Understanding Additives 9783748602385

Citation preview

Bodo Müller

Understanding Additives 2nd Revised Edition

Cover: kwanchaift – AdobeStock.com

Bibliographische Information der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliographie; detaillierte bibliographische Daten sind im Internet über http://dnb.d-nb.de abrufbar.

Bodo Müller Understanding Additives, 2nd Revised Edition Hanover: Vincentz Network, 2019 European Coatings Library ISBN 3-7486-0167-0 ISBN 978-3-7486-0167-8 © 2019 Vincentz Network GmbH & Co. KG, Hanover Vincentz Network, Plathnerstr. 4c, 30175 Hanover, Germany This work is copyrighted, including the individual contributions and figures. Any usage outside the strict limits of copyright law without the consent of the publisher is prohibited and punishable by law. This especially pertains to reproduction, translation, microfilming and the storage and processing in electronic systems. Discover further books from European Coatings Library at: www.european-coatings.com/shop Layout: Vincentz Network, Hanover, Germany Printed by: BWH GmbH, Hanover, Germany

European Coatings Library

Bodo Müller

Understanding Additives 2nd Revised Edition

Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

THERE’S MORE!

Discover our other bestsellers: MICROBICIDES IN COATINGS · This book from Frank Sauer provides a comprehensive overview of the working mechanisms and possible applications of microbicides for coatings – invaluable for formulators and technicians as well as for business people with a basic knowledge of chemistry and biology.

Dr Frank Sauer finalised his PhD in Organic Chemistry at the University of Cologne, Germany, in 1993. Ever since, he has been working in the coatings industry – since 2012 as a Manager Regulatory Affairs for the Lanxess Group.

2017, 156 pages, hardcover, 149 € order no. 400, eBook: 20041

www.european-coatings.com/shop

Brock + Groteklaes + Mischke EUROPEAN COATINGS HANDBOOK

Mannari + Patel UNDERSTANDING COATINGS RAW MATERIAL

Wernfried Heilen ADDITIVES FOR WATERBORNE COATINGS

2010, 432 pages, hardcover 169 €, order no. 544 eBook: 544_EBOOK

2015, 287 pages, hardcover 99 €, order no. 599 eBook: 599_EBOOK

2009, 224 pages, hardcover 149 €, order no. 473 eBook: EBOOK_473

EUROPEAN

C OATINGS library

Vincentz Network · P.O. Box 6247 · 30062 Hannover · Germany T +49 511 9910-033 · [email protected]

Foreword The author was not surprised by the success of the first edition, as there had not been a comparable textbook on the market that provided detailed explanations of paint additives. The author has therefore gladly taken on the task of revising and updating the first edition. To this end, he has eliminated errors, added several new figures and the completely new Chapter 11. Additives are substances which are added to a coating composition in small amounts (commonly less than one percent) to alter the properties of a liquid paint or the resultant solid coating in a particular direction. For example, one property of a liquid paint is its ease of application, which can be influenced by rheological additives (e.g. sag control agents). One property of solid coatings is their anti-corrosion property, which can be influenced by corrosion protection additives or adhesion promoters. Additives act like “medicines for paints” and so can have unwanted side-effects as well. The present book is primarily a textbook and originated from lectures given by the author at the University of Applied Sciences in Esslingen, Germany. Its objective is to provide a description of the chemistry and applications technology behind various paint additives. It does not cover test methods. On account of the sheer diversity of additives that exist, the book must further restrict itself to the most important types – but it looks at these in detail. It also contains numerous photographs/diagrams to illustrate the various kinds of damage which additives prevent, as a picture says more than a thousand words. In specific cases, coatings formulations (starting formulations obtained from suppliers of raw materials) are presented to explain the specific use of an additive, thereby also illustrating the needs of the coatings industry. Patent restrictions and registered trademarks (e.g. ™ or ®) are not mentioned explicitly. Product and trade names are only used where otherwise unavoidable. It should also be noted that product or trade names can change as a result of mergers and acquisitions. Nonetheless, most of the raw materials described herein or their equivalents are available worldwide. The present textbook seeks to familiarise laboratory assistants, technicians, graduates, engineers, bachelors, masters and chemists with this class of raw materials for paints. It presupposes a basic knowledge of chemistry. Moreover, it will serve as a reference work for all readers interested in raw materials, paints and coatings. Würzburg, June 2019 Bodo Müller [email protected]

5

Exclusion of liability It should be noted that this book reflects the author's personal views, based upon his own knowledge. This does not absolve readers of the responsibility to perform their own tests with respect to the uses and applications of various processes or products described herein, and/or of obtaining additional advice regarding the same. Any liability of the author is excluded to the extent permitted by law, subject to legal interpretation.

FR CROS®

6

Protecting what really matters in life 

Halogen-free fire protective ingredients



Anti-corrosion pigments



Dispersing aids

Additives for paints and coatings Chemische Fabrik Budenheim KG | Budenheim | Germany www.budenheim.com | [email protected]

Contents

Contents 1 Wetting and dispersing agents.....................................................................11 1.1 Dispersing process................................................................................................... 11 1.2 Stabilisation of dispersions................................................................................... 14 1.2.1 Electrostatic stabilisation....................................................................................... 16 1.2.2 Steric stabilisation............................................................................................................... 17 1.3 Typical wetting and dispersing agents................................................................ 19 1.3.1 Dispersing agents..................................................................................................... 19 1.3.2 Wetting agents (surfactants)................................................................................. 24 1.4 Surface treatment of pigments and fillers......................................................... 31 1.4.1 Organic coloured pigments................................................................................... 31 1.4.2 Titanium dioxide...................................................................................................... 34 1.4.2.1 Inorganic after-treatment....................................................................................... 34 1.4.2.2 Organic after-treatment.......................................................................................... 35 1.4.3 Fillers........................................................................................................................... 36 1.4.4 Fumed or pyrogenic silica...................................................................................... 37 1.4.5 Lamellar metal pigments........................................................................................ 39 1.4.5.1 For solvent-borne paints......................................................................................... 39 1.4.5.2 For water-borne paints............................................................................................ 43 2 Substrate-wetting additives/levelling agents.............................................47 2.1 Surface defects......................................................................................................... 47 2.2 Silicone additives..................................................................................................... 50 2.2.1 Polydimethyl siloxanes........................................................................................... 50 2.2.2 Chemically modified silicone additives............................................................... 51 3 Defoamers..........................................................................................................55 3.1 Foam types and foam stabilisation...................................................................... 55 3.2 Types and mode of action of defoamers............................................................ 59 3.2.1 Defoamers for water-borne paints....................................................................... 59 3.2.2 Defoamers for solvent-borne paints.................................................................... 62 3.3 Deaeration of coating powders............................................................................. 63 4 Rheologically active additives........................................................................65 4.1 Rheological additives for water-borne and latex paints.................................. 68 4.1.1 Layer silicates............................................................................................................ 68 4.1.2 Fumed (pyrogenic) silica........................................................................................ 72 7

Contents 4.1.3 Polymeric rheology modifiers............................................................................... 73 4.1.3.1 Acrylates.................................................................................................................... 73 4.1.3.2 Hydrophobically modified ethylene oxide urethanes (HEURs)..................... 75 4.1.3.3 Polysaccharides........................................................................................................ 79 4.2 Rheological additives for solvent-borne paints................................................. 83 4.2.1 Organoclays............................................................................................................... 83 4.2.2 Fumed (pyrogenic) silica........................................................................................ 85 4.2.3 Rheological additives based on urea................................................................... 87 4.3 Rheological additives in stoving enamels........................................................... 90 5 Catalysts........................................................................................................... 93 5.1 Driers.........................................................................................................................93 5.1.1 Anti-skinning agents..............................................................................................97 5.2 Catalysts for polyurethane coatings.................................................................... 99 5.3 Catalysts for two-component epoxy coatings..................................................105 5.4 Acid catalysts for stoving enamels.....................................................................107 6 Adhesion promoters..................................................................................... 113 6.1 Silane adhesion promoters..................................................................................115 6.2 Silane adhesion-promoting primers..................................................................117 6.3 Thin polymeric adhesion layers..........................................................................119 6.4 Aminosilane as hardener in two-component epoxy paints...........................120 7 Corrosion protection additives................................................................... 123 7.1 Corrosion of metals...............................................................................................123 7.2 Distinction from anticorrosive pigments..........................................................125 7.3 Organic corrosion protection additives.............................................................128 7.4 Corrosion inhibitors for metal pigments in aqueous alkaline paint media....129 8 Protection of coatings from weathering.................................................. 133 8.1 Photooxidation/UV degradation.........................................................................134 8.1.1 Absorption and emission of light.......................................................................134 8.1.2 Photooxidation of polymers and paint resins..................................................136 8.2 Damage done to coatings by weathering.........................................................140 8.2.1 Chalking...................................................................................................................140 8.2.1.1 Photocatalytic oxidation cycle.............................................................................141 8.2.2 Fading of organic colorants.................................................................................143 8.2.3 Embrittlement, crack formation and delamination........................................144 8.2.4 Special aspects of wood weathering.................................................................145 8

Contents 8.2.5 Damage specific to two-coat metallic coatings...............................................147 8.3 Stabilisation of coatings against photooxidation...........................................148 8.3.1 Pigmentation...........................................................................................................148 8.3.2 Light stabilisers......................................................................................................151 8.3.2.1 UV absorbers..........................................................................................................151 8.3.2.2 Free-radical scavengers.........................................................................................155 9 Photoinitiators/UV initiators...................................................................... 161 9.1 UV-curing coatings................................................................................................161 9.2 Mode of action of photoinitiators......................................................................163 10 Coalescing agents.......................................................................................... 167 10.1 Film formation by primary dispersions (latices).............................................167 10.2 Mode of action of coalescing agents.................................................................171 11 Neutralising agents....................................................................................... 173 11.1 Neutralising agents for binders bearing carboxyl groups............................173 11.2 Neutralising agents for binders bearing amino groups................................175 Author.............................................................................................................. 177 Index................................................................................................................. 179

9

Dispersing process

1 Wetting and dispersing agents 1.1 Dispersing process Before the dispersing process is discussed, it would be helpful to characterise pigment particles (Figure 1.1). Small, single crystals which are formed during pigment synthesis are called primary particles. Aggregates are primary particles which grow together via their faces; the dispersing process does not generally separate them. Agglomerates are associ­ ates of primary particles and/or aggregates joined by their edges (Figures 1.1 to 1.3). Figure 1.2 shows a scanning electron micrograph and Figure 1.3 an optical micrograph of pigment Red 3 as a typical example of agglomerates of an organic coloured pigment. Grinding processes are classified as true size reduction (crushing of primary particles) and deagglomeration (grinding of agglomerates to yield aggregates and/or primary particles).

Figure 1.1: Simplified diagram of primary particles, aggregates and agglomerates

Figure 1.2: Scanning electron micrograph of pigment Red 3

Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

11

Wetting and dispersing agents In the coatings sector, dispersing or milling is taken to mean the homogeneous distribution of disperse solid particles (e.g. pigments) in a liquid medium (mostly a solution of binder), i.e. deagglomeration. As the dispersing of pigments is the most important step in paint manufacturing and is necessary for an understanding of dispersing agents, the underlying process will be discussed here briefly. During dispersing, adhesive forces (e.g. van der Waals) acting between the pigment particles must be overcome [1]. The purpose of dispersing agents is to stabilise the defloccu­ lated pigment dispersions, which were produced by the dispersing process, for a protracted

Figure 1.3: Optical micrograph of pigment Red 3 (scale bar 1 mm)

Figure 1.4: Coatings of pigment Blue 60 mixed with titanium dioxide after different dispersing periods in an aqueous latex gloss enamel (development of colour strength)

Figure 1.5: Optical micrographs (transmitted light) of a dispersion of phthalocyanine blue pigment after different dispersing periods

12

Dispersing process period of time; i.e. the pigment particles must be prevented from flocculating (see Chap­ ter 1.2) [2]. The purpose of the dispersing process is to separate agglomerates. Ideally, the outcome is a dispersion of primary particles and aggregates. In practice, though, it is often only de­ agglomeration of large agglomerates to smaller ones which occurs [3]. Excessive dispersing (leading in extreme cases to crushing of primary particles) should be avoided at all costs, because after-treated pigment surfaces can be damaged and application properties can be impaired (see Chapter 1.4). Dispersing proceeds in three steps: – Wetting of pigment agglomerates – Deagglomeration of pigment agglomerates – Stabilisation of the resulting dispersion against flocculation (see Chapter 1.2 below). Pigment agglomerates are wetted in two steps. The liquid phase spreads over their surface and then penetrates the pores or voids, displacing air. Mechanical deagglomeration of the pigment agglomerates increases the tinting strength of coloured pigments (Figure 1.4) and renders them more economical. Figure 1.5 shows optical micrographs (transmitted light) of a dispersion of phthalocy­ anine blue pigment after different dispersing periods; the qualitative decrease in agglom­ erate size is clearly evident.

Figure 1.6: Particle size distribution of an organic red pigment in an aqueous binder (blank, d50 = 1.3 μm; see also Figure 1.8)

13

Wetting and dispersing agents Figures 1.6 and 1.7 show the quantitative decrease in agglomerate size with increase in dispersing duration as measured by the particle size distribution. After 10 minutes, the agglomerates have disappeared (Figure 1.7).

1.2 Stabilisation of dispersions During dispersing, adhesive forces (e.g. van der Waals) acting between the pigment parti­ cles must be overcome. Dispersing agents are necessary for stabilising the deflocculated pigment dispersion which was produced by the dispersing process; i.e. flocculation caused by these forces of attraction must be inhibited [2]. It is essential to distinguish between the terms dispersing agent and dispersion medium. A dispersing agent is an additive which improves the stability of pigment dispersions whereas a dispersion medium is a liquid phase wherein pigments are dispersed. Flocculation is the association of pigment particles, which have been dispersed in a liquid paint medium, and is the result of forces of attraction (e.g. van der Waals) between the particles. Flocculation reduces the size of the phase boundary between pigment and dispersion medium. Most disperse systems are thermodynamically unstable, resulting in a decreased interfacial surface area due to flocculation [3]. Why?

Figure 1.7: Particle size distribution of an organic red pigment in an aqueous binder (after 10 minutes’ dispersing, d50 = 0.51 μm; see also Figure 1.8)

14

Stabilisation of dispersions The work (energy input) needed to move a molecule from inside a phase to the surface is calculated by: where

W = γ · ΔA W γ ΔA

is the work [energy] is the surface tension [force/length = energy/area] [3] is the increase in interfacial area [area]

The work W is proportional to the increase in interfacial area ΔA and to the surface tension γ. Consequently, the larger the interfacial area and the higher the surface tension, the higher is the energy content of the disperse system and the less thermodynamically stable it is. Flocculation lowers the energy content of a disperse system because the interfacial area ΔA decreases; i.e. it shifts the disperse system towards a thermodynamically stable state (Figure 1.9). Pigment and polymer dispersions are often stable for years because dispersions they are metastable. The term metastability is illustrated in Figure 1.9. So, why do pigment dispersions have a long shelf life, i.e. why they are metastable? The answer is that electrostatic or steric stabilisation can prevent them from flocculating. These terms are explained below. The combination of both these stabilisation mechanisms is called electrosteric stabilisation.

Figure 1.8: Reduction paste of an organic red pigment in an aqueous binder: blank and after 10 minutes’ dispersing time (see also particle size distributions in Figures 1.6 and 1.7)

Figure 1.9: Thermodynamic states of pigment dispersions

15

Wetting and dispersing agents Table 1.1:  Isoelectric points (IEPs) of certain oxides [5] Oxide

IEP (pH)

MgO

12.5

Al2O3

9

ZnO

9

Cr2O3

7

Fe3O4

6.5

Fe2O3

6.7

TiO2

4.5 to 6.5

Stabilisation is a kinetic effect. Transfor­ mation into the thermodynamically stable state by flocculation is prevented or re­ tarded by a high energy of activation (Fig­ ure 1.9).

1.2.1 Electrostatic stabilisation

Inorganic pigments and fillers dispersed in water especially (which has a high dielec­ tric constant) mostly bear electric charges. 2 SiO2 Since the overall disperse system is un­ charged, the liquid phase must contain an equal number of counter ions in close proximity to the particles (Figure 1.10). An electric double layer (ion cloud) forms and flocculation is prevented because of the electrostatic repulsion of like charges (this is known as electrostatic stabilisation) [3]. It is important for the forces of repulsion (Coulomb’s law) to extend further into the dispersion medium than the forces of attraction (van der Waals forces; Figure 1.10). Wher­ ever there are opposite charges (e.g. in the case of different pigments), co-flocculation occurs because of electrostatic attraction. Adding electrolytes (salts) causes the electric double layer (ion cloud) to contract and may lead to flocculation [3]. The Schulze-Hardy law describes how the ionic charge of added salts influences their flocculating power: – In negatively charged dispersions (the most common), the flocculating power of cations increases with increase in ca­ tionic charge: Na+ < Ca2+ < Al3+ or Fe3+ – In positively charged dispersions (rare, e.g. cathodic electrodeposition primers), the flocculating power of anions increa­ ses with increase in anionic charge: Cl− < SO42− < PO43−. Figure 1.10: Simplified diagram of the electric double layer of two pigment particles

16

In practice, therefore, water-borne paints should not contain any electrolytes (salts).

Stabilisation of dispersions There are various mechanisms by which the surfaces of pigment or filler particles can ac­ quire charges: – Dissociation of functional groups on the particle surfaces – Adsorption of ions (mostly polyanions); see dispersing agents (Chapter 1.3.1).

Dissociation of functional groups on particle surfaces

Oxides have hydroxyl groups on their surfaces [3] that may react as acids or bases, in ac­ cordance with the oxide concerned. The pH at which the surface charge is zero (point of zero charge) is also called the isoelectric point (IEP) [4]. Above the IEP, the oxide surface is negatively charged due to deprotonation while, below the IEP, it is positively charged due to protonation [3]. The charge density increases with increase in distance between the pH of the dispersion medium and the IEP. The data shown in Table 1.1 apply to chemically pure oxides. There may be significant deviations from these IEP values in the case of industrial pigments or fillers whose surfac­ es have been after-treated (e.g. titanium dioxide; see Chapter 1.4.2). The type of crystalline structure, too, can influence the IEP. Electrostatic stabilisation is especially important in water-borne paints and latex paints because of the high dielectric constant of water. However, studies have shown that the elec­ tric charge on pigment surfaces also plays important role in solvent-borne paints [6]; the same pigment dispersed in different paint resins may exhibit opposite charges or none at all.

1.2.2 Steric stabilisation It has been known for more than 100 years that aqueous disper­ sions (colloids) can be readily stabilised against flocculation by adding water-soluble poly­ mers (so-called protective col­ loids), such as gelatin, casein, and polyvinyl alcohol. In con­ trast to electrostatic stabilisa­ tion, stabilisation effected with polymers is insensitive to the addition of electrolytes (salts). As a rule, adding suitable pol­ ymers is the only way to stabilise

Figure 1.11: Steric stabilisation (simplified diagram; not to scale)

17

Wetting and dispersing agents dispersions in organic solvents. The polymers (oligomeric dispersing agents or paint resins) must adsorb on the surface of the dispersed particles; i.e. they must displace adsorbed sol­ vent and/or surfactant molecules. As pigment particles approach each other, the polymer segments become restricted in their movement and the entropy decreases; the outcome is a repulsive force (Figure 1.11). For this reason, steric stabilisation is also referred to as entropic repulsion or deflocculation. Studies [7, 8] have shown that, for steric stabilisation of dispersed particles up to 10 μm (10,000 nm) in diameter, the steric barrier need only be 10 nm (Figure 1.12). The diagram in Figure 1.11 suggests that surfactants (wetting agents) can also act as steric stabilisers (especially if it is not realised that Figure 1.11 is not to scale). Low-molecular surfactants (wetting agents) do not act as steric stabilisers as their molecules are much too small (see Chapter 1.3.2). Only oligomers or polymers have the necessary molecular size.

Requirements for steric stabilisation

a) The polymers must strongly adsorb on the pigment surface by appropriate func­ tional groups (anchoring groups). b) The polymers must have sufficiently long chain segments (barrier groups) which readily dissolve in the dispersion medium (organic solvents or water), a process that leads to widening of polymer chains. The addition of poor solvents can cause these polymer chains to coil and so lead to flocculation (see “Dispersing agents” in Chapter 1.3.1). c) Polymers (oligomers) of me­ dium molar mass are optimal: – If the molar mass is too low, the chain is not long enough. – If the molar mass is too high, bridging flocculation may oc­ cur (see “Dispersing agents” in Chapter 1.3.1). Moreover, if the molar mass is too high, incompatibility may occur, or the viscosity may increase. d) A minimum polymer concentra­ tion is necessary; if the concen­ tration is too low, flocculation Figure 1.12: Size relationships involved in steric stabimay occur, especially in the case lisation; the black rim around the particles shows the 10 nm-thick steric barrier (diagram roughly to scale) of high molar masses (see c).

18

Typical wetting and dispersing agents

1.3 Typical wetting and dispersing agents Unfortunately, manufacturers provide hardly any information about the chemical compo­ sition of their wetting and dispersing agents, and that makes it difficult to explain the ef­ fectiveness of these additives. Moreover, they often employ the terms wetting and dispers­ ing agents indiscriminately because the distinction between them is fuzzy. In this book, they were distinguished by molar mass: – Wetting agents (surfactants) are low-molecular amphiphilic substances. – Dispersing agents are oligomers or polymers which stabilise dispersions of pigments and fillers against flocculation. Let us look at dispersing agents first.

1.3.1 Dispersing agents Polyanionic dispersing agents for electrostatic stabilisation

Polyanions serving as dispersing agents or dispersing additives adsorb primarily on inorgan­ ic pigment and filler surfaces, which they charge by means of their own electric charges. The purpose of such polyanionic dispersing agents is to increase the level of repulsion and, therefore, to improve electrostatic stabilisation. They achieve this by: – Boosting the ‘like’ charge – Reversing the charge of oppositely charged pigments (prevention of co-flocculation) – Complexing multivalent cations (e.g. Ca2+; thereby increasing the radius of action of electrostatic forces) Typical polyanionic dispersing agents are polycarboxylates (mostly salts of polyacrylic acids; Figure 1.13). Figure 1.14 schematically shows the adsorption of a salt of polyacrylic on zinc oxide (IEP at pH 9). Advantages of polycarboxylates + Hydrolytic resistance + Similarity to paint resins (compatibility, film quality) Disadvantages of polycarboxylates – Higher cost (compared to polyphosphates) – Sensitive to multivalent cations – Relatively high addition level.

19

Wetting and dispersing agents Further examples of polyanionic dis­ persing agents are polyphosphates – Linear polyphosphates: Nan+2PnO3n+1 (Calgon) n = 2: pyrophosphate Na4P2O7 n = 4: tetraphosphate Na6P4O13 – Cyclic metaphosphates: (NaPO3)n The charge and, therefore, the dis­ persing efficiency increase with in­ crease in n. Figure 1.13: Salt of a polyacrylic acid as a dispersing agent for water-borne paints

Advantages of polyphosphates + Lower cost (compared to polycarboxylates) + Complexing of multivalent cations (e.g. Ca2+) + Relatively low addition level (0.2 to 0.5 wt.%) Disadvantages of polyphosphates – Slow hydrolysis to (mono-)phosphate – Soluble phosphate salts may lead to leaching and crystallisa­ tion on coatings A combination of polyacrylate and polyphosphate is often used in latex paints.

Polymeric dispersing agents for steric stabilisation

Figure 1.14: Charging of zinc oxide pigment at pH 9 by a polyacrylic acid salt (simplified diagram, not to scale)

20

From Chapter 1.2.2 “Steric stabilisa­ tion”, it follows that block and graft copolymers (Figure 1.15) are better at stabilisation than random copoly­ mers or homopolymers. Given significant difference in polarity between blocks A and B

Typical wetting and dispersing agents (Figure 1.15), AB block copoly­ mers can have amphiphilic properties (see polymeric sur­ factants); in such cases, the term “wetting and dispersing agents” which is often used by the manufacturers is correct. A precise description of the chemical composition of poly­ meric dispersing agents is rarely found in the literature. Steric stabilisation of titani­ um dioxide in methyl ethyl ke­ tone by AB block copolymers of 2-vinylpyridine and methyl methacrylate (Figure 1.16) is de­ scribed in [9]. Even a 2-vinylpyri­ dine content of just 18 mol-% in the block copolymer is enough to effectively stabilise titanium dioxide dispersions (rutile) against flocculation. The degree of polymerisation of 2-vinylpyri­ dine in the block copolymer con­ taining 18 mol-% of 2-vinylpyri­ dine computes to n ≈ 27 while the corresponding figure for methyl methacrylate is m ≈ 122 (Figure 1.16). A second example of a poly­ meric dispersing agent em­ ployed for steric stabilisation is shown in Figure 1.17. Polyure­ thane chemistry can be used to build up this AB block copoly­ mer step by step. The nitrogen atoms in both AB block copolymers (Figures 1.16 and 1.17) are Lewis bases

Figure 1.15: Block and graft copolymers as dispersing agents (multiblock copolymers could also be used)

Figure 1.16: AB block copolymer as a dispersing agent for titanium dioxide in an organic solvent

21

Wetting and dispersing agents that are capable of interacting with metal ions (Lewis acids). They act as anchoring groups, especially for inorganic pigments and fillers. Diluting sterically stabilised pigment dispersions with unsuitable solvents can cause the solvated polymer chains of the dispersing agent to coil and, therefore, to flocculate (Figure 1.18).

Figure 1.17: AB block copolymer as a dispersing agent

Figure 1.18: Effect of solvents on steric stabilisation (diagram not to scale)

22

Figure 1.19: Principle behind bridging flocculation (diagram not to scale)

Typical wetting and dispersing agents Bridging flocculation is especially likely to occur when the polymer concentration is too low and the molar mass is very high (Figure 1.19). The use of certain dispersing agents to effect deliberate but reversible bridging flocculation is called controlled flocculation (see below).

Mixtures of different pigments

At this stage, a brief discussion of the properties of different mixed pigments is needed. Since paints generally contain more than one pigment, the various pigments may sepa­ rate on account of their different mobility (due to different sizes and/or densities). This separation occurs when different pig­ ments in the fresh, still flowable paint are no longer distributed uniformly. There are two different types of pigment separation: – Vertical pigment separation (floating): Local differences in con­ centration on the paint surface is called vertical pigment separation (Figure 1.20), and often leads to Bénard cells (Figure 1.21) or to strea­ king. The surface of the coating does Figure 1.20: Vertical pigment separation and not have a uniform colour, but instead cell formation (optical micrograph of a coalooks speckled or streaky. ting containing titanium dioxide and phthalo– Horizontal pigment separation cyanine blue) scale bar 0.5 mm (flooding): If the differences in con­ centration occur not on the paint surface but rather perpendicularly to it, this is called horizontal pigment se­ paration. In this case, the coating has a uniform colour. This paint defect only becomes apparent when, e.g., the rub-out test is carried out or a coated glass plate is viewed from above and below. Both types of pigment separation may be prevented or reduced by controlled floccu­ lation or – preferably – controlled co-floc­ culation (Figure 1.22).

Figure 1.21: Bénard cells (scanning electron micrograph of a coating containing titanium dioxide and pigment Red 3)

23

Wetting and dispersing agents

Advantages of controlled flocculation + N  o separation of pigments of different density → Little or no pigment separation → Little or no occurrence of Bénard cells or streaking.

+ Rheological effect: Increase in non-Newtonian flow (reversible three-dimensional network) → Improved application properties (e.g. less sagging) → Less settling of pigments.

Figure 1.22: Controlled flocculation (diagram simplified and not to scale)

+ No direct contact between the pigments in the flocs → Readily redispersible (mostly by stirring) → “Soft” sediment. Finally, it should be stated that in many cases paint resins (mostly oligomers) can stabilise pigment dispersions sterically.

1.3.2 Wetting agents (surfactants)

Figure 1.23: Sodium stearate

24

This chapter covers the “classic” low-molecu­ lar wetting agents. Wetting agents, which are also called surfactants, emulsifiers, sur­ face-active, and amphiphilic compounds, are substances which decrease surface or interfa­ cial tension at phase boundaries. They there­ fore improve wetting. The terms wetting agent and surfactant are used synonymously in this book. The first wetting agents to be employed were salts of fatty acids (soaps); a typical example is sodium stearate (salt of stearic acid; Figure 1.23). Sodium stearate has a

Typical wetting and dispersing agents Table 1.2:  Influence of the hydrophobic group on surface activity in water (surface tension) Surface tension [mN/m] Water

73

+ hydrocarbon surfactants

40 to 25

+ silicone surfactants

30 to 20

+ fluorinated surfactants

25 to 15

hydrophilic group (carboxylate) and a hy­ drophobic group (long-chain alkyl). Each molecule has both hydrophilic and hydro­ phobic properties (“head-and-tail structure”) and is said to be amphiphilic. The maximum length of the stearic acid molecule is about 2.2 nm (Figure 1.23). Sodium stearate is therefore unable to stabilise pigment dispersions sterically because that requires a chain length of at least 10 nm (Chapter 1.2.2) [7, 8]. Wetting agents align themselves at phase boundaries, where they lower the interfacial tension and improve wetting. Figure 1.24 shows the self-alignment of wetting agents at the water/air phase boundary. The hydrophilic groups align themselves with the water, and the hydro­ phobic groups, with the air. The water sur­ face now no longer consists of water mole­ cules (high surface tension) but of hydro­ carbon chains of lower surface tension. Usually, wetting agents are classified by the charge on the hydrophilic group. There are anionic, cationic, amphoteric and nonionic types (Figure 1.25). Wetting agents can be further classified on the basis of the chemical composition of the hydrophobic group, namely as hydrocarbon, silicone and fluorinated sur­

Figure 1.24: Self-alignment of wetting agents at the water/air phase boundary

Figure 1.25: Classification of wetting agents by the charge on the hydrophilic group

25

Wetting and dispersing agents Table 1.3:  Influence of the hydrophilic group on surface activity (surface tension) of different nonylphenol ethoxylates* (see Figure 1.27) Surface tension [mN/m] Water + surfactant

73 0.01 g/l

0.1 g/l

1 g/l

n=6

38

30

29

n = 10

45

31

30

n = 30

54

40

36

* It should be noted that concerns have recently been raised about the toxicity of nonylphenol ethoxylates.

factants. The greater the hydropho­ bicity of the “tail”, the more the sur­ face tension of water decreases when treated with wetting agents; i.e. the more efficient the wetting agent becomes (Table 1.2). A typical hydrocarbon surfactant is sodium stearate (Figure 1.23). Examples of silicone and fluorinated surfactants are presented in Figure 1.26. The influence of the hydrophilic group on surfactant activity (sur­ face tension) of nonylphenol ethox­ ylates (Figure 1.27) is shown in Ta­ ble 1.3. The longer the hydrophilic polyether chain, the more hydro­ philic these nonionic surfactants become. At the same time, the sur­ face tension of the aqueous sur­ factant solution increases. In other words, as the “head” becomes more hydrophilic, the surfactant becomes less efficient. Furthermore, Table 1.3 shows the influence exerted by the sur­ factant concentration. As expected, increasing the concentration causes the surface tension of the aqueous solution to decrease; i.e. the effec­ tiveness of the added surfactant in­ creases. Table 1.3 also shows clearly that 0.1 g/l is the optimal level of surfactant addition; increasing it fur­ ther to 1 g/l does not provide much additional help.

