decay values while attaining near-quantitative conversions. The multiple curves over time are illustrated in Figure 4.8 using the example given in Figure 4.7 as demonstrative spectra. Obtaining similar curves for all copolymers, including the pure PGI, infers that the propagation rate (expressed as mol⋅L–1⋅s–1) increases with ene-functionality as expected since the density of ene scales up proportionally with globalide content. The existence of comparable final conversions to all reaction systems irrespective of comonomer feed ratios indicates that the step-wise growth mechanism leading to delayed gel-point, in association
30
with high chain mobility owed to low Tg values (Table 4), allows the reactive groups to be accessible always throughout the cure until very high conversions are attained. A dynamic balance between kinetics and mobility established within the network seems therefore to exist. This distinctive feature has beneficial implications in the network formation process since the resulting properties can be tailored to specific purposes without interference of structure restraints, i.e., changes in cure behaviour caused by rigidity effects. In addition to this, it was found that these 1,2-disubstituted enes can be cross-linked to high conversions at reasonable reaction rates. Conversion up to ∼87% can be seen already at 5.7 min reaching a maximum of ∼95% in the end (Figure 4.8). Also, from Table 3, relatively high thiol–ene film conversions (76–92%) with only minute amounts of sol-content extracted (≤11%), indicates the formation of final C–S linkages in high isolated yields for all resulting networks. This was further supported from exploratory NMR tests of leached sol-fractions which showed mostly thiol–ene coupling with no evidence of important side-reactions such as homopolymerization of the ene monomer (results not shown). Indeed, it is well known that 1,2-disubstituted enes exhibit very poor ability for homopolymerization [66, 73]. All these results agree well with the conclusions achieved in the previous section on monounsaturated oils thus allowing an extensive array of natural alkenes of this kind to be exploited in new cross-linkable systems.
Scheme 4.1. Hypothesized effect of double-bond dilution on photocure behaviour. Low content ene-density copolymers (dashed line). High content ene-density copolymers (solid line). Arrow denotes the start of the UV-induced thiol–ene polymerization.
Time [s] or dose [mJ⋅cm–2]t0
Nor
mal
ized
trans
-alk
ene
peak
area
100%
0
31
Wavenumber (cm−1)600800100012001400160018002000220024002600280030003200
time
incr
ease
Wavenumber (cm−1)
940945950955960965970975980985990995
time increase
968 cm−1
(a) (b)
Figure 4.7 Changes observed in the IR-spectra at selected times of the photoinduced thiol–ene cross-linking involving the copolymer P(G:C) 32:68 as monitored by RT–FTIR (repre-sentative spectra). (a) overall spectra taken at time = 0, 60, 98, 116, 172, 278 and 1800 seconds, (b) region of interest showing the time-consumption of trans-double bonds.
Time (secs)
0 200 400 600 800 1000 1200 1400 1600 1800
Res
idua
l tra
ns-e
ne p
eak
(a.u
.)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
P(G:C) 10:90P(G:C) 22:78P(G:C) 32:68P(G:C) 40:60P(G:C) 47:53P(G:C) 100:0
Figure 4.8 Kinetic RT–FTIR profiles showing the time-decay of residual trans-isomer for all P(GI-co-CL) copolymers upon photoinduced thiol–ene reaction with the trithiol cross-linker. Original trans-peak area curves were corrected using the analytical procedure append-ed in Paper II. Arrow indicates the start of the photoinduced polymerization after a waiting period of 1-min.
32
4.2.2. Thermoset Evaluations
To evaluate the properties of the cross-linked thiol–ene networks the final films were characterized with respect to curing performance (with use of the UV-fusion lamp), residual sol-content (leaching tests), and thermal and viscoelastic behaviour (DSC and DMTA measurements). Polymers are often cross-linked to enhance the thermal and physicomechanical properties of the resulting materials.
The different thiol–ene films, denoted by f-TE(G:C) x:y, were cured and the result-ing conversions measured by FT–Raman. This was accomplished by measuring the area under the curve for the thiol band at ∼2580 cm–1 and the ene band at ∼1640 cm–1 before and after cure. The ester carbonyl peak (∼1735 cm–1) did not change during the reaction and was used as internal standard for all thiol–ene films. It can be seen from Figure 4.9 that the degree of conversion in most of the films is higher than 80% and superior than thermally cured copolymers [52]. This indicates that the incorporation of caprolactone does not have a significant impact on enhancing the overall conversion of the 1,2-disubstituted ene functionality in the thiol–ene reaction, as was theorized initially.
Figure 4.9 Post-cure conversions of the thiol–ene films as determined by FT–Raman.
