Novel [4 + 2] cycloaddition reactions of alkyne and enyne key-units: Direct access to bicyclic aromatic and heteroaromatic products. A theoretical mechanistic study

Abstract

A new efficient approach was developed for the synthesis of aromatic and heteroaromatic compounds based on [4 + 2] cycloaddition of unsubstituted and heteroatom-substituted alkyne and enyne units. The developed approach provides a practical Green chemical route to several types of important bicyclic products (indane, cyclopentapyridines, indole, isoindole, indolizine, isophosphindole, benzofuran, benzothiophene, benzoselenophene and corresponding dihydro derivatives) starting from simple linear compounds. The mechanism of the reactions was revealed by theoretical calculations using different methods, including CCSD(T) and MP4(SDTQ) for energy calculations and B3LYP, M052X, B3PW91, BLYP and MP2 levels for evaluation of molecular structures.

---

1. Introduction

Development of new Green strategies for the synthesis of aromatic and heteroaromatic compounds is the problem of key importance in modern organic chemistry. Numerous bioactive compounds, pharmaceutical substances and natural products are based on aromatic and heteroaromatic cores. Equally important, several classes of aromatic and heteroaromatic derivatives are in demand in material science applications and govern “smart” features of new generations of advanced materials.

The classical approach for the synthesis of (hetero)aromatic compounds employs a substitution pathway, which is always accompanied by production of waste (as with any substitution reaction). Cycloaddition methodology is an attractive Green alternative with 100% atom efficiency of the addition reaction, which avoids waste.¹ The cycloaddition strategy has an important practical advantage, because it is a single step route to substituted products where the position of the substituent can be easily controlled by the structure of the initial reagent. In fact, a challenging problem of selective preparation of cyclic (as well as bicyclic and even polycyclic) products with all the substituents in place can be solved using a cycloaddition strategy.^2,3

[4 + 2] Cycloaddition of enyne and alkyne units is a fascinating direct route to construct six-membered aromatic rings. In a simple model case the intermolecular cycloaddition reaction of vinylacetylene and acetylene leads to the formation of benzene (Scheme 1). Experimental and theoretical studies were carried out to investigate the nature of the reaction and have suggested intermediate involvement of strained cyclic alleneIII (1,2,4-cyclohexatriene).⁴ In spite of great mechanistic interest, synthetic application of this reaction was limited due to rather harsh experimental conditions required to carry out the intermolecular transformation.

Scheme 1 Intermolecular [4 + 2] cycloaddition reaction between acetylene and vinylacetylene (but-1-en-3-yne).

Excellent practical utilization of this cycloaddition strategy was demonstrated by Danheiser and co-workers via introducing yne-yne link and facilitating intramolecular cycloaddition (Scheme 2).⁵ Further studies in the field resulted in the development of an efficient and flexible modular approach for the synthesis of important classes of aromatic and heteroaromatic compounds.⁶ The mechanistic studies of both intermolecular (Scheme 1) and intramolecular (Scheme 2) reactions were carried out and have revealed preferable factors for the cycloaddition step involving an intramolecular pathway.^7,8 Smaller activation barriers due to lowered activation entropy and favorable intrinsic structural arrangements of reagent I afforded formation of cycloaddition products in good to high yields under simple experimental conditions (Scheme 2).^5,6

Scheme 2 Intramolecular [4 + 2] cycloaddition reaction governed by yne-yne link in non-1-ene-3,8-diyne (I).

Inspired by Schreiner's recent concept on the description of reaction space of addition reactions⁹ and in light of our previous studies^7,8 we have carried out a systematic theoretical investigation of the interconversions of alkyne and enyne key-units. Surprisingly, we have found six new intramolecular [4 + 2] cycloaddition reactions governed by an ene-yne link (Scheme 3). A detailed study of the reaction mechanism was carried out using density functional theory (B3LYP, BLYP, B3PW91 and M052X), Møller–Plesset methods (MP2, MP4(SDTQ)) and coupled cluster CCSD(T) calculations. Reliable synthetic pathways to aromatic and heteroaromatic products were designed based on the theoretical mechanistic study.

Scheme 3 Proposed in the present study an intramolecular [4 + 2] cycloaddition reaction governed by ene-yne link in non-3-ene-1,8-diyne and corresponding heteroatom derivatives

2. Results and discussion

A detailed study of the [4 + 2] cycloaddition reaction was carried out for both isomers: E-non-3-ene-1,8-diyne 1 and Z-non-3-ene-1,8-diyne 11 (Scheme 4). Calculated ΔE, ΔH and ΔG data are provided in Table 1 and Table S1 (Supporting Information†), and optimized geometry parameters are listed in Table 2. The energy surface and molecular structures are shown on Fig. 1 and 2, respectively.

Scheme 4 The studied mechanism of the intramolecular [4 + 2] cycloaddition reaction of E- (1) and Z- (11) isomers of non-3-ene-1,8-diyne.

