A Case of Competitive Aromatization vs. Sigmatropic [1,5]-Hydrogen Atom Migration in a 1,2,4-Cyclohexatriene Intermediate Derived from a Bis-Enyne Cyclization

Abstract

1,2,4-Cyclohexatrienes are strained, reactive intermediates often formed by the tetradehydro-Diels–Alder (TDDA) reaction of a conjugated enyne bearing a tethered alkyne as the enynophile. The ene component is commonly the π-bond of an aromatic group. In this focused study, we investigated the reactivity of a symmetrical substrate in which the pair of terminal ene moieties were simple 2-propenyl groups. The intermediate 1,2,4-cyclohexatriene, now bearing a 5-isopropenyl group, underwent competitive aromatization (the most usual outcome of the strain-relieving event of the cyclohexatriene), along with an intramolecular [1,5]-hydrogen atom migration, ultimately producing a non-benzenoid, pyrrole derivative. This represents a previously unknown process for a 1,2,4-cyclohexatriene derivative. Mechanistic aspects of the competitive processes were revealed by experiments performed in the presence of various protic additives (MeOD and BHT).

1. Introduction

1,2,3-Cyclohexatriene (1, Figure 1a), a reactive intermediate and isomer of benzene, was first formed and intercepted in 1990 [1]. The isomeric 1,2,4-cyclohexatriene (2) was first generated and trapped in 1992 [2]. These species and their derivatives remain of contemporary interest [3,4]. Members of this class of compounds are highly reactive because they house a highly strained cyclic allene within a small ring. The strain energies in 1 and 2 have been evaluated computationally and are estimated to be 50 and 34 kcal mol⁻¹, respectively [5]. The two pairs of terminal substituents on acyclic allenes (1,2-propadiene derivatives, 3) ideally, of course, have an orthogonal relationship. The constraints of any rings (with the exception of very large ones) that house a cyclic allene dictate that the 1,2-propadiene unit is unable to achieve that ideal geometry, hence introducing increasing amounts of strain for lower homologs within smaller rings [5,6].

1,2,4-Cyclohexatriene derivatives are often formed as transient intermediates in, for example, net (4 + 2) cycloadditions between an alkyne and a 1,3-enyne in a process that can be categorized as a tetradehydro-Diels–Alder (TDDA) reaction [7,8] (cf. 4 to 5, Figure 1b [9,10]). The transient 1,2,4-triene is known to undergo several types of processes to relieve the strain of the significantly twisted allene that it houses. Hydrogen atom migration that results in a benzenoid product (5 to 6, Figure 1b) is most common. Rarer are processes that result in non-aromatized products (7 to 8, [11] Figure 1c; or 9 to 10, [12] Figure 1d).

Figure 1. (a) Strained, cyclic cyclohexatrienes (1 and 2) vs. the strain-free parent, propa-1,2-diene (also known as allene, 3), whose most stable geometry has orthogonal (90°) a-C-b and c-C-d planes. (b) Formation of 1,2,4-cyclohexatriene derivatives by a tetradehydro-Diels–Alder (TDDA) reaction and their aromatization to benzene derivatives by formal [1,5]-hydrogen atom migration [9,10]. (c) Relief of ring-strain by skeletal rearrangement [11]. (d) Relief of ring-strain by migration of an exocyclic group [12]. (e) Computed lowest energy pathway for the net [1,5]-hydrogen atom migration via two, sequential (blue then green) 1,2-H-atom migrations [13,14,15].

Figure 1. (a) Strained, cyclic cyclohexatrienes (1 and 2) vs. the strain-free parent, propa-1,2-diene (also known as allene, 3), whose most stable geometry has orthogonal (90°) a-C-b and c-C-d planes. (b) Formation of 1,2,4-cyclohexatriene derivatives by a tetradehydro-Diels–Alder (TDDA) reaction and their aromatization to benzene derivatives by formal [1,5]-hydrogen atom migration [9,10]. (c) Relief of ring-strain by skeletal rearrangement [11]. (d) Relief of ring-strain by migration of an exocyclic group [12]. (e) Computed lowest energy pathway for the net [1,5]-hydrogen atom migration via two, sequential (blue then green) 1,2-H-atom migrations [13,14,15].

