Diels–Alder reactions between hexafluoro-2-butyne and bis-furyl dienes: kinetic versus thermodynamic control

Abstract

The tandem [4+2] cycloaddition between hexafluoro-2-butyne and bis-furyl dienes, like difurfuryl ester, at room temperature leads to the kinetically controlled “pincer”-adducts – annulated 4a,8a-bis(trifluoromethyl)hexahydro-1,4:5,8-diepoxynaphthalenes. On the other hand, if these reactions proceed at 140 °C, only the thermodynamically controlled “domino”-adducts – annulated 2,3-bis(trifluoromethyl)hexahydro-1,4:5,8-diepoxynaphthalenes – are formed. Therefore, a very rare and unexpected example of full kinetic and thermodynamic control in the Diels–Alder reaction is reported in this paper.

---

Domino and tandem intramolecular Diels–Alder reactions of furans (IMDAF reaction) are widely used in organic synthesis for different purposes due to their simplicity, reliability and controlled stereochemistry.¹ In particular, the products of the IMDAF reaction are often applied as starting materials for the synthesis of functionally substituted naphthalenes, indoles, isoindoles, quinolines and isoquinolines.^1c

The products of the reaction between furans and alkynes have a 7-oxabicyclo[2.2.1]heptene scaffold. The structural motif has great potential as a tool for the design and synthesis of a broad diversity of substances with various practically useful properties. For example, recently these cycloadducts have been used for the construction of polycyclic aromatic hydrocarbons – fragments of graphene, which can serve as a model for new carbon based electronic materials.² The annulated 7-oxabicyclo[2.2.1]heptane moiety acts as a framework for molecular tweezers,³ various supramolecular systems,⁴ donor–bridge–acceptor molecules,⁵ a wide range of bioactive and natural compounds,⁶ high-molecular weight materials,^6e,7etc.

Taking into account such widespread actual or potential applications of the IMDAF reaction products, design of new efficient molecular systems for these processes and the understanding of intimate mechanistic details and driving forces underlying these reactions are of obvious general importance.

Various alkynes were used in the tandem intramolecular [4+2] cycloaddition, but only limited information concerning the reactivity of fluorinated acetylenes is available.⁸ In 1985, this journal published pioneering work by Visnick and Battiste,⁹ who reported an unprecedented example of the total kinetic and thermodynamic control in the course of the domino [4+2] cycloaddition between hexafluorobut-2-yne and N,N′-dipyrrolylmethane (1, Scheme 1). When the reaction was carried out under kinetically controlled conditions (room temperature), only the pincer-type polycycle 3 was formed through the open intermediate 2. Upon heating, however, the pincer adduct 3 underwent an intramolecular rearrangement into the domino-type adduct 4. Both reactions proceeded in quantitative yields, unlike other known cycloadditions of this type,¹⁰ which usually lead to a mixture of products of kinetic and thermodynamic control, regardless of the reaction temperature. For a long time, the transformation discussed in that paper had been attractive in synthetic¹¹ and theoretical¹² chemistry research, but it remained unique due to the limited accessibility of the suitable starting bis-dienes.

Scheme 1 Example of the total temperature control in the tandem intramolecular Diels–Alder reaction.^9a

In 1997, the second example similar to system 1 appeared in a report by Lautens and Fillion who demonstrated that difuryl bis-dienes in a reaction with alkynes are able to give pincer-cycloadducts analogous to 3 under kinetically controlled conditions.¹³

Having obtained a significant insight from previous reports about the importance of kinetic and thermodynamic control in intra/intermolecular [4+2]-cycloaddition reactions, our main goal in this study was to expand the available classes of dienes and dienophiles for the implementation of the tandem Diels–Alder reaction under kinetic or thermodynamic control conditions. In this work, we focused on bis-furyldienes 5 that can be easily obtained in a two-step process,^10a,13a and on hexafluorobutyne as a dienophile. The latter, apparently, is the most active known dienophile and usually ensures excellent yields of the cycloaddition products at low temperatures. However, the experimental treatment of this alkyne has some peculiarities due to its gaseous state under normal conditions.

