The incorporation of pigments and extenders was carried out by milling using a dissolver (fitted as vertical pearl mill) and zirconium dioxide beads 1.6-2.5 mm in diameter. The lacquers were hand-applied using a 100 ?m device (i.e., ca. 50 ?m dry film thickness) on fiberboard sheets covered either with a white melamine coating or with ash veneer sealed twice with a UV-curing transparent filler. The surfaces were sanded shortly before application. For comparison, all samples were tested on glass panels and some were also tested on steel.

For the most part, the films were cured at a belt speed of 5 or 10 m/min using a combination of two electrode powered 80 W/cm medium pressure mercury bulbs (one Gallium-doped and one standard radiator). For comparison, two 120 W/cm microwave powered lamps (one Iron-doped and one standard radiator, elliptical reflectors) were also used. Similar results were obtained by the latter curing device but they are not discussed in the following text.

Surface hardness was checked by simple test procedures such as empirical evaluation of scratch and solvent resistance (ethyl acetate/butyl acetate 1:1 and ethanol/water 1:1). Through-curing of the coating was evaluated by its adhesion on the substrate. An adhesive tape was used to assess the adhesion. Before sticking on the tape a strip of the cured film was removed by scratching with a spatula (this procedure proved to be more suitable than the cross-hatch test). Additionally, pendulum hardness on glass or the cupping test on steel were used for selected formulations.

Coloristic investigations were carried out with samples cured on glass over black and over white backgrounds. A spectrophotometer "MA 68" (X-Rite; only 0?/45? geometry results evaluated) was used to determine the degree of jetness MY and the blue-dependent degree of jetness MC.

Introductory Investigations and Formulation Principles

• A low-viscosity trifunctional polyether acrylate without amine modification,
  
• Very reactive amine-modified polyether acrylates (not used in all formulations),
  
• A flexible polyester acrylate,
  
• An aromatic epoxy acrylate (type with lower viscosity),
  
• Limited amounts of an aliphatic hexafunctional urethane acrylate, and, partially,
  
• Flexibilizing aliphatic bifunctional urethane acrylates.

Through-curing behavior was not only dependent on formulation, thickness, and radiation dose but also on the substrate reflectance. This is explained by UV absorption by pigments and photoinitiators, and by scattering and back reflection from the pigment particles.

The extent to which the substrate absorbs residual radiation, or reflects it back into the lower layers of the coating has a significant influence on the curing in these lower layers. This correlates well with results from coating materials containing high levels of white pigments.1,2,4,6,7 This effect sometimes allows an opaque black-pigmented film to have better curing performance on a white or pale substrate than on a dark (or black) substrate.

On the other hand, it is also possible that a UV-curable formulation that gives sufficiently cured coatings with good flexibility on poorly reflecting wood substrates appears too hard and somewhat brittle when applied and cured on glass or metal substrates with higher reflectance. Furthermore, scratch and solvent resistance of the surface had to be taken into consideration.

For application on substrates with medium or poor reflectance, it was necessary to find a compromise formulation between sufficient opacity at the required film thickness and through-cure sufficient for adherence on the substrate.

Black pigment concentrations had to be minimized as far as possible without compromising opacity. Use of extenders improved opacity of formulations containing minimum amounts of black pigment, without affecting the through-cure performance.

A reduction in belt speed (i.e., increased radiation dose) can result in somewhat improved through-curing. Formulations with different photoinitiator blends containing both short-wave absorbing and long-wave absorbing (i.e., also absorbing in the near UV and visible regions) compounds were cured using the combination of a Gallium-doped and a non-doped 80 W/cm medium pressure mercury bulb (10 m/min). However, during the introductory investigations it was difficult to distinguish between their curing properties.

On the other hand, the materials clearly performed better than a comparable formulation containing only a short-wave absorbing photoinitiator (2-hydroxy-2-methyl-l-phenylpropane-l-one; cured at only 5 m/min using the single non-doped radiator). This is remarkable in so far as carbon black does absorb visible light as well as near and far UV radiation.

Formulation Improvements

During further experiments the formulations were improved step by step. Table 1 gives an overview of the most important compositions. A liquid solution of 25% bis-(2,6-dimethoxybenzoyl)-2,4,4-trimethylpentyl phosphine oxide in 75% 2-hydroxy-2-methyl-1-phenyl-propane-1-one was used as photoinitiator.

Two different types of fiberboard sheets were used during investigations. They were covered either with an industrially applied white hard coating or with ash veneer and a self-applied transparent UV cured filler. Frequently, poor adhesion had been found on these substrates. There are two possible explanations for these adhesion problems, which are opposite to each other and, hence, would require two different strategies for improvement:

1. Through-curing and crosslinking density are somewhat reduced in the lower layers of films which are applied and cured on poorly reflecting substrate. Up to a certain level this will give no problems with respect to adhesion. The cured films, however, are both flexible and hard (but not brittle in contrast to the curing result of the same enamel on substrates with higher reflectance).

However, if the crosslinking density at the lower surface becomes too low, insufficient adhesion results. Moreover, liquid residues at the bottom side and wrinkled surfaces can occur if acrylate conversion becomes even worse.

To solve this problem it is, at first, necessary to adjust pigment type and amount and to use long-wave absorbing photoinitiators with better through-curing. However, if further improvement is required a higher acrylate functionality of the formulation becomes necessary (formulations no. 34 and 36, see Table 1: use of the hexafunctional urethane acrylate, of the trifunctional polyether acrylate and of trimethylolpropane triacrylate). This gives higher reactivity with respect to chaingrowing reaction, and, hence the potential to crosslink a greater section beginning from a single generated starting radical. Compared with the film surface, this effect is substantially more important near the substrate where (due to the presence of light-absorbing pigments) the number of starting radicals is low.7

2. When a coating is cured sufficiently its adhesion properties are influenced by mechanical parameters. In particular a certain flexibility is required for good adhesion.