2. Results and Discussion

Of the three modes of reaction shown in Figure 1b–d, the isomerization of the cyclic allene to its isomeric benzenoid product (cf. 5 to 6) is by far the most common and, therefore, well studied. The generally accepted mechanism of the aromatization of 1,2,4-cyclohexatriene (and its derivatives, at least under conditions where there is no proton source in the reaction medium), is portrayed in Figure 1e [11 to benzene (13)]. Computations [13,14,15] suggest that this thermal-only pathway involves two consecutive 1,2-H-atom shifts rather than a single [1,5]-H-atom migration. The first 1,2-shift gives a species (carbene or orthogonal diradical, cf. 12/12′) that undergoes a rapid, second C–H insertion/migration event.

In the course of studying the thermal cyclization chemistry of bis-enyne 14, we observed, in addition to the normal aromatized benzenoid TDDA product, 16, an unexpected side product [16], 17 (Figure 2). The ratio of 16:17 under these conditions was ca. 2:1. Although this ratio changed under other reaction conditions (see below), 17 was never the major product. The assignment of the structure of the expected 16 was straightforward from analysis of its ¹H and ¹³C 1D and 2D NMR spectra (see Supporting Information (SI)). However, the structure of 17 was not at all clear. Only after determining its structure would we be in a position to hypothesize a pathway for its formation.

It was straightforward to confirm (GCMS) that 17 had the same mass as 16 (i.e., they were isomers). The chromatographic behaviors on silica gel of 16 and 17 were nearly the same, but by shaving the fractions eluting from an MPLC separation, we were able to obtain 17 in a 96:4 ratio with 16. In the proton NMR spectrum of 17, it was clear that the NTs moiety was intact. It also contained three singlets, whose chemical shifts (δ 7.04, 6.83, and 6.07 ppm) were indicative of protons attached to sp² hybridized carbon atoms. Another significant difference in the spectrum (vs. that of 16) was the presence of three, rather than two, methyl singlets (in addition to the Ts-Me), judged to be either allylic or benzylic in nature. The remaining protons were a singlet methylene pair (2.99 ppm) that, among other things, indicated that 17 possessed an achiral structure.

The NOESY NMR spectrum was particularly informative; enhancements (double-headed arrows in Figure 3a) led us to propose structure 17 as the identity of the unknown. More specifically, the NOEs between what are assigned there as H1/H7 and H3/H1′′ (red) were quite diagnostic. To further support this assignment, we located in the literature compound 18 [17], a structurally analogous, fused N-tosyl pyrrole that possesses quite similar chemical shifts of the indicated protons in 18 vs. the analogous protons in 17 (Figure 3b). We also carried out a DP4+ analysis [18] of the NMR chemical shifts of the protons and carbons in 17 vs. an isomeric alkene having the endocyclic cyclohexene double bond between C5 and C6; this analysis also supported the assignment of 17 (see SI). Finally, we performed resolution enhancement of the (many) “singlets” in 17, mentioned in the above paragraph; this revealed (many) small coupling constants that further reinforced the structure assignment (see J values in the ¹H NMR line listings in Section 3).

With the structure of the unusual side product 17 established, we were obliged to propose a mechanism to account for its formation, especially so because, as far as we are aware, it represents an unprecedented type of transformation for a product arising from a TDDA reaction via a 1,2,4-cyclohexatriene-containing intermediate such as 5. We show in Figure 3c a proposed pathway that branches following the initial formation of the intermediate 15 from an initial TDDA cycloisomerization. The 1,2-hydrogen atom migration to 19 and final carbene C–H insertion account for the major product, 16. But in addition, the isopropenyl alkene in intermediate 15 is poised to undergo a sigmatropic rearrangement, namely [1,5]-hydrogen atom migration to produce 20. A final net 1,3-hydrogen atom migration (tautomerization) of the strained cyclic allene in 20 would establish the pyrrole ring in the minor product, 17. As indicated by the arrows in structure 20, there is the possibility that this aromatization could be mediated (catalyzed) by a protic molecule (e.g., water) acting as a proton shuttle; this conversion could be either concerted or stepwise, with the latter following an initial protonation of the central carbon of the allene.

