In this chapter we will discuss the various thermoset chemistries, ranging from petro‐ to bio‐based thermosets and new developments with renewable raw materials for thermosets. We will see that, although this type of chemistry has a long history, still significant developments are made relatively recently, which renders either processes more flexible or more environmentally friendly.¹

3.1 Petro‐based Thermoset Resins

Until relatively recently, thermoset polymers as used for coatings mostly originated from petro‐based chemistries. For these petro‐based thermoset chemistries, the main types used are:

We will discuss only the last five types here (for phenolics, see [2] and for aminos, see [3]). From the dates shown, it is clear that this type of chemistries has already a century‐long history. In particular, the 1930s is regarded as the rich period of the plastics industry with many new chemistries invented [4].

3.2 Epoxy Systems

Each type of chemistry has its own nomenclature. For epoxies we distinguish between an epoxy group (formally an oxirane ring, 1,2‐ethyleneoxide) and an epoxy resin (i.e. a resin‐containing epoxy groups), mostly in the form of glycidyl groups formed by reaction with (highly reactive and highly toxic) epichlorohydrin (ECH) (Figure 3.1). The reaction of epoxy compounds occurs via ring opening of the oxirane ring by nucleophiles, optionally catalyzed by Lewis acids or special amines, such as tris (dimethylaminomethyl) phenol (DMP‐30) or dimethylbenzylamine (BDMA). Herewith OH‐groups are generated. Examples of nucleophiles are NH‐, COOH‐, and OH‐groups, so in this case a self‐reaction occurs. A schematic and an example of such a reaction are given in Figure 3.2.

For epoxy systems we distinguish between 100% epoxies and coepoxies. In the former case the network is exclusively formed by the epoxy compounds (Figure 3.3). For the latter we distinguish between OH‐epoxies with OH‐functional resins, COOH‐epoxies with acid‐functional resins, and anhydride epoxies, which react with the OH‐group first and generally provide a more reactive system than acids. A somewhat special case is the NH‐epoxies. A primary amine can react first with an epoxy to a secondary amine, which subsequently reacts again with an epoxy to yield a tertiary amine (Figure 3.4). In fact, diamines are ideal 4‐functional crosslinkers for epoxies (Figure 3.4).

By far the most important epoxy system is based on the reaction between bisphenol‐A (BPA) and ECH, as shown in Figure 3.5. Higher bisphenol‐A epoxy resins use the so‐called advancement process, which typically involves high temperatures, low viscosity, long reaction times, and good agitation, resulting in less branching. One example is the chain extension of bisphenol‐A diglycidyl ether (DGEBPA) with BPA (Figure 3.6). Here there are two epoxy groups per polymer chain. The final properties, like the glass transition temperature T_g, can be tuned by selecting the proper ratio of components, as illustrated in Table 3.1.

When uncured, epoxy resins can be found in a large variety of forms, ranging from low viscosity liquids to tack‐free solids, which offer a wide range of processing conditions. The crosslinking can be done without the formation of low molecular weight products, no (or very low) emission of volatiles and with relatively low shrinkage during curing. Furthermore, there is also a wide choice of curing agents, and it is possible to cure at ambient as well as elevated temperatures. All these aspects provide an enormous versatility in terms of processing (e.g. wide choice of curing parameters or easy tuning of viscosity and rheological behavior), which has made epoxy resins very popular among coatings producers and users. It is not strange that epoxy resins are slightly more expensive than some of their counterparts (e.g. phenolic resins used in coatings or laminates), which is anyway compensated by their superior properties [5].

In view of the variety of binders, hardeners, and processing conditions available for epoxy resins, a wide range of properties is also attainable. It is possible to obtain high toughness, good chemical resistance, mechanical properties ranging from extreme flexibility to high strength and hardness, high adhesive strength, heat and moisture resistance, and high electrical resistance [6]. Furthermore, epoxy systems are also known for their good adhesion to most of the substrates (including metals, concrete, glass, ceramics, stone, wood, leather, etc.) and low water sensitivity. The resulting coatings typically show some flexibility, are highly resistant to chemicals, and are rather insensitive to ambient moisture.