Theoretical DFT calculations (IEFPCM-M06-2X/6-311++G**) of the reaction between F₃CC≡CCF₃ and three dienes (5a, 5c, and 5g) were carried out (for computational details, see the ESI†). The reaction starts with the cycloaddition of F₃CC≡CCF₃ at one of the furan moieties which occurs in a concerted fashion viaTS1 (Fig. 1) and represents the rate-limiting step of the whole process with the activation barrier of 23.1–26.8 kcal mol⁻¹ (in terms of ΔG^≠).

Fig. 1 Gibbs free energy profile for the reaction between bis-dienes 5a, c, and g and F₃CC≡CCF₃. The relative energies are given in kcal mol⁻¹ for X = CH₂ (plain text), S (italic) and NC(O)CF₃ (bold).

Further, the reaction could proceed via two competing channels, i.e. either leading to the pincer-type products 7viaTS2k or resulting in the formation of the domino product 8viaTS2t. The calculations showed that the first channel is more kinetically favourable (by 5.7–5.9 kcal mol⁻¹), which is in agreement with the published experimental observations^9,13 on the exclusive formation of the pincer products at room temperature. Meanwhile, the domino products 8 are more thermodynamically stable than 7 by 4.2–4.7 kcal mol⁻¹, and this fact may cause the isomerization of 7 into 8 at elevated temperature. Indeed, the calculated activation barriers for the 7 → 8 isomerization via the retro-Diels–Alder reaction of 7 followed by the intramolecular [4+2]-cycloaddition in the chain intermediate 6 to give 8 are 34.0–34.4 kcal mol⁻¹. These values are reasonable to permit such a process at a sufficiently high temperature. Taking into account these predictions, the following experiments were performed.

Under kinetically controlled conditions, the reaction of bis-dienes 5 with hexafluorobutyne was carried out in sealed glass ampules at −70 °C, with subsequent increase of temperature to ∼21–24 °C during 2–3 h. Toluene was chosen as a solvent due to its inertness, relatively low volatility and low melting point. Under these conditions, the reaction can occur within five days, but in order to reach a true equilibrium, all reactions were run for 10 days at r.t.

The ¹H NMR data of the crude reaction mixtures obtained after gentle evaporation of the solvent are given in Table 1. In most cases, the cycloaddition occurs chemospecifically leading to the kinetically controlled adducts 7a, d, g, and h only in quantitative yields or proceeds with a high level of chemoselectivity; the ratio of pincer/domino adducts is 7/8 > 95/5 (entries e, f, and i). The ratio of diene/alkyne varied from 1/1 to 1/1.4, and in most cases did not significantly affect both the yield of the products and the composition of the reaction mixtures.

Table 1 Conditions of kinetic and thermodynamic control and yields of adducts 7 and 8

Entry	X	Molar ratio of 5/C4F6^a	Ratio^b of 7/8	7 (%)	8 ^, (%)
a All reactions were carried out in sealed ampules at r.t. for 10 days. b Ratio of 7/8 according to the ^1H NMR analysis of the reaction mixtures obtained after solvent evaporation. c Isolated yields after recrystallization. d Domino-adducts 8 were obtained by heating of pincer-adducts 7 in o-xylene for 2 h. e See Table 2.
a	CH2	1/1.1	100/0	69	62
b	O	1/1.1	—^e	60	67
c	S	1/1.1	—^e	76	76
d	N-Bn	1/1.4	100/0	71	66
e	N-Ac	1/1.1	>95/5	62	68
f	N-Bz	1/1.1	>96/4	70	76
g	N-COCF3	1/1.4	100/0	56	77
h	N-COCCl3	1/1	100/0	62	74
i	N-CO2Me	1/1	>95/5	79	71

---

However, two examples in Table 1 (rows b: X = O and c: X = S) drop out of the series and gave rise to a lower chemoselectivity (Scheme 2).

Scheme 2 Synthesis of pincer and domino adducts 7 and 8.