It should be noted that allenes are known to undergo reaction with Brønsted acids via kinetic protonation at either their terminal or central atoms to give vinylic or allylic carbocations, respectively [19]. Particularly relevant here are the differences seen between unstrained, acyclic allenes vs. strained, cyclic allenes. The former allenes tend to prefer kinetic protonation at a terminal carbon atom, although even for this class of unstrained allenes, this selectivity can often be reversed by the nature of substituents on the terminal atoms. In contrast, strained, cyclic allenes, tend to protonate preferentially at their central carbon atom, with there being a reduced amount of twist needed to attain the more stable allylic cation. An early report that brought this difference to light described the addition of HCl to cyclonona-1,2-diene to give only 3-chlorocyclononene [20]. The 1983 review by Smadja [19] contains a collection of many dozens of examples of “protonation of allenic derivatives” from which these general trends can be gleaned.

Several experiments were performed to probe the possible role that a hydroxyl-containing species might be playing. Saá and coworkers have used CD₃OD (or CH₃OD) as an additive and observed incorporation of a deuterium atom at the central carbon of TDDA-derived cyclic allene intermediates generated in their experiments (further discussed below) [15,21]. This established that a protic additive provided a pathway for aromatization that was faster than the unimolecular series of 1,2-H migrations computed for the parent system (cf. Figure 1e). We performed an analogous experiment with our substrate 14. Heating the dienyne 14 in a mixture of DCE and CD₃OD (90:10, v:v) at 130 °C for 16 h led to the formation of 16-d and 17-d, deuterated to ca. 91% and 90% (¹H NMR) and in a ratio of 2:1, that ratio being the same as the reaction performed without added methanol. An analogous experiment using CH₃OD as the additive also showed that both 16-d and 17-d were produced, indicating that the hydroxyl group, and not the methyl group, was the source of the hydrogen incorporated into each of those products.

These labeling experiments do not differentiate between the possible stepwise (via 15-d⁺ or 20-d⁺) vs. concerted net 1,3-tautomerizations shown in Figure 4a. A fuller discussion of the Saá experiments is instructive because they shed light on this question. As outlined in Figure 4b, heating the TDDA substrate 22 in toluene alone produced solely the naphthalene product 29-angular. This can be rationalized by the sequence of electrocyclic and isomerization events commencing from the initial cyclic allene 23 and proceeding through 24–26. Computations supported the main features of this pathway, although whether the final aromatization occurred via sequential intramolecular 1,2-H shifts (cf. Figure 1e) or a protic molecule-assisted event (e.g., via 27) was not discerned. However, the role of the protic additive phenol was clearly demonstrated because, when present, it resulted in the formation of 29-linear as the only product. This is explained by protonation of allene 23 by PhOH more quickly than the rate of unimolecular ring-opening of 23 to 24.

We performed a similar experiment with 14 using butylated hydroxytoluene (BHT) as a phenolic, protic additive. In the presence of 0.1 equivalent of BHT, the ratio of 16 to 17 changed to 6:1. Using 1.0 equivalent of BHT, we observed formation of only the aromatic product 16. This suggests that even with the use of a small amount of BHT, the cyclic allene was protonated to give 15-h⁺ at a rate competitive with its [1,5]-H atom migration, and that the larger amount of BHT promoter fully suppressed the [1,5]-H atom migration. Recall that even in the presence of 10 vol% of methanol, the ratio of products 16 to 17 was the same as when no methanol was present. The different effect of the BHT additive is most likely a reflection of its greater acidity relative to that of methanol (MeOH pKa_(DMSO) = 29.0 [22], PhOH pKa_(DMSO) = 18.0 [23]), thereby accelerating the production of 15-h⁺.

Finally, we prepared the ynoate/enyne substrate 30 (Figure 5). When heated, it reacted at a similar rate to that of 14. It produced the aromatized compound 33 as the major product. As expected, when this reaction was performed in the presence of CD₃OD, a deuterium atom was incorporated into the product 33-d. In principle, the carbonyl group of the ester in the intermediate cyclic allene 31 could have engaged in a [1,5]-hydrogen atom migration to give 32. However, the relative instability of the enol tautomer of an ester makes this a sufficiently high-barrier process, so there is little opportunity for that to occur enroute to a pyrrole derivative analogous to 17. This further emphasizes the unique role that the isopropenyl substituent is playing in diverting the reactivity of 15.