Secondary effects Figure 1.26: Typical silicone and fluorinated surfactants

26

Wetting agents (like all paint addi­ tives) have an optimal addition level

Typical wetting and dispersing agents which should not be exceeded because wet­ ting agents (again, like all paint additives) can have unwanted side-reactions. For exam­ ple, wetting agents increase the hydrophilic­ ity of coatings. An additional unwanted side-reaction of surface-active compounds, such as wetting agents, is foaming in water, with the foam lamellas (which have large surface areas) being stabilised by wetting agents. One remedy for this is to add a de­ foamer (see Chapter 3). To sum up so far, the more hydrophobic the “tail” and the less hydrophilic the “head”, the more effective is the wetting agent. However, solubility acts as a limiting condition in water. For example, in the case of the nonylphenol ethoxylates (Figure 1.27), the surfactant having n = 4 is not completely soluble in water. The nonylphenol ethoxylates presented in Figure 1.27 are the most important sub­ set of the alkylphenol ethoxylates (APEO) and are nowadays not more used in Central Europe because of ecotoxicity. As with this group of wetting agents the relationships between structure, application quantity and effect can be demonstrated impressive­ ly (Table 1.3), these APEOs are still men­ tioned in this book.

Figure 1.27: Nonylphenol ethoxylates (nonionic hydrocarbon surfactants)

Figure 1.28: Adsorption of a cationic surfactant on a negatively charged metal oxide surface (highly simplified; not to scale)

Orientation at phase boundaries Wetting agents align themselves at phase boundaries and are able to alter the polarity (and the wettability) of solid surfaces. For ex­ ample, a cationic surfactant can hydropho­ bise a hydrophilic, negatively charged metal oxide surface (pigment or filler; Figure 1.28).

Figure 1.29: Adsorption of an anionic surfactant on a nonpolar organic pigment surface (highly simplified; not to scale)

27

Wetting and dispersing agents While surfactants can be used to improve wetting of hydrophobic organic pigments in water-borne paints (Figure 1.29), they are also often employed in solvent-borne paints. Furthermore, Figure 1.29 shows that adsorption of an anionic surfactant causes the surfa­ ce of an organic pigment to become negatively charged, and that leads to electrostatic stabilisation of the pigment. Pigment wetting is a requirement for successful pigment dis­persion (see Chapter 1.1). Paint properties, such as gloss and tinting strength, can be improved by wetting agents and, as a result, the time necessary for dispersing can be re­ duced (savings on energy and costs). In Western Europe, 2.5 million tonnes of surfactants are used, of which 2 % (= 50,000 tonnes) are used for paints, coatings and plastics [22].

Examples of wetting agents which are important in coatings technology Three important wetting agents for paints are described in detail below. The first one is the amphoteric (zwitterion) lecithin, a nontoxic renewable raw material shown in Figure 1.30. Note that there are two hydrophobic groups (two “tails”). Lecithin is produced from soybean oil (soybean lecithin) and is chiefly used in solvent-borne paints. Soybean fatty acids occur as a natural mixture (C16:0 10 %, C18:0 4 %, C18:1 21 %, C18:2 56 %, C18:3 8 % and others, variable content). Such mixtures often serve as raw materials for paints because they have little tendency to crystallise. Naphthenates (e.g. calcium salts of naphthenic acids; Figure 1.31) are an industrial mixture of substances obtained from mineral oil and are added to solvent-borne paints as

Figure 1.30: Lecithin as an amphoteric surfactant

28

Typical wetting and dispersing agents wetting agents and driers (see Chapter 5.1). Additives often have dual functions – users should be aware of this. A nonionic surfactant based on butynediol (Figure 1.32) is recommended as low-foam­ ing wetting agent for water-borne paints. Moreover, it is claimed that it improves levelling (dual function). The two hydrophobic groups have relatively short chains, with the hydro­ philic group (butynediol) having a strong affinity for pigment surfaces.

Reactive surfactants

The concept of reactive surfactants will be discussed using two-component, water-borne epoxy paints as an example [3]. Epoxy-reactive surfactants (ERS) enable aqueous emulsions to be made from liquid epoxy resins. ERS react with amine hardeners during curing to become a part of the crosslinked net­ work. As a result, they lose their emulsify­ ing and hydrophilic properties. The synthesis of epoxy-reactive sur­ factants starts with polyoxyalkylene mono­ amines (“Jeffamine”; Figure 1.33) [10]. These polyoxyalkylene monoamines are made to react with an excess of liquid bisphenol A-epoxy resin (epoxy equivalent mass 188, epoxy value 0.53) [10]. It may be concluded from the data for the epoxy resin that it consists predominantly of bisphenol A-di­ glycidyl ether (epoxy equivalent mass 170, epoxy value 0.59; Figure 1.34). Put simply, this is a reaction between 2 Figure 1.31: Naphthenates moles bisphenol A-diglycidyl ether and 1 mole (mono)amine-terminated polyether to yield an amphiphilic molecule, the epoxy-reactive surfactant (Figure 1.35). This is a nonionic emulsifier and its two epoxy groups can react with amine hard­ eners. These epoxy-reactive surfactants are synthesised “in situ”, i.e. in a large excess of epoxy resin. The reaction between 100 parts by weight of liquid epoxy resin (epoxy Figure 1.32: Nonionic surfactant based on equivalent mass 188) and 10 parts by butynediol

29

Wetting and dispersing agents weight “Jeffamine” M-1000 or M-2070 (Figure 1.33) at 125 °C (under nitrogen) yields after one hour a conversion rate of more than 95 % for the primary to the tertiary amine (Figure 1.35). The outcome is self-emulsifying epoxy resins with good film-forming properties [10]. The term emulsion needs to be explained here (especially in contrast to dispersions). Emulsions are two-phase systems with a liquid disperse phase (e.g. binder) and mostly water as the dispersion medium. By physical drying only, binder emulsions form perma­ nently tacky films (like the coatings on flycatchers). Solid coatings can be formed by binder emulsions only via curing (chemical reactions); ex­ amples are the emulsions of liquid low-molecular epoxy resins, which cure (crosslink) after addition of polyamines (two-components sys­ tems [3]). Dispersions, however, are two-phase systems with a solid dis­ perse phase (e.g. binder) and mostly water as the dispersion medium (see above). By physical drying only, dis­ persions form solid films (e.g. latex gloss enamels and latex paints [3]) – in contrast to emulsions. Figure 1.33: Polyoxyalkylene monoamines

Figure 1.34: Bisphenol A-diglycidyl ether

30

Figure 1.35: Simplified diagram of epoxy-reactive surfactants

Surface treatment of pigments and fillers

1.4 Surface treatment of pigments and fillers Frequently, pigments and fillers are after-treated with specific substances to improve cer­ tain properties (e.g. dispersing or flow properties). As these substances often have amphi­ philic properties, it makes sense to discuss this after-treatment in connection with wetting and dispersing agents.

1.4.1 Organic coloured pigments As a rule, manufacturers of organic coloured pigments do not provide any information about the after-treatment (finish, surface treatment) of their pigments, because these are closely guarded secrets. One excellent reference book on organic coloured pigments [11] contains some information on after-treatment while another more recent book [12] is a bit more generous in this regard. The following references are exceptions to this [13, 14, 27]. Mostly, the final reaction step in pigment manufacturing is to wash the aqueous pigment suspension with water to free it of electrolytes. Were this suspension to be filtered and dried, the result in many cases would be non-redispersible pigment agglomerates. As a rule, organ­ ic pigments are more disperse ( larger specific surface area) than the common inorganic pigments. Therefore, organic pigments form more cohesive agglomerates during manufac­ ture and drying; capillary forces may even lead to aggregates. Consequently, substances are added to pigment suspensions that prevent the formation of “hard” agglomerates during drying and lead to readily dis­ persible pigments (possible grinding processes during the last steps of pigment manufac­ ture are neglected here for the sake of simplicity). The cohesion of the powder particles (agglom­ erates) is reduced by these addi­ tives and so readily dispersible pigment particles are obtained. In some cases, certain addi­ tives are incorporated while the pigments are crystallising to in­ fluence their crystal size and shape; these additives mostly re­ Figure 1.36: Adsorption of calcium salts of abietic acid main on the pigment surface. (diagram simplified and not to scale)

31

Wetting and dispersing agents The additive demand of a pigment is 1 to 2 mg/m2. When 2 parts by weight of a fine organic pigment (specific sur­ face area of about 100 m2/g) are dis­ persed in a paint, 0.2 to 0.4 parts by weight of an (usually unknown) addi­ tive are entrained into the paint for­ mulation at the same time that may trigger unwanted side reactions.

Typical surface treatments

Figure 1.37: Salt of dodecylbenzenesulphonic acid and a long-chain alkyl amine on a pigment surface (diagram simplified and not to scale)

– Addition of resins (e.g. rosin) Resins may be added during or after pigment crystallisation. Performing surface treatment in conjunction with crystallisation has two advantages: + Growth of the primary particles is blocked by the resin. + After drying, the pigment ag­ glomerates are readily dispersi­ ble because the surface is cove­ red by the resin.

Figure 1.38: Four different pigment-specific additives for copper phthalocyanine (CuPc); one to four side chains are possible per pigment-specific additive

32

Surface treatment of pigments and fillers Problem: Since not all the resin is strongly adsorbed, the resin may partially dissolve in the paint medium, altering its properties. Besides rosin resins, amphiphilic calcium, barium and cyclohexyl ammonium salts of abi­ etic acid are used (Figure 1.36). For this, a readily soluble alkali salt of abietic acid is dis­ solved in the pigment suspension and then precipitated on the pigment surface, e.g. by adding a calcium salt ( calcium soap). – Addition of surfactants Figure 1.37 shows a sparingly soluble salt of a sulphonic acid-based surfactant and a longchain alkyl amine (bonding of the dodecyl group to the benzene ring at C2 to C6: technical mixture). For steric stabilisation, the molecular size (chain length) of the surfactant is not suffi­ cient. However, the surface treatment modifies the polarity and wettability of the pigment surface and reduces cohesion between the agglomerates. – Adsorption of substances chemically related to the pigment For example, consider so-called pigment-specific additives (pigment derivatives) for copper phthalocyanine (CuPc) (Figure 1.38) [13]. These pigment-specific ad­ ditives optimize stability to floc­ culation, especially in sol­ vent-borne paints. If used in water-borne paints, the pig­ ment-specific additive (CuPcSO3-) may possibly go into solu­ tion and act undesirably as a dye. Because their structure is similar to that of the pigment, these pigment-specific additives probably adsorb on the mole­ cule “stacks” (Figure 1.39). Again (Figure 1.39), the mo­ lecular size (chain length) of the additives is not sufficient to effect steric stabilisation. However, the polarity and wettability of the cop­ per phthalocyanine surface are changed, and cohesion between Figure 1.39: Adsorption of copper phthalocyanine-specithe agglomerates is reduced. fic additives (diagram simplified and not to scale)

33

Wetting and dispersing agents

Figure 1.40: Adsorption of sulphonated quinacridone (diagram simplified and not to scale)

Sulphonated pigment-specific ad­ ditives are by no means limited to copper phthalocyanines. Quinacri­ done pigments, too, have been treated with sulphonated quinacri­ done derivatives, either as the sul­ phonic acid (Figure 1.40) or as met­ al sulphonate salts, with various metals being possible [27]. As with copper phthalocyanine (Figure 1.39), the planar sulphonated quin­ acridone molecule appears to lie flat on the pigment surface (Figure 1.40) [27].

1.4.2 Titanium dioxide Titanium dioxide is by far the most important pigment. In this Chapter, we will only discuss its more impor­ tant rutile modification.

1.4.2.1 Inorganic after-treatment Figure 1.41: Organic substances on commercial titanium dioxide pigments

Figure 1.42: Example for a polymethylhydrosiloxane (R : C12H25 to C14H29) [31]

34

The two most important substances for inorganic after-treatment of tita­ nium dioxide pigments are silica (hydrated silica: SiO2 · x H2O) and aluminium oxide (hydrated alumina: Al2O3 · x H2O). Both are precipitated onto the pigment surface from aqueous solution [15]. Non-after-treated titanium diox­ ide pigments contain about 0.25 % silica by way of manufacturing im­ purity. The after-treated rutile types

Surface treatment of pigments and fillers on the market contain up to 11 % silica. The following groups can be distinguished with respect to silica content: – 0.5 to 4 % silica: Pigments offering improved gloss retention and resistance to chalking, greying and yellowing – 6 to 11 % silica: Pigments offering particularly improved weatherability and special grades having exceptional light scattering (for latex paints) The morphology of the applied silica layers can be differentiated as follows: – totally uniform, complete encapsulation: For reducing photoactivity (improved weatherability), see Chapter 8.2.1.1 – loose, spongy precipitation: For special grades having exceptional light scattering (latex paints) As a rule, alumina (1 to 5 %) is precipitated after the silica and has a loose, spongy structure (improved weatherability; see Chapter 8.2.1.1). Furthermore, zinc oxide (ZnO) or – preferably – zirconium oxide (ZrO2) may be integrated into the surface layers of the pigment particle. A striking feature is that after-treated titanium dioxide pigments have a higher specific surface area (7 to 20 m2/g) than their non-after-treated counterparts (4 to 9 m2/g). The rea­ son lies in the loose, spongy structure of the precipitated metal oxides. One consequence of the inorganic surface treatment is a change in the isoelectric point (IEP; see Chapter 1.2.1, Table 1.1 [5]) of the titanium dioxide pigment surface. Only sili­ ca-treated pigments have isoelectric points in the acidic pH range, i.e. they have an acidic pigment surface. In water-borne paints (pH ≈ 8), the silica-treated types have negative surface charges. In contrast, alumina-treated pigments have isoelectric points in the alkaline pH range, i.e. they have an alkaline pigment surface. In water-borne paints (pH ≈ 8) the sili­ ca-treated types may have positive surface charges. These changes in surface charge are easy to measure and have consequences for the stability of pigment dispersions, especially in aqueous media [13, 16].

1.4.2.2 Organic aftertreatment Most commercial titanium diox­ ide pigments nowadays are af­ ter-treated with organic sub­ stances [17, 27]; the addition level is below 1 % (expressed in terms of titanium dioxide). The follow­ ing classes of substance are used (Figure 1.41).

Figure 1.43: Simplified presentation of the reaction of a polymethylhydrosiloxane with the surface-hydroxyl groups of titanium dioxide

35

Wetting and dispersing agents To disperse hydrophilic titanium dioxide in hydrophobic polymeric materials the surface of titanium dioxide has to be hydrophobized, for example by silicones. For silicone-post­ treatment of titanium dioxide [31] polymethylhydrosiloxanes [32] which have Si-H-bonds can be used; an example is presented in Figure 1.42 [31]. The chemical reaction of polymethylhydrosiloxanes with the surface-hydroxyl groups of titani­ um dioxide (Figure 1.43) can be explained easily by the concept of electronegativity according to Pauling: oxygen (3.4) is more electronegative than hydrogen (2.2) whereas hydrogen is more electronegative compared to silicium (1.9); so the partial charges as presented in Figure 1.43 are formed (Si-H δ+ and O-H δ-); thus, the hydrogen abstraction is evident. In paints, the titanium dioxide pigments used are after-treated with polyalcohols, derivatives thereof (commonly ethers) or with alkanolamines (mostly triethanol amine); see Figures 1.41 and 1.44. This after-treatment improves dispersibility and gloss. The polyalcohols are most­ ly pentaerythritol and neopentyl glycol (Figure 1.44).

1.4.3 Fillers

Figure 1.44: Polyalcohols and derivatives for after-treatment of titanium dioxide

36

Inorganic fillers (extenders) such as talcum can be surface-treated with silane adhesion promoters (Chapter 6). Judicious selection of the silane for the respective binder can im­ prove the bond between the (cured) binder matrix and the filler. This re­ tards water vapour diffusion through the coating and improves corrosion protection [18]. After-treating talcum with ami­ no-functional silanes improves the corrosion protection of a two-com­ ponent, water-borne epoxy coating to a much greater extent than is pos­ sible with a vinyl-functional silane. The amino groups of the silane (on the talcum surface) can react with the epoxy groups of the binder (Fig­ ure 1.45) and effect a durable bond between filler and binder. In con­ trast, vinyl groups cannot react with epoxy groups.

Surface treatment of pigments and fillers A further possibility is to after-treat calcium carbonate fillers with fatty acids to form cal­ cium soaps on the filler surface [19]. Generally, after-treatment of fillers does not play an important role in coatings technology, as any after-treatment increases the costs of the filler (whose purpose is mainly to reduce costs).

1.4.4 Fumed or pyrogenic silica Highly disperse (i.e. nanoparticle) fumed silica serves as a rheological additive in paints (Chapter 4.1.2). Fumed silica is made by hydrolysing silicon tetrachloride in a hydrogen flame [20]:  2 H2O 2 H2 + O2  SiO2 + 4 HCl SiCl4 + 2 H2O SiCl4 + 2 H2 + O2  SiO2 + 4 HCl During this combustion, the first thing to form is molten spheres of silica (SiO2). The diam­ eter of these nanoparticle primary particles can be controlled and is only 7 to 15 nm, a fact which leads to the high specific surface area of 130 to 400 m2/g in the end product. In the hydrogen flame, the primary particles form as liquid spheres which sinter irreversibly and cool to yield chain-like, branched aggregates. Below the melting point of silica (about 1710 °C), reversible agglomeration of the aggregates occurs. Fumed silica with a specific surface area of 200 m2/g has about 1,000 silanol groups (Si-OH) per primary particle and is therefore hydrophilic. Treatment of the surface silanol groups with appropriate reagents

Figure 1.45: Reaction of chemisorbed amino silane with an epoxy resin (simplified diagram, not to scale)

37

Wetting and dispersing agents affords surface-silylated silicas which are hydrophobic and have modified application properties (Figures 1.46 to 1.48).

Application properties of hydrophobic, fumed silicas

Hydrophobic, fumed silica (about 2 parts by weight, expressed in terms of solid binder) modifies the flow properties of paints, e.g. tendency to sag on vertical surfaces. Hydropho­ bic, fumed silica is wetted not by water, but by water-borne paint resins; it therefore makes a suitable additive for influencing rheology, especially of water-borne paints. Furthermore, it is an excellent antisettling agent in solvent-borne and water-borne zinc-rich primers [21]. Special attention is drawn to the defoaming action of hydrophobic, fumed silica, especial­ ly in water-borne paints (see Chapter 3.2.1).

After-treated fumed silica as reinforcing filler

After-treated fumed silica nowadays serves not only as rheological additive but also as reinforcing filler in scratchproof coatings. Silica nanoparticles are surface-treated with

Figure 1.46: Hydrophobisation of silica with hexamethyldisilazane (simplified diagram, not to scale; formula of the reaction product unequivocal)

38

Figure 1.47: Hydrophobisation of silica with dimethyldichlorosilane (simplified diagram, not to scale; formula of the reaction product not quite unequivocal)

Surface treatment of pigments and fillers functional silanes (compare silane adhesion promoters, Chapter 6) in such a way that they are incorporated into the binder matrix during crosslinking. There, they act as reinforcing filler, improving scratch resistance of coatings (e.g. in mechanical car-washes) [20]. On ac­ count of their small particle size and low refractive index, they do not scatter light and any applied clearcoat remains transparent [refractive index of quartz (SiO2): 1.54; refractive indices of solidified paint resins: 1.5 to 1.7]. For example, two-component polyurethane clearcoats for OEM automotive coatings can be modified by adding 7 % of a special­ ly after-treated fumed silica. In the meantime, special surface-treat­ ed inorganic nanoparticles are used in UV-curing coatings to improve scratch re­ sistance. “Nano” additives are already avail­ able on the market. Surface-treatment with the silane in Figure 1.49 could be useful for incorporating inorganic nanoparticles into UV-curing coatings.

1.4.5 Lamellar metal pigments 1.4.5.1 For solvent-borne paints Metal pigments are used for all kinds of ap­ plications in paints and coatings (including printing inks; Table 1.4). Other metals (e.g. stainless steel) are used to make metal pigments but are of secondary importance. It should be noted that the term “alu­ minium bronze” which is sometimes used for lamellar aluminium pigments is not only outdated but also objectively wrong. Aluminium pigments have a silver colour and consist of highly purified aluminium. Aluminium bronze, by contrast, has a gold­ en colour and is a copper alloy containing about 10 % aluminium [28].

Figure 1.48: Hydrophobisation of silica with trimethoxyoctylsilane (simplified diagram, not to scale; formula of the reaction product not quite unequivocal)

Figure 1.49: Silane for surface-treating silica for use in UV-curing coatings

39

Wetting and dispersing agents Table 1.4:  Applications of different types of metal pigments Applications Effect

Function

metallic coatings

corrosion protection

printing inks

reflecting coatings *

Copper

printing inks

shielding coatings **

Brass (10 to 30 % zinc)

printing inks

Metal pigments 1. Flakes (lamellar) Aluminium

Zinc

corrosion protection

2. Spherical Zinc dust

corrosion protection

* Light and heat ** Electromagnetic radiation

Figure 1.50: Scanning electron micrograph of “cornflake” aluminium pigment  source: Eckart-Werke

Figure 1.51: Scanning electron micrograph of “silver dollar” aluminium pigment

Figure 1.52: Lubricants for lamellar metal pigments (examples)

40

source: Eckart-Werke

Surface treatment of pigments and fillers A simplified treatment of the most important manufacturing process for lamellar metal pigments (aluminium, zinc, copper, brass) now follows; readers interested in further infor­ mation will find it in the literature [23]. Atomisation of (highly purified) molten metals yields more or less spherical metal powders. These are then suspended in white spirit and ground to flakes in a ball mill. The type of grinding employed determines whether the flakes are the “cornflake” type (Figure 1.50) or the superior “silver dollar” type (Figure 1.51). During the forming process, a lubricant must be added to avoid cold welding of the metal flakes in the ball mill. The choice of lubricant determines the distribution of the metal flakes in an organic coating. If the pigments are lubricated with stearic acid, which is amphiphilic (Figure 1.52), the flakes will float more or less on the surface of an organic liquid (leafing effect; Figure 1.53). In contrast, pigments lubricated with oleic acid (Figure 1.52) are distributed uni­ formly in an organic liquid (non-leafing ef­ fect; Figure 1.53). In the past, fatty amines also served as lubricants. Leafing pigments often contain more lubricant (up to 7 %) than non-leafing types (1 to 2 %). Furthermore, leafing pigments Figure 1.53: Leafing and non-leafing effects

Figure 1.54: Optical micrograph (x500) of a cross-section through an automotive coating (silver metallic)

41

Wetting and dispersing agents act as a barrier layer against radiation and moisture (applications in, e.g., corrosion protec­ tion, printing inks, reflective coatings). However, coatings with leafing pigments cannot be overpainted (due to poor adhesion). Leafing pigments are commonly used in corrosion protection coatings (zinc, aluminium), reflective coatings (aluminium) and printing inks (aluminium, copper, brass). Metallic coatings are made with non-leafing aluminium pigments only, because a large number of shades can be obtained through combination with transparent coloured pig­ ments. Figure 1.54 clearly shows how uniformly the aluminium pigments (non-leafing) are distributed in the basecoat. The parallel orientation of the aluminium pigments with re­ spect to the substrate (Figure 1.54) is necessary for achieving the metallic effect. During the grinding process, the fatty acids adsorb in various ways on the metal (oxide) surfaces (Figure 1.55). In the case of zinc, which is more reactive, direct salt formation (zinc soaps) has been detected. For the other metals, partial chemisorption should at least be considered; the exact mechanism still remains to be clarified. The conformation or configura­ tion of both fatty acids dictates the leafing or non-leafing properties of the metal pigments. This “fur model” is now being disputed in some quarters. It postulates that oleic acid forms only a monolayer on the pigment surface on account of the angularity of the cis-double bond (Figure 1.56): The re­ sultant non-polar (oleophilic) pigment surface is completely wetted by the paint resin solution, giving rise to the non-leafing effect. In contrast, straightchain stearic acid forms a double layer on the pigment surface (Figure 1.56): the resultant polar pigment surface (-COOH) is incompletely wetted by the paint resin solution, giving rise to the leafing effect. During grinding (local applica­ tion of high pressure and tempera­ ture) in the presence of oxygen, a number of mostly unclarified chem­ ical reactions occur between the lu­ Figure 1.55: Three ways in which fatty acids can adsorb on metal pigment surfaces bricants (e.g. oxidation).

42

Surface treatment of pigments and fillers Conversion of the metal particles to flakes is followed by screening to modify the particle size distribution. Commonly, the white spirit present in metal pigment suspensions is re­ moved in filter presses; often, metal pigments serve as non-dusting pigment pastes (e.g. for solvent-borne paints: 65 wt.% in white spirit). However, dusting pigment powders are used, too, especially for printing inks. The crucial point here is that the requisite (amphiphilic) fatty acid also acts as a mould-re­ lease agent and can impair adhesion of the binder matrix on the metal pigment surfaces. This can impair the basecoat bond (cohesion) and hence the stone-chip resistance.

1.4.5.2 For water-borne paints

Replacing organic solvents with wa­ ter for the purposes of environmen­ tal protection (by reducing emis­ sions of organic solvents) is both important and necessary. Owing to the chemical reactivity of water (e.g. with regard to corrosion reactions [3]), there are problems with wa­ ter-borne paints and printing inks containing metal pigments that need to be resolved before organic solvents can be replaced by water. The automotive industry has been using water-borne metallic paints (pH about 8) for several years now. The dispersed alumini­ um pigments undergo corrosion in this aqueous, slightly alkaline paint, according to the following equation [24]:

Figure 1.56: Adsorption of stearic respectively olecic acid on metal pigment surfaces (simplified diagram not to scale)

2 Al + 6 H2O  2 Al(OH)3 + 3 H2 Corrosion is an interface reaction and is accelerated by the high specif­ ic surface area of aluminium pig­ ments (about 5 m2/g). A conversion rate of just 2 ‰ aluminium pigment is sufficient to effect greying of the

Figure 1.57: Salt of a phosphoric acid partial ester (stabilising agent for metal pigments in aqueous milieu)

43

Wetting and dispersing agents metallic colour. Hydrogen, the corrosion product, is unwanted for safety reasons (build-up of overpressure and oxyhydrogen). There are several ways to effect the necessary stabilisa­ tion (additional surface treatment) of the pigments. The oldest method is organo-phosphorus stabilisation (Figure 1.57), which is still used today for simple applications. However, it is unsuitable for (water-borne) two-coat metallic coatings, as the stabilising agent (Figure 1.57), which is a fairly hydrophilic surfactant, lowers the moisture resistance of the coatings. The non-chemisorbed part of the stabilising agent (Figure 1.57) can dissolve in the alkaline paint, a fact which leads to increased water adsorption by the coating after contact with moisture. Furthermore, this stabilising agent can act as adhesion-inhibiting layer on the pigment surface ( poor bond). Another way to stabilise aluminium pigments in water-borne paints involves chromate conversion coating [25]. It not only prevents hydrogen corrosion but also improves adhesion of the binder on the pigment surface. Thus, cohesion of the coating is improved, especial­ ly when exposed to moisture. Other, modern passivation methods include sol-gel silicate coatings and polymeric coating of pigments [23]. Finally, it should be noted that water-soluble paint resins bearing appropriate function­ al groups adsorb on the surfaces of aluminium pigments; they can thus partly and, in in­ dividual cases, completely prevent hydrogen corrosion [26, 29, 30]; some more details follow in Chapter 7.4.

1.5 Literature [1] J. Bieleman et al., Additives for Coatings, Wiley-VCH, Weinheim, 2000 [2] H. Kittel †, M. Ortelt et al., Lehrbuch der Lacke und Beschichtungen – Vol. 4: Lösemittel, Weichmacher und Additive, S. Hirzel Verlag, Stuttgart, 2nd ed. 2007 [3] B. Müller, U. Poth †, Coatings Formula­ tion, Vincentz Network, Hanover, 3rd ed 2017 [4] H. Kittel †, H. W. Ritter, W. Zöllner et al., Lehrbuch der Lacke und Beschichtungen – Vol. 8: Herstellung von Lacken und Beschichtugsstoffen, Arbeitsicherheit, Umweltschutz, S. Hirzel Verlag, Stuttgart, 2nd ed. 2004 [5] K. Köhler, C. W. Schläpfer, Chemie in unserer Zeit, 27 (1993) No. 5, p. 248– 255 [6] M. Knospe, W. Scholz, Farbe & Lack 96 (1990) p. 120 ff

44

[7] R. Jérôme, Farbe & Lack 98 (1992) p. 325–329 [8] A. Bouvy, Europ. Coat. Journ. No. 11 (1996) p. 822–826 [9] J. M. Reck, L. Dulog, Farbe & Lack 99 (1993) p. 95 ff [10] Brochures of Huntsman “The Jeffamine® Polyoxyalkyleneamines” and “WaterReducible Epoxy Coatings via Epoxy Resin Modification with Jeffamine® MATP’s – In situ Epoxy Reactive Surfactant” [11] W. Herbst, K. Hunger, Industrial Organic Pigments, Wiley-VCH, Weinheim, 3rd ed. 2004 [12] E. B. Faulkner, R. J. Schwartz (eds.) High Perfomance Pigments, Wiley-VCH, Weinheim, 2009 [13] J. Schröder, Farbe & Lack 93 (1987) p. 715–720

Literature [14] B. G. Hays, American Ink Maker, June 1984, p. 28–50 [15] H. Weber, Kieselsäure als Bestandteil der Titandioxid-Pigmente, FATIPEC-Kongress­ buch 1978 [16] M. Osterhold, K. Schimmelpfennig, Farbe & Lack 98 (1992) p. 841–844 [17] H.-H. Luginsland, Organische Behandlung von Titandioxid-Pigmenten, FATIPECKongressbuch 1986 [18] N. Wamser, E. Urbino, Farbe & Lack 95 (1989) p. 109 ff [19] D. Gysau, Fillers for Paints, Vincentz Network, Hanover 2006 [20] H. Kittel †, J. Spille et al., Lehrbuch der Lacke und Beschichtungen – Vol. 5: Pigmente, Füllstoffe, Farbmetrik, S. Hirzel Verlag, Stuttgart, 2nd ed. 2003, p. 403 ff [21] B. Müller, P. Kienitz, Farbe & Lack 102 (1996) No. 8, p. 76–80 [22] Die fleißigen Verbindungen – eine kurze Einführung in die Welt der Tenside, Verband TEGEWA e.V., Frankfurz a. Main, 2014

[23] P. Wißling, Metallic Effect Pigments, Vincentz Network, Hanover 2006 [24] B. Müller, M. Gampper, Werkst. Korros. 45 (1994) p. 272–277 [25] R. Treutlein, B. Müller, P. Mayenfels (BASF Lacke und Farben AG), DE OS 3636183 A1 (disclosed 3rd March 1988); Chem. Abstr. 109 (1988) 8106d [26] B. Müller, Europ. Coat. Journ., No. 5 (2001) p. 81 ff [27] D. Satas, A. A. Tracton (eds.), Coatings Technology Handbook, Marcel Dekker, New York, Basel, 2001, p. 620 ff [28] http://en.wikipedia.org/wiki/Aluminium_ bronze [29] B. Müller, Surface Coat. Int., Vol 85, B2 (2002) p. 111–114 [30] B. Müller, S. Fischer, Corros. Sci. 48 (2006) p. 2406–2416 [31] US-Pat. 4,810,305 [32] https://en.wikipedia.org/wiki/ Polymethylhydrosiloxane

45

Surface defects

2 Substrate-wetting additives/ levelling agents Figure 2.1 tries to classify qualitatively the most important additives active on the paint surface; moreover, intersecting sets between the certain types are shown. For clarification it should be noted that slip additives decrease the sliding resistance of coatings whereas rub respectively antislip additives increase the sliding resistance of coatings. In this chapter levelling agents, substrate wetting agents and, briefly, hammer finish as well as slip additives are presented. Matting agents and rub respectively antislip additives cannot be discussed in detail.

