How to speed up the UV curing of water-based acrylic coatings.

Water-based polyurethane-acrylate (PUA) coatings have been cured, after drying, by a short exposure to UV radiation in the presence of a radical-type photoinitiator. The light induced polymerization has been followed quantitatively by means of real-time infrared spectroscopy, by monitoring continuously the disappearance of the acrylate double bonds. The UV curing of the dry film was markedly accelerated by adding to the PUA formulation a reactive plasticizer (diacrylate monomer), by operating in a humid atmosphere or by raising the temperature. The neutralizer of the carboxylic groups, needed to get a stable dispersion, plays a key role in both the polymerization kinetics and in the hydrophilic character of the UV-cured polymer, the best performance being achieved by using a volatile tertiary amine. Water-based UV-cured PUA coatings combine hardness and flexibility and are, therefore, quite resistant to abrasion and scratching.

Keywords: Acrylics, EB UV radiation cure, FTIR-ATR, photoinitators, reaction kinetics, waterborne

**********

Water-based systems are increasingly used in the coating industry because of environmental considerations and the ever more drastic regulations regarding the emission of volatile organic compounds (VOCs). (1-3) These types of resins have raised great expectations, even in solvent-free technologies like UV radiation curing. (4-6) Most UV-curable resins are made of functionalized oligomers, which control the physicochemical properties of the final polymer, and acrylic monomers which serve to adjust the formulation viscosity. To reach the low viscosity needed for spray application, a solvent is sometimes used, thus, requiring a flash-off step before the UV exposure which results in the emission of VOCs. By using water to reduce the formulation viscosity, one avoids VOC emission, as well as the odor and irritation problems inherent to the use of acrylic reactive diluents; while at the same time reducing unwanted shrinkage, brittleness, and internal stresses caused by the high crosslink density. Water-based UV-curable coatings have been shown to combine flexibility and hardness, (7,8) thus making such materials quite resistant to both shock and scratching. When compared to conventional UV-cured coatings, water-based systems show a number of advantages, such as low viscosity applications and recycling of overspray, less oxygen inhibition, a lower amount of extractables, easier mattability, and better adhesion on various substrates, in particular wood.

The basic principle of UV radiation curing can be represented schematically as follows, for a monomer-free acrylic coating:

[FORMULA NOT REPRODUCIBLE IN ASCII]

The potential of UV-curable water-based coatings and their overall performance has been thoroughly investigated in the past decade. (7-16) They consist generally of an aqueous dispersion or emulsion of a telechelic oligomer end-capped by the very reactive acrylate double bond. After an initial drying stage to remove the water, the solid sample is crosslinked by a short exposure to intense UV radiation to make it hard and chemically resistant. The final properties of the UV-cured polymer depend primarily on the chemical structure of the oligomer chain (polyurethane, polyester, polyether), as well as its molecular weight, i.e., the crosslink density. As the oligomer is usually not soluble in water, a few carboxylic groups are incorporated into the oligomer chain to make it dispersible in alkaline aqueous solution.

In UV-curable water-based systems, photocrosslinking is performed on a dry film, i.e., in a solid material where molecular mobility is highly restricted. Consequently, the chain reaction will proceed at a slower pace and less extensively than in a conventional UV-curable liquid resin. (11) The objective of the present work was, therefore, to study the UV-curing process of some typical water-based acrylic coatings and try to speed up the crosslinking reaction through an increase of the mobility of the polymer chains in order to get a more completely cured material.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Infrared spectroscopy was used to follow quantitatively the light-induced polymerization by monitoring the disappearance of the acrylate double bonds, and thus, evaluate the amount of residual unsaturation in the UV-cured polymer. Various factors were shown to play a key role in the curing kinetics, in particular the temperature, humidity, and the presence of a reactive plasticizer. We have also examined the influence of the hydrophilic group, the neutralizer, light intensity, and atmospheric oxygen on the polymerization kinetics. Low molecular weight aliphatic polyurethane-acrylates (PUA) have been selected as telechelic oligomers because of their high impact and tensile strength, and their superior resistance to scratching, abrasion, and weathering (17, 18) which makes them well suited for coating applications.

