UV powder coatings are making strong inroads into the powder coatings market, especially for temperature-sensitive substrates. The existing UV-curing powder coatings systems generally build on two binders. This paper presents an option to the existing systems and consists of using unsaturated amorphous or amorphous-crystalline polyesters comprising maleic/fumaric moieties as reactive double bonds in combination with crystalline crosslinkers containing unsaturated reactive groups. The study further presents a comparison of the crosslinking performance for Zn diacrylate, hexahydro-1,3,5-tris(1-oxo-2-propenyl)-1,3,5-triazine (1,3,5-triacryloil-hexahydro-triazine) Tradename M-1530) and unsaturated derivatives from 3,9-bis(2-hydroxy-1,1-dimethyl)-2,4,8,10-tetraoxispiro[5,5]undecane (penta-spiroglycol) suitable for this application.

The latest development in powder coatings is the combination of powder coating technology and UV technology. This approach shows some very important benefits. The benefits due to powder coating technology versus liquid coatings are:

• Elimination or reduction of intermediate sanding steps;

• High performance at high utilization rate, reclaim leading to 95% powder usage;

• No VOC during the curing process; and

• Easy recycling and no waste treatment.

New benefits due to the combined technology of UV and powder coating are:

• Split between melting by IR or convection and the curing process;

• Application of complete board in one step in vertical line;

• Better and undisturbed levelling by separation of the flow and cure phases;

• No cooling of the coated object required;

• Good storage stability below 50 ºC;

• Good adhesion on different substrates; and

• Design with good ODD.

State of the Art

Although there are a few limitations, UV-curable powder coatings have made progress, and several options are now available in the marketplace. These include:

• Maleic/fumaric polyester combined with a vinyl ester;

• Acryl/methacryl functional polyesters;

• Unsaturated polyester combined with acrylic functional polyurethane;

• Acrylic functional polyacrylic combined with acrylic functional polyurethane; and

• Maleic/fumaric polyester combined with allyl functional polyesters.

All these UV-curing systems have limited applications. Some of them have a very narrow range of the ratio between components.

Crosslinkers for UV Powder Coatings

The most suitable approach for a coating formulation is the use of a major binder and a crosslinker. The cross¬linker may control the network density for the coating, while the binder determines properties of the coating such as discoloration, outdoor stability, mechanical properties, etc. Furthermore, this approach will lead to a more homogenous concept in the powder coatings applications as a category bringing similitude to thermosetting coatings where crosslinkers such as TGIC and -hydroxyl amides are used (Table 1). A crosslinker should present properties quite specific for the application intended: molecular weight; high functionality; and physical properties compatible to the application.

Crosslinkers for UV-coating powders must somehow match the properties of the above-mentioned crosslinkers used for thermosetting powder coatings. We have synthesized and studied the crosslinking performance of three acrylic and one allyl derivative, suitable for UV curing powder coatings.

The structures (Table 2) chosen for study were:

• pentaspiroglycol diacrylate;

• hexahydro-1,3,5-tris(1-oxo-2-propenyl)-1,3,5-triazine (1,3,5-triacryloil-hexahydro-triazine);

• zinc acrylate; and

• pentaspiroglycol diallylether.

Experimental Pentaspiroglycol Diacrylate

PSG diacrylate (CAS number: 85286-82-4) has been synthesized from pentaspiroglycol (purity 98.5%) obtained by re-crystallization of crude pentaspiroglycol, which resulted from the reaction of pentaerythritol and hydroxyl-pivalic aldehyde.

In a 2-liter, four-neck round-bottom flask, 0.60 mols of pentaspiroglycol was reacted with 9 mols of ethyl acrylate in the presence of 265 ml n-heptane, phenothiazine 500 ppm, nitrobenzene 1,000 ppm, and 1% tert-butyl titanate (Tyzor TBT). The reaction was performed under reflux removing a part of the distillate at a ratio of 5:1 until the vapour temperature rose to 85 ºC. The synthesis blend was cooled down to 5 ºC and the spiroglycol diacrylate isolated by filtration. The raw product was re-crystallized from a 20% solution in ethanol, washed with n-heptane and dried overnight at 60 ºC. The purity of the prepared product was 98% and the yield was 77%.

Pentaspiroglycol Diallylether

PSG diallylether (no CAS number yet) was synthesized for the first time. The pentaspiroglycol (purity 98.5%) was obtained by re-crystallization of crude pentaspiroglycol, which resulted from the reaction of pentaerythritol and hydroxyl-pivalic aldehyde.

