Highly Conjugated π‐Systems Arising from Cannibalistic Hexadehydro‐Diels–Alder Couplings: Cleavage of C−C Single and Triple Bonds

Abstract We have investigated the cannibalistic self‐trapping reaction of an ortho‐benzyne derivative generated from 1,11‐bis(p‐tolyl)undeca‐1,3,8,10‐tetrayne in an HDDA reaction. Without adding any specific trapping agent, the highly reactive benzyne is trapped by another bisdiyne molecule in at least three different modes. We have isolated and characterized the resulting products and performed high‐level calculations concerning the reaction mechanism. During the cannibalistic self‐trapping process, either a C≡C triple bond or an sp–sp3 C−C single bond is cleaved. Up to seven rings and nine C−C bonds are formed starting from two 1,11‐bis(p‐tolyl)undeca‐1,3,8,10‐tetrayne molecules. Our experiments and calculations provide considerable insight into the variety of reaction pathways which the ortho‐benzyne derivative, generated from a bisdiyne, can take when reacting with another bisdiyne molecule.

Lastly, the COSY-45 spectrum shows strong coupling of the bridgehead H atom in position 1 (4.71 ppm) with the H atom in position 6 (6.93-6.87 ppm) as well as strong coupling between the bridgehead H atom in position 4 (4.84 ppm) and the H atom in position 5 (6.85-6.79   (4.91 ppm) shows a NOESY cross-peak to the CH2 group of the five-membered ring in position 7 (3.11-2.99 ppm) (Fig. S5). The H atom in position 4 resonates at 4.60 ppm, showing a NOESY cross-peak to the CH3 group, resulting from the trapped toluene molecule, in position 3 (1.86 ppm) and the aromatic H atoms in position 8 (7.38-7.28 ppm) (Fig. S6). Furthermore, a NOESY cross-peak between the CH3 group in position 3 (1.86 ppm) and the H atom in position 2 (6.38 ppm) was observed (Fig.   S7). Continuing from the H atom in position 2 (6.38 ppm), a through-space coupling to the bridgehead H in position 1 (4.91 ppm) was detected (Fig. S8). With the help of the COSY-45 spectrum, the coupling of the H atom in position 2 (6.38 ppm) to the two bridgehead H atoms 1 (4.91 ppm) and 4 (4.60 ppm) was confirmed (Fig. S9).
Lastly, the COSY-45 spectrum shows strong coupling of the bridgehead H atom in position 1 (4.91 ppm) with the H atom in position 6 (6.93-6.87 ppm) as well as coupling between the bridgehead H atom in position 4 (4.60 ppm) and the H atom in position 5 (6.85-6.79   Isomer III has one bridgehead H atom in position 4, which resonates at 4.87 ppm, showing a NOESY cross-peak to the aromatic H atoms in position 8 (7.38-7.28 ppm) ( Fig. S6). Furthermore, a NOESY cross-peak between the CH3 group in position 1 (2.20-2.10 ppm) and the H atoms in position 2 and 6 (6.58 ppm) was observed (Fig.   S7). The cross-peak expected for the coupling of the CH3 group in position 1 (2.20-2.10 ppm) and the CH2 group in position 7 (3.11-2.99 ppm) overlaps with the cross-peaks of the CH2 groups in positions 7 and 9 (2.20-2.10 ppm). With the help of the COSY-45 spectrum, the coupling of the H atoms in positions 3 and 5 (6.85-6.79 ppm) to the bridgehead H atom in position 4 (4.87 ppm) was confirmed ( Fig. S9). Lastly, the COSY-45 spectrum shows coupling of the bridgehead H atom in position 4 (4.87 ppm) with the H atoms in position 2 and 6 (6.58 ppm) (Fig. S9).

Photophysical measurements
General photophysical measurements. All measurements were performed in standard quartz cuvettes (1 cm x 1 cm cross-section). UV-visible absorption spectra were recorded using an Agilent 8453 diode array UV-visible spectrophotometer. The emission spectra were recorded using an Edinburgh Instruments FLSP920 spectrometer equipped with a double monochromator for both excitation and emission, operating in right-angle geometry mode, and all spectra were fully corrected for the spectral response of the instrument. All solutions used in photophysical measurements had concentrations lower than 5•10 -6 M to minimize inner filter effects during fluorescence measurements.
Fluorescence quantum yield measurements. The fluorescence quantum yields were measured using a calibrated integrating sphere (inner diameter: 150 mm) from Edinburgh Instruments combined with the FLSP920 spectrometer described above.
For solution-state measurements, the longest-wavelength absorption maximum of the compound in the respective solvent was chosen as the excitation wavelength, unless stated otherwise.
Lifetime measurements. Fluorescence lifetimes were recorded using the time-correlated single-photon counting (TCSPC) method using an Edinburgh Instruments FLS980 spectrometer equipped with a high speed photomultiplier tube positioned after a single emission monochromator. Measurements were made in right-angle geometry mode, and the emission was collected through a polarizer set to the magic angle. Solutions were excited with a pulsed diode laser at a wavelength of 316 nm (for 4, 6,8,9) and 472 nm (for 11). The full-width-at-half-maximum (FWHM) of the pulse from the diode laser was ca. 75-90 ps with an instrument response function (IRF) of ca. 230 ps FWHM. The IRFs were measured from the scatter of an aqueous suspension of Ludox at the excitation wavelength. Decays were recorded to pass of the emission monochromator and a variable neutral density filter on the excitation side were adjusted to give a signal count rate of <60 kHz. Iterative reconvolution of the IRF with one decay function and non-linear least-squares analysis were used to analyze the data. The quality of all decay fits was judged to be satisfactory, based on the calculated values of the reduced χ 2 and Durbin-Watson parameters and visual inspection of the weighted residuals.

