Ortho effect in the Bergman cyclization: Comparison of experimental approaches and dissection of cycloaromatization kinetics

Article abstract of DOI:10.1021/jo0520801

Four different experimental sources of kinetic information were combined to study the effect of ortho substituents on the rate of Bergman cycloaromatization. All methods confirm that the cyclization barrier is highly sensitive to the nature of the ortho substituents. However, the measured activation energies strongly depend on the choice of experimental technique: even the relative trends provided by the different methods agree with each other only in the case of acceptor substituents. Both the onset peaks and the activation energies determined by differential scanning calorimetry (DSC; either in neat enