Optimization of UV curable acrylated polyester-polyurethane/polysiloxane ceramer coatings using a response surface methodology.

Dual UV and moisture curable acrylated polyester, organic/inorganic hybrid coatings were prepared using a coupling agent and tetraethylorthosilicate (TEOS) oligomers. An acrylated polyester resin based on adipic acid, neopentyl glycol, trimethylolpropane, and acrylic acid was synthesized. TEOS oligomers were prepared through the hydrolysis and condensation of TEOS with water and 3-(triethoxysilyl)propylisocyanate (TEOSPI) was used as the coupling agent between organic and inorganic phases. Trimethylolpropane triacrylate (TMPTA) was used as a reactive diluent. The formulations were cured into films by utilizing UV, followed by moisture curing. The resultant coatings were evaluated in terms of tensile and fracture toughness properties by using ANOVA. Effects of

the TEOS oligomers, TEOSPI, and TMPTA on the free radical UV polymerization kinetics were also investigated. The experiments were planned according to the three-factor, three-level Box-Behnken design. Reaction kinetics, fracture toughness, and tensile properties were evaluated in terms of the concentrations of TEOS oligomers, TEOSPI, and TMPTA concentrations.

Keywords: Photochemistry, organosilane, UV, EB, radiation cure, adhesion, design of experiments, statistical design, isocyanate, aluminum

**********

In early UV-cure coatings, unsaturated polyesters were used extensively in wood finishing as fillers, sealers, and topcoats. (1) Styrene is usually incorporated into the coating formulations as a reactive diluent to reduce viscosity, increase the crosslink density, and decrease cost. Current UV-cure coatings are mostly based on acrylated oligomers due to their relatively higher reactivity and lower volatility. Acrylated resins are preferred over methacrylated ones due to higher cure rates at room temperature and lower oxygen inhibition. (2) The main types of acrylate oligomers are: epoxy acrylates, polyether acrylates, urethane acrylates, polyester acrylates, and silicone acrylates. Polyester acrylates are synthesized by reacting either the hydroxyl terminated polyesters with acrylic acid or the carboxyl terminated polyesters with hydroxyethyl acrylate. (3) They are used in wood coatings, varnishes, and lithographic and screen inks. (4)

Another class of UV-curing resins are polyurethanes synthesized via the reaction of hydroxyl and isocyanate functionalized monomers or oligomers, (5) as shown below.

[FORMULA NOT REPRODUCIBLE IN ASCII]

The high reactivity of the isocyanate group allows for ambient temperature preparation of the polyurethane coatings. Due to the balance of hydrogen bond forming and flexible segments, polyurethanes show superior abrasion resistance. (1) UV-curing polyurethane acrylates are prepared by reacting isocyanates with hydroxyl functional acrylates. (6) Versatile chemistry of polyurethane acrylates allow for coatings with good flexibility, toughness, abrasion resistance, adhesion, chemical resistance, and excellent weathering resistance. (4)

Free radical UV-curing acrylates, including polyester and polyurethane acrylates, show a typical autoacceleration-autodeceleration polymerization behavior. As the conversion increases, the mobility of polymer radicals decreases due to the increased viscosity and entanglements, which causes a slower rate of termination and higher rate of polymerization (autoacceleration). (7) At higher conversion rates, the effect of crosslinking on the chain propagation reaction becomes more predominant and the rate of polymerization decreases (autodeceleration). Autoacceleration-autodeceleration behavior can be characterized by the conversion at maximum rate of polymerization ([[chi].sub.mrp]). (8)

Organofunctional silanes have been utilized as coupling agents for a variety of organic binders including polyesters and polyurethanes in composites and primer coatings as adhesion and corrosion protection promoters. (2,9,10) Alkoxysilanes are common sol-gel precursors that hydrolyze and condense to form Si-O-Si bonds in water. The sol-gel process can be depicted by the following reactions:

[FORMULA NOT REPRODUCIBLE IN ASCII]

Further hydrolysis and condensation reactions form either crosslinks or a silicon oxide network. Soucek and coworkers (11-14) showed that alkoxysilanes functioned as a compatalizer and coupling agent in TEOS-based inorganic/organic polyurethane hybrid materials. They also reported that alkoxysilane groups functioned as a nucleation site for silicon-oxo-cluster growth, thus providing a template for uniform dispersion of silicon-oxo-nanophase.