2.1 Surface defects Wetting defects

Adequate wetting of a solid substrate by a liquid phase (in our case, liquid paint) is contingent on the surface tension of the liquid paint’s being lower than the surface tension of the solid substrate (for details, see [1]).

Figure 2.1: Attempt of a roughly qualitative classification of additives acting on the paint surface Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

47

Substrate-wetting additives/levelling agents Wetting defects occur when the surface of the substrate contains contaminants whose surface tension is lower than that of the paint. Common contaminants are oils and greases, and in certain cases mould-release agents, abrasive dust and spray mist. Contaminants which occur in very small, locally restricted areas may give rise to cratering (Figure 2.2). Such contaminants are primarily located in the centre of the crater and sometimes are as easy to see as in Figure 2.2. For the most part, though, contaminants are so tiny that they can only be detected with highly sophisticated analytical equipment. Particularly feared are contaminants composed of high-molecular, incompatible silicones or perfluorinated hydrocarbons, because they have extremely low surface tension. Often, craters occur not alone but rather in groups (Figure 2.3). Furthermore, areas of the substrates which have been ground or wiped off (leading to a change in surface tension) can leave marks in the applied paint (“ghosting”).

Figure 2.2: Optical micrograph of a crater in a coating

Figure 2.3: Multitude of craters in a limited coating area

Figure 2.4: Wetting defects of a co-solvent-free, oxidative-cure, water-borne primer on steel

Figure 2.5: Flow due to solvent evaporation in a freshly applied paint film

48

Surface defects Figure 2.4 shows wetting defects of a co-solvent-free water-borne primer on steel; in this case the wetting defect has a rare dendritic shape. The defects here were completely eliminated by adding 2 % butyl glycol (co-solvent).

Bénard cells

Pigment separation and Bénard cells are discussed in Chapter 1.3.1 in connection with controlled flocculation (see Figures 1.20 and 1.21 earlier). Solvent evaporating from a fresh coat of paint “draw” low-viscosity paint material from the deeper layers to the surface (Figure 2.5). There, the low-viscosity paint spreads out, the solvents evaporate, and because the viscosity and density increase, the paint sinks to the deeper layers. Solvent evaporation and flow in the paint film cause local (minor) differences in surface tension, temperature and density of the paint film; flow and eddies cause pigments of different mobility to separate (they have different sizes and/or densities). On horizontal surfaces, so-called Bénard cells are formed (see Figures 1.20 and 1.21 in Chapter 1) while, on vertical surfaces, streaks are formed (this is called “silking”). As drying progresses and – therefore – the viscosity of the paint film rises, flow comes to an end and the resultant pigment separation becomes permanent. There are various other types of damage, e.g. “telegraphing” (sensitivity to drafts) that can all be explained by local changes in surface tension. Most of the aforementioned problems can be resolved with the help of silicone additives that accumulate at the paint surface where they create a surface layer of uniform viscosity and surface tension. Figure 2.6 illustrates how silicone additives affect pigment separation.

Figure 2.6: (a) Defects in a coating containing pigment red 88 and titanium dioxide, before and after addition of a silicone additive. (b) Magnified view of the separated area

49

Substrate-wetting additives/levelling agents

2.2 Silicone additives The best way to avoid wetting defects is to eliminate the contaminants causing them. As this is not always possible in every case (e.g. areas that have been ground), the surface tension of the paint can be lowered by adding surface-active agents; these improve wetting of the substrate and often also lead to good levelling. Silicone additives are particularly adept at this; however, they can cause unwanted side reactions, such as adhesion failure when overpainted or turbidity in the paint film. As silicone additives are generally more or less incompatible with paint resin solutions, they accumulate on the paint surface (at the interface of the paint film and air), thereby lowering the surface tension of the paint film; that is why they can be called surface-active [2, 3]. As a result of this surface activity, they can also improve substrate wetting (Figure 2.6).

2.2.1 Polydimethyl siloxanes The first silicone additives used were polydimethyl siloxanes (“silicone oils”) whose effectiveness depends on their molar masses (Figure 2.7): the higher the molar mass, the greater the incompatibility. The low-molecular polydimethyl siloxanes (M < 5000) act as levelling agents and are discussed in this section. The slightly higher-molecular types (M ≈ 6000) lower the sli-

Figure 2.7: Applications of polydimethyl siloxanes (“silicone oils”) as a function of molar mass

50

Silicone additives ding resistance of coatings and are called slip additives; they are not discussed in detail here. Much higher-molecular types (M ≈ 90,000) have a defoaming effect; more information on them is presented in Chapter 3. Those having the highest molar mass (M > 100,000) are the most incompatible of all and, in combination with metal effect pigments, create what is known as a “hammer finish” (Figure 2.8).

2.2.2

Chemically modified silicone additives

Chemical modification of polydimethyl siloxanes yields more readily compatible “tailor made” silicone additives. The first class of modified silicone additive is that of methylalkyl polysiloxanes (Figure 2.9). The relative reduction in surface tension decreases with increase in chain length of the alkyl groups (Figure 2.9), because the molecules have longer alkyl groups (increasingly resembling those of paint resins) and so become more compatible. This affords a means of selectively modifying the surface tension and compatibility. Modification is also possible with polyethers (Figure 2.10). Polyether-polysiloxanes are thermostable up to 150 °C; at 170/180 °C and

Figure 2.8: Photograph of a “hammer finish” coating

Figure 2.9: Methylalkyl polysiloxanes

51

Substrate-wetting additives/levelling agents above, the polyethers thermally decompose and give rise to adhesion problems. However, polyester-modified polysiloxanes and aralkyl-alkyl polysiloxanes (Figure 2.11) are heat resistant up to 230 °C. A reactive hydroxy-functional polyester-polysiloxane (R’’’ = H in Figure 2.11) can be incorporated chemically into a binder matrix (e.g. polyurethane) and cannot be removed by wiping. All modification groups are bonded via non-hydrolysable Si-C bonds. In the past, modified polysiloxanes had hydrolysable Si-O-C bonds, which could have led to a deterioration in coating properties (e.g. cratering), especially after protracted storage (Figure 2.12). Splitting off the organic modificaFigure 2.10: Polyether-modified polysiloxanes tion R by hydrolysis (Figure 2.12) increases incompatibility and promotes formation of reactive silanol groups (Si-OH). Subsequent condensation yields even more incompatible polysiloxanes of higher molar mass and may cause cratering. Hydrolysis can also happen in solvent-borne paints because technical organic solvents often contain about 0.5 % water by way of impurity. As modified polysiloxanes are generally incompatible with paint resin solutions, they accumulate on the paint surface and, in thereby lowering the surface tension, are surface-active; the modification orients itself towards the paint resin solution while the silicone points toFigure 2.11: Polyester-modified polysiloxanes and aralkyl-alkyl polysiloxanes wards the air (Figure 2.13).

52

Silicone additives Table 2.1:  How a silicone additive typically works Surface tension

γ [mN/m]

A: substrate (non-pretreated steel)

29

B: liquid paint (with solvent)

28

C: solid coating (without solvent)

38

D: solid coating (without solvent) + silicone additive

26

Wetting of the substrate

Δ γ [mN/m]

B-A: good

-1

C-A: poor

+9

D-A: good

-3

The surface tension values in Table 2.1 illustrate the influence of the solvents on the substrate wetting properties of paints. Evaporation of solvents causes the surface tension of the paint to rise and wetting defects to occur. A suitable silicone additive ensures that the surface tension of the paint is so low after solvent evaporation that the substrate is wetted (Table 2.1). Silicone-free levelling agents, based on acrylate, exist as well (e.g. poly-n-butyl acrylate). Besides polyether-modified polysiloxanes (see above), relatively low-molecular silicone surfactants (Figure 1.26 in Chapter 1.3.2) can be added to water-borne systems; the silicone surfactants serve only to wet the substrate. The formation of Bénard cells can be prevented: – During dispersing, by adding suitable dispersing agents; e.g.

Figure 2.12: Hydrolysis and condensation of firstgeneration, modified polysiloxanes

53

Substrate-wetting additives/levelling agents

Figure 2.13: Interface orientation of modified polysiloxanes

controlled flocculation impedes separation of pigments of different density (see Figure 1.22 in Chapter 1.3.1). – After dispersing, by adding suitable silicone additives. These form a thin layer of uniform (low) surface tension and uniform viscosity on the paint surface. As a result, no areas of different surface tension can form – which are the cause of Bénard cells.

2.3 Literature [1] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed. 2017 [2] J. Bieleman et al., Additives for Paints, Wiley-VCH, Weinheim, 2000

54

[3] H. Kittel †, M. Ortelt et al., Lehrbuch der Lacke und Beschichtungen – Vol. 4: Lösemittel, Weichmacher und Additive, S. Hirzel Verlag, Stuttgart, 2nd ed. 2007

Foam types and foam stabilisation

3 Defoamers Foam can occur in all liquid paint systems and can cause defects, such as cratering and pinholes (Figure 3.1). Air can be incorporated into a paint during manufacture (e.g. by the agitator) or during application (e.g. spraying). Foam occurs particularly often in water-borne paints. Air can also be incorporated into paint by porous substrates, such as wood and mineral substrates, because liquid paint displaces the air from the pores. Figure 3.2 clearly shows the pores in beech wood.

3.1 Foam types and foam stabilisation Generally, pure liquids will not foam (Figure 3.3). Since paints are multi-substance mixtures, however, foaming can occur during manufacture and application. Surface-active substances, in particular, can be very effective at stabilising foam bubbles (Figure 3.4) [1, 2]. In a pure liquid (containing no surface-active substances), foam bubbles burst when they reach the surface of the liquid (Figure 3.3). If the liquid contains a surface-active substance (e.g. wetting agent), a (monomolecular) film is formed on the inner and outer surface of the liquid. Bubbles which rise to the surface become covered with a stabilising surfactant film – the foam lamella (Figure 3.4). In foams which contain ionic surfactants, the foam bubbles can be stabilised electrostatically (see Chapter 1.2.1). First, the repulsion of electric double layers (Figure 3.5) prevents fusion of smaller foam bubbles to larger ones

Figure 3.1: Depth profile (x1000) of a pinhole in a water-borne stoving enamel Source: J. Domnick, D. Gruseck, University of Applied Sciences Esslingen) a) plan view; b) side view

Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

55

Defoamers and, therefore, rapid bubble ascent. Second, a minimum thickness of foam lamellae is stabilised and so the foam bubbles do not burst. When air bubbles are introduced into a liquid, they form spheres (least surface area combined with the largest volume) because of the interfacial tension between the air and liquid (surface tension). When gas bubbles form within a liquid, the surface area of the liquid is increased; i.e. work W has to be done (see Chapter 1.2). This work is done e.g. by the dispersing units or during application. where

W = γL · ΔA W is the work [energy] γ L is the surface tension of the liquid phase [force/length = energy/surface area] [3] ΔA is the increase in surface [area]

The work W is proportional to the increase in surface area ΔA and to the surface tension of the liquid phase γL. Therefore, smaller foam bubbles have higher energy content than larger ones and so smaller foam bubbles can fuse to larger ones and thereby gain in energy (reduction in surface area). Often small foam bubbles are called microfoam, while large ones are called macrofoam. Surfactants lower the surface tension of the liquid phase, facilitating foam formation (W decreases).

Figure 3.2: Scanning electron micrograph of beech wood (scale bar: 100 μm = 0.1 mm)

56

Figure 3.3: Air bubbles in a pure liquid

Foam types and foam stabilisation Moreover, surface tension γL (force/length) generates an internal pressure p inside the foam bubbles; according to Laplace, the following simplification holds: p ~ γL / r In other words, the smaller the radius r of the bubbles (the larger the surface area) and the higher the surface tension of the liquid phase, the higher is the internal pressure.

Figure 3.4: Foam formation and stabilisation

Figure 3.5: Electrostatic stabilisation of a foam lamella by anionic surfactants (the positive counter ions are neglected for the sake of clarity)

57

Defoamers The speed of ascent v (negative rate of descent) of foam bubbles can be calculated from Stokes’ law: ​

Figure 3.6: Optical micrograph (100x) of microfoam bubbles in a coating of water-borne, black latex paint; scale bar 1 mm

Figure 3.7: Surface defects in a coating that were caused by foam

Figure 3.8: Spherical and polyhedral foam bubbles

58

​r​ 2​ ∙ ​(​ ​ρ​ G​− ​ρ​ L​)​       ​ v = c ∙  ​_   η

where c is a constant r is the radius of foam bubbles ρG s the density of air ρL is the density of the liquid phase (e.g. paint resin solution) η i s the viscosity of the liquid phase (e.g. paint resin solution) As the density of air is always smaller than that of the liquid, v becomes negative (v < 0). The following simplification then applies: v ~ r2 / η From this, it follows that large foam bubbles (macrofoam) rise faster than small bubbles (microfoam). Figure 3.6 shows microfoam bubbles in a coating of water-borne, black latex paint. The greater the viscosity of the liquid phase (paint resin solution), the more slowly the bubbles ascend. After the paint has been applied, the viscosity of the resin solution increases markedly and rapidly due to solvent evaporation. Thus, small foam bubbles (microfoam) are often unable to escape the paint completely and so may cause pinholes in the coating’s surface (see Figure 3.1). Figure 3.7 shows surface defects, including pinholes, in a coating that were caused by foam.

Types and mode of action of defoamers On the surface, the bubbles form a foam crest of hexagonally close-packed spheres. Gravitational draining of liquid from the foam lamellae converts the foam bubbles into polyhedra (Figure 3.8) [4]. For geometrical reasons, foam composed of spherical bubbles has a maximum gas content of 74 vol.% (assuming the foam bubbles are the same size); foam made up of polyhedral bubbles has a far greater gas content (up to 99.x vol.%). An example of polyhedral foam bubbles of very high gas content is presented in Figure 3.9; the thinness of the lamellae is clearly apparent [4]. Draining of liquid from the of polyhedral foam lamellae causes them to become increasingly thinner until they burst at a critical thickness. The critical thickness for foam lamellae is approximately 30 nm for water and 70 nm for hexane. The rate at which the liquid drains depends on its viscosity, the critical thickness does not. Foam lamellae having thicknesses below the wavelength of visible light (400 to 800 nm) exhibit interference colours which are commonplace in everyday life (e.g. rainbow effect). Consequently, the occurrence of these rainbow colours indicates that the thickness of a foam lamella is less than the wavelength of light.

3.2 Types and mode of action of defoamers Some manufacturers make a distinction between defoamers and deaerators: defoamers act at the paint surface and destroy the foam crest (macrofoam) or the polyhedral foam. Deaerators (sometimes also called air-release additives) act inside the paint film, causing foam bubbles (microfoam) rise faster to the surface. Defoamers are often combined with deaerators. It is also possible for one substance to act both as a defoamer and a deaerator; in these cases a distinction is not always useful.

3.2.1 Defoamers for waterborne paints Defoamers are mostly liquids of low surface tension which must meet three requirements:

Figure 3.9: Polyhedral foam bubbles of very high gas content

59

Defoamers 1. They should be virtually insoluble in the paint resin solution. 2. They should have a positive penetration coefficient E, whereby

E = γL - γE + γLE > 0 where γL is the surface tension of the paint resin solution (liquid phase) γE is the surface tension of the liquid defoamer is the interfacial tension between γLE  paint resin solution and defoamer 1. They should have a positive spreading coefficient S, which is calculated as follows: S = γL - γE - γLE > 0  γE + γLE < γL

Figure 3.10: Mode of action of defoamers (liquid defoamer: black): a) foam bubbles and defoamer droplets; b) defoamer droplets penetrate into the foam lamella and displace foamstabilising surfactants; c) foam bubbles burst

60

If the penetration coefficient E is positive, the defoamer can penetrate into the foam lamella. If, additionally, the spreading coefficient S is positive, the defoamer can spread out at the interface of the lamella, displacing the foam-stabilising surfactants and destabilising the lamella because the defoamer has a low surface tension (low cohesion; Figures 3.10 a to c). The effect of such liquid defoamers can often be improved by adding finely dispersed hydrophobic solid particles [e.g. hydrophobic fumed silica (Chapter 1.4.4) or polyureas]. The liquid defoamer transports the hydrophobic solid particles into the foam lamella. The hydrophobic solid particles act in the hydrophilic foam lamella as a foreign body, reducing cohesion and so destabilising

Types and mode of action of defoamers the lamella; furthermore, surfactants may adsorb on their surface and cause further destabilisation.

Silicone defoamers

Table 3.1:  Formulation of a defoamer based on mineral oil Mineral oil

75

Hydrophobic solid particles

10

Emulsifier(s)

5.5

Modified polysiloxanes

7.5

Frequently, modified polydimeOther additives 2 thyl siloxanes are used. ExtremeTotal 100 ly hydrophobic polydimethyl siloxanes (“silicone oils” see Chapter 2.2.1) are highly incompatible and the level of addition must be controlled very carefully (risk of cratering). Modification of the polysiloxane with relatively hydrophilic polyether (Figure 3.11; see also Chapter 2.2.2) can improve compatibility with paint, without changing the desired incompatibility with the foaming medium (aqueous surfactant solution) [5]. The molar mass of the polysiloxane and the type and quantity of polyether (Figure 3.11) determine the properties of the silicone defoamer in the paint (e.g. compatibility, defoaming action). Possible adverse side-effects of silicone defoamers include: – adhesion failure upon overpainting – surface defects (craters)

Mineral oil defoamers The principle behind mineral oil defoamers is often employed in the home. When noodles are boiled, a persistent foam forms in which soluble starch components act as a protective colloid. Adding some drops of edible oil destroys the foam – mineral oil defoamers act in much the same way. Mineral oil defoamers are cheaper than silicone defoamers; an example is presented in Table 3.1. Mineral oil has a lower surface tension than water and acts as a defoaming agent.

Figure 3.11: Polyether-modified polysiloxane as a defoamer for water-borne paints (EO = ethylene oxide; PO = propylene oxide)

61

Defoamers Possible deleterious side-reactions of mineral oil defoamers include: – lowering of gloss (silicone defoamers do not affect gloss) – surface defects – separation of mineral oil

3.2.2 Defoamers for solvent-borne paints In solvent-borne paints, too, the spreading coefficient S should be greater than zero. S = γL - γE - γLE > 0  γE + γLE < γL As the surface tension γL of solvent-borne paints is significantly lower than that of water-borne paints, mineral oils cannot be used; the only recourse here is to use silicone defoamers (extremely low γE). As is the case in water-borne systems, incompatible polydimethyl siloxanes (“silicone oil”: risk of cratering) having molar masses from 10,000 to 100,000 are used. Even higher-molecular polydimethyl siloxanes are so incompatible that they give rise to the “hammer finish” effect (see Figure 2.8 in Chapter 2.2.1).

Figure 3.12: How to find the right level of defoaming

62

Deaeration of coating powders Again, mostly modified polysiloxanes are used to strike a balance between poor compatibility and targeted defoaming action [6]. If the defoamer is too compatible, the foam becomes stabilised. If it is too incompatible, defects may occur (Figure 3.12) in the form of craters, haze or adhesion failure. Therefore, paint formulators must be adept at choosing the right (optimum) defoamer for their specific paint from the large range available on the market. Also available are silicone-free polymeric defoamers, mostly based on acrylate.

3.3 Deaeration of coating powders Coating powders are solid, solvent-free coatings for industrial applications. They are applied as powders by different methods and are then stoved. Film formation (solidification) occurs first by melting the applied powder particles on the substrate and then cooling the possibly pigmented polymer melt. A distinction is drawn between thermoplastic coating powders and thermosetting coating powders which cure (crosslink) after melting [3, 7]. In this book, only high-performance thermosetting coating powders are discussed. These involve the reaction of a molten base resin with a molten curing agent (crosslinker) to yield a thermosetting coating. A serious drawback of all thermosetting coating powders is deaeration, because the air present between the powder particles must be released during film formation (Figure 3.13). Release of air (deaeration) is possible only below a certain viscosity. This is closely connected to the problem of levelling by thermosetting powder coatings, which also requires the viscosity of the powder coating film to be below a critical level (ηc) (open time, Figure 3.14). Figure 3.14 shows the how the viscosity of two powder coatings changes during melting and curing. Initially, the viscosity decreases as Figure 3.13: Cross-section through a powder-coated the oven temperature rises; once cu- substrate (the small triangles and squares represent ring starts, the viscosity increases the pigments in coating powder particles)

63

Defoamers markedly and rapidly (Figure 3.14). Coating powder 1 is more reactive (crosslinks faster) than coating powder 2; i.e. the open time during which deaeration and levelling are possible is shorter for coating powder 1 than for coating powder 2. For comparison, Figure 3.14 also shows the change in viscosity of the pure base resin without curing agent; it behaves like a thermoplastic (the viscosity drops to the level of the melt viscosity). In other words, the open time during which deaeration and levelling occur can be controlled via the reactivity of the coating powders. Where the reactivity cannot be changed, viscosity-lowering additives can be used. One example of a deaerator which also improves levelling is benzoin (Figure 3.15). Addition levels of about 0.5 % will lower the melt viscosity of coating powders. Benzoin (melting point: 132 to 137 °C) acts like a solid solvent. By means of as-yet unclarified reactions, benzoin is often incorporated into the polymer network; i.e. it could also be called a solid reactive diluent. To further improve levelling, it is quite common to additionally employ levelling agents based on acrylate.

3.4 Literature

Figure 3.14: Change in viscosity during stoving of thermosetting coating powders

Figure 3.15: Benzoin

64

[1] F. Sebba, Foams and Biliquid Foams – Aphrons, Wiley 1987 [2] http://de.wikipedia.org/wiki/ Schaum (German version is superior to http://en.wikipedia.org/wiki/ Foam) [3] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed 2017 [4]  www.maths.tcd.ie/~foams/gallery.htm [5] K.-H. Käsler et al., Phänomen Farbe, No. 9 (1996) p. 25 f [6] Byk, Technical Information, Defoamers and Air Release Additives [7] P. G. de Lange, Powder Coatings – Chemistry and Technology, Vincentz Network 2004

Rheologically active additives

4 Rheologically active additives This chapter discusses only the most important members of the vast number of rheologically active additives available. Other paint additives, such as dispersing agents, can modify the rheological properties of paints by means of controlled flocculation (Chapter 1.3.1); the border between dispersing and rheological action is sometimes fuzzy. In general, the action of all paint additives (here rheologically active additives or dispersing agents) can vary greatly with the paint formulation (e.g. type of pigments and fillers). Rheologically active additives (rheology modifiers) effect changes in viscosity over a specific range of shear rate, and this leads to non-Newtonian flow. In contrast, thickeners effect an increase in viscosity over the whole range of shear rate by increasing only the viscosity of the liquid phase (liquid phase thickeners). Ideally, they do not (or only slightly) change the flow (rheology); i.e. cause little or no shear thinning (pseudoplasticity). Unfortunately, both types cannot always be clearly distinguished. The oft-used, outdated term “thixotropic agent” should be avoided. Figure 4.1 shows Newtonian flow, in which the viscosity does not depend on the shear rate (viscosity is constant). As an example of non-Newtonian flow, Figure 4.1 shows shear thinning (pseudoplasticity), in which the viscosity decreases with increase in shear rate [1]. Another example of non-Newtonian flow that is important in coatings technology is thixotropy (Figure 4.2), in which, again, the viscosity decreases with increase in shear rate (gel curve) and increases with decrease in shear rate (sol curve), but Figure 4.1: Two examples of viscosity curves (viscosinot to the same extent as in the case ty as a function of shear rate) of the gel curve. If the system is all1: Newtonian liquid (viscosity = constant) owed to stand for some time, it re2: Shear thinning (pseudoplastic) liquid: viscosity decreases with turns to its initial viscosity (complete increase in shear rate; one example of non-Newtonian flow Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

65

Rheologically active additives reversibility). The greater the difference between the sol and gel curves, the greater is the degree of thixotropy. Unlike shear thinning, which is time-independent (Figure 4.1), thixotropy is time-dependent. Rheology modifiers can become active at different shear rates encountered during manufacture or application of the paint (Figure 4.3). At rest, many rheological additives build up a three-dimensional network (gel) in paints (solvent-borne or water-borne). During shearing (e.g. stirring), this network is broken temporarily (sol) but re-forms fairly rapidly at rest (reversible sol-gel transition). Pigments and fillers can also be incorporated into these rheologically active network structures, but this is not necessary. A common problem with paints is the settling of pigments and fillers. The sedimentation rate (settling rate) v of pigments and fillers is governed by Stoke’s law: ​r​ 2​ ∙ ​(​ ​ρ​ S​− ​ρ​ L​)​       ​ ​v = c ∙  ​_   η

Figure 4.2: Thixotropic flow

Figure 4.3: Shear rates during manufacture or application of paints

66

Rheologically active additives where c r ρS ρL η

is a constant is the radius of pigment and filler particles (approximation: spherical) is the density of the pigment and filler particles is the density of the liquid phase (e.g. paint resin solution) is the viscosity of the liquid phase (e.g. paint resin solution)

For a given pigment or filler, the sedimentation rate v is inversely proportional to the viscosity of the paint resin solution; this statement applies to the viscosity at low shear rate (see Figure 4.3: storage). Fur­ thermore, the sedimentation rate v is proportional to the square of the radius of the pigment and filler particles. From Stoke’s law, two conclusions can be drawn as to how settling of pigments and fillers may be prevented:

Figure 4.4: Sagging of a paint on a vertical surface of a car body

Figure 4.5: Attempt of a systematical classification of rheologically active additives

67

Rheologically active additives – Dispersing of agglomerates (r is large), where possible, to primary particles (r is small); also through using appropriate dispersing agents (Chapter 1.3.1). – Use of rheology modifiers which are active only at low shear rates (conferring, e.g. shear thinning on the paint). A basic principle is that the settling of pigments and fillers in a paint, the sagging of an applied paint on a vertical surface (Figure 4.4) and levelling on a coated substrate are conflicting rheological properties (all three occur at low shear rates). Consequently, compromises must be struck during paint formulation. Rheologically active additives can be classified as presented in Figure 4.5. Waxy additives (Figure 4.5) cannot be discussed in this book because it is limited to the more impor­ tant types. Because of the practical application this chapter structured according to the type of paint medium (water-borne and solvent-borne).

4.1 Rheological additives for water-borne and latex paints In a departure from the general systematic approach adopted so far, this chapter will dis­ cuss aqueous systems first, because (unmodified) layer silicates are used in water-borne paints, whereas organically modified “refined” layer silicates (organoclays) are added to solvent-borne paints.

4.1.1 Layer silicates Layer silicates are clay minerals that can be swelled by water because of their layer structure. Smectites are a group of layer silicates which are composed of three elementary

Figure 4.6: Simplified diagram of the elementary layers of smectite

68

Rheological additives for water-borne and latex paints layers (three-layer clay minerals) [2, 3]. The middle layer consists of alumina/magnesium oxide octahedra and is enclosed on both sides by layers of silica tetrahedra (sometimes containing alumina; Figure 4.6) [4]. Other cations may be incorporated, too, depending on the type of mineral and its source. These three-layer platelets constitute the elementary layers (primary platelets) of the smectites; the primary platelets in turn are stacked in layers, one above the other. The negative charges of the elementary layers must be compensated by cations between the elementary layers (Figure 4.7). A scanning electron micrograph is presented in Figure 4.25 as an example of an organoclay. Two industrially important smectites are montmorillonite and hectorite. Montmorillonite contains some magnesium oxide in the octahedra layer [4]: [(Al3.34Mg0.66)Si8O20(OH)4]Na0.66 The negative charges are countered by (exchangeable) sodium ions. Approximate size of the nanoparticle primary platelets: Ø 10 to 100 nm; thickness about 1 nm. Montmorillonite (composition fluctuates) is not usually found pure in Nature, but rather occurs in a mixture with other minerals. Furthermore, besides sodium ions, calcium ions or mixtures of the two are often present. Layer silicates with a high content of montmorillonite are called bentonites. Hectorite is a magnesium layer silicate containing lithium and fluoride ions: [(Mg5.34Li0.66)Si8O20(OH)3F]Na0.66 Approximate size of primary platelets of natural hectorites: length 300 nm, width 50 nm, thickness 1 nm. Two different types of layer silicates are manufactured, namely purified natural products (sometimes chemically modified)

Figure 4.7: Structure of smectites

69

Rheologically active additives and wholly synthetic types. Wholly synthetic layer silicates are nano-particles and form totally clear dispersions in water; the primary platelets measure roughly 40 · 10 · 1 nm and have a specific surface area of about 370 m2/g (BET). All layer silicates cause a certain reduction in gloss; the finer the layer silicate particles, the less pronounced is this reduction.

Dispersing of layer silicates When layer silicates are dispersed in water, the exchangeable cations are eluted from the stacks of primary platelets. As a result, the negatively charged primary platelets repel each other and a fine electrostatically stabilised dispersion (sol) is formed (Figure 4.8) [5].

 el formation of G layer silicates

Figure 4.8: Aqueous dispersion of layer silicates (two-dimensional, simplified diagram)

Figure 4.9: Highly simplified diagram of the primary smectite platelets (elementary layers)

70

The surfaces of the sol’s primary platelets have pronounced negative charges, whereas the edges, with their hydroxyl groups (Figure 4.9), have a relative positive charge (the overall charge of a primary platelet is, of course, negative) [4]. In concentrated sols at rest, electrostatic attraction can cause an interaction between edge and surface that leads to so-called house-of-cards structures (Figure 4.10) which yield thixotropic gels that often have a yield point. This gel breaks down under shear stress (e.g. stirring, spraying) but forms again during relaxation; the sol-gel transition is reversible. Figure 4.10 clearly shows that an optimum gel

Rheological additives for water-borne and latex paints structure requires a layer silicate sol to have been dispersed as far as the primary platelets, where possible. Layer silicate sols can form gels at relatively low concentrations (some types from just 2 %). To manufacture more highly concentrated silicate sols as stable, flowable (pumpable) intermediate products, so-called peptising agents are sometimes added. The peptising agents employed then are often polyanionic dispersing agents, such as tetrasodium pyrophosphate (Na4P2O7), which make dispersions containing up to 10 % possible. Adsorption of the pyrophosphate tetra-anions on the edges of the platelets causes the edges to become negatively charged and repulsed by the surfaces, which are also negatively charged (electrostatic stabilisation). Pigments and fillers added to the peptised silicate sol also adsorb the pyrophosphate via adsorption-desorption equilibria, and gel formation can take place (after a time delay) in the paint [5]. Peptised layer silicates are used in latex paints. Furthermore, more highly concentrated sols of layer silicates can also be produced by adding certain water-soluble oligomers such as polypropylene glycol (molar mass: about 900; solubility in water: about 2 %) [6]. The molar mass of the polypropylene glycol is very important, because lower-molecular types are completely water-soluble (enrichment occurs in the aqueous phase and not on the silicate surface) whereas higher-molecular types are sparingly soluble. Compared with tetrasodium pyrophosphate, nonionic polypropylene glycol has the advantage of not introducing any additional electrolyte into the paint; thus, such layer silicate sols can also serve as a rheological additive for high-performance, moisture-resistant industrial paints [6].

