- Top of page
- Results and Discussion
- Supporting Information
Plant oils and the thereof derived fatty acids are important renewable raw materials for polymer chemistry.1 Especially castor oil is a very versatile renewable feedstock for all kinds of polymeric materials, including, e.g., polyesters, polyamides, polyurethanes, and many others.2 One of the commercially available castor oil derived platform chemicals, 10-undecenoid acid 1,2, 3 was recently used to prepare polyamides X,20,4 a variety of acyclic diene metathesis derived polyesters,5–7 as well as different cross-linked materials,8–10 thus demonstrating its broad range of application possibilities. Moreover, methyl 10-undecenoate 2 was shown to be a suitable starting material for the preparation of α,ω-bisfunctional fatty acids for polycondensation polymers via cross-metathesis with methyl acrylate,11 acrylonitrile,12 and allyl chloride.13 Within this contribution, we will discuss a completely complementary approach to α,ω-bisfunctional fatty acids using thiol-ene additions to 2, and use the thus resulting materials as renewable monomers for polyester synthesis.
It is important to note that the addition of thiyl radicals to olefins (Equation 2) is reversible and can thus lead to cis/trans isomerization of the starting material.16 The extent of reversibility of CS bond formation with terminal olefins, with olefins that lead to resonance stabilized free-radical intermediates, and with monosubstituted alkynes is much less pronounced than that of the addition to olefins with non-terminal double bonds.15 Thus, terminal double bonds, as in 2, are more reactive than internal ones.16 Finally, the formed C-radical reacts with a thiol molecule to give the final product together with a new thiyl radical (Equation 3) in the rate determining step of the reaction.15
Even if these reactions are well understood, only the introduction of the concept of click chemistry by Sharpless and coworkers17 led to a revival of this efficient functionalization method. Today, thiol-ene additions (or thiol-ene click reactions) are considered as a versatile and broadly applicable tool in polymer science.18 Nevertheless, already early on it was discovered that dithiols and dienes are suitable monomers for polymer synthesis.19 More recent investigations have shown that quite high molecular weight polythioethers can be obtained in that way,20 that initiators are not necessary for the reaction to proceed,21 but radical inhibitors suppress the polymerization for some time.21 Today, the reaction is used in the sense of a click reaction for the synthesis of dendrimers,22 the grafting of side-groups to reactive polymers,23, 24 or the synthesis of star polymers.25
Also in the field of oleochemistry thiol-ene additions are known for a long time.26, 27 More recently, Johansson and coworkers28 studied the kinetics of photo-initiated thiol-ene addition to methyl oleate and methyl linoleate using trifunctional thiols. In the same sense, allyl-, acrylate-, as well as vinyl ether derivatives of castor oil were used together with multifunctional thiols for the preparation of UV curable resins.29 Moreover, Bantchev et al.30 investigated thiol-ene addition reactions of canola and corn oil with butanethiol to yield sulfur-containing compounds with possible applications as lubricants to improve wear and friction properties. Very recently, thiol-ene chemistry was also applied for the preparation of bio-based telechelics31 and soy-based thiols and enes were formulated with allyl triazine resulting in tack-free coating films after UV curing.32 No detailed studies on the preparation of monomers derived from fatty acids and/or their subsequent polymerization is described in the literature. Therefore, we describe here for the first time the use of methyl-10-undecenoate 2 as a castor oil derived renewable platform chemical in the context of thiol-ene click reactions. A variety of renewable monomers was thus obtained in high yields, their polymerization was subsequently studied and the material properties of the resulting polyesters were investigated. It is important to note that the presented results offer yet another efficient and sustainable entry to renewable raw materials derived from plant oils and are complementary to recently reported results on olefin cross-metathesis of fatty acid derivatives to yield renewable monomers.33
Results and Discussion
- Top of page
- Results and Discussion
- Supporting Information
In a starting set of experiments we examined the synthesis of renewable monomers from 2via thiol-ene additions with different thiols (Scheme 1). First, a set of model reactions was performed in order to investigate some important parameters concerning the reaction conditions, especially the reaction temperature. The reaction temperature for the synthesis of M1 had a marginal effect on the observed conversions of 2 and essentially full conversions were obtained after 2 h reaction time for the performed small scale reactions. Only above 40 °C conversions were somewhat lower (≈80% after 2 h). Therefore, 35 °C was considered as an optimum temperature for the synthesis of M1 and also M2.