In these cases, in addition to the side domino-adducts 8b and c, the products of the double cycloaddition 9b and c and 10b and c were formed as inseparable diastereomeric pairs with a practically equal ratio of components (Scheme 3 and Table 2).

Scheme 3 Compositions of the reaction mixtures in the case of the [4+2] cycloaddition between F₃CC≡CCF₃ and bis-furans 5b and c.

Table 2 Compositions of the crude reaction mixtures obtained after interaction of bis-dienes 5b and c with F₃CC≡CCF₃ according to ¹H NMR analysis

Entry	X	Ratio of 5/C4F6	5 (%)	7 (%)	8 (%)	9 + 10^b (%)
a The isolated yield of pincer-adducts 7 after recrystallization is given in Table 1, rows b and c. b Ratio of 9/10 is ∼50/50.
b	O	1/1	14.9	64.9	5.1	15.1
b′	O	1/1.1	4.4	63.2	4.5	27.9
c	S	1/1	8.3	82.3	3.6	5.8
c′	S	1/1.1	5.2	78.0	2.9	13.9

---

The formation of bis-adducts 9 and 10 suggests that in these examples, the intermolecular Diels–Alder reaction involving a subsequent interaction of the second molecule of hexafluorobutyne with intermediates 6b and c takes place. In other words, the intermolecular [4+2] cycloaddition between C₄F₆ and intermediate 6 starts to compete with the intramolecular Diels–Alder reaction in the chain intermediate.

Initially, the reactions of 5b and c were carried out with 1.0 mol equivalent of C₄F₆ under similar conditions: 10 days at r.t. (rows b and c in Table 2). At the equimolar ratio of 5/C₄F₆, the reaction mixtures contained 8–15% of unreacted starting bis-furans 5b and c, therefore the molar ratio of 5/C₄F₆ was changed by 1/1.1. Under these conditions (rows b′ and c′), the percentage of the initial 5b and c in the crude reaction mixtures was reduced to 4–5%. Simultaneously, a sharp increase (up to 28%) in the content of the double-adducts 9 and 10 was observed. Due to this, we did not perform further experiments with a larger excess of hexafluorobutyne. The data shown in Table 2 allow us to conclude that the excess of alkyne exerts a significant influence on the fraction of the double cycloaddition adducts 9 and 10, but at the same time, the ratio of 7/8 remains virtually constant (it changes only from 96/4 to 93/7).

The rearrangement of the pincer-adducts 7 to the domino-products 8 is observed at temperatures higher than 110 °C (¹H NMR experiments with a temperature gradient were carried out) and proceeds rather quickly at 120 °C. According to NMR spectroscopic data in boiling o-xylene, the products of thermodynamic control 8 were formed in 100% yields at 1–2 h, but after recrystallization they were isolated in moderate yields (Table 1).

Since the last transformation 7 → 8 occurred in reasonable time without any by-products, the half-life and rate constant have been calculated using dynamic ¹H NMR spectroscopy assuming first-order kinetics. The reaction involving the rearrangements of 7b, c, and g to 8b, c, and g was carried out in an NMR tube in CDCl₂–CDCl₂ (for details, see the ESI†). At 120 °C, the half-life (τ_1/2) was 37.5 min, 80 min, and 42.5 min (for X = O, S, and N–COCF₃, respectively), which corresponds to the first-order reaction rate constants of 0.0185 min⁻¹, 0.0087 min⁻¹, and 0.0163 min⁻¹. The transformation of 7b to 8b (X = O) is the quickest one along the series. As expected, at 140 °C in the same solvent, the reactions proceed even faster. For example, for the synthesis of compound 8b, τ_1/2 was ∼4.6 min with the first-order reaction rate constants of 0.151 min⁻¹. The rearrangement is so quick at 140 °C that it is difficult to accurately measure the reaction rate by the NMR method.

Our attempts to carry out the analogical dynamic NMR experiment for the first step of the reaction under discussion (the cycloaddition of C₄F₆ to 5a) in a sealed NMR tube failed due to its insufficient strength against the elevated pressure.