Table 1 Calculated energy surfaces at various theory levels (in kcal mol⁻¹).^a

Entry	Level	1	2-TS	3	4-TS	5	6-TS	7-TS	8	9-TS	10	11	12-TS	13	14-TS	Av. Dev.^e
a For structures see Scheme 4, for optimized molecules see Fig. 2. b Energy evaluation at indicated theory level using 6-311+G(d) basis set, geometry optimization at B3LYP/6-311+G(d) level. c Energy evaluation and geometry optimization at indicated theory level using 6-311+G(d) basis set. d Energy evaluation and geometry optimization at B3LYP/6-311++G(d,p) level. e Average deviation of corresponding ΔE values with those obtained at CCSD(T) level; calculated for points 1–10.^31
1	B3LYP (ΔE)^b	0.0	31.4	−16.3	33.9	4.2	17.2	8.7	−11.0	−7.6	−100.0	−1.3	35.6	−13.8	2.3	2.9
2	CCSD(T) (ΔE)^b	0.0	29.9	−23.9	28.0	−4.7	15.2	9.3	−10.6	−8.1	−99.9	−1.5	32.6	−19.9	1.8	0.0
3	MP2 (ΔE)^c	0.0	23.0	−20.5	22.2	−3.5	16.8	6.1	−7.4	−7.4	−104.1	−1.8	29.6	−17.1	2.1	2.8
4	MP4(SDTQ) (ΔE)^b	0.0	25.9	−22.6	24.7	−4.2	18.6	7.5	−9.8	−8.6	−100.8	−1.5	30.6	−18.8	0.7	1.6
5	BLYP (ΔE)^c	0.0	27.6	−13.0	33.7	6.2	15.4	12.3	−7.4	−2.6	−93.0	−1.2	29.6	−11.4	5.9	4.9
6	B3PW91 (ΔE)^c	0.0	27.5	−25.8	21.0	−6.1	6.8	−2.0	−19.6	−17.7	−109.7	−1.3	30.9	−23.2	−6.6	5.7
7	M052X (ΔE)^c	0.0	28.1	−24.1	28.0	−2.8	13.0	2.6	−15.4	−14.0	−108.3	−1.9	36.0	−20.3	−1.7	3.1
8	B3LYP (ΔE)^d	0.0	32.5	−14.7	34.2	5.6	18.1	9.8	−9.0	−6.4	−98.7	−1.3	37.1	−12.1	4.2	3.9

---

Table 2 Optimized geometry parameters at B3LYP/6-311+G(d) level (bond length in Å, angles in deg); for atom numbering see Fig. 2.^a

	1	2-TS	3	4-TS	5	6-TS	7-TS	8	9-TS	10	11	12-TS	13	14-TS
a Imaginary frequencies for the transition states: 476.6 i cm^−1 (2-TS), 1542.0 i cm^−1 (4-TS), 625.1 i cm^−1 (6-TS), 1001.8 i cm^−1 (7-TS), 783.8 i cm^−1 (9-TS), 205.3 i cm^−1 (12-TS), 2340.0 i cm^−1 (14-TS).
C1–C2	1.207	1.252	1.331	1.408	1.494	1.471	1.409	1.423	1.444	1.396	1.206	1.284	1.341	1.423
C2–C3	1.421	1.386	1.329	1.359	1.364	1.435	1.422	1.497	1.459	1.397	1.422	1.387	1.327	1.390
C3–C4	1.341	1.370	1.510	1.483	1.479	1.408	1.475	1.485	1.447	1.392	1.343	1.373	1.528	1.488
C4–C5	1.499	1.494	1.538	1.520	1.507	1.517	1.516	1.507	1.509	1.514	1.500	1.496	1.545	1.547
C5–C6	1.540	1.541	1.541	1.546	1.551	1.547	1.547	1.551	1.551	1.548	1.540	1.546	1.549	1.543
C6–C7	1.544	1.541	1.549	1.552	1.552	1.547	1.551	1.552	1.551	1.548	1.543	1.563	1.565	1.550
C7–C8	1.461	1.472	1.514	1.509	1.511	1.511	1.512	1.514	1.513	1.514	1.461	1.449	1.512	1.506
C8–C9	1.203	1.242	1.343	1.401	1.486	1.424	1.385	1.426	1.419	1.392	1.203	1.264	1.352	1.361
C4–C8	3.220	2.447	1.558	1.452	1.378	1.410	1.414	1.359	1.369	1.400	3.954	2.841	1.519	1.480
C1–C9	5.129	2.054	1.489	1.417	1.363	1.384	1.410	1.385	1.378	1.397	4.886	1.673	1.466	1.431
C1–H	1.064	1.073	1.087	1.085	1.085	1.090	1.090	1.089	1.088	1.086	1.064	1.081	1.088	1.093
C3–H	1.089	1.088	1.087	1.090	1.096	1.093	1.090	1.103	1.090	1.087	1.088	1.089	1.087	1.094
							1.315	1.107	1.203
C4–H	1.088	1.086	1.102	1.463	2.434	2.243	1.286	—	—	—	1.089	1.091	1.103	1.117
C2–H	—	—	—	1.391	1.096	1.137	2.268	—	1.457	1.086	—	—	—	—
C9–H	1.064	1.073	1.087	1.086	1.088	1.088	1.089	1.091	1.089	1.087	1.064	1.085	1.087	1.088
C1–C2–C3	178.3	138.4	130.8	123.0	112.0	118.1	113.2	113.3	111.8	120.4	177.6	134.0	129.3	111.3
C2–C3–C4	124.4	116.1	111.6	94.7	106.5	115.3	122.8	121.2	123.6	119.2	125.6	124.1	115.2	125.0
C9–C8–C7	178.2	163.0	129.7	130.7	124.9	125.6	131.1	128.7	128.9	129.2	178.9	171.9	127.5	130.8

---

Fig. 1 Calculated ΔG surface of the studied reaction (see Scheme 4 and Fig. 2 for structures).

Fig. 2 B3LYP/6-311+G(d) optimized molecular structures of initial compounds, transition states, intermediates and products of the studied reaction. Displacement vectors corresponding to imaginary frequency are shown for each transition state (see Scheme 4 for structures and Table S2† for optimized geometry parameters).