Various researchers used reactive monomers and acrylate or methacrylate functionality in the organic phase to UV-cure organic/inorganic hybrid coatings. Gilberts et al. (15) prepared UV-curing hybrid coatings on polycarbonate using hexanedioldiacrylate as the reactive diluent, and 3-(trimethoxysilyl) propoxymethacrylate as the coupling agent. It was found that incorporation of prehydrolyzed TEOS greatly improved the abrasion resistance. Gigant et al. (16) formed hybrid films based on vinyltriethoxysilane and tetraethoxysilane and polyfunctional acrylates and showed structure-property correlations between microhardness and normalized Raman intensities. Soppera et al. (17) studied the effect of silicate backbone on photopolymerization kinetics. Colloidal silica acrylates and methacrylates were also used by various researchers to form abrasion resistant inorganic/organic hybrid coatings. (18-20)

Investigation of the kinetic and mechanical properties of UV-curing organic-inorganic hybrid systems is relatively hard due to the complexity and number of variables. Simple experiments can take long times to explore complex systems, and the results mostly depend on the skill and experience of the researcher. Designed experiments can find and quantify how the interaction of two or more factors affects the system. It is also easier and faster with the designed experiments to demonstrate nonlinear relationships. The object of the response surface methodology (RSM) is to form a mathematical model of the system by using statistical analysis of the experimental results. When nonlinearity is concerned, more than two level experiments are required to describe a response curve. (21) A nonlinear quadratic response function for three variables can be described by the following equation:

[FIGURE 1 OMITTED]

[gamma] = [[beta].sub.0] + [[beta].sub.1][X.sub.1] + [[beta].sub.2][X.sub.2] + [[beta].sub.3][X.sub.3] + [[beta].sub.1][[beta].sub.2][X.sub.1][X.sub.2] + [[beta].sub.1][[beta].sub.3][X.sub.1][X.sub.3] + [[beta].sub.2][[beta].sub.3][X.sub.2][X.sub.3] + [[beta].sub.11][X.sub.1.sup.2] + [[beta].sub.22][X.sub.2.sup.2] + [[beta].sub.33][X.sub.3.sup.2]

A three-factor, three-level Box-Behnken design is one of many response surface designs used to construct the response surface for the analysis and optimization of the coatings formulations. Box-Behnken designs are spherical, nearly rotatable, and require fewer experimental runs than central composite designs. These designs should not be preferred for predicting responses at the extremes. (21) The three-factor, three-level Box-Behnken designs are depicted in Figure 1. Data obtained from the designed experiments can be analyzed using analysis of variance (ANOVA), collection of statistical models which compare means by subdividing the overall observed variance into different parts. ANOVA shows whether model variance is significant when compared to experimental variance. (22)

EXPERIMENTAL

Materials

Adipic acid (ADA), trimethyiol propane (TMP), and neopentyl glycol (NPG), were purchased from Aldrich. Photoinitiator, 1:1 mixture by weight of 1-hydoxy-cyclohexyl-phenyl-ketone and benzophenone (IRGACURE 500), was obtained from Ciba Specialty Chemicals. Tetraethyl orthosilicate (TEOS), trimethylolpropane triacrylate (TMPTA), and the coupling agent, 3-(triethoxysilyl) propyl isocyanate (TEOSPI), were also obtained from Aldrich. All the formulations were prepared under an argon atmosphere in a glove box.

