Biodegradation and ageing of bio-based thermosetting resins from lactic acid

(1)

B IODEGRADATION AND

A ^{GEING OF} B ÎO -B ÂSED T HERMOSETTING R ÊSINS

FROM L ^ACTIC A ^CID

MSc in Resource Recovery Technology Lara Lopes Gomes Hastenreiter

(2)

Programme: Master in Resource Recovery

English title: Biodegradation and ageing of bio-based thermosetting resins from lactic acid

Authors: Lara Lopes Gomes Hastenreiter Supervisor: Dan Åkesson and Akram Zamani Examiner: Dan Åkesson

Date: 2019-06-20

Key words: bio-based polymer, lactic acid, glycerol, ethylene glycol, pentaerythritol, biodegradation, ageing, characterisation, DSC, TGA, FTIR, tensile.

_________________________________________________________________

Abstract

The need for replacing petroleum-based polymers has been increasing and bio-based polymers prove to be a suitable solution. The aim of this thesis was to synthesize bio-based resins with different chemical architectures to evaluate the effect of the structure on the properties and on their response to ageing and biodegradation. For this, three different bio- based thermoset resins have been synthesised by reacting one of three distinct core-molecules with lactic acid. The options of core-molecules chosen for this work were ethylene glycol, glycerol and pentaerythritol. Lactic acid was first reacted with a core-molecule by direct condensation, the resulting branched molecule was then end-functionalized with methacrylic anhydride. The amount of moles of lactic acid varied according to which core-molecule it was reacted with, but the chain length (n) was always maintained as three. Part of the samples were characterised by Fourier-transform infrared spectroscopy (FT-IR), differential scanning calorimetry (DSC), thermogravimetric analysis (TGA) and tensile test. DSC and TGA were used for determining the thermal behaviour. FT-IR was used to verify the first and second stage of the reaction and to ascertain the occurrence of the crosslinking reaction. Tensile test was done for investigation of mechanical properties. The ageing and biodegradation tests are useful to ascertain the material possible applications. Therefore, the samples that went through the process of ageing or biodegradation were also characterised in the end of the procedures to further check the effect of those processes on the specimens. The test results indicated that the PENTA/LA cured resin was the most stable thermally. The cured resin’s mechanical properties were similar to each other, so there was no comparison to make in this area. The samples proved to be affected by the biodegradation and the ageing processes, both in visual and structural aspects.

(3)

TABLE OF FIGURES

Figure 1. Material coordinate system of bioplastic (European bioplastics, 2016) ... - 1 -

Figure 2. Global production capacities of bio-based polymers by region in 2018 and 2023 (excluding polyurethanes, epoxy resins and cellulose acetate) (Chinthapalli, 2019). ... - 2 -

Figure 3. Shares of the produced bio-based polymers in different market segments in 2018 and 2023 (Chinthapalli, 2019). ... - 3 -

Figure 4. Lactide production routes from biomass-derived lactic acid (Upare et al., 2016). - 4 - Figure 5. Synthesis of itaconic acid-based polyesters (Dai et al., 2015). ... - 6 -

Figure 6. Reaction scheme of polycondensation of itaconic acid with 1,2-propandiol, and molecular structure of reactive diluents (Panic et al., 2017). ... - 7 -

Figure 7. Reaction schemes for (a) the sysnthesis of star-Ita-Gly resins, and (b) end- functionalization with ethanol to synthesis Tstar-Ita-Gly resins (Jahandideh, Esmaeili and Muthukumarappan, 2018). ... - 8 -

Figure 8. The FTIR spectra of the Tstar-Ita-Gly resin and the Tstar-Ita-Gly cured sample (Jahandideh, Esmaeili and Muthukumarappan, 2018). ... - 8 -

Figure 9. The synthetic route to EIA (Ma et al., 2013). ... - 9 -

Figure 10. Allylation route of itaconic acid and Synthetic route of trifunctional epoxy resin from itaconic acid (TEIA) (Kumar at al., 2018). ... - 9 -

Figure 11. Chemical structure of a triglyceride molecule (Adekunle, Åkesson and Skrifvars, 2010). ... - 10 -

Figure 12. Structure of a synthesized glycolide/lactide based polyester resin (Xie et al., 2007). ... - 11 -

Figure 13. Structure of poly(lactide-co-propylene glycol) dimethacrylate adhesives (Ho and Young, 2006). ... - 11 -

Figure 14. Idealized structure of the obtained end-functionalized oligomer (Åkesson et al., 2010). ... - 12 -

Figure 15. Thermogravimetric analysis of the cross-linked resin (Åkesson et al., 2010). .. - 12 -

Figure 16. Molecular structures of butanediol, pentaerythritol and polyglicerine-06 and -10 (Helmine, Korhonen and Seppala, 2002). ... - 13 -

Figure 17. Functionalization of lactic acid oligomers with methacrylic anhydride and crosslinking scheme (Helmine, Korhonen and Seppala, 2002)... - 13 -

Figure 18. FT-IR spectra of GLY/LA resin with chain length n = 3. The spectra for first synthesis step, last synthesis step and for cured resin are shown (Bakare, 2015). ... - 14 -

Figure 19. Stress at break of bidirectional ﬁber type composites with 65 % ﬁber load with and without styrene before and after ageing in climate chamber (Esmaeili, 2015). ... - 15 -

Figure 20. Synthesis scheme (Chen et al., 2014). ... - 16 -

(5)

Figure 21. Chemical composition of polylactide (PLA) and polyhydroxyalkanoate (PHA) (Masek and Latos-Brozio, 2018). ... - 16 - Figure 22. Chemical composition of rutin and hesperidin(Masek and Latos-Brozio, 2018).- 17 -Figure 23 . Chemical structure of 3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionic acid (IrganoxCOOH) and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) (Pérez Amaro et al., 2016) ... - 18 - Figure 24. Tensile strength of dry samples after 1000 h of ageing, and after 24 h of reconditioning (Åkesson et al., 2011). ... - 19 - Figure 25. Tensile modulus of dry samples after 1000 h of ageing, and after 24 h of reconditioning (Åkesson et al., 2011). ... - 19 - Figure 26. Hydrolysis mechanism of PLA in soil (Lv et al., 2017). ... - 20 - Figure 27. Surface morphologies of starch/wood flour/PLA blends before and after the degradation in outdoor soil (Lv et al., 2017). ... - 20 - Figure 28. The evolution of mechanical strength of the starch/wood flour/PLA blends during the soil burial degradation process (Lv et al., 2017). ... - 21 - Figure 29. FTIR-ATR spectra of PLA 75 film: initial (top) and after 11 months of soil burial (bottom) in Spata (Rudnik and Briassoulis, 2011). ... - 21 - Figure 30. Soil burial experiment plot ground – Buried samples in the soil (Tajau et al., 2016).

... - 23 - Figure 31. Percentage of weight loss in palm oil-based films and petrochemical-based films after 12 months of biodegradation – Effect of oligomer and photoinitiator (Tajau et al., 2016).