As a result of this versatility, epoxy resins have found use in nearly all fields of application, although mainly indoor because of their aromatic content and therefore limited outdoor durability, and, specifically, as protective primer coatings in the automotive and marine industries, on the overall use as adhesives, in solders and caulking products, in flooring protection, compounding for molds, as insulating layers and components for electronics, in the composition of synthetic textiles, and in fiber‐reinforced plastics and composites for the aerospace industry [7].

3.3 Acrylates and Acrylics

An acrylate is an ester of (meth)acrylic acid in which the vinyl group is still present, while an acrylic is such an ester for which the vinyl group has reacted (Figure 3.7). These are chain‐growth polymers, where polymerization proceeds via the acrylate group. For acrylates we have polyacrylate polymers, known as acrylic resins, and polyvinyl polymers, known as vinyl resins. Acrylate‐functional resins, oligomers, and dimers are generally designated as acrylates, as they still carry an unreacted acrylate group that can be used for following reactions, such as crosslinking.

The formation of polyacrylates (acrylic resins) proceeds via free radical polymerization of (meth)acrylic monomers in solution or aqueous dispersion (emulsion polymerization). A general scheme is:

where R^• denotes a radical and R^* a dormant species. Apart from acrylic and methacrylic acid, other monomers are also used. For polyacrylate resins we distinguish between nonfunctional monomers and functional monomers. Functional acrylate monomers may still have another reactive group that does not interfere with the free‐radical polymerization. A few examples of nonfunctional monomers are shown in Figure 3.8, while Figure 3.9 shows some functional monomers.

Nonfunctional monomers lead to thermoplastic acrylic resins and also form the backbone engineering of thermoset acrylic resins. The of the copolymer can be estimated by the Fox equation:

3.1

using the individual T_gs of the various homopolymers j and W_j denotes the weight fraction of the monomer of type j (see Chapter 2).

The topology of acrylic resins is the result of a random monomer distribution, which leads to a molecular mass distribution, typically with an M_w/M_n ratio of 2 to 5, as well as a compositional distribution. The probability of incorporation of a certain kind of monomer is determined by the copolymerization constants and the relative concentration of unreacted monomers in the reaction medium. For a binary system we have P_1,2 (and P_1,1) labeling the probability of incorporating a monomer of type 2 (or type 1) on a chain of type 1. Since these probabilities are not equal, during polymerization a composition drift is possible. Nevertheless, this (almost) random incorporation of functional monomers leads to a functionality distribution: large chains have a high probability of having (multiple) functional groups, while small chains may be nonfunctional.

Of the polyvinyl resins polyvinylacetate is the cheapest and more abundant polymer. It is often copolymerized with other monomers, for example, with acrylic monomers to adjust the T_g, with vinyl chloride to enhance barrier properties, or with maleic anhydride to enhance adhesion. Another example is polyvinylidenedifluoride (PVDF). While the PVDF can be made in emulsion form, polytetrafluoroethylene (PTFE) (Teflon^TM) is hardly processable. These fluorinated (co)polymers appear to be extremely durable (against UV, heat, and moisture) and are water and oil repellent (as they have a low surface energy). The various components involved in both processes are shown in Figure 3.10.

Acrylate groups are used for crosslinking reactions. Although an acrylate monomer is 2‐functional, so that it results in no extra crosslinks, they increase the M_c of the elastically active network (EAN) chains. Usually two or more acrylate groups are incorporated per resin/molecule. Diacrylates or triacrylates are used to that purpose (Figure 3.11). They are called reactive diluents since they act as solvents before curing. Acrylate‐functional resins will be discussed further in Chapter 4 dealing with radiation curing.

3.4 Isocyanates

Another important class of reactive molecules for binders is the isocyanates (R-N=C=O), which are formed as a result of the reaction of amines with phosgene. They are very reactive with nucleophiles, leading to a clean addition reaction (catalyzed by Lewis acids). A thermally reversible reaction with a sacrificial nucleophilic group or blocking is also possible and can be used for the masking of the high reactivity of the isocyanates for premature reactions during storage with desired or undesired reaction partners (Figure 3.12).