33
Poly
mer
co
mpo
siti
onEn
e–en
eco
nver
sion
(mol
.%) a
Thi
ol–e
ne c
onve
rsio
n(m
ol.%
±SD
) bR
eact
eddo
uble
-bon
ds(m
ol.%
)cG
el c
onte
nt(w
t.%) c
Sol c
onte
nt
(wt.%
±SD
) d
P(G
:C) 1
00:0
14.2
76.7
±2.
3⎯
⎯1.
0 ±
0.6
P(G
:C) 5
0:50
3.3
85.3
±5.
4⎯
992.
7 ±
0.7
P(G
:C) 4
0:60
12.6
91.5
±1.
813
973.
6 ±
0.2
P(G
:C) 3
0:70
52.4
83.5
±4.
116
984.
1 ±
0.04
P(G
:C) 2
0:80
31.9
86.8
±4.
233
965.
9 ±
0.5
P(G
:C) 1
0:90
12.2
82.1
±4.
258
9510
.6 ±
1.3
Tab
le 3
. Rel
ativ
e co
mpa
rison
of c
onve
rsio
n de
gree
s for
the
hom
opol
ymer
ized
ene
film
s and
thio
l–en
e fil
ms w
ith v
alue
s rep
orte
d in
the
liter
atur
e (g
rey
colu
mns
with
val
ues i
n ita
lic) f
rom
pol
yeste
rs c
ross
-link
ed w
ith d
icum
yl p
erox
ide
[52]
.
a Mea
sure
d by
FT
–Ram
an fo
r the
hom
opol
ymer
ized
film
s (cu
red
with
out t
rithi
ol in
the
initi
al m
ixtu
res)
. b Det
erm
ined
by
FT–R
aman
: nu
mbe
rs a
re e
xpre
ssed
as t
he a
vera
ge ±
SD
of i
ndiv
idua
l con
vers
ion
valu
es o
f thi
ol a
nd e
ne g
roup
s sho
wn
in F
igur
e 4.
9. c V
alue
s tak
en
from
the
lite
ratu
re (
used
with
per
miss
ion
from
the
aut
hors
). D
oubl
e-bo
nd c
onve
rsio
ns f
or t
he t
herm
ally
cur
ed p
olym
ers
wer
e de
ter-
min
ed b
y FT
IR [5
2]. d
Det
erm
ined
gra
vim
etric
ally
.
33
34
The copolyesters were also analyzed with respect to ene-ene polymerization ability by free-radical polymerization using the exact same reaction conditions as for the thiol–ene photo cross-linking but without the presence of thiol (Table 3). The cured films revealed overall poor ene homopolymerization efficiency except for the f-H (G:C) 30:70 and f-H(G:C) 20:80 where slightly higher conversions were at-tained. This could be due to an optimal balance between chain mobility and func-tional group density if compared to f-H (G:C) 10:90 where conversion decreases again. Furthermore, conventional chain-growth homopolymerizations are known to constrain mobility at much lower conversions than for step-growth reactions. For the thiol–ene reaction the extent of homopolymerized ene was insignificant since the thiol–ene coupling route usually proceeds much more rapidly than any homo-polymerization occurring in parallel.
To further understand the network structure, sol-gel measurements were performed by swelling the films in a favourable solvent (THF) and then gravimetrically deter-mining the weight loss after drying (Table 3). The results indicate that the level of functionality is correlated with the amount of sol present in the film. Even though high conversion is observed for the f-TE(G:C) 10:90 film, given the low functional group density, a larger amount of sol-content, ∼11.0 wt.%, is observed. Conversely, for the f-TE(G:C) 100:0 film a relatively small percent of sol is observed at about 1.0 wt.% as a result of higher density of functional groups. It is reasonable that a system having increased functionality would result in a lower fraction of unbound monomers at the same degree of conversion. The sol-fraction increases proportion-ally to caprolactone content, meaning that a balance should be attained in formula-tions between the amounts of caprolactone resulting in a modification of properties, and increased sol-content. It is worth noting, however, that the sol-fraction is rather low in all cases showing equivalent gel-content to thermally cured films.
