The Effects of Acrylate Secondary Functionalities on The Kinetics of Epoxide during Epoxide-Acrylate Hybrid Photopolymerizations

Transcription

1 The Effects of Acrylate Secondary Functionalities on The Kinetics of Epoxide during Epoxide-Acrylate Hybrid Photopolymerizations Ho Seop Eom and Julie L. P. Jessop Chemical & Biochemical Engineering, University of Iowa Iowa City, IA Abstract Epoxide-acrylate hybrid systems mitigate oxygen inhibition and moisture sensitivity of free-radical and cationic photopolymerizations, respectively. The control of interpenetrating networks (IPNs) produced by chemically independent free-radical and cationic polymerizations allows tuning of physical/mechanical properties of final hybrid polymers. Here, highly viscous urethane acrylate oligomers were first combined with epoxides. However, the epoxide cationic photopolymerizations were considerably suppressed in the presence of these urethane acrylate oligomers. To determine the factors causing the sluggish cationic polymerizations, low-viscosity mono-functional acrylates with various secondary functional groups were then examined in hybrid systems containing diepoxides. Using Raman spectroscopy, the epoxide polymerization rate and final conversion in hybrid systems were shown to be affected significantly by the acrylate structures and their molar ratios. Acrylates containing ether or urethane groups negatively affected the epoxide kinetics for higher molar ratios of acrylate to epoxide. This detrimental effect is caused by the fixation or abstraction of protons generated from the photolysis of photoinitiators by ether or urethane groups, and these secondary groups and compositions should be taken into account when tuning epoxide conversion and ultimate strength of hybrid IPNs. Introduction The effects of atmospheric conditions on photopolymerizations Photopolymerization, which uses light energy to initiate a chemical reaction, is one of the most efficient ways to produce a variety of polymer structures. The major advantages of this technology lie with its energy savings, low production of volatile compounds, and ultra-fast curing process at ambient temperature, as well as ease of temporal and spatial control of reactions. These unique attributes have led to its adoption within the fields of coatings, optoelectronics, adhesives, membranes, stereolithography, composites, and biomaterials [1, 2]. In industrial applications, ultraviolet (UV) light-curable systems are most prevalent and facilitate the rapid free-radical and

2 cationic polymerizations of a wide range of acrylates and epoxides, respectively. The rapid polymerizations allow desirable physical properties to be developed in the course of the reaction, depending on curing conditions and formulation of the reaction system. However, both free-radical and cationic systems are not free of shortcomings for the photopolymerizations, since free-radical and cationic species generated from the photolysis of photoinitiators are affected detrimentally by atmospheric conditions, such as oxygen and humidity. In the presence of oxygen, the excited triplet-state photoinitiator/sensitizer molecules can be quenched 3 by interaction with triplet-state oxygen ( 0 2 ), and radical species can form stable, less reactive peroxy radicals after reacting with oxygen [3, 4]. As a result, acrylate free-radical photopolymerization is inhibited and delayed until the concentration of dissolved oxygen in the reaction system is consumed to a certain degree [5]. In addition, continued diffusion of oxygen into curing systems interferes with polymer chain growth at the air/coating interface, resulting in a undesirable tacky surface [6]. Comparably, atmospheric moisture can have an adverse effect on epoxide cationic photopolymerization, since the basicity of water molecules can inhibit, retard, or induce chain-transfer reactions with cationic active species [7]. nce the internal concentration of water increases over a certain threshold, the majority of propagating cationic active centers competes with water, resulting in the reduction of kinetic chain length or termination of growing epoxide polymers, thereby deteriorating the end-product performance. Many efforts to address the problems with atmospheric factors have been approached by various means. For acrylate systems, expensive inerting