EXPERIMENTAL

Materials

The water dispersible PUA oligomers were prepared by reacting an aliphatic diisocyanate (hexamethylene or isophorone) with a telechelic diol oligomer and a hydroxyacrylate monomer, as described in U.S. Patent 6,444,721 (2002). A small amount of dimethylolpropionic acid was added in order to introduce carboxylic acid groups in the PUA chain. By working under stoichiometric conditions, all the isocyanate and hydroxyl groups were shown by IR spectroscopy to have reacted after heating the mixture for three hours at 60[degrees]C. Two types of diol oligomers were used: polycaprolactone and cycloaliphatic. The carboxylic acid was neutralized with NaOH, or an amine, so as to convert it into the water dispersible carboxylate salt, and water was added to form a stable aqueous dispersion with a 35 wt% solid content. Thus, for example, PUA-1 was built from a caprolactone diol, isophorone diisocyanate, hydroxyethylacrylate, and dimethylolpropionic acid. For comparison, PUA-2 with the same structure as PUA-1 was dissolved in butyl acetate instead of being dispersed in water, and PUA-3, also dissolved in butyl acetate, had the same structure as PUA-1, but without the dimethylolpropionic acid content. The viscosity of the dispersion was typically on the order of 40 mPas and the size of the micelle particle was in the 50 to 100 nm range, depending on the carboxylic acid content (with an acid number of 10 to 23 mg KOH/g). A water-soluble photoinitiator (Irgacure 2959 from Ciba Specialty Chemicals) was added to the dispersion at a concentration of 0.3 to 1 wt% (1 to 3 wt% in the dry film).

Drying and UV Curing

The formulation was cast onto a barium fluoride crystal at a typical thickness of 70 [micro]m to attain, after water removal, a dry film of 25-[micro]m thickness. The drying was performed by placing the sample in an oven at 80[degrees]C for three minutes, after which time a tack-free film was attained. The residual water content was less than 1 wt%, as measured by IR spectroscopy. If necessary, a faster drying can be achieved by using infrared radiation or by casting the formulation on a hot plate.

[FIGURE 4 OMITTED]

The UV curing of the dry sample was performed by passage under a medium pressure mercury lamp (Minicure IST--power of 80 W/cm) at a speed of typically 20 m/min, which corresponds to a 0.3 sec exposure and a UV dose of 170 mJ [cm.sup.-2] per pass, as measured by radiometry (International Light IL-390). When the photopolymerization was followed in real time by IR spectroscopy, we used a less intense irradiation device equipped with an optical guide (Novacure from EFOS). The light intensity at the sample position could be varied in a large range and was typically set at 120 mW [cm.sup.-2]. Most of the experiments were performed at ambient temperature, in the presence of air. To quantify the inhibitory effect of atmospheric oxygen on such a radical-type polymerization, a special cell was designed to allow the UV irradiation, as well as the IR analysis, to be carried out in a controlled atmosphere. The cell, equipped with two Ba[F.sub.2] windows transparent to both UV and IR radiation and containing the sample, was flushed with nitrogen for 10 min and then sealed off, before being exposed to UV light.

Analysis

Upon UV radiation curing of the PUA resin, the acrylate double bond disappeared rapidly after radical attack, with the formation of a three-dimensional polymer network, represented schematically in Figure 1:

[FORMULA NOT REPRODUCIBLE IN ASCII]

Such high-speed reaction can be conveniently followed by real-time infrared (RTIR) spectroscopy. (19) An FTIR spectrophotometer (Brucker IFG-66), equipped with an MCT detector, was employed to monitor continuously the decrease, upon UV exposure of the IR band at 1410 [cm.sup.-1] of the acrylate double bond. Up to 30 spectra were recorded per second, thus allowing an accurate in situ monitoring of high-speed reactions. The 25-[micro]m thick sample was exposed simultaneously to the UV beam that induced the polymerization of the acrylate oligomer and to the IR beam which analyzed in real time the resulting chemical modifications. As the degree of polymerization is strictly proportional to the decrease of the IR absorbance, conversion versus time curves were directly recorded by following in situ the disappearance of the IR band of the acrylate double bond.

[FIGURE 5 OMITTED]

The hardness of the UV-cured polymer was evaluated on 30-[micro]m thick films coated onto glass plates by means of a Persoz pendulum, measuring the damping time required to pass from an initial oscillation angle of 12[degrees] to a final angle of 4[degrees]. Persoz values are strongly dependent on the glass transition temperature of the coating, (7) and range from 30 sec for soft elastomeric materials up to 400 sec for very hard and glassy polymers.

RESULTS AND DISCUSSION

One of the great advantages of UV-curing technology lies in the rapidity of the process which transforms a liquid resin into a solid material within less than one second under intense illumination, at ambient temperature. (5,20) The UV-curing line can, thus, be operated at high speeds, up to a few meters per second. This is not true for UV-curable water-based systems since the prerequisite drying step slows down the production line considerably.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

In the case of the polyurethane-acrylate used in this study, the sample emerging from the drying oven is already quite hard (Persoz value of 240 sec), but it is not water resistant and needs, therefore, to be crosslinked to make it resistant to organic solvents and chemicals, as well as to improve its mechanical properties. As the UV radiation curing of the dry sample occurs in the solid state, it proceeds much slower and less extensively than in a typical UV-curable liquid resin. (16) To achieve a faster and more complete polymerization, the mobility of the reactive species (free radicals and acrylate double bonds) has to be considerably increased. This can be done by increasing the temperature or by plasticizing the solid coating before UV exposure.

Influence of the UV-Curing Temperature

When the PUA coating was exposed to UV radiation (light intensity of 100 mW [cm.sup.-2]) at ambient temperature in dry air, the polymerization of the acrylate double bond was found to proceed slowly, reaching a conversion of only 22% after a 20-sec exposure (UV dose of 2 J [cm.sup.-2]). This is due to the low molecular mobility which results from the presence of hydrogen bonds between the polyurethane chains. Increasing the sample temperature up to 100[degrees]C will break these bonds and restore the chain mobility; thus allowing the polymerization to proceed both faster and more extensively. This temperature effect is clearly illustrated in Figure 2 which shows the polymerization profiles recorded by RTIR spectroscopy for different temperatures. At 100[degrees]C the curing reaction was found to proceed initially 45 times as fast as at 21[degrees]C, while the 20 sec conversion value was rising concomitantly from 22 to 80% (Figure 3). It should be noted that there is no thermal contribution to the observed polymerization process, as the acrylate double bond content was found to remain unchanged after a 20-min storage of the dry PUA sample at 100[degrees]C in the dark.

Based on these results, it is highly recommended to perform the UV irradiation on the hot sample, immediately after it emerges from the drying oven. An even more complete curing was achieved by increasing the light-intensity up to 600 mW [cm.sup.-2], a value typically found on industrial UV-curing lines. A single pass under the lamp at a web speed of 20 m/min proved to be sufficient to polymerize 80% of the acrylate double bonds when the sample temperature was set at 80[degrees]C (Figure 2). This value was found to hardly increase upon further UV exposure, probably because the polymer had already reached its glassy state. The more extensive cure achieved, at equal UV dose, by operating under intense illumination was attributed to the heat evolved by the exothermal polymerization. At high light intensity, the curing proceeded very rapidly and the heat was evolved in a short time (0.3 sec), thus, making the sample temperature rise substantially more than by operating at low light intensity. This explanation has been fully confirmed by recording in real time the temperature profiles of PUA samples undergoing ultrafast curing at different light intensities. (21,22) An important consequence of such behavior is that the irradiation conditions (sample temperature, light intensity, type of substrate), by controlling the ultimate degree of conversion and, therefore, the crosslink density, will directly govern the physico-chemical properties of the UV-cured polymer.

Influence of a Reactive Plasticizer

When water-based UV-cured coatings are used to protect heat sensitive substrates, the drying and UV curing must be performed at a moderate temperature. A possible way to speed up the polymerization is by introducing into the formulation a diacrylate monomer, which will act as a reactive plasticizer. This additive will both increase the molecular mobility and participate with the polymerization, so that it will be incorporated into the polymer network. This is an important requirement in order to prevent a softening of the cured material and avoid exudation of liquid plasticizers upon long-term use. Hexanedioldiacrylate (HDDA) was selected because of its high reactivity, but also because it is partly miscible in water and can be added directly to the aqueous dispersion. As HDDA was used at a relatively low concentration and in aqueous solution, it does not impart any odor or irritancy to the formulation.

At a weight concentration of 10% in the dry PUA film, HDDA was found to accelerate substantially the UV-curing reaction performed at 25[degrees]C, as shown in Figure 4. After a single pass at a web speed of 20 m/min, nearly 70% of the acrylate double bonds have already polymerized in the HDDA-plasticized coating, compared to only 30% in the neat PUA coating. The polymerization kinetics were followed more accurately by means of RTIR spectroscopy, and by operating at a lower light intensity (40 mW [cm.sup.-2]). The boosting effect of HDDA is clearly apparent in Figure 5, with a 10-fold increase of the initial polymerization rate in the presence of 10 wt% HDDA, for UV-curing experiments performed at ambient temperature. At 80[degrees]C, the effect is less pronounced because the HDDA-free film already possesses some chain mobility. Increasing the radiation intensity up to values typically found on an industrial UV line will not only speed up the curing process, but it will also raise the sample temperature; thus allowing a nearly complete cure to be achieved for the HDDA-plasticized PUA coating UV-irradiated at 80[degrees]C. It should be mentioned that the addition of HDDA also has the advantage of reducing the hydrophilic character of the UV cured polymer, the contact angle of a droplet of water increasing from 50 to 57[degrees]. A more pronounced effect can be obtained, if needed, by adding small amounts of a fluoroacrylate monomer to the resin before dispersion in water.