Calculations at the B3LYP level will be considered first using a mechanistically reliable ΔG energy surface (Sections 2.1), followed by discussion of the CCSD(T) energy surface, calculations at various theory levels (MP2, MP4(SDTQ), BLYP, B3PW91 and M052X) as well as computations applying a larger basis set (Section 2.2).

Comparison with available experimental and theoretical data on similar reactions will be provided next (Section 2.3). Finally, the reaction leading to different heterocyclic products will be revealed using the studied [4 + 2] cycloaddition strategy (Sections 2.4 and 2.5).

2.1 B3LYP study of the [4 + 2] cycloaddition reaction of non-3-ene-1,8-diyne

Starting from the E-non-3-ene-1,8-diyne 1 transition state of the [4 + 2] cycloaddition reaction 2-TS was successfully located at the B3LYP/6-311+G(d) level. Backward IRC calculations were performed to locate the structure of the initial compound 1, and forward IRC calculations have revealed cyclic allene3 as the product of this cycloaddition step.

Starting from 3 one of the possible pathways involves a 1,3-hydrogen shift (4-TS) leading to formation of structure 5 with a “Möbius benzene”^7,8 arrangement of the six-membered ring. The isomerization of 5 through the transition state 6-TS led to the final product 10. Another pathway includes a 1,2-hydrogen shift (7-TS) with intermediate formation of cyclic carbene8 and affords the same final product 10 after the second 1,2-H shift through the 9-TS.

The calculated energy surface indicated the activation barrier of ΔG^≠ = 34.5 kcal mol⁻¹ for the cycloaddition step, and formation of allene3 resulted in an energy gain of ΔG = −9.7 kcal mol⁻¹ (Fig. 1). Comparison of the two possible pathways of further transformation of 3via the 4-TS, 6-TS or via the 7-TS, 9-TS led to selection of the latter as more energetically favored. The activation barrier of 1,2-H shift ΔG^≠ = 23.1 kcal mol⁻¹ (3 → 7-TS) was calculated as much lower compared to the barrier of 1,3-H shift ΔG^≠ = 47.2 kcal mol⁻¹ (3 → 4-TS). Cyclic carbene8 was more stable by ΔG = −6.6 kcal mol⁻¹ relative to initial compound 1 and less stable by ΔG = 3.1 kcal mol⁻¹ relative to cyclic allene3. The last 1,2-H shift through the 8 → 9-TS step was characterized with a very small activation barrier of ΔG^≠ = 3.8 kcal mol⁻¹. Formation of the final product 10 was calculated to be highly exothermic ΔG = −91.4 kcal mol⁻¹, thus providing the necessary driving force for the transformation.

Starting from Z-non-3-ene-1,8-diyne 11 as the initial compound the transition state of the [4 + 2] cycloaddition reaction 12-TS was located. The calculated energy barrier was somewhat higher, ΔG^≠ = 40.5 kcal mol⁻¹, and formation of cyclic allene13 was less exothermic, ΔG = −6.0 kcal mol⁻¹, compared to the reaction involving E-geometry of the double bond (ΔG^≠ = 34.5 kcal mol⁻¹, ΔG = −9.7 kcal mol⁻¹). Rearrangement of the hydrogen atoms at C3 and C4 centers from syn- to anti- orientation took place via the inversion-like transition state 14-TS, which connected 13 with more stable isomer 3. The calculated 13 → 14-TSactivation barrier was relatively small, ΔG^≠ = 15.1 kcal mol⁻¹. The further transformation involving 1,2-H shifts led to the final product 10 as discussed above.

Optimized structures of 2-TS and 12-TS and IRC calculations confirmed concerted nature of the [4 + 2] cycloaddition step: both carbon-carbon bonds C1–C9 and C4–C8 were formed upon movement along the reaction coordinate toward allene intermediates 3 and 13. However, molecular geometries of the studied cycloaddition transition states significantly differ (Table 2 and Fig. 2). A more synchronous transition state 2-TS was located for the cycloaddition involving E-isomer 1, C1–C9 = 2.054 Å and C4–C8 = 2.447 Å, whereas a highly asynchronous transition state 12-TS was found for the cycloaddition of Z-isomer 11, C1–C9 = 1.673 Å and C4–C8 = 2.841 Å. Clearly, the degree of asynchronicity with the Z-geometry of the double bond was much larger compared to the same reaction starting with E-geometry: Δ = 0.393 Å and Δ = 1.168 Å for 2-TS and 12-TS, respectively.¹⁰

Formation of a six-membered ring resulted in elongation of the triple bonds which was more pronounced for the Z-isomer, C1–C2 = 1.252 Å (2-TS), 1.284 Å (12-TS) and C8–C9 = 1.242 Å (2-TS), 1.264 Å (12-TS). In contrast, the C=C unit was more conservative and the length of the double bond changed only slightly, C3–C4 = 1.370 Å (2-TS), 1.373 Å (12-TS). Thus, a less favorable Z-geometry arrangement of the conservative C=C unit was compensated by more flexible terminal triple bonds. In agreement with calculated geometry trends, the earlier transition state 2-TS was characterized by a smaller activation barrier compared to more late transition state 12-TS (Scheme 4).

