2022 Volume 70 Issue 5 Pages 359-361
Despite the long history of the ene reaction between 1,2,4-triazoline-3,5-diones (TADs) and alkenes, its efficiency has always been hampered by competing side reactions, including the overreaction of ene adducts. In this communication, we demonstrate that this inherent limitation can be overcome by using 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) as the solvent. HFIP uniquely facilitates the desired azo–ene process between alkene with TADs presumably through hydrogen-bonding interactions. In addition to TADs, diethyl azodicarboxylate is a competent azo compound that undergoes a sluggish ene reaction with terminal alkenes.
Allylic C–H amination is a powerful method for the streamlined synthesis of valuable allylic amine derivatives.1) Among these methods, azo–ene reaction, where the reaction between alkenes and azo compounds affords allylic hydrazine derivatives, has served as a unique alternative to well-established transition metal-catalyzed technologies2) (Chart 1A). Since the initial report by Diels and Alder in 1926,3) several azo compounds have been utilized to facilitate the otherwise sluggish ene reaction of conventional azo compounds such as diethyl azodicarboxylate (DEAD).4) For instance, Leblanc et al. discovered that bis(2,2,2-trichloroethyl) azodicarboxylate was a very reactive azo compound that underwent the ene reaction under mild conditions.5) 1,2,4-Triazoline-3,5-diones (TADs) are among the most reactive azo compounds that rapidly react with various unsaturated hydrocarbons including alkenes,6) dienes,7) and arenes.8,9) Owing to their markedly high reactivity, TADs are extensively applied in the realm of chemical sciences spanning from polymer chemistry to biochemistry.10) Although remarkable, their overwhelming reactivity represents several challenges in the ene reaction with a specific class of alkenes. For instance, the TAD–ene reaction of unactivated terminal alkenes usually results in low yields, presumably due to competing overreaction of the resulting ene adducts, as has been documented by Du Prez and colleagues.11)
Since the seminal report of the TAD–ene reaction in 1967 by Pirkle and Stickler,6) halogenated solvents such as CH2Cl2 have been routinely used. The superiority of CH2Cl2 over other common organic solvents such as acetonitrile (MeCN), benzene, and tetrahydrofuran was experimentally validated by Ohashi and Bulter.12) In this study, we sought to establish whether inherent limitations of the TAD–ene reaction could be addressed using unconventional organic solvents. In particular, we were drawn to the upsurge of reports documenting the remarkable effect of fluorinated alcohols, including 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) and 2,2,2-trifluoroethanol (TFE) over the past two decades.13–18) Based on these reports, we speculated that the unique properties of fluorinated alcohols (e.g., high hydrogen bonding donor ability19)) may have beneficial effects on the TAD–ene reaction. In this Communication, we report the successful execution of this idea to provide a simple yet powerful solution to the longstanding challenge in the TAD–ene reaction. The use of HFIP as a solvent was crucial because it dramatically attenuated the overreaction while accelerating the desired ene reaction with alkene starting materials.
We began our investigation by examining the ene reaction between 4-phenyl-1-butene (1a) and 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD, 2) (Table 1). When 2 was treated with a slightly excess amount of 1a (1.1 equivalent (equiv.)) in CH2Cl2 at 0 °C, the characteristic carmine color of 2 disappeared after 8 h, affording ene adduct 3a in 41% NMR yield. In addition to the double ene adduct, the formation of the triple ene adduct was confirmed by the electrospray ionization (ESI) mass spectrometry of the crude mixture (see the Supplementary Materials). As expected, the yield of 3a increased proportionally with increasing equivalents of alkene 1a (entries 2 and 3). Furthermore, treatment of ene adduct 3a with an equimolar of 2 led to substantial consumption of 3a, corroborating the competing overreaction (see the Supplementary Materials). Based on these results, we compared several solvents with CH2Cl2 to improve the efficiency of the reaction. Consistent with a previous study,12) switching CH2Cl2 to MeCN, MeNO2 or toluene had deleterious effects both in terms of the reaction rate and the yield of 3a (entries 4–6). We were pleased to observe the remarkable effect of fluorinated alcohols (entries 7 and 8). In particular, HFIP was outstanding and delivered an excellent yield (92%) of the desired ene adduct 3a (entry 7). Notably, the use of fluorinated alcohols as solvents resulted in a significant rate enhancement while increasing the yield of 3a (entries 1, 7, and 8). In stark contrast, 2-propanol did not yield the ene product, highlighting the advantage of fluorinated alcohols (entry 9).
 | ||||
|---|---|---|---|---|
| Entry | 1a (equiv) | Solvent | Time (h) | Yield of 3a (%)b) |
| 1 | 1.1 | CH2Cl2 | 8.0 | 41 |
| 2 | 3.0 | CH2Cl2 | 4.5 | 58 |
| 3 | 5.0 | CH2Cl2 | 2.5 | 70 |
| 4 | 1.1 | MeCN | 8.0 | 9 |
| 5 | 1.1 | MeNO2 | 8.0 | 35 |
| 6 | 1.1 | Toluene | 8.0 | <5 |
| 7 | 1.1 | HFIP | 1.0 | 92 (88)c) |
| 8 | 1.1 | TFE | 4.0 | 72 |
| 9 | 1.1 | 2-Propanol | 8.0 | <5 |
a) Reaction conditions: PTAD (2) (0.20 mmol), 1a, solvent (1.0 mL) at 0 °C. b) Yields determined by 1H-NMR analysis using 1,3,5-trimethoxybenzene as an internal standard. c) Isolated yields in parentheses.