3. Experimental Section

3.1. Chemistry

3.1.1. General Experimental Protocols

¹³C and ¹H NMR spectra were obtained using a Bruker HD-500 spectrometer. All spectra were recorded in CDCl₃. The reported proton chemical shifts are referenced to CHCl₃ in (δ = 7.26 ppm) in CDCl₃. Resonances in the line listings of ¹H NMR spectral data are given in this format: chemical shift (δ, ppm) [multiplicity, coupling constant(s) J (in Hz), integral value (to the nearest integer)]. First-order coupling constants were extracted using methods we have described elsewhere [24,25]. An nfom, nfod, or nfot refers to a non-first order multiplet, doublet, or triplet in a ¹H NMR spectrum. Coupling constants for these resonances are given as apparent values (J_app); for example, the observed spacing between the two most intense lines in an nfod is the value of J_o + J_p and not the true doublet coupling constant. The reported carbon chemical shifts in ¹³C spectra are referenced to CDCl₃ (δ = 77.16 ppm). Infrared spectra were recorded using a using a Bruker Alpha II Spectrometer. The data were collected in attenuated total reflectance (ATR) mode. Samples were measured as thin films that were deposited onto a diamond ATR window by evaporative loss of solvent, often CDCl₃. Absorption maxima are given as cm⁻¹. High-resolution mass spectrometry (HRMS) data were collected in the ESI mode with a Thermo Orbitrap Velos instrument having mass accuracy of ≤3 ppm. Pierce^TM LTQ was used as an external calibrant. Samples were injected directly into the ion source. Melting points were measured on a Köfler hot-stage and polarizing microscope and are uncorrected. New substances were often purified by medium-pressure liquid chromatography (MPLC) at 25–200 psi. Silica gel columns (hand-packed, normal-phase, Teledyne RediSep Rf Gold^®, 20–40 μm, 60 Å pore size) were used. The unit was assembled from a Waters HPLC pump (model 510) affixed to a Waters differential refractive index detector (R401). Preparative flash column chromatography was performed on columns hand-packed with Agela silica gel (230–400 mesh). Thin-layer chromatography (TLC) was carried out on silica gel-coated plates (plastic-backed), visualized by UV light and/or by staining with a solution of potassium permanganate. Reaction temperatures refer to the temperature of a pre-equilibrated and external heating oil bath. Any reaction performed at a temperature higher than that of the reaction solvent’s boiling point was carried out in a threaded culture tube, which was closed with an inert Teflon^®-lined screw-cap.

3.1.2. Synthesis of Precursors 14 and 30 (Structures of Intermediates Not Specifically Shown in Any of Figure 1, Figure 2, Figure 3, Figure 4 and Figure 5 Are Given S#s Here)

N,N-bis(4-hydroxy-4-methylpent-2-yn-1-yl)-4-methylbenzenesulfonamide (S2). In a 100 mL round-bottomed flask equipped with a magnetic stir bar, 4-methyl-N,N-di(prop-2-yn-1-yl)benzenesulfonamide (S1 [26], 2.3 g, 9.4 mmol, 1 equiv) was added. THF (47 mL) was added at −78 °C under an atmosphere of N₂. n-BuLi (7.9 mL, 2.5 M in hexanes, 2.1 equiv) was added dropwise. After 1 h of stirring at −78 °C, acetone (2.1 mL, 28 mmol, 3 equiv) was added dropwise. The reaction mixture was stirred at −78 °C for another 1 h. The reaction was quenched with saturated NH₄Cl aqueous solution. THF was removed under vacuum, and the aqueous layer was extracted with DCM, washed with brine, dried (Na₂SO4), concentrated, and purified by flash column chromatography (10:1 to 1:1 Hex:EtOAc) to give S2 (2.9 g, 8.0 mmol, 85%) as a white solid. mp: 98–100 °C. ¹H-NMR (500 MHz, CDCl₃): δ 7.76 (nfod, J_app = 8.3 Hz, 2H), 7.33 (nfod, J_app = 8.0 Hz, 2H), 4.17 (s, 4H), 2.43 (s, 3H), 1.89 (s, 2H), and 1.37 (s, 12H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 144.0, 136.1, 129.7, 128.3, 90.5, 74.9, 65.1, 37.1, 31.2, and 21.6. IR (neat): ν_max 3524 (br O-H), 3265 (br O-H), 2981 (CH₃), and 2924 (CH₂) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₉H₂₆NO₄S⁺ [M + H]⁺, 364.1577; found, 364.1560 (9%). Calcd. for 346.1471 C₁₉H₂₄NO₃S⁺ [M + H⁺ – H₂O]; found, 346.1455 (100%).