... - 23 - Figure 32. Structural formulas of ethylene glycol, glycerol and pentaerythritol (Sigma Aldrich, 2018). ... - 23 - Figure 33. Reaction scheme of first and second stages for the synthesis of methacrylate functionalized ethylene glycol/lactic acid resins. ... - 24 - Figure 34. Reaction scheme of first and second stages for the synthesis of glycerol/lactic acid resins with lactic acid chain lengths of 3. ... - 24 - Figure 35. Reaction scheme of first and second stages for the synthesis of methacrylate functionalized pentaerythritol/lactic acid resins. ... - 25 - Figure 36. Cured resin in dog bone mold (Photo Lara Hastenreiter). ... - 29 - Figure 37. FT-IR spectra of EG/LA, GLY/LA and PE/LA resins with chain length n = 3. The spectra for first stage, second stage and for cured resin are shown... - 29 - Figure 38. FT-IR spectra of EG/LA, GLY/LA and PE/LA cured resins with chain length n = 3 ... - 30 - Figure 39. DSC analysis for EG/LA, GLY/LA and PE/LA cured resins with chain length n = 3. ... - 30 - Figure 40. Thermogravimetric analysis of the cross-linked resins. ... - 31 - Figure 41. Curve Stress versus Strain for cured EG/LA resin. ... - 32 - Figure 42. Comparison of mechanical properties of each resin type: (a) Tensile modulus; (b) Tensile strength at break; (c) Maximum elongation at break. ... - 33 - Figure 43. DSC curves for cured samples after the ageing process. ... - 34 - Figure 44. Comparison of FT-IR spectra of EG/LA, GLY/LA and PENTA/LA cured resins without and with ageing influence. ... - 35 - Figure 45. Curve Stress versus Strain for cured EG/LA resin after ageing. ... - 35 - Figure 46. Comparison of tensile results for samples before and after ageing: (a) Tensile modulus; (b) Tensile strength at break; (c) Maximum elongation at break. ... - 36 - Figure 47. Visual aspects of specimens before and after the biodegradation process: (a) EG/LA; (b) GLY/LA; (c) PENTA/LA (Photo Lara Hastenreiter). ... - 37 -

(6)

Figure 48. Microscopic aspects of specimens before and after the biodegradation process: (a) EG/LA; (b) GLY/LA; (c) PENTA/LA (Photo Lara Hastenreiter) ... - 38 - Figure 49. Weight average before and after biodegradation. ... - 38 - Figure 50. Heat flow versus temperature curves for EG/LA, GLY/LA and PENTA/LA cured resins after biodegradation process. ... - 39 - Figure 51. Comparison between EG/LA, GLY/LA and PENTA/LA cured resins without and with biodegradation influence. ... - 39 - Figure 52. Comparison of FT-IR spectra of EG/LA, GLY/LA and PENTA/LA cured resins without and with biodegradation influence. ... - 40 -

TABLE OF TABLES

Table 1. Thermal characterization results for the resins (Bakare, 2015). ... - 14 - Table 2. Thermal characterization results for the cured resins... - 31 - Table 3. Comparison of the tensile modulus, tensile strength at break and maximum elongation at break for each resin type. ... - 33 - Table 4. Comparison of the tensile modulus, tensile strength at break and maximum elongation at break for each resin type before and after ageing ... - 36 -

(7)

- 1 -

1. INTRODUCTION 1.1. Bio-based resin

In recent years, bio-based resins, prepared from renewable resources, have been catching attention from both the academic and industrial fields, in order to substitute petroleum-based polymers (Avella et al., 2009). Petroleum-based materials have benefited humankind in various ways, by being disposable and highly durable. However, the massive use of fossil materials is becoming an issue in the polymer industry due to the continue deficiency of crude oil and its frequent price oscillation. Besides that, fossil resources raise several environmental concerns with sustainability, carbon dioxide emissions, disposal, and recyclability. In this context, efforts are on made to design polymeric materials derived from renewable resources instead, providing a sustainable development of economical and eco-friendly materials (Bakare et al., 2014).

The term bio-based was defined as “derived from biomass”, by the Technical Report 16575, drawn up by the European Committee for Standardization (CEN) in August 2014. The report also defines biomass as “material of biological origin such as plants, trees, algae, marine organisms, microorganisms, animals, etc.” It is important to notice that this definition does not entirely applies to ‘bioplastics’, a term that can generate some confusion with ‘bio-based’.

The prefix ‘bio-’ express a bio-functionality of the material, meaning that a plastic material can be defined as a bioplastic if it is either bio-based, biodegradable, or both (Kabasci, 2014).

Therefore, bio-based polymers are bioplastics but not all bioplastics are bio-based, as shown in Figure 1.

Figure 1. Material coordinate system of bioplastic (European bioplastics, 2016)

The bio-based resins can be used in many thermoplastic and thermoset applications.

Thermoplastic polymers can melt and flow when heated above the polymer’s melting point.

Thermoset, on the other hand, do not flow on heating, hence cannot be reshaped (Carraher, 2007). This different behavior of thermosets and thermoplastics when heated comes from their chemical structures. Thermosets are cross-linked polymers that remain on solid state as long as the covalent chemical bonds are intact (Guo, 2017). Thermoset polymers, in general,

(8)

- 2 -

have high strength and good thermal and chemical resistances. Therefore, they are broadly used in coatings, electronic packaging, adhesives, composites, and several other applications (Ma et al., 2016). Thermoset resins are attractive for use as bio-based composite applications, due to their low viscosity (Bakare et al., 2014). The production of thermosetting polymers is usually accomplished with two stages. The first stage consists in the production of an incompletely reacted pre-polymer. The other stage gives the final cross-linked product (Cowie, 2008). In this way, bio-based thermoset resins can be changed irreversibly by curing reactions, not being able to melt it down again (Ma, 2016).

Unsaturated polyester resins, epoxy resins and phenol formaldehyde resins are some examples of bio-based thermoset resins. Unsaturated polyesters are normally prepared by condensation of a dicarboxylic acid with one or more di-alcohols, such as ethylene glycol or propylene glycol. After the condensation reaction, the carbon–carbon double bonds are able to form a highly cross-linked network by free radical polymerization (Ma, 2016).

1.1.1. Bio-based market

Bio-based materials are rapidly catching industries attentions, since they minimize the risk to human health and the environment (Wool, 2005). Therefore, the bio-based polymer market have been growing globally. Goldstein Research released a report in 2018 about the Europe sustainable polymers market, informing that the production capacity of bio-based polymers was 4.2 million tons in 2016, and it increased at the rate of 4.7% annually from the period of 2014 to 2016 (Goldstein Research, 2018). According to data provided by the German Nova- Institute about the new market, the total production volume of bio-based polymers reached 7.5 million tons in 2018. The same report stated Asia as the leading region when it comes to installation of bio-based production capacities. Europe is next in line with a promising rise for 2023, as observed in Figure 2 (Chinthapalli, 2019).

Figure 2. Global production capacities of bio-based polymers by region in 2018 and 2023 (excluding polyurethanes, epoxy resins and cellulose acetate) (Chinthapalli, 2019).

According to this same trend report, this increase in production capacity is mainly because of the poly(lactic acid) (PLA) expanding market in Thailand and the polytrimethylene terephthalate (PTT) and starch blends in USA. Moreover, among all the bio-based polymers, PLA and starch blends display a promising growth until 2023. However, the bio-based polymers market remains challenging due to the low prices on crude oil and little political support (Chinthapalli, 2019).

(9)

- 3 - 1.1.2. Bio-based resins applications

Many of the bio-based polymers have good film forming properties, being suitable for both high performance and traditional applications. Food containers, agriculture film and waste bags are some examples of where to apply them. Those polymers can also be widely found in medical field applications (Van de Velde and Kiekens, 2002). Figure 3 shows a summary of the applications for bio-based polymers in 2018 as well as a projection for 2023 (Chinthapalli, 2019). It is noticeable at the Figure that consumer goods made up the largest share of produced bio-based polymers with 28% in 2018. Building and construction sector follows with 21%, automotive and transport with 19%, packaging (flexible and rigid) with 15% and textiles (woven and nonwoven) with 11%. It is also informed by Figure 3 that no significant changes are expected for 2023 with regard to market application shares (Chinthapalli, 2019).

Figure 3. Shares of the produced bio-based polymers in different market segments in 2018 and 2023 (Chinthapalli, 2019).