Figure 4.10: Gel formation by layer silicate dispersions

71

Rheologically active additives Like all electrostatically stabilised dispersions, layer silicate sols are sensitive to multivalent cations, such as Ca2+, Fe3+ or Al3+, which usually cause them to undergo irreversible flocculation (Chapter 1.2.1).

4.1.2 Fumed (pyrogenic) silica The production and surface treatment of fumed silica is described in Chapter 1.4.4. Because of their larger surface area, the finer fumed silicas are more rheologically active than the coarse types. The surface of fumed silica has 3.5 to 4.5 silanol groups per nm2 and thus somewhat fewer than is theoretically possible (about 8); see Figure 4.11. Fumed silica is slightly acidic (isoelectric point approx. 5; see Chapter 1.2.1); the pH of a 4 % aqueous dispersion is about 4. In other words, in water-borne paints (pH 8 to 9), fumed silica has a negative surface charge (Figure 4.11). Near the isoelectric point (pH 4 to 7), aqueous dispersions of fumed silica exhibit maximum viscosity. The uncharged or only slightly charged silica particles form a network in a manner akin to controlled flocculation. The reason for this is poor electrostatic stabilisation. This gel formation is reversible; the gels can be liquefied (sol formation) by shearing (e.g. stirring). If the alkalinity of the aqueous dispersion medium increases, the rheological action of fumed silica decreases and falls to zero at a pH of about 11 (complete electrostatic stabilisation of the sol). In the pH range 8 to 9 of interest here, the rheological action is still good, however. Generally, though, water is not the best medium for fumed silicas, as it is highly polar and solvates the hydrophilic silica completely. Consequently, no hydrogen bonds can form between the individual silica particles. Thus, hydrophobic fumed silicas (some 70 % of the silanol groups are silylated; see Chapter 1.4.4) are often more effective in aqueous alkaline paints, all the more Figure 4.11: Surface charge of fumed silica (simpliso because of their lower surface fied diagram, not to scale)

72

Rheological additives for water-borne and latex paints charge. It should be noted that certain paint ingredients (e.g. nonionic surfactants) can change the rheological action of aqueous silica dispersions.

4.1.3

Polymeric rheology modifiers

The group of rheology modifiers based on acrylates or polyurethanes requires that attention be paid to their interaction with other ingredients (e.g. pigments, fillers, surfactants, dispersing agents) of the water-borne paint composition. Essentially, the rheological action can be enhanced (synergism) or reduced (antagonism) [7]. Synthetic polymeric rheology modifiers can be classified by their mode of action into two groups: – liquid phase thickeners (no associative action) – ASE: alkali soluble/swellable emulsion (or dispersion) – certain cellulose ethers – associative rheology modifiers – HASE: hydrophobically modified anionic soluble emulsion (or dispersion) – HEUR: hydrophobically modified ethylene oxide urethane rheology modifiers The differences between the two types are fuzzy, however, and depend on the specific formulations. The following discussion describes rheology modifiers in terms of their chemical composition.

4.1.3.1 Acrylates

Acrylate-based rheology modifiers fall into high-molecul aracrylate copolymers of high acid number (ASE) and hydrophobically modified acrylate copolymers (HASE), with no sharp dividing line between the two (Figure 4.12).

Figure 4.12: Acrylate copolymers as rheology modifiers

73

Rheologically active additives Rheology modifiers based on acrylates (ASE and HASE) are mostly supplied as low-viscosity emulsions or dispersions. In this form, the acrylate copolymers are not fully neutralised, are insoluble in water and are not rheologically active, as the viscosity of polymer emulsions or dispersions is independent of the molar mass of the polymer. It takes complete neutralisation of the carboxyl groups (at pH 8 to 9  complete dissociation to anionic carboxylate groups) before the macromolecules either increase in volume by swelling or dissolve. The molecular chains expand owing to electrostatic repulsion of the now negatively charged carboxylate groups. Only then do the acrylate copolymers become rheologically active. After complete neutralisation, ASE acrylates swell in water or dissolve, giving rise to a marked increase in viscosity (Figure 4.13). The thickening effect mostly stems from the high molar mass of the water-soluble acrylates (ASE). Water becomes immobilised as water of hydration and intermolecular hydrogen bonds form between the macromolecules; reversible entanglement of polymer chains may also be a factor. In other words, in the ideal case, ASE acrylates would increase the viscosity only of the water phase. However, high-molecular acrylate copolymers of high acid number (e.g. ASE) have a quite pronounced affinity for metal oxide surfaces, such as inorganic pigments and fillers; this explains the dispersing effect of low-molecular polyacrylic acids (Chapter 1.3.1). High-molecular acrylate copolymers of high acid number (e.g. ASE) may have an opposite effect to low-molecular types on the stability of aqueous pigment dispersions; i.e. they can cause flocculation, especially at low concentrations (bridging flocculation as described in Chapter 1.2.2). According to the manufacturers, this should not actually happen with acrylates of the ASE type, but it should be borne in mind. This potential flocculating action forms the basis of the industrial use of (other) high-molecular acrylate copolymers of high acid number as flocculating agents in waste water treatment. To summarise, the rheological/thickening action of acrylates can be influenced by the type and quantity of pigFigure 4.13: Viscosity of an acrylic thickener as a function of pH (illustrative) ments or fillers.

74

Rheological additives for water-borne and latex paints This dependence of the rheological action on the type and quantity of pigments or fillers and the polymer dispersion used as binder applies especially to the hydrophobically modified acrylate copolymers (HASE). HASE acrylates combine the modes of action of both ASE and HEUR types (see Chapter 4.1.3.2). The characteristics of hydrophobically modified, water-soluble polymers are described in the next Chapter in the context of hydrophobically modified ethylene oxide urethanes (HEUR). Acrylate copolymers can be used not only in latex gloss enamels and latex paints but also to some extent in high-performance stoving enamels. During stoving, the ammonium salts of the acrylates undergo thermal decomposition and the hydrophilicity of the dried coating is reduced: Pol-COO−   +HNR3      Pol-COOH  +  NR3 

4.1.3.2

 ydrophobically modified ethylene oxide urethanes H (HEURs)

“Polyurethanes” or “PUR thickeners” are vague terms for this group of rheological additives, which are actually Hydrophobically modified Ethylene oxide Urethanes (HEURs) [8]. The molar masses range from 10,000 to 50,000 and so are relatively low. In contrast to acrylates, HEURs are nonionic, water-soluble (and so overall hydrophilic) polymers; i.e. the hydrophilicity is more or less retained in the dried coating. The urethane groups afford a simple way of linking hydrophilic and hydrophobic molecular segments (Figure 4.14). Rheological HEUR additives are amphiphilic block copolymers;

Figure 4.14: Linear, hydrophobically modified ethylene oxide urethane (HEUR); simplified diagram, not to scale

75

Rheologically active additives the polyethylene oxide chain is hydrophilic while the terminal alkyl groups (e.g. oleyl, stearyl, dodecylphenol, nonylphenol) are hydrophobic. This amphiphilic structure confers a certain degree of surface and interfacial activity. Hydrophobically modified ethylene oxide urethanes can also be branched (Figure 4.15).

Figure 4.15: Branched hydrophobically modified ethylene oxide urethane (HEUR); simplified diagram, not to scale

Figure 4.16: Association (network formation) of HEUR additives; simplified diagram, not to scale

76

Rheological additives for water-borne and latex paints

Rheological action of HEUR additives

1. Increase in viscosity due to dissolving of the polymeric HEUR additive (binding of water by the polyethylene oxide chain). 2. Association of the terminal, hydrophobic alkyl chains occurs in a manner akin to micelle formation in surfactants (Figure 4.16); this association yields a reversible, three-dimensional network which is responsible for the rheological action. 3. Furthermore, association takes place between the terminal, hydrophobic alkyl chains and the surfaces of polymer dispersion particles (in latex paints) and other hydrophobic particles (Figure 4.17). This interaction is deemed crucial to the formation of rheological structures, but it depends heavily on the polymer dispersion employed.

Interaction of HEUR additives with other paint ingredients

1. Low-molecular surfactants or wetting agents (Chapter 1.3.2) compete directly with terminal, hydrophobic alkyl chains on the HEUR additives and can heavily influence the rheological properties of a paint. It must be remembered that amphiphilic compounds (emulsifiers from emulsion polymerisation) can be entrained via the (primary) polymer dispersion. 2. Water-soluble organic co-solvents lower the interfacial tension of water and impair micelle formation (compare surfactants). This leads to a decrease in viscosity at low shear. 3. Water-insoluble ingredients, such as defoamers (Chapter 3) and coalescing agents (Chapter 10), become solubilised in the micelles and can increase the viscosity [9].

Figure 4.17: Association of HEUR additives with polymer particles; simplified diagram, not to scale

77

Rheologically active additives Table  4.1:  Water-solubility of coalescing agents (Chapter 10) and co-solvents Substance 2,2,4-trimethyl-1,3-pentanediol-1-isobutyrate Butyldiglycol acetate Butylglycol Butyldiglycol

Water-solubility Abbreviation (room temperature) “Texanol”

0.05 wt %

BDGA

6.5 wt %

BG

unlimited

BDG

unlimited

“Texanol” = trade name

The influence of coalescing agents and co-solvents [10] (addition level 3.0 wt.%; see Table 4.1) on a styrene-acrylate dispersion thickened with HEUR additives is shown in Figure 4.18 (viscosity at low shear). Clearly, the largely water-insoluble “Texanol” increases the viscosity of the dispersion while the sparingly water-soluble butyl glycol acetate does not, despite “dilution”. As expected, the completely water-soluble co-solvents (butyl glycol and butyl diglycol) lower the viscosity (Figure 4.18). These results, obtained for pure polymer dispersions (Figure 4.18), can be transferred to formulated latex paints, as was shown in laboratory tests (Figure 4.19) [10]. Guide formulation for a latex façade paint: – pigment/binder ratio P/B = 4 : 1 – pigment: titanium dioxide and fillers – binder: finely divided styrene-acrylate dispersion (e.g. “Acronal” 290 D) – addition of coalescing agent (Chapter 10): 4.0 parts by weight – two HEUR additives (“Rheolate” grades, addition: 1.0 part by weight, as supplied) – other additives, such as defoamers, dispersing agents and so on Figure 4.18: Influence of coalescing agents and co-solvents (see Table 4.1) on the viscosity of a styrene-acrylate dispersion thickened with HEUR additives (schematic diagram)

78

Figure 4.19 shows that the viscosity (at low shear) decreases with increase in water solubility of the coalescing

Rheological additives for water-borne and latex paints agent (Table 4.1). The rheology measurements (Figure 4.19) tallied with practical application via a lambskin roller; the paint containing added “Texanol” was the easiest to apply, while that containing butyl diglycol was the hardest. Once a paint containing HEUR additives has been made, it should be allowed to “ripen” overnight to allow sufficient time for the network structures to form in full. Generally, every paint or latex paint should be allowed to ripen overnight. Further advantages of HEUR additives include biostability and good levelling. They are particularly advantageous in latex gloss enamels intended to have good levelling properties [10].

4.1.3.3 Polysaccharides Cellulose ethers

Cellulose is a linear, water-insoluble biopolymer (molar mass 50,000 to 500,000) consisting of β-1,4-linked D-glucose units (Figure 4.20). It is obtained from, e.g., wood and cotton. Cellulose is insoluble in water because of strong intermolecular hydrogen bonds. However, it swells in about 20 percent sodium hydroxide solution and then lend itself to chemical reactions. Chemical modification (polymer-type conversion) of swollen cellulose in sodium hydroxide solution affords derivatives (Table 4.2 and Figure 4.21). In these derivatives (ethers), the intermolecular hydrogen bonds are sterically hindered by the substituents, and water-solubility is thus conferred. Three hydroxyl groups, at most, can react per glucose unit; i.e. the ma- Figure 4.19: Influence of coalescing agents on the ximum degree of substitution is DS = viscosity of latex paints containing two different 3. In the reaction with ethylene oxide HEUR additives

Figure 4.20: Cellulose (simplified diagram)

79

Rheologically active additives Table 4.2:  Degree of substitution of cellulose ethers for use in latex paints Cellulose ether

Abbreviation

Degree of substitution (DS)

Methyl cellulose

MC

1.5 to 2

Carboxymethyl cellulose

CMC

0.5 to 1

Hydroxyethyl cellulose

HEC

>2

to yield hydroxyethyl cellulose (HEC), the addition can continue in the side chain bearing the hydroxyl group, leading to DS values > 3. Therefore, hydroxyethyl cellulose must be characterised by an additional molar degree of substitution (MS; Figure 4.22). There also exist hydroxypropyl cellulose (HPC; reaction with propylene oxide) and mixed etherified types, such as methylhydroxypropyl cellulose (MHPC), methylhydroxyethyl cellulose (MHEC) and ethylhydroxyethyl cellulose (EHEC). A hydrophobically modified hydroxyethyl cellulose (hm HEC) is commercially available too (which provides associative thickening in the manner of HEUR). The various cellulose ethers on the market differ in their degree of substitution and molar mass (viscosity). They imbue latex paints with more or less shear thinning flow characteristics; shear-thinning increases with increase in molar mass of the cellulose ethers. It should be noted that synergistic rheological effects may also arise where different cellulose ethers are combined (e.g. 1 : 1 mixture of NaCMC and HEC).

Figure 4.21: Important cellulose ethers (simplified diagram)

80

Rheological additives for water-borne and latex paints The rheological or thickening action stems mainly from the high molar mass of the cellulose ethers. The water is bound as water of hydration and intermolecular hydrogen bonds form between the macromolecules. Reversible entanglement of polymer chains is of importance here. The thickening action of non-ionic cellulose ethers, such as HEC and MHEC, is independent of the pH, to a first approximation. In contrast, the thickening action of anionic carboxymethyl cellulose increases with increase in pH and reaches a maximum at about pH 8 (complete dissociation of the carboxyl groups). Cellulose ethers have a broad range of properties: – immobilisation of water through swelling of structures – rheological action in strongly alkaline milieu – water retention – interaction with solid particles by adsorption (dispersing action) – stabilisation of pigment and filler dispersions – binder effect (in special cases) Advantages of cellulose ethers: + renewable raw materials (chemically modified) + universal use in latex paints Disadvantages: – nutrient medium for bacteria and fungi (fungicides and biocides required) – dried films can be swelled by water The characteristic value quoted for cellulose ethers is not the molar mass but rather the viscosity [mPa s] of a 2 percent solution in water, which can rise to about 100,000. The addition level for cellulose ethers to latex paints and latex silicate paints is about 0.2 % (expressed in terms of the total formulation). Cellulose ethers can also be combined with layer silicates or HEUR additives.

Figure 4.22: Simplified structure of hydroxyethyl cellulose (HEC), where DS = 1.5 and MS = 2.5

81

Rheologically active additives

Xanthan gum

In contrast to cellulose, starch consists of some 80 % branched, water-swellable amylopectin (molar mass > 300,000) and some 20 % linear, water-soluble amylose (molar mass 50,000 to 150,000); both are composed predominantly of α-glycosidically linked D-glucose units. A major difference between cellulose (β) and starch (α), therefore, is the type of linkage between the D-glucose units. Xanthan gum is produced by biotechnological fermentation of starch. It has a molar mass of about 2 · 106, is branched and contains carboxyl groups. Aqueous solutions of xanthan gum undergo much more pronounced shear thinning than those of cellulose ethers. A 1 : 1 mixture of xanthan gum and hydroxyethyl cellulose exhibits a synergistic effect. The thickening action of xanthan increases with increase in pH (more so than in the case of carboxymethyl cellulose) and reaches a maximum at about pH 9 (dissociation of carboxyl groups). The main disadvantage of xanthan is its high price compared to that of cellulose ethers. Xanthan is used in latex paints and latex silicate paints (addition level: about 0.2 %, expressed in terms of the total formulation). It is also added to certain high-performance industrial stoving enamels, either alone [12, 13] or in combination with layer silicates [14]. Finally, xanthan serves as a thickener in the food industry (e.g. in jelly) and thus can be considered toxicologically harmless.

Figure 4.23: Quaternary ammonium ions for organic modification of layer silicates

82

Rheological additives for solvent-borne paints

4.2 Rheological additives for solvent-borne paints 4.2.1 Organoclays Unmodified hydrophilic layer silicates are described in Chapter 4.1.1. The metal cations (mostly sodium) on the surfaces of layer silicates can be replaced by other cations. In the case of bentonite and hectorite, ion exchange in aqueous medium goes to completion within 24 hours (highly simplified equation): [R4N+ Cl−] + [Na+ layer silicate−]  [R4N+ layer silicate−] + NaCl Quaternary ammonium ions containing at least one long-chain alkyl group (minimum C12; Figure 4.23) have proved themselves for practical coating applications; they enable hydrophilic layer silicates to be converted into hydrophobic organo-modified layer silicates, also known as organoclays (Figure 4.24). The adsorption of the long-chain ammonium ions increases the spacing of the primary platelets from its original value of 1 to 1.5 nm to 1.7 to 2.5 nm (see Figure 4.7 earlier). The long-chain alkyl groups are arranged regularly in the crystal (platelet stack). The long-chain alkyl groups hydrophobise the surfaces of the primary platelets and render them organophilic. However, there is no steric stabilisation as the alkyl groups are too short for that (for steric stabilisation, 10 nm would be necessary; see Chapter 1.2.2). Organoclays for paints are classified into standard types (e.g. “Bentone” 27, 34, 38) and readily dispersible types (e.g. “Bentone” SD 1 to SD 3). Figure 4.25 shows a scanning electron micrograph of the latter type (“Bentone” SD 1); the platelet structure can be clearly seen.

Figure 4.24: Primary platelets of organoclays (schematic diagram, not to scale)

Figure 4.25: Scanning electron micrograph of a readily dispersible organoclay “Bentone SD” type

83

Rheologically active additives For complete dispersion, the standard types need a polar activator, e.g. ethanol/ water (95/5), addition level 50 %, expressed in terms of organoclay. This activator (e.g. ethanol) diffuses between the stacked primary platelets and increases the distance between them by swelling. The stacks of platelets can be dispersed as far as primary platelets, ideally. The small quantity of water introduced by the activator adsorbs on the hydroxyl groups on the edges of the platelets and is crucial to gel formation by the organoclay dispersions in organic solvents (Figure 4.26); a minimum 3 % of water, expressed in terms of organoclay, is necessary. It should be noted that industrial organic solvents mostly contain a few tenths of one percent of water by way of impurity. Hydrogen bonds link the individual edges of platelets to form a three-dimensional network, i.e. a rheologically active, reversible gel structure (Figure 4.26). The standard organoclay types can be incorporated into paint in two ways. Preparation of basic paste (intermediate product): 1. to 87 parts by weight solvent 2. admix 10 parts by weight organoclay, and disperse for 10 minutes (agitator) 3. add 3 parts by weight activator/disperse for 5 to 10 minutes (agitator) During paint production, the basic paste is added to the mill base before the pigments. Alternatively, the organoclay can be dispersed directly in the mill base: 1. to solvents and paint resin 2. admix organoclay/disperse for 10 minutes (agitator) 3. add activator/disperse for 5 to 10 minutes (agitator) 4. add surfactants, if necessary

Figure 4.26: Gel formation by organoclays (schematic diagram, not to scale)

84

Rheological additives for solvent-borne paints 5. admix pigment and disperse 6. let down Process variants are possible depending on the type of paint. The readily dispersible organoclay types can be added as powder to the mill base direct (as per the second process above), without addition of activator. The readily dispersible organoclay types, too, require a certain (small) quantity of water for the formation of the gel structure; commonly, the requisite water is present in industrial organic solvents by way of impurity. Furthermore, paint batches made with readily dispersible organoclay types can be corrected. For this, the additional quantity of organoclay is first dispersed in a portion of the paint batch with the aid of an agitator and is then added to the remainder of the batch. Organoclays unfold their rheological activity particularly at low shear rates; they prevent or retard settling of pigments and fillers and sagging of fresh paint on vertical surfaces (see Figure 4.4). In contrast, paint levelling is impaired. This necessitates compromises during formulation. The addition levels for organoclays (as powder) are 0.2 to 1 %, expressed in terms of overall formulation; they find application in industrial and trade paints as well as corrosion protection. Not every organoclay type is suitable for every type of paint.

4.2.2 Fumed (pyrogenic) silica Certain aspects of fumed silicas are described in Chapters 1.4.4 and 4.1.2. The surface of fumed silica contains siloxane groups (O-Si-O), isolated silanol groups (Si-OH) and silanol groups linked by hydrogen bonds (Figure 4.27).

Figure 4.27: Surface of (unmodified, hydrophilic) fumed silica

85

Rheologically active additives The silanol groups on the surface crucially determine the rheological action of fumed silica in organic solvents. Fumed silica is most effective at forming rheologically active networks in organic solvents which form few, if any, hydrogen bonds (i.e. mostly non-polar media). The reason for this is that a network can be established by means of hydrogen bonds between individual particles (Figure 4.28); this process could also be viewed as controlled flocculation. On the other, hand, paint media with a tendency to form hydrogen bonds reduce network formation by fumed silica. Here, molecules of the paint medium compete with the silanol groups of the fumed silica; such liquids require significantly more fumed silica in order that the desired rheological properties may be obtained. Thus, the addition level and effectiveness of fumed silica depend heavily on the other ingredients of the paint. Furthermore, certain paint additives can greatly influence the rheological action of fumed silica. For example, in strongly polar systems, cationic surfactants can improve the effect of fumed silica. In non-polar systems, a further increase in the rheological action can by effected by small quantities (15 to 25 % expressed in terms of silica) of short-chain molecules bearing at least two functional groups capable of forming hydrogen bonds (e.g. with glycols; see Figure 4.29). During paint manufacture, fumed silica is best incorporated together with the pigment into the mill base; the addition level is 0.5 to 3 parts by weight, expressed in terms of the overall formulation.

Figure 4.28: Network of fumed silica (simplified diagram, not to scale)

86

Figure 4.29: Fumed silica network formed through addition of ethylene glycol (simplified diagram, not to scale)

Rheological additives for solvent-borne paints

4.2.3 Rheological additives based on urea These types of additives have been known for some time [11]. They used to be prepared by chemical reaction of amines with diisocyanates directly (in situ) in the corresponding solution of paint resin (Figure 4.30). During preparation, 2 moles benzylamine are dissolved in the paint resin, for example, and then 1 mole hexamethylene diisocyanate is slowly added dropwise under intensive (!) stirring. The diurea (Figure 4.30) forms immediately, precipitating out in the form of very tiny needle-shaped rheologically active crystals (Figure 4.31) [11]. The interaction (gel formation) between these crystals takes the form of hydrogen bonding; urea derivatives form very strong hydrogen bonds. The presence of very small urea crystals sometimes manifests itself as turbidity in the paint resin. The advantage of this method is that the particulate rheological urea additive is already optimally dispersed in the paint resin directly after synthesis. The disadvantage is that smaller paint manufacturers may possibly have problems with chemical synthesis in the production line. A further advantage of these urea additives is that, unlike all other particulate rheological additives, they do not reduce gloss in stoving enamels. In melamine formaldehyde resin (MF) stoving enamels, the urea additive reacts with the binder during stoving (Figure 4.32) and does not exist in particulate form afterwards. Consequently, it does not lower the gloss of the coating.

Figure 4.30: In situ preparation of a rheologically active urea additive

87

Rheologically active additives

Figure 4.31: SEM of the crystalline urea-derivative obtained from benzyl amine and hexamethylene diisocyanate in butyl acetate (and dried afterwards); scale bar 2 µm

New urea derivatives For some years, liquid urea additive solutions are available which can be easily incorporated into the liquid paint medium without chemical reaction: “Byk” 410, 411 and others. These are solutions of solid urea derivatives in N-methylpyrrolidone (NMP); NMP-free versions are now available, too. It should be noted that NMP is nowadays considered to be toxicologically harmful. On being admixed to paints, these additives precipitate in the form of very tiny crystals, because urea deri-

Figure 4.32: Reactions of urea additives in stoving enamels based on MF

88

Rheological additives for solvent-borne paints vatives are insoluble in common organic paint solvents. These urea crystals form a three-dimensional, rheologically active network as described before. The advantage of these new liquid types is that they can be incorporated into paint without undergoing a chemical reaction. However, intensive stirring is required, as otherwise specks can form. Figure 4.33 shows a model experiment of a horizontal glass bottle containing precipitated 5.4 % “Byk” 410 in xylene/butanol = 4 : 1; an almost transparent, non-sag gel can be seen in the lower part of the bottle.

Figure 4.33: Glass bottle containing a precipitate of 5.4 % “Byk” 410 in xylene/ butanol = 4 : 1

Figure 4.34: Urea additive which can be easily incorporated in liquid paints [15]; simplified presentation

89

Rheologically active additives Figure 4.34 shows a patent example [15] for such a new urea additive. Nowadays there exist also this kind of urea additives for water-borne paints (e.g. “Byk” 420).

4.3 Rheological additives in stoving enamels Stoving enamels, also called thermosetting coatings, are cured at elevated temperatures (80 to 250 °C) in an oven. They find particular application in industrial coating processes, e.g. automotive coatings. Paints can run (sag) when applied to vertical surfaces (see Figure 4.4 before), for example, if they are applied too thickly. But sagging can occur in the oven, too, when high-solids paints containing low-molecular paint resins are stoved. Figure 4.35 tries to illustrate this. Each solvent-borne stoving enamel (and especially water-borne types) must be flashed off before stoving to allow solvents to evaporate and to prevent pores from forming while in the oven (Figure 4.36); this defect is also called popping or solvent popping. In industrial application processes, attempts are naturally made to minimise the flash-off time.

Figure 4.35: Schematic diagram of the change in viscosity of solvent-borne stoving enamels after application and during stoving (high-solids paint, with and without rheological additive)

90

Rheological additives in stoving enamels Figure 4.35 shows that solvent evaporates from a conventional paint during the six-minute flash-off period, leading to an increase in viscosity (logarithmic scale). In the oven, the temperature of the painted object to be stoved rises until, after nine minutes, the crosslinking temperature is reached. In the interval between six and nine minutes after application, two processes happen simultaneously in the oven: 1. further solvents evaporate ( increase in viscosity) and 2. the object heats up, lowering the viscosity of the coating. The overriding effect is a lowering of viscosity due to heating (Figure 4.35). After reaching the crosslinking temperature (at about nine minutes), the viscosity rises markedly and rapidly due to incipient crosslinking. In the case of high-solids stoving enamel, the situation becomes interesting. As it contains lower-molecular, lower-viscosity binders and there­ fore less solvent, its viscosity does not increase so much during the flash-off period. However, the initial decrease in viscosity in the oven is all the greater, and this can lead to sagging on vertical surfaces if a critical viscosity (ηc) is not achieved (Figure 4.35). This can be prevented in the oven by adding suitable rheology modifiers, as shown clearly by the middle curve in Figure 4.36: Popped pores caused by too short a Figure 4.35. flash-off time in a stoving enamel

91

Rheologically active additives

4.4 Literature [1] G. Meichsner, T. Mezger, J. Schröder, Lackeigenschaften messen und steuern, Vincentz Network, Hanover, 2nd ed. 2016 [2] G. Lagaly, R. Fahn, Ton und Tonminerale, Ullmann’s Encyklopädie der technischen Chemie, Verlag Chemie, Weinheim, 4th ed. (1983), Vol. 23, p. 311 ff [3] H. van Olphen, An Introduction to Clay Colloid Chemistry, 2nd ed., Wiley 1977 [4] J. Bieleman et al., Additives for Coatings, Wiley-VCH 2000 [5] B. S. Neumann, K. G. Sansom, Farbe & Lack 81 (1975) p. 514 ff [6] R. Berg, B. Müller (BASF Lacke + Farben AG) US Pat. 5,198,490 and 5,290,845, Chem. Abstr. 110, 25453c (1989) [7] E. A. Johnson, Farbe & Lack 100 (1994) p. 759 ff [8] H. Reimann, B. Joos-Müller, K. Dirnberger, T. Schauer, C. D. Eisenbach, Farbe & Lack,

92

108 (2002) No. 5, p. 44–55; No. 8, p. 59– 63 and No. 9, p. 91–97 [9] J. Müller, U. Thies, H. Kober, J. Mazanek, Farbe & Lack, 104 (1998) No. 7, p. 30–40 [10] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed 2017 [11] C. A. M. Vijverberg, JOCCA, No. 7 (1988) p. 204–207 [12] B. Müller, E. Martin (BASF Lacke + Farben AG) EP 0 301 300 B1, Chem. Abstr. 111, 41442z (1989) [13] B. Müller, E. Martin (BASF Lacke + Farben AG) DE 38 32 142 C2, Chem. Abstr. 113, 154373n (1990) [14] B. Müller, E. Martin (BASF Lacke + Farben AG) EP 0 302 240 B1, Chem. Abstr. 110, 194743c (1989) [15] EP 1 188 779 A1

Driers

5 Catalysts Generally, catalysts (accelerators) are substances which lower the activation energy of a certain chemical reaction; catalysts therefore accelerate the reaction rate without appearing in the end product (Figure 5.1). They participate in the chemical reaction, but remain unchanged at the end, i.e. they are not consumed [16]. Every type of curing undergone by coatings can be catalysed as a chemical reaction, i.e. accelerated, and this has great practical consequences.

5.1 Driers Mostly, oxidative curing of alkyd resins is called oxidative drying or air drying. Strictly speaking, this is wrong [1]. To explain, consider the following two definitions. Drying of an applied coating is the transformation from the liquid into the solid state by solvent evaporation (physical drying).