4-Methyl-N,N-bis(4-methylpent-4-en-2-yn-1-yl)benzenesulfonamide (14). In a 100 mL round-bottomed flask, equipped with a magnetic stir bar, S2 (790 mg, 2.2 mmol, 1 equiv) was added. DCM (13 mL) was added under a N₂ atmosphere. To this solution, triethylamine (10.6 mL, 76 mmol, 35 equiv) was added at 0 °C. Methanesulfonic anhydride (6.7 g, 38 mmol, 18 equiv) dissolved in DCM (26 mL) was added dropwise. The reaction mixture was stirred at RT for 2.5 h. The reaction mixture was quenched with water and extracted with DCM. The combined organic layers were washed with brine, dried (Na₂SO₄), filtered, and concentrated; and the residue was purified by flash column chromatography (10:1 Hex:EtOAc) to give 14 (511 mg, 1.6 mmol, 72%) as a pale-yellow oil. ¹H-NMR (500 MHz, CDCl₃): δ 7.76 (nfod, J_app = 8.0 Hz, 2H), 7.31 (nfod, J_app = 8.2 Hz, 2H), 5.19 (br s, 2H), 5.13 (br s, 2H), 4.29 (s, 4H), 2.42 (s, 3H), and 1.76 (br s, J = 1.3 Hz, 6H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 143.8, 135.6, 129.7, 128.1, 126.1, 122.5, 87.0, 80.8, 37.3, 23.2, and 21.6. IR (neat): ν_max 3003 (C_sp2-H), 2925 (H₂C_sp3-H), and 1598 (C=C) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₉H₂₂NO₂S⁺ [M + H]⁺ 328.1366; found 328.1353 (100%).

4-Methyl-N-(4-methylpent-4-en-2-yn-1-yl)-N-(prop-2-yn-1-yl)benzenesulfonamide (S4). In a culture tube, equipped with a magnetic stir bar, N-(4-hydroxy-4-methylpent-2-yn-1-yl)-4-methyl-N-(prop-2-yn-1-yl)benzenesulfonamide (S3 [27], 305 mg, 1 mmol, 1 equiv) was added. DCM (5.9 mL) was added under an atmosphere of N₂. Triethylamine (2.4 mL, 17.6 mmol, 17.6 equiv) was added at 0 °C. Methanesulfonic anhydride (1.5 g, 8.8 mmol, 8.8 equiv) was added dropwise. The reaction mixture was stirred at RT for 1 h and then poured into ice-cold concentrated aqueous HCl (2 mL). The aqueous layer was extracted with DCM. The combined organic layers were washed with brine, dried (Na₂SO₄), filtered, concentrated, and purified by MPLC (15:1 Hex:EtOAc) to give S4 (162.88 mg, 0.56 mmol, 57%) as a pale-yellow oil. ¹H-NMR (500 MHz, CDCl₃): δ 7.72 (nfod, J_app = 8.4 Hz, 2H), 7.29 (nfod, J_app = 8.1 Hz, 2H), 5.15 (dq, J = 1.6, 1.6 Hz, 1H), 5.07 (dq, J = 1.8, 1.0 Hz, 1H), 4.29 (s, 2H), 4.13 (dd, J = 2.5, 0.7 Hz, 2H), 2.41 (t, J = 0.9 Hz, 3H), 2.16 (t, J = 2.5 Hz, 1H), and 1.71 (dd, J = 1.6, 1.1 Hz, 3H). ¹³C{¹H} (126 MHz, CDCl₃): δ 144.0, 135.4, 129.7, 128.1, 126.0, 122.5, 87.2, 80.4, 76.6, 74.0, 37.1, 36.5, 23.2, and 21.7. IR (neat): ν_max 3278 (C_sp-H), 2982 (H₂C_sp3-H), 2927 (RHC_sp3-H), and 1597 (C=C) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₆H₁₈NO₂S⁺ [M + H]⁺ 288.1053; found 288.1060 (100%).