1.2. Lactic Acid

Lactic acid (2-hydroxypropanoic acid) was identified in sour milk by the Swedish chemist Carl Wilhelm Scheele in 1780, and exists as two optical isomers: D- and L-lactic acid (Auras, 2010). Lactic acid can be produced by fermentation of sugar-based carbohydrates or by chemical synthesis. (Adekunle, Åkesson and Skrifvars, 2010). This organic acid is used industrially in several fields as pharmaceutical, food, textile and polymers, because of its optically active isomers and chemical structure (Yadav, Chaudhari and Kothari, 2011; Gao et al., 2012). Both isomeric forms of lactic acid can be polymerized into polymers with different properties (Hofvendahl and Hahn–Hägerdal, 2000). Lactic acid based polymers are a great substitute for petrochemical polymers in terms of environmental performance. Kozlovskiy, Shvets, and Kuznetsov (2017) reported that lactic acid can be used for the synthesis of biodegradable solvents as butyl lactate and for the production of nontoxic propylene glycol, for instance.

The demand for lactic acid has grown because of its use in the synthesis of poly(lactic acid) (PLA), a biodegradable and biocompatible polymer. According to Aeschelmann (2015), PLA is a well-established bio-based polymer produced by numerous companies worldwide. PLA is also known for having a reasonable price, competing well with fossil-based polymers (Belgacem, 2008). This polymer has been employed in many industrial applications, such as packaging – to reduce the risk of pathogen contamination in fruit and vegetable packages –

(10)

- 4 -

and biomedical materials (Farrán et al., 2015). Industrially, PLA is synthesised by ring- opening polymerization from lactide, a six-membered dimeric cyclic ester of (usually L-) lactic acid. Figure 4 shows the process to obtain lactide from lactic acid (Upare et al., 2016).

Figure 4. Lactide production routes from biomass-derived lactic acid (Upare et al., 2016).

As mentioned before, the PLA market plays a crucial role in the increase of bio-based polymers production capacity. The company named NatureWorks®, for instance, has the goal to “turn greenhouse gases into a portfolio of poly(lactic acid) (PLA) performance materials called Ingeo”. They create the lactic acid by submitting plants to a milling process that extracts the starch (glucose) which is then converted to dextrose via hydrolysis by enzymatic action. The fermentation of this dextrose results in lactic acid. After that, they transform the lactic acid molecules into rings of lactide which are opened and linked together to form the long chain of PLA polymer they call Ingeo. They process the Ingeo into pellets forms and sell it to other companies who transform those into products as coffee capsules, yogurt cups, baby wipes and appliances (NatureWorks, 2019).

1.3. Biodegradation

Biodegradation is a chemical process that consists in the breakdown of matter by microorganisms that are available in the environment, such as fungi and bacteria. This converts materials into lower molecular weight products like water and carbon dioxide (Garrison, 2016). The main processes of environmental biodegradation are hydrolysis, oxidation, and composting (Polymer Nanocomposites Biodegradation, 2016). The microorganisms are unable to transport polymeric materials directly into their cells, where most of the biochemical process take place. Instead, they excrete enzymes that depolymerise the polymers outside the cells. Those enzymes generally act only on the polymer surface, for being too large to penetrate deeply into the polymer (Biodegradable polymers-2., 2012).

The process of biodegradation does not depend on the resource basis of a material; it is, in reality, closely linked to its chemical structure. Which makes it understandable the fact that there are bio-based plastics that are non-biodegradable, as well as fossil-based plastics that biodegrades (European bioplastics, 2016). Some fossil-based polymers, like polycaprolactone (PCL), or poly(butylene adipate terephthalate) (PBAT), for instance, are biodegradable. On the other hand, polyethylene (PE), a bio-based polymer from sugar cane, is resistant to biodegradation, as many others (Kabasci, 2014).

(11)

- 5 -

The two categories of bioplastics that are interesting for this thesis is bio-based/non- biodegradable and bio-based/biodegradable. The first one includes polymers such as bio- based polyamides (PA), polyesters (e.g. PTT, PBT), polyurethanes (PUR) and polyepoxides.

The bio-based/biodegradable category consists of starch blends made of thermo-plastically modified starch and other biodegradable polymers as well as polyesters such as polylactic acid (PLA) or polyhydroxyalkanoate (PHA) (Chinthapalli, 2019). In the discussion of bio- based materials, the biodegradability is an important factor when it comes to applications. For structural applications, for instance, is preferable to have a more durable material, while for disposable goods it is desirable a biodegradable one (Garrison, 2016). Therefore, some typical technical applications for the bio-based/non-biodegradable group are textile fibers for seat covers and carpets, foams for seating, cables and covers used in the automotive industry. For those cases, as mentioned before, biodegradability is not a sought-after property, since the products operating life needs to last several years. For the second category of bioplastics, on the contrarily, the aim is to use them in short-lived materials applications such as packaging (European bioplastics, 2016). The fact that those biodegradable polymers are also bio-based, brings the advantage that they can biodegrade without leaving behind microplastics, unlike the fossil-based ones (Chinthapalli, 2019).

1.4. Ageing

Polymers are always exposed to various outer ageing factors during their use, like temperature, UV radiation, oxygen, ozone moisture, fuels, oils or chemicals, for instance (Johlitz, Diercks and Lion, 2014). This can affect their mechanical properties, such as stiffness and strength, and limit their operating time (Creus, 2015). There are several ageing mechanisms, as thermal decomposition, hydrolysis, photo- and thermo-oxidation (Arrigo, 2016). Those mechanisms have the purpose of checking the durability performance of the polymer. They do that by accelerating the natural ageing process of the material, so it quickly reaches an end-state as a real time aged material (Bakare, 2015). One of the most commonly used accelerative condition for polymer degradation is the thermo-oxidative ageing, where temperature effect is coupled with oxygen effect (Celina, 2013). Ouyang et al. (2017) stated that for most polymer types, oxidative ageing is dominated by oxidation reactions that can lead to chain scission resulting mainly in carbonyl groups. In the bio-based polymers area, aliphatic biopolymers such as PLA are characterised by having poor resistance to climatic factors, according to Masek and Latos-Brozio (2018). The same researches stated that the presence of polar ester groups in their chemical structure makes them more susceptible to hydrolysis.

1.5. Literature review of current research

This section presents a research in the field of bio-based resins in general and, subsequently, is given focus to lactic acid-based resins. Besides that, some characterization methods and ageing and biodegradation tests are reviewed.

1.5.1. Itaconic acid-based resins

(12)

- 6 -

Itaconic acid (IA), or methylenesuccinic acid, is obtained by fermentation of carbohydrates, such as glucose. This renewable source has been widely used in emulsion paints and paper coating latexes fields, as well as in other applications (Ma et al., 2013).

Dai et al. (2015) reported the synthesis of itaconic acid-based polyesters by pre- polymerization and polycondensation, as shown in Figure 5. Their chemical structures were confirmed with the help of FTIR and H-NMR, subsequently the resins were copolymerized with acrylated epoxidized soybean oil (AESO). The resulting thermosets had their thermal and mechanical properties studied by DSC, tensile testing and Dynamic Mechanical Analysis (DMA). They obtained a tensile strength of 2.9 MPa ± 0.6, young’s modulus of 33.3 MPa ± 1.1, elongation at break equals 12.4 % ± 2.2 and glass transition temperature (Tg) of 32.2 °C, for the AESO resin. Their results indicated that the polyesters derived from itaconic acid effectively enhanced the properties and the bio-based content of soybean oil-based thermosets.

Figure 5. Synthesis of itaconic acid-based polyesters (Dai et al., 2015).