[FIGURE 10 OMITTED]

Influence of Humidity

Another simple way to increase the molecular mobility of the dried sample is by placing it in a humid environment before performing the UV curing. The water absorbed by the hydroscopic uncured PUA, which can be followed quantitatively by IR spectroscopy, (11) leads to a softening of the coating. This plasticizing effect causes a drastic increase in both the polymerization rate and the final cure extent. Figure 6 shows the polymerization profiles recorded by RTIR spectroscopy for a PUA coating UV irradiated at 25[degrees]C under different relative humidity conditions (0, 46, and 100%). Here again, a 10-fold rate increase was observed by passing from a dry to a water saturated atmosphere. A similar improvement in cure speed was achieved by working with an incompletely dried sample, as well as by performing the curing at high light intensity on a UV line (Figure 7). It can be seen that as much as 80% of the acrylate double bonds have polymerized after a one-second UV exposure at an ambient temperature under high humidity conditions. On a production line, where the curing will be performed on the sample emerging from the drying oven, it is therefore essential to control the drying stage (temperature and duration) well, as the amount of residual water will affect both the cure speed and the final properties of the crosslinked polymer.

Similar experiments have been carried out with a solventborne PUA coating with the same chemical structure, but without the carboxylic groups. It is interesting to note that essentially the same behavior was observed as with the water-based PUA coating (Figure 8), even though this sample was much less hydrophilic. It was still found to pick up water when placed in a highly humid atmosphere, before being cured, which explains the accelerating effect of humidity. By contrast, the UV curing of a liquid polyurethane-acrylate resin (Laromer 8987 from BASF) was shown not to be affected at all by the atmosphere humidity, superimposed polymerization profiles being recorded at 0 and 100% relative humidity. To make sure that the boosting effect of humidity was really due to an increase in the polymer chain mobility and not from an increase in the initiation rate, we monitored the disappearance of the photoinitiator upon UV irradiation under dry and humid conditions. Figure 9 clearly shows that the atmosphere humidity had essentially no effect on the photolysis of Irgacure 2959.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

[FIGURE 13 OMITTED]

Influence of Oxygen

The photoinitiated polymerization of acrylate resins is known to be strongly inhibited by oxygen which reacts rapidly with the initiator R* and polymer ([P*.sub.n]) radicals to generate inactive peroxy radicals:

[FORMULA NOT REPRODUCIBLE IN ASCII]

The inhibitory effect is particularly pronounced in thin liquid films which can only be cured in air under intense illumination so as to reduce as much as possible the irradiation time during which air diffuses into the film. Different methods have been proposed to overcome [O.sub.2] inhibition in UV-curable systems, the most efficient one being obviously to operate under [O.sub.2]-free conditions, i.e., in an [N.sub.2] or C[O.sub.2] atmosphere. (23) In the case of water-based coatings, oxygen inhibition was expected to be less important because of the much slower diffusion of air into a solid coating than into a liquid film. Surprisingly, the polymerization of water-based PUA coatings was found to proceed faster and more extensively in the presence of air than in a pure nitrogen atmosphere, as shown in Figure 10. Such an unusual positive effect of atmospheric oxygen on the UV curing of water-based PUA coatings was already observed by Menhert et al. (13) It was attributed to an additional formation of initiating radicals generated by decomposition of hydroperoxides:

[FORMULA NOT REPRODUCIBLE IN ASCII]

[FIGURE 14 OMITTED]

If this were true, there would be no inhibitory effect of atmospheric oxygen anymore in the UV curing of acrylate resins, which is unfortunately not the case. Actually, these hydroperoxides are unlikely to decompose significantly in a timescale of seconds at a relatively low temperature ([less than or equal to] 80[degrees]C). Their photolysis is also problematic, given their low UV absorptivity, their very low concentration, and the inner radiation filter effect of the photoinitiator. The faster cure observed by Menhert and by us in the presence of air, compared to a dry nitrogen atmosphere, was actually due to the humidity present in air and its plasticizing effect on the UV curing of the solid PUA coating. Indeed, by performing the photocrosslinking reaction under the same humidity conditions (in a dry atmosphere or under high humidity conditions), the polymerization of the acrylate double bond was always found to proceed significantly faster in the absence of oxygen (Figure 11). However, the small difference between the two polymerization profiles clearly indicates that oxygen inhibition is not a crucial issue in the UV curing of PUA coatings cast from aqueous dispersions, as expected from the much slower diffusion of oxygen into dry films. This is a distinct advantage of water-based systems, especially for thin films, which can be cured in air even at low radiation intensity, e.g., with fluorescent lamps or sunlight.

Influence of the Hydrophilic Group

A few carboxylic acid groups have to be introduced into the polyurethane-acrylate chain in order to make the telechelic oligomer dispersible in water, after neutralization. To evaluate the influence of this functional group on the cure kinetics, we followed the polymerization of the acrylate double bond for the three following PUA systems based on the same polycaprolactone-isophorone diisocyanate structure: PUA-1 is a water-based polyurethaneacrylate with carboxylate groups (dimethylolpropionic acid neutralized by NaOH), PUA-2 has the same structure as PUA-1 and was dissolved in butyl acetate instead of dispersed in water, and PUA-3, also dissolved in butyl acetate, has the same structure as PUA-2, but without the dimethylolpropionic acid group.