equipment, waxes, or shielding films can be used to prevent the diffusion of oxygen into the system. Higher concentrations of a photoinitiator or high-intensity irradiation sources can be used to produce a larger number of radicals to consume the dissolved oxygen faster and allow the polymer chains to grow. ther chemical species such as amines, thiols, and a combination of singlet oxygen generator and trapper can be added to acrylate photopolymerization systems in order to capture the oxygen, transform the normally inactive peroxy radical into a propagating active center, or consume dissolved oxygen [8]. For epoxide systems, highly hydrophobic compounds, such as epoxide monomers containing silicon or long hydrocarbon chains, can prevent the diffusion of water from environments with high moisture content [9]. Another promising approach is to develop hybrid photopolymerization systems containing epoxide and acrylate monomers. In the epoxide-acrylate hybrid systems, free-radical and cationic polymerizations take place in a simultaneous or sequential step, depending on curing conditions and the kinetics of the respective monomers. In addition, improved conversion and polymer properties conditions can be realized by the complementary combination, since the radical species are not sensitive to nucleophilic agents and the cationic active species are not scavenged by oxygen [6]. Epoxide-acrylate hybrid photopolymerizations Free-radical and cationic hybrid photopolymerizations have been reported by formulating acrylate with epoxide or vinyl ether monomers and by synthesizing monomers containing both epoxide and acrylate moieties [6, 10]. The kinetic synergy has been demonstrated between these two distinguishable polymerization mechanisms for epoxide-acrylate hybrid systems using real-time infrared [10] and Raman [6] spectroscopies. In these systems, the formation of epoxide polymer domains at the coating surface prevented further diffusion of atmospheric oxygen into the sample. In addition, physical and mechanical properties of hybrid polymers can be tuned, not only by

3 composition of acrylate and epoxide mixtures, but also by production of interpenetrating networks (IPNs) based on monomer functionality [11]. The development of IPNs is significantly affected by the kinetics of the two independent reactions, such as the rate of polymerization, final conversion, and kinetic sequences. Thus, the degree of entanglement of two different polymer domains varies at the micro/nano-scale during phase separation, and the interaction of the two domains, (i.e., non-covalent bonding interaction) will have a direct effect on the ultimate performance of the polymer products. In previous studies, urethane acrylate oligomers were combined with difunctional cycloaliphatic epoxide (DCE) monomer systems to mitigate the atmospheric sensitivity by taking advantage of the high viscosity and hydrophobicity of the selected urethane acrylate oligomers, which hinder diffusion of oxygen and water into the photopolymerizing system. [12]. Improved processibility of the formulation at room temperature was provided by the dilution of the highly viscous acrylate oligomers with low molecular weight epoxides. The addition of the acrylate to the epoxide formulation could also enhance the fracture toughness and impact resistance of the final hybrid polymers over that of the neat epoxide polymers, since the elasticity of the urethane acrylate polymers could surmount the brittleness of the epoxide polymers. Despite all expected benefits from these proposed hybrid systems, increasing the initial viscosity of hybrid photopolymerization systems containing oligomeric acrylates affected the kinetics of the epoxide cationic photopolymerizations, resulting in slower polymerization rates compared to the free-radical kinetics of acrylates. [12]. Furthermore, rapid formation of acrylate polymer domains significantly suppressed propagation of cationic active centers during the hybrid photopolymerizations. However, it was not clear what variables (e.g., viscosity, chemical structure, and/or compositions of epoxide and acrylate) primarily caused the sluggish cationic photopolymerizations of the epoxides in the hybrid systems. Since the knowledge of how variations in acrylate structures and compositions affect the kinetics of epoxide cationic