Further transformation of cyclic allene3via an energetically favored pathway involved two transition states (7-TS and 9-TS) with geometry corresponding to a 1,2-H shift process (Table 2). In the final product 10 a benzene ring was formed with C–C bonds lengths 1.39–1.40 Å typical for an aromatic system. Selected geometry parameters of all optimized stationary points are shown in Table 2 and will not be discussed in detail here. It is interesting to point out that related allowed 1,5-shifts for the out-of-plane π-system were reported by Alabugin and co-workers.¹¹ Such shifts accompanied with aromatization are known to be relatively fast.¹¹

2.2 Calculations at different theory levels

To check the reliability of the calculations the energy surface was computed at different theory levels and basis sets. Comparing ΔE energies at CCSD(T) (Entry 2, Table 1) and B3LYP (Entry 1, Table 1) levels we may notice excellent agreement in the activation barriers of the cycloaddition step, ΔE^≠ = 29.9 kcal mol⁻¹ and 31.4 kcal mol⁻¹. Cyclic allene intermediate 3 was found to be even more stable at the CCSD(T) level, ΔE = −23.9 kcal mol⁻¹ and −16.3 kcal mol⁻¹, respectively. The average deviation between the CCSD(T) and B3LYP energy surfaces was rather small for the studied reaction, 2.9 kcal mol⁻¹ (Table 1).

In terms of average deviation between the calculated energy values, very good agreement was obtained between CCSD(T)//B3LYP and MP4(SDTQ)//B3LYP surfaces (1.6 kcal mol⁻¹). The M052X level performed well (3.1 kcal mol⁻¹), while worse agreement was observed with BLYP (4.9 kcal mol⁻¹) and B3PW91 (5.7 kcal mol⁻¹) functionals (Table 1). Re-optimization of the energy surface with a larger basis set led to only a minor change of the calculated energy surface (Entries 1, 8; Table 1).

Thus, the mechanistic conclusions were found to be valid at all theory levels used in the present study. The studied [4 + 2] cycloaddition reaction takes place with the activation barrier of 26–32 kcal mol⁻¹ and the formation of the allene intermediate is an exothermic process. Further transformation of the allene3via the 1,2-H shifts is a more favorable pathway and formation of the final product 10 is a highly exothermic and irreversible reaction.

2.3 Comparison with known intermolecular and intramolecular reactions of non-1-ene-3,8-diyne

Since for the cycloaddition reaction directed by yne-yne link both experimental and theoretical studies are available it would be worthwhile to make a comparison. Theoretical studies of the intermolecular reaction (Scheme 1) and intramolecular cycloaddition of non-1-ene-3,8-diyne (Scheme 2) were reported eariler,^7,8 however with a smaller basis set. For a proper comparison both reactions were re-optimized at B3LYP/6-311+G(d) level. Calculated relative energies of stationary points at B3LYP/6-311+G(d) level were ΔG = 0.0 (I), 41.8 (II-TS) and −11.6 (III) kcal mol⁻¹ for intermolecular reaction (Scheme 1) and ΔG = 0.0 (I), 36.7 (II-TS) and −8.2 (III) kcal mol⁻¹ for the corresponding intramolecular reaction (Scheme 2). Obtained energy parameters and optimized molecular structures well agree with those reported previously. Calculated relative energies at the CCSD(T)/6-311+G(d) level were ΔE = 0.0 (I), 29.9 (II-TS) and −32.7 (III) kcal mol⁻¹ for intermolecular reaction (Scheme 1) and ΔE = 0.0 (I), 30.4 (II-TS) and −22.9 (III) kcal mol⁻¹ for the corresponding intramolecular reaction (Scheme 2).

From experiment it is known that cycloaddition reactions corresponding to Scheme 2 took place at 180–250 °C and resulted in product formation in good to high yields.⁵ Several other examples of this cycloaddition strategy were demonstrated under similar conditions and provided a useful practical route to substituted bicyclic derivatives.^4,6 The proposed mechanism of the reaction involving the cyclic allene and carbene intermediates was confirmed by trapping experiments and fragmentation reactions.^4–6 For convenient practical application microwave-accelerated experimental procedures were reported for heating reaction mixtures of cycloaddition reactions (100–250 °C).³

The reaction explored in the present study (Schemes 3, 4) was characterized by the calculated activation energy for the cycloaddition step and stability of the allene key-intermediate very similar to the experimentally proven transformations (Schemes 1, 2). Thus, it is natural to expect the cycloaddition approach designed in the present study and governed by the ene-yne link to take place as a thermal reaction similar to the process involving an yne-yne link. As a further practical extension of new transformations described in the present study, known reagents and metal catalysts may be added to carry out the reactions under more mild conditions.^2,3

2.4 [4 + 2] Cycloaddition reactions leading to N-heteroaromatic compounds

Synthesis of heteroaromatic compounds was studied starting from amines (15, 19, 23), nitriles (26, 30) and imines (34, 37) as initial reagents. For convenient discussion calculated energy is reported relative to the starting point in each reaction (Scheme 5). Relative stability of reagents, intermediates and products reflecting the energy difference between each other is given in Supporting Information (Figures S1–S3).

Scheme 5 Cycloaddition reactions leading to heterocyclic products with calculated ΔE values at CCSD(T)/6-311+G(d)//B3LYP/6-311+G(d) level (without parenthesis) and ΔG values at B3LYP/6-311+G(d) level (in parenthesis); in kcal mol⁻¹ (see Supporting Information† Table S4 and Scheme S1 for another representation of the energy surface).^13,14

Transition states 16-TS, 20-TS, and 24-TS were located for the [4 + 2] cycloaddition of amines with the activation barrier at CCSD(T) level of 29.1–30.3 kcal mol⁻¹ (Scheme 5a–5c). In spite of similar activation energies, stability of cyclic allene intermediates was different: −18.5 kcal mol⁻¹ (17), −25.7 kcal mol⁻¹ (21), and −33.0 kcal mol⁻¹ (25). Indoline 18 and isoindoline 22 were calculated as the final heteroaromatic products, which can be formed after a series of H-shifts.¹² Comparison with the 1 → 2-TS → 3cycloaddition step discussed above, the activation barrier of 29.9 kcal mol⁻¹ and stability of cyclic allene of −23.9 kcal mol⁻¹ were calculated in the absence of nitrogen atoms (Entry 2, Table 1).