Methyl 4-{[4-Methyl-N-(4-methylpent-4-en-2-yn-1-yl)phenyl]sulfonamido}but-2-ynoate (30). In a 10 mL round-bottomed flask, equipped with a magnetic stir bar, S4 (100 mg, 0.35 mmol, 1 equiv) was added. THF (3.5 mL) was added under a N₂ atmosphere. nBuLi (0.17 mL, 0.42 mmol, 1.2 equiv) was added dropwise at −78 °C. The reaction mixture was stirred at −78 °C for 1 h. Methyl chloroformate (0.08 mL, 1.0 mmol, 3 equiv) was added dropwise. The reaction mixture was stirred at −78 °C for another 1.5 h. The reaction mixture was quenched with NH₄Cl and extracted with EtOAc. The combined organic layers were washed with brine, dried (Na₂SO₄), and concentrated; and the residue was purified by flash column chromatography (30:1 to 10:1 Hex:EtOAc) to give 30 (21.3 mg, 0.0615 mmol, 18%) as a pale-yellow oil. ¹H-NMR (500 MHz, CDCl₃): δ 7.72 (nfod, J_app = 8.4 Hz, 2H), 7.31 (nfod, J_app = 8.3 Hz, 2H), 5.17 (m, 1H), 5.09 (m, 1H), 4.27 (s, 2H), 4.26 (s, 2H), 3.73 (s, 3H), 2.41 (s, 3H), and 1.71 (dd, J = 1.3, 1.4 Hz, 3H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 153.2, 144.3, 135.0, 129.9, 128.0, 125.9, 122.9, 87.7, 80.6, 80.0, 77.1, 52.9, 37.7, 36.5, 23.1, and 21.7. IR (neat): ν_max 2954 (C_sp3-H), 2922 (C_sp3-H), 2847 (C_sp3-H), 2242 (C≡C), and 1717 (C=O) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₈H₂₀NO₄S⁺ [M + H]⁺ 346.1108; found 346.1094 (100%).

3.1.3. Synthesis of Products of the TDDA Reactions

6-Methyl-4-(prop-1-en-2-yl)-2-tosylisoindoline (16) and 6-methyl-4-(propan-2-ylidene)-2-tosyl-4,5-dihydro-2H-isoindole (17). In a 10 mL culture tube, equipped with a magnetic stir bar, 14 (21 mg, 0.064 mmol) was added. 1,2-DCE (2 mL) was then added. The reaction mixture was heated at 130 °C for 16 h. The reaction mixture was concentrated and purified by flash column chromatography (30:1 Hex:EtOAc) to give, in order of elution, a small amount of the pyrrole derivative 17, contaminated with a small amount of co-eluting remaining starting material 14, and the major product, 16 (6.3 mg, 0.019 mmol, 30%) as a white crystalline solid. The ratio of 17:16 in this experiment was ca. 1:2 (analysis of the ¹H NMR spectrum of the crude product mixture prior to chromatographic purification). The pyrrole derivative showed evidence of slowly giving rise to several new compounds over time and handling (TLC and NMR). The experiment was repeated on a larger scale [50 mg 14 in 1.5 mL of DCE in the presence of added BHT (0.1 equiv); 130 °C, ca. 16 h]. The ratio of products 17:16 was now 1:6 (crude NMR). MPLC (5:1 Hex:EtOAc) provided a purer sample of the pyrrole derivative 17 (10 mg, 10%, corrected for residual EtOAc and hexanes, which were purposely left to minimize loss of the somewhat volatile product; this is the sample for which the characterization data were recorded) and the isoindoline 16 (30 mg, 60%), respectively. 16: mp: 138–142 °C. ¹H-NMR (500 MHz, CDCl₃): δ 7.76 (nfod, J_app = 8.2 Hz, 2H), 7.31 (nfod, J_app = 8.0 Hz, 2H), 6.95 (dq, J = 1.6, 0.8 Hz, 1H), 6.88 (br dq, J = 1.6, 0.8 Hz, 1H), 5.18 (dq, J = 1.5, 1.5 Hz, 1H), 4.92 (dq, J = 1.4, 1.0 Hz, 1H), 4.60 (nfom, 2H), 4.57 (nfom, 2H), 2.40 (s, 3H), 2.30 (dd, J = 0.8, 0.8 Hz, 3H), and 2.04 (dd, J = 1.5, 1.0 Hz, 3H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 143.8, 143.1, 138.4, 138.0, 136.9, 133.8, 130.4, 129.9, 127.7, 127.2, 121.9, 115.5, 53.70, 53.69, 23.4, 21.6, and 21.3. IR (neat): ν_max 3016 (C_sp2-H), 2920 (C_sp3-H), and 2847 (C_sp3-H) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₉H₂₂NO₂S⁺ [M + H]⁺ 328.1366; found 328.1351 (100%). 17: ¹H-NMR (500 MHz, CDCl₃): δ 7.72 (nfod, J_app = 8.3 Hz, 2H), 7.25 (nfod, J_app = 8.2 Hz, 2H), 7.03 (br d, J = 2.3 Hz, 1H), 6.82 (d, J = 2.1 Hz, 1H), 6.07 (tqd, simulated as J = 2.1, 1.3, 0.8 Hz, 1H), 3.05–2.97 (m, ΣJs = ~14 Hz, 2H), 2.38 (t, J = 0.8 Hz, 3H), 1.93 (br t, J = 2.2 Hz, 3H), 1.82 (br m, 3H), and 1.81 (dt, J = 1.3, 1.3 Hz, 3H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 144.7, 136.6, 136.3, 130.0, 129.4, 126.84, 126.78, 124.0, 120.6, 117.8, 114.0, 113.3, 35.7, 24.1, 23.4, 22.3, and 21.7. HRMS (ESI-TOF): Calcd for C₁₉H₂₂NO₂S⁺ [M + H]⁺ 328.1366; found 328.1357. This compound showed signs of oxidative decomposition upon handling and storage, and the HRMS data showed ions consistent with such processes (M + H⁺–H₂, M + H⁺ + O, M + H⁺ – H₂ + O).