[FIGURE 2 OMITTED]

Preparation of Acrylated Polyester

A hydroxyl terminated polyester with a functionality of 3 was synthesized using ADA, NPG, and TMP with molar amounts 2, 2, and 1, respectively, via a two-stage polycondensation method. For the first stage, ADA and NPG (2:1) were reacted at 185[degrees]C for three hours, and then a mixture of an equimolar NPG and TMP (1:1) was added to the reaction flask for the second stage. Temperature of the flask was kept at 185[degrees]C for two hours and was then increased to 215[degrees]C. The reaction was performed in a three-neck flask equipped with a mechanical stirrer and Dean-Stark trap under a nitrogen purge until the resultant polyester resin had an acid value of 10 mg KOH/g (ASTM D 1639-90) and hydroxyl number of 150 mg KOH/g (ASTM D 4274-94) resin. The polyester was characterized by Fourier transform infrared (FTIR) and nuclear magnetic resonance (NMR) spectroscopy. The number average molecular weight was 1100 with a polydispersity of 1.5 according to gel permeation chromatography (GPC). Acrylated polyester (AcPE) was prepared by the reaction of acrylic acid (AA) with the polyester at 115[degrees]C for 12 hr. Molar amount of acrylic acid was 0.65 x (hydroxyl equivalent number of the polyester). In addition, 0.5% p-toluenesulphonic acid (p-TSA) and 0.1% hydroquinone were used as the catalyst and the inhibitor. The yield of the reaction was determined from H-NMR spectra as 90%. Preparation of acrylated polyester (AcPE) is depicted in Figure 2.

Instrumentation

Tensile tests were performed on an Instron 5567 (Instron Corporation). FTIR spectra were obtained on a Mattson Genesis Series FTIR. [.sup.1]H-NMR spectra were recorded on a Gemini-300 spectrometer (Varian) in CDC[l.sub.3] solvent with tetramethylsilane (TMS) as a reference. A Waters system was used for gel permeation chromatography (GPC) with an HR4, HT2, HR1, HR0.5 Styragel, and 500 [Angstrom] ultrastyragel columns connected in series. Tetrahydrofuran was applied as the mobile phase and delivered at a rate of 1.0 mL/min. Photopolymerization kinetic studies were performed using a Q 1000 differential scanning calorimeter (DSC) (TA Instruments) equipped with a photocalorimeter accessory (PCA). The intensity of the exposed UV light was 33 mW/[cm.sup.2]. Obtained data were analyzed using the TA Instruments Universal Analysis software. Rate of polymerization ([R.sub.p]) and conversion (C) were calculated by integrating the area under the DSC exotherm. Acrylate conversion was calculated via the equation C = [DELTA][H.sub.t]/[DELTA][H.sub.o], where [DELTA][H.sub.t] is the heat of reaction calculated by integrating exothermic peak area up to time "t." The theoretical heat for 100% conversion, [DELTA][H.sub.o], was taken as 86.25 kJ/mol for an acrylate double bond. (23) [R.sub.p] was directly calculated from the heat flow (dH/dt) by utilizing equation [R.sub.p] = (dH/dt)/[DELTA][H.sub.o]. The thin wet films were cured in a Fusion F300 UV processor (H-bulb; Fusion UV System, Inc.).

Formulations and Film Formation

The coating formulations were prepared according to the Box-Behnken response surface design lattice. Table 1 summarizes the factors and their levels.

TEOSPI was used as a coupling agent between the organic phase and the inorganic phase. The reaction between TEOSPI and AcPE was performed at ambient temperature and was monitored by utilizing FTIR spectroscopy. Coatings were formulated with NCO/OH ratios of 1, 0.5, and 0 to observe the effect of the coupling agent. After the reaction of the coupling agent and AcPE, the rest of the formulation ingredients, as listed in Table 2 (TEOS, TMPTA, and photoinitiator), were added. To eliminate the evaporation of TEOS, oligomers were prepared by reacting TEOS with water and hydrochloric acid. (14) A unimolecular photoinitiator mixture (Irgacure 500), 4% of the total amount of TMPTA and acrylated polyester, was incorporated into the formulations for UV curing. The chemistry of the preparation of the coating film is depicted in Figure 3. The coatings were casted on glass panels by a drawdown bar with a thickness of 127 [micro]m. Films were formed in a UV processor and then further cured at 90[degrees]C for 20 hr at 100% humidity with moisture.

RESULTS AND DISCUSSION

The main aim of this study was to design a UV-moisture dual curable organic/inorganic hybrid system by utilizing a polyester acrylate, isocyanate functional alkoxysilane coupling agent, and sol-gel precursor. The free radical reaction kinetics and mechanical properties of the hybrid coatings were investigated using an experimental design response surface methodology. The effect of the sol-gel precursor, TEOS oligomers, the coupling agent, TEOSPI, and the reactive diluent, TMPTA, on the free radical polymerization kinetics and mechanical properties of the UV-moisture dual curable inorganic-organic hybrid coatings was investigated.