The formulations containing 3 wt% Irgacure 2959 were dried at 80[degrees]C for five minutes to remove water or the butyl acetate solvent. The 20-[micro]m thick dry films were then cured online (web speed of 20 m/min) at ambient temperature and a relative humidity of 36%. The polymerization profiles obtained (Figure 12) clearly show that the curing proceeds faster and more extensively when the carboxylate group is replaced by a carboxylic acid (PUA-2), and even more so when it is completely removed (PUA-3), 90% conversion being then reached at a UV dose of 3J [cm.sup.-2].

These results suggest that the molecular mobility was greater in the solventborne coatings than in the water-based coating. This was confirmed by hardness measurements which indicate that PUA-1 was much harder than PUA-3 (Persoz value of 240 and 66 sec, respectively). It was partly due to some residual solvent which was completely eliminated from the coating by a one-hour drying at 130[degrees]C (Table 1). The great difference in hardness of the uncured films was attributed to intermolecular hydrogen bonds between the polyurethane chains and the carboxylic groups in PUA-2, as well as to ionic interactions involving the carboxylate groups which increased the polarity of the medium (in PUA-1). Infrared spectroscopy analysis of the 3400 [cm.sup.-1] region revealed that hydrogen bonding was indeed more important in PUA-1 than in PUA-3, and that H-bonds disappeared upon heating at 80[degrees]C. (24) Such an explanation would also account for the fact that an increase in the curing temperature from 25[degrees] to 80[degrees]C had a more pronounced accelerating effect on the photopolymerization of water-based coatings (cleavage of H bonds) than on the solventborne coating.

It can be seen in Table 1 that the three coatings show very similar hardness after UV curing at ambient temperature (Persoz hardness of 350 sec), even though their degree of conversion is quite different. These results can be explained by considering that the light-induced hardening of the relatively hard water-based PUA-1 coating will require less acrylate double bonds to polymerize than that of the soft solventborne PUA-3 coating. Owing to their great hardness, these UV-cured polyurethane-acrylate coatings showed a superior resistance to scratching, while they still remained flexible and impact resistant because of their relatively low crosslink density (1 mol [kg.sup.-1] compared to 3 mol [kg.sup.-1] for conventional UV-cured coatings).

[FIGURE 15 OMITTED]

In Table 2, we compared the scratch and chemical resistance of a water-based PUA coating and of a typical polyurethane-acrylate coating (Laromer LR-8987) which were UV cured under the same conditions. It can be seen that the water-based PUA coating shows slightly better performance than the UV-cured PUA coating, especially regarding its flexibility.

The hydrophilic character of the UV-cured PUA coating was directly related to its acid content, as was shown by water uptake and surface energy measurements. (11) The softening of the polymer placed in a humid environment, which is fully reversible, had to be minimized by introducing on the PUA chain just the amount of carboxylate groups needed to get a stable dispersion. Another way to reduce the hydrophilic character of water-based PUA coatings is by a proper selection of the neutralizer used to disperse the functionalized oligomer in water.

[FIGURE 16 OMITTED]

Influence of the Neutralizer

To make the polyurethane-acrylate resin dispersible in water, the carboxylic acid group had to be converted into a carboxylate group by the addition of NaOH or an amine:

[FORMULA NOT REPRODUCIBLE IN ASCII]

Given the pK value (4.9) of the propionic acid used, the neutralization can be considered complete for amines having a pK value superior to 8. The volatility of the amine plays a critical role because its evaporation during the drying stage will shift the equilibrium toward the left, thus, regenerating the less hydrophilic carboxylic acid group. Such an effect was indeed observed when we used various tertiary amines having different boiling points as neutralizers (Table 3). The carboxylic acid content of the dried film, characterized by its IR band at 1580 [cm.sup.-1], was found to increase with the amine volatility. Consequently, the UV-cured coating became less hydrophilic, as shown by the increase of the contact angle of water ([[theta].sub.w]) and the decrease of the polar component of the surface energy ([[gamma].sub.p]), while the water uptake by the sample placed in a humid atmosphere decreased concomitantly.