The [4 + 2] cycloaddition of nitriles26 and 30 was characterized with high activation barriers of 43.4 and 39.0 kcal mol⁻¹, respectively (Scheme 5d–5e). Formation of allene intermediates was found to be endothermic by 14.3 and 14.0 kcal mol⁻¹, respectively. The reaction leading to pyridine products was calculated to be exothermic by −58.9 and −59.5 kcal mol⁻¹ with a smaller energy gain when compared to formation of indoline (18), isoindoline (22) and indane (10) products −98.5–−105.5 kcal mol⁻¹ (Scheme 5, 4 and Table 1).

For the synthesis of the pyridine system 33 the reaction involving imine34 should be a more useful alternative than transformation of the nitrile30. A smaller activation energy of 32.0 kcal mol⁻¹ (35-TS) and better stability of the intermediate allene −23.3 kcal mol⁻¹ (36) were found to be more favorable for this transformation. Larger energy gains in the formation of product 33 (cf.Scheme 5e and 5f) reflects a contribution from relative stability of the reagents 30 and 34 (see Figure S1†).

Cycloaddition of imine37 was found to be unique among the studied reactions (Scheme 5g). The most stable cyclic allene39 (−40.2 kcal mol⁻¹) and the most stable N-heterocylcic carbene intermediate 40 (−56.1 kcal mol⁻¹) were found in the calculations. Possible rearrangement of the carbene may involve H-shifts from five-membered rings leading to dihydroindolizine derivates 41 and 42.

Comparing calculated energy surfaces at CCSD(T) and B3LYP levels we may notice the tendency outlined already for the reaction leading to product 10 (Table 1). Calculated activation energies for the cycloaddition step were in a good agreement, while cyclic allene intermediates were found to be more stable at the CCSD(T) level. Both theory levels confirm the exothermic nature of the studied cycloaddition reactions on the final product formation stage. Optimized geometries of all stationary points for the transformations involving heteroatom-substituted molecules as well as reactions of cis-34 and cis-37 are provided in the Supporting Information (Scheme S2, Figures S4–S6, Table S5†).

It is interesting to point out, that compounds 36 and 39 significantly differ in relative stability: ΔE = E(39) − E(36) = −24.2, −16.9 and −23.0 kcal mol⁻¹ at the B3LYP/6-311+G(d), CCSD(T)/6-311+G(d)//B3LYP/6-311+G(d) and MP2(full)/6-311++G(d,p) levels of theory, respectively. Thus, compound 39 with a tertiary nitrogen atom is much more thermodynamically preferred, than molecule 36 with two carbon substituents at the nitrogen atom. Compound 39 has more formal “aromatic” attributes compared to 36: a six-membered heterocycle is planar (36 is not planar) and interatomic distances of six-membered heterocycles are close to one another (in 36 there is a marked difference). We have applied Bader quantum chemical theory of atoms in molecules¹⁵ to reveal the difference between structures 36 and 39 and have found that molecule 39 possess more similar electronic features to the aromatic pyridine molecule, rather than molecule 36, in total agreement with observed thermodynamic stability.¹⁶ This electronic structure also suggests that the reaction is somewhat mechanistically different from the other cases discussed in the present study. It may be also considered as a pseudopericyclic process, which starts with nucleophilic attack of the nitrogen at the yne terminus. A similar consideration has been reported for radicals “masquerading as electrophiles” and this mechanism would make the process a new and quite interesting type of cycloaromatization reaction.^17,18

It should be pointed out that indoline (18), isoindoline (22), and dihydroindolizine (41, 42) can be easily converted to indole, isoindole and indolizine derivatives using well-established experimental methods.¹⁹ Thus, providing an easy route to bicyclic heteroaromatic products.

2.5 [4 + 2] Cycloaddition reactions leading to P, O, S and Se-heterocyclic compounds

Several different heteroatoms can be incorporated in the starting material and provide an easy access to heterocyclic products in the framework of the investigated cycloaddition approach. Compound 19 is of particular interest as far as the library of corresponding reagents should be easily available for various heteroatoms.

The calculated energy surfaces for the reactions of phosphorus (43), oxygen (47), sulfur (51) and selenium (55) analogs indicated activation energy 28–31 kcal mol⁻¹ and relative stability of the allene intermediate of about −25 kcal mol⁻¹ at CCSD(T) level (Scheme 6). All transformations were characterized with a large energy gain of about −103 kcal mol⁻¹. Higher energy barriers and slightly lower stability of cyclic allenes and products were found on the ΔG energy surface calculated at the B3LYP level. The energy parameters well agree with those calculated for the nitrogen derivatives (Scheme 5b).

Scheme 6 Cycloaddition reactions leading to heterocyclic products calculated ΔE values at CCSD(T)/6-311+G(d)//B3LYP/6-311+G(d) level (without parenthesis) and ΔG values at B3LYP/6-311+G(d) level (in parenthesis); in kcal mol⁻¹ (see Supporting Information† Table S6 for energy surface, and Figure S7 and Table S7 for optimized structures).^20,14

Indeed, the calculations have shown that the studied cycloaddition reactions should be suitable for preparation of a series of heterocyclic compounds, such as 2,3-dihydro or 1,3-dihydro derivatives of isophosphindole (46), 2-benzofuran (50), 2-benzothiophene (54), and 2-benzoselenophene (58). Removal of two hydrogen atoms and formation of heteroaromatic rings can be carried out using known experimental procedures.¹⁹

3. Conclusions

To summarize, a family of new [4 + 2] cycloaddition reactions was designed in the present study for a convenient practical synthesis of aromatic and heteroaromatic products. Several novel mechanistic findings were revealed which are worthy of note.