[FIGURE 3 OMITTED]

Effect of TEOS oligomer concentration, TEOSPI concentration, and TMPTA concentration on the conversion at maximum rate of polymerization ([[chi].sub.mrp]) was investigated to characterize the autoacceleration-autodeceleration behavior. Design Expert software (Stat-Ease Inc.) was used to obtain the statistical modeling of the design space. The experimental data were analyzed to fit into a reduced quadratic model. It was observed that there is a relationship between [[chi].sub.mrp] and concentrations of TEOS oligomers, TEOSPI, and TMPTA. Best fit mathematical model in terms of coded factors for the response [[chi].sub.mrp] and variables, concentrations of TEOS oligomers (A), TEOSPI (B), and TMPTA (C), was:

[[chi].sub.mrp] = 28.59 - 2.072A - 1.350 B - 0.4025C - 5.026 [A.sup.2] - 3.272[C.sup.2] + 3.374AB + 2.872AC

The model F-value of 16.6 implied that the model was significant. There was only a 0.07% chance that a model F-value this large could occur due to noise. The lack of fit F-value of 0.78 showed the lack of fit was not significant relative to the pure error. Nonsignificant lack of fit was desirable because it implied that the model was fit. Adequate precision is a measure of the ratio of predicted response to its associated error. A ratio greater than 4 is desirable. The model adequate precision ratio of 13.52 indicated an adequate signal. Therefore this model could be used to navigate the design space.

Figures 4 and 5 show interaction graphs of [[chi].sub.mrp] at different levels of TMPTA and TEOSPI, respectively. At high TMPTA or TEOSPI concentrations, an increase in TEOS amount increases [[chi].sub.mrp], but a further increase results in lower [[chi].sub.mrp]. Concave shaped interaction curves imply that TEOS oligomers have two competing effects: one that increases and one that decreases [[chi].sub.mrp] values. Also, interaction graphs suggest strong synergistic and antagonistic effects between TEOS oligomers, TEOSPI, and TMPTA. At high TMPTA and low TEOSPI concentrations, TEOS had almost no increasing effect, but with the increase of either TMPTA or TEOSPI, influence of the increasing effect was more pronounced. Inorganic network formation of siloxanes while UV-curing process was earlier observed by Soppera et al. (17) It was attributed to the breakage of the hydrogen bonds and availability of the free silanol moieties to react with other silanol or alkoxysilane functionalities. The results presented in this article were also in agreement with the findings of Soppera et al. Inhibition and inorganic network formation effects can be proposed to explain the influence of TEOS oligomers on [[chi].sub.mrp]: (1) Inhibition effect: Inorganic network formation inhibits the reaction of acrylates and formation of the organic network, to some extent, increases [[chi].sub.mrp]. This effect is more pronounced at higher TMPTA values; (2) Inorganic network formation effect: As TEOS oligomer concentration increases, inorganic network by itself increases the viscosity of the system and so decreases [[chi].sub.mrp]. In the absence of TEOS oligomers, with the increase of the coupling agent, TEOSPI, [[chi].sub.mrp] decreases due to the formation of the hard urethane segments in the backbone structure of the resin. At 20% TEOS oligomer concentration, the effect of the TEOSPI on [[chi].sub.mrp] is reversed due to the increase in the inhibition effect of the TEOS oligomers at high TEOSPI concentrations.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

The response function equation obtained for tensile strength of the films in terms of coded factors was:

[Tensile Strength (MPa)][.sup.-1.77] = 0.053 - 0.012A - 0.010B - 0.025C - 0.022 [A.sup.2] + 0.016AB + 0.021AC