Figure 13 shows some typical water uptake profiles obtained by following the increase of the IR band at 3500 [cm.sup.-1] of the OH group for UV-cured PUA coatings placed in a water saturated atmosphere at 25[degrees]C. It can be seen that the amine neutralized PUA samples were less hydroscopic than the NaOH neutralized sample, and nearly as good as the solventborne PUA sample. Measurements of the contact angle of water appeared to be a simple way to evaluate the degree of neutralization, which was maximum with NaOH and minimum with the volatile triethylamine. By drying the dispersion containing dimethylethanolamine (BP = 134[degrees]C) at a temperature of 130[degrees]C instead of 80[degrees]C, the [[theta].sub.w] value was found to increase from 64[degrees] to 70[degrees], i.e., the value of the solventborne PUA sample containing no carboxylic group.

[FIGURE 17 OMITTED]

In order to further reduce the hydrophilic character of the UV-cured coating, we have introduced into the PUA resin, before dispersion, a fluorinated acrylate monomer (2 wt% of FX-13 from Atochem). After drying at 130[degrees]C and UV curing, a [[theta].sub.w] value of 74[degrees] was obtained ([[gamma].sub.p] = 5.6 mJ [m.sup.-2]), making this coating the most hydrophobic of all the water-based coatings studied so far, which should be quite beneficial with respect to its resistance to outdoor weathering. Another effective way to decrease the hydrophilicity of water-based coatings was to replace, in the PUA oligomer, the poly(vinyl caprolactone) chain with a biscyclohexane structure. With an amine neutralizer, the water uptake was reduced (Figure 13) and the value of the polar component of the surface energy (6.7 mJ [m.sup.-2]) was comparable to that of the solventborne PUA sample.

The type of neutralizer used also has some effect on the polymerization kinetics, as it affects the molecular mobility of the dry film. Replacing NaOH with an amine caused a two-fold increase of the resin reactivity and a more extensive curing, the acrylate conversion after a 20 sec UV exposure at ambient temperature rising from 28 to 47%. The plasticizing effect of triethanolamine on the cure kinetics is illustrated in Figure 14, for a PUA-1 resin exposed to UV radiation (140 mW [cm.sup.-2]) in dry air and at a relative humidity of 33%. While polymerization hardly occurred in dry air at ambient temperature for the NaOH neutralized PUA, it proceeded much more effectively for the same polymer neutralized with triethanolamine, but it was then less affected by humidity upon UV curing. Chain transfer reactions promoted by the amine may have also contributed to the faster and more complete polymerization by allowing trapped polymer radicals to initiate new kinetic chains:

[FORMULA NOT REPRODUCIBLE IN ASCII]

A ten-fold increase in cure speed was achieved by raising the temperature from 25[degrees] to 80[degrees]C, the amine neutralized PUA remaining more reactive than the NaOH neutralized PUA, as shown in Figure 15. Similar polymerization profiles were recorded by RTIR spectroscopy for the three resins containing the different amine neutralizers. The accelerating effect of the temperature was found to be less pronounced for the solventborne PUA-3 coating which is softer (Persoz hardness = 66 sec) and has therefore a greater molecular mobility at ambient temperature (Figure 16). A performance analysis of the three types of PUA samples (waterbased + NaOH, water-based + amine, and solventborne), UV cured either at 25[degrees] or at 80[degrees]C, is given in Figure 17 as histograms showing their reactivity (initial polymerization rate) and the final cure extent.

[FIGURE 18 OMITTED]

These results were fully confirmed by curing experiments performed under intense illumination on an industrial UV line at ambient temperature or at 80[degrees]C, as shown in Figure 18. A faster and more complete polymerization was always achieved with the amine neutralized PUA sample, the amount of residual unsaturation dropping to a remarkably low value of 4% for the coating cured immediately after the drying stage at a UV dose of 0.5 J [cm.sup.-2]. Such curing performance is quite similar to that reached with conventional liquid UV-curable PUA resins by working at ambient temperature. (5) Besides their superior reactivity, amine neutralized water-based polyurethane-acrylate coatings exhibited after UV curing a greater resistance to moisture and proved therefore to be very resistant to wet cycle accelerated weathering, (24) as will be reported in a forthcoming article. (25)

CONCLUSION

UV radiation curing has been successfully used to rapidly crosslink water-based polyurethane-acrylate systems and obtain hard and chemically resistant coatings. As the polymerization of the acrylate double bonds occurred in the solid state after drying of the coating, it proved to be very sensitive to the mobility restrictions imposed on the growing polymer chains in such conditions. A number of mobility related factors were shown to greatly affect the polymerization kinetics, by controlling both the cure speed and the final conversion of the acrylate double bonds. Given the great accelerating effect of the temperature, it is recommended that the UV irradiation be performed on the hot sample emerging from the drying oven. Faster and more complete polymerization can also be achieved by adding a diacrylate monomer to the dispersion to act as a reactive plasticizer. A similar trend was observed by performing the UV exposure in a humid atmosphere, because the coating became softer in such an environment. An incomplete drying of the sample will have the same beneficial effect on the polymerization process, without being detrimental towards the properties of the UV-cured coating, as long as the residual water content is kept low enough. One of the distinct advantages of such UV-curable systems is their low sensitivity towards oxygen, thus allowing thin films to be rapidly cured in the presence of air.