The thermal properties were also investigated by DSC and are summarized in Table 4. It can be seen that the Tg of PGI cannot be observed, either due to the high crys-tallinity or it is below the DSC detection limit (–60°C for the instrument in use). The same lack of Tg is observed for the homopolymerized film due to the low con-version. However, the thiol–ene films have a Tg that decreases from –32°C of f-TE(G:C) 100:0 to around –45°C of f-TE(G:C) 30:70 due to the incorporation of caprolactone. The Tg could not be detected once the caprolactone content reaches 80 and 90 mol.%. The melting temperature is highest for polyglobalide and 90% caprolactone at 46°C and then decreases as the mixture goes to 50:50. The homo-polymerized films show a decrease in crystallinity with cross-linking that inhibits
35
chain mobility to crystallize, but the conversion was low so that a large degree of crystallinity remains. The thiol–ene films with 0 to 70% caprolactone lose all crys-tallinity, but once the distance between cross-links increases, and the content of globalide and the 1,2 disubstituted alkene it carries decreases, with 70 and 80 % caprolactone content, crystallinity appears. The exotherm of the crystalline melt illustrates the amount of crystallinity in the materials and further demonstrates these points.
The viscoelastic properties of the thiol–ene films were investigated in order to determine the effect of the volume of intercalated caprolactone on the modulus and the cross-link homogeneity (Figs. 4.10 and 4.11). The glassy modulus of the films is around 1.0 GPa and is typical for polymeric films. It is apparent from the storage modulus that the film containing the highest volume percent of caprolactone, f-TE(G:C) 10:90 has a much lower Tg, and increasing the amount of polyglobalide an increase to ∼0°C is observed. Even though it was impossible to detect the Tg of f-TE(G:C) 100:0 by DSC due to the high crystallinity or because it was lower than the detection temperature of –60°C, it can be assumed that the Tg is inferior than caprolactone owed to a larger content of hydrocarbonated chains. The increase in Tg of the thiol–ene films with the removal of caprolactone indicates that this is caused by and increased cross-link density. The film, f-TE(G:C) 10:90, also exhibits a melt temperature around 50°C, characteristic of caprolactone, that is not observed in the other films containing smaller amounts. Additionally, this creates a large amount of noise and the temperature region was omitted in the plotting of the tan δ (Figure 4.11). The higher rubbery plateau of the f-TE(G:C) 100:0 in comparison to f-TE(G:C) 50:50 is due to a high cross-link density in comparison to the sample containing only 50 mol.% globalide. The tan δ plot clearly confirms this increase in Tg with increasing globalide content, but also the half-width of the Tg peaks increases with caprolactone addition indicating that a more inhomogeneous cross-linked network is formed [74]. However, there is no demarked difference in the two films with 50 and 90 mol.% caprolactone in the macrolactone based copolymer. Furthermore, the film containing 50 mol.% caprolactone shows a bimodal band which could be the result of a small degree of crystallinity that was undetected by DSC.
36
Com
posi
tion
Tg
( °C
)T
m( °
C)
Cry
stal
linem
elte
xoth
erm
(J/g
)
Poly
mer
Hom
o-en
efil
ms
TE-
film
sPo
lym
erH
omo-
ene
film
sT
E-fil
ms
Poly
mer
Hom
o-en
efil
ms
TE-
film
s
G:C
100
:0⎯
⎯-3
246
.244
.3⎯
73.7
56⎯
G:C
50:
50-4
9⎯
-46
31.9
33⎯
39.9
47⎯
G:C
40:
60⎯
⎯-4
229
.526
.8⎯
50.7
44⎯
G:C
30:
70⎯
⎯-4
634
.730
.8⎯
6356
⎯G
:C 2
0:80
⎯⎯
⎯38
.634
.5-2
60.9
576
G:C
10:
90⎯
⎯⎯
45.8
4228
55.6
5529
Tab
le 4
. The
rmal
pro
perti
es a
s det
erm
ined
from
DSC
mel
ting
curv
es (s
econ
d he
atin
g ru
n).
36
37
Figure 4.10 Dynamic mechanical analysis plots of storage modulus as a function of temper-ature for thiol–ene networks photocured from polymers bearing different comonomer ratios.
Figure 4.11 Representative plots of loss tangent (tan δ) vs. temperature for thiol–ene films photocured from different copolymers as measured by DMTA.
38
5. CONCLUSIONS
n Paper I the effect of initial cis/trans-configurations of FAMEs on their reactivi-ty with a tri-functional thiol was evaluated in bulk conditions via the thiol–ene
reaction. The stoichiometric reaction of an internal cis-alkene (MO) with TMP-trimercapto propionate was found to consist of a sequential insertion–isomerization–elimination process catalyzed by photogenerated thiyl radicals, which led to a fast formation of the corresponding trans-isomer. The reversible cis/trans-isomerization dominated the early instants of the reaction and was subsequently followed by thiol addition yielding the final thiol–ene product without major influ-ence of side-reactions. When a trans-alkene (ME) was used instead, thiol addition took place at a much higher rate showing insignificant conversion into the cis-isomer. Since isomerization occurred more rapidly than thiol addition, it was found that the rate-limiting step, in the two cases, was the hydrogen abstraction from the labile thiol group promoted by the intermediate alkyl radical. This indicated that although trans-structures have kinetic preference in the formation of end-product, aliphatic cis-alkenes can still be employed as starting substrates since isomerization exceeds hydrogen transfer and the slowest step controls the overall reaction. Even though no cross-linked structures were obtained at this point and the reduced reac-tion rates in comparison with monosubstituted olefins, this study provides relevant information about the reactivity of monounsaturated fatty acids (and related com-pounds) as potential bio-based monomers for the thiol–ene reaction.