The model F-value of 23.8 implied that the model was significant. There was only 0.06% chance that a model F-value this large could occur due to noise. "Prob > F" values less than 0.05 indicate model terms are significant. In this model, A, B, C, [A.sup.2], AB, and AC were the significant terms. The lack of fit F-value of 0.49 implied nonsignificant lack of fit. The model adequate precision ratio of 14.59 indicated an adequate signal. Therefore, this model was used to navigate the design space. The interaction and 3D response surface graphs shown in Figures 6 and 7 reveal that tensile strength increases as the reactive diluent, TMPTA, and the coupling agent, TEOSPI, concentrations increase. At high TMPTA and/or TEOSPI concentrations, tensile strength of the films decreases extensively after TEOS oligomers concentration exceeds 5%. At low TMPTA and/or TEOSPI concentrations, TEOS oligomer concentration does not affect the tensile strength of the films appreciably.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

The response surface equation obtained for elongation-at-break of the films in terms of coded factors were:

[Elongation-at-break (%)][.sup.1.77] = 17.85 + 6.930A + 14.99B - 30.25C + 19.34[A.sup.2] - 11.24[B.sup.2] + 17.57[C.sup.2] + 17.09BC

The model F-value of 14.6 implied that the model was significant. There was only 0.47% chance that a model F-value this large could occur due to noise. The lack of fit F-value of 2.32 implied nonsignificant lack of fit. The model adequate precision ratio of 12.3 indicated an adequate signal. Therefore, this model was used to navigate the design space. Figures 8 and 9 reveal interaction and 3D elongation-at-break response surface graphs. Elongation-at-break decreases as the concentration of the reactive diluent increases, whereas it slightly increases with the coupling agent. Elongation-at-break first decreases as TEOS oligomer concentration increases, but with further increase in TEOS oligomer concentration, it increases.

The model equation for tensile modulus was linear:

Tensile Modulus (MPa) = 144.2 - 0.1500A - 9.420B + 100.9C

The model F-value of 26.1 implied that the model was significant. There was only 0.01% chance that a model F-value this large could occur due to noise. The lack of fit F-value of 0.47 implied nonsignificant lack of fit. The model adequate precision ratio of 14.2 revealed an adequate signal. Figure 10 depicts a 3D tensile modulus response surface. Tensile modulus appeared to be very weakly correlated to the coupling agent and TEOS oligomer concentrations, but increased drastically with the reactive diluent.

[FIGURE 10 OMITTED]

The fracture toughness ([K.sub.c]) is a measure of resistance to crack extension. Best fit model equation for [K.sub.c] in terms of coded factors was:

[[K.sub.c] (MPa[m.sup.1/2])][.sup.-1.53] x [10.sup.4] = 9.641 - 0.8601A - 5.561B - 13.29C + 6.644[A.sup.2] + 11.31[C.sup.2] + 4.526AB - 6.614AC + 6.232BC

The model F-value of 61.0 implied that the model was significant. There was only 0.01% chance that a model F-value this large could occur due to noise. The lack of fit F-value of 3.15 implied nonsignificant lack of fit relative to the pure error. The model adequate precision ratio of 21.9 indicated that there was an adequate signal. Therefore, this model could be used to navigate the design space. In Figures 11 and 12, it was observed that with the increase of the reactive diluent concentration, [K.sub.c] also increased. Amount of the increase in [K.sub.c] was not linear and depends on both TEOS oligomer and the coupling agent concentrations. The change of [K.sub.c] with respect to the coupling agent concentration was also similar and depends on both TEOS oligomer and the reactive diluent concentration. At low reactive diluent or coupling agent concentrations, [K.sub.c] was loosely dependent (can be considered independent) on TEOS oligomer concentration. At high TMPTA concentration, the highest [K.sub.c] value was obtained around 15% TEOS oligomer concentration. Similarly, at high TEOSPI concentration, 8% TEOS oligomer concentration gave the highest [K.sub.c] value.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

Utilization of the response surface methodology on the kinetics and mechanical properties of free radical UV-curing ceramer films were not previously reported in the literature. In this study, synergistic and antagonistic effects of the sol-gel precursor, coupling agent, and reactive diluent on curing kinetics, tensile, and fracture-toughness of the hybrid films were evaluated with a Box-Behnken design lattice. Complex, nonlinear relationships in UV-curing ceramer coatings were successfully revealed via utilizing response surface methodology. Inhibition and inorganic network formation effects of the sol-gel precursor on the conversion at maximum rate of polymerization were proposed to describe the complex autoacceleration-autodeceleration behavior of UV-curing polyester-polyurethane/polysiloxane hybrid coatings.