Figure 5.1: Simplified diagram of activation energy, with and without addition of a catalyst (C) - - - 1: activation energy without catalyst, — 2: activation energy with catalyst (C) Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

93

Catalysts Curing of an applied coating is the transformation from the liquid into the solid state by increase in molar mass and crosslinking via chemical reactions. The term “oxidative” clearly indicates a chemical reaction; so, the term curing is correct. Furthermore, the term “drier” is misleading too; a drier is a catalyst which accelerates the reaction rate in oxidatively curing coatings. Unfortunately, however, the term “drier” is so deeply entrenched in coatings technology (e.g. [15]) that it must be used in this book, against the author’s better judgement. Oxidative curing is a crosslinking reaction triggered by the diradical, atmospheric oxygen; crosslinking occurs in paint resins bearing multi-unsaturated fatty acid side chains (e.g. alkyd resins, oils, and epoxy ester resins). The mechanism underlying these crosslinking reactions is still not fully resolved. Ideally, such crosslinking reactions yield three-dimensionally crosslinked coatings. Generally, the more double bonds there are in a molecule, the faster oxidative curing occurs. However, the type of double bond (isolated, conjugated) plays an important role [1]. The first step in the crosslinking reaction is autoxidation of an activated C-H bond of an unsaturated fatty acid by oxygen to form a hydroperoxide (Figure 5.2). Autoxidation occurs at allylic double bonds and is particularly fast at methylene groups between two

Figure 5.2: Formation and decomposition of hydroperoxides

94

Driers double bonds, so-called isolene fatty acids [1]; the ratio of the respective reaction rates is about 1 : 30. At elevated temperatures or at ambient temperature in the presence of redox catalysts (siccatives), hydroperoxides decompose to free-radicals, thereby initiating curing (Figure 5.2). Further information on oxidative curing may be found in references [1–3]. Driers are metal salts of long-chain organic acids (metal soaps) that are soluble in organic solvents. Driers are added to oxidatively curing paints to accelerate the crosslinking reaction (solidify the coating). The organic acids necessary for salt formation are either naturally occurring fatty acids or synthetic carboxylic acids, such as naphthenic acids (see Figure 1.31) and 2-ethylhexanoic acid. The effect of driers is to decompose hydroperoxides to free-radicals at ambient temperature (Figure 5.2). Especially effective are salts of metals capable of forming two or more oxidation states [e.g. Co(II)/Co(III), Mn(II)/Mn(III)/Mn(IV) or Ce(III)/ Ce(IV)] and thus acting as redox catalysts; they are also called surface driers. The most important redox catalyst is cobalt, whose mode of action is shown in Figure 5.3. In the first step, cobalt(II) is oxidised to cobalt(III) while in the second step cobalt(III) is reduced to cobalt(II); i.e. cobalt(II) is not consumed and thus is a true catalyst. Nowadays, the use of cobalt compounds is becoming more and more controversial due to possible toxicological effects. Cobalt-free driers are already available on the market. However, metal salts which form only one oxidation state [e.g. Ca(II), Ba(II), Zn(II) or Zr(IV)] serve as driers, too. These have a weak catalytic effect, if any, when used alone; but they can enhance the effect of redox catalysts. This group of driers can be divided into “through-driers” (or “coordination driers”) and “auxiliary driers” (Table 5.1) [4]. “Auxiliary driers” probably derive their effect (enhancement of the action of redox catalysts) from their Lewis acid character (see Friedel-Crafts catalysts). The mode of action of “through driers” (or “coordination driers”) is shown in Figure 5.4. The metal ions Figure 5.3: Mode of action of cobalt driers link paint resin molecules together by salt formation in a manner akin to ionomer crosslinking. Thus, “through-driers” are not true catalysts, as they are consumed. Oxidative curing accelerated by driers never comes to an end but becomes progressively slower. In the Figure 5.4: Mode of action of “through driers”

95

Catalysts Table  5.1:  Classification of driers Redox catalysts (“active” driers) “Through” driers

“Auxiliary” driers

Acceleration of peroxide decomposition

crosslinking of carboxy groups interaction with redox (and hydroxy groups) catalysts

Cobalt, manganese, cerium, iron, vanadium

(lead*), zirconium, lanthanum, neodymium, aluminium, bismuth, barium, strontium

calcium, potassium, lithium, zinc

* For ecological reasons, lead is hardly used nowadays

long term, excessive crosslinking may cause embrittlement (see also Chapter 8.2.3). At some stage, embrittlement leads to cracking, as shown in Figure 5.5. Dosing of driers must be done carefully as it can influence the life-time of the coating; overdosing should be avoided at all costs. Driers should be soluble in the paint, which is why metal soaps are used mostly. Metal soaps are surface-active and can accumulate, e.g., on the paint surface (see discussion on skin formation below). Furthermore, driers can lower the moisture resistance of coatings. The inherent colour of the driers can be a drawback in clearcoats or pale colours (e.g. Co(II) is an intense blue-violet). Since lead compounds have been classified as toxic or environmentally dangerous for a long time they should be replaced in paints. Especially zirconium is recommended as a substitute for lead. In the meantime, also, cobalt compounds are classified as dangerous for health and environment. So, cobalt has to be replaced in paints as far as possible. As cobalt siccatives are very effective, appropriate substitutes are difficult to find; manganese, iron and vanadium siccatives are offered for this purpose (all three Figure 5.5: Cracks in an old coat of oil paint on a wooden door are redox catalysts).

96

Driers Commercial driers are mostly characterised by their metal content [%]. As certain metal soaps selectively accelerate surface-curing and others enhance through-curing (see above), combinations of different metal soaps are most often used to achieve optimum curing and minimise addition of driers. Thus, premixed multi-metal driers whose composition has been optimised by the manufacturer are available on the raw material market. For styrene-modified alkyd resins and epoxy esters, multi-metal driers are not recommended; the right choice in this case is cobalt, as a single drier. Recently toxicological concerns have been expressed with respect to cobalt driers. Thus, alternatives such as manganese or vanadium driers should be taken into account. Certain inorganic pigments or fillers can also act as driers; examples are zinc dust, zinc oxide and calcium carbonate (zinc(II) and calcium(II) soaps are driers) as well as iron oxide pigments [redox catalyst Fe(II)/Fe(III)]. In particular, transparent iron oxide pigments can exert a noticeable influence because of their large surface areas (up to 100 m2/g). In contrast, carbon black pigments can retard curing, with metal soaps adsorbing onto their surface (up to 500 m2/g); the carbon black here acts as an adsorbent in the manner of charcoal. It is also possible for driers to lose their effectiveness after protracted storage (due to adsorption); one solution here, albeit questionable, is to add further driers. During paint manufacture, it is common for driers to be admixed to the cooled mill base after the pigment has been dispersed. Only calcium soaps are added before dispersing, on account of their pronounced surfactant properties. With some water-borne paint systems, all the driers are added to the mill base prior to dispersing; the shear forces created during dispersing distribute the driers in the disperse paint resin droplets (where they are also protected from hydrolysis).

5.1.1 Anti-skinning agents Anti-skinning agents are not catalysts at all. However, since they are used together with driers as additives for oxidatively curing paints, it is appropriate to discuss them here. Anti-skinning agents are a kind of inhibitor that temporarily prevent or retard oxidative curing; i.e. they are the opposite of a catalyst, which accelerates a reaction. Gradual accumulation of the surface-active metal soaps on the paint surface in the can and/or rapid surface curing can also lead to skin formation (skinning) on the liquid paint. Skin formation occurs most often in cans which have been opened, partly emptied and then re-sealed (large air volume!). To prevent skin formation in oxidative-curing paints, anti-skinning agents are incorporated. These anti-skinning agents may be certain derivatives of phenols and/or oximes (Figure 5.6). Oximes form labile complexes with metal ions, and so temporarily deactivate driers. Under paint-curing conditions, these labile complexes are cleaved again, and the drier

97

Catalysts becomes active. Oximes are often highly volatile. Methyl ethyl ketone oxime, widely used in the past, was classified as a class 3 carcinogen (suspected of causing cancer) by the European Union in 2001; for alternative complexing agents, see [5]. For this reason, oximes will not be discussed any further. Sterically hindered phenol derivatives, such as 2,6-di-tert.-butyl-4-methylphenol (common name: butylated hydroxytoluene, BHT, Figure 5.6) [6] are not volatile and act as so-called antioxidants; they scavenge free-radicals, thereby terminating free-radical chain reactions. Oxidative curing of the paint starts only when the phenol has fully reacted. BHT is also commonly employed as an antioxidant in plastics technology. A large number of further sterically hindered phenol derivatives acting as antioxidants are available on the market [7]. The mode of action of the simplest example, 2,6-di-tert.-butyl-4-methylphenol, is shown in Figure 5.7. R-O· is a binder free-radical, for example, produced by the decomposition of a hydroperoxide (see Figure 5.2). Two of these are consumed (scavenged) until the antioxidant is spent (Figure 5.7). For this reason, sterically hindered phenols are also called free-radical scavengers. Steric hindrance must be effective enough to prevent two phenol free-radicals (Figure 5.7) from dimerising; on the other hand, steric hindrance should not be so effective as to prevent the desired reaction with the sterically not hindered free-radical R-O·. The principle behind the antioxidizing action of sterically hindered phenol derivatives is found in Nature; the lipophilic antioxidant vitamin E is nothing other than a sterically hindered phenol derivative (Figure 5.8) [8]. Phenol derivatives retard curing of the paint films; retardation of surface drying can give rise to improve through-curing because the film is “open” for longer. Phenol derivatives can thus improve levelling and gloss in oxidatively curing paints. The strong (dark) inherent colour of certain oxidation products of phenol derivatives can be disadvantageous as it can cause yellowing. With water-borne paints, tests should be carried out to ascertain whether anti-skinning agents are necessary at all – alkyd emulsions (strictly two-phase systems) normally do not require them. Another group of free-radical scavengers based on sterically hindered amines is discussed under light stabilisers (Chapter 8.3.2.2). Figure 5.6: Typical anti-skinning agents

98

Catalysts for polyurethane coatings

5.2 Catalysts for polyurethane coatings By polyurethane coatings here is meant two-component systems which are based on polyol/polyisocyanate and cure at ambient temperature, as well as one-component stoving enamels based on blocked isocyanates [1]. The basic reactions of polyurethane chemistry are presented in Figure 5.9. Figure 5.9 shows the addition of an alcohol to an isocyanate to yield a urethane; this is the curing reaction undergone by two-component polyurethane coatings. However, urethanes are thermally unstable and can cleave at about 200 °C, which is a requirement for

Figure 5.7: Mode of action of 2,6-di-tert.-butyl-4-methylphenol as antioxidant

Figure 5.8: Simplified formula of a-tocopherol (one constituent of vitamin E)

99

Catalysts

Figure 5.9: Basic reactions of polyurethane chemistry (simplified diagram)

Figure 5.10: Reaction mechanism for the addition of an alcohol to an isocyanate

100

Catalysts for polyurethane coatings the one-component stoving enamels based on blocked isocyanates [1]. Figure 5.9 (lower part) shows the most important side reaction with water; this undesirable reaction can occur in two-component, solvent-borne polyurethane coatings in contact with moisture or in water-borne systems.

Solvent-borne two-component polyurethane coatings

Aliphatic (or cycloaliphatic) polyisocyanates react more slowly with polyols than aromatic polyisocyanates [1]. Two-component polyurethane coatings containing (cyclo)aliphatic

Figure 5.11: Mechanism of polyurethane catalysis (simplified diagram)

101

Catalysts polyisocyanates therefore usually have to be catalysed (but not systems containing aromatic polyisocyanates). The pot-life (processing time) of two-component polyurethane coatings is the period of time during which a mixture of polyol and polyisocyanate can still be applied. It is commonly accepted that the pot-life has elapsed once the viscosity has doubled. If two-component polyurethane coatings containing (cyclo)aliphatic polyisocyanates are to be cured rapidly at ambient temperature, it is usually necessary to add a catalyst (accelerator) to compensate for the lower reactivity relative to aromatic isocyanates. Naturally, adding a catalyst shortens the pot-life. Generally, a catalyst accelerates both the forward reaction (addition, urethane formation) and the reverse reaction (thermal cleavage of urethanes); see Figure 5.9. This means that catalyst addition lowers the thermostability of polyurethane coatings. For this reason, no more catalyst should be added than is absolutely necessary. To aid an understanding of the catalytic action, Figure 5.10 presents the chemical reaction mechanism behind the addition of an alcohol to an isocyanate. In this reaction, the nucleophilic oxygen atom (δ-) of the alcohol adds to the electrophilic carbon atom (δ+) of the isocyanate. To accelerate this reaction, either the nucleophilicity of the oxygen (alcohol) or the electrophilicity of the carbon (isocyanate) has to be increased. The addition reaction can be accelerated by both Lewis acids (electron acceptors, such as metal salts) and Lewis bases (electron donors, such as tertiary amines); see Figure 5.11. Lewis bases (tertiary amines) increase the nucleophilicity of the oxygen (alcohol) while Lewis acids (meFigure 5.12: Typical catalysts tal salts) increase the electrophilicity

102

Catalysts for polyurethane coatings Table  5.2:  Relative activity of some polyurethane catalysts Catalyst Concentration [%] Standard (without)

–

Relative reactivity 1

DABCO

0.1

130

DABCO

0.2

260

DABCO

0.3

330

DBTL

0.1

210

DBTL

0.5

670

0.1 + 0.2

1000

DBTL + DABCO

of the carbon (isocyanate); see Figure 5.11. The two catalyst types can also be combined, which may prove synergistic [9]. Some examples of catalysts are shown in Figure 5.12. In the case of DABCO (Figure 5.12), both free electron pairs (“centre” of nucleophilicity) on both nitrogen atoms are held in place by the bicyclic system, and the inversion that otherwise usually occurs with amines is impossible. Furthermore, note that, with the amines, there are always precisely two methylene groups between two nitrogen atoms; this is no accident, and serves to boost the catalytic effect. Table 5.2 shows the relative activity of some polyurethane catalysts [10]. It may be clearly seen that doubling the DABCO addition level from 0.1 to 0.2 % effects a doubling of reactivity. Trebling the DABCO addition level to 0.3 % does not treble reactivity, though; i.e. higher addition levels do not bring commensurate

Figure 5.13: Triethylamine and 2-amino-2-methyl-1propanol and the reaction of the latter with isocyanate

103

Catalysts improvements. The sum of the reactivities of 0.1 % DBTL and 0.2 % DABCO is 470 (Table 5.2); however, 1000 has been measured (Table 5.2) – clearly real synergism exists here. The production of dibutyltin dilaurate (DBTL) may give rise to toxic triorganic tin compounds by way of impurity [11]. For this reason, strenuous efforts are being made to replace stannous catalysts, such as DBTL, with safer alternatives, such as zinc or bismuth carboxylates, in industrial applications [11].

Two-component, water-borne polyurethane coatings

Two-component, water-borne polyurethane coatings [1] suffer from the problem of a side reaction between the isocyanate groups and water, with elimination of carbon dioxide (see Figure 5.9). This undesirable side reaction is catalysed by, amongst others, tertiary amines. Tertiary amines, such as triethylamine (Figure 5.13), serve as neutralising agents in water-borne paints [1]; neutralising agents are discussed in detail in Chapter 11. Where it becomes necessary to use too much triethylamine as a neutralising agent, the side reaction can become the principal reaction and the paint batch will start foaming due to evolution of carbon dioxide (Figure 5.14). There are two ways to resolve this problem: 1. Employ a polyol of lower acid number and thus lower amine demand; however, this leads to a completely new paint formulation (time-intensive). 2. Employ a reactive secondary or primary amine (e.g. 2-amino-2-methyl-1-propanol; Figure 5.13). Such an amine acts in the desired manner as a neutralising agent in the base paint component before mixing takes place with the isocyanate curing agent [1]. Upon mixing with the curing agent, the amine reacts very quickly with the isocyanate to yield a urea derivative (Figure 5.13) which is not catalytically active. However, urea derivatives are not basic and thus cannot act as a neutralising agent. Whether the second way works in a particular case can only be established by tests.

Stoving enamels based on blocked isocyanates

Figure 5.14: Foaming of a wrongly formulated two-component, water-borne polyurethane paint

104

Up to now, the curing reaction of stoving enamels based on blocked polyisocyanates was commonly catalysed with DBTL [1]. Due to the possible toxicity described above, alternatives must be sought for this, too [11].

Catalysts for two-component epoxy coatings

5.3 Catalysts for two-component epoxy coatings Epoxy resins crosslink with amino-functional curing agents at ambient temperature by an addition reaction (sometimes also called cold setting or room temperature vulcanising) [1]. As a result, the carbon atom of the oxirane ring acts as the electrophile while the amino-functional curing agent is the nucleophile. Hydroxyl groups can form hydrogen bonds with the oxygen atom of the oxirane ring and thus increase the electrophilicity and reactivity of the epoxy resin (Figure 5.15). Generally, all compounds bearing hydroxyl groups have an accelerating (catalysing) effect, e.g. salicylic acid, lactic acid and phenol (toxic). In two-components, water-borne epoxy paints water takes on the accelerating function; this is why pot-lives of two-component, water-borne epoxy paints are often shorter than those of solvent-borne systems. As hydroxyl groups are formed (Figure 5.15) during addition reactions, the reaction is self-accelerating (auto-catalysis) [12]. However, tertiary amines, too, accelerate ring opening of the oxirane ring by interacting with the electrophilic terminal carbon atom of the oxirane ring. As a result, the already polar C-O bond is further polarised and reactivity is increased (Figure 5.15).

Figure 5.15: Catalysed reactions between epoxy resins and amino-functional curing agents (simplified diagram)

105

Catalysts Catalysts for two-component, solvent-borne epoxy paints prepared by the Mannich reaction of phenol, formaldehyde and dimethyl amine are available on the market (Figure 5.16). Apart from the catalytic acidic aromatic hydroxyl group, these Mannich bases contain catalytic tertiary amino groups (Figure 5.16). There are also amine curing agents bearing nonreactive tertiary amino groups or hydro­xyl groups as “built-in” catalysts (Figure 5.17). The last example in Figure 5.17 shows

Figure 5.16: Accelerators for two-component, solvent-borne epoxy paints 1

These Mannich-bases are available as “Ancamine” 1110 and “Ancamine” K54

Figure 5.17: Typical amine curing agents which can act as accelerators

106

Acid catalysts for stoving enamels that reactive curing agents, too, can be prepared with the help of the Mannich reaction; for details, see [1]. The weakly acidic benzyl alcohol is often used in two-component epoxy paints to protect the amine curing agent from carbamate formation by protonation. Carbamate formation occurs during curing and is a side reaction of primary amines with carbon dioxide in the presence of moisture (Figure 5.18). The resultant carbamate can no longer react with the epoxy resin, leading to curing defects and film defects. Furthermore, benzyl alcohol has a weak accelerating action and also serves as solvent.

5.4 Acid catalysts for stoving enamels This Chapter deals with one-component paint systems which crosslink by polycondensation reactions at elevated temperatures and are thus called stoving enamels (and thermosetting coatings); polycondensation reactions can be catalysed by acids. In particular, this chapter will focus on stoving enamels based on amino resins [1]; stoving enamels based on blocked isocyanates are discussed above. During specific acid catalysis with strong acids, such as sulphonic acids Figure 5.18: Reaction between primary amines and and phosphoric acid, the reaction rate carbon dioxide (carbamate formation) is accelerated only by adding the conjugated acid of the solvent (in the case of aqueous solutions, the rate is a function of the pH). The poorly reactive HMMM resins (Figure 5.19) react by this mechanism. In the case of general acid catalysis by weak acids, such as carboxylic acids or also ammonium ions (from ionically blocked acids; see below), the reaction rate is accelerated by increasing the concentration of any acid (even very weak ones). Highly reactive melamine resins react by this mechanism. Thus, the relatively weakly aci- Figure 5.19: Simplified diagram of a HMMM resin dic carboxyl groups of alkyd resins, (hexamethoxy methyl melamine)

107

Catalysts

Figure 5.20: Typical acid catalysts

Figure 5.21: Ionic blocking of strong acids using para-toluene sulphonic acid as an example

108

saturated polyesters or acrylates (acid number!) can act catalytically, being sufficiently effective for most stoving enamels based on reactive amino resins. In other words, stoving enamels based on the widely employed alkyd resins, saturated polyesters and acrylates, each combined with a reactive amino resin, mostly do not need the addition of a separate acid catalyst. Some exceptions to this rule can be stoving enamels based on combinations of high-molecular epoxy resins and reactive amino resins because the acid number of epoxy resins is close to zero. Curing of the relatively weakly reactive HMMM resins can be accelerated significantly by strong acids [1]. HMMM resins without added catalyst crosslink with hydroxyl-functional paint resins only at temperatures above 180 °C but adding suitable acid catalysts can lower the stoving temperature to about 140 °C (saving on energy and costs). Various acids are available for use as additional catalysts in stoving enamels (mostly based on HMMM); these acids must naturally be soluble in organic solvents and compatible with paint resins. The acid strength in organic solution varies with the type of acid, akin to the situation regarding dissociation in aqueous solutions. The objective (specified stoving temperature and/ or stoving time) or amino resin to be catalysed determines

Acid catalysts for stoving enamels the type of acid catalyst to choose. High acid strength and/or high catalyst addition level can reduce storage stability. For low stoving temperatures (e.g. for repair coatings), a limited pot-life is acceptable. Acid catalysts are non-volatile and thus remain in the paint film; they are hydrophilic, polar substances and so may lower the moisture resistance of the coating. Acids particularly catalyse self-crosslinking of reactive melamine resins more than co-crosslinking. Thus, catalysed coatings may possess inferior flexibility and weatherability. The acids presented in Figure 5.20 are used for catalysing crosslinking reactions of amino resins. In this regard, sulphonic acids and phosphoric acid are strong acids whereas maleic acid monobutyl ester is only a weak acid. The storage stability of paints with added strong acids (sulphonic acids, phosphoric acid) is improved by using so-called blocked acids. The first type of blocking consists in neutralising the acids with amines to form ammonium salts; this also improves solubility in water-borne paints (ionic blocking, Figure 5.21). For example, the following amines can be used: diisopropanolamine and 2-amino-2-methyl-1-propanol (AMP) [1]. At ambient temperature, there exist among neutralised acids only the very weak acidic ammonium ions, which have hardly any catalytic effect on the weakly reactive HMMM resins. To an extent depending on the acid and amine, the ammonium salts decompose at elevated temperatures in the oven, the amines evaporate, the acids are released and act catalytically. The time and temperature at which decomposition occurs determine not only the crosslinking reaction but also the degree of levelling and sagging on vertical surfaces. As the weakly acidic ammonium ions of ionically blocked acids (Figure 5.21) catalyse the reaction of highly reactive melamine resins and thus may reduce storage stability, a second type of blocking is needed in particular cases, namely non-ionic Figure 5.22: Principle behind non-ionic blocking of strong blocking of acids (Figure 5.22). acids using an aromatic sulphonic acid as example

109

Catalysts Non-ionic blocking of strong acids uses esters of these acids which are not acidic. These esters decompose in the stoving oven into free acid and a volatile alkene via an electrocyclic reaction (Figure 5.22). For example, one such ester could be a β-hydroxy ester (from acid and oxirane), as presented in Figure 5.23. Non-ionic blocking is used if a stoving enamel contains a combination of weakly reactive HMMM resin and a highly reactive melamine resin; the weakly reactive HMMM resin must be catalysed during stoving by a strong acid whereas the highly reactive melamine resin must be explicitly prevented from being catalysed by ammonium ions in the can. An example of this is presented in the following formulation (Table 5.3) [13]. The water-borne clearcoat in Table 5.3 is a one-component stoving enamel based on acrylate/melamine resin. However, a combination of two differently reactive melamine resins is used here; this is rarely done but is not unusual. Table 5.3 shows a weakly reactive HMMM resin which needs a catalyst, and a highly reactive melamine resin which does not need a catalyst because the acrylate has an acid number of 36 mg KOH/g. In this case, it is imperative that a non-ionically blocked acid catalyst (item 6 in Table 5.3) be used because an ionically blocked acid would reduce storage stability of the reactive melamine resin by means of the ammonium Figure 5.23: β-Hydroxy ester (from acid and oxirane) ions present. as a non-ionically blocked acid catalyst

Figure 5.24: Light stabiliser based on a sterically hindered amine (free-radical scavenger)

110

Acid catalysts for stoving enamels Table 5.3:  Water-borne one-component, stoving clearcoat Item

Raw material

Parts by weight

Solids* 30.3

01

acrylic resin

70.56

02

butyl diglycol

5.68

03

HMMM resin

5.83

5.8

04

reactive melamine resin

2.44

2.2

05

levelling agent

0.09

0.1

06

curing catalyst

0.57

0.5

07

N-methyl pyrrolidone (NMP)

2.75

08

dipropylene glycol monomethyl ether

1.37

09

HALS

0.38

0.4

10

UV absorber

0.38

0.4

11

deionised water

9.95

Total

100

39.7

* Non-volatile matter 01:  water-borne acrylic resin, 43 % in water (+2.3 % “Texanol” and 7.8 % butyl glycol), pH 8.0 to 9.5 (DMEA), hydroxyl number about 110 mg KOH/g (solids), acid number 36 mg KOH/g (solids), “Macrynal” VSM 6285w 02:  cosolvent 03:  “Maprenal” MF 904 (solvent-free) 04:  reactive methanol-etherified melamine formaldehyde resin, 90 % in isobutanol, “Cymel” 327 05:  levelling agent (modified silicone), “Byk” 333 06:  ester of para-toluene sulphonic acid, 90 % active substance, acid number max. 10 mg KOH/g (as supplied), “Additol” VXK 6357 07:  cosolvent, for toxicological reasons it should be replaced 08:  cosolvent, evaporation number about 400, “Dowanol” DPM 09:  light stabiliser, free-radical scavenger based on a sterically hindered amine (HALS), no influence on the activity of the acid catalyst, emulsifiable in water, “Tinuvin” 123 (see Chapter 8.3.2.2 and Figure 5.24) 10:  light stabiliser, UV absorber based on benzotriazole, emulsifiable in water, “Tinuvin” 1130 (see Chapter 8.3.2.1) 11:  deionised water

The formulation contains a second interesting detail as regards catalysis by strong acids: the light stabiliser is based on a sterically hindered amine (item 09 in Table 5.3, Figure 5.24); light stabilisers [14] are explained in detail in Chapter 8. Light stabilisers based on sterically hindered amines are usually bases and can block acids permanently (!), i.e. have a disabling effect (Chapter 8.3.2.2). To prevent this, a special type of amine must be used (Figure 5.24) based on =N-O-R; i.e. an O-alkylhydroxylamine derivative which is not basic, on account of the electronegativity of the oxygen; it is unable to block acids. In acid-catalysed systems, therefore, all other paint ingredients must be checked for possible interaction with acids and, if necessary, individual ingredients may have to be replaced.

111

Catalysts

5.5 Literature [1] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed 2017 [2] U. Poth †, Polyester und Alkydharze, Vincentz Network, Hanover 2014 [3] W. J. Muizebelt et al., Progr. Org. Coat. 24 (1994) p. 263 ff, Progr. Org. Coat. 31 (1997) p. 331 ff and Journ. Coat. Technol. 70, No. 876 (1998) p. 83 ff [4] R. Wissmann, R. W. Hein, Farbe & Lack 106, No. 3 (2000) p. 38 ff [5] D. Edelmann et al., Farbe & Lack, 107, No. 12 (2001) p. 54–66 [6] https://en.wikipedia.org/wiki/Butylated_ hydroxytoluene [7] R. Gächter, H. Müller, Taschenbuch der Kunststoff-Additive, Hanser Fachbuch­ verlag, 3rd ed. 1989 [8] S. Böhmdorfer, T. Rosenau, Nachrichten aus der Chemie, 56, No. 4 (2008) p. 411–417

112

[9] L. Thiele, R. Becker, H. Frommelt, Zeitschr. für Polymerforschung 28 (1977) p. 405 ff [10] Kunststoffhandbuch, Vol. 7, p. 95, Hanser Verlag 1983 [11] D. Oberste, A. Steinert, New tin-free catalysts as alternatives to DBTL and amine-based compounds in modern solventborne polyurethane clearcoat systems, brochure of Borchers GmbH [12] P. Mischke, Film Formation, Vincentz Network, Hanover 2009 [13] Starting formulation of the former Vianova [14] A. Valet, Lichtschutzmittel für Lacke, Vincentz Network, Hannover 1996 [15] J. Bieleman et al., Additives for Coatings, Wiley-VCH 2000, chapter 7 [16] https://en.wikipedia.org/wiki/Catalysis

Adhesion promoters

6 Adhesion promoters Adhesive strength is a measure of the resistance of a coating to mechanical removal from the substrate. Permanent adhesion (under wet conditions, too) of the coating to the substrate and within a multi-coat system is a basic prerequisite for the protective effect (e.g. corrosion protection). One exception is strippable coatings, which are temporary coatings for protecting goods during transport. To illustrate, Figure 6.1 shows large scale loss of adhesion of a coating on a hot-dip-galvanized rainwater drainpipe.

Definition of adhesion/cohesion

Adhesion is defined as the effect of the forces of attraction at the interface of two different solid phases. It is expressed in units of energy/area, while the units of adhesive strength are force/area [1]. The counterpart to adhesion is cohesion; this is the effect of forces of attraction within the same phase. If the adhesion is inadequate, the interface between the substrate and the coating must be modified. There are various ways of doing this: 1. Modify the substrate surface, namely by surface preparation such as sanding, or by pre-treatment, such as flame treatment (plastics) or phosphating (metals). 2. Apply an adhesion-promoting layer between the substrate and coating, e.g. silane adhesion promoting primers (Chapter 6.2) or thin polymeric adhesive layers (Chapter 6.3). 3. Modify the coating by adding adhesion-promoting additives, e.g. silane adhesion promoters Figure 6.1: Loss of adhesion of a coating on a (Chapter 6.1). hot-dip-galvanized rainwater drainpipe

Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

113

Adhesion promoters On metal substrates the distinction between 1 (phosphating) and 2 (thin polymeric adhesive layers) is fuzzy, however, as both are so-called conversion layers.

Figure 6.2: Typical silane adhesion promoters

Figure 6.3: Reaction of silane adhesion promoters

Figure 6.4: Reaction between an amino-functional silane adhesion promoter and epoxy resin

114

Silane adhesion promoters Adhesion promoters are substances which improve adhesion between two solid phases, for example substrate and coating. Various substances are claimed to promote adhesion [2]. However, this book will deal only with the most important type: silane adhesion promoters.

6.1 Silane adhesion promoters Silane adhesion promoters are well established in adhesives technology [2–5]. They are also known as organo-functional trialkoxysilanes (Figure 6.2).  ow silane adhesion promoters work H The reactivity of the three alkoxy groups is similar to that of esters, as tetraalkoxysilane is the tetraalkyl ester of silicic acid. Hydrolysis of alkoxy groups by atmospheric moisture yields highly reactive silanol groups (Figure 6.3). Silanol groups can react by condensation with hydroxyl groups on the surfaces of metals (Me-OH), glass (Si-OH), mineral pigments, fillers and construction materials (e.g. Si-OH, Ti-OH, Al-OH), and with each other (Si-OH). The reaction with hydroxyl groups on surfaces effects a change in the functionality of the inorganic surface (Figure 6.3). Functional group X, chosen to suit the functionality of the binder (paint resin), can react with the binder to produce a covalent bond between the binder and the inorganic surface. Figure 6.4 shows an example of this type of covalent bonding produced by the reaction of aminosilane with epoxy resin.

Figure 6.5: Silane adhesion promoters as coupling agents

115

Adhesion promoters

Figure 6.6: Effectiveness of silane adhesion promoters for different substrates

Figure 6.7: Side reaction of silane adhesion promoters

116

Amino-functional silane adhesion promoters are unsuitable for stoving enamels based on amino resins because, as bases, they inhibit the acid-catalysed curing reaction and they permanently block acid catalysts (Chapter 5.4). In contrast, amino-functional silane adhesion promoters are eminently suitable for stoving enamels based on blocked polyisocyanates because the curing reaction is accelerated somewhat by amines. This illustrates the need to adapt the silane adhesion promoters precisely to the binder. Silane adhesion promoters are also called coupling agents [6] (see Figure 6.5). The spacing group mostly consists of three methylene groups (γ-silanes). Relatively new are types featuring only one methylene group (α-silanes); the alkoxy groups which they contain offer increased reactivity [7]. However, the efficiency of silane adhesion promoters depends heavily on the substrate (Figure 6.6) [2]. Applications of silane adhesion promoters a) Promoting adhesion of adhesives and sealants (important) [2, 3] Adhesive bonding and sealing of glass, metals and mineral construction materials b) Promoting adhesion between paint resin and mineral fillers Inorganic fillers, such as talcum, can be surface-treated with

Silane adhesion-promoting primers silane adhesion promoters (Chapter 1.4.3). Choosing the right silane for the paint resin concerned enables the bond between the (cured) binder matrix and the filler to be improved. This impedes water vapour diffusion through the coating and boosts corrosion protection [8]. c) Promoting adhesion of coatings (of little importance to date) Primary goal here is to improve adhesion under wet conditions [9].