As a follow-up study to the first article, Paper II dealt with the investigation of thin-film formation by using a library of lipase synthesized poly(globalide-caprolactone) copolymers, P(GI-co-CL), bearing different stoichiometric composi-tions in internal main-chain double-bonds. All the unsaturated random heteropol-ymers were successfully cross-linked in the melt with TMP-trimercapto propionate via the photoinduced thiol–ene reaction affording fully amorphous and insoluble elastomeric materials with different thermal and mechanical properties depending on functionality. The addition of ε-caprolactone to globalide represents a good method to increase chain mobility, while providing high conversions (>80%) and maintaining the cure behaviour irrespective of the functionality, although at the expense of some residual crystallinity and more heterogeneous cross-linked network. The reasonable reaction rates achieved demonstrates the capability of the thiol–ene reaction in curing these functional groups in a timely fashion.
I
39
To summarize, it is clear from both studies that the thiol–ene coupling reaction represents a chemically simplistic tool for the efficient cross-linking of internal 1,2-disubstituted alkenes without a need for pre-modification of the double-bonds. However, this is a relatively under studied area due to the general assumption that 1,2-disubstituted enes are very slow in reacting. The results presented in this re-search serve as model example for the sustainable use of unsaturated C=C com-pounds from renewable resources in making semi-synthetic thermosets with high conversions and a range of properties. Potential application areas include: polymeric coatings, packaging materials and biodegradable polymers for medical purposes.
40
6. FUTURE WORK
he research conducted in this contribution underlines the importance of utili-zation of internal main-chain alkenes in the thiol–ene reaction despite their
usual slow reactive nature. Additionally, this type of chemistry has demonstrated to be a versatile and robust step-wise polymerization route by exploiting this class of enes thus serving as an important ‘green’ platform for future studies. It would how-ever be of interest to pursue different classes of bio-based olefins, such as terpenes (e.g., limonene) and itaconates (e.g., itaconic acid), and assess if they are also suita-ble candidates for this reaction as polymer (or network) forming monomers, investi-gating: (a) reactivity (mechanism and kinetics), (b) effect of substituints, (c) occur-rence of simultaneous competitive reactions, and (d) impact on the thermal and viscoelastic properties of the resulting cross-linked materials. For example, prelimi-nary studies carried out in-house (unpublished results) have revealed for the case of limonene (which contains two non-isomerizable double-bonds) that it is possible to functionalize the –SH groups of TMP-trimercapto propionate selectively at the terminal double-bond of R-(+)-limonene via the thiol–ene reaction if the monoter-pene is charged in high stoichiometric excess with respect to the thiol functionality (Scheme 6.1.). This allows, after proper purification, the remaining unreacted endo-cyclic unsaturation – as represented in the scheme – to be characterized in a second reaction step with respect to reactivity and efficiency in network formation processes involving multifunctional thiols, further illustrating the possible usability of natural-ly occurring terpenes in thermoset synthesis.
R-(+)-limoneneTMP–trimercaptopropionate
AIBN or DMPA
ethyl acetate, Δ or UV
Thiol–ene coupledproduct
Scheme 6.1. Selective thiol–ene functionalization of the trithiol cross-linker at the exo-olefinic bond of limonene.
Itaconic acid is another promising building-block that can be obtained in kiloton quantities from the glucose fermentation industry [75]. However, it has been re-ported for this unsaturated di-carboxylic acid that structural isomerization–
T
41
rearrangement to citraconic and mesaconic acids takes place in aqueous solution when elevated temperatures are employed (Scheme 6.2.). The dynamic equilibrium established for the itaconate moiety is proposed to affect the overall rate of addition adduct formation and is thus expected to have a significant impact in a polyfunc-tional network formation system. Therefore, it would be of interest to examine the influence of this rearrangement, if it still occurs in bulk conditions, in addition to the homopolymerization propensity of the double-bond, for thiol–ene photopoly-merizations and assess its potential use in the creation of non-toxic biodegradable thiol–ene networks aiming, for instance, at biomedical applications.
itaconic acid citraconic acid
mesaconic acid
Scheme 6.2. Thermal isomerization-rearrangement equilibrium established between itaconic acid and its two isoforms citraconic and mesaconic acids [76].