The mechanism of the reaction of non-3-ene-1,8-diyne (1) (Scheme 3) was proposed to involve cycloaddition and a series of H-shifts with the final formation of the indane core (10). Both E- and Z- derivatives can be involved in the reaction and furnish the same final product. According to the calculations, the E-isomer was predicted to be more reactive, although the difference in reactivity was small and the E-/Z- mixture may be utilized in the reaction as a starting material. The reliability of the calculations was confirmed at various theory levels.

Using nitrogen-substituted reagents all possible cycloaddition reactions were investigated to develop a new strategy for the synthesis of heteroaromatic products (Scheme 4). The calculations revealed the following order of relative reactivity in the studied cycloaddition reaction: amines > imines > nitriles. Thus, the reaction of amines with the final formation of indoline (18) and isoindoline (22) cores should be possible in more mild conditions at lower temperature compared to the synthesis of pyridine systems (29, 33) from imines and even less reactive nitriles.

Cycloaddition reactions involving imines are of much interest and can result in the formation of pyridine (33) or dihydroindolizine (41, 42) cores depending on the position of nitrogen atom in the starting reagent. It is worthwhile mentioning that synthesis of compounds 18, 22, 41 and 42 may be followed by hydrogen abstraction with well-known experimental procedures to prepare indole, isoindole and indolizine products.

Important to note, according to the theoretical predictions formation of stable cyclic allene (39) and carbene (40) intermediates can be expected in the reaction of imine37. In a certain case such intermediate species can be directly observed by experimental methods or analyzed using known trapping techniques.

A series of new reactions were developed in the present study based on [4 + 2] cycloaddition of alkyne and enyne key units as well as various substituted units. The designed reactions utilize Green chemistry potential of cycloaddition approach and provide a convenient synthetic route to important classes of aromatic and heteroaromatic products. We encourage implementing the developed reactions and investigating their scope under experimental conditions.

4. Computational details

The calculations were carried out with the standard 6-311+G(d) basis set for H, C, N, P, O, S and Se.²¹ Geometry optimization and energy calculations were also performed with additional polarization and diffusion functions on the hydrogen atoms using the 6-311++G(d,p) basis set²¹ (Entry 8, Table 1).

Geometry optimization of the initial compounds, transition states and products of the reactions was carried out using the B3LYP hybrid density functional method.²² For the studied structures normal coordinate analysis was performed to characterize the nature of the stationary points and to calculate thermodynamic properties (298.15 K and 1 atm). Transition states were confirmed with IRC (Intrinsic Reaction Coordinate) calculations using the standard method.²³

Geometry optimization and energy calculations were also carried out with MP2 (Entry 3, Table 1),²⁴BLYP (Entry 5, Table 1),²²B3PW91 (Entry 6, Table 1),^22,25 and M052X (Entry 7, Table 1)²⁶ methods using 6-311+G(d) basis set.

Single point energy calculations were carried out at CCSD(T)/6-311+G(d) level²⁷ were performed for all studied reactions (Entry 2, Table 1; Sections 2.3–2.5) and single point calculations at MP4(SDTQ)/6-311+G(d) level²⁸ were executed for the cycloaddition of non-3-ene-1,8-diyne (Entry 4, Table 1). In both cases B3LYP/6-311+G(d) optimized geometries were used for the single point calculations.

In the previous studies it was established B3LYP, MP4(SDTQ) and CCSD(T) levels of theory reasonably well describe the energy and geometry parameters of the cycloaddition reactions of interest,⁸ the utility of M052X for studying cycloaddition reactions was reported recently.²⁹

All calculations were performed without any symmetry constraints using the Gaussian09 program.³⁰