photopolymerizations is lacking, a series of low-viscosity acrylate monomers with similar molecular weights as DCE monomer was selected for epoxide-acrylate hybrid formulations, thereby excluding the effect of initial viscosity. A better understanding of these two effects will ultimately facilitate optimization of formulations and reaction conditions for thin film and coating applications using epoxide-acrylate hybrid photopolymerizations. Experimental methods Materials A difunctional cycloaliphatic epoxide 3,4-epoxycyclohexylmethyl 3,4-epoxycyclohexane carboxylate (EEC) was used for cationic photopolymerizations and supplied from Union Carbide. Three mono-functional acrylate monomers were selected: hexyl acrylate (HA) and 2-butylaminocarbonyloxyethyl acrylate (BCEA) from Sigma-Aldrich; 2-butoxyethyl acrylate (BEA) from Scientific Polymer Products. Each acrylate contains a different secondary group in its backbone adjacent to the mono-functional acrylate group: hydrocarbon tail (HA), ether (BEA), and urethane (BCEA). A non-reactive ethyl hexanoate was purchased from Sigma-Aldrich. The α-cleavable free-radical photoinitiator 2,2-dimethoxy-2-phenyl-acetophenone (DMPA, Aldrich) and the cationic photoinitiator diaryliodonium hexafluoroantimonate (IHA, Sartomer) were used to initiate the

4 reaction of C=C double bonds and epoxide rings, respectively. All materials were used as received and are shown in Table 1. Epoxides and acrylates were mixed thoroughly by molar ratios. The concentration of each photoinitiator (PI) was calculated to be 32 mm for the corresponding monomer fraction. Hence, the relative concentration of the individual PIs was kept constant to the fraction of the respective acrylates and epoxides (Table 2). The photoinitiator powders were completely mixed with individual acrylates, epoxides, or their hybrid mixtures until the formulated liquids were transparent. Table 1. Structure and selected physical properties of monomers, oligomeric elastomers, and photoinitiators used for this study. Materials Abbreviation Viscosity f* (MW, g/mole) [cp] Structure HA H 2 C 25 C Acrylates Difunctional cycloaliphatic epoxides BEA (172.2) BCEA (215.2) EEC (252.31) 1 25 C 1 NA 2 25 C H 2 C H 2 C CH 3 CH 3 NH CH 3 Ethyl hexanoate EH (144.21) NA NA Photoinitiators DMPA (256.30) IHA (745.23) NA NA SbF 6 *f symbolizes the number of functional groups that undergo polymerization for monomers used in this study. Methods Raman spectroscopy was used for real-time monitoring of acrylate and epoxide functional groups during hybrid photopolymerizations. Real-time Raman spectra were collected using a holographic fiber-coupled stretch probehead (Mark II, Kaiser ptical Systems, Inc.) attached to a modular research Raman spectrograph (HoloLab 5000R, Kaiser ptical Systems, Inc.). A 10x non-contact sampling objective with 0.8-cm working distance was used to deliver ~200 mw 785 nm near-infrared laser intensity to the sample, thereby inducing the Raman scattering effect. The exposure time for each spectrum was 100 ms, and the time interval between data points was 6 s. Samples were illuminated for 30 min at 25 C in sealed 1 mm ID quartz capillary tubes using a 100 NA NA H CH 3(CH 2) 10CH 2CHCH 2 C CH 3 C CH 3 I +

5 W high-pressure mercury lamp (Acticure Ultraviolet/Visible Spot Cure System, EXF Photonic Solutions, Inc.). The full-spectrum effective irradiance of the system was set to 100 mw/cm 2 as measured by a radiometer for the wavelength range of nm (R5000, EFS). The Raman spectrum of each acrylate and epoxide was acquired first to identify bands in the fingerprint region that could be used to calculate conversion (see Figure 1). The reactive band representing the acrylate C=C double bond is located at 1638 cm -1 and is associated with the C=C stretching vibrations; the reactive band representing the epoxide ring is located at 789 cm -1 and is associated with the asymmetric epoxide ring deformation. An internal reference band was selected at 605 cm -1, which represents the skeletal bending of the non-reactive acrylate carbonyl group. The peak areas under each band were integrated and used to calculate the concentration of related functional groups. Since the spectral baselines and the reference band intensity remained constant in these real-time reaction studies, the conversion of each functional group was calculated