42
7. ACKNOWLEDGEMENTS
n first place, I wish to express my deepest gratitude to my main supervisor Professor Mats Johansson and co-supervisor Professor Mats Jonsson for accepting me as a
PhD-student at the Division of Coating Technology of KTH. Your advisory efforts and patience in dealing with my questions have been nothing short than exceptional. Thank you both for the excellent scientific guidance, support and encouragement over these three years. I have learnt so much from you!
I am also very grateful to The Swedish Research Council (VR) for the financial support of this project.
Professor Anders Hult, Professor Eva Malmström, Professor Mats Johansson, Assistant Professors Michael Malkoch and Anna Carlmark Malkoch are gratefully acknowledged for promoting an outstanding and stimulating research environment within the group where students are encouraged to develop their own initiative, creativi-ty and self-confidence as scientists. A dedicated word of acknowledgement goes to Prof. Eva Malmström for giving me the chance to come to Sweden under the BioMaDe pro-ject and for introducing me to the extraordinary field of polymer science. Inger is thanked for all the help with the administrative tasks and practical issues. I would also like to thank Sonny Jönsson for the helpful discussions and knowledge regarding photo-chemistry.
I thank all members of the Department of Fibre and Polymer Technology, in particular to my ‘Ytgruppen’ friends, colleagues and roommates (former and present) for creating such a nice and pleasant workplace atmosphere; for all the parties, the funny ‘gris’ soccer kicks after lunch, ‘fika’ breaks, and the great joyful moments spent outside the laboratory.
A sincere thank you to Dr. Stacy M. Trey, for all your friendship, fruitful dis-cussions and collaboration which resulted in significant contributions to Paper II.
Alireza Salehi is greatly acknowledged for sharing with me the responsibility on the FTIR/Raman instrumentation and for being a great friend. Thank you for all your help.
Huge thanks to my family, especially my parents, Eduardo and Mariana, who have always supported my desire to pursue a higher degree education, and have coped well with losing their son to Sweden.
And last, but not least, I would like to say thanks to my dear wife, Leila, for all your love, never-ending support and kindness – You are the best Wife a man like me can wish for!
Obrigado! Stockholm, August 22nd of 2011 Mauro Claudino
I
43
8. REFERENCES
1. Odian, G., Principles of Polymerization. 4th Ed., 2004: Johan Wiley & Sons, Inc., Hoboken, N.J. 768.
2. Chanda, M., Introduction to Polymer Science: a problem solving approach, ed. T.F. Group, 2006, New York: CRC Press. 640.
3. Desroches, M., et al., Synthesis of Biobased Polyols by Thiol−Ene Coupling from Vegetable Oils. Macromolecules, 2011. 44(8): p. 2489-2500.
4. Samuelsson, J., et al., Thiol-ene coupling reaction of fatty acid monomers. Journal of Polymer Science Part A: Polymer Chemistry, 2004. 42(24): p. 6346-6352.
5. Stevens, M.P., Polymer Chemistry1999, New York: Oxford University Press. 6. Posner, T., Beitrage zur Kenntniss der ungesattigten Verbindungen. 11. Ueber die
addition von meroaptsnen an ungesättigte kohlenwasserstoffe. Ber Dtsch Chem Ges, 1905. 38: p. 646-657.
7. Braun, J.V. and R. Murjahn, Haftfestigkeit organischer reste (IV). Chem Ber, 1926. 59: p. 1202-1209.
8. Kharasch, M.S., A.T. Read, and F.R. Mayo, Chem. and Ind., 1938(57): p. 752.
9. Griesbaum, K., Problems and possibilities of the free-radical addition of thiols to unsaturated compounds. Angew. Chem., Int. Ed. Engl., 1970. 9(4): p. 273-87.
10. Jacobine, A.F., Thiol-ene photopolymers. Thiol-Ene Photopolymers, ed. J.D.R. Fouassier and F.F. Rabek. Vol. Chapter 7. 1993, London: Elsevier. 219-268.
11. Chatgilialoglu, C., M.P. Bertrand, and C. Ferreri, Sulfur-centered radicals in organic synthesis. S-Cent. Radicals, 1999: p. 311-354.
12. Hoyle, C.E., A.B. Lowe, and C.N. Bowman, Thiol-click chemistry: a multifaceted toolbox for small molecule and polymer synthesis. Chemical Society reviews, 2010. 39(4): p. 1355-87.