Isocyanates react with alcohols to urethanes (carbamates) and at elevated temperature further to allophanates. They also react with amines to urea and at elevated temperature further to biurets. Finally, with water they react to amines and further to urea (Figure 3.13), also referred to as moisture cure.

Isocyanates are used for various purposes, such as monomers for polyurethanes (next section) or as crosslinkers. For example, diisocyanates (from diamines) or triisocyanates (isocyanurate trimers of diisocyanates, Figure 3.14) have the advantage of having a reduced volatility in comparison with the former. Blocked di‐ and triisocyanates are also used, for example, in powder/coil coatings and for resins reacting with polyols to polyurethanes. For an extensive review on the mechanisms, chemistry, and applications of blocked isocyanates, see [8, 9]. Note that in view of the toxicity and health risks of isocyanates, mainly linked to respiratory problems caused by direct exposure [10](ISO‐3), restrictions are applied to their use in certain coating application areas, and there are very strict regulations and requirements for specific safety and health procedures. Although lately a great deal of work has been carried out to provide isocyanate‐free crosslinking alternatives [11], isocyanates remain largely involved in the chemistry of coatings.

3.5 Polyurethanes

Polyurethanes are normally made by polyaddition of polyols (mostly diols) and diisocyanates. Normally there is an excess of either polyols, that is, the chains will end up carrying only OH‐groups at their ends, or isocyanates, that is, resulting in NCO end groups. Their diol/diisocyanate monomer ratio n/n + 1 ratio determines the eventual average chain length (M_n). Branching is possible using triols or triisocyanates (see Chapter 2). The mechanism is illustrated in Figure 3.15. The NCO end groups are used for crosslinking (e.g. with OH‐functional polyesters or polyacrylates). As monomers a wide variety of polyols and isocyanates is available, a few of which are shown in Figure 3.16. When a polyurethane is built up mainly from aromatic or aliphatic diisocyanates, it is designated as an aromatic or an aliphatic polyurethane, respectively.

The wide applicability of polyurethanes on coatings is due to the versatility in selection of monomeric materials from a very extensive list of macrodiols, diisocyanates, and chain extenders available, and, consequently, the wide range of properties offered. Furthermore, a good agreement can be reached on the high properties–low cost compromise, owing to the good mechanical properties, and generally reasonable UV, heat, and water resistance capacity of polyurethane coatings [12, 13].

Polyurethanes can be found as coatings on many different materials and devices, mainly to improve their appearance, performance, and life‐span. For example, on automotive and aerospace applications, aliphatic polyurethanes can provide the desired exterior high gloss, improved color retention, and scratch and corrosion resistance. In the construction field, building floors, steel tethers, and concrete supports are typically spray‐coated with aromatic polyurethanes to make them more durable against environmental deterioration and less costly to maintain. Yet other examples come from the furniture and flooring industry and from the medical, food, cosmetics, and pharmaceutical packaging areas, which have been gaining quite some attention lately and where waterborne (WB) polyurethane dispersions are used for coating and ink formulations. Ultimately, one would say that polyurethane coatings are ubiquitously present in all materials areas, and the field keeps expanding and upgrading. For an extensive review on the chemistry and engineering applications of polyurethanes in coatings, see [12, 14].

3.6 Polyesters

Polyesters are made by polycondensation of polyols (mostly diols) and polyacids under the removal of water (Figure 3.17). Again, normally an excess of either polyols, that is, resulting in chains with only OH end groups or polyacids, that is, resulting in COOH end groups are employed. Also here, the diol/diacid monomer ratio n/n + 1 ratio determines the average chain length (M_n), and branching is possible using triols or triacids (see Chapter 2).

Polyesters are often incorporated as macrodiols in other types of resins, and their end groups can be used for crosslinking with, for example, phenolics, aminos, epoxies, or isocyanates, some of which are discussed in Section 4.2 for coil/powder coatings. Examples of polyol and polyacid monomers often used are shown in Figure 3.18, while Figure 3.19 shows a schematic view of a polycondensation reactor and the effect of reaction time on some properties. When a polyester is built up mainly from aromatic or aliphatic diacids, it is designated as an aromatic or aliphatic polyester, respectively.