Panic et al. (2017) also performed polycondensation of itaconic acid. Their aim was to prepare fully bio-based unsaturated polyester resins (UPRs) as alternate nonpetroleum-based monomers. The IA was mixed with 1,2-propandiol as core-molecule and diluted with dialkyl itaconates. The reaction was performed within a temperature range of 110−190 °C with temperature increase of 10 °C/hour, under a nitrogen atmosphere. After that, the resin was cooled down to 110 °C and dissolved in reactive diluents. Dimethyl, diethyl, diisopropyl and dibutyl itaconates, all bio-based in nature, were used as reactive diluents to investigate their possible role as styrene replacement for the bio-based UPRs synthesis. Figure 6 shows the reaction scheme as well as the molecular structure of the reactive diluents.

(13)

- 7 -

Figure 6. Reaction scheme of polycondensation of itaconic acid with 1,2-propandiol, and molecular structure of reactive diluents (Panic et al., 2017).

The researchers performed the curing process by adding methyl ethyl ketone peroxide (MEKPO) to the resin. The mixture was poured into Teflon molds and left in an air oven for 3h at 80 °C and 1 h at 120 °C. DMA analysis returned glass transition temperatures around 65−118 °C and tensile test of the cured samples showed moderate stiffness (270−660 MPa) and high break stress (21−54 MPa). By data analysis, it was concluded that the synthesized bio-based resin UPR-DMI was the one with most comparable applicative properties to the UPR with styrene. Besides that, this resin had the advantage of absence of toxic and volatile petrochemical components in the product.

Jahandideh, Esmaeili and Muthukumarappan (2018) synthesised a star-shaped bio-based thermoset resin by direct condensation reaction of itaconic acid and glycerol, named star-Ita- Gly. The end-functionalization was carried out with the addition of ethanol, in order to decrease the viscosity of the resin, producing the Tstar-Ita-Gly. This resin was then transferred to a rotary evaporator to remove the residual ethanol and toluene, under pressure of approximately 10 mbar, for 1 h at 90 °C. Figure 7 presents the reaction and end- functionalization scheme. Chemical structures of the resins were studied by performing ¹H and ¹³C NMR and FT-IR tests. Thermomechanical properties were obtained by DSC, DMA and TGA analyses. FT-IR results (Figure 8) confirmed the occurrence of the cross-linking process by the presence and the absence of the C=C peak at 1635 cm⁻¹ (stretching C=C) and 815 cm⁻¹ (bending CH2). On the TGA analysis the Tstar-Ita-Gly presented two major thermal decomposition stages with maximum rates at 200 and 380 °C. The test also showed a 25%

solid residue of Tstar-Ita-Gly at 450 °C, making this resin suitable for development of flame retardant systems, according to the authors. The Tg of this crosslinked thermoset was 122 °C, presented based on the peak of tan δ in DMA curves.

(14)

- 8 -

Figure 7. Reaction schemes for (a) the sysnthesis of star-Ita-Gly resins, and (b) end-functionalization with ethanol to synthesis Tstar-Ita-Gly resins (Jahandideh, Esmaeili and Muthukumarappan, 2018).

Figure 8. The FTIR spectra of the Tstar-Ita-Gly resin and the Tstar-Ita-Gly cured sample (Jahandideh, Esmaeili and Muthukumarappan, 2018).

Polycondensation was proven by other researchers to not be the only way to synthesise itaconic-based resins. Ma et al. (2013), for instance, synthesized an itaconic acid-based epoxy resin with curable double bonds (EIA) by the esterification reaction with epichlorohydrin (ECH), see Figure 9. The curing was performed by addition of methyl hexahydrophthalic anhydride (MHHPA). The results for the cured EIA analyses returned glass transition temperature, tensile strength, flexural strength and modulus of 130.4 °C, 87.5 MPa, 152.4 MPa and 3400 MPa, respectively. Those results were considered comparable or better than those of diglycidyl ether of bisphenol A (DGEBA), also cured by MHHPA. Moreover, the bio-based epoxy resin was also copolymerized with divinyl benzene (DVB) and acrylated epoxidized soybean oil (AESO) and it was found that the properties of the cured copolymerized EIA could be regulated further. It was concluded that EIA has great potential to replace some petroleum-based thermosetting resins in determined fields.

(15)

- 9 -

Figure 9. The synthetic route to EIA (Ma et al., 2013).

Other distinct way of synthesising itaconic acid-based resins was reported by Kumar at al.

(2018). They synthesised a trifunctional epoxy resin from itaconic acid (TEIA) by first performing allylation of IA to form diallyl itaconate (DAI). After that, metachloroperoxybenzoic acid (m-CPBA) was used for epoxidation of DAI. Figure 10 presents the synthesis process.

Figure 10. Allylation route of itaconic acid and Synthetic route of trifunctional epoxy resin from itaconic acid (TEIA) (Kumar at al., 2018).

1.5.2. Plant oils-based resins

Plant oils are renewable natural resources and another replacement for fossil resources because of its abundance, low price and availability in various oil seeds. These oils are normally triglycerides of differing fatty acids and degrees of unsaturation (Islam, Beg, and Jamari, 2014). Their chemical structure contains multiple C=C bonds, which make them great building blocks for the synthesis of several polymeric materials. This synthesis can be achieved by either fatty acid C=C bond functionalization and following copolymerization or by direct copolymerization of the fatty acid C=C bonds with alkene comonomers (Andjelkovic et al., 2005).

(16)

- 10 -

Researchers investigated many ways to functionalize the plant oils. Some examples are epoxidation and acrylation (Bakare, 2015). For instance, Petrović et al. (2002) epoxidized soybean oil in a toluene solution with peroxoacetic and peroxoformic acids as oxidizing agents, to study the reactions kinetics. Lu et al. (2010) also performed epoxidation, with the aim of investigating the effects of different free fatty acids on the enzymatic epoxidation of soybean oil methyl esters.

The acrylation of epoxidized soybean oil modifies the epoxidized triglycerides in order to increase its molecular weight. This allows the triglyceride to go through free-radical polymerisation. Fu et al. (2010) reacted epoxidized soybean oil with acrylic acid as ring opener to obtain acrylated epoxidized-soybean oil (AESO). Adekunle, Åkesson and Skrifvars (2010) arranged three different modifications of epoxidized soybean oil. The idea of the research was to functionalize the triglycerides of epoxidized soybean oil with methacrylic acid, acetyl anhydride, and methacrylic anhydride. Soybean resins are based on triglycerides, which have three fatty acid chains linked to a glycerol molecule (Figure 11). Methacrylated epoxidized soybean oil (MSO) was obtained by ring-opening polymerization of epoxidized soybean oil. The MSO was mixed with 2 wt % t-butyl peroxybenzoate and cured thermally at 160 °C for 24 h. The characterization by FT-IR confirmed that there was no structural changes in the cross-linked polymer, despite of the long curing time. The TGA curve of the MSO sample showed a single degradation in the range of 320–480 °C, which corresponded to the decomposition of the cross-linked structure.

Figure 11. Chemical structure of a triglyceride molecule (Adekunle, Åkesson and Skrifvars, 2010).

1.5.3. Lactic acid-based resins

Lactic acid-based thermoset resins can be synthesised by introduction of reactive groups into lactic acid oligomers. This can be achieved by end-capping the oligomers with methacrylic or acrylic groups, for instance, in order to enable their free-radical polymerization. Many studies have been made about this processes, for example Xie et al. (2007) synthesised a three-arm oligomeric polyester from glycolic acid and L-lactic acid. The product was then end- functionalized with methacryloyl chloride. The result was a methacrylated copolymer, see Figure 12, with compressive yield strength of about 20.1 MPa and modulus of 730.2 MPa.

(17)

- 11 -

Figure 12. Structure of a synthesized glycolide/lactide based polyester resin (Xie et al., 2007).

Ho and Young (2006) also used methacryloyl chloride for the end-functionalization. A mixture of propylene glycol and lactides were end-capped into a triblock poly(lactide-co- propylene glycol) dimethacrylate adhesive, as shown in Figure 13.