13. Magenau, A.J.D., et al., Facile polyisobutylene functionalization via thiol-ene click chemistry. Polym. Chem., 2010. 1(6): p. 831-833.
14. Hoyle, C.E., T.Y. Lee, and T. Roper, Thiol-enes: Chemistry of the past with promise for the future. Journal of Polymer Science Part A: Polymer Chemistry, 2004. 42(21): p. 5301-5338.
15. Hoyle, C.E. and C.N. Bowman, Thiol-ene click chemistry. Angewandte Chemie, 2010. 49(9): p. 1540-73.
16. Gress, A., A. Voelkel, and H. Schlaad, Thio-Click Modification of Poly[2-(3-butenyl)-2-oxazoline]. Macromolecules (Washington, DC, U. S.), 2007. 40(22): p. 7928-7933.
44
17. Gress, A. and H. Schlaad, Thio-click polymer chemistry. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 2008. 49(1): p. 230-231.
18. Killops, K.L., L.M. Campos, and C.J. Hawker, Robust, Efficient, and Orthogonal Synthesis of Dendrimers via Thiol-ene "Click" Chemistry. J. Am. Chem. Soc., 2008. 130(15): p. 5062-5064.
19. Kade, M.J., D.J. Burke, and C.J. Hawker, The power of thiol-ene chemistry. J. Polym. Sci., Part A Polym. Chem., 2010. 48(4): p. 743-750.
20. Fossey, J., D. Leffort, and J. Sorba, Free Radicals in Organic Chemistry, 1995: John Wiley & Sons.
21. Parsons, A.F., An Introduction to Free Radical Chemistry, 2000: Blackwell Science. 238.
22. Reddy, S.K., N.B. Cramer, and C.N. Bowman, Thiol-Vinyl Mechanisms. 1. Termination and Propagation Kinetics in Thiol-Ene Photopolymerizations. Macromolecules, 2006. 39(10): p. 3673-3680.
23. Reddy, S.K., N.B. Cramer, and C.N. Bowman, Thiol-Vinyl Mechanisms. 2. Kinetic Modeling of Ternary Thiol-Vinyl Photopolymerizations. Macromolecules, 2006. 39(10): p. 3681-3687.
24. Reddy, S.K., O. Okay, and C.N. Bowman, Network Development in Mixed Step-Chain Growth Thiol-Vinyl Photopolymerizations. Macromolecules, 2006. 39(25): p. 8832-8843.
25. Chatgilialoglu, C., A. Altieri, and H. Fischer, The Kinetics of Thiyl Radical-Induced Reactions of Monounsaturated Fatty Acid Esters. Journal of the American Chemical Society, 2002. 124(43): p. 12816-12823.
26. Chatgilialoglu, C., et al., The kinetics of Z/E isomerization of methyl oleate catalyzed by photogenerated thiyl radicals. Chemphyschem: a European journal of chemical physics and physical chemistry, 2005. 6(2): p. 286-91.
27. Walling, C. and W. Helmreich, Reactivity and reversibility in the reaction of thiyl radicals with olefins. J. Am. Chem. Soc., 1959. 81: p. 1144-8.
28. Chatgilialoglu, C., et al., Geometrical isomerism of monounsaturated fatty acids: thiyl radical catalysis and influence of antioxidant vitamins. Free Radical Biol. Med., 2002. 33(12): p. 1681-1692.
29. Sgoutas, D.S. and F.A. Kummerow, Cis-trans isomerization of unsaturated fatty acid methyl esters without double bond migration. Lipids, 1969. 4(4): p. 283-7.
30. Ferreri, C. and C. Chatgilialoglu, Geometrical trans lipid isomers: a new target for lipidomics. Chembiochem: a European journal of chemical biology, 2005. 6(10): p. 1722-34.
45
31. Chatgilialoglu, C., et al., (Z)-(E) Interconversion of Olefins by the Addition-Elimination Sequence of the (TMS)3Si.bul. Radical. J. Org. Chem., 1995. 60(12): p. 3826-31.
32. Chatgilialoglu, C. and C. Ferreri, Trans Lipids: The Free Radical Path. Acc. Chem. Res., 2005. 38(6): p. 441-448.
33. Ferreri, C., M. Panagiotaki, and C. Chatgilialoglu, Trans Fatty Acids in Membranes: The Free Radical Path. Molecular Biotechnology, 2007. 37(1): p. 19-25.
34. Belfield, K.D. and J.V. Crivello, Photoinitiated Polymerization. ACS Symposium Series 847, ed. K.D. Belfield and J.V. Crivello, 2003: American Chemical Society (ACS), Distributed by Oxford University Press. 565.