Since polyesters have a very rich chemistry [15] and can be made out of various building blocks, they allow a broad range of molecular weights and compositions, which can be played to deliver enhanced properties in what concerns adhesion, abrasion resistance, overall hardness, and a particularly enhanced hydrolytic and outdoor stability. They find application in nearly all different coatings areas and particularly on metals for construction and transportation because of their excellent adhesive and mechanical properties. Examples are in metal building facades and roofing plates, automotive and public transport vehicles, household appliances, and lighting.

3.7 Renewable Raw Materials

In the framework of sustainability, renewable raw materials for binders receive increasing attention, associated with a trend from petro‐based toward agro‐ or bio‐based raw materials. Natural resins have been used since ancient times. An example is the use of bee wax and vegetal gums already in the Egyptian empires, say, 4000 BCE. Also the oriental lacquers as used in China and Japan are in use since 2000 BCE. For these lacquers the sap from the Rhus vernicifera plant is used. Their hardening comprises the autoxidative drying via unsaturated phenols. Another example is shellac, which originates from India (about 1000 BCE) and was used heavily as electrical insulator coatings and is still used in, for example, confection food. It is made from an insect excrement (Laccifer Lacca) and consists mainly of polyhydroxy acids. Finally, we mention rosin or colophonium, which came in to use in Northern Europe around 400 CE and is still used in inks. It is extracted from pine tree (trunks) and contains as monomer abietic acid (Figure 3.20a), which when mixed with drying fatty acids, polymerizes using the diene bond (esters) in autoxidative drying.

More recent natural resins originate from drying oils and mixtures with other monomers. For example, as linseed oil‐based paints dry rather slowly, paint boiling was introduced to speed up the polymerization process by adding amber, which contains mainly acids, to heated linseed oil. This process was used by Dutch painters in the sixteenth century for their paints [16]. This process later evolved to alkyd paints. Also purified cellulose (Figure 3.20b) became available in the twentieth century, for example, from flax (linen) fibers and yields a high molecular mass M_w and high glass transition temperature T_g material. Its processing is difficult, and therefore the raw material was chemically modified to cellulose esters. These esters also show less yellowing. Cellulose acetate ester, with 3 acetates per unit, shows a high transparency and is used nowadays in liquid crystal displays (LCDs). Also cellulose acetate mixed esters, such as cellulose acetate/butyrate (CAB), are used, for example, as flow improver in acrylic paints. Cellulose nitrate (<2.25 NO₃ per unit), another modified cellulose, was invented in 1869 by Eastman Kodak (celluloid) during the industrial revolution. It is used as a spray coating with solvents even though it contains extreme low solids content, typically less than 15–20% in so‐called spirits (a mixture of organic solvents). It has a very good wood coloring (“Anfeuerung”), flow, and gloss and is therefore still used for wooden furniture.

Clearly bio‐based monomers do not form directly a resin but provide building blocks for resins. Note that bio‐based means in this context obtained from agricultural sources, and this can be done directly, for example, by extraction, distillation, or purification or indirectly using intermediates for which (bio)chemical processing steps are involved, such as esterification, phosgenation, cyclization, hydrogenation, or oxidation. In a number of cases, enzymatic reactions are also used or even fermentation (in which yeasts or bacteria metabolize sugar into the desired building blocks). The first generation of bio‐based monomers from agricultural sources competed with food supply, either directly by using food (e.g. corn) as raw material or indirectly by using arable land. The second generation employs waste streams (i.e. uses corn stovers or lignin instead of corn). The subsequent polymerization of bio‐based monomers proceeds in the usual way.

It should be noted that bio‐based does not necessarily imply free of footprint. Energy is needed and there are still waste and pollution in the whole chain. A life cycle analysis (LCA) is needed to assess whether bio‐based raw materials have a lower footprint than petro‐based raw materials. It should also be noted that bio‐based does not also (necessarily) imply biodegradable, as normal performance criteria still apply.