The spatial structures of the regioisomer pairs 7c, d, and g/8c, d, and g have been determined by single-crystal X-ray diffraction using synchrotron radiation (see the ESI†).

In conclusion, a rare example of full kinetic and thermodynamic reaction control in the process of the tandem IMDAF reaction of bis-furyl dienes with hexafluoro-2-butyne has been discovered. At room temperature, the reaction occurs chemoselectively leading to adducts of the pincer-[4+2] cycloaddition. The exclusive formation of domino-adducts is observed at elevated temperatures. This reaction could be useful for the construction of natural product and supramolecular frameworks and can serve as an outstanding example of reaction reversibility to be used in physical chemistry courses.

Funding for this research was provided by the Ministry of Education and Science of the Russian Federation (award No. 4.1154.2017/4.6) and by the Fundação para a Ciência e a Tecnologia (FCT, Portugal) (project UID/QUI/00100/2013).

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. For selected reviews on tandem intramolecular Diels–Alder cycloaddition in organic synthesis, see: (a) J. E. Sears and D. L. Boger, Acc. Chem. Res., 2016, 49, 241–251; (b) P. T. Parvatkar, H. K. Kadam and S. G. Tilve, Tetrahedron, 2014, 70, 2857–2888; (c) A. Padwa and S. K. Bur, Tetrahedron, 2007, 63, 5341–5378; (d) J. Wu, L. Sun and W.-M. Dai, Tetrahedron, 2006, 62, 8360–8372; (e) K.-I. Takao, R. Munakata and K.-I. Tadano, Chem. Rev., 2005, 105, 4779–4807; (f) J. D. Winkler, Chem. Rev., 1996, 96, 167–176.
2. (a) S. Eda, F. Eguchi, H. Haneda and T. Hamura, Chem. Commun., 2015, 51, 5963–5966; (b) A. Criado, M. Vilas-Varela, A. Cobas, D. Peréz, D. Peña and E. Guitián, J. Org. Chem., 2013, 78, 12637–12649; (c) F. Furrer, A. Linden and M. C. Stuparu, Chem. – Eur. J., 2013, 19, 13199–13206.
3. (a) R. B. Murphy, R. E. Norman, J. M. White, M. V. Perkins and M. R. Johnston, Org. Biomol. Chem., 2016, 14, 8707–8720; (b) R. N. Warrener, D. Margetic, A. S. Amarasekara, D. N. Butler, I. B. Mahadevan and R. A. Russell, Org. Lett., 1999, 1, 199–202.
4. (a) T.-C. Chou and Y.-J. Li, Tetrahedron, 2015, 71, 5620–5633; (b) C. H. Oh, H. J. Yi and K. H. Lee, Bull. Korean Chem. Soc., 2010, 31, 683–688; (c) M. Eckert-Maksić, D. Margetić, S. Kirin, D. Milić and D. Matković-Čalogović, Eur. J. Org. Chem., 2005, 4612–4620.
5. S. Chakrabarti, M. Liu, D. H. Waldeck, A. M. Oliver and M. N. Paddon-Row, J. Am. Chem. Soc., 2007, 129, 3247–3256.
6. (a) S. Roscales and J. Plumet, Nat. Prod. Commun., 2017, 12, 713–732; (b) V. S. Enev, W. Felzmann, A. Gromov, S. Marchart and J. Mulzer, Chem. – Eur. J., 2012, 18, 9651–9668; (c) A. Gromov, V. Enev and J. Mulzer, Org. Lett., 2009, 11, 2884–2886; (d) C. S. Schindler and E. M. Carreira, Chem. Soc. Rev., 2009, 38, 3222–3241; (e) P. Vogel, Tetrahedron, 1999, 55, 13521–13642.