References

1. (a) Green Chemistry: Theory and Practice, P. T. Anastas and J. C. Warner, Oxford University Press, Oxford, 2000; (b) R. A. Sheldon, Pure Appl. Chem., 2000, 72, 1233–1246; (c) B. M. Trost, Angew. Chem., Int. Ed. Engl., 1995, 34, 259–281; (d) R. A. Sheldon, Chem. Ind., 1992, 903–906.
2. Selected recent reviews: (a) R. Hua, M. Abrenica and P. Wang, Curr. Org. Chem., 2011, 15, 712–729; (b) H. Pellissier, Adv. Synth. Catal., 2011, 353, 189–218; (c) F. Inagaki, S. Kitagaki and C. Mukai, Synlett, 2011, 594–614; (d) P. Standen and M. Kimber, Curr. Op. Drug Disc. Devel., 2010, 13, 645–657; (e) T. Kitamura, Aust. J. Chem., 2010, 63, 987–1001; (f) K. J. Cahill, A. Ajaz and R. P. Johnson, Aust. J. Chem., 2010, 63, 1007–1012; (g) A. Pla-Quintana and A. Roglans, Molecules, 2010, 15, 9230–9251; (h) J. M. Palomo, Eur. J. Org. Chem., 2010, 6303–6314; (i) A. V. Ivachtchenko, Y. A. Ivanenkov, V. M. Kysil, M. Y. Krasavin and A. P. Ilyin, Russ. Chem. Rev., 2010, 79, 787–817; (j) N. A. Bokach, Russ. Chem. Rev., 2010, 79, 89–100; (k) V. Estévez, M. Villacampa and J. C. Menéndez, Chem. Soc. Rev., 2010, 39, 4402–4421; (l) B. Alcaide, P. Almendros and C. Aragoncillo, Chem. Soc. Rev., 2010, 39, 783–816; (m) H. Pellissier, Tetrahedron, 2009, 65, 2839–2877; (n) P. Merino, E. Marqués-López, T. Tejero and R. Herrera, Synthesis, 2010, 1–26; (o) K. Tadano, Eur. J. Org. Chem., 2009, 4381–4394; (p) B. B. Touré and D. G. Hall, Chem. Rev., 2009, 109, 4439–4486; (q) S. Perreault and T. Rovis, Chem. Soc. Rev., 2009, 38, 3149–3159; (r) M. Juhl and D. Tanner, Chem. Soc. Rev., 2009, 38, 2983–2992; (s) P. Wessig and G. Müller, Chem. Rev., 2008, 108, 2051–2063; (t) K. Takao, R. Munakata and K. Tadano, Chem. Rev., 2005, 105, 4779–4807; (u) G. Domínguez and J. Pérez-Castells, Chem. Soc. Rev., 2011, 40, 3430–3444.
3. P. Appukkuttan, V. P. Mehta and E. V. Van der Eycken, Chem. Soc. Rev., 2010, 39, 1467–1477.
4. (a) R. C. Burrell, K. J. Daoust, A. Z. Bradley, K. J. DiRico and R. P. Johnson, J. Am. Chem. Soc., 1996, 118, 4218–4219; (b) T. L. Gilchrist, J. Chem. Soc., Perkin Trans. 1, 1999, 2849–2886; (c) M. Christl, M. Braun and G. Müller, Angew. Chem., Int. Ed. Engl., 1992, 31, 473–476; (d) R. Janoschek, Angew. Chem., Int. Ed. Engl., 1992, 31, 476–478; (e) H. Hopf, H. Berger, G. Zimmermann, U. Nüchter, P. G. Jones and I. Dix, Angew. Chem., Int. Ed. Engl., 1997, 36, 1187–1190; (f) W. R. Roth, H. Hopf and C. Horn, Chem. Ber., 1994, 127, 1765–1779; (g) Z. Li, D. W. Rogers, F. J. McLafferty, M. Mandziuk and A. V. Podosenin, J. Phys. Chem. A, 1999, 103, 426–430; (h) K. M. Merz Jr. and L. T. Scott, J. Chem. Soc., Chem. Commun., 1993, 412–414; (i) H. F. Bettinger, P. R. Schreiner, H. F. Schaefer III and P.v. R. Schleyer, J. Am. Chem. Soc., 1998, 120, 5741–5750; (j) R. P. Johnson, Chem. Rev., 1989, 89, 1111–1124; (k) G. Zimmermann, Eur. J. Org. Chem., 2001, 457–471; (l) M. Prall, A. Wittkopp and P. R. Schreiner, J. Phys. Chem. A, 2001, 105, 9265–9274.
5. R. L. Danheiser, A. E. Gould, R. F. de la Pradilla and A. L. Helgason, J. Org. Chem., 1994, 59, 5514–5515.
6. (a) M. S. B. Wills and R. L. Danheiser, J. Am. Chem. Soc., 1998, 120, 9378–9379; (b) A. R. Renslo and R. L. Danheiser, J. Org. Chem., 1998, 63, 7840–7850; (c) D. T. Amos, A. R. Renslo and R. L. Danheiser, J. Am. Chem. Soc., 2003, 125, 4970–4971; (d) J. R. Dunetz and R. L. Danheiser, J. Am. Chem. Soc., 2005, 127, 5776–5777; (e) W. F. Austin, Y. Zhang and R. L. Danheiser, Org. Lett., 2005, 7, 3905–3908; (f) M. E. Hayes, H. Shinokubo and R. L. Danheiser, Org. Lett., 2005, 7, 3917–3920; (g) W. F. Austin, Y. Zhang and R. L. Danheiser, Tetrahedron, 2008, 64, 915–925; (h) X. Y. Mak, A. L. Crombie and R. L. Danheiser, J. Org. Chem., 2011, 76, 1852–1873.
7. V. P. Ananikov, J. Phys. Org. Chem., 2001, 14, 109–121.
8. V. P. Ananikov, J. Phys. Org. Chem., 2003, 16, 253–263.
9. N. Graulich, H. Hopf and P. R. Schreiner, Chem.–Eur. J., 2011, 17, 30–40.
10. The difference in the forming bonds lengths C1–C9 and C4–C8 (△) in the corresponding transition states was used as a measure of the degree of asynchronicity.
11. I. V. Alabugin, M. Manoharan, B. Breiner and F. D. Lewis, J. Am. Chem. Soc., 2003, 125, 9329–9342.
12. The present and previous studies^7,8 indicated that H-shifts may occur easily and need not to be recalculated. Note, that intermolecular H-transfer may also take place. High thermodynamic stability of the products facilitates H-shifts towards the formation of final structures.