Table 6.1: Formulation for the GLYMO adhesionpromoting primer [12] parts by wt. Methoxypropanol

mmol

88.8

986

Deionised water

10

556

Glacial acetic acid (catalyst)

0.2

GLYMO Sum

1

4.2

100

However, caution must be exercised when using silane adhesion promoters in paints, as they can react with traces of water in solvent-borne paints or with atmospheric moisture (Figure 6.7). These side reactions (hydrolysis + condensation, Figure 6.7) yield high-molecular polysiloxanes which may possibly give rise to cratering in coatings. Therefore, where silane adhesion promoters are used in paints, appropriate additives should be incorporated to bind traces of water (e.g. molecular sieves); excessive moisture should be avoided, too.

6.2 Silane adhesion-promoting primers These adhesion promoting primers make it possible to produce ultra-thin adhesion layers just nanometres thick. They can be produced at ambient temperature, as follows (Tables 6.1 and 6.2) [12]: – start with methoxypropanol – add water and, if necessary, acetic acid – slowly add silane, with stirring – stir for one hour (in the lab, using a magnetic stirrer) Two formulations are shown by way of typical examples in Tables 6.1 and 6.2; the silanes used are presented in Figure 6.8. The storage stability of the GLYMO primer solution (Table 6.1) is reported as being 50 days [12], because the large excess of methoxypropanol (glycol ether) shifts the equilibria in the direction of the reactants (Figure 6.9). Glacial acetic acid (concentrated acetic acid) is needed as a catalyst. Catalysts are often added to silanes, with acids as well as bases generally able to catalyse hydrolysis and condensation reactions respectively (Figure 6.9).

117

Adhesion promoters The GLYMO adhesion-promoting primer can, for example, be coated with two-component polyurethane paints because the epoxy group of GLYMO is able to react with the isocyanate groups of the hardener (Figure 6.10). But it would be wrong to state that the GLYMO adhesion-promoting primer will yield positive results with two-component polyurethane paints generally. Each case needs to be examined on its own merits. The DAMO adhesion-promoting primer (Table 6.2) would be also suited to two-component polyurethane paints. A catalyst is not necessary for the DAMO adhesion-promoting primer (Table 6.2), as the amino groups of DAMO (bases) act catalytically. As a result of this autocatalysis, the storage stability of the DAMO primer solution is reported as being just five days [12], but this is sufficient because these primers can easily be produced on a just-in-time basis. Application of the adhesion-promoting primers: Apply primer solution thinly to degreased metal paFigure 6.8: (3-Glycidyloxypropyl)trimethoxysilane nels and immediately wipe off. Ap(GLYMO) and N-2-aminoethyl-3-aminopropyltri methoxysilane (DAMO) ply the paint, and cure both primer

Figure 6.9: Equilibria in aqueous silane solutions, with an excess of alcohol and glycol ether (R-OH)

118

Thin polymeric adhesion layers and paint. Wiping off is important Table 6.2: Formulation for the DAMO adhesion[12] as it is the only way to obtain ul- promoting primer tra-thin primer layers, which is a reparts by wt. mmol quirement for their effectiveness. Methoxypropanol 93 1032 Overly high expectations should Deionised water 5 278 not be imposed on these adhesion DAMO 2 8.9 promoting primers; they can provide Sum 100 very good adhesion, especially under wet conditions, but not generally. Each case needs to be examined on its own merits. Reference [12] describes a large number of similar adhesion-promoting primers based on different silane adhesion promoters.

6.3 Thin polymeric adhesion layers Thin layers of poly(meth)acrylic acids can improve the adhesion of cured coatings (e.g. AK/MF or two-component polyurethane), e.g. to steel under wet conditions [10]. These layers are produced by dipping sheet steel into dilute (about 1 wt.%) aqueous solution of poly(meth)acrylic acid (for 1 to 2 min) at a pH around 3. The iron is converted into iron (II), and an organic conversion coating is obtained (Figure 6.11). Deposition of the poly(meth) acrylic acids is a form of autophoresis or chemiphoresis; the thickness of these layers is in the range of na- Figure 6.10: Reaction between isocyanates and noscale, ultrathin layers: 20 to 30 epoxy groups to produce oxazolidones nm. For polyacrylic acid, the optimum effect has been found at a molar mass of about 100,000 [10]. The approximate chemical composition of such adhesive layers is presented in Figure 6.11. Phenolic resins (resols) and epoxy resins can also generate thin adhesive layers [11]. Poly(meth)acrylic acids are not typical paint additives, but rather a Figure 6.11: Composition of an adhesive layer of polyacrylic acid on steel form of metal pre-treatment.

119

Adhesion promoters

6.4 Aminosilane as hardener in two-component epoxy paints In this particular paint system, aminosilane is used in amounts substantially greater than 1 part by weight and strictly speaking is not an additive but rather a part of the binder. As it fits in well with the silane chemistry described in this chapter, it will be discussed here briefly. The systems presented in this chapter work as follows. First, a silicone containing alkoxysilane groups is condensed with an epoxy resin containing hydroxy groups [13]. The condensation product is then made to react with an aminosilane hardener, whereupon two different crosslinking reactions take place: a) addition of the amino groups of the aminosilane to the oxirane rings of the epoxy resin b) condensation (influenced by humidity) of the alkoxysilane groups of the siloxane/silicone joined to the epoxy resin with those of the aminosilane (similar to Figure 6.7).

Figure 6.12: Schematic, simplified crosslinking of the silicone-epoxy resin hybrid with 3-aminopropyltriethoxysilane (AMEO) in the presence of humidity [13, 16]

120

Aminosilane as hardener in two-component epoxy paints Table 6.3: Guide formulation [17] and two formulation variants [16] for two-componentssilicone-epoxy resin hybrid systems Raw material

Guide formulation

Less crosslinking

More crosslinking

Silicone-epoxy resin

56.8

59.0

54.8

Component 1 Titanium dioxide

22.5

22.5

22.6

Talc

1.7

1.7

1.7

Hydrophobic fumed silica

0.9

0.9

0.9

Butyl acetate

4.3

4.4

4.0

AMEO

13.8

11.5

16.0

Sum

100.0

100.0

100.0

Component 2

Silicone-epoxy resin: Silikopon EF (98 %) Titanium dioxide: Kronos 2310 Talc: Finntalc M15-AW Hydrophobic fumed silica: Aerosil R972

This “dual” curing yields a highly crosslinked siloxane/silicone-epoxy resin hybrid [13, 14]. Essentially, what is being used here is 3-aminopropyltriethoxysilane (AMEO) as an aminosilane and a silicone-epoxy resin system based on a cycloaliphatic epoxy resin. Figure 6.12 is an attempt to schematically show the crosslinking of the silicone-epoxy resin hybrid with AMEO [13, 16]. As cycloaliphatic epoxy resins are not a commonly employed group of binders, they are described briefly below. A glycidyl ether of hydrogenated bisphenol A and epichlorohydrin serves as cycloaliphatic epoxy resin. These react with amines during curing in a manner akin to aromatic epoxy resins based on bisphenol A. However, there are some differences. These cycloaliphatic epoxy resins show a much lower viscosity per equivalent molecular mass compared to the corresponding aromatic binders and, naturally, they have better lightfastness. However, they do not have the same good level of anticorrosion properties as the aromatic binders [15].

Formulations A guide formulation [17] was used to produce a white paint. To examine the influence of the crosslink density, this formulation [17] was systematically crosslinked with AMEO [16] to greater and lesser extents. All three formulations, presented in Table 6.3, are ultra-high-solids paints (non-volatile matter > 90 %).

121

Adhesion promoters The paints are applied to steel panels and cured for one week at ambient temperature. Adhesion, adhesion after one week of immersion in water, pendulum hardness, flexibility (Erichsen cupping) and MEK resistance were tested. Within the limits of the measuring accuracy of these test methods, no significant differences were observed between the three coatings in Table 6.3 [16]. Interesting, but not unexpected, was the fact that additional stoving for 45 minutes at 180 °C increased MEK resistance and pendulum hardness (though flexibility decreased, of course, and a certain amount of yellowing was observed). Further experiments showed that adhesion after one week of immersion in water can be improved by partially replacing the titanium dioxide with the barrier filler, mica.

6.5

Literature

[1] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed 2017 [2] E. M. Petrie, Handbook of Adhesives and Sealants, McGraw-Hill, New York 2000 [3] B. Müller, W. Rath, Formulating Adhesives and Sealants, Vincentz Network, Hanover 2010 [4] E. P. Plueddemann, Silane Coupling Agents, Plenum Press, New York 1982 [5] U. Deschler et al., Angew. Chem. 98 (1986) p. 237–253 [6] B. Meyer-Roscher, S. Wellman, H. Brockmann, kleben & dichten, Adhäsion, 39 (1995) No.4, p. 34–37 [7] Organofunctional Silanes from Wacker, www.wacker.com/cms/media/ publications/downloads/6085_EN.pdf [8] N. Wamser, E. Urbino, Farbe & Lack 95 (1989) p. 109 ff [9] W.-D. Kaiser, A. Rudolf, S. Pietsch, Farbe & Lack 103, No. 7 (1997) p. 80–88

122

[10] Z. Gao, H. Yamabe, B. Marold, W. Funke, Farbe & Lack, 98 (1992) p. 917 ff [11] J. Gähde, Farbe & Lack, 101 (1995) p. 689 ff [12] pdf-file “Innovative Sol-Gel Coatings with Sivento Silanes” formerly located at www.sivento-silanes.de (internet address no longer valid) [13] M. Dornbusch, U. Christ, R. Rasing, Epoxy Resins, Vincentz Network, Hanover, 2016, p. 94ff [14] M. Heuer, M. Siemens, F. Eichenberger, Farbe & Lack, 122, No. 7 (2016) p. 58ff [15] U. Poth †, Synthetische Bindemittel für Beschichtungssysteme, Vincentz Network, Hanover, 2016, p. 289 [16] D. Beljaeva, M. Wrenger, project work, University of Applied Sciences Esslingen, 2017 [17] Evonik, Guide Formulation No. 4005, 7/2007

Corrosion of metals

7 Corrosion protection additives 7.1 Corrosion of metals Corrosion is a (mostly chemical) reaction between a (in this case, metallic) material and its surroundings (corrosive medium) which causes a measurable change in the material and may lead to corrosion damage (Figure 7.1). The corrosion reactions described here are interface reactions; i.e. the corrosion rate depends on the size of the phase boundary. All utility metals, such as iron, are thermodynamically unstable with regard to their oxides und thus are always covered with a (hydrated) oxide layer (dotted line in Figure 7.1). These oxide layers can be very thin and thus nearly invisible; in certain cases, they can protect thermodynamically unstable metals from corrosion (e.g. chromium, aluminium); i.e. transform them into a metastable state. In the following discussion of atmospheric corrosion, the corrosive medium is moist air; mild steel (approx. 99.5 % iron) corrodes in an electrochemical reaction (Figure 7.2). Atmospheric corrosion of iron proceeds by consumption of oxygen and is called oxygen corrosion. Electrochemical corrosion of iron proceeds in two steps (Figure 7.2). First, iron is oxidised to the Figure 7.1: Schematic diagram of the corrosion of sparingly soluble iron(II)hydroxi- metallic materials (the dotted line indicates the de[Fe(OH)2, colourless], and then oxide layers on metals)

Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

123

Corrosion protection additives

Figure 7.2: Electrochemical corrosion of iron

Figure 7.3: Corrosion and osmotic blistering at the tank filler spout of a car

124

to red hydrated iron(III) oxide – rust. The fact that the intermediate iron(II) hydroxide is sparingly soluble slows down the second reaction because the sparingly soluble product of the first reaction is the starting material for the second. The cathodic partial reactions are the same in the first and second reactions and yield alkaline hydroxyl ions (Figure 7.2). These hydroxyl ions are formed in the cathodic region and, in the case of a (damaged) coating on steel, lead to blistering by osmosis (Figure 7.3) because the cathodic region beneath the coating is close to the damage (so-called cathodic delamination). This causes the rust to penetrate underneath the coating (subsurface corrosion, Figure 7.3). A further consequence of the cathodic partial reaction is that binders for corrosion protective coatings must be resistant to saponification because otherwise they will be hydrolysed by hydroxyl ions. The amount of metal lost due to corrosion corresponds to some 5 % of metal production, and also depletes energy resources. Annual losses due to metal corrosion in industrialised countries are estimated to be about 3.5 % of gross domestic product.

Distinction from anticorrosive pigments

7.2 Distinction from anticorrosive pigments Adhesion of a coating is a necessary but on its own insufficient requirement for corrosion protection. Additional protective measures methods should therefore be considered, such as the use of corrosion inhibitors and corrosion protection additives. Strictly speaking, anticorrosive pigments do not belong in this chapter; however, by way of introduction and to differentiate them from corrosion protection additives, they are discussed briefly below. Anticorrosive pigments can be classified as follows: a) Chemical anticorrosion pigments (e.g. zinc oxide) bind corrosion stimulators, such as chloride and sulphate, by forming insoluble compounds and/or stabilising the pH of a coating in contact with the corrosion meFigure 7.4: Scanning electron micrograph of talcum dium. Therefore, a slight degree of solubility in the corrosion medium is necessary. b) Electrochemically active anticorrosive pigments (e.g. zinc chromate (toxic), zinc phosphate) passivate metal surfaces by forming thin layers, such as chromate and phosphate layers. Again, slight solubility in the corrosion medium is necessary. The distinction from chemically active pigments (a) is sometimes unclear because anticorrosive pigments exist which can act in both ways (e.g. zinc phos- Figure 7.5: Barrier action of lamellar, passive anti­ phate). corrosive pigments (diagram not to scale)

125

Corrosion protection additives c) Active, cathodic protective anti-corrosion pigments (e.g. zinc dust) act as sacrificial anodes and thus protect the metal substrate. They are “classic” pigments and insoluble in the application medium. d) Passive anticorrosive pigments, barrier pigments (lamellar particles, e.g. micaceous iron oxide, talcum: Figure 7.4) lengthen the diffusion pathways for corrosion stimulators and therefore improve the corrosion protection afforded by a coating (Figure 7.5). Again, they are “classic” pigments (or fillers) and insoluble. Anticorrosive pigments can be classified with respect to their solubility in corrosion media as slightly soluble (a, b) and insoluble (c, d). Some highly active anticorrosive pigments (e.g. the carcinogenic zinc chromate or red lead) should not be used, for ecological reasons. However, the relatively ecologically safe active anticorrosive pigments (e.g. phosphates) are less effective. Thus, where possible, non-toxic organic additives are needed to improve the corrosion protection afforded by coatings. Such additives are commonly termed corrosion inhibitors [1, 2]; a more correct term is corrosion protection additives [3, 4]. The difference between the two terms is explained below. Corrosion inhibitors in particular are substances dissolved in a liquid medium that concentrate at the metal (oxide)/medium interface. In other words, a protective layer at the interface is generated by adsorption, thereby preventing or decreasing corrosion (see electrochemically active anticorrosive pigments). This definition characterises the primary property of

Figure 7.6: Sets of corrosion inhibitors, corrosion protection additives and various anticorrosive pigments (simplified diagram)

126

Distinction from anticorrosive pigments corrosion inhibitors used in the water of heating and cooling circuits. For application in anticorrosive coatings, corrosion inhibitors have to meet additional requirements such as: – Adequate solubility in the liquid paint – Compatibility with the binder(s) – Low water solubility in the dried or cured coating – No adverse side effects, such as discoloration or increased hydrophilicity of the coating These demands have led to the development of corrosion inhibitors that specifically possess secondary properties for coatings [1], namely corrosion protection additives. Corrosion protection additives are thus a subset of corrosion inhibitors (Figure 7.6). In other words, a corrosion protection additive is a corrosion inhibitor, but the converse is not true. Figure 7.6 uses set theory to visualise the differences between corrosion inhibitors, corrosion protection additives and four different types of anticorrosive pigments (a to d, see Figure 7.6). For environmental reasons, red lead (Pb3O4), a chemically and electrochemically active anticorrosive pigment, is rarely used nowadays in Western Europe; however, it is still commonly encountered in aged prime coats (Figure 7.7). Red lead can be easily recognised by its orange-red

Figure 7.7: Prime coat containing red lead (exposed by wear)

Figure 7.8: Two organic corrosion protection additives MBTS: “Irgacor 252” [19]; also available as bis(tridecyl ammonium) salt “Irgacor 153” [1] Basic zinc salt of nitroisophthalic acid: “Heucorin RZ” [5]

127

Corrosion protection additives colour (Figure 7.7). Where red lead has been used in prime coats, the lead-containing debris generated by abrasive blasting during renovation must be disposed of properly.

7.3 Organic corrosion protection additives While a number of corrosion protection additives are available on the market, the suppliers do not provide enough details of their composition and mode of action. Two corrosion protection additives of known structure are presented in Figure 7.8. Details of the mode of action of 2-mercaptobenzothiazolyl succinic acid (MBTS) are known. On one hand, MBTS acts like a weak surfactant, improving the wetting of steel by an alkyd resin; the contact angle of an alkyd resin solution (60 % in xylene) falls from 114° to 78° upon 2 % addition of MBTS. On the other, MBTS reacts with iron ions during the corrosion process, inhibiting corrosion by forming an insoluble iron-MBTS complex (Figure 7.9). A key prerequisite for the corrosion-inhibiting effect of MBTS is the formation of an iron-MBTS complex at the iron oxide/coating interface, i.e. chemisorption of MBTS on the iron oxide surface. Furthermore, this iron-inhibitor complex must be insoluble; if it were readily soluble; the corrosion reaction would be accelerated (stimulated). If formation of the iron-inhibitor complex is faster than rust formation, corrosion is inhibited (Figure 7.9). Particularly good inhibition has been observed for a combination of MBTS (corrosion protection additive) and zinc phosphate (active anticorrosive pigment).

Figure 7.9: Reaction between 2-mercaptobenzothiazolyl succinic acid (MBTS) and iron ions

128

Corrosion inhibitors for metal pigments in aqueous alkaline paint media MBTS acts in two ways: 1. As a corrosion protection additive in the dried coating (Figure 7.9). 2. As a “classic” corrosion inhibitor in liquid (!) water-borne paints by preventing flash rust formation during flash-off.

Table 7.1:  Weight lost by mild steel after 200 h in an aqueous solution of sodium chloride (0.2 g NaCl/l) Sample

pH value Loss in weight [%]

Standard (without MBTS)

7.0

2.6

8.5

1.8

9.5

1.0

Addition of

7.0

0

A model experiment clearly revea2 % MBTS* 8.5 0 led the corrosion inhibiting effect of 9.5 0 MBTS in water-borne paints; here, * Ammonium salt the weight loss of mild steel after 200 hours in an aqueous sodium chloride solution (0.2 g NaCl/l) was determined at different pH values (Table 7.1). Table 7.1 shows two results. First, MBTS inhibits the corrosion of iron in an aqueous sodium chloride solution. Second, raising the pH passivates the iron (this has been known for a long time from the use of mild steel in alkaline concrete). A more recent group of corrosion protection additives for water-borne paints are complexes of 3-toluoylpropionic acid (TPA, Figure 1.10) [1], such as the TPA-amine complex [(TPA)2 · N-ethylmorpholine] and a TPA-zirconium complex.

7.4 Corrosion inhibitors for metal pigments in aqueous alkaline paint media The need to inhibit hydrogen corrosion of aluminium pigments in alkaline water-borne metallic paints is mentioned in Chapter 1.4.5.2. This chapter will summarise results obtained over the years in numerous studies of the stabilisation of aluminium pigments by corrosion inhibitors. A model system of water and butyl glycol in the ratio 9 : 1 was adjusted with DMEA to a pH between 8 and 10 [6]. The following classes of compounds were found to have corrosion inhibiting properties (see references for details). a) Low-molecular corrosion inhibitors – chelating agents [6, 7] – antioxidants [8] – fluorinated surfactants [9] (Figure 7.11) b) Oligomeric and polymeric corrosion inhibitors – polymers [10, 11] – paint and printing ink resins [10, 12] Figure 7.10: 3-Toluoylpropionic acid (TPA)

129

Corrosion protection additives Figure 7.11 shows the fluorinated surfactants studied (an example for a book about paint additives). The results were transferable to “paint-like” systems (mixtures of paint resins) [13]. The principles, too, were transferable in modified form to corrosion inhibition of other metal pigments, such as zinc [9, 12, 14, 15], copper and brass [16]. It should be noted that aluminium and zinc corrode by evolution of hydrogen whereas the more noble copper and brass react by absorption of oxygen; all these corrosion reactions can be measured gas volumetrically. Aluminium: Zinc: Copper: Brass:

2 Al + 6 H2O  2 Al(OH)3 + 3 H2  Zn + 2 H2O  Zn(OH)2 + H2  2 Cu + O2  2 CuO Cu/Zn + O2  CuO + ZnO

The equations for the chemical reactions of copper and brass are not trivial, as theoretically two different copper oxides (CuO or Cu2O) can be formed. However, gas volumetrically revealed that only copper(II) oxide (CuO) is formed. At pH 8, certain paint resins are capable of totally inhibiting corrosion [12]. Paint resins are crucial to the stability of aluminium pigments in water-borne paints. Studies of the interaction of paint resins and zinc pigments [14] have led to the development of storage-stable, water-borne, one-component zinc-rich primers [17].

Figure 7.11: Fluorinated surfactants [9]

130

Literature In all the references cited so far, gasvolumetry served as the measuring method (hydrogen evolution or oxygen absorption). In one case, it proved possible to study the adsorption of certain polymers and paint resins on aluminium with the aid of x-ray photoelectron spectroscopy (XPS) and to correlate the results with the gas volumetrically findings [18].

7.5 Literature [1] A. Braig, Corrosion Inhibitors, in Additives for Coatings (editor J. Bielemann), Wiley-VCH, Weinheim (2000), p. 291–305 [2] G. Schulte, W. Greß und R. Höfer, Welt der Farben (1995) No. 2, p. 17–22 [3] B. Müller, Farbe & Lack, 106 (2000) No. 11, p. 156 [4] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed. 2017 [5] www.heubachcolor.com/de/produkte/ pigmentsuche/heucorin/ [6] B. Müller, M. Gampper, Werkstoffe Korrosion, 45 (1994) p. 272–277 [7] B. Müller, K. Franze, Werkstoffe Korrosion, 45 (1994) p. 467–473 [8] B. Müller, M. Müller, I. Löhrke, Farbe & Lack, 100 (1994) p. 528–532 [9] B. Müller, Tenside Surf. Det., 37 (2000) p. 241–244 [10] B. Müller, Reactive & Functional Polymers, 39 (1999) p. 165–177

[11] B. Müller, M. Schubert, C. Oughourlian, Pigment & Resin Technology, 30, No. 1 (2001) p. 6–12; in 2002 honoured with the “Highly Commended Award” of the Emerald Literati Club [12] B. Müller, Europ. Coat. Journ., No. 5 (2001) p. 81–83 [13] B. Müller, C. Oughourlian, D. Triantafillidis, Journ. Coat. Tech., 73, No. 917 (2001) p. 81–84 [14] B. Müller, P. Kienitz, Farbe & Lack 102 (1996) p. 49–55 [15] B. Müller, J. Langenbucher, Corros. Sci., 45 (2003) p. 395–401 [16] B. Müller, C. Oughourlian, M. Schubert, Farbe & Lack, 105, No. 5, (1999) p. 48 ff [17] B. Müller, P. Kienitz, Farbe & Lack 102 (1996) p. 76–80 [18] B. Müller, Surface Coat. Int. Chapter B: Coat. Transact., Vol. 85, B2 (2002) p. 111–114 [19] www.chemicalbook.com/ChemicalProductProperty_EN_CB9186907.htm

131

Protection of coatings from weathering

8 Protection of coatings from weathering Weathering of coatings

Weathering is the stress experienced by coatings outdoors. The harmful components of the weather are light (especially UV radiation), atmospheric oxygen, heat and cold, and temperature changes, as well as water in the form of liquid (rain, dew) and vapour (moisture). Water can accelerate the weathering damage by causing swelling and drying. Industrial emissions (“acid rain”) of corrosive gases (SO2, NOx, HCl) play their part, too. Coatings not exposed to weathering can last a very long time. Examples include the stone-age cave paintings in Altamira and Lascaux (about 14,000 B.C.); however, a certain degree of sintering with the stone substrates must be factored in here. Further testament to the durability of coatings are the magnificent paintings in the sepulchres of ancient Egypt (3rd and 2nd millennium B.C.), see Figure 8.1; for the most part, these sepulchres are artificial rock caves without light and an extremely dry microclimate. Coatings are classified by their base binder as inorganic, mineral systems (e.g. lime and silicate paints) and as organic, polymeric systems (e.g. oil and synthetic resin paints). Organic coatings are particularly damaged by light and oxygen. By contrast, inorganic coatings are damaged by water (not usually by light) and are thus more durable than organic systems. Modern organic coating systems (e.g. heavy-duty corrosion protection) have a service life of about 25 years and must then be renovated. Inorganic coatings often last longer. Examples of extreme weather resistance are the relatively well-preserved pain- Figure 8.1: Wall painting inside the sepulchre tings (on pit lime plaster) on the outer (!) of Sennofer, Thebes, Egypt (1500 B.C.) Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

133

Protection of coatings from weathering façades of the Moldavian monasteries in Bucovina (Romania) [3]; these paintings mostly date back to the 16th century A.D. (Figure 8.2). The following discussions will be limited to organic coatings.

8.1 Photooxidation/ UV degradation 8.1.1 Absorption and emission of light Figure 8.2: Painting (1535 A.D.) on the outer façade of the church of the Humor monastery in Romania

Figure 8.3: Simplified Jablonski diagram absorption and emission of light

134

Familiarity with photochemical processes is important not only for photooxidative degradation of coatings and light stabilisers but also for UV curing (chain initiation, Chapter 9); a brief discussion therefore follows. Only that fraction of light which is absorbed by the exposed substance can initiate chemical reactions (Grotthus-Draper law). Light (electromagnetic radiation) can be absorbed by a substance only in the form of whole quanta (photons, E = h ν). The Jablonski diagram (Figure 8.3) shows the processes which occur when light is absorbed and emitted. Usually, a molecule is in the singlet ground state (S0), in which the electrons are paired, and their spins are non-parallel. Absorption of light requires only 10-15 seconds and causes the molecule to leave the S0 state and transfer to the first excited singlet state (S1); see Figure 8.3. Absorption of light by an organic molecule requires the presence of a chromophore (molecular segment capable of absorbing light). Generally, chromophores are more or less

Photooxidation/UV degradation extended π-electron systems. As the Table 8.1:  Radiation impinging on the earth’s S1 state has only a very short life- surface (global radiation) Wave length Proportion time of 10-10 to 10-7 seconds, only Radiation [nm] [energy-%] monomolecular chemical reactions UV: UV-B 285 to 315 0.4 are possible (e.g. intramolecular  UV-A 315 to 400 5.6 scission of a chemical bond). Bond scission occurs when the amount of VIS 400 to 780 50 energy absorbed is greater than the IR 780 to 1400 44 binding energy of a chemical bond in the molecule. Homolytic bond scission leads to the formation of two free-radicals (see also UV curing, Chapter 9). Internal conversion (IC) is a so-called radiationless deactivation process which lasts 10-12 to 10-6 seconds. During fluorescence (duration: 10-12 to 10-6 seconds), light is emitted, and the molecule returns to the S0 state (Figure 8.3). Fluorescent light usually has less energy than the absorbed light (hν’ < hν). During intersystem crossing (ISC), the molecule transfers through spin reversal to the first excited triplet state (T1). Strictly speaking, this transition is forbidden in quantum mechanics, yet it still takes place. This explains the long-life time (in the order of seconds) of the T1 state. The T1 state is diradical by nature. As the T1 state has a relatively long-life time, bimolecular (intermolecular) chemical reactions can occur there. In this connection, energy can be transferred (by collision) to other molecules which themselves cannot absorb light (this is called sensitisation). During phosphorescence (duration: in the order of seconds) light is emitted again, and the molecule returns to the S0 state (Figure 8.3). Phosphorescent light usually has less energy than fluorescent light (hν” < hν’). Visible light (VIS) has wavelengths from 400 nm to roundabout 780 nm. Non-visible radiation below 400 nm is called ultraviolet radiation (UV) while that above 780 nm is known as infrared radiation (IR). Above the outer atmospheric layers (in space), the energy spectrum of the solar radiation is continuous. As radiation passes through the atmosphere, some of the long-wave infrared fraction is absorbed by water vapour (H2O) and carbon dioxide (CO2) and so only short-wave infrared radiation (thermal radiation) impinges on the earth’s surface. The short-wave UV radiation below 175 nm (so-called vacuum UV) is absorbed by oxygen (O2) at altitudes greater than 100 km. UV radiation of wavelengths between 175 and 290 nm is absorbed by the ozone layer (O3) at an altitude of 25 to 30 km. The radiation ultimately impinging on the earth’s surface is presented in Table 8.1.

135

Protection of coatings from weathering

8.1.2 Photooxidation of polymers and paint resins

Figure 8.4: General definition of corrosion

Figure 8.5: Schematic comparison of the electrochemical corrosion of metals and the photooxidation of organic materials The dotted line on the surface of the metallic material represents the oxide layer

136

Like all other organic compounds, polymers are thermodynamically unstable with regard to their oxidation products (CO2, H2O and others) [1]. Thermodynamics do not reveal anything about the period over which this oxidation occurs. Under common conditions of use, a high energy of activation leads to thermodynamic metastability (this term is explained in [2]). In other words, polymers exhibit thermodynamic relationships similar to those of utility metals, which are also thermodynamically unstable with regard to their oxides (Chapter 7.1). Organic materials such as polymers or coatings are slowly oxidised by atmospheric oxygen (autoxidation, Table 8.2); in everyday parlance, this phenomenon is also called “ageing”. A simultaneous effect exerted by light (especially of energy-rich UV radiation) is acceleration of these oxidation reactions (photooxidation). The photooxidative degradation of polymers (UV degradation) is a chemical reaction between a polymeric material and its environment (UV radiation, oxygen) and may be described as a corrosion reaction in a broader sense (Figure 8.4). If the term corrosion is expanded to the extent shown in Figure 8.4, then all materials are subject to (different) corrosion reactions (Table 8.2). Figure 8.5 schematically compares the electrochemical corrosion of metals (Chapter 7.1) with the photooxidation of organic materials and reveals the following substantial differences, among others:

Photooxidation/UV degradation Table 8.2:  Corrosion reactions of materials (in the broadest sense) Inorganic materials Organic materials Nonmetallic materials (e.g. glasses, enamel)

metallic materials

polymeric materials (coatings, plastics)

• Chemical corrosion (leaching)

• chemical corrosion • electrochemical corrosion

• autoxidation • photooxidation

– Electrochemical corrosion of metals occurs only in the interface between metal (oxide)/medium. Photooxidation also occurs in deeper layers close to the interface (as far down as the light and oxygen can penetrate). As coatings are very thin polymer layers, this fact is significant. – Light does not play any role in the electrochemical corrosion of metals but does so in photooxidation. – Utility metals, unlike organic materials, are covered by oxide layers which can significantly influence corrosion resistance (Chapter 7.1) An impressive example of the influence of sunlight is shown in Figure 8.6. The fronts of both road signs are extensively exposed to sunlight and exhibit all the various types of weathering damage. Clearly (Figure 8.6, centre), photochemical degradation of the coating of the upper sign cancels out its corrosion protective effect on the substrate; the steel starts rusting. The rear of the signs is always in shadow and thus are nearly intact (Figure 8.6, right). Figure 8.7 shows a scanning electron micrograph of an extensively degraded railcar

Figure 8.6: Weathering of road signs; front extensively exposed to sunlight (left), detail there of (centre) and rear, always in shadow (right)

137

Protection of coatings from weathering coating. The surface of the coating shows exposed particles (inorganic pigments and fillers); this defect is caused by photooxidative degradation on the organic binder (Figure 8.7). An extreme example of weathering damage to automotive coatings is shown in Figure 8.8; as the car bonnet is “heated” from below by the engine while the sunlight impinges at a steep angle of incidence, car bonnets often suffer pronounced weathering damage.