The alkaline compound needed to neutralize the carboxylic acid groups and get a stable aqueous dispersion was also shown to play a key role in both the polymerization kinetics and the hydrophilic character of the UV-cured polymer. The best performance was obtained by using a volatile tertiary amine as neutralizer which, by being removed during the drying step, allowed the carboxylate anion to be converted into the less hydrophilic carboxylic acid group. Thus, polyurethane-acrylate dispersions have been designed which, once UV cured, showed nearly the same low water uptake as solvent-based PUA coatings containing no carboxylate groups.

UV radiation curing is an environment-friendly technology which offers a number of advantages, mainly solvent-free formulation, low energy consumption, and high-speed processing. The main applications of the UV-curable water-based systems examined here are expected to be found in the coating industry, in particular for protecting porous substrates like paper or wood and improving their surface properties, as well as their weathering resistance for outdoor end-uses.

Table 1 -- Hardening Upon UV Exposure at 25[degrees]C of a Water-Based
Polyurethane-Acrylate (PUA-1) and Solventborne Polyurethane-Acrylates
with (PUA-2) or without (PUA-3) Carboxylic Acid Groups. I = 600 mW
[cm.sup.-2]

                  Persoz Hardness (sec)                 Conversion (%)
           Heating       Heating         UV Curing        UV Curing
          10 min at      1 hr at     (3 J [cm.sup.-2])      (3 J
Sample  80[degrees]C  130[degrees]C                     [cm.sup.-2])

PUA-1       240            250              350               60
PUA-2       104            180              350               71
PUA-3        66             90              360               88

Table 2 -- Scratch and Chemical Resistance and Flexibility of UV-Cured
Polyurethane-Acrylate Coatings

                           Laromer LR 8987  Water-Based PUA

Gloss loss (%) (a)             34              32
Chemical resistance (b)         0.9             0.75
Pendulum hardness (s) (c)     172             153
Erichsen cupping (mm) (d)       3.4             6.6

(a) Gloss loss measured in a Scotch Brite test after 50 double rubs
scratching with a Scotch Brite fleece under a load of 750 g.
(b) Chemical resistance is characterized with an average mark (from 5 =
strong film destruction to 0 = no film destruction) after 1 hour soaking
of 10 typical household chemicals (mustard, red wine, ink, etc.) on the
film surface.
(c) Konig hardness of the film.
(d) Erichsen cupping reflects the flexibility of a film.

Table 3 -- Influence of the Neutralizer Volatility on the Hydrophilic
Character of a UV-Cured Water-Based PUA Coating

Neutralizer           Boiling Point  pK    [[theta].sub.w] (a)
                      ([degrees]C)

Sodium hydroxide         --          14       53[degrees]
Triethanolamine          335          7.8     59[degrees]
Dimethylethanolamine     134          9.2     64[degrees]
Triethylamine             89         10.8     66[degrees]
None (PUA-3)             --          --       70[degrees]

Neutralizer           [[gamma].sub.p] (mJ [cm.sup.-2]) (b)

Sodium hydroxide                   15.8
Triethanolamine                    12.5
Dimethylethanolamine                9.6
Triethylamine                       8.6
None (PUA-3)                        6.1

(a) [[theta].sub.w]: contact angle of a droplet of water.
(b) [[gamma].sub.p]: polar component of the surface energy.

ACKNOWLEDGMENTS

Two of the authors (F. Masson and C. Decker) express their thanks to BASF (Ludwigshafen) for a research grant and to Dr. W. Paulus and T. Jaworek for innovating ideas and fruitful discussions.

Presented at the Spring 2003 Meeting of the American Chemical Society, March 23-27, New Orleans, LA.

References

(1) Doren, K., Freitag, P., and Stoye, D., Waterborne Coatings. The Environmentally-Friendly Alternative, Hanser Verlag, p. 204 (1994).

(2) Nicholson, J., "Performance of Water-Based Coatings," in Waterborne Coatings and Additives, Karsa, D. and Davies W. (Ed.), Royal Soc. Chem. Cambridge, p. 73, 1995.

(3) Thomas, P., Waterborne and Solvent-based Surface Coatings and Their Applications, Vol. 3, Wiley & Sons, Chichester, 1998.

(4) Decker, C. and Moussa, K., "Recent Advances in UV-Curing Chemistry," JOURNAL OF COATINGS TECHNOLOGY, 65, No. 819, 49 (1993).

(5) Decker, C., "Photoinitiated Crosslinking Polymerization," Prog. Polym. Sci., 21, 593 (1996).