6-Methyl-4-(prop-1-en-2-yl)-2-tosylisoindoline-7-d (16-d) and 6-Methyl-4-(propan-2-ylidene)-2-tosyl-4,5-dihydro-2H-isoindole-7-d (17-d). In a 10 mL culture tube, equipped with a magnetic stir bar, 14 (21 mg, 0.064 mmol) was added. 1,2-DCE (1.8 mL) and CD₃OD (0.2 mL) were then added. The reaction mixture was heated at 130 °C for 16 h. The reaction mixture was concentrated and purified by flash column chromatography (30:1 Hex:EtOAc) to give 16-d (6.12 mg, 0.019 mmol, 29%) as a white crystalline solid. The ratio of 17-d:16-d in this experiment was ca. 1:2 (analysis of the ¹H NMR spectrum of the crude product mixture prior to chromatographic purification). 16-d: mp: 136–140 °C. ¹H-NMR (500 MHz, CDCl₃): δ 7.76 (nfod, J_app = 8.3 Hz, 2H), 7.31 (nfod, J_app = 8.1 Hz, 2H), 6.95 (q, J = 0.8 Hz, 1H), 6.88 (s, 0.09H), 5.18 (dq, J = 1.6, 1.6 Hz, 1H), 4.92 (dq, J = 1.4, 1.0 Hz, 1H), 4.60 (nfom 2H), 4.57 (nfom 2H), 2.40 (s, 3H), 2.30 (d, J = 0.8 Hz, 3H), and 2.04 (dd, J = 1.5, 1.1 Hz, 3H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 143.7, 143.1, 138.4, 137.9, 136.8, 133.8, 130.5, 129.9, 127.8, 127.2, 115.5, 53.71, 53.68, 23.4, 21.6, and 21.3. (No resonance was observed at 122.0 ppm for the now-deuterated carbon.) IR (neat): ν_max 3016 (C_sp2-H), 2919 (C_sp3-H), and 2848 (C_sp3-H) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₉H₂₁DNO₂S⁺ [M + H]⁺ 329.1429; found 329.1414 (100%). 17-d: ¹H-NMR (500 MHz, CDCl₃): δ 7.72 (nfod, J_app = 8.3 Hz, 2H), 7.25 (nfod, J_app = 8.2 Hz, 2H), 7.04 (br d, J = 2.1 Hz, 1H), 6.82 (d, J = 2.1 Hz, 1H), 6.07 (m, 0.1H), 3.00–2.97 (m, ΣJs = ~14 Hz, 2H), 2.38 (t, J = 0.9 Hz, 3H), 1.93 (td, J = 2.0, 0.9 Hz, 3H), 1.82 (br m, 3H), and 1.81 (t, J = 1.2 Hz, 3H).

