Synthesis of UV-Curable yperbranched Urethane Acrylates Ahmet ebioglu, James. Aerykssen, R. David Zopf, Igor V. Khudyakov, Bomar Specialties, Torrington, USA 1. Introduction yperbranched, highly branched, dendritic, arborescent polymers are synonyms or almost synonyms. These polymers/oligomers attract much attention in the last decades due to the interesting properties resulting from their architecture and a high number of functional groups. They are usually prepared in one pot which leads to a one or another molecular weight distribution (MWD). 1 Monodisperse (MWD=1.0) hyperbranched polymers with a defined architecture are dendrimers. Usually dendrimers are prepared in a rather complex multistage synthesis and cost ~$100/g. UV-curable urethane acrylates (UA) are widely used in the coating industry. We expect dendritic UA and UA dendrimers of acceptable cost to be promising new oligomers for coatings formulations. The present work is devoted to synthesis of such UA oligomers and their properties. 2. Synthesis of urethane acrylates Common UA oligomers are prepared usually in two stages: the first is a reaction of polyol (P) with diisocyanate (D) taken in an excess. The residual isocyanate groups of a prepolymer are capped with hydroxyl substituted acrylate or methacrylate. The most common in the industry are the known EA and EMA. 2 This synthesis is called direct addition. 3 A synthesis which starts with a reaction between D with hydroxy acrylate followed by a reaction with P as a second stage is called reverse addition. 3 Perstorp manufactures multifunctional polyols Boltorn which find a wide application in the coatings industry. It seems obvious to run a direct synthesis with Boltorns, get a multifunctional prepolymer and cap it with acrylate. owever, in our experience a gelation occurs in the case of a reaction of D (functionality f C = 2) with hyperbranched polyol P of f 4 due to branching and chain extension. It is known that the probability of gelation p is proportional to product of functionalities of monomers (oligomers) and inversely proportional to a ratio r of a number of equivalents of a reagent in excess to a deficient reagent: 4 p ~ f *f C /r (1) ur goal was to find ways to get hyperbranched UA oligomers. 3. Extreme reverse addition It is apparent that using monofunctional isocyanates (f C = 1) avoids gelation via chain extension and branching. Taking the reverse addition scheme described above to its extreme, we prepared a monofunctional isocyanate (f C =1.0) by reacting EA with an unsymmetrical D having different reactivity of -C groups. 5 The reaction conditions were structured making all attempts to get predominately monofunctional isocyanate acrylate of a generic structure EA-D. As a byproduct we got the D-EA-D adduct which often has beneficial properties in formulations or at least does not have an evident negative impact.
We used the known in the industry D = IPDI. Surprisingly, a reduction of temperature from ambient to 0 o C did not lead to an increase in the relative concentration of IPDI-EA. owever, the addition of the known catalyst DBTDL led to a yield of IPDI-EA of ~75-80% with the remainder being the EA-IPDI-EA byproduct. In this way a formal C functionality in the system becomes f C = 1. We made all efforts to get the maximum possible yield of IPDI-EA in a reaction of EA with IPDI and leave negligible amount of non-reacted IPDI. Capping polyols of high functionality with such isocyanate/adduct mixture allowed us to avoid gelation and get UVcurable oligomers useful for industrial applications. This method we named the extreme reverse addition. The properties of an oligomer prepared by this method and named XDR-1406 are presented in the Table 1: Table 1 Properties of the oligomer XDR-1406 prepared by the extreme reverse addition* Acrylate Functionaliy strength, Elongation to break, % ardness MEK Double rubs η,** P 8.4 61 4 1834 87D 200 138 * Determination error of values presented in the Table is 10-15% ** η is the viscosity of the formulation with diluent at 25 o C 4. Reactions with monoisocyanate-acrylates 6. f course, using a pure monofunctional isocyanate-acrylate would be better suited for capping high functionality polyols than an EA-D/EA-D-EA mixture. Such monofunctional isocyanates-acrylates