Figure 13. Structure of poly(lactide-co-propylene glycol) dimethacrylate adhesives (Ho and Young, 2006).

Another example is Åkesson et al. (2010) who synthesized a cross-linkable resin based on lactic acid reacted with pentaerythritol and itaconic acid (IT) by a direct condensation reaction. The obtained oligomers were end-functionalized by methacrylic anhydride, see Figure 14. DMA analysis showed a Tg equals to 83 °C. The TGA analysis returned results showed in Figure 15. The polymer presented relatively stability up to 200 °C, losing 10 wt%

at about 279 °C and 50 wt% at 358 °C, as observed in the Figure. It is also recorded at the graph the maximum degradation at 319 °C and a second peak of degradation at 443 °C.

(18)

- 12 -

Figure 14. Idealized structure of the obtained end-functionalized oligomer (Åkesson et al., 2010).

Figure 15. Thermogravimetric analysis of the cross-linked resin (Åkesson et al., 2010).

Helmine, Korhonen and Seppala (2002) synthesized polylactide oligomers with co-initiators containing different numbers of hydroxyl groups and investigated the effect of the chemical architecture on their properties. The co-initiators molecular structures used by them are shown in Figure 16.

(19)

- 13 -

Figure 16. Molecular structures of butanediol, pentaerythritol and polyglicerine-06 and -10 (Helmine, Korhonen and Seppala, 2002).

The polylactide oligomers were functionalized with methacrylic anhydride and crosslinked with thermal initiation at 90 °C, see Figure 17. It was concluded that oligomers with more arms and shorter lactide blocks had higher mechanical strength, being the cured polymer prepared from pentaerythritol the one with the best mechanical properties (compressive yield strength = 120 MPa, modulus = 2800 MPa).

Figure 17. Functionalization of lactic acid oligomers with methacrylic anhydride and crosslinking scheme (Helmine, Korhonen and Seppala, 2002)

Bakare (2015) investigated the possibility of producing bio-based thermoset resin from lactic acid and glycerol, synthesizing resins with three different chain lengths. The synthesis was performed in two stages. In the first condensation reaction stage, it was prepared a star-shaped oligomer of glycerol and lactic acid. The flask was immersed in an oil bath and the reaction temperature was increased to 145 °C. At this moment, there was a production of water by the reaction, which was collected by azeotropic distillation. After the initial reaction, the temperature was raised to 165 °C and finally to 195 °C. The product was then end- functionalized with methacrylic anhydride and submitted to a rotavapor distillatory to remove the by-product (methacrylic acid) and the remaining toluene in the resin mixture. FT-IR was performed in order to check the crosslinking reaction after curing the resins. The test

(20)

- 14 -

registered carbon-carbon double bonds (C=C) at the second step of the reaction, indicating the occurrence of the end-functionalization by methacrylic anhydride. Those same double bonds were not present on the cured resin spectra, indicating complete reaction of all double bonds during the curing process. In the spectra it is also visible the almost disappearance of the OH- group peak in the uncured resin which confirms the reaction of almost all the hydroxyl groups. Those results can be seen in the Figure 18 bellow.

Figure 18. FT-IR spectra of GLY/LA resin with chain length n = 3. The spectra for first synthesis step, last synthesis step and for cured resin are shown (Bakare, 2015).

Bakare (2015) also performed DSC analysis, that detected a glass transition temperature of 76

°C for the glycerol/lactic acid resin (n=3), as shown in Table 1. It is also shown the results for the TGA analysis, where it was noticeable that the resin with chain length of three had better thermal stability.

Table 1. Thermal characterization results for the resins (Bakare, 2015).

Bakare (2015) concluded that it was possible to produce a resin comparable to commercial unsaturated polyester resins. Between the three resins with different chain lengths, the one with n=3 proved to have better mechanical, thermal, and rheological properties.

Esmaeili (2015) synthesized thermosetting resins prepared from lactic acid and glycerol following the same method used in Bakare (2015) research. They investigated the possibility of using those resins as a matrix for regenerated cellulose ﬁber reinforcements. They noticed that ageing deteriorated the mechanical properties of the glycerol-lactic acid-viscose

(21)

- 15 -

composites but the addition of styrene somewhat improved the composite ageing properties and the tensile max stress before ageing (Figure 19).

Figure 19. Stress at break of bidirectional ﬁber type composites with 65 % ﬁber load with and without styrene before and after ageing in climate chamber (Esmaeili, 2015).

Moreover, interests in the use of lactic acid thermoset resin using different types of natural fiber have been growing. Chen et al. (2014) produced a ramie fabric-reinforced thermoset composite from lactic acid, pentaerythritol and methacrylic anhydride. The fiber content varied from 37 wt-% to 59 wt-%. In the first step of the synthesis, L-lactic acid and pentaerythritol were reacted, producing a four-armed star-shaped poly lactic acid. In the second step, the resin was end capped with methacrylic anhydride (MAAH), as shown in Figure 20. The tensile strength, flexural strength and impact strength results showed an increase for the resins up to 48 wt-% of raime addition. However, further increase in fiber content resulted in a decrease of those properties due to possible poor dispersion and interfacial bonding between the fiber and matrix.

(22)

- 16 -

Figure 20. Synthesis scheme (Chen et al., 2014).

Åkesson et al. (2011) also acted on the fiber reinforcement field, with the purpose of investigate if a LA-based thermoset resin can be used to prepare natural fiber-reinforced composites. For that, a bio-based thermoset resin from methacrylated star-shaped oligomers of lactic acid was produced with addition of flax fibers. The tests returned an increase in the tensile strength and tensile modulus by addition of fiber up to 70wt-%. The storage modulus of this flax/lactic acid composite was also reported, being equal to 9.32 GPa.

1.5.4. Ageing review

Masek and Latos-Brozio (2018) investigated the thermo-oxidative ageing of aliphatic polyesters – Poly(hydroxyalkanoates) (PHAs) and poly(lactic acid) (PLA) – aiming to improve their stability by extending and controlling their lifetime. For this purpose, different flavonoids were used as stabilizer additives and their stabilization efficiency was studied.

Figure 21 shows the molecular structures for the PHA and PLA and Figure 22 presents the natural flavonoids used in their work (rutin and hesperidin). Flavonoids can interact directly with the free radicals, acting as an antioxidant. When free radicals are mixed with natural antioxidants, the material becomes protected from free radical damage and interrupting free radical reactions. Besides that, the fact that the antioxidants are natural results in a more environmental friendly material.

Figure 21. Chemical composition of polylactide (PLA) and polyhydroxyalkanoate (PHA) (Masek and Latos-Brozio, 2018).

(23)

- 17 -

Figure 22. Chemical composition of rutin and hesperidin(Masek and Latos-Brozio, 2018).

The researchers performed UV ageing during 288 h and consisted of two alternately repeating segments: day (radiation intensity = 0.7 W/m², temperature = 60 °C, and period = 8 h) and night (no UV radiation, temperature = 50 °C, and period = 4 h). Weathering test was also carried out simulating a day cycle (radiation intensity = 0.4 Wm⁻², temperature = 60 °C, period = 240 min, humidity = 80%, and rain water on) and a night cycle (no radiation, temperature = 50 °C, humidity = 60%, and period = 120 min). Moreover, it was performed a thermo-oxidative ageing test for 10 days with the help of a dryer, with thermo-circulation of air and elevated temperature of 383 K. The samples were characterised by DSC analysis before and after the thermo-oxidative ageing process. The glass transition temperature registered for PLA and PHA before the ageing were, respectively, 59.66 °C and 36.79 °C.