Bio‐based monomers have a few typical problems, which we briefly discuss below:

• Availability and purity. Both are poor, generally scale dependent, and depend on time and site of production. While the typical purity of industrial scale terephthalic acid is >99.99%, the typical purity of bio‐based monomers is only >95%.
• Polarity. Most bio‐based monomers contain more heteroatoms, like oxygen and nitrogen, than petro‐based monomers and are therefore more polar. Consequently, they generally take up more moisture and hydrolyze more easily. Moreover, they also oxidize more easily via hydrogen abstraction followed by O₂ addition.
• Aromatic building blocks are less abundant from bio‐based sources than from petro‐based sources. As bio‐based sources furans and lignin‐derived compounds can be used, but bio‐based analogs are often “softer,” that is, have a lower T_g than their petro‐based originals.

Obvious examples of bio‐based polymers are polyester based on bio‐based diol and diacid building blocks. Easily accessible diacids comprise aliphatic diacids (saturated as well as unsaturated) from the citric acid cycle, succinic acid, malic acid, maleic acid, itaconic acid, citric acid, and lactic acid, as branching acid citric acid itself can be used. Further, aromatic diacids, such as furanedicarboxylic acid (obtained from oxidation of hydroxymethylfurfuryl alcohol), and polyols, such as sugars, can be used. The direct heat and acid treatment of hemicellulose (C5 sugars) containing waste results contains furfuryl alcohol (and/or derivates), which can be distilled out and with further hydrolysis leads to levulinic acid (Figure 3.21b).

As branching agent a polyol‐like glycerol is also an option, obtainable from hydrolysis of fats (Figure 3.22). For a more specific example, we mention the bio‐derived polyester [17] as obtained by polycondensation (like done for petro‐based monomers) using a combination of monomers to match the properties of well‐established, fully petro‐based polyesters, like functionality and T_g (Figure 3.23).

In this example a bio‐based diol intermediate is used that is already commercially available: isosorbide obtained from D‐glucose by hydrogenation, concomitant with a heat and acid treatment (Figure 3.21a). It has a rigid, cycloaliphatic structure, yielding a high T_g, so that it can act, even together with bio‐based aliphatic diacids, as replacement for the combination of aromatic diacids and aliphatic diols.

For epoxies, examples of partly bio‐based monomers are epoxidized fatty oils (Figure 3.24a) and diglycidyl ether of isosorbide or diglycidyl ether of bis‐furyl‐acetone (BFA) (Figure 3.24b).

Bio‐based monomer examples for acrylics are the dimethyl esters from the citric acid cycle‐derived unsaturated diacids, that is, itaconates (although they polymerize difficultly), bio‐acrylic acid, which is acrylic acid obtained from dehydration of glycerol or lactic acid, and esters from bio‐alcohols, such as (m)ethanol, butanol, and glycerol or glycol. There are currently no obvious bio‐sources for methacrylates and styrenes (Figure 3.25).

Bio‐based monomer examples for polyurethanes are the amino acid‐derived diisocyanates, for example, lysine esters, the dimer fatty acid diamine‐derived diisocyanates, the dimer fatty acid‐derived diols, and the polyols from sugars, such as isosorbide and isomers (Figure 3.26).

3.8 Drying Oils

Drying oils are natural oils obtained from vegetal or animal (mostly fish) sources containing unsaturated C=C groups. Since, in principle, all components originate from natural sources, the chemistry based on these raw materials could be designated as the ultimate renewable chemistry. They are called drying oils because they air dry, which is actually due to the fact that a chemical network formation occurs by the presence of ambient oxygen, which initiates and drives the crosslinking reactions of unsaturated groups by oxidation. We therefore speak of autoxidative drying, that is, the process occurs autonomously, although it is usually catalyzed by certain metals for reasons of shortened drying time. The concentration of the unsaturated groups is characterized by the iodine number (IN), that is, the number of grams I₂ taken up by (to color) 100 g oil. Different sources of oils give a different drying potential. An oil providing the required chemical film formation and a hard, dry film is called a drying oil and has IN > 140. A semidrying oil results in some film formation with remaining stickiness and has typically IN = 120–140, while a nondrying oil shows no film formation with IN < 120.

Typical drying oils are the triglycerides that are triesters of glycerol and fatty acids. As fatty acid saturated C₁₈, single‐C_18:1, double‐C_18:2, and triple‐C_18:3 unsaturated acids are used. The most reactive ones are the C_18:2 and C_18:3 fatty acids. In their natural structure the unsaturated bonds are separated by methylene groups, which are actually the reactive centers for the autoxidation process. A special case is formed by the conjugated dienes and trienes, which form by isomerization when the oils are heated. They are typically higher in reactivity toward autoxidative drying than the natural structures (Figure 3.27).