Figure 8.7: Scanning electron micrograph of a heavily degraded railcar coating

Figure 8.8: Weathering of a car bonnet

Figure 8.9: Photooxidation of polymers (highly simplified diagram)

Figure 8.10: Free-radical chain branching and polymer chain scission

138

Photooxidation/UV degradation Photooxidation has been studied in great detail for the simplest polymer, polyethylene [(-CH2-CH2-)n]; photo-oxidation is a free-radical chain reaction [1, 4] (Figure 8.9). The reader might initially have some difficulty understanding the chain initiation in Figure 8.9. Ultrapure polymers (e.g. polyethylene) often have very good weather resistance because they lack chromophores; as a result, they cannot absorb high-energy radiation, which is necessary for free-radical formation (chain initiation). Without absorption, photooxidation cannot be initiated [1]. Engineering polymers mostly contain enough process-related additives or impurities which act as chromophores and transfer the absorbed radiation-energy to the macromolecules especially by collision in the long-lasting T1-state. These impurities can thus act as sensitisers (Figure 8.9, chain initiation). It should be noted that the chain termination reaction yields a crosslink (Figure 8.9) which causes the polymer to embrittle. Once the free-radical chain reaction is underway, it starts accelerating because photochemical scission of the formed hydroperoxides yields further free-radicals. This is called free-radical chain branching (Figure 8.10). Furthermore, polymer chain scissions may occur. This has two consequences: – Polymer chain scission forms an aldehyde which can absorb UV radiation (n-π* transition). – Polymer chain scission (Figure 8.10) leads (like crosslinking, Figure 8.9) to polymer embrittlement. Paint resins have a much more complicated chemical structure than polyethylene. Thus, the photooxidation reactions undergone by paint resins have not been wholly

Figure 8.11: Photooxidation of an aromatic poly­urethane based on MDI (simplified diagram)

Figure 8.12: Yellowing of a foamed material based on aromatic polyurethanes

139

Protection of coatings from weathering elucidated yet. Generally, aromatic components in paint resins (e.g. epoxies, polyurethanes) lower the resistance to light because they can absorb long-wave UV radiation and thus can initiate photochemical reactions. A sensitiser is then no longer required. Two-component polyurethane coatings containing aliphatic isocyanates are light resistant whereas those with a content of aromatic isocyanates yellow rapidly on exposure to radiation containing relatively long-wave UV light (335 to 410 nm; Figure 8.11). This yellowing is readily seen in cheap foamed materials based on aromatic polyurethanes (Figure 8.12). Chlorinated paint resins (chlorinated rubber, PVC-copolymers) are also relatively unstable to weathering because of the ready formation of chlorine free-radicals.

8.2 Damage done to coatings by weathering 8.2.1 Chalking Chalking is the detachment of pigments and fillers which become exposed due to binder degradation on the coating’s surface. In particular, chalking refers to the photooxidative degradation of coatings containing titanium dioxide [5, 6]. Figure 8.13 shows chalking on the plastic module of a car; comparison with the intact metal coating of the car bonnet clearly reveals the extent of the damage. Figure 8.14 shows pronounced chalking on an outdoor metal gate. A finger-tip run over the coating turns white due to titanium dioxide; the coating is becoming “bluer” (Figure 8.14). Figure 8.15 shows a scanning electron micrograph of this coating. The surface consists only of titanium dioxide particles which are no longer bound to each other by the binder (Figure 8.15).

Figure 8.13: Chalking of the coating on a plastic module for a car

140

Figure 8.14: Chalking on an outdoor metal gate (for SEM, see Figure 8.15)

Damage done to coatings by weathering Figure 8.16 is a simplified diagram of binder degradation caused by chalking. After chalking, the titanium dioxide particles are no longer bound to each other by binder but rather are bound only to the coating by a thin base of binder (in the shaded area) and thus can be wiped off easily. During chalking, binder is degraded photooxidatively by two different mechanisms: – “normal” photooxidation (Chapter 8.1.2) – photocatalytic oxidation cycle in which titanium dioxide catalyses (accelerates) photooxidative degradation of the binder.

8.2.1.1 Photocatalytic oxidation cycle Anatase, a titanium dioxide modification, is photocatalytically more active than rutile, which is why rutile is used in coatings as a rule. Most inorganic coloured pigments (e.g. iron or chromium oxides) are photo-inactive; i.e. they even stabilise the coating against photooxidative degradation (see Chapter 8.3.1). Figure 8.17 schematically shows the mechanism of the photocatalytic oxidation cycle on titanium dioxide surfaces. In moist air, there are always surface hydroxyl groups (Ti-OH groups) on the surface of titanium dioxide. The photocatalytic oxidation cycle presented in Figure 8.17 can be explained in four steps: 1. The primary step in chalking is the absorption of UV-A radiation by titanium dioxide to yield an electron-hole pair (exciton, Figure 8.18), as titanium dioxide is an extrinsic semiconductor. The electron is promoted from the filled valence band to the empty conduction band by the absorbed radiation energy.

Figure 8.15: Scanning electron micrograph (SEM) of an extensively chalked coating surface (metal gate in Figure 8.14)

Figure 8.16: Chalking (highly simplified diagram)

141

Protection of coatings from weathering 2. The OH free-radical initiates a free-radical chain reaction in the binder directly beside the titanium dioxide particle because the life time of the free-radical is very short(10-4 to 10-5 seconds). 3. Atmospheric oxygen reoxidises titanium(III) to titanium(IV) and the free-radical anion O2- is adsorbed. 4. Protons from the dissociation equilibrium of water release the peroxide free-radical (HO2·) which also reacts with the binder directly beside the titanium dioxide particle.

Figure 8.17: Photocatalytic oxidation cycle (the charges on the titanium represent oxidation numbers)

Figure 8.18: Absorption of UV radiation by titanium dioxide (energy band model)

142

The remaining hydroxyl ion reacts with the titanium(IV) to re-form the surface hydroxyl group (Ti-OH); the photocatalytic oxidation cycle is thus complete (Figure 8.17). Concurrently with “normal” photooxidation on the coating’s surface, internal degradation takes place (photocatalytic oxidation cycle on titanium dioxide and UV degradation), because firstly moisture and oxygen can easily diffuse into the coating. Secondly, the UV radiation necessary for degradation is reduced only by 30 to 40 energy-% at a depth of 30 μm in a coating pigmented with 15 vol.% titanium dioxide. Measures for avoiding chalking of rutile pigments: – Incorporate photo-inactive zinc or zirconium oxide into the upper layers of the pigment particle. – Additionally stabilise chalking by covering the pigment surface with

Damage done to coatings by weathering precipitated silica (hydrated silica: SiO2 · x H2O) and alumina hydrate (Al2O3 · x H2O); see Chapter 1.4.2.1. These precipitates have high surface areas and catalyse recombination of free-radicals (chain termination). Intentional chalking There are now proposals for using photocatalysts to accelerate photooxidation of coatings on façades as a way of reducing dirt pick-up and infestation by algae or fungi; increased photooxidative degradation should render the coating self-cleaning [13]. So, the damage pattern of chalking is also considered to be beneficial. Photooxidative uncovering of pigments and fillers at the surface of façade coatings and road-marking paints induces a kind of self-cleaning (intentional chalking). Pigmentation with nanoscale titanium dioxide (anatase) as photocatalyst can yield self-cleaning surfaces; however, the working life of the coatings is reduced.

8.2.2 Fading of organic colorants Fading is the photochemical degradation of organic colorants by absorbed UV radiation (Figure 8.19). The photochemical stability of organic coloured pigments depends on their chemical composition and their concentration in the coating; mass tones are often (apparently) more stable than pastel shades. Furthermore, the photochemical stability of organic colorants decreases with decrease in particle size; molecularly dissolved organic dyes are commonly not lightfast (Figure 8.20). Coloured organic pigments become photochemically more stable with increase

Figure 8.19: Fading of an organic red pigment (upper traffic sign)

Figure 8.20: Organic colorants

143

Protection of coatings from weathering in the number of inter-molecular and intramolecular hydrogen bonds [7]. Indanthrone blue (pigment blue 60) is shown in Figure 8.21 as a simple example of two intramolecular hydrogen bonds. These can transform absorbed light energy in portions into (harmless) heat energy (see more in Chapter 8.3.2.1).

Figure 8.21: Indanthrone blue (Pigment Blue 60) containing two intramolecular hydrogen bonds

8.2.3 Embrittlement, crack formation and delamination

Figure 8.22: Schematic diagram of changes to networks during UV degradation

Photooxidative scission of long and thus tough-elastic polymer chains into smaller fragments (Figure 8.10 before) and subsequent crosslinking of these fragments (PolO-O-Pol, Figure 8.9 before) causes weathered coatings to become hard and brittle over the course of time. As a result, the glass temperature (Tg) and surface hardness increase. Figure 8.22 is an attempt to portray these changes to networks; in this regard, the number of additional crosslinks is equal to the number of chain scissions. Additionally, low-molecular (plasticising) film components or cleavage products can be washed out. This lowers the binder mass and raises the PVC and Tg. Embrittlement proceeds from the coating’s surface to the interior of the film. It yields a brittle upper layer over a tough-elastic lower layer (compare crackle finish); over time, thermal and hygral movements of the film lead to cracking (Figure 8.23 and 8.24). The example shown in Figure 8.23 is extreme but very rare in weathered automotive coatings.

Figure 8.23: Cracks in an automotive coating (crack width in the order of millimetres)

144

Damage done to coatings by weathering Ultimately, progressive embrittlement may even lead to delamination of the coating (Figure 8.25).

8.2.4 Special aspects of wood weathering Wood generally has a porous structure (see Figure 3.2 in Chapter 3 before). Its chemical composition is presented in Table 8.3. Irradiation (300 to 540 nm wavelength) of uncoated wood causes photochemical degradation of the aromatic component lignin [15] which manifests itself as a brown discoloration. A simplified chemical equation for this photooxidation reaction to yield a chromophore (brown discoloration!) is shown in Figure 8.26. Over time, the degradation products are leached out by rain water, leaving behind colourless cellulose. The wood starts to turn grey and may ultimately become colonised by fungi. Outdoor coatings on metallic or mineral substrates can last up to 25 years; wood protection coatings on the sides of buildings exposed to sunlight tend to crack after just a few years (Figure 8.27). The causes include photochemically induced embrittlement (see Chapter 8.2.3 above) and hygral movement of wood: swelling in high moisture and drying (shrinkage) in low moisture, due to the wood’s porosity. Figure 8.28 shows a fracture in a plywood sheet coated with a very thin layer of a clearcoat based on an oxidatively curing alkyd resin and illustrates the complex structure of coated wood.

Figure 8.24: Cracks in a railcar coating (crack width in the order of centimetres) [14]

Figure 8.25: Delamination of a railcar coating [14]

Figure 8.26: Photooxidation of lignin (highly simplified diagram)

145

Protection of coatings from weathering Table 8.3:  Composition of wood Portion [%] Compound

Characteristics/function

40 to 50

cellulose* (fibres)

supporting and structural substance (absorbing tensile forces)

20 to 30

lignin

binder

20 to 40

hemicelluloses

cementing substances, flexibilisers (lower-molecular than cellulose)

1 to 10

sugar, starch, protein extractable substances

0.2 to 0.8

minerals

* See Figure 4.20 in Chapter 4

Table 8.4:  Crack formation by one-component clearcoats Base resin Crack formation after (always + melamine resin) Florida [a] WOM KFA [h]* AK

1

200

AK/AY

2

1000

AY

>5 **

1500

AY + light stabilisers

>5 **

3200

* Weather-O-Meter, cut-off filter A ** Test finished after 5 years

Figure 8.27: Coated wooden beam on an open-roofed balcony; west-facing (exposed to sunlight) and north-facing (not exposed to direct sunlight)

146

Damage done to coatings by weathering In contrast to wood, metallic or mineral substrates do not undergo hygral movement [8] and coatings on them are thus more stable. In other words, the substrate too can influence the weatherability of coatings.

8.2.5

Damage specific to two-coat metallic coatings

Possible damage of the clearcoat of two-coat metallic coatings takes the form of gloss reduction, yellowing and cracking (Figure 8.29). As the refractive index of the clearcoat binder (n ≈ 1.5) differs significantly from that of air (n ≈ 1.0) and at the same time the metallic pigments of the basecoat reflect light, light can be reflected repeatedly inside the clearcoat, especially in the case of light metallic shades. In other words, a clearcoat over a (light) metallic basecoat is exposed to a higher intensity of light than – for example – a clearcoat over a basecoat having a black solid-colour shade. In the clearcoat/basecoat boundary layer, osmosis may be induced by weathering, leading to blistering and ultimately delamination (Figures 8.30 and 8.31). In coloured pigmented basecoats (especially light shades), colour changes (fading) are possible. Dark shades have the appearance of being more stable because the photochemical degradation of the organic coloured pigments is not visible. Primer surfacers can contain small quantities of non-weatherstable (aromatic) epoxy resins. Thus, especially if the basecoat is not totally opaque, photochemical degradation processes may occur in the basecoat/primer-surfacer boundary layer. This can be regarded as chalking of the primer surfacer and may lead to delamination of the basecoat. Here, it should be noted that the hiding power of pearlescent pigments is significantly lower than that of metallic pigments. A brief review of the development of one-component clearcoats for two-coat metallic coatings is presented below and reveals the improvements made in weatherability (Table 8.4). The thermosetting clear­ coats (stoving enamels) of the first generation in the 1970s were based on alkyd/melamine resin. This binder system of one-coat solid colour topcoats was regarded as relatively weatherstable. However, the stabilising effect of the pigments had been overlooked (Chapter 8.3.1). Accordingly, Figure 8.28: Fracture in a plywood sheet coated with cracks formed after a relatively short a very thin layer of a clearcoat based on oxidatively time (Table 8.4). Acrylic resins were curing alkyd resin; scanning electron micrograph

147

Protection of coatings from weathering known to be weatherable, but their gloss was poor at that time. Therefore, for a transitional period, a mixture of alkyd and acrylic resins was used (with added weatherstable melamine resin, of course). However, the weatherability, though improved, was still inadequate (Table 8.4). The break-through came with the development of glossy acrylic resins for clearcoats. A further improvement was the addition of light stabilisers. Table 8.4 shows that weathering tests in Florida and in a Weather-O-Meter trend in the same direction but at different rates. Table 8.4 greatly simplifies the conditions; for example, an overly high styrene content in acrylates reduces their weatherability, especially with regard to crack formation. For such tests, as shown in Table 8.4, the clearcoat needs to be of uniform thickness, because the time to crack formation increases with increase in thickness. In polished coatings, cracking occurs prematurely in the direction of polishing. The weatherability of automotive coatings and especially of two-coat metallic coatings is generally very high nowadays. Damage of the kind seen in Figures 8.29 and 8.30 is seldom in Germany, for example. The photograph in Figure 8.30 was taken on the Croatian island of Brač; it is common knowledge that the solar radiation there is more intense and weathering damage more prevalent.

8.3 Stabilisation of coatings against photooxidation 8.3.1 Pigmentation An old rule of thumb says that pigmented coatings have twice to four times the lifetime of clearcoats made with the same binder. Iron oxides are very good stabilising pigments, including the transparent types used in transparent wood preservers. Mention should also

Figure 8.29: Hairline cracks in an automotive clearcoat (macroscopic image)

148

Figure 8.30: Blistering and delamination of an automotive clearcoat (macroscopic image)

Stabilisation of coatings against photooxidation be made of carbon black, which is Table 8.5:  Chemical composition of carbon black effective even when added in small (limiting values) Element wt% amounts. On one hand, it acts as a Carbon 80 to 99.5 non-postoxidised UV absorber by absorbing UV, VIS carbon blacks: C > 95 wt % and IR radiation. On the other, it serHydrogen 0.3 to 1.3 ves as a free-radical scavenger. This property stems from the functional Oxygen 0.5 to 15 (!) postoxidation increases oxygen content up to 15 wt % groups on its surface (compare sterically hindered phenol derivatives Nitrogen 0.1 to 0.7 as anti-skinning agents, Chapter Sulfur 0.1 to 0.7 5.1.1). Figure 8.32 schematically shows the oxygen functional groups on carbon black surfaces. Table 8.5 Table 8.6:  Degradation rate in weathered coatings shows that post-oxidised carbon based on long oil soya alkyd resin Coating Degradation rate [μm/year] black, especially, has a high content Clearcoat 10 of oxygen. Table 8.6 presents the degrada15 vol % TiO2 3 tion rate in weathered coatings ba1.5 15 vol % Fe2O3 sed on long oil soya alkyd resin. It can be clearly seen that pigmentation (especially with iron oxide) improves weatherability. Figure 8.33 shows three different types of pigmentation. It should be remembered that glossy coatings far below the critical pigment volume concentration (CPVC) form a kind of clearcoat layer on the surface that is responsible for the gloss. This clearcoat layer is not protected by pigments and degrades first during weathering. Loss of gloss is therefore commonly the first type of weathering damage. Pigment 1 in Figure 8.33 is composed of isometric particles (e.g. titanium dioxide or iron oxide); here, the incident light is attenuated by scattering and absorption. Pigments 2 and 3 are lamellar (anisometric) particles, e.g. aluminium pigments (Figure 1.50 in Chapter 1.4.5.1) and micaceous iron oxide (Figure 8.34). These protect the coating by reflection (and, in the case of coloured or micaceous iron oxide, additionally by absorption). Figure 8.31: Extensive clearcoat delamination

149

Protection of coatings from weathering Leafing aluminium pigments (Chapter 1.4.5.1), especially, seal the coating’s surface, protecting the bulk against incident light (Figure 8.33, pigment 3). At a depth of 30 μm, a coating pigmented with 15 vol.% titanium dioxide will have absorbed only 30 to 40 % of the energy of incident UV-A radiation. Possible consequences are chalking of non-lightfast primers followed by delamination after exposure to moisture. More advantageous with respect to light protection are lamellar pigments (e.g. aluminium, micaceous iron oxide) which reflect light (Figure 8.33). Amine-cured epoxy coatings [2] have poor lightfastness because of their aromatic character. In heavy-duty corrosion protection, epoxy resins are widely used on account of their good corrosion protective action; here, epoxy topcoats are often pigmented with micaceous iron oxide (Figure 8.34) to improve the poor weatherability of the binder. Non-pigmented clearcoats do a poor job of protecting the substrate (e.g. wood) against light, as only the inherent absorption of the binder comes into play (Figure 8.35). The inherent absorption of the binder is governed by the Lambert-Beer law: Figure 8.32: Oxygen functional groups on carbon black surfaces (schematic diagram) E = lg(I0/I) = ε · c · d

Figure 8.33: Influence of pigmentation on light stability

150

Stabilisation of coatings against photooxidation where E is the extinction I0 is the incident light intensity I is the residual light intensity after passage through the coating (Figure 8.35) ε  is the extinction coefficient [measure of the inherent absorption of the binder; ε = f(λ)] c is the concentration (e.g. chromophore groups in the binder) d is the thickness of the layer through which the light is passing The Lambert-Beer law applies only to substances molecularly dissolved in the binder, such as UV absorbers (see Chapter 8.3.2.1 following).

8.3.2 Light stabilisers 8.3.2.1 UV absorbers

UV absorbers are light stabilisers. They absorb (harmful) UV radiation and transform it into (harmless) thermal energy (Figure 8.36). According to the Lambert-Beer law (see Chapter 8.3.1 above), the effectiveness of a UV absorber increases with increase in its concentration c, its inherent absorption ε = f(λ) and its film thickness d. This means that a UV absorber protects the substrate and perhaps deeper layers of the coating but not the surface of the coating. On account of their chemical structures, there are many different types of UV absorbers on the raw materials market [9]; a selection is presented in Figure 8.37.

Figure 8.34: Scanning electron micrograph of micaceous iron oxide

Figure 8.35: Passage of a light beam through a clearcoat film

151

Protection of coatings from weathering Note that all four different chemical structures have intramolecular hydrogen bonds which, as will be seen in the following, are key to their effectiveness.

Figure 8.36: Basic mode of action of UV absorbers

Mode of action of UV absorbers based on benzotriazole The mode of action of UV absorbers has been thoroughly investigated using benzotriazoles [10]. As an aid to understanding this mechanism, proton transfer in hydroxyphenyl benzotriazoles is explained first (Figure 8.38); proton transfer can occur readily for steric reasons, due to the presence of the intramolecular hydrogen bond.

Figure 8.37: Different types of organic UV absorbers (simplified diagram)

Figure 8.38: Proton transfer in hydroxyphenyl benzotriazoles

152

Stabilisation of coatings against photooxidation In the ground state S0, there is no proton transfer (Figure 8.38). Upon absorption of light, the benzotriazole is promoted to the electronically excited state S1 with this transition causing significant changes in the acid and base properties. The aromatic hydroxyl group of the benzotriazole becomes more acidic (7 pK units) while the basicity of the aza-nitrogen atom increases in the same magnitude (pK units are logarithms). In other words, in the S1 state, proton transfer as shown in Figure 8.38 takes place quickly: S1  S’1. The molecule S’1, formed by proton transfer in the excited state reverts non-radiatively (internal conversions: IC) to the ground state S’0 (Figure 8.39) whereupon proton transfer to S0 ensues. The term “non-radiative” refers to light radiation: the energy is emitted as (harmless) IR/thermal radiation (Figure 8.39).

Applications of organic UV absorbers in coatings technology

One of the most important applications of UV absorbers is clearcoats for two-coat automotive metallics (see Chapter 8.2.5). The requirements imposed on industrially useful UV absorbers fall into two types, namely primary and secondary [9]. Primary requirements are: – high efficiency (high extinction coefficient) – broad absorption band in the UV-B and UV-A radiation ranges – photochemical stability. Secondary requirements concern applications in coatings technology: – Good solubility in paint solvents; the best UV absorbers are liquid and non-crystallisable. In winter, when temperatures are below freezing solid, crystallisable UV absorbers can crystallise out of a solvent-borne clearcoat composition. – For coating powders, however, UV absorbers must be solid and should melt at the standard extrusion temperatures.

Figure 8.39: Mode of action of UV absorbers (simplified model for benzotriazoles)

153

Protection of coatings from weathering – – – –

Good compatibility with paint resins in the liquid as well as in the crosslinked state. Easy incorporation into water-borne systems. Thermal stability (low volatility) in thermosetting clearcoats. No interaction with other reactive paint components. It should be noted that hyroxyphenyl-benzotriazoles are potential chelating agents; in individual cases, problems may arise with certain metals or metal compounds (catalysts). – Resistance to extraction: On one hand, UV absorbers should be readily incorporated into water-borne paints but, on the other, they should not be water-soluble, as otherwise they may be extracted by rainwater over time. All these secondary requirements are controlled by the R groups in the formulae of Figure 8.37. Figure 8.40 shows a specific commercial product of a UV absorber based on benzotriazole which was used in the clearcoat formulation in Chapter 5.4 (Table 5-3). This UV absorber is not a homogeneous substance but rather the product of the reaction between benzotriazole and polyethylene glycol (Figure 8.40), which involves transesterification and the release of methanol; on account of the bifunctionality of polyethylene glycol, transesterification is not a uniform process. The inhomogeneous product of the reaction is liquid, and it cannot crystallise. This is very advantageous for this application. The second advantage of this UV absorber derives from the hydrophilicity of polyethylene glycol: It is insoluble in water but can be incorporated into water-borne systems (see Table 5-3). It is thus a light stabiliser for both solvent-borne and water-borne systems. Particulate inorganic UV absorbers Carbon black and transparent iron oxide were mentioned in Chapter 8.3.1.

Figure 8.40: UV absorber based on benzotriazole (“Tinuvin” 1130)

154

Stabilisation of coatings against photooxidation Organic UV absorbers can suffer from the disadvantage of reduced long-term photochemical stability. Inorganic compounds, from titanium dioxide to cerium dioxide to zinc oxide, also absorb UV radiation, but have the advantage that the oxides are stable over the long term. However, one drawback is reduced film transparency, which stems from their particulate structure (haze induced by light scattering). The undesirable effect of haze can be minimised by reducing the particle size to the nanometre scale. The smaller the particle size, the less light scattering occurs and the greater is the transparency. In addition to particle size, the refractive index of the particles plays an important role. The smaller the difference in the refractive indices of the binder matrix (about 1.5) and the solid particles dispersed therein, the less light scattering and haze there is and the greater is the transparency. Zinc oxide (refractive index 2.0) is much better than titanium dioxide (refractive index 2.7). Furthermore, titanium dioxide is photoreactive (see Chapter 8.2.1 Chalking). Such additives are offered as aqueous nanoparticle dispersions for wood varnishes (e.g. “Nanobyk” 3840).

8.3.2.2

Free-radical scavengers

UV absorbers do not protect the surface of a coating; so, a second type of light stabiliser is needed that acts on the coating’s surface. These are free-radical scavengers based on sterically hindered amines (hindered amine light stabilisers: HALS; Figure 8.41). Photooxidation causes HALS to form a nitroxide free-radical which serves as the active agent (Figure 8.41) [9]. The mode of action of this nitroxide free-radical is presented in Figure 8.42. Ultimately, HALS catalyse chain termination reactions, because polymer free-radicals are transformed into non-free-radical compounds. Thus, the mode of action of HALS is similar to (but not identical with) that of antioxidants based on sterically hindered phenols (anti-skinning agents, Chapter 5.1.1). Like HALS, sterically hindered phenols

Figure 8.41: Free-radical scavengers based on sterically hindered amines (HALS)

155

Protection of coatings from weathering terminate free-radical chains but, unlike HALS, they are consumed and thus are not catalysts. Furthermore, as phenols absorb UV radiation and as they have a tendency to turn dark, they do not make suitable light stabilisers. In contrast to UV absorbers, HALS act both on the surface and in the bulk of the coating. The UV absorber prevents damage at depth, whereas HALS act only if free-radicals are formed in the coating by photooxidation. As HALS are piperidine derivatives, they are alkaline, a fact which gives rise to blocking of acid catalysts in acid-catalysed paints. Certain commercial HALS products are available which have electron-withdrawing substituents at the nitrogen atom (e.g. acetyl); these products are not alkaline (Figure 8.43) and can be used in acid-catalysed paints. A specific example is the

Figure 8.42: Simplified diagram of the mode of action of HALS (R: polymeric groups)

Figure 8.43: Basicity of HALS (R3)

156

Stabilisation of coatings against photooxidation water-borne clearcoat formulation in Chapter 5.4 (Table 5-3), where a HALS with -OC8H17 at the nitrogen atom was used (Figure 5.24) and does not block the acid catalyst permanently. Clearcoats for two-coat effect paints contain a combination of UV absorber and HALS. It has been shown experimentally that, after stoving (curing), both light stabilisers are uniformly distributed in the base and the clearcoat by diffusion [11]. Practical example: aircraft coating An important requirement for aircraft exterior topcoats is resistance to UV radiation because UV radiation is much more intense at cruising altitudes (of up to 12,000 m) [12]. Table 8.7 shows a sample formulation for an exterior aircraft coating. This pigmented coating contains a UV absorber and HALS, a combination which is actually only employed in clearcoats. An acrylate/polyester mixture cured with an aliphatic polyisocyanate serves as a light-resistant two-component polyurethane binder system [2]. As an exercise, the characteristic values of the formulation will now be calculated [2]. For a given formulation, the NCO/OH ratio is calculated as follows. wt% polyisocyanate ∙ 17 ∙ %NCO

______________________      ​ =​ ​NCO/OH ratio =  ​      wt% polyol ∙ 42 ∙ %OH

= (15 · 17 · 23) : [(6.4 · 42 · 3.3) + (15.2 · 42 · 7.4)] ≈ 1; i.e. stoichiometric crosslinking Pigment/binder ratio: P/B = 26.7 : (6.4 + 15.2 + 15) = 26.7 : 36.6 = 0.7 : 1 Pigment volume concentration [2]: PVC = (26.7/4.1) : [26.7/4.1 + (6.4 + 15.2 + 15)/1.2] · 100 % = 17.6 %

Figure 8.44: HALS from Table 8.7

157

Protection of coatings from weathering Table 8.7:  Two-component polyurethane exterior topcoat (solvent-borne) for aircraft Base component Item 01

Raw material OH-acrylate (74 % BA)

Parts by weight

Solids

8.7

6.4

02

polyester resin (80 % BA)

19

15.2

03

light stabiliser HALS

1.0

1.0

04

UV-absorber (50 %)

2.0

1.0

05

wetting and dispersing agent

0.8

06

titanium dioxide rutile

26.7

26.7

07

organoclay (10 %)

3.8

0.38

08

butyl acetate

15.8

09

methoxypropyl acetate

6.7

10

levelling agent

0.2

11

tin catalyst (1 %)

0.3

Total

85

50.7

15

15

100

5.7

Hardener 12

HDI-polyisocyanate (100 %)

Total

01:  acrylic resin with hydroxyl groups, 74 % in butyl acetate, 3.3 % OH on solids, “Setalux” 1905 BA-74 (Nuplex) 02:  polyester resin, 80 % in butyl acetate, 7.4 % OH on solids, “Synthoester” 1773 (Synthopol) 03:  “Tinuvin” 292 (BASF) see Figure 8.44 04:  “Tinuvin” 384-2 (BASF) see Figure 8.45 05:  “Disperbyk” 110 07:  “Bentone” 38 (Elementis), rheology modifier 10:  polyethermodified silicone, “Byk” 301 11:  butyltin carboxylate, “Tinstab” BL 277 (Akcros) 12:  “Tolonate” HDT-LV (Perstorp) 23 % NCO

Both the UV absorber (Figure 8.45) and the HALS (Figure 8.44) are liquid; the UV absorber is liquid probably on account of it relatively long octyl side chain. The HALS product is a liquid because it is deliberately a mixture of two compounds (Figure 8.44). Figure 8.45: UV absorber from Table 8.7

158

Literature

8.4 Literature [1] [2] [3] [4] [5] [6] [7] [8]

U. Schulz, Accelerated Testing, Vincentz Network, Hanover 2008 B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed. 2017 http://en.wikipedia.org/wiki/Painted churches of northern Moldavia R. Gächter, H. Müller, Taschenbuch der Kunststoff-Additive, Hanser Fachbuch­ verlag, 3rd ed. (1989) G. Kämpf, W. Papenrot, R. Holm, Farbe & Lack 78 (1972) p. 606 ff H. Völz, G. Kämpf, A. Klaeren, Farbe & Lack 82 (1976) p. 805 ff and 86 (1980) p. 1047 ff E. B. Faulkner, R. J. Schwartz (eds.), High Performance Pigments, Wiley-VCH, Weinheim, 2nd ed. (2009) V. E. Schmid, Farbe & Lack, 98 (1992) p. 330 ff

[9] A. Valet, Lichtschutzmittel für Lacke, Vincentz Network, Hanover (1996) [10] H. Kramer, Farbe & Lack 92 (1986) p. 919 ff [11] H. Böhnke, E. Hess, Farbe & Lack 95 (1989) p. 715 ff [12] S. Schröder in H. Kittel †, H.-J. Streitberger, Lehrbuch der Lacke und Beschichtungen – Vol. 6: Anwendungen von Lacken und sonstigen Beschichtungen, 2nd ed., S. Hirzel Verlag, Stuttgart (2008) p. 313-322 [13] U. Christ et al., Farbe & Lack 115 (2009) No. 11, p. 30–35 [14] Ph. Hofmann, R. Hofmann, Student Project Work, University of Applied Sciences, Esslingen (2009) [15] http://en.wikipedia.org/wiki/Lignin

159

UV-curing coatings

9 Photoinitiators/ UV initiators 9.1 UV-curing coatings Binders containing polymerisable double bonds can be cured by UV radiation. Curing mostly occurs by free-radical (co-)polymerisation. As free-radical reactions are by far the most common reactions during UV curing, they are explained in Chapter 9. During chain initiation, UV radiation intervenes as a “clean reagent” to generate the necessary initiator free-radicals with the help of photoinitiators (UV initiators) [1, 2].