35. O'Brien, A.K., N.B. Cramer, and C.N. Bowman, Oxygen inhibition in thiol-acrylate photopolymerizations. J. Polym. Sci., Part A Polym. Chem., 2006. 44(6): p. 2007-2014.
36. Cramer, N.B., et al., Initiation and kinetics of thiol-ene photopolymerizations without photoinitiators. Journal of Polymer Science Part A: Polymer Chemistry, 2004. 42(22): p. 5817-5826.
37. Shin, J., S. Nazarenko, and C.E. Hoyle, Enthalpy Relaxation of Photopolymerized Thiol-Ene Networks: Structural Effects. Macromolecules (Washington, DC, U. S.), 2008. 41(18): p. 6741-6746.
38. Meier, M.A.R., J.O. Metzger, and U.S. Schubert, Plant oil renewable resources as green alternatives in polymer science. Chemical Society Reviews, 2007. 36(11): p. 1788.
39. Montero de Espinosa, L. and M.A.R. Meier, Plant oils: The perfect renewable resource for polymer science?! European Polymer Journal, 2011. 47(5): p. 837-852.
40. Belgacem, M.N., A. Gandini, Monomers, Polymers and Composites from Renewable Resources, 2008. 552.
41. Black, M. and J.W. Rawlins, Thiol–ene UV-curable coatings using vegetable oil macromonomers. European Polymer Journal, 2009. 45(5): p. 1433-1441.
42. Bexell, U., et al., A tribological study of a novel pre-treatment with linseed oil bonded to mercaptosilane treated aluminium. Surf. Coat. Technol., 2003. 166(2-3): p. 141-152.
43. Bantchev, G.B., et al., Free Radical Addition of Butanethiol to Vegetable Oil Double Bonds. J. Agric. Food Chem., 2009. 57(4): p. 1282-1290.
46
44. Türünç, O. and M.A. Meier, Fatty Acid Derived Monomers and Related Polymers Via Thiol-ene (Click) Additions. Macromolecular rapid communications, 2010. 31(20): p. 1822-6.
45. Ates, Z., P.D. Thornton, and A. Heise, Side-chain functionalisation of unsaturated polyesters from ring-opening polymerisation of macrolactones by thiol–ene click chemistry. Polymer Chemistry, 2011. 2(2): p. 309.
46. Stemmelen, M., et al., A fully biobased epoxy resin from vegetable oils: From the synthesis of the precursors by thiol-ene reaction to the study of the final material. Journal of Polymer Science Part A: Polymer Chemistry, 2011.
47. Türünç, O., et al., Clicking renewable resources: thiol-ene additions for the synthesis of monomers and polymers derived from plant oils and terpenes. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 2010. 51(2): p. 724-725.
48. Lluch, C., et al., Rapid Approach to Biobased Telechelics through Two One-Pot Thiol-Ene Click Reactions. Biomacromolecules, 2010. 11(6): p. 1646-1653.
49. Chen, Z., et al., Soy-based UV-curable thiol–ene coatings. Journal of Coatings Technology and Research, 2010. 7(5): p. 603-613.
50. Türünç, O. and M.A.R. Meier, Thiol-ene vs. ADMET: a complementary approach to fatty acid-based biodegradable polymers. Green Chemistry, 2011. 13(2): p. 314.
51. van der Meulen, I., et al., Polymers from Functional Macrolactones as Potential Biomaterials: Enzymatic Ring Opening Polymerization, Biodegradation, and Biocompatibility. Biomacromolecules, 2008. 9(12): p. 3404-3410.
52. van der Meulen, I., et al., Copolymers from unsaturated macrolactones: toward the design of cross-linked biodegradable polyesters. Biomacromolecules, 2011. 12(3): p. 837-43.
53. http://www.gepasi.org/. 54. Mendes, P., GEPASI: a software package for modelling the dynamics, steady states
and control of biochemical and other systems. Comput Appl Biosci, 1993. 9(5): p. 563-71.
55. Mendes, P., Biochemistry by numbers: simulation of biochemical pathways with Gepasi 3. Trends Biochem. Sci., 1997. 22(9): p. 361-363.
56. Mendes, P., GEPASI, ver. 3.30, 2002, The University of Walles, Aberystwyth, UK.
57. Petzold, L.A.H., LSODA (Livermore Solver of Ordinary Differential Equations). Computing and Mathematics Research Division. Lawrence Livermore National Laboratory1997, Livermore, CA.
47
58. Lalevee, J., et al., Role of the Medium on the Reactivity of Cleavable Photoinitiators in Photopolymerization Reactions. Macromolecules, 2006. 39(5): p. 1872-1879.