by rationing the peak area of the reactive band at any given time [A rxn (t)] to the peak area of the reactive band prior to illumination [A rxn (0)]: CCCCCCCCCCCCCCCCCCCC, αα = 1 AA rrrrnn (tt) (1) AA rrrrrr (0) epoxide ring HA-EEC BEA-EEC BCEA-EEC Arbitary acrylate C=C Raman shift (cm -1 ) Figure 1. Comparison of Raman spectra for epoxide-acrylate hybrid mixture systems before photopolymerization. The reactive band representing the acrylate C=C double bond is located at 1638 cm -1 ; the reactive band representing the epoxide ring is located at 789 cm -1. Results and Discussions The kinetics of neat monomer systems Kinetic studies were performed using Raman spectroscopy for each neat monomer system with only the free-radical photoinitiator (DMPA for acrylates) or the cationic photoinitiator (IHA for diepoxide). The conversion profiles of neat acrylate monomers with different secondary functional groups in their backbone are compared in Figure 2A. The rate of polymerization for neat BEA,

6 which contains ether groups, was higher than that for neat HA, which contains hydrocarbon tail groups only. Stable peroxy radicals are formed through hydrogen abstraction on the ether groups of BEA, accelerating the consumption of oxygen dissolved in the reaction system or diffused from the atmosphere, These peroxy radicals can then reabstract the hydrogen from monomers or polymer chains [13]. This recursive reaction efficiently eliminates the oxygen from the BEA polymerization system quicker, resulting in an increased polymerization rate. Since neat HA monomer has a low chance of hydrogen abstraction due to the absence of abstractable hydrogen in its backbone, its rate of polymerization is lower. However, BCEA, which contains urethane groups, showed the highest rate of polymerization of the three acrylates studied. The faster polymerization and higher reactivity of BCEA can be attributed to hydrogen abstraction from a two-carbon aliphatic spacer between the acrylic moiety and the secondary urethane group, as well as its higher viscosity than other two acrylates. With phenyl carbamate acrylates, hydrogen abstraction mostly occurs at the α carbon, which results in a considerably increased rate of polymerization, while hydrogen abstraction from the β carbon minimally influences the rate of polymerization [14]. In addition, the higher viscosity of BCEA (due to the intermolecular interaction via hydrogen bonding between urethane groups) may contribute to a decrease of radical termination, and thereby an increase in radical concentration, subsequently resulting in the early autoacceleration of the polymerization. Thus, the structure of the secondary functional groups considerably affects the reactivity of acrylate monomers during free-radical polymerizations. In comparison, neat EEC exhibited relatively slow cationic polymerization and low conversion compared to the neat acrylates (Fig 2A). The effect of acrylate structure on epoxide cationic polymerization Acrylate and epoxide functional groups of the respective acrylate and EEC monomers were monitored simultaneously using real-time Raman spectroscopy. Acrylate-epoxide hybrid systems with 1:1 molar ratio were compared to determine the effect of the secondary functional groups on EEC cationic polymerizations (Figure 2B). In the presence of EEC, the conversion profiles of the three acrylates showed the same trend as neat acrylates. The free-radical polymerization of the acrylates generally proceeded much quicker than the EEC cationic polymerization in the corresponding hybrid systems. The polymerization rate and conversion of EEC increased when a low viscosity HA was added to EEC relative to EEC neat system due to dilution effect of those acrylates. However, the reactivity of EEC also varied with the secondary functional groups of the acrylates used. Compared to HA-EEC system, the rate of polymerization and conversion of EEC are extensively influenced by the presence of BEA and BCEA. Comparison of BEA-EEC system with HA-EEC systems indicated that the ether group of BEA caused sluggish cationic polymerization of EEC in the hybrid systems. This phenomenon could be explained to an extent by a decrease in the diffusion of initiating protons generated from the photolysis of IHA via their fixation by the crown ether-like acrylate polymer chains. In the presence of a crown ether with the smallest ring, 12-crown-4, the retardation of EEC cationic photopolymerization was observed, with an accompanying decreased rate of polymerization [15]. In addition, in cationic polymerization systems with diols and polyethers with oxygen in the 1,2 position, protons can be fixed by such conformation of the crown ether-like structure [15]. In Figure 2B, EEC cationic photopolymerization was completely inhibited by the presence of BCEA, which contains a secondary urethane group. Although it is not yet clear how the urethane