Methyl 6-Methyl-2-tosylisoindoline-4-carboxylate (33). In a 10 mL culture tube, equipped with a magnetic stir bar, 30 (30 mg, 0.087 mmol) was added. 1,2-DCE (2 mL) was added. The reaction mixture was heated to 130 °C for 16 h. The reaction mixture was concentrated and purified by flash column chromatography (10:1 Hex:EtOAc) to give 33 (12.61 mg, 0.037 mmol, 42%) as a white crystalline solid. mp: 175–180 °C. ¹H-NMR (500 MHz, CDCl₃): δ 7.78 (nfod, J_app = 8.0 Hz, 2H), 7.72 (s, 1H), 7.30 (nfod, J_app = 7.9 Hz, 2H), 7.16 (s, 1H), 4.87 (s, 2H), 4.60 (s, 2H), 3.89 (s, 3H), 2.39 (s, 3H), and 2.35 (s, 3H). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 166.4, 143.8, 138.3, 137.9, 135.7, 134.0, 130.2, 130.0, 127.73, 127.72, 125.2, 54.9, 53.2, 52.3, 21.6, and 21.2. IR (neat): ν_max 2952 (C_sp3-H), 2923 (C_sp3-H), 2856 (C_sp3-H), and 1721(C=O) cm⁻¹. HRMS (ESI-TOF): Calcd for C₁₈H₂₀NO₄S⁺ [M + H]⁺ 346.1108; found 346.1094 (100%).

Methyl 6-Methyl-2-tosylisoindoline-4-carboxylate-7-d (33-d). In a 10 mL culture tube, equipped with a magnetic stir bar, 30 (21 mg, 0.064 mmol) was added. 1,2-DCE (1.8 mL) and CD₃OD (0.2 mL) were then added. The reaction mixture was heated at 130 °C for 16 h. The reaction mixture was concentrated and purified by flash column chromatography (10:1 Hex:EtOAc) to give methyl 6-methyl-2-tosylisoindoline-4-carboxylate-7-d (33-d, 5.90 mg, 0.017 mmol, 27%) as a white crystalline solid: mp: 186–190 °C. ¹H-NMR (500 MHz, CDCl₃): δ 7.78 (nfod, J_app = 8.3 Hz, 2H, H2), 7.72 (q, J = 0.8 Hz, 1H, H5′), 7.30 (nfod, J_app = 8.2 Hz, 2H, H3), 7.16 (s, 0.1H, H7′), 4.87 (nfot, J_app = 2.1 Hz, 2H, H2′), 4.60 (nfot, J_app = 2.2 Hz, 2H, H9′), 3.89 (s, 3H, CO₂CH₃), 2.39 (s, 3H, C4CH₃), and 2.35 (d, J = 0.9 Hz, 3H, C6′CH₃). ¹³C-NMR{¹H} (126 MHz, CDCl₃): δ 166.4, 143.8, 138.2, 137.8, 135.7, 134.0, 130.2, 130.0, 127.74, 125.3, 54.9, 53.2, 52.3, 21.7, and 21.1. (No resonance was observed at 127.72, the largely deuterated C7′). IR (neat): ν_max 2952 (C_sp3-H), 2921 (C_sp3-H), 2850 (C_sp3-H), and 1721(C=O) cm⁻¹. HRMS (ESI-TOF): [M + H]⁺ Calcd for C₁₈H₁₉DNO₄S⁺ 347.1170; Found 347.1155 (100%).

4. Conclusions

The symmetrical bis-enyne 14 represents a rare type of substrate used for study of a thermally driven tetradehydro-Diels–Alder reaction. As such, the initially produced strained 1,2,4-cyclohexatriene intermediate 15 uniquely had an alkene (a 2-propenyl group) attached to C5. Heating 14 gave rise to not only the expected aromatic benzenoid product, 16, but a second, unusual isomeric side product, 17, whose structure was not immediately apparent. After detailed NMR analyses, it was concluded that 17 was a pyrrole derivative. Its formation is best rationalized by a [1,5]-hydrogen atom migration within 15 (or 21; cf. Figure 3). Deuteration studies using CD₃OD or (CH₃OD) showed that the presence of a protic additive was involved in forming both products 16 and 17. When the more acidic protic additive BHT was used, formation of the pyrrole side product 17 was completely suppressed, suggesting that the acidity of the OH additive played an important role in determining the fate of the 1,2,4-cyclohexatriene derivative 15.