Quite interestingly, although 2 and 1-thioglycerol are not miscible at this temperature, the reaction still proceeds and in time the mixture becomes homogeneous, probably due to the amphiphilic nature of the formed product. All these reactions were performed under vacuum since deoxygenated reaction mixtures showed better results. However, since methyl thioglycolate has a low boiling point of 42 °C, the synthesis of M3 was performed at room temperature without the application of vacuum, also resulting in high conversions. M4 and M5 on the other hand had to be prepared at 60 and 70 °C, respectively, in order to avoid crystallization of the products and high viscosities. Up to that time all reactions were carried out using a 1:1 ratio of thiol and the ene reacting groups. Having the basic reaction parameters established, we also performed thiol-ene addition reactions using 1.1 and 1.2 equiv. of the thiol compounds. For M4 and M5 we investigated an excess amount of the ene compound (e.g., 2:butanedithiol = 1.2:0.5) in order to avoid the formation mono-addition-products. The results clearly showed that a 1:1 ratio of reactive groups provided similar or better results than the respective reactions with excess of one of the reactants. Therefore, we carried out all monomer syntheses with a feed ratio of ene:thiol = 1:1. We then went on to upscale the reactions and had to note that much longer reaction times were required (see Experimental Section in the Supporting Information). However, later on, we observed that the prolonged reaction times were due to inefficient stirring of the large scale reactions. For instance, the synthesis of M1 at 10 and 20 mmol scale of 2, and three different magnetic stirrers (cross and elliptical) provided unsatisfactory results (≈50% conversion after 8 h), but the same reaction performed with a mechanical stirrer (half-moon, ≈400 rpm) resulted in a satisfactory conversion of 92% after 8 h. Thus, the obstacle of long reaction times for large scale reactions can easily be overcome by using efficient mechanical stirring.
The thus obtained monomers were then polymerized using TBD as a catalyst, which is known to be a highly effective catalyst for the trans-esterification of plant oils and others.34 The test polymerizations of M3 and M4, both with M5, were carried out at 120 and 140 °C, and with 5 and 10 mol-% of TBD per ester group of the monomers. According to GPC results obtained from these tests, the most suitable conditions for polymerization were 120 °C and 5 mol-% catalyst. Therefore, all further polymerization reactions were performed at these conditions and with continuous vacuum to efficiently remove the released methanol.
Polymer P1 is different from all other prepared polymers, since it produces a hyperbranched polymeric architecture. A first polymerization reaction (Table 1, entry 1) gave polymer with considerable molecular weight, which is most likely underestimated by GPC due to its branched structure. We then investigated the use of glycerol as a core molecule and observed that the addition of small amounts of glycerol increased the observed molecular weight values (Table 1). The further addition of more glycerol however decreased the observed molecular weights, as expected from theory.35
The thus obtained hyperbranched polymers were also characterized by 1H-NMR to obtain their degree of branching (DOB) by comparison of the integral of the signals for linear and branched units. As discussed in the literature,36 the DOB is commonly calculated according to the following equation:
where D, T, and L are the fractions of dendritic, terminal, or linearly incorporated monomers in the resulting hyperbranched polymers. The values commonly reported for DOB are in the range of 0.4–0.8. Here, we observed DOBs between 0.40 and 0.47 (see Table 1). As shown in Table 1, the DOB slightly increased with the addition of the core molecule, but remained relatively constant with the addition of higher amounts of the core molecule.