After ageing the Tg had a slight decrease being 58.37 °C for PLA and 35.14 °C for PHA. This study also revealed that the addition of rutin and hesperidin enhanced the resistance to oxidation of both PLA and PHA. They stated that the hydroxyl groups in the flavonoids’

structure provided the polymers with great antioxidant properties. It was concluded that the improvement of polyester resistances to the oxidation could indicate a longer life time for the material.

Pérez Amaro et al. (2016) analysed the thermo-oxidative stabilization of poly(lactic acid) with antioxidant intercalated layered double hydroxides (LDHs). LDHs are usually resistant to degradation by light, temperature, oxygen and other medias, which makes it an ideal additive for storage materials of bioactive molecules and drugs. The antioxidant-modified LDHs (AO- LDHs) were obtained by intercalation between the inorganic layers via anion exchange of IrganoxCOOH and Trolox. Figure 23 shows their chemical structures. Poly(lactic acid) (PLA) Ingeo™ produced by NatureWorks® was used as a polymer matrix to prepare the PLA/IrganoxCOOH-LDH and PLA/Trolox-LDH composites. In order to determine the effect of the AO-LDHs against the thermo-oxidative degradation of PLA, it was performed accelerated thermo-oxidative experiments of all the PLA/AO-LDH composites. The experiment took place in an air oven at 200 °C for different intervals. It was indicated by size exclusion chromatography (SEC) analysis that both the AO-LDHs protected the PLA from thermo-oxidative ageing.

(24)

- 18 -

Figure 23 . Chemical structure of 3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionic acid (IrganoxCOOH) and 6- hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) (Pérez Amaro et al., 2016) .

Badía et al. (2011) performed an accelerated thermal ageing of PLA over its glass transition temperature, in order to simulate its service life. They also obtained the PLA pellets from NatureWorks®. Thermo-oxidative ageing was performed on forced-ventilation oven under air atmosphere at 60 °C. According to the authors, ageing at temperatures above the glass transition temperature of PLA (ca. 55 °C) can increase the molecular mobility of the chains.

Two stages were observed on the thermo-oxidative ageing for their work. The first stage presented the occurrence of intramolecular transesterifications and the second showed chain- scission reactions. At this last stage there was observed a strong reduction of cyclic species as well as noticeable increment of H-[LAL]n-OH species, which was explained by the authors as product of hydrolytic reactions induced by present water.

Lai and Liau (2003) investigated the thermo-oxidative degradation of poly(ethylene glycol)/poly(L-lactic acid) (PEG/PLLA) blends by characterizing it with FT-IR, DSC, Gel permeation chromatography (GPC) and TGA. The thermo-oxidative ageing test was carried out for 24 hours in air at 80 °C. The results showed that the addition of PLLA promoted the thermo-oxidative degradation of PEG until a certain amount (PEG/PLLA 50/50). After this amount, the addition of more PLLA resulted on the decrease of the thermo-oxidative degradation degree of PEG. The authors suggested that this could be explained by the dilution effect, meaning that the PLLA can dilute the concentration of free radicals.

Åkesson et al. (2011), whose work was mentioned before, evaluated the durability of the composites produced by an ageing test. Test samples with 70 wt% flax fiber were placed in a climate chamber at a temperature of 38 °C and relative humidity of 95% for 1000 h. Tensile test was carried out before the ageing test, directly after, and after 24 h of reconditioning at room temperature. The results showed that by the end of the test the composites had lost around 70% of their tensile strength, see Figure 24. The authors stated that the reconditioning of the samples did not have any significant effect, when analysed the standard deviation. The tensile modulus was also affected, dropping from 9 GPa to 2.5 GPa, see Figure 25. It was also pointed out the change on the elongation, which increased from 1.4% to 3.1%.

(25)

- 19 -

Figure 24. Tensile strength of dry samples after 1000 h of ageing, and after 24 h of reconditioning (Åkesson et al., 2011).

Figure 25. Tensile modulus of dry samples after 1000 h of ageing, and after 24 h of reconditioning (Åkesson et al., 2011).

1.5.5. Biodegradation review

Sahoo et al. (2018) reported the degradation behaviour of bio-based polyester polyurethane nanocomposites (PEPUNC) under soil burial, UV exposure and hydrolytic-salt water medium. The samples were exposed to degradation for a period of 20 and 40 days. The FTIR spectra revealed that the degradation behaviour might be due to the breakage of ester linkage, C–N bond and C–O bond respectively. It also showed that among those three types of biodegradation, the PEPUNC samples went maximum degradation under the soil burial medium.

Lv et al. (2017) investigated the biodegradation behavior of Poly(lactic acid) (PLA) blended with starch and wood flour in the campus of the Northeast Forestry University, Harbin, China.

Three different specimens were obtained (P-S3W7, P-S5W5 and P-S7W3) by varying the wood and starch composition from 9 to 21 wt% while the PLA amount remained 70 wt%. The researchers mentioned the complexity of the soil degradation in outdoor soil process, since it involves hydrolysis and bacterial degradation. According to their study, PLA is first hydrolysed, with the presence of water; this process causes the cleavage of ester bonds, resulting in low molecular weight PLA. Then the microorganism metabolize the low molecular weight material to form water, carbon dioxide and other metabolic biomass. The researchers performed an element analysis during the soil burial degradation and noticed that the carbon atom content decreased after 60 days of degradation but the oxygen atom content increased. They stated that the increase of oxygen atom content could have been a consequence of the increase of carboxylic acid end group numbers in the molecular structure, due to the hydrolysis of ester bond, see Figure 26.

(26)

- 20 -

Figure 26. Hydrolysis mechanism of PLA in soil (Lv et al., 2017).

Their soil biodegradation test lasted for a total of 105 days (from April 2015 to August 2015) under outdoor conditions with average precipitation of 95 mm. The degraded samples were retrieved, after different exposure time, washed and dried at room temperature. They were then characterised by tensile test and scanning electron microscopy (SEM). Figure 27 shows the surface morphologies of the samples analysed by SEM. As observed, the surfaces of the blends were smooth before degradation, but seriously eroded after 105 days of process, presenting cavities. According to the authors, this surface damage resulted in the decrease of the mechanical strength. Figure 28 shows that both the tensile and flexural strengths of the blends had a steep decline during the first days of degradation.

Figure 27. Surface morphologies of starch/wood flour/PLA blends before and after the degradation in outdoor soil (Lv et al., 2017).

(27)

- 21 -

Figure 28. The evolution of mechanical strength of the starch/wood flour/PLA blends during the soil burial degradation process (Lv et al., 2017).

Rudnik and Briassoulis (2011) studied the degradation behaviour of poly(lactic acid) films in soil under Mediterranean field conditions and soil burial laboratory simulations testing. FT-IR analysis of PLA before and after 11 months burial in the soil under real field conditions were carried out to evaluate chemical modifications occurring on film surface (see Figure 29). The researchers reported that there was a decrease in the intensity of absorbance at 1748 cm−1, corresponding to carbonyl C=O stretching band. This can be observed at Figure 8. They concluded that for both laboratory simulated soil burial and real field conditions the degradation of the mechanical properties started early, however the degree of disintegration of the PLA was still rather low. Besides that, variation of thickness and also form of material presented a crucial influence in the biodegradation process under soil burial. They stated that the thinner films evidenced more general changes.

Figure 29. FTIR-ATR spectra of PLA 75 film: initial (top) and after 11 months of soil burial (bottom) in Spata (Rudnik and Briassoulis, 2011).

Gallet, Lempiäinen and Karlsson (2000) buried films of poly(l-lactide), PLLA, in outdoor environment in south Finland during two years. Their aim was to identify the small oligomers

(28)

- 22 -

present in PLA films after degradation in soil and to monitor the changes in PLLA matrix.

The paper concluded that the degradation of the material involved an assimilation of small products by microorganisms while hydrolysis was taking place. It was also confirmed by DSC analysis that the thermal properties of the films (Tg and Tm) were affected during the second year in soil.