Different types of oils have different compositions in terms of fatty acids (Figure 3.28), and a typical composition and the associated IN are given in Table 3.2.

Autoxidative drying contains a number of stages. The first one is the induction period. A certain, small amount of natural antioxidants is always present and when they are consumed by oxidation, the second stage starts. This is the stage of oxygen take‐up via the formation of hydroperoxides on reactive diene and triene groups (Figure 3.29a) by hydrogen abstraction from the center methylene groups. This can be a spontaneous reaction but also may be metal catalyzed. An often used explanation for the catalyzed reaction is the Haber–Weiss cycle using the cobalt(II/III) equilibrium. The steps in this cycle are

The overall reaction is therefore 2ROOH → ROO^• + RO^• + H₂O. A similar pathway is used for Pb(II/III), Mn(II/III), and Fe(II/III). The third stage is the decomposition of the hydroperoxides to radicals (RO^• and ROO^•, Figure 3.29b), followed by the addition of radicals to double bonds (RO^• + C=C → RO-C-C^•). Finally, we have the recombination of radicals (crosslinking) to ether (R-O-R) or peroxide (R-O-O-R) bonds (Figure 3.29c). As Co is toxic, the trend is to use nontoxic Mn‐based catalysts and this topic will be revisited in Chapter 5.

3.9 Alkyds

Alkyds are polyesters made with fatty monoacids or drying oils (glycerides). The usual polyester monomers are polyols, for example, glycerol, trimethylolpropane and glycols, and diacids, such as phthalic anhydride (PA) and isophthalic acid. A simplified alkyd structure is given in Figure 3.30.

Crosslinking proceeds via autoxidation like for drying oils. They are catalyzed by siccatives (Co(II)octoate (primary) and auxiliary Ca(II)/Zr(IV)octoates). The polyester backbone increases T_g with respect to drying oils. Moreover, they show a faster (physical) drying and increased hardness. They are classified with respect to their fatty acid content by the oil length (OL), the weight percentage of triglyceride in the resin (calculated even when using fatty monoacids instead of triglycerides). A short oil has an OL of 30–42% and contains 38–46% of PA, while a medium oil has an OL of 43–54% and contains 30–37% of PA. Finally, a long oil has an OL of 55–72% with 15–30% PA.

For the alkyd polycondensation process, one has two options. The first is the fatty acid process, proceeding as for polyester resins but using fatty acid as monofunctional monomer. Branching is provided by trimethylolpropane or pentaerythritol. They have the better hydrolysis resistance of the alkyds as there is no glycerol in the backbone. The second is the glyceride‐oil process. This is a cheaper and a bio‐based process. It uses glycerol from oil in the polyester backbone and proceeds by having first a transesterification of oil with excess glycerol, followed by polycondensation with diacids (Figure 3.31). Direct addition of glycerol and oil to diacids leads to glyptal formation (i.e. nonmiscibility oil/polyester) and should be avoided. As a disadvantage, the resulting material shows less resistance to hydrolysis.

Fatty oils and fatty acids are as such already bio‐based raw materials, but they can be food competitive. A solution is to use nonfood‐competitive raw materials, for example, castor bean oil (which is not suitable for human consumption). Here glycerol can be used as branching agent and combined with, for example, abietic acid (rosin), while phthalic acid could be replaced by furanedicarboxylic acid (Figure 3.32).

Although WB latex paints have taken over the architectural coating market, for metal and wood parts, alkyd emulsions are used as an alternative to alkyd paints, thereby avoiding the use of (neurotoxic) solvents and solvent emissions. Alkyd emulsions are dispersions of solventless alkyd resin droplets in water. When the paint dries, the water evaporates, and the emulsion changes from oil in water (O/W) to water in oil (W/O), while further evaporation of water leads to the dry film. The properties of the dried and autoxidatively crosslinked film are largely comparable with those derived from solventborne alkyd paints [18].