(6) Roffey, C., Photogeneration of Reactive Species for UV-Curing, Wiley, New York, 1997.

(7) Schwalm, R., Hausling, L., Reich, W., Beck, E., Enenkel, P., and Menzel, K., "Tuning of the Mechanical Properties of UV-Coatings Towards Hard and Flexible Systems," Prog. Org. Coat., 32, 191 (1997).

(8) Reich, W., Enenkel, P., Keil, E., Lokai, M., Menzel, K., and Shrof, W., "Waterbased Radiation-Curable System. Newest Investigations," Proc. of RadTech North America, p. 258 (1998).

(9) Arnoldus, R., "Radiation Curable Aqueous Emulsions," Proc. of RadTech Europe, p. 121 (1989).

(10) Philips, M., Loutz, J.M., Peeters, S., Destexhe, R., and Lindekens, L., "Radiation Curable Water Dilutable Polyester-Acrylates," Proc. of RadTech North America, p. 157 (1992).

(11) Masson, F., Decker, C., Jaworek, T., and Schwalm, R., "UV-Radiation Curing of Waterbased Urethane-Acrylate Coatings," Prog. Org. Coat., 39, 115 (2000).

(12) Awad, R. and Lunger, F., "New Developments of Waterborne UV-Resins for Wood Coatings," Proc. of RadTech Europe, p. 415, 2001.

(13) Tauber, A., Scherzer, T., and Menhert, R., "UV-Curing of Aqueous Polyurethane Acrylate Dispersions. A Comparative Study by Real-Time FTIR Spectroscopy and Pilot Scale Curing," JOURNAL OF COATINGS TECHNOLOGY, 72, No. 911, 51 (2000).

(14) Decker, C., Masson, F., Jaworek, T., and Schwalm, R., "High-Speed Curing of Waterborne Coatings by UV Irradiation," Polym. Mater. Sci. Eng., 85, 414 (2001).

(15) Jaworek, T., Menzel, K., Paulus, W., and Schwalm, R., "Water-Based UV-Curable Urethane-Acrylate Clearcoats," FATIPEC Congress, Vol. 2, 363, 2000.

(16) Decker, C., Masson, F., and Schwalm, R., "Dual-Curing of Waterborne Urethane-Acrylate Coatings by UV and Thermal Processing," Macromol. Mater. Eng., 288, 17 (2003).

(17) Decker, C., Moussa, K., and Bendaikha, T., "Photodegradation of UV-Cured Coatings II.--Polyurethane-Acrylate Networks," J. Polym. Sci., Polym. Chem. Ed., 29, 739 (1991).

(18) Decker, C., Zahouily, K., and Valet, A., "Weathering Performance of Thermoset and Photoset Acrylate Coatings," JOURNAL OF COATINGS TECHNOLOGY, 74, No. 924, 87 (2002).

(19) Decker, C. and Moussa, K., "Real-Time Monitoring of Ultrafast Curing by UV-Radiation and Laser Beams," JOURNAL OF COATINGS TECHNOLOGY, 62, No. 786, 55 (1990).

(20) Decker, C., "Kinetic Study and New Applications of UV Radiation Curing," Macromol. Rapid. Commun., 23, 1067 (2002).

(21) Decker, C., Elzaouk, B., and Decker, D., "Kinetic Study of Ultrafast Photopolymerization Reactions," J. Macromol. Sci., A33(2), 173 (1996).

(22) Decker, C., Decker, D., and Morel, F., "Light Intensity and Temperature Effect in Photoinitiated Polymerization," in Photopolymerization Fundamentals and Applications, Scranton, A.B., Bowman, C.N., and Peiffer, R.W. (Ed.), ACS Symp. Series G73, American Chemical Society, Washington D.C., p. 63, 1997.

(23) Studer, K., Decker, C., Beck, E., and Schwalm, R., "Overcoming Oxygen Inhibition in UV-Curing of Acrylate Coatings by Carbon Dioxide Inerting," Prog. Org. Coat., 48, 32 and 111 (2003).

(24) Masson, F., "Etude de la Polymerisation Photoamorcee de Revetements Obtenus a Partir de Dispersions Aqueuses de Polyurethanes-Acrylates," Ph.D. Thesis, University of Mulhouse (2001).

(25) Decker, C., Masson, F., and Schwalm, R., Polym. Degrad. Stab. (in press).

C. Decker** and F. Masson--Ecole Nationale Superieure de Chimie de Mulhouse, Universite de Haute-Alsace*

R. Schwalm--BASF-AG ([dagger])

*Departement de Photochimie Generale (UMR-CNRS No. 7525), 3, rue Werner--68200 Mulhouse, France.

([dagger]) Polymer Research Laboratory, 67056 Ludwigshafen, Germany.

**Author to whom correspondence should be addressed: C. [e-mail].