CONCLUSION

A hydroxyl functional aliphatic polyester was synthesized and acrylated by using acrylic acid. Coating formulations were prepared according to Box-Behnken response surface design with the independent variables: TEOS oligomers, 3-(triethoxysilyl) propylisocyanate (TEOSPI) as the coupling agent, and trimethylolpropane triacrylate (TMPTA) as the reactive diluent. It was found that the concentration of TEOS oligomers plays an important role in the autoacceleration-autodeceleration behavior, which was explained by the inhibition and inorganic network formation effects. Tensile strength values increase as the reactive diluent and the coupling agent concentrations increase. At high TMPTA and/or TEOSPI concentrations, tensile strength of the films decrease after TEOS oligomers concentration exceeds 5%. At low reactive diluent or coupling agent concentrations, fracture toughness was independent of TEOS oligomer concentration. At high TMPTA concentration, the highest fracture toughness value was obtained around 15% TEOS oligomer concentration. Similarly, at high TEOSPI concentration, 8% TEOS oligomer concentration gave the highest fracture toughness value. Thus, the properties of the inorganic/organic hybrid networks were sensitive to optimizing the organic crosslinker and coupling agent.

References

(1) Chiocchetti, P., "Radiation Curing Coatings in the Wood Industry," Polym. Paint Col. J., 178, 4216 (1988).

(2) Wicks, Z.W., Jones, F.N., and Pappas, S.P., Organic Coatings Science and Technology, 2nd Ed., John Wiley and Sons, New York, 1999.

(3) Mehnert, R., Pincus, A., Janorsky, I., Stowe, R., and Berejka, A., Chemistry and Technology of UV and EB Formulations for Coatings, Inks, and Paints: UV and EB Curing Technology and Equipment, John Wiley and Sons, New York, 1998.

(4) Webster, G., Chemistry and Technology of UV and EB Formulation for Coatings, Inks, and Paints: Prepolymers and Reactive Dilutents for UV and EB Curable Formulations, John Wiley and Sons, New York, 1997.

(5) Pourjavadi, A., Rezai, N., and Zohuriaan-M, M.J., "A New Homologous Series of Linear Aliphatic Unsaturated Hydroxypolyesters as Polyol Soft Segments for Polyurethanes: Synthesis and Characterization," J. Appl. Polym. Sci., 68, 173-183 (1998).

(6) Swaraj, P., Surface Coatings: Science and Technology, 2nd Ed., John Wiley and Sons, New York, 1996.

(7) Odian, G.G., Principles of Polymerization, John Wiley and Sons, New York, 2004.

(8) Avci, D., Nobles, J., and Mathias, L.J., "Synthesis and Photopolymerization Kinetics of New Flexible Diacrylate and Dimethacrylate Crosslinkers Based on C18 Diacid," Polymer, 44, 963-968 (2003).

(9) Walker, P., "Organo-Silanes as Adhesion Promoters for Organic Coatings," J. COAT. TECHNOL., 52, No. 670, 49 (1980).

(10) Witucki, G.L.," A Silane Primer: Chemistry and Applications of Alkoxy Silanes," J. COAT. TECHNOL., 65, No. 822, 57 (1993).

(11) Ni, H., Skaja, A.D., and Soucek, M.D., "Acid-Catalyzed Moisture-Curing Polyurea/Polysiloxane Ceramer Coatings," Prog. Org. Coat., 40, 175-184, (2000).

(12) Ni, H., Skaja, A.D., Sailer, R.A., and Soucek, M.D., "Moisture-Curing Alkoxysilane-Functionalized Isocyanurate Coatings," Macromol. Chem. Phys., 201, 722-732 (2000).

(13) Ni, H., Aaserud, D.J., Simonsick, W.J., and Soucek, M.D., "Preparation and Characterization of Alkoxysilane Functionalized Isocyanurates," Polymer, 41, 57-71 (2000).

(14) Ni, H., Simonsick, W.J., Skaja, A.D., Williams, J.P., and Soucek, M.D., "Polyurea/Polysiloxane Ceramer Coatings," Prog. Org. Coat., 38, 97-110 (2000).