UV/Vis absorption spectrum of 3 in THF
Quantum yield and lifetime were not measured for compound 5 because of the mixture of four isomers (see text). Absorption between 350 and 450 nm is due to trace quantities of 11. The excitation spectrum (em = 352 nm) shows no band between 350 and 450 nm, because compound 11 does not emit at 352 nm. Cryostream 700 or 800). Diffraction data were collected on Bruker Apex II 4-circle diffractometers with CCD area detectors using Mo-Kα radiation monochromated by graphite (for 3, 5I, 6, and 8) or multi-layer focusing mirrors (for 4 and 11) at 100 K.
Only the major part (88%) of the disordered (CH2)3 moiety is shown here.    Only one of the four symmetry-independent molecules is shown here.

Computational details 1
The geometries of 4, 5, 6, 8, 9 and 11 were optimized without symmetry constraints using the Gaussian 09 and Gaussian 03 program packages. [7] The optimized structures were verified as minima on the potential energy surface by calculation of the vibrational frequencies. The DFT computations were performed using the B3LYP functional and the 6-31+g(d) basis sets. [8] Excited state properties were obtained from time-dependent DFT calculations at the CAM-B3LYP/6-31G(d,p) level. [9] The simulated gas-phase and CH2Cl2 solution (PCM model) absorption spectra are shown in the following Figures using Gaussian functions with halfband-widths of 1000 cm -1 .