Figure 9.1: Typical reactive diluents in UV-curing coatings

Figure 9.2: 2-Hydroxy-2-methyl-1-phenyl-1-propanone as an example of an a-hydroxyalkylphenyl ketone photoinitiator of type (a)

Figure 9.3: Benzophenone as an example of a type (b) photoinitiator Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

161

Photoinitiators/UV initiators The most important binders for solvent-free UV-curing coatings are: – Unsaturated polyester resins (UP) dissolved in styrene (reactive diluent) – Prepolymers (oligomers) bearing acrylate end groups (which are liquid because of the low molar masses of 500 to 1,500) – Low-molecular, low-viscosity mono-, di-, tri- and tetra-acrylates (as reactive diluents), Figure 9.1 Free-radical polymerisation of monomers, oligomers and polymers containing one double bond yield linear polymers, while those with more than one double bond yield crosslinked products. Typical applications of UV-curing systems are wood coatings [3] (furniture and floor coatings), plastic-film coatings and paper coatings (including printing inks). The advantages of UV-curing coatings are extremely high rates of cure (!), cure at ambient temperature and the absence of volatile organic compounds (solvents).

Figure 9.4: Possible synthesis of a urethane-acrylate secondary dispersion (binder for water-borne UV-curing coatings) The sequence of addition (1) and (2) may be changed to ensure complete reaction of the hydroxyethyl acrylate so as to minimise free monomer.

162

Mode of action of photoinitiators

9.2 Mode of action of photoinitiators Binders or reactive diluents such as monomers, oligomers and polymers commonly used for UV-curing coatings are poor absorbers of long-wave UV radiation; thus, so-called photoinitiators (UV initiators) must be added. By irradiation with long-wave UV photoinitiators generates the free-radicals necessary for chain initiation in two different ways: a) By intramolecular homolytic scission of bond in an excited state (e.g. α-hydroxyalkyl­ phenyl ketones, Figure 9.2). A bond is cleaved in an excited state if the absorbed radiation energy exceeds the binding energy of the weakest bond in the molecule. b) By intermolecular hydrogen abstraction in the T1 state (e.g. benzophenone, Figure 9.3). Here, co-initiators may be necessary; i.e. compounds which readily release a hydrogen atom (e.g. tertiary amines). Only in the relatively long-lived T1 state (see Jablonski diagram in Figure 8.3), bimolecular reactions are also possible. Other types of photoinitators are commercially available [2, 4]. Water-borne UV-curing coatings UV-curing, water-borne coatings [5] require complete evaporation of water prior to curing, a fact which increases the time required and/or increases energy consumption. Moreover, UV-curing water-borne coatings have numerous advantages: + can be produced without reactive diluents or monomers, + little film shrinkage due to the absence of monomers, + adjustment of viscosity with water and/or rheology modifiers, + depending on the binder, physical drying before curing (repair possible), + application equipment can be cleaned with water, + reduction in fire and explosion hazards. Furthermore, in the case of wood, low-molecular, reactive paint ingredients are unable to penetrate into the pores (see Figure 3.2 in Chapter 3) which are shadow zones; there, the

Figure 9.5: Photoinitiator (in Table 9.1): 2,4,6-Trimethylbenzoyl-diphenyl-phosphine oxide

163

Photoinitiators/UV initiators Table 9.1:  Guide formulation for a water-borne UV-curing primer for wood Raw material

Parts by weight

Solids 11.7

1. Formulation of the mill base Urethane-acrylate dispersion

30.0

Desalinated water

5.0

Wetting and dispersing agent

1.0

AMP 90

0.2

HEUR-additive (10 %)

0.5

Levelling and degasing agent

0.05

Talcum

8.0

8.0

Titanium dioxide rutile

5.0

5.0

Urethane-acrylate dispersion

38.6

15.1

Substrate wetting agent

0.2

Defoamer

0.2

2. Completion

Desalinated water

10.25

UV-initiator

1.0

Total

100

39.8

Urethane-acrylate dispersion: “Bayhyrol” UV VP LS 2280, 39 % in water (compare Figure 9.4) Wetting and dispersing agent: “Borchi Gen” SN 95 (Borchers) AMP 90: 2-Amino-2-methyl-1-propanol, neutralising agent HEUR-additive: “Borchi Gel” PW 25 (Borchers), 25 % in water/propylene glycol = 4 : 6 “Talc IT” extra: filler, lamellar silicate, density = 2.8 g/cm3 Titanium dioxide rutile, white pigment, density = 4.1 g/cm3 Levelling and degasing agent: “Baysilone” paint additive 3468 (Borchers), polyether-modified methylpolysiloxane Substrate wetting agent: “Baysilone” paint additive 3739 (Borchers), polyether-modified methylpolysiloxane “Defoamer T” (Borchers): defoamer based on tri-n-butyl-phosphate UV-initiator: “Darocur” 4265 (BASF) 1:1-mixture from 2-hydroxy-2-methyl-1-phenyl-propanone-1 (Figure 9.2) and 2,4,6-trimethylbenzoyl-diphenyl-phosphine oxide (Figure 9.5)

reactive paint ingredients cure incompletely, if at all, in UV radiation, and an unpleasant odour is generated. Only water penetrates into the pores of wood; an oligomeric binder (Figure 9.4) stays on its surface. For illustration, Table 9.1 shows a guide formulation for a water-borne, physically drying, UV-curing primer for wood [6]. The liquid photoinitiator used in the formulation (Table 9.1) is only slightly water-soluble but can be incorporated into water-borne paints; it dissolves in the disperse paint resin (urethane-acrylate). The freshly applied coating (Table 9.1) is dried physically for 5 minutes

164

Literature at ambient temperature and then for 10 minutes at 60 °C. Thereafter, curing is carried out with two lamps, each rated at 80 W/cm, and a conveyor speed of 2.5 m/min [6]. As an exercise, the pigment/binder ratio and the PVC of this coating (Table 9.1) are calculated below (the additives can be more or less neglected):

P/B = (8 + 5) : (11.7 : 15.1) = 13 : 26.8 ≈ 0.5 : 1 PVC = [(8/2.8 + 5/4.1) : (8/2.8 + 5/4.1 + 26.8/1.2)] · 100 % = 15.4 %

Monoacyl phosphine oxides (example in Figure 9.5) are another major group of type (a) photoinitiators; they utilise intramolecular homolytic scission of a bond in an excited state.

9.3 Literature [1] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed. 2017, Chapter IV.2 [2] P. Glöckner et al., Radiation Curing, Vincentz Network (2008) [3] H.-H. Bankowsky, P. Enenkel, M. Lokai, K. Menzel, Farbe & Lack 105 (2000) No. 10, p. 50 ff

[4] J.-P. Fouassier, Europ. Coat. Journ. No. 6 (1996) p. 412–419 [5] W. Reich, K. Menzel, W. Schrof, Farbe & Lack 104 (1999) No. 12, p. 73 [6] Guide formulation RR 3.12, Borchers GmbH (2004)

165

Film formation by primary dispersions (latices)

10 Coalescing agents 10.1 Film formation by primary dispersions (latices) A well-known and naturally occurring polymer dispersion is rubber latex [1], which is an aqueous dispersion of natural rubber (high-molecular cis-1,4-polyisoprene; particle size up to 1.5 μm) stabilised by proteins acting as protective colloids. The bark of rubber trees is scored, and the milky turbid latex is collected in small pots (Figure 10.1). The latex is coagulated in platters (a thermodynamically favoured process; see below), e.g. with formic acid, and the water is squeezed out. The term latex (plural latices) is also often used instead of primary dispersion. In coatings technology, primary dispersions (latices) are polymer dis­persions which are produced by emulsion polymerisation; during emulsion polymerisation the liquid monomers react to a solid polymer. Primary dispersions (latices) have a discrete disperse phase and a sharp interface boundary between the disperse phase (polymer) and the dispersion medium (water). As a rule, primary dispersions are thermodynamically unstable with regard to coagulation. They therefore have to be stabilised electrostatically or sterically; i.e. transferred into a thermodynamically metastable state [2]. The term emulsion needs to be explained here (especially in contrast to dispersion). Emulsions are two-phase systems with a liquid dis- Figure 10.1: Tapping latex from a rubber tree in perse phase (e.g. binder) and water Southern India Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

167

Coalescing agents as the dispersion medium. By physical drying only, binder emulsions form permanently tacky films (like the coatings on flycatchers). Solid coatings can be formed by binder emulsions only via curing (chemical crosslinking reactions). The here used primary dispersions (latices), however, are two-phase systems with a solid disperse polymer phase and water as the dispersion medium (see above). By physical drying only, dispersions form solid films – in contrast to emulsions. Important types of primary dispersions in coatings technology include pure acrylic dispersions (copolymers of various esters of acrylic and methacrylic acid) and styrene-acrylic dispersions (copolymers of styrene and esters of acrylic acid) [3]. The backbone of these macromolecules, commonly produced by free-radical copolymerisation, contains only unsaponifiable C-C bonds while the (meth)acrylate side chains contain ester groups which are difficult to saponify. This is the basis of the high stability of coatings made therefrom. Primary dispersions are often used as binders in physically drying coatings for the protection of buildings: for high-pigmented latex paints (e.g. façade coatings) and for low-pigmented latex gloss enamels (e.g. wood coatings) [2]. As the water evaporates, the polymer particles approach one another more closely until spherical packing takes place (Figure 10.2). Further evaporation of water causes capillary forces of the very thin water lamellas to deform spherical polymer particles, yielding

Figure 10.2: Film formation by primary dispersions

168

Figure 10.3: Transmission electron micrograph of a contrasted thinsection of a film obtained from a pure acrylic primary dispersion

Film formation by primary dispersions (latices) rhombic dodecahedra. These fuse together at the particle boundaries (coalescence). Interdiffusion of polymer chains across the particle boundaries may occur. Primary dispersions are thermodynamically unstable with regard to coagulation (see also rubber latex above); i.e. film formation leads irreversibly to a thermodynamically stable state. Film formation by primary dispersions occurs by physical drying and the individual polymer particles are still identifiable in the coating (Figure 10.3). Figure 10.3 shows a transmission electron micrograph of a contrasted thin-section of a film obtained from a pure acrylic primary dispersion; the two-dimensional diagram of the rhombic dodecahedra looks like honeycomb (Figure 10.3). Coatings of primary dispersions are non-crosslinked and thus are thermoplastic and swellable by solvents [2]. For a deeper understanding of film formation, the optical properties of primary dispersions will now be explained. When light interacts with particles which are smaller than its wavelength λ, the light will be scattered: this is known as Rayleigh scattering (opalescence, Tyndall effect). According to Rayleigh, the scattering coefficient is proportional to λ-4; i.e. short-wave, blue light is scattered ten times as strongly as longwave, red light. This explains why fine (colourless) polymer dispersions have a bluish shimmer when observed from the side (Figure 10.4) but are reddish in transmitted light (Figure 10.5). The bluish shimmer is particularly evident in the lower

Figure 10.4: Spatula-applied film of a colourless primary dispersion on a glass plate, as observed from the side (bluish shimmer) immediately after application

Figure 10.5: Film of a colourless primary dispersion on a glass plate in transmitted light (from the sky) immediately after application (left); after one hour’s drying (right), the film has already become clear at the edge (red arrow)

169

Coalescing agents picture of Figure 10.4; the bluish reflection colour is clearly visible against the black background on the hiding power chart. Conversely, if this “colour effect” is observed, the particle size of the polymer dispersion is smaller than the wavelength of visible light. Dispersed particles larger than the wavelength of visible light also scatter light (Mie scattering); such dispersions are colourless and milky (see rubber latex in Figure 10.1); coarse polymer dispersions have particle sizes up to 5 μm. When evaluating the appearance of primary dispersions, it should be remembered that disperse systems mostly have more or less broad particle size distributions. A few large particles at the limits of the particle size distribution can significantly increase the turbidity of the dispersion (arrow in Figure 10.6). Moreover, dispersions may appear clear and transparent if there is no significant difference between the refractive indices of the dispersion medium (water) and the disperse phase (polymer). During film formation, the polymer-water phase boundary disappears, and no more light is scattered; the film becomes clear and transparent (Figure 10.7). The series of colour pictures in Figures 10.4, 10.5 and 10.7 shows the same film of a colourless styrene-acrylate

Figure 10.6: Schematic diagram of the particle size distributions of two primary dispersions of the same mean particle size but narrow and broad particle size distribution; the arrow indicates the (few) coarse particles at the upper limit of the broad particle size distribution.

170

Figure 10.7: Completely dry and thus clear film of a primary dispersion in transmitted light (from the sky); because the layer is thick and uneven, there are cracks and air bubbles trapped in the film

Mode of action of coalescing agents Table 10.1:  Examples of important coalescing agents and cosolvents Substance

Evaporation number* Solubility in water (RT)

1. Coalescing agents 2,2,4-Trimethyl-1,3-pentanediol

5000

0.05 wt. %

Butyl glycol acetate

137

1.5 wt. %

Butyl diglycol acetate

3000

6.5 wt. %

Butyl glycol

119

unlimited

Butyl diglycol

1200

unlimited

1000

unlimited

1-isobutyrate (Figure 10.8)

2. Cosolvents

3. Others 1,2-Propylene glycol**

* Diethyl ether = 1 ** 1,2- Propylene glycol is neither a classic cosolvent nor a coalescing agent. In latex paints, it often serves as a water-retention agent; i.e. it prolongs the “open time”. It also has antifreeze properties.

primary dispersion (particle size 0.1 μm) at different viewing angles and at different times (because of thick and uneven application with a spatula, the film contains some cracks and trapped air bubbles).

10.2 Mode of action of coalescing agents Film formation (coalescence) of primary dispersion particles only takes place above the minimum film forming temperature (MFFT). Below the MFFT, the polymer is in the glassy state and unable to form a film. The MFFT of primary dispersions for coatings is often between -5 and 25 °C; in special applications (latex gloss enamels for industrial uses), the MFFT can be much higher. A low MFFT means that the polymer is soft, and this can lead to poor mechanical film properties. Mostly, the MFFT is slightly lower than the glass temperature Tg of the dispersed polymers [3]. Primary dispersions having a MFFT above 5 °C require external plasticisation with added permanent plasticisers (e.g. phthalates); even Figure 10.8: 2,2,4-Trimethyl-1,3-pentanediol-1better is the addition of coalescing isobutyrate (“Texanol”)

171

Coalescing agents agents (low-volatility organic solvents) to act as temporary plasticisers. In contrast to permanent plasticisers, coalescing agents leave the film after a certain time and do not confer lasting plasticisation. The requirements for coalescing agents are: – optimal dissolving or swelling of the polymer particles – no influence on the stability of the polymer dispersion – minimal solubility in water; i.e. dissolution in the polymer – fast evaporation after film formation – odour-free (very important nowadays for indoor applications) – environmental compatibility The difference between cosolvents [2] and coalescing agents is fuzzy (Table 10.1). For example, butyl glycol can act as both cosolvent and coalescing agent. White spirit is totally immiscible with water, but it can act as coalescing agent, alone or in combination with other solvents (e.g. 1:1 mixture with “Texanol”). It should be noted that coalescing agents and cosolvents can affect the viscosity: – by swelling polymer particles – by interacting with rheology modifiers, especially HEUR additives (compare Chapter 4.1.3.2). Coalescing agents are added to latex paints and latex gloss enamels in levels of about 2 wt.%.

10.3 Literature [1] http://en.wikipedia.org/wiki/Latex [2] B. Müller, U. Poth †, Coatings Formulation, Vincentz Network, Hanover, 3rd ed. 2017

172

[3] R. Baumstark, M. Schwartz, Waterborne Acrylates for Decorative Coatings, Vincentz Network 2001

Neutralising agents for binders bearing carboxyl groups

11 Neutralising agents Generally, a distinction has to be drawn between binders bearing carboxyl groups which need bases for neutralisation and binders bearing amine groups which need acids for neutralisation.

11.1 Neutralising agents for binders bearing carboxyl groups Systems bearing carboxyl groups are the most important type of water-borne binders. As a rule, amines serve as the neutralising agents in this case (Figure 11.1). The amines have to volatilise during drying or curing in order that the hydrophilicity of the coatings may be reduced (Figure 11.2). Thermal dissociation of the ammonium salts formed can take place even at ambient temperature if the neutralising amine is volatile at room temperature (e.g. ammonia).

Figure 11.1:  Examples of amines important in coatings technology AMP = 2-amino-2-methyl-1-propanol, AEPD = 2-amino-2-ethyl-propanediol, DMEA = dimethylethanolamine (skull-and-crossbones pictogram), TEA = triethylamine (skull-and-crossbones pictogram) Bodo Müller: Understanding Additives, 2nd Revised Edition © Copyright 2019 by Vincentz Network, Hanover, Germany

173

Neutralising agents For stoving enamels, DMEA has been the most important neutralising agent. Nowadays, DMEA is labelled with the skull-and-crossbones pictogram; where possible, therefore, DMEA should be replaced in water-borne stoving enamels by the less toxic AMP. For stoving enamels, the toxicologically more favourable AEPD and the solid and therefore odour-free diisopropanolamine may prove to be less harmful alternatives (Figure 11.1). It is recommended that the solid diisopropanolamine be dissolved in water (50 wt  %) before use. The slow-evaporating and therefore low-odour butyldiethanolamine may also be considered because, additionally, it serves as emulsifier and corrosion inhibitor [2]; here, again, this is a case of a multifunctional additive. For paints that dry or cure at ambient temperature, the more volatile amines ammonia and TEA have been used; but nowadays TEA is also labelled with the skull-and-crossbones picFigure 11.2:  Neutralisation reaction of binders togram and so should not be used bearing carboxyl groups with an amine and thermal anymore. In oxidatively curing wadissociation of the ammonium salt formed ter-borne paints, ammonia should serve as neutralising agent. Neutralisation of water-borne two-component polyurethane paints is described in Chapter 5.2 (see Figure 5.13). Many amines have a strong and unpleasant odour. For latex paints, non-volatile and odourless potassium hydroxide (KOH) may be used. Recently, the use of saccharides bearing amino groups has been proposed, because they are renewable raw materials [1]. Various amines also influence the solubility of paint resins to different extents. In this case, their basicity is of lesser importance than their solubilising properties. In the Figure 11.3:  Simplified diagram of the change in series DMEA > TEA > ammonia, the viscosity of a neutralised water-borne paint resin during dilution with water solubility of paint resins decreases

174

Neutralising agents for binders bearing amino groups or the turbidity of the (colloidal) aqueous resin solution increases, because DMEA (amino alcohol) also acts as a cosolvent. The amount of amine (mass: mamine; molar mass: Mamine) for complete neutralisation of a paint resin (mass: mresin) can be calculated from the acid number (AN, mg KOH/g) of the paint resin: mamine = (Mamine · SZ · mresin) : 56100 It should be noted that the amines may adsorb on the surfaces of pigments and fillers (especially on post-oxidised acidic carbon blacks). In the case of pigmented paints, the amount of amine for adjusting the pH to 8 may be much higher than calculated. When water-borne paints are being produced, the pH value should be checked at every step, in order that unpleasant surprises, such as coagulation of the paint resin, may be avoided. When neutralised paint resins containing carboxyl groups (e.g. AK, SP, AY) are diluted with water, abnormal viscosity behaviour is often observed [3, 4]. This viscosity anomaly is caused by two opposing effects (Figure 11.3): – The diluting effect of water lowers the viscosity. – Water is a poor solvent for organic paint resins (exceptions are some melamine and urea resins). Thus, the solvent power of the medium decreases upon dilution with water; this leads to increasing association of resin molecules (similar to the case for secondary dispersions) and a rise in viscosity (“viscosity peak”). The judicious choice of neutralising agent can minimise this viscosity anomaly (“viscosity peak”); amino alcohols are preferred here.

11.2 Neutralising agents for binders bearing amino groups Cathodic electrodeposition primers contain aromatic epoxy resins which are well-known for their good anticorrosion protective properties. Epoxy resins are suitable for an addition reaction with amines (Figure 11.4); this addition reaction also produces the hydroxyl groups necessary for the later crosslinking reaction. Such amine-modified epoxy resins can be neutralised with organic acids, yielding a water-dilutable binder (Figure 11.5). These amine-modified epoxy resins form a kind of base binder for the other water-insoluble formulation ingredients, such as crosslinkers (commonly blocked polyisocyanates), pigments and fillers. The neutralised amine-modified epoxy resins form aqueous colloidal particles which incorporate the aforementioned water-insoluble formulation ingredients (Figure 11.6) and transport them to the cathode during subsequent electrophoresis. The preferred neutralising agents are acetic acid, lactic acid and formic acid. Furthermore, due to the low volatility of organic acids compared with that of amines (Chapter 11.1),

175

Neutralising agents the stoving temperatures are relatively high: 165 to 180 °C for cathodic electrodeposition primers. More details on cathodic electrodeposition primers may be found in reference [5].

Figure 11.4:  Addition of a secondary amine to an aromatic epoxy resin to produce an aminemodified epoxy resin as binder for cathodic electrodeposition primers (simplified)

Figure 11.5:  Simplified presentation of the neutralisation of an amine-modified epoxy resin with acetic acid

Figure 11.6:  Colloidal particle of an aminemodified epoxy resin and the other water-insoluble formulation ingredients of a cathodic electrodeposition primer (highly simplified)

11.3 Literature [1] S. Ziebold, J. Rüger, Farbe & Lack, 123, No. 3 (2017) p. 90–95 [2] B. Müller, P. Kienitz, Farbe & Lack, 101, No. 11 (1995) p. 919–921 [3] J. Dörffel, Farbe & Lack, 81 (1975) p. 10–15

176

[4] H. J. Luthardt, Farbe & Lack, 87 (1981) p. 456–460 [5] B. Müller, U. Poth †, Coatings Formulation, 3rd ed., Vincentz Network (2017) Chapter III.4.4.3

Author

Author Prof Dr rer. nat. Bodo Müller, born 1954, was first employed in industry in the development of water-borne metallic coatings and later adhesives and sealants. Since 1990, he has been professor for coatings technology and adhesives and sealants at the Esslingen/Germany University of Applied Sciences from 1996 to 2017. There, he was Head of the Chemical Engineering/Colour-Coating-Environment course. He has been retired since 2017 and has published over 100 papers and three books to date.

177

Index

Index Chemical names 2-amino-2-methyl-1-propanol (AMP)  104, 109, 174 2-mercaptobenzothiazolyl succinic acid (MBTS)  128

A absorption of light  134 acetic acid  175 acid catalyst  108 acrylate  162 acrylic dispersion  168 addition reaction  99 adhesion  113 adhesion promoter  117 adhesive layer  119 adsorption  17 agglomerate  11 aggregate  11 aircraft exterior topcoat  157 alkanolamine  36 aluminium pigment  42, 129 anatase  141 anticorrosive pigment  125 anti-skinning agent  97 ASE/alkali soluble/swellable emulsion  73 association  77 associative rheology modifier  73 atmospheric corrosion  123 auto-catalysis  105 automotive coating  90, 144, 148 autophoresis  119 autoxidation  94, 136

B basecoat (metallic)  42 Bénard cell  24, 49, 54 benzoin  64 benzotriazole  152 benzyl alcohol  107

BHT  98 binder degradation  141 binder emulsion  30 blistering  147 block copolymer  21 blocked acid  109 blocked isocyanate  99, 104, 175

C capillary force  168 carbamate formation  107 carbon black  149 catalysts (accelerator)  93 cathodic delamination  124 cathodic electrodeposition primer  175 cellulose ethers  73, 80 chain scissions  144 chalking  140 chemiphoresis  119 chemisorption  128 chromophore  134, 139 clearcoat  147, 153 coalescence  171 coalescing agent  172 coating powder  63 cobalt  95, 96 controlled flocculation  65 conversion layer  114 copolymer  20 corrosion  123 corrosion inhibitor  126 corrosion protection  113, 126 corrosion protection additive  126 corrosion reaction  43 corrosion stimulator  125 cosolvent  172 coupling agent  116 crack formation (cracking)  144, 148 cratering  48, 52, 55, 117 crosslinker  144

179

Author

D

G

DABCO  103 DBTL  104 deaerator  59 deagglomeration  12 defoamer  59 delamination  145, 147, 150 diisopropanolamine  109, 174 dispersing  12 dispersing agent  18 dispersing process  11 DMEA 174 drier (siccative)  94 dyes  143

gasvolumetry  131 glass temperature  171 grinding  11 grinding process  42

E electrophoresis 175 electrostatically stabilised dispersion  72 electrostatic stabilisation  17 embrittlement  139, 144 emulsion  30, 167 emulsion polymerisation  167 entropy  18 epoxy resin  105

F fading  143 fatty acids  42 film formation  169 film shrinkage  163 flash off  90 flocculation  14 fluorescence  135 fluorinated surfactant  129 foam  55 foam lamellae  56, 59 free-radical chain reaction  139, 142 free-radical polymerisation  162, 168 free-radical scavenger  149, 155 fumed silica  37, 72, 85

180

H hammer finish  47, 51 HASE/hydrophobically modified anionic soluble emulsion  73 HEUR/hydrophobically modified ethylene oxide urethane rheology modifiers  73 high-solids paints  90 hindered amine light stabilisers\HALS  155 HMMM  107, 110 house-of-cards structure  70 hydrogen corrosion  44 hydroperoxide  94, 139 hydroxyethyl cellulose (HEC)  80

I infrared radiation (IR)  135 intentional chalking  143 interface reaction  43, 123

J Jablonski diagram  134

L Lambert-Beer law  150 lamellar (anisometric) particle  149 lamellar metal pigment  40 latex gloss enamel  168 latex paint  168 layer silicate  68 leafing  41 leafing aluminium pigment  150 lecithin  28 levelling  64, 68 levelling agent  47 Lewis acid  102 Lewis base  102

Index light scattering  155 light stabiliser  111, 151 lignin  145 lubricant  41

M macrofoam  58 Mannich base  106 melamine  107 metal flake  41 metallic basecoat  147 metallic pigment  39, 147 metal soap  95 methylalkyl polysiloxane  51 microfoam  58 mild steel  123 mineral oil defoamer  61 minimum film forming temperature (MFFT)  171 multi-metal drier  97

N nanoparticle  155 neutralising agent  104, 173 Newtonian flow  65 non-leafing  42 nonylphenol ethoxylate (APEO)  27

O organic colorant  143 organoclay  83, 85 organo-functional trialkoxysilane  115 oxidative curing  94, 98 oximes  97 oxirane  105 oxygen  133, 142

P pearlescent pigment  147 peptising agent  71 phase boundary  27 phenol derivative, sterically hindered  98 phosphorescence  135

photocatalyst  143 photocatalytic oxidation cycle  141 photochemical degradation  137 photochemical stability  155 photoinitiator (UV initiator)  163 photooxidation  136 photooxidative degradation  134 physical drying  168, 169 pigment separation  23, 49 pigment-specific additive  33 pinhole  55 plasticisation  171 plasticiser  171 point of zero charge (IEP)  17 polyanion  19 polycarboxylate  19 polydimethyl siloxane  61 polydimethyl siloxane (silicone oil)  50 polyester-polysiloxane  52 polyether-polysiloxane  51 polyisocyanate  99 polymer dispersion  167 poly(meth)acrylic acid  119 polymethylhydrosiloxane  36 polyphosphate  19 polypropylene glycol  71 polyurethane coating  99 popping  90 powder coating  63 prepolymer (oligomer)  162 primary dispersion  167, 168 primary particle  11 primer surfacer  147 pseudoplasticity  65

R radical  95, 135 Rayleigh scattering  169 reactive diluent  64, 162 reactive melamine resin  110 reactive surfactant  29 refractive index  170 reinforcing filler  38 rheologically active additive  65 rheology modifier  65 rubber latex  167 rutile  141

181

Index

S

U

sagging  90 Schulze-Hardy law  16 secondary dispersion  162 sedimentation  67 self-cleaning  143 sensitisation  135 sensitiser  139 shear-thinning  80 silane adhesion promoter  115 silane  36 silanol group  85 silicone additive  49 silicone  48 slip additive  47 sol-gel transition  70 solvent evaporation  58, 90 stabilisation  15 sterically hindered amine  111 steric stabilisation  18 Stokes’ law  58, 66 stoving enamel  91, 99, 104, 107, 147 styrene-acrylic dispersion  168 surface area  56 surface tension  15, 25, 48, 56 surface treatment of pigment  31

ultraviolet radiation (UV)  133, 135 unsaturated polyester resin (UP)  162 urea derivative  87 urethane-acrylate  162 UV absorber  151, 153, 154 UV absorber, inorganic  155 UV absorber, organic  155 UV-curing  162 UV degradation  136 UV initiator  161

T thermosetting  63, 90, 107 thickener  65 thin layer  119 thixotropy  66 three-dimensional network (gel)  66 titanium dioxide  140, 141 two-coat metallic coating  147, 148 two-component epoxy  120 two-component polyurethane  157 two-component  99, 105

182

V van der Waals force  16 viscosity  65 visible light (VIS)  135

W water-borne clearcoat  110, 157 water-borne metallic paint  129 water-borne paint  43, 130 water-retention agent  171 weathering  133, 138 wetting  50 wetting agent (surfactant)  19, 24, 47, 55 wood  145 wood coating  145, 162

X xanthan gum  82

Z zinc-rich primer  130