When the starting point formulations give coatings that are too brittle, the use of acrylate oligomers and monomers with lower functionality and higher flexibility will lead to improved substrate adhesion (formulation nos. 30A-F; use of lower functional monomers instead of polyether triacrylate and trimethylolpropane triacrylate).

Formulations nos. 34 and 36 clearly showed superior adhesion. In contrast, the use of the lower-functionality reactive thinners in no. 30A-F gave no improvement but in some cases led to deterioration.

Oxygen Inhibition Can Lead to Poor Scratch Resistance

With respect to scratch resistance no. 34 and no. 36 performed equally or better compared to nos. 33 and 35. Obviously, the possibility to crosslink a greater section beginning from starting radicals generated in a deeper region of the film also gives the possibility of "reverse through-curing" from bottom to top and, thus, contributes to improved surface resistance. Therefore, nos. 34 and 36 were used as comparative systems for further changes in formulation.

Good Results when Extenders Are Added

In contrast to both black pigments mentioned above, the standard carbon black "FW 200" did not prove to be suitable for use in UV-curable wood coatings. From the opacity point of view it was not significantly better than "Special Black 100". However, through-curing and substrate adhesion were insufficient even at pigment concentrations below the opacity limit (no. 22, nos. 67-70. Note: This limitation is dependent on the individual system. It is valid for the systems investigated here but, for example, not for coatings on more reflecting substrates. In UV-curable metal coatings, in principle, "FW 200" can also be used with sufficient through-curing).

However, the results mentioned above were only obtained in the presence of 10 weight percent of various extenders. No significant differences were found between these fillers, especially barium sulphate and calcium carbonate (In contrast to that, some talc-containing formulations prepared during the pre-investigations had shown inferior opacity compared to the other extenders).

Possible explanations for this improved opacity in the presence of extenders are:

• The extender particles cause particle-size dependent light scattering effects and, therefore, longer pathways of the light beams through the film, i.e., a higher pigment efficiency.
  
• The presence of extender makes the milling process more intensive and, thus, improves pigment distribution.
  
• The distance between the black pigment particles becomes larger and more constant.
  
• The density of extenders is considerably higher than that of the binder. This results in an increase in overall density of the coating material and, therefore, an increase in the pigment volume concentration of carbon black or aniline black.

The degree of jetness MY and the blue-dependent degree of jetness MC allow an evaluation of color strength and opacity (Table 2). The addition of extender leads to a decrease in jetness, which is in contrast to the improved opacity.

However, the use of Aniline Black "BS 890" gives superior jetness compared to "Special Black 100". It became evident that this improvement is considerably more significant than the effect of the extender.

Attempts to Find an Optimized Photoinitiator System

1 part isopropyl thioxanthone + 4 parts 2-benzyl-2-N,N-dimethylamino-1-(4-morpholinophenyl)-butane-1-one. This combination is known to be very effective in dark formulations (e.g., printing inks). However, it proved to be less efficient in the wood coating material tested.

Other combinations of long-wave absorbing phosphine oxides and short-wave absorbing hydroxyketones:

• 1 part 2,4,6-trimethylbenzoyldiphenylphosphine oxide combined with 1 part 2-hydroxy-2-methyl-1-phenyl-propane-1-one (liquid solution)
  
• 1 part bis-(2,6-dimethoxybenzoyl)-2,4,4-trimethylpentyl phosphine oxide combined with 1 part or 3 parts 1-hydroxycyclohexylphenyl ketone (solid mixtures)
  
• 1 part phenyl-bis-(2,4,6-trimethylbenzoyl)-phenylphosphine oxide (solid) combined with 1 part or with 3 parts of the hydroxyketones (liquid or solid)
  
• phenyl-bis-(2,4,6-trimethylbenzoyl)-phenylphosphine oxide without hydroxy-ketone

Generally, the liquid photoinitiator blend mentioned above (used in nos. 34 and 36) seemed to be the most efficient type within the formulations tested. However, it should be taken into consideration that compositions with very low monomer contents were used, which may not have been suitable for achieving an optimum solubility and effectiveness of solid photoinitiators.

Conclusion

On wooden substrates the latter could be achieved both with "Special Black 100" and with Aniline Black "BS 890" whereas the formulations with carbon black "FW 200" failed. The films obtained with aniline black were significantly more black and brilliant than those with "Special Black 100". However, a sufficient opacity required the tenfold pigment percentage (by weight); this also had influence on gloss and viscosity.

The hiding power of black coating materials with a pigmentation level low enough to ensure through-cure even on poorly reflecting substrates could be improved by extenders (~ 10 wt. % barium sulphate, calcium carbonate, within certain limitations also talc). This made it possible to achieve opaque coatings at approx. 50 ?m dry film thickness with only 0.7 wt. % "Special Black 100" or by 7.0 wt. % Aniline Black "BS 890". Without these extenders the pigment content had to be increased considerably to obtain sufficient hiding power. The insertion of these extenders caused a certain decrease in jetness but the decrease was outweighed by the advantages of using aniline black.

Acknowledgment

For more information on black pigments, contact Maria Nargiello-Tetreault, Applications Manager, Degussa Corp., Akron Technical Center, 3500 Embassy Parkway, Suite 100, Akron, OH 44333; phone 888/SILICAS; e-mail maria.nargiello-tetreault@degussa.com; or Circle Number 92.

References