7 structure interacts with cationic photopolymerization systems, protons may be fixated when nucleophilic alkoxy oxygen groups in the crown ether-like BCEA polymer chains surround them. In addition, the strong nucleophilicity of the carbonyl groups in the urethane structure is more likely to affect protons and growing cationic active centers, severely interfering with the initiation and cationic propagation of epoxide [16]. However, it is not clear whether, during the EEC cationic photopolymerization, the urethane amines on the acrylate monomers act as proton scavengers under moderately or highly acidic condition. Although some amine compounds can fully inhibit cationic photopolymerization depending on their pk b and oxidation potential [17], amines in the urethane structure are known to be electrophilic, as well as proton donors [18]. To determine the role of the amines in this system, further studies were performed and will be discussed later in this paper. (A) BCEA BEA HA Figure 2. profiles of (A) neat monomer systems and (B) hybrid systems of 1:1 ratio. [DMPA] = 32 mm for acrylates, and [IHA] = 32 mm for epoxide; photopolymerizations conducted at room temperature with effective irradiance = 100 mw/cm 2. The effects of acrylate-epoxide molar ratios on the kinetics of EEC diepoxide The effect of molar ratios of acrylate to epoxide in hybrid systems was evaluated using real-time Raman spectroscopy. In the presence of HA, EEC cationic photopolymerizations were not influenced, irrespective of their molar ratios varied from 2:1 to 1:1 to 1:2 (Figure 3A), although slightly decreased polymerization rate were observed with 4:1 HA-EEC system. In contrast, in the presence of BEA and BCEA, the rate of polymerization and conversions of EEC are significantly affected by acrylate-epoxide molar ratio (Figure 3B and 3C, respectively). As more BEA was added to the hybrid systems, lower polymerization rates and conversions of EEC were observed. With BCEA monomers, the EEC cationic photopolymerization was completely inhibited for 2:1 and 1:1 molar ratios, while it started to proceed to a low degree up to 10% conversion for the 1:2 molar ratio. To obtain higher EEC conversion, BCEA should be incorporated less than 5wt% (the BCEA-EEC ratio =63) for the given condition. Therefore, the suppression of the EEC cationic photopolymerizations is more pronounced in the presence of higher concentration of acrylates with secondary ether or urethane groups. The effects of photoinitiators on the kinetics of EEC diepoxide EEC In the previous sections, it was demonstrated that the kinetics of the acrylates used are not affected by the presence of epoxide and their molar ratios. The insensitivity of acrylate kinetics to the presence of epoxide can be attributed to the dual photoinitiator system containing DMPA and IHA, (B) s BCEA BEA HA EEC (HA) EEC (BEA) EEC (BCEA)

8 (A) (B) (C) EEC HA EEC BEA EEC BCEA EEC Figure 3. profiles of acrylate and epoxide during hybrid photopolymerizations: (A) HA EEC systems, (B) BEA EEC systems, and (C) BCEA EEC systems. The molar ratio of acrylate to diepoxide is labeled for the corresponding EEC conversion profiles. [DMPA] = 32 mm for acrylates, and [IHA] = 32 mm for epoxide. Photopolymerizations were conducted at room temperature with effective irradiance = 100 mw/cm 2. each initiating free-radical and cationic polymerizations, respectively. Although each photoinitiator is responsible for their respective polymerization mechanisms, they can also initiate their counter polymerizations [12]. For instance, it was verified that with comparable hybrid systems acrylate free-radical photopolymerizations can be initiated using the cationic PI (IHA) only due to the production of aryl radicals as a by-product of the cationic PI photolysis. Increasing the