7. (a) D. Margetić, M. Eckert-Maksić, P. Trošelj and Z. Marinić, J. Fluorine Chem., 2010, 131, 408–416; (b) R. N. Warrener, D. Margetić, P. J. Foley, D. N. Butler, A. Winling, K. A. Beales and R. A. Russell, Tetrahedron, 2001, 57, 571–582.
8. For cycloaddition reactions between furans (pyrroles) and fluorinated alkynes, see: (a) D. Kozai, Y. Kabasawa, M. Ebert, S. Kiyonaka, Firman, Y. Otani, T. Numata, N. Takahashi, Y. Mori and T. Ohwada, Mol. Pharmacol., 2014, 85, 175–185; (b) A. A. Kislukhin, C. J. Higginson, V. P. Hong and M. G. Finn, J. Am. Chem. Soc., 2012, 134, 6491–6497; (c) K. M. Guckian, E. Y.-S. Lin, L. Silvian, J. E. Friedman, D. Chin and D. M. Scott, Bioorg. Med. Chem. Lett., 2008, 18, 5249–5251; (d) T. Yanagimoto, T. Toyota, N. Matsuki, Y. Makino, S. Uchiyama and T. Ohwada, J. Am. Chem. Soc., 2007, 129, 736–737; (e) G.-D. Zhu, M. A. Staeger and S. A. Boyd, Chem. Lett., 2000, 3345–3348; (f) L. R. Domingo, M. T. Picher and M. J. Aurell, J. Phys. Chem. A, 1999, 103, 11425–11430.
9. (a) M. Visnick and M. A. Battiste, J. Chem. Soc., Chem. Commun., 1985, 1621–1622; (b) G. Sun, D. N. Butler, R. N. Warrener and D. Margetić, J. Heterocycl. Chem., 2015, 52, 1195–1200.
10. For selected reports on kinetic and thermodynamic control in the Diels–Alder reaction, see: (a) R. C. Boutelle and B. H. Northrop, J. Org. Chem., 2011, 76, 7994–8002; (b) C. Taffin, G. Kreutler, D. Bourgeois, E. Clot and C. Périgaud, New J. Chem., 2010, 34, 517–525; (c) J. D. White, F. W. J. Demnitz, H. Oda, C. Hassler and J. P. Snyder, Org. Lett., 2000, 2, 3313–3316; (d) A. P. Marchand, B. Ganguly, W. H. Watson and S. G. Bodige, Tetrahedron, 1998, 37, 10967–10972; (e) M. Manoharan and P. Venuvanalingam, J. Chem. Soc., Perkin Trans. 2, 1997, 1799–1804; (f) S. G. Bott, A. P. Marchand and K. A. Kumar, J. Chem. Crystallogr., 1996, 26, 281–286; (g) D. Suarez, T. L. Sordo and J. A. Sordo, J. Org. Chem., 1995, 60, 2848–2852; (h) P. D. Bartlett and C. Wu, J. Org. Chem., 1985, 50, 4087–4092; (i) L. A. Paquette, M. J. Wyvratt, H. C. Berk and R. E. Moerck, J. Am. Chem. Soc., 1978, 100, 5845–5855.
11. L. F. Tietze, G. Brasche and K. M. Gericke, in the book Domino reactions in organic synthesis, Wiley-VCH Verlag GmbH & Co. KgaA, 2006, pp. 280–336.
12. For the density functional theory study of the reaction between hexafluorobut-2-yne and 1, see: (a) L. R. Domingo, M. Arnó and J. Andrés, J. Am. Chem. Soc., 1998, 120, 1617–1618; (b) L. R. Domingo, M. T. Picher, M. Arnó, J. Andrés and V. S. Safont, THEOCHEM, 1998, 426, 257–262.
13. (a) M. Lautens and E. Fillion, J. Org. Chem., 1997, 62, 4418–4427; (b) L. R. Domingo, M. T. Picher and J. Andrés, J. Org. Chem., 2000, 65, 3473–3477.

---

Footnote

† Electronic supplementary information (ESI) available: Single-crystal X-ray descriptions for 7c, d, and g and 8c, d, and g; computational details and tables with energies and atomic coordinates; detailed synthetic procedures and spectral data for compounds 5, 7, and 8. CCDC 1570123–1570128. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7cc09466c

---