Oprea (2014) investigated the effect of introducing modified and crude soybean oil into polyurethane matrix composition on the biodegradation process. For that, a series of polyurethane acrylate/AESO pre-polymer mixtures were prepared by varying the weight ratio of the components. The samples were subjected to biodegradation by soil burial under natural environment conditions during one year from September 2012 to September 2013. After a period of 6 and 12 months, the degraded samples were cleaned with distilled water, dried at 24°C for 48 hours and characterised by FT-IR, tensile test and observation of surface chemical and morphological changes. FT-IR spectra showed decreases of absorption peaks, which were due to the action of microorganisms, bacteria and fungi. Regarding the tensile properties there was an increase in the young’s modulus and a significantly decrease on the tensile stress and strain of the samples. After 12 months, the samples were completely week.

It was concluded that these reductions of the physical properties and those changes on the chemical structure indicate a complete polymer biodegradation, which was accelerated by the presence of relatively high amounts of soybean oil in their structure.

Guleria, Singha and Rana (2017) have studied the biodegradation of starch-based biocomposites reinforced with mercerized lignocellulosic fibers. Starch is a polymeric carbohydrate produced by plants and corn is a major commercial source of starch. The samples of dimensions 30 mm x 8 mm x 2 mm were buried under natural soil for 120 days.

After pre-determinate periods, samples were cleaned and dried at 70°C for 24 hours. From this process, it was concluded that by increasing the percentage of the fiber in the composites, a decrease in the biodegradation rate occurs. That could be observed by the weight analysis of the samples. The fiber loaded composite lost around 40% of the weight after seven weeks of test, while the cross-linked corn starch matrix lost 60% after six weeks.

Tajau et al. (2016) studied the soil burial biodegradation of palm oil-based UV-curable films.

The samples were synthesised from epoxidized palm oil acrylated (EPOLA) resin and polyurethane palm oil (POBUA) resin. Square films were weighted and buried in the experimental plot ground for 12 months, see Figure 30. The samples were washed with distilled water and acetone, and left to dry under ambient temperature before weighing and characterization. The biodegradability of those samples were compared with petrochemical resin based films from the commercial grade oligomer type such as Ebecryl-264 (EB264) and Ebecryl830 (EB830), see Figure 31. As observed in the Figure, the EPOLA type uv-curable film presented an approximated 90% of weight loss for the films synthesised with the photoinitiator benzophenone and 93% for the ones with IRR819. The petrochemical-based films results, on the other hand, showed weight loss between 59% - 76% after 12 months in soil. Meanwhile, the POBUA-IPDI showed slow degradation, indicating that the aromatic type structure of urethane oligomer influence the process. The researchers concluded that the palm oil-based UV-curable films could be considered sustainable and environmental friendly, when compared to petrochemical-based products.

(29)

- 23 -

Figure 30. Soil burial experiment plot ground – Buried samples in the soil (Tajau et al., 2016).

Figure 31. Percentage of weight loss in palm oil-based films and petrochemical-based films after 12 months of biodegradation – Effect of oligomer and photoinitiator (Tajau et al., 2016).

Finally, it is noticeable that several raw materials and new processes in the bio-based polymer area have been studied. However, renewable resources still play a very small role and, despite of the high research activity, there are still unknown areas in this field (Garrison, 2016).

1.6. Purpose and limitations

This thesis is a continuation study from Bakare (2015) research with the aim of synthesising bio-based polyester resins from lactic acid with not only glycerol but also with other two different core-molecules – ethylene glycol and pentaerythritol. The difference between these three molecules is the number of hydroxyl (O-H) endings, see Figure 32. This will provide three different chemical architectures resins with 2, 3 and 4 arms.

Figure 32. Structural formulas of ethylene glycol, glycerol and pentaerythritol (Sigma Aldrich, 2018).

Ethylene glycol Glycerol Pentaerythritol

(30)

- 24 -

Figures 33, 34 and 35 shows the chemical reactions for the first and second stages of the synthesis. The amount of lactic acid varies according to the number of hydroxyl (O-H) endings in the core-molecule. Ethylene glycol has two arms and all two arms should be reacted with lactic acid, as observed on Figure 33, so one mole of ethylene glycol should react with 2x3 moles of lactic acid. The same goes for glycerol (3x3 arms; Figure 34) and pentaerythritol (4x3 arms; Figure 35). For the second stage reactions, taking the glycerol/lactic acid resin synthesis as an example, to end-functionalize three arms of glycerol, three moles of methacrylic anhydride should be added. The same principle goes for ethylene glycol/lactic acid (2 moles; Figure 33) and pentaerythritol/lactic acid (4 moles; Figure 35).

Figure 33. Reaction scheme of first and second stages for the synthesis of methacrylate functionalized ethylene glycol/lactic acid resins.

Figure 34. Reaction scheme of first and second stages for the synthesis of glycerol/lactic acid resins with lactic acid chain lengths of 3.

(31)

- 25 -

Figure 35. Reaction scheme of first and second stages for the synthesis of methacrylate functionalized pentaerythritol/lactic acid resins.

The molecular architectures and chemical composition of the bio-based polymers influence their properties, as well as their response through biodegradation and ageing. Therefore, the overall idea with this study is to determine how the chemical architecture will affect thermal and mechanical properties by analysing those three different resins before and after the processes of ageing and biodegradation, separately. To accomplish that, the cured resins were tested according to their response to heat, mechanical tension, pressure, stress and IR radiation.

2. EXPERIMENTAL WORK

Three structurally different resins were synthesised from lactic acid: ethylene glycol/lactic acid (EG/LA) resin, glycerol/lactic acid (GLY/LA) resin and pentaerythritol-lactic acid (PENTA/LA) resin. This section describes the synthesis and the curing processes for each resin, and presents the characterization methods applied to them.

2.1 Resin synthesis

2.1.1 Materials

L-Lactic acid (88-92%, Sigma-Aldrich) was purified by using a rotary evaporator. Ethylene glycol (99.5%; Sigma Aldrich), glycerol (99%; Acros Organics) and pentaerythritol (98%;

Sigma Aldrich) were used as received. Toluene was used as solvent (99.99%; Fisher Scientific) and methanesulfonic acid (98%; Alfa Aesar) was used as the catalyst in the condensation reaction. For the end-functionalization, was used hydroquinone (99%; Acros Organics, supplied by Fisher Scientific) as inhibitor and methacrylic anhydride (94%; Alfa Aesar) as reagent.

(32)

- 26 - 2.1.2 Methods

All three resins were synthesised by the same methods, changing only the amount of chemicals used in each. This section, therefore, provides a general description of methods followed by specifications of the substances’ amounts.

Lactic acid was first poured in a 500 mL one-neck round bottom flask connected with a rotary evaporator to remove the water content. The evaporator apparatus for this thesis contained a heater, a vacuum pump and a condenser connected to it. To evaporate the water, the substance was heated to 90 °C, which is well below the boiling point of lactic acid. As soon as the temperature stabilised, it was necessary to turn on the pump so the water boiling point could be reached, but with caution to avoid lactic acid evaporation. The condenser had continuous water supply with the purpose of maintaining a low temperature in it so the water vapour could condensate and be collected. The evaporation process was held within 2 hours and the pure lactic acid was stored dry at room temperature.

The purified lactic acid was then ready for the synthesis, which was performed in two stages.

In the first stage (the condensation reaction stage), linear oligomers were prepared by reacting lactic acid with ethylene glycol, and star-shaped oligomers were obtained from reaction with glycerol and pentaerythritol. In the second step (the functionalization step), the resin was reacted further with methacrylic anhydride to add end-groups. The result was three distinct resins with different chemical architectures.