concentration of IHA relative to the acrylate concentration resulted in increased final acrylate conversion during acrylate-epoxide hybrid photopolymerizations [12]. In the case of dual photoinitiators, dissolved oxygen in the polymerizing systems could be further consumed by the IHA-based aryl radicals at initial stage, leaving more of the reactive DMPA-based radicals to initiate polymerization. Hence, the dual photoinitiator can accelerate the acrylate free-radical photopolymerization due to the increased concentration of primary radical species from both DMPA and IHA PIs. Although the free-radical PI (DMPA) cannot initiate epoxide cationic polymerizations since the photolysis of DMPA does not produce protonic acids, radical products of the radical PI photolysis can induce cationic polymerizations. In this so-called radical-promoted cationic polymerization mechanism, one of the primary radical products generated from the DMPA photolysis reduces the cationic PI, resulting in a carbocationic active center [16, 17]. Neat EEC monomer systems were examined to determine how the DMPA-IHA molar ratios affect the cationic photopolymerization kinetics. The DMPA-IHA molar ratios were varied from 2:1 to 1:1 to 1:2, which correspond to the EEC partial volumes used in the hybrid systems above. In Figure 4, the highest polymerization rate and conversion of EEC occurred with the 2:1 DMPA-IHA ratio. These increases can be ascribed to the faster photolysis of the radical PI (DMPA) than the cationic PI (IHA), resulting in more radical-promoted cationic polymerizations for the higher DMPA-IHA ratio. However, DMPA did not considerably influence the kinetics of EEC when DMPA-IHA molar ratios were 1:1 or 1:2. A positive

9 correlation could exist between the concentration of each PI and the photolysis rate of both PIs during illumination, since they undergo photon absorption competition in the spectral range of nm. Thus, the kinetic behavior of EEC with different acrylates shown in Figure 3 can be further understood by considering these results from varying the DMPA-IHA molar ratio. With higher acrylate concentrations in the hybrid systems, the relative DMPA-EEC concentration results in the higher DMPA-IHA molar ratio. In theory, the polymerization rate and conversion of EEC in the hybrid systems should increase, although primary radical species generated from DMPA photolysis are also consumed by acrylates. However, this increase is not seen when ether/urethane secondary functionalities are incorporated to the acrylates (Figures 3B and 3C, respectively). These results underscore the importance of carefully selecting acrylate structure and acrylate-epoxide molar ratio for the formulation of hybrid photopolymerization systems. 2: :1 0.3 IHA only 1:2 DMPA:IHA Figure 4. profiles of neat epoxide (EEC) with various ratios of [DMPA] to [IHA]. The molar ratio of [DMPA] to [IHA] is labeled for the corresponding EEC conversion profiles: [DMPA]=32mM and [IHA]=16mM for 2:1 ratio; [DMPA]=[IHA]=32mM for 1:1 ratio; and [DMPA]=16mM and [IHA]=32mM for 1:2 ratio. Photopolymerizations were conducted at room temperature with effective irradiance = 100 mw/cm 2. The effect of urethane structure on the kinetics of EEC diepoxide The complete inhibition of EEC cationic photopolymerization by the urethane functionality was further understood by incorporating a non-reactive compound containing ester carbonyl groups (EH) with the EEC diepoxide formulations. Here, the amine groups in the urethane structure were excluded to simplify the analysis. The concentration of DMPA was calculated based on the EH concentration in the EH-EEC mixtures. Similar to the acrylate-eec hybrid systems, the DMPA-IHA molar ratios ranged from 2:1 to 1:1 to 1:2 in EH-EEC mixture systems. Figure 5 shows the decreased EEC polymerization rate in the presence of EH, compared to neat EEC containing the dual photoinitiator system. The concentration of primary radical species generated from the DMPA photolysis should be much higher for the 2:1 EH-EEC system since they are not consumed by the non-reactive EH. However, the expected increase in rate due to radical-promoted cationic photopolymerization was not observed even for the 2:1 EH-EEC system with the 2:1 DMPA-IHA