13. IUPAC names: (1E)-N-but-3-yn-1-ylbut-1-en-3-yn-1-amine (15), indoline (18), (2E)-pent-2-en-4-yn-1-yl(prop-2-yn-1-yl)amine (19), isoindoline (22), (3E)-N-ethynylhex-3-en-5-yn-1-amine (23), (2E)-oct-2-en-7-ynenitrile (26), 6,7-dihydro-5H-cyclopenta[c]pyridine (29), (5E)-oct-5-en-7-ynenitrile (30), 6,7-dihydro-5H-cyclopenta[b]pyridine (33), ethynyl[(1E)-hex-5-yn-1-ylidene]amine (34), pent-4-yn-1-yl[(1E)-prop-2-yn-1-ylidene]amine (37), 3,5-dihydroindolizine (41), and 1,5-dihydroindolizine (42).
14. The ΔH and ΔG energy surface constructed from the CCSD(T)/6-311+G(d)//B3LYP/6-311+G(d) single point energies and B3LYP/6-311+G(d) frequencies calculations are given in Table S8 in the Supporting Information†.
15. R. F. W. Bader, Atoms in Molecules - A Quantum Theory, Oxford University Press, Oxford, 1990.
16. See Supporting Information for full details on QTAIM calculations†.
17. C. H. Schiesser, U. Wille, H. Matsubara and I. Ryu, Acc. Chem. Res., 2007, 40, 303–313.
18. I. V. Alabugin, B. Breiner, M. Manoharan, Electronic Effects in Cycloaromatization Reactions: The Melting Pot of Theory and Experiment in J. Richard (Ed.), Advances in Physical Organic Chemistry, 2007, 42, 1–35. Elsevier.
19. (a) Y. Kawashita and M. Hayashi, Molecules, 2009, 14, 3073–3093; (b) T. Hara, K. Mori, T. Mizugaki, K. Ebitani and K. Kaneda, Tetrahedron Lett., 2003, 44, 6207–6210; (c) S. Bayindir, E. Erdogan, H. Kilic and N. Saracoglu, Synlett, 2010, 1455–1458; (d) R. P. Kreher and J. Kalischko, Chem. Ber., 1991, 124, 645–654; (e) H. G. Tsai, M. Chung, Y. Chou and D. Hou, J. Phys. Chem. A, 2008, 112, 5278–5285.
20. IUPAC names: ((2E)-pent-2-en-4-yn-1-yl(prop-2-yn-1-yl)phosphine (43), 2,3-dihydro-1H-isophosphindole (46), (3E)-5-(prop-2-yn-1-yloxy)pent-3-en-1-yne (47), 1,3-dihydro-2-benzofuran (50), (3E)-5-(prop-2-yn-1-ylthio)pent-3-en-1-yne (51), 1,3-dihydro-2-benzothiophene (54), (3E)-5-(prop-2-yn-1-ylseleno)pent-3-en-1-yne (55), 1,3-dihydro-2-benzoselenophene (58).
21. (a) A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72, 5639–5648.
22. (a) A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–3100; (b) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–789; (c) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
23. (a) H. P. Hratchian and H. B. Schlegel, J. Chem. Phys., 2004, 120, 9918–9924; (b) H. P. Hratchian and H. B. Schlegel, J. Chem. Theory Comput., 2005, 1, 61–69; (c) C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1989, 90, 2154–2161; (d) C. Gonzalez and H. B. Schlegel, J. Phys. Chem., 1990, 94, 5523–5527.
24. (a) M. Head-Gordon, J. A. Pople and M. J. Frisch, Chem. Phys. Lett., 1988, 153, 503–506; (b) M. J. Frisch, M. Head-Gordon and J. A. Pople, Chem. Phys. Lett., 1990, 166, 275–280; (c) M. Head-Gordon and T. Head-Gordon, Chem. Phys. Lett., 1994, 220, 122–128.
25. (a) J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson, D. J. Singh and C. Fiolhais, Phys. Rev. B: Condens. Matter, 1992, 46, 6671–6687; (b) J. P. Perdew, K. Burke and Y. Wang, Phys. Rev. B: Condens. Matter, 1996, 54, 16533–16539.
26. (a) Y. Zhao, N. E. Schultz and D. G. Truhlar, J. Chem. Theory Comput., 2006, 2, 364–382; (b) Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157–167; (c) Y. Zhao and D. G. Truhlar, J. Chem. Theory Comput., 2008, 4, 1849–1868.
27. (a) J. A. Pople, M. Head-Gordon and K. Raghavachari, J. Chem. Phys., 1987, 87, 5968–5975; (b) R. J. Bartlett and G. D. Purvis, Int. J. Quantum Chem., 1978, 14, 561–581; (c) R. J. Bartlett and G. D. Purvis, Phys. Scr., 1980, 21, 255–265; (d) K. Raghavachari, G. W. Trucks, J. A. Pople and M. Head-Gordon, Chem. Phys. Lett., 1989, 157, 479–483.
28. (a) R. Krishnan and J. A. Pople, Int. J. Quantum Chem., 1978, 14, 91–100; (b) R. Krishnan, M. J. Frisch and J. A. Pople, J. Chem. Phys., 1980, 72, 4244–4245.
29. (a) Y. Zhao and D. G. Truhlar, J. Chem. Theory Comput., 2011, 7, 669–676; (b) E. R. Johnson, P. Mori-Sanchez, A. J. Cohen and W. Yang, J. Chem. Phys., 2008, 129, 204112; (c) Y. Zhao and D. G. Truhlar, J. Phys. Chem. A, 2006, 110, 10478–10486.
30. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, GAUSSIAN 03 (Revision C.02), Gaussian, Inc., Wallingford, CT, 2009.
31. Since geometry optimization at CCSD(T) level was not possible, B3LYP optimized geometry was utilized. Thus, the analysis should be used for estimation purposes only and it involves comparison between CCSD(T)//B3LYP values and corresponding theory level.

---

Footnote

† Electronic supplementary information (ESI) available: Tables S1–S7, Figures S1–S7, and Schemes S1–S2. See DOI: 10.1039/c1sc00380a

---