We investigated the generality of the optimized conditions by exploring the scope of terminal alkenes. As summarized in Table 2, the use of HFIP as a solvent was observed to be effective for various monosubstituted terminal alkenes, providing ene adducts in good yields (3b–3d, a range of 89–92% isolated yields). In addition, we were pleased to observe that the developed conditions functioned well for disubstituted alkene 3e (91% isolated yield). As a notable example, ene adduct 3f bearing a trisubstituted alkene functionality, which should be far more reactive toward PTAD (2) than the starting material 1f,11) was obtained in a synthetically useful yield (55% isolated yield). Remarkably, control experiments underscored the dramatic effect of HFIP, as diminished yields of ene adducts 3 were observed in all cases when CH2Cl2 was used as the solvent.
 |
a) Reaction conditions: PTAD (2) (0.20 mmol), 1 (0.22 mmol), solvent (1.0 mL) at 0 °C for 1 h (HFIP) or 8 h (CH2Cl2). b) Yields determined by 1H-NMR analysis using 1,3,5-trimethoxybenzene as an internal standard. c) Isolated yields in parentheses.
The effect of HFIP was further validated in the ene reactions of varying classes of azo compounds. As shown in Chart 2, the developed conditions were observed to be effective for the ene reaction of DEAD with alkene 1a, and the reaction proceeded smoothly in HFIP to provide ene adduct 7 in 74% yield, whereas replacement of HFIP with CH2Cl2 resulted in the recovery of alkene 1a. Although Brimble and Heathcock demonstrated that the stoichiometric use of SnCl4 can facilitate the DEAD–ene reaction, the use of HFIP should be beneficial in terms of practicality.20)
a) Reaction conditions: DEAD (0.20 mmol), 1a (0.22 mmol), and solvent (1.0 mL) at room temperature (15–18 °C). Isolated yield.
To gain insight into the effect of HFIP, competition experiments were performed (Chart 3). When an equimolar mixture of alkene 1c and ene adduct 3d was treated with PTAD (2) in CH2Cl2, ene adduct 3c was obtained in 46% yield along with recovered 1c and 3d both in 50% yields, hinting that the overreaction between ene adduct 3d and PTAD (2) competed in a comparable rate. In contrast, the same competition reaction in HFIP furnished 3c in 81% yield, and 90% of 3d was recovered, establishing that HFIP promoted the reaction between alkene 1 and PTAD (2) more effectively than the overreaction of ene adduct 3. While the accurate role of HFIP is unclear at this stage, we speculate that the observed promotion effect was attributed to the ability of HFIP to stabilize the ionic transition state of azo–ene reactions21–23) presumably through hydrogen bonding interactions.17,24) However, we recognize that such a stabilization should similarly promote the undesired overreaction of ene adduct 3. Accordingly, we surmise that HFIP might dissuade ene adduct 3 from undergoing the overreaction. In this context, Costas and colleagues observed a remarkable deactivation effect of HFIP capable of inhibiting the oxidation of otherwise reactive C–H bonds nearby basic functionalities presumably through hydrogen bonding.25) On this basis, we analogously speculate that HFIP would engage in hydrogen bonding with ene adduct 3, thereby attenuating the reactivity of 3 toward azo compounds.
a) Reaction conditions: PTAD (2) (0.10 mmol), 1c (0.10 mmol), 3d (0.10 mmol), solvent (1.0 mL) at room temperature (15–18 °C) for 5 h. b) Yields determined by 1H-NMR analysis using 1,3,5-trimethoxybenzene as an internal standard.
In conclusion, we provided a simple yet powerful solution to the persistent problem in the TAD–ene reaction by employing HFIP as a solvent. Our method allows the rapid conversion of unactivated terminal alkenes to provide the corresponding allylic urazoles in good to excellent yields. In addition, other classes of azo compounds, such as DEAD, were observed to be competent under the optimized reaction conditions. Control experiments revealed that HFIP uniquely facilitates the desired azo–ene process of alkene 1 more effectively than the undesired overreaction of ene adduct 3. We speculate that this remarkable solvent effect of HFIP would be attributed to its high hydrogen bonding donor ability.
This study was supported by a JSPS KAKENHI Grant (21K15237) and by the MEXT Leading Initiative for Excellent Young Researchers. I thank Dr. Kin-ichi Tadano (ITSUU Laboratory) and Dr. Mitsuaki Ohtani (ITSUU Laboratory), and professors Shuji Akai (Osaka University), Masayuki Inoue (The University of Tokyo), Takeo Kawabata (International University of Health and Welfare), Tomohiko Ohwada (The University of Tokyo), and Hideaki Kakeya (Kyoto University) for valuable discussions.
The author declares no conflict of interest.
This article contains supplementary materials.