59. Caceres, T., M.V. Encinas, and E.A. Lissi, Photocleavage of benzil. J. Photochem., 1984. 27(1): p. 109-14.
60. Allen, N.S., P.N. Green, and W.A. Green, Photochemistry of carbonyl photoinitiators. Photopolymerization, flash photolysis and spectroscopic study. Eur. Polym. J., 1986. 22(1): p. 49-56.
61. Gruber, H.F., Photoinitiators for free radical polymerization. Prog. Polym. Sci., 1992. 17(6): p. 953-1044.
62. Segurola, J., et al., Photoyellowing and discoloration of UV cured acrylated clear coatings systems: influence of photoinitiator type. Polym. Degrad. Stab., 1999. 64(1): p. 39-48.
63. Segurola, J., et al., Photochemistry and photoinduced chemical crosslinking activity of acrylated prepolymers by several commercial type I far UV photoinitiators. Polym. Degrad. Stab., 1999. 65(1): p. 153-160.
64. Cramer, N.B. and C.N. Bowman, Kinetics of thiol-ene and thiol-acrylate photopolymerizations with real-time fourier transform infrared. Journal of Polymer Science Part A: Polymer Chemistry, 2001. 39(19): p. 3311-3319.
65. Roper, T., Photopolymerization of pigmented thiol–ene systems. Polymer, 2004. 45(9): p. 2921-2929.
66. Roper, T.M., et al., Influence of the alkene structure on the mechanism and kinetics of thiol-alkene photopolymerizations with real-time infrared spectroscopy. Journal of Polymer Science Part A: Polymer Chemistry, 2004. 42(24): p. 6283-6298.
67. Matsushima, H., et al., Thiol-isocyanate-acrylate ternary networks by selective thiol-click chemistry. J. Polym. Sci., Part A Polym. Chem., 2010. 48(15): p. 3255-3264.
68. Decker, C. and T. Nguyen Thi Viet, Photocrosslinking of functionalized rubbers. Part 7. Styrene-butadiene block copolymers. Macromol. Chem. Phys., 1999. 200(2): p. 358-367.
69. Decker, C. and T.N.T. Viet, Photocrosslinking of functionalized rubbers. Part 8. The thiol-polybutadiene system. Macromol. Chem. Phys., 1999. 200(8): p. 1965-1974.
70. Decker, C. and T.N.T. Viet, High-speed photocrosslinking of thermoplastic styrene-butadiene elastomers. J. Appl. Polym. Sci., 2000. 77(9): p. 1902-1912.
48
71. Decker, C. and T.N.T. Viet, Photocrosslinking of functionalized rubbers IX. Thiol-ene polymerization of styrene-butadiene-block-copolymers. Polymer, 2000. 41(11): p. 3905-3912.
72. Jacobine, A.F., et al., Photocrosslinked norbornene-thiol copolymers: synthesis, mechanical properties, and cure studies. J. Appl. Polym. Sci., 1992. 45(3): p. 471-83.
73. Johansson, M. and A. Hult, Influence of allyl ethers in coating resins. J. Polym. Sci., Part A Polym. Chem., 1991. 29(11): p. 1639-44.
74. Park, J.-G., et al., Dynamic mechanical analysis and esterase degradation of dentin adhesives containing a branched methacrylate. J. Biomed. Mater. Res., Part B, 2009. 91B(1): p. 61-70.
75. Molhoek, L.J., et al. Isomerization of Itaconic Acid During Direct Polycondensation. in Polycondensation 2010 Meeting. 2010. Roldduc Abey, The Netherlands.
76. Sakai, M., Studies of the isomerization of unsaturated carboxylic acids. II. The thermal rearrangement of citraconic acid to itaconic acid in aqueous solutions. Bull. Chem. Soc. Jpn., 1976. 49(1): p. 219-23.
49
Corrections to Paper I
• Page 2322 in the Materials subsection where it is written ‘Ciba Speciality Chem-icals’ should be read ‘Ciba Specialty Chemicals’.
• Table 1 (page 2325) for the isomerization rate coefficients:
Correct: = × 7
f 2.0 10Zk s–1 (incorrect: = × 7
a 2.0 10Zk s–1);
Correct: = × 5
a 2.9 10Ek M–1⋅s–1 (incorrect: = × 5
f 2.9 10Ek M–1⋅s–1).
• Page 2326 in first paragraph where it is written ‘mol L–1⋅s–1’ should be replaced
by ‘mol⋅L–1⋅s–1’.
• Page 2326 in second paragraph where it is written ‘TMP-mercapto propionate’ should be read ‘TMP-trimercapto propionate’.