Using a 3-neck round-bottom flask, 0.05 mol pentaspiroglycol was added to 152 g freshly distilled DMF and heated to 40 ºC. Then 0.05 mol NaH was added to the solution, and hydrogen gas starts to evolve from solution. After 5 minutes reaction time, 0.1 mol allyl chloride was added. The temperature was raised to 60 ºC in order to complete the reaction. The reaction blend was then cooled down to 40 ºC, and 0.08 mol NaH was further added at 40 ºC in small portions to control the exothermic reaction. After 5 minutes, 0.07 mols allyl chloride was further added at 53 ºC and the temperature increased to 60 ºC. After 30 minutes at 60 ºC, an additional quantity of 0.16 mol allyl chloride was added. After 6.5 hours, the reaction was completed and NaCl had precipitated. 100 g water was added to the reaction blend and then the pH adjusted to 6.5 with 0.3 g 10% HCl solution. The organic phase was further extracted twice with xylene, washed with water and dried at 50 ºC under vacuum. The obtained crystals are then re-crystallized twice from warm methanol. Pentaspiroglycol diallyl ether, having a purity of 97% and a melting point of 68 ºC, has been obtained at a yield of 72%.

Zinc Acrylate

Zn acrylate (CAS number 14643-87-9) has been prepared by the reaction between acrylic acid (3 mols) in 216 g water and Zn oxide (0.5 mols) in the presence of 4-metoxy phenol 200 mg and strong air bubbling. The reaction blend was submitted to vacuum distillation (10 mm Hg) at 40 ºC. Small amounts of toluene were added and removed by vacuum in order to completely eliminate all water from the product. 114.5 g product containing approximately 31% Zn was obtained.

Evaluation of Crosslinking Properties

A polyester binder was produced using the raw materials shown in Table 3 and having physical properties also noted in the table.

The evaluation of the crosslinking performance was made on blends at different ratios (by mol) of double bound C=C in polyester versus acrylic unsaturation in the crosslinker. The performance was monitored by UV-FTIR, UV-Rheology and UV-DSC.

UV-DSC Test Method

UV-DSC was performed in a Mettler DSC. About 15 mg of powder sample, in an open, 40-µl DSC crucible, was placed in a Mettler DSC with DSC20S HVP with a glass cover instead of the normal Ag lid The oven was purged with 10 ml/min N2 under the UV operation. The sample was placed in a 140 °C isothermal oven and the DSC heat flow monitored for 12 minutes. After the 5 minutes necessary to attain thermal equilibrium, the UV lamp was switched on for 5 minutes (the last 2 minutes without UV light).

UV irradiation was at 365nm using a handheld UV lamp UVGL25 above the cell in the Mettler DSC. The lamp was 52 mm above the crucible. Measurements at this distance show a UV-A intensity of 1.3 mW/cm2.

After the isothermal UV curing was completed, the sample was removed and cooled down to +20 °C and a normal DSC test with deep Ag lid was performed at 10 °C/min from -10 to +150 °C to determine the Tg after curing.

UV-DSC Results

UV-DSC test results are shown in Table 4 and Figure 1 for various powder mixtures. The higher reported enthalpy for blend F than for blend E is the effect of difficulties in drawing the base line.

UV-FTIR Test Method

The samples were melted on a KBr disk placed on a 140 °C heated plate. After melting, another KBr disk was set above it, so the sample formed a film between the two KBr discs. The KBr disk was placed in a heatable KBr holder and positioned in the FTIR instrument together with a handheld UV Lamp UVGL25. The holder was heated to 140 °C isothermal and then the UV lamp turned on. Several FTIR spectra were recorded up to 8' UV. UV-A irradiation at 365 nm on the sample is about 1.2 mW/cm2.

UV-FTIR Results

No significant changes could be detected in the UV-FTIR curves from Blend A (reference) and B.

Blend C has peaks at 809 cm-1 (Figure 2) and 1635 cm-1 (Figure 3) that decrease with UV irradiation. The 809 peak is probably from the PSG diacrylate. The 1635 peak is probably from the C=C in the polyester.

Blend D has peaks at 828 cm-1 (Figure 4) and 1646 cm-1 (Figure 5) that decrease with UV irradiation. The 828 peak probably comes from the Zn acrylate. The 1646 is probably from the carbon-carbon double bond in the polyester.