Lowest energy singlet electronic transitions of 4 in the gas phase
computations. [15] Figure S19. Detailed reaction mechanism for Steps 3 and 4 ( Figure 10 main text), modelled by the reaction of benzocyclobutadiene with butadiyne. Figure S20. Top: Computed energies for the reaction mechanism given in Figure S19 at the CCSD(T)/aug-cc-pVDZ level of theory. For the labelling of the various stationary points, see Figure   S19. Bottom: Computed energies for the reaction mechanism given in Figure S19 at the    Figure S19. All energies are given with respect to the reactants benzocyclobutadiene and butadiyne. The geometries were determined employing UB3LYP-D3/ 6-311++G(d,p). For the labelling of the various stationary points, see Figure S19. In addition, we obtained single-point energies using M06-2X-D3, CCSD(T), and CAS-OVB-MP2 (8,8) in combination with the aug-cc-pVDZ basis set.
M06-2X-D3 CCSD(T) D1-amplitude CAS-OVB-MP2 (8,8)  The first step of the overall reaction represents a formal intramolecular [4+2]-cycloaddition within a 1,11-bis(p-tolyl)undeca-1,3,8,10-tetrayne 3. In this reaction, as we have described in  [14] employing the CCSD(T)/6-311+G(d,p)//B3LYP/6-311+G(d,p) +ZPVE level of theory. The data are summarized in Figure S17. The calculations predicted that the biradical intermediate The total reaction energy of Step 3 and Step 4 amounts to -124.7 kcal/mol, i.e. the overall reaction is strongly exothermic. Two key differences exist between the model system investigated by Jones and Krebs [15] and our present system. First, our system contains a benzocyclobutadiene unit instead of a cyclobutadiene. Going from cyclobutadiene to benzocyclobutadiene, one double bond of the cyclobutadiene subunit is integrated into the benzene ring. Consequently, an electronic structure as indicated in Ia/b' is highly unfavorable and, as such, the four-membered ring can no longer act as a diene, but only as a double bond.
According to our CCSD(T)/aug-cc-pVDZ calculations, the activation barrier increases from ca. 20 kcal/mol, which was predicted for cyclobutadiene, [15] to more than 60 kcal/mol. A second difference in our system is that the cyclobutadiene subunit reacts with a butadiyne instead of methylacetylene. To investigate effects resulting from this difference, we recomputed the reaction paths for our system. The details of the mechanisms are depicted in Figure S19, and the computed relative energies are summarized in Table S2 and S3. Figure   S20 gives followed by a ring closure to int3 via TS3. M06-2X-D3-calculations predict that the six-membered ring formation of the benzyne derivative int3' is kinetically favored because the barrier connected with TS3' (29 kcal/mol) is slightly lower than the combined barriers for TS2 (ca. 24 kcal/mol) and TS3 (ca. 32 kcal/mol) leading from int1 to the four-membered ring.
It is important to note that M06-2X-D3 predicts that both possible second steps (TS2/TS3 and TS3') possess a higher barrier than the formation of int1. MP2, SCS-MP2, CC2, SCS-CC2 and CCSD calculations lead to a similar conclusion (Table S2). In contrast, CCSD(T) and CAS-OVB-MP2 (8,8) calculations indicate that the pathway leading to the Dewar benzene derivative int3 via int2 is the energetically favored one. CCSD(T) predicts a barrier of only ca. 2 kcal/mol for the conversion of int1 to int2 which is ca. -2 kcal/mol lower in energy than int1. Additionally, the barrier TS3 is lower in energy than TS3'. CCSD(T) predicts that TS3 is ca. 13 kcal/mol more stable than int2, so that the reaction can directly form the four-membered ring from int2 without any further barrier. It is important to note that CCSD(T) computes the barriers for the second step (TS2/TS3 and TS3') to be considerably lower than the barrier for the first step (TS1). The significant difference between CCSD(T) and the other single-reference approaches indicates that the electronic structures of the intermediates are very difficult to compute. This is also indicated by the D1-amplitudes [16] ( Table S3). They are a measure for the multi-reference character of a wave function. For values larger than 0.05, experience shows that multi-reference effects become so strong that all single-reference methods should be handled with care. To obtain more reliable insight, we employed the CAS-OVB-MP2 approach. The CAS-OVB-MP2 (8,8) results predict no barrier between int1 and int2, and place int2 slightly lower in energy (-1.1 kcal/mol) than int1.
The barrier to form the six-membered ring is slightly higher (TS3' 4 kcal/mol); i.e., CAS-OVB-MP2 (8,8) also favors the formation of the four-membered ring, which is in line with our experimental findings.
According to our CAS-OVB-MP2 (8,8) computations, the conversion of int3 to the naphthalene product via TS4 should possess an activation barrier of ca. 24 kcal/mol. The final product is predicted to lie -92.9 kcal/mol below TS4. The overall reaction energy is computed to be -131 kcal/mol. The barrier is, therefore, ca. 7 kcal/mol below the results of the ring opening in the Dewar benzene derivative Jones_int3 investigated by Jones and Krebs. [15] Considering the various very exothermic steps along the overall reaction path (Figure 10 main text) up to IIa/b, the system should contain sufficient energy to surmount this barrier quite quickly. This underpins the validity of the proposed mechanism up to IIIa / 8. The computed data supports the formation of a Dewar benzene derivative, but the differences in the barrier heights are so small that the formation of the benzyne derivative (int3') as a reactive side product cannot be excluded. It may be responsible for some of the additional as of yet unidentified side products.  Figure 11 and Figure 12 in the main text) and the biradical structure SB 3, i.e. in a first step,only the C 2 -C 3 bond is formed (see Figure 12 main text). In comparision the concerted reaction path via SB 4 has to overcome a barrier of 6.9 kcal/mol (25.9 kcal/mol CAS-OVB-MP2 (8,8)/aug-cc-pVDZ). A full optimization starting from SB 3 leads to structure SB 3' which, using CAS-OVB-MP2 (8,8), is lower in energy than the structure SB 3 (see Table S4). Structure SB 3' represents a complete biradical as indicated in Figure 11 (main text). Using the most accurate CAS-OVB-MP2 (8,8)  (structure SB 5) has to be overcome to get to the next minimum SB 9, which is about 25 kcal/mol lower in energy than the starting structure SB 1. In structure SB 9, the hydrogen is already attached to the C 1 center, but the C 3 -C 4 bond is not yet broken, i.e. in the UB3LYP/6-311++G(d,p) calculations, the C 3 -C 4 bond does not break spontaneously if the hydrogen is shifted to C 1 . To estimate the barrier between structure SB 9 and the final product SB 10, i.e. to estimate the height of barrier which has to be overcome to break the C 3 -C 4 bond, we scanned the potential energy curve starting from structure SB 9 and stretched the C 3 -C 4 bond within the UB3LYP/6-311++G(d,p) level of theory. Structure SB 9'', which represents the top of the barrier on the reaction path from SB 9 to the final product SB 10, is less than 1 kcal/mol higher in energy than SB 9 according to CAS-OVB-MP2 (8,8) single-point calculations, i.e. the final C 3 -C 4 bond cleavage has no further real barrier. Similar scans were also started from the structures SB 6 -SB 8. Starting from SB 6, the barrier is also less than 0.5 kcal/mol (Table S4, structure SB 6'': -9.7 kcal/mol). The scan starting from structure SB 7 already indicated no further barrier, i.e. from structure SB 7, the system can relax barrier-free to structure SB 9, but also directly to SB 10. In summary, our CAS-OVB-MP2 (8,8) calculations support our proposal that the final product SB 10 can be formed via the mechanisms depicted in Figure 11 (main text