The linear polymers derived from M2 to M5 (P2 derived from M2; P35 derived from M3 and M5 and P45 derived from M4 and M5) showed moderate to good molecular weights and unimodal molecular weight distributions (compare Figure 1 and Table 1). Concerning the 1H-NMR spectra of the linear polymers, we did not observe any side products and all peaks were easily assigned (compare Supporting Information). Moreover, end-groups were only observed for P2, but their integrals were too small for a reliable end-group calculation. This indicates high molecular weights of P2 and especially of P35 and P45.
Since the aim of this study was to obtain plant oil derived renewable polymers and to substitute materials synthesized from petroleum-originated monomers, it is also very important to investigate the thermal behavior and stability of the prepared polymers. DSC analysis of the polymers at a heating rate of 10 °C · min−1 revealed sharp melting endotherms (Table 1, Figure S1 Supporting Information) thus putting the materials in the class of semicrystalline polymers, similar to other thio-ether containing polyesters.37 Figure S1 shows multiple melting points for all polymers, except for P45. This phenomenon for sulfide-ether-containing homopolyesters was already described in the literature38 and ascribed to a melting and recrystallization processes occurring during the calorimetric run.39
However, Figure S1 clearly reveals the melting points of P35 and P45 as 61.5 and 71.3 °C, respectively. The higher Tm of P45 is due to the longer chain length of M4 (if compared to M3) and a more regular structure of P45 (spacing of the ester groups), both allowing a better crystallization of P45. As already mentioned, P1 and P2 exhibited very sharp double melting endotherms and we investigated this effect in more detail by DSC. It could be expected that a lower heating rate would allow the polymers to rearrange themselves to a single crystalline morphology. However, DSCs performed at a heating rate of 5 °C · min−1 still showed two melting points, and, in addition, pronounced exothermal peaks (higher than the base line) between the two endothermal peaks. This let us conclude that P1 and P2 first undergo a melting process. This causes an increase in chain mobility leading to the observed crystallization, which is followed by a second melting. The melting temperatures detected from this experiment were 37.9 and 50.1 °C for P1 and 59.1 and 67.2 °C for P2 (all data from the second heating run). Subsequently, annealing was investigated as a pre-treatment of P1 and P2. Thus, samples were heated inside the calorimeter to different temperatures, kept at this temperature for 2 h, subsequently cooled to −50 °C and finally heated to 150 °C with 20 °C · min−1 heating rate. The annealing temperatures chosen for P1 are 25 and 40 °C, and for P2 50 and 56 °C. The thermograms obtained from these experiments are presented in Figure 2. It is obvious that annealing the samples had a pronounced influence on the thermal behavior of P1 and P2. Considering the relative intensities of the endotherms, annealing of P1 at 25 °C favored the first endotherm peak, but annealing at 40 °C resulted in a polymer with single type crystalline structure which has a single, broad melting point. Similarly, annealing P2 at 50 °C also favored the first peak, but annealing at 56 °C resulted in one sharp endotherm peak at the same temperature of the initial second peak. In both cases the single peak appeared at the temperature of the second peak before annealing was applied, indicating that the lower temperature endothermic peaks were a result of meta-stable states40 and ultimately the polymers reorganized to more stable crystal structures upon annealing. The discussed thermal properties are summarized in Table 1 and the inset of Figure 2.
Finally, the thermal stabilities of the prepared polymers were studied by TGA under nitrogen atmosphere with 10 °C·min−1 heating rate. The obtained results are summarized in Table 1. The linear polymers possess good thermal stability showing 5% weight loss between 344 and 353 °C. The hyperpranched polymer P1 showed a somewhat smaller, but still good, thermal stability with 5% weight loss occurring at 299 °C. This could be a result of the lower molecular weight of P1 and its branched structure with a significantly higher number of end-groups. A similar trend was observed for the maximum weight loss temperatures (compare Table 1).