UV-Rheology Test Method

To follow the rheology changes during the curing, 0.2g of sample was placed in a Stressech rheometer with 22 mm Plate/Plate measuring system held at 140 °C and gap 0.5 mm. The upper plate was made from a 0.1-mm-thick diameter 22 mm glass disk glued with cyanoacrylate glue to a metal rod that fits the Stresstech. UV hits the sample at 45° from a UVGL25 UV lamp at a 15 cm distance.

The Stresstech was set at oscillation with stress=500 Pa. Delay time for each measurement was set from standard 1" to 0" to minimize time for each measurement. Measurements were done and the measurement interval set to 3". This UV-rheology method is newly developed for this test.

UV-Rheology Results

UV rheology data is shown in Table 5 for various samples.

Conclusions

The PSG-diacrylate is rated as having the highest UV-DSC reactivity in powder coatings formulations, followed by Zn-acrylate and (lowest reactivity) hexahydro-1,3,5-tris(1-oxo-2-propenyl)-1,3,5-triazine. The PSG diallyl ether gives no crosslinking at all.

The high reactivity with the PSG-diacrylate can also be seen in the UV-rheology tests. The onset time is little longer than for hexahydro-1,3,5-tris(1-oxo-2-propenyl)-1,3,5-triazine but the network is more crosslinked. That can be seen from the lower phase angle and high storage modulus G'. The Zn-acrylate is slower to crosslink than hexahydro-1,3,5-tris(1-oxo-2-propenyl)-1,3,5-triazine in the start but shows both lower phase angle and higher storage modulus G' at the end of the measurement.

UV-FTIR gives rather small changes during the curing.

Coatings using a crystalline acrylic ester may be easily formulated using as a starting point any unsaturated polyester of maleic/fumaric type. It has been shown that addition of a small quantity of PSG-diacrylate may improve the performance in commercial systems of lower reactivity. The achieved level of the crosslinked systems in terms of double MEK rubs and Tg of cured film strongly recommend this approach as a reliable option in UV powder coating formulations. The approach makes possible the use of typical unsaturated polyesters as binders in powder coatings formulation.

Acknowledgments

The authors would like to thank S. Lundmark (Perstorp),Stavros Apostolatos (Interchem), Ronan Van Rjisbergen (Oxyplast), Ralph Nussbaum (IKEA), Carmen Fourar (Protech) for their help, and also the European Community for financial support of this project (Growth project LOTEC GRD 1-2000-25541).

Bibliography

>

1. Ericsson, J; Nilsson, M; Lundmark, S; Svensson, L; Jönsson, S; Lindgren, K. Synthesis and photoinduced copolymerization of novel maleimides, combined with vinyl and/or styrene ethers. In Radtech 2000, Technical proceedings, Baltimore, MD, US, 2000, p.173-195.
2. Subramanian, R; Sullivan, C.J. Unsaturated polyesters in low temperature cure powder coatings. In Powder Coatings Europe 2002, Congress Papers, Nürnberg 2002, p 231-239.
3. Research disclosure 463047, November 2002.
4. Matsumoto, A. Polymerisation of multiallyl monomers. Prog Polym Sci., 2001, 26, 189-257.
5. Schildknecht, C.E. Allyl compounds and their polymers. Wiley Interscience, 1973.
6. Johansson, M; Hult, A. Influence of allyl ethers in coating resins. J. of Polymer Chemistry, 1991, 29, 1639-1644.
7. Pohl, H. U; Traenckner, H-J; Grünewald, H. Analytishe untersuchungen an photochemish vernetzbaren allylethermodifizierten Fumarsäureestern, Die Angewaldte Makromolekulare chemie 98, 1981 1-34 (nr.1516)
8. Rehnberg, N ; Harden A; Lundmark, S; Manea, A; Svensson, L. Allyl ethers in the Thiol-Ene reaction, ACS, Chicago, Polymer preprints, 42(2) August 26-31, 2001, 701-702
9. Knapczyk, Jerome W: Polyallyl glycidyl ether resins for very fast curing high performance coatings, The 65th Annual Meeting of the Federation of Societies for Coatings Technology, 1987, Dallas, TX.
10. Nishijima, T; Hata, S: Allyl ester oligomer and its applications, Showa Denko KK, November 2001.
11. Burget, D Mallen C, Fouassier, J-P, Manea, M, Lundmark, S. Kinetics and Photochemistry of visible light induced polymerization of allyl ether. Radtech 2003, Berlin, ECJ, 12/ 2003.

This paper was presented at the 8th Nürnberg Congress, Creative Advances in Coatings Technology, April 2005 in Nürnberg, Germany. The Congress is sponsored by FPL, PRA and the Vincentz Network.