are produced by Showa Denko, and their names are abbreviated as AI and MI, cf. below: C 2 C C C 2 C 2 C AI C 3 C 2 C C C 2 C 2 MI C The one-stage capping of any hyperbranched P by AI/MI will lead to a hyperbranched acrylate of the same functionality as that of P. owever, notice that such UA oligomers will have twice less urethane (carbamate) links than UA capped by a standard way by D and by EA/EMA. It is well-known that hard segments provided by D in polyurethanes are important. In combination with soft segments of P, hard segments provide the remarkable properties of polyurethanes (resilience, toughness, high elongation, abrasion resistance, etc. 5 ) Before moving directly to high functionality polyols, we explored the effect of using AI/MI by reacting them with common difunctional polyols (f = 2) of varying molecular weights (M W ). We then compared these products to the corresponding UAs prepared in the standard way. As seen in the Table 2, regardless of the M W of P, AI/MI-capped oligomers have much lower viscosities than standard UAs of similar structure. This property is believed to be attributable,
first, to the absence of any chain extension. AI/MI-capped UAs have only one P molecule in their structure, and thus lower molecular weights than standard UAs. Second, as mentioned above, AI/MI-capped UAs have one-half the number of links that are present in standard UAs. The reduced number of these links leads to reduced hydrogen bonding between UA molecules and therefore reduced viscosity. Surprisingly, these same factors contribute to marked differences in final properties of the cured oligomers depending on the M W of P. As depicted in Figure 1, the tensile strength of the cured oligomer is drastically reduced when the M W of P is increased to above 500-650 g/mol. The elongation to break of AI/MI-capped oligomers is also lower than the standard UAs, especially with the higher molecular weight polyols. Both differences are likely due to the altered combination of hard and soft segments within the oligomer chain. Properties of cured formulations of oligomers named 4-1 thru 4-5 * Strength, Elongation to break, % Table 2 M w of P D ** / Methacrylate ***, P ligomer 4-1A 250 MI 2.3 41.6 5 1090 ligomer 4-1B 250 DesW/EMA 1,030 56.4 6 1335 ligomer 4-2A 650 MI 3 5.8 70 7.6 ligomer 4-2B 650 DesW/EMA 150 32.3 98 35.2 ligomer 4-3A 1000 MI 4.3 2.0 38 6.3 ligomer 4-3B 1000 DesW/EMA 235 35.3 245 9.4 ligomer 4-4A 2000 MI 10 0.5 18 3.9 ligomer 4-4B 2000 DesW/EMA 500 16.8 353 2.4 ligomer 4-5A 2900 MI 24 1.8 176 0.2 ligomer 4-5B, 2900 DesW/EMA 4150 17.2 441 2.2 * Determination error of the presented values is 15% ** DesW stands for the known in the industry D Desmodur W, a.k.a. 12 MDI 5 *** is the viscosity of an uncured formulation with diluent Strength () 60 50 40 30 20 10 0 AI capped UA Standard UA 0 500 1000 1500 2000 2500 3000 M w, g/mol Figure 1. Dependence of a tensile strength of cured oligomers on M w of diol
5. Products of reactions of high functionality polyols with MI 6 With some understanding of AI/MI based UAs, we moved on to the capping of hyperbranched polyols with f 4. Table 3 below presents properties of the multifunctional polyols used in the present examples: Table 3 Properties of high f polyols * Boltorn P1000 Boltorn P500 Boltorn 2004 CAPA 4101 f 14 16 6 4 M W 1313 1048 3960 1613 M n 458 363 2017 1284 MWD 2.87 2.89 1.96 1.26 * Determination error of M W, M n and of MWD is 15% From these starting polyols, the following oligomers were produced: Boltorn P1000 capped with MI, wherein 80% of the -groups, on an equivalent basis, were capped (designated ligomer 5-1); Boltorn P500 capped with MI, wherein 70% of the -groups, on an equivalent basis, were capped (designated ligomer 5-2); Boltorn 2004 capped with MI, wherein 95% of the -groups, on an equivalent basis, were capped (designated ligomer 5-3), and CAPA 4101 capped with MI, wherein 100% of the -groups, on an equivalent basis, were capped (designated ligomer 5-4). Properties of the liquid and cured oligomers are summarized in the Tables 4 and 5: Table 4 Properties of oligomers prepared with high functionality polyols and MI * ligomer 1-1 ligomer 1-2 ligomer 1-3 ligomer 1-4 M W 2374 2213 4378 2267 M n 880 875 2270 2020 MWD 2.70 2.61 1.93 1.12 @ 25 C, P** Color (APA) 310 1060 360 140 * Determination error of M W, M n, MWD and of is 15% ** η is the viscosity of undiluted, neat oligomer 10 0 Light yellow, slight haze 0 Table 5 Properties of cured formulations prepared with high functionality polyols and MI * strength, Elongation to break, % ardness MEK double rubs @ 25 C, P ligomer 5-1, 46 13 1,190 85 D 200 13 ligomer 5-2, 58 5 1,560 87 D 200 27 ligomer 5-3, 15 39 113 66 D 8 13 ligomer 5-4, 37 10 901 82 D 200 12 * Determination error of the values presented is 15% ** is the viscosity of a formulation with diluent In a sense, these UAs combine the best properties of the oligomers described in Sections 3, 4 above. They are not diluted by any unintended byproducts and there is little to no chain extension evidenced by the near identical MWDs of the starting P and the final UA. Like the oligomers prepared with diols, these oligomers are of quite low viscosity. Yet despite having a reduced number of urethane links, the higher functionality, and therefore expected high crosslink
density of the cured films, imparts exceptional tensile and hardness properties to these oligomers. These properties are quite similar to the oligomer XDR-1406 (Section 3) which contains the typical number of two urethane links per capped hydroxyl of the polyol. 6. Dendrimer [G1]-ene 6 6 ur work presentation thus far has focused on hyperbranched UAs with poorly defined structure and relatively high MWD. In this section we describe the preparation of a true urethane acrylate dendrimer almost monodispersed (MWD = 1.07) and with acrylate functionality f =6. We managed to get it in two simple stages. First we reacted 1,3,5-tris(2- hydroxyethyl)isocyanurate triacrylate (SR-368 of Sartomer) with 1-thioglycerol of Evans- Chemetics. We found that unsaturated compounds and especially acrylates participate in dark reactions of addition with different primary thiols. 7 MR of this Michael addition product confirmed that it occurs in a manner identical to a free-radical mechanism, that is in an anti- Markovnikov way. 7 In a second stage we reacted the obtained hexol with MI to produce UA of the following structure: S S S Figure 2. exafunctional urethane acrylate dendrimer [G1]-ene 6 The dendrimer was diluted with 50% TRPGDA and cured the usual way. The properties reproduced in the Table 6 below show the material to be hard, strong, and highly temperature resistant. A photochemistry approach to synthesis of a similar dendrimer was used in ref. 8.
Table 6* Temp. @ 2% wt. loss, C strength, Properties of the cured and liquid [G1]-ene 6 Elongation Durometer to break, % ardness MEK double rubs η** @ 25 C, P T g, C 292 51 3 2013 87 D 200 13 67 and 159 * Determination error of the presented values is 15% ** is the viscosity of the liquid formulation with diluent 7. Conclusions The goal of this work was capping of high functionality (f 4) dendritic polyols P with AI (MI) and getting UA oligomers. Syntheses were done a way which avoided the almost inevitable gelation typical of reactions of dendritic polyols with common diisocyanates. This enabled valuable hyperbranched urethane acrylate oligomers to be produced in one stage. In addition to common applications of dendritic UA as UV-curable protective coatings, the unique dendritic structure opens up opportunities for nanotechnology applications based on specific confinement of functional units and the formation of pores and cavities. We expect that a new dendrimer of modest cost will be useful for such purpose. 8. References 1. Voit, B.I.; Lederer, A., Chem. Rev. 2009, 109, 5924. 2. The examples given below are valid for both acrylates and methacrylates. We specify what is what when necessary. Also, we will use known in the industry abbreviations of other common P and D. 3. Swiderski, K.W.; Khudyakov, I.V. Ind. Eng. Chem. Res. 2004, 43, 628. 4. iemenz, P.C.; Lodge, T.P. Polymer Chemistry, CRC Press, Boca Raton, 2007, ch. 10. 5. Szycher, M. Szycher s andbook of Polyurethanes; CRC Press: Boca Raton, 1999, ch. 4,6. 6. Leon, J. A., ebioglu, A., Aerykssen, J.., Khudyakov, I.V.; Zopf, R. D. Patent is pending. 7. Aerykssen, J..; Leon, J.A.; Zopf R.D.; Khudyakov, I.V. Proc. RadTech Europe Conference, ice, 2009. 8. Killops, K.L.; Campos, L.M.; awker, C.J. J. Amer. Chem. Soc. 2008, 130, 5062.