(15) Gilberts, J., Tinnemans, A.H.A., Hogerheide, M.P., and Koster, T.P.M., "UV Curable Hard Transparent Hybrid Coating Materials on Polycarbonate Prepared by the Sol-Gel Method," J. Sol.-Gel Sci. Technol., 11, 153-159 (1998).

(16) Gigant, K., Posset, U., Schottner, G., Baia, L., Kiefer, W., and Popp, J., "Inorganic-Organic Cross-Linking in UV Curable Hard Coats Based Upon Vinyltriethoxysilane-Tetraethoxysilane-Polyfunctional Acrylate Hybrid Polymers: A Raman Spectroscopic Study" J. Sol.-Gel Sci. Technol., 26, 369-373 (2003).

(17) Soppera, O. and Croutxe-Barghorn, C., "Real-Time Fourier Transform Infrared Study of Free-Radical UV-Induced Polymerization of Hybrid Sol-Gel. I. Effect of Silicate Backbone on Photopolymerization Kinetics," J. Polym. Sci., Polym. Chem., 41, 716-724 (2003).

(18) Guang-Way, J., Ren-Jye, W., Yuung-Ching, S., Ya-Hui, L., and Chi-Jung, C., "UV Curable Organic-Inorganic Hybrid Materials," Mater. Res. Soc. Symp. Proc., Vol. 628, 14.1-14.8 (2000).

(19) Vu, C., LaFerte, O., and Eranian, A., "Colloidal Silica Acrylates Use in UV Coatings," Eur. Coat. J., 1-2, 64-70 (2002).

(20) Chrisholm, B., Resue, J., "UV-Curable Hybrid Organic-Inorganic Coatings," Proc. Int. Waterborne, High-Solids, and Powder Coatings Symposium, New Orleans, LA, 261-281, 2003.

(21) Myers, R.H. and Montgomery, D.C., Response Surface Methodology, Process and Product Optimization Using Designed Experiments, John Wiley and Sons, New York, 2002.

(22) Del Vecchio, R.J., Understanding Design of Experiments: A Primer for Technologists, Hanser/Gardner, Cincinnati, 1977.

(23) Anseth, K.S., Wang, C.M., and Bowman, C.N., "Reaction Behavior and Kinetic Constants for Photopolymerizations of Multi(meth)acrylate Monomers," Polymer, 35, 3243 (1994).

Ahmet Nebioglu and Mark D. Soucek ([dagger]) -- The University of Akron*

Presented at the 82nd Annual Meeting of the Federation of Societies for Coatings Technology, October 25-27, 2004, in Chicago, IL.

* Department of Polymer Engineering, 250 S. Forge St., Akron, OH 44325.

([dagger]) Author to whom correspondence should be addressed. Email: [e-mail].

Table 1 -- Factors and Their Levels for Box-Behnken Response Surface
Design

                                            Level
Factors                                 -1  0      1

A: TEOS oligomer concentration (%) (a)   0  10    20
B: TEOSPI concentration (%) (a)          0  13.5  27
C: TMPTA concentration (%) (a)          20  35    50

(a) Percentage values are based on the amount of AcPE.

Table 2 -- Formulations for Acrylated Polyester-Polyurethane Ceramer
Coatings

          TEOS  TEOSPI  TMPTA  Acrylated Polyester  Irgacure 500
Sample #  (g)   (g)     (g)    (g)                  (g)

 1        0     2.7     3.5    10                   0.54
 2        2     1.35    2      10                   0.48
 3        1     2.7     2      10                   0.48
 4        1     0       5      10                   0.6
 5        2     1.35    5      10                   0.6
 6        2     0       3.5    10                   0.54
 7        1     1.35    3.5    10                   0.54
 8        0     1.35    5      10                   0.6
 9        0     0       3.5    10                   0.54
10        1     2.7     5      10                   0.6
11        1     1.35    3.5    10                   0.54
12        2     2.7     3.5    10                   0.54
13        1     1.35    3.5    10                   0.54
14        1     0       2      10                   0.48
15        0     1.35    2      10                   0.48