In the first stage, all components were placed in a 250 mL three-neck round bottom flask. An amount of 130.00 g of lactic acid was added for each resin reaction. For the EG/LA resin it was weighted 14.93 g of ethylene glycol, 49.27 g of toluene and 131.12 µL of methanesulfonic acid. The GLY/LA had similar amounts: 14.76 g of glycerol, 49.22 g of toluene and 130.99 µL of methanesulfonic acid. For the PENTA/LA it was taken 16.37 g of pentaerythritol, 49.76 g of toluene and 132.41 µL of methanesulfonic acid. The flask necks were connected to an azeotropic distillation apparatus, a nitrogen gas inlet – to create an inert atmosphere – and a thermometer. The solution was heated for 2 hours in an oil bath with a set temperature of 145 °C. The water produced in the reaction was collected by azeotropic distillation. After the initial reaction, the temperature was raised to 165 °C for further 2 h, and finally increased to 195 °C for 1 h under constant stirring.

In the second stage, the resins were end-functionalized with methacrylic anhydride (EG/LA - 78.35 g, GLY/LA - 78.40 g and PENTA/LA - 77.08 g). This anhydride introduces reactive double bonds in the structure, which enables the resin to be cross-linked when it is later cured.

Hydroquinone (0.1 wt%) was also added, in order to stabilize the reaction and avoid premature curing. The apparatus used in this second step is similar to the one used for the first step but with a condenser connected directly to the main neck of three-neck round bottom flask. Methacrylic anhydride was added slowly in drops using a dropping funnel. The mixture was stirred continuously for 4 hours with constant temperature of 90°C.

The by-product methacrylic acid that had formed and the residual toluene were removed by rotary evaporator distillation at a temperature of 90°C and pressure of 30 mbar to purify the resin before polymerization.

The amount of main reactants used for the first step were calculated from the mass ratios:

ethylene glycol: lactic acid = 62.07: 540.48; Glycerol: lactic acid = 92.094: 810.72 and

(33)

- 27 -

pentaerythritol: lactic acid = 136.15: 1080.96. To avoid starvation of methacrylic anhydride, 1.1 mole of methacrylic anhydride for each arm was used, so 2.2 moles for EG/LA, 3.3 for GLY/LA and 4.4 for PENTA/LA.

The percentage amount of toluene, methanesulfonic acid and hydroquinone used were based on Bakare (2015) research.

2.2 Curing procedure

The sample was obtained by mixing the resin with dibenzoyl peroxide as the free radical initiator and N,N-Dimethylaniline as the accelerator. The sample mixture was left to cure at room temperature for 1 h. After this, it was placed in a drying oven (Termarks, series TS900) to postcure. Several tests were performed varying the quantity of peroxide and accelerator as well as the post curing time and temperature. Finally, the best parameters found were of 2 wt

% of dibenzoyl peroxide and 0.5 wt % of N, N-dimethylaniline with 8 hours in the oven at 130°C. The specimens were then grinded in a MetaServ 2000 grinder, with the purpose of polishing them to achieve smother surfaces and similar dimensions.

2.3 Characterisation

The different analysis methods that were used to characterise the specimens produced in this project will be described below.

2.3.1 FT-IR analysis

The Fourier transform infrared (FT-IR) spectroscopy was used to verify the resin functionalization with methacrylic anhydride in the second stage of the reaction by comparing to the resin spectra from the first stage. The cured resin were also analysed, so the crosslinking reaction could be verified from the spectral bands corresponding to the carbon- carbon double bonds. The neat and cured resins were analysed using on a Nicolet IS10 spectrometer (Thermo Fisher Scientific). The procedure was repeated three times for each resin type, resulting in the analysis of nine samples from each stage. The FTIR were first cleared of background, to avoid interference from the surrounding environment. Spectra were collected for all the samples at a resolution of 4 in the range 4000–500 cm^-1 for a total of 64 scans. The machine was cleaned with ethanol after each set of samples.

2.3.2 DSC analysis

The cured resins were also analysed by differential scanning calorimetry (DSC) on a TA Instrument Q 2000, from TA Instruments. Samples of cured resin were placed in sealed aluminium pans and heated from 0 °C to 200 °C at 10 °C /min in a nitrogen atmosphere. This was done with the purpose of determining the glass transition temperature (Tg) of the cured samples. The procedures were repeated three times for each resin type, meaning nine samples in total.

(34)

- 28 - 2.3.3 TGA analysis

Thermogravimetric analysis (TGA) on the cured resin was done on a TA Instrument Q500 from TA Instruments to investigate the thermal stability of the resins by recording the percentage weight loss of the cured samples. Samples of about 20 mg were heated from 0 °C to 600 °C at a heating rate of 10 °C/min in a nitrogen purge stream. The procedure was repeated three times for each composite type, totalizing in nine samples.

2.3.4 Tensile test

The tensile test was conducted for characterisation of the mechanical properties of the cured sample. For tensile testing dog bone shaped specimens were tested on a H10KT (Elastocom) machine equipped with a 2.5 kN load cell and a mechanical extensometer, at a crosshead speed of 1 mm/min. Nine specimens of each resin were tested in the tensile test.

2.3.5 Ageing

This thesis will focus on the thermo-oxidative ageing by submitting the samples to specific temperature and humidity with the help of a climate chamber. For the ageing under humid conditions, it was separated nine samples of each resin. The cured resins were placed in a climate chamber (NUVE, TK120) with temperature of 40°C and 75% relative humidity for 695 h. Those samples were then dried at 70 °C for 24 hours and characterised by FT-IR, DSC and tensile test.

2.3.6 Biodegradation

The biodegradation test was carried out by soil burial, using similar parameters of Guleria, Singha and Rana (2017) research. For this process, it was separated three samples of each resin. The specimens were buried in pots between two layers of soil, after being wrapped by plastic nets. The pots were kept at outdoor ambient conditions from March 2019 to May 2019.

The specimens were taken to weight after every 15 days to visual and microscopic analysis.

The humidity of every checking day was taken on account for. After 53 days of procedure, the samples were carefully cleaned with distilled water before being dried at 70°C for 24 hours.

The composted samples were then characterised by FT-IR and DSC.

3. RESULTS AND DISCUSSION

In this section, results from synthesis of the resins and characterization analyses are presented.

3.1 Resin synthesis

The resins obtained weighted around 200 g each. With that amount it was possible to cure and grind 25 samples for each resin type with some uncured resin to spare. Figure 36 shows the dog bones shaped specimens obtained still in the mold used for the curing process.

(35)

- 29 -

Figure 36. Cured resin in dog bone mold (Photo Lara Hastenreiter).

Part of the samples was separated for initial FT-IR, DSC, TGA and tensile analysis. The other part was submitted to biodegradation and ageing processes and then characterised by DSC and FT-IR. Moreover, tensile characterization was performed for the aged samples.

3.2 FT-IR analysis

Figure 37 shows the FT-IR spectra of the first stage, second stage and cured resins for each resin. In the second stage spectra it can be seen appearance of unreacted carbon-carbon double bonds at about 1640 cm^-1 (stretching, C=C) and at 816 cm^-1 (bending, =CH2), these bands were not present in the spectra for the first stage. Besides that, it is also evident a near disappearance of the OH-group absorbance at 3500cm^-1 which indicates that almost all the hydroxyl groups were reacted. This confirms that the end-functionalization by methacrylic anhydride had proceeded as intended. Those same carbon-carbon double bonds disappear again on the spectra for the cured resins. This clearly indicates that all double bonds had reacted when cross-linked, and therefore the resins could be considered as completely cured.

Figure 37. FT-IR spectra of EG/LA, GLY/LA and PE/LA resins with chain length n = 3. The spectra for first stage, second stage and for cured resin are shown.

OH-group

C=C

=CH2

Biodegradation and ageing of bio-based thermosetting resins from lactic acid