10 molar ratio. It strongly indicated that the decreased EEC polymerization rate is most likely due to the higher nucleophilic and basic character of the ester carbonyl groups in EH [18, 19]. This basicity would lead to the decrease of available protonic acids for epoxide chain propagation by protonating the ether and carbonyl groups of EH. The extent of this effect was such that there was no difference in the degree of the decreased EEC polymerization rate for any of the EH-EEC systems studied. In addition, when comparing the EH-EEC systems to the 2:1 BEA-EEC system, the negative effect of the EH ester carbonyl groups on EEC cationic photopolymerization is more severe than that of the BEA ether groups, especially considering there is no formation of ether-crown like structure with the non-reactive EH during polymerizations (Figure 5). Thus, the inhibition of EEC cationic photopolymerization is more likely to be affected by carbonyl groups with higher nucleophilicity and basicity than ether groups. The role of amine groups in the BCEA secondary urethane structure was also evaluated by predicting pk a values of conjugated acids for amines and ester carbonyl groups of BCEA using SPARC n-line Calculator V4.5 ( The SPARC n-line Calculator combines linear free energy calculations with perturbation theory and structure activity relationships to predict ionizing pk a values [20, 21]. Based on the predicted pk a values of conjugated acids for amines and ester carbonyl groups, the amine groups are quantitatively 2.5 times more basic than carbonyl groups. These results provide a strong support for why EEC cationic photopolymerizations were greatly inhibited in the presence of BCEA. In these BCEA-EEC systems, the combined properties of the carbonyl ester groups and amine groups result in complete inhibition of EEC cationic photopolymerizations through the scavenging and fixing of protons or carbocation active centers (Figures 4C and 5) Neat EEC (DMPA:IHA 2:1) 2:1 BEA:EEC 2:1 1:1 1:2 EH:EEC 0.1 2:1 BCEA:EEC Figure 5. profiles of epoxide (EEC) in the presence of ethyl hexanoate (EH). The molar ratio of EH to EEC is labeled in the plot. [DMPA] = 32 mm for EH, BEA and BCEA; [IHA] = 32 mm for EEC; [DMPA]=[IHA]=32mM for neat EEC. Photopolymerizations were conducted at room temperature with effective irradiance = 100 mw/cm 2. Conclusion This research has demonstrated that the kinetic of the epoxide cationic photopolymerizations with the presence of the acrylates used in this study are significantly controlled by the structure of

11 the acrylate secondary functional groups as well as acrylate-epoxide molar ratios. A low viscosity acrylate with a hydrocarbon tail reduced the viscosity of the epoxide-acrylate hybrid systems. This reduced viscosity allowed epoxide cationic polymerizations to be much faster due to its dilution effect, compared to neat epoxide systems. However, acrylates containing ether or urethane groups negatively affected the epoxide kinetics. A low viscosity acrylate with ether groups was not effective in providing dilution effect for the epoxide as acrylate-epoxide ratios increased. This was attributed to a decrease in the diffusion of initiating protons generated from the cationic PI photolysis via their fixation by the crown ether-like acrylate polymer chains. In addition, the synergy of a dual photoinitiator at higher DMPA-IHA ratio on the cationic polymerization was negligible in the presence of acrylates with the secondary ether or urethane groups. The role of the secondary urethane groups was determined by separately characterizing the ester carbonyl and amine groups in the urethane structure. As a result, the nucleophilicity and basicity of carbonyl groups greatly affected the epoxide cationic polymerizations, resulting in the decreased polymerization rates and conversions. Quantitatively, the basicities of the carbonyl and amine groups were compared using predicted pk a values of their conjugated acids, strongly suggesting that amine groups are 2.5 times more basic than carbonyl groups. Thus, the combined properties of the ester carbonyl groups and amine groups result in complete inhibition of the epoxide cationic photopolymerizations through the scavenging and fixing of protons or carbocation active centers. Moreover, this detrimental effect caused by the secondary ether or urethane groups becomes worse as the acrylate-epoxide molar ratio increase. Therefore, acrylate secondary functional groups and acrylate-epoxide compositions should be taken into account when tuning epoxide conversion and ultimate strength of hybrid IPNs. These findings provide fundamental guidelines to optimize industrially relevant formulations and reaction conditions for thin-film and coating applications using epoxide-acrylate hybrid photopolymerizations. Acknowledgements This material is based upon work supported by the National Science Foundation under Grant No We would like to acknowledge Union Carbide for providing the materials used in this study. References [1] Fouassier JP. Photoinitiation, Photopolymerization, and Photocuring. Munich, New York: Hanser Publishers; [2] Fisher JP, Dean, D., Engel, P. S., Mikos, A. G. Photoinitiated polymerization of biomaterials. Annual Review of Materials Science 2001;31: [3] Gou L, pheim B, Scranton AB. The effect of oxygen in free radical photopolymerization. Recent Research Developments in Polymer Science 2004;8: [4] Christophersen AG, Jun H, Jørgensen K, Skibsted LH. Photobleaching of astaxanthin and canthaxanthin: quantum-yields dependence of solvent, temperature, and wavelength of irradiation in relation to packageing and storage of carotenoid pigmented salmonoids. Z Lebensm Unters Forsch 1991;192: [5] Gou L, Coretsopoulos CN, Scranton AB. Measurement of the dissolved oxygen concentration in acrylate monomers

12 with a novel photochemical method. Journal of Polymer Science Part A: Polymer Chemistry 2004;42: [6] Cai Y, Jessop JLP. Decreased oxygen inhibtion in photopolymerized acrylate/epoxide hybrid polymer coatings as demonstrated by Raman Spectroscopy. Polymer 2006;47: [7] Hartwig A, Schneider B, Luhring A. Influence of moisture on the photochemically induced polymerisation of epoxy groups in different chemical environment. Polymer 2002;43: [8] Decker C. Kinetics Study and New Applicatioins of UV Radiation Curing. Macromolecular Chemistry and Physics 2002;23: [9] Chen Z, Zhang Y, Chisholm BJ, Webster DC. A humidity blocker approach to overcoming the humidity interference with cationic photopolymerization. Journal of Polymer Science Part A: Polymer Chemistry 2008;46: [10] Decker C, T Nguyen Thi Viet, D Decker, Weber-Koehl E. UV-radiation curing of acrylate/epoxide systems. Polymer 2001;42: [11] Sperling LH, Mishra V. The Current Status of Interpenetrating Polymer Networks. Polymers for Advanced Technologies 1996;7: [12] Eom HS, Jessop JLP. Epoxide/urethane acrylate hybrid photopolymerizations to minimize atmospheric sensitivity: Understanding the interplay between viscosity and dual photoinitiator system. In: J. P. Fouassier,X. Allonas editors. Basics and Applications of Photopolymerization Reactions; accepted 05/31/09. [13] Lee TY, Guymon CA, Jonsson ES, Hoyle CE. The effect of monomer structure on oxygen inhibition of (meth)acrylates photopolymerization. Polymer 2004;45: [14] Beckel ER, Stansbury JW, Bowman CN. Effect of Aliphatic Spacer Substitution on the Reactivity of Phenyl Carbamate Acrylate Monomers. Macromolecules 2005;38: [15] Hartwig A, Koschek K, Luhring A, Schorsch. Cationic polymerization of a cycloaliphatic diepoxide with latent initiators in the presence of structurally different diols. Polymer 2003;44: [16] Braun H, Yagci Y, Nuyken. Copolymerization of butyl vinyl ether and methyl methacrylate by combination of radical and radical promoted cationic mechanisms. European Polymer Journal 2002;38: [17] M. Degirmenci YH, Y. Yagci. ne-step, one-pot Photoinitiation of Free Radical and Free Radical Promoted Cationic Polymerizations. Journal of Applied Polymer Science 2002;85: [18] James VC, Ulrike V. Mechanistic study of the reactivity of 3,4-epoxycyclohexylmethyl 3prime,4prime-epoxycyclohexancarboxylate in photoinitiated cationic polymerizations. Journal of Polymer Science Part A: Polymer Chemistry 1995;33: [19] Crivello JV, Narayan R. Novel Epoxynorbornane Monomers. 2. Cationic Photopolymerization. Macromolecules 1996;29: [20] Hilal SH, Karickhoff SW, Carreira LA. A Rigorous Test for SPARC's Chemical Reactivity Models: Estimation of More Than 4300 Ionization pka. Quantitative Structure-Activity Relationships 1995;14: [21] Sipil J, Nurmi H, Kaukonen AM, Hirvonen J, Taskinen J, Yli-Kauhaluoma J. A modification of the Hammett equation for predicting ionisation constants of p-vinyl phenols. European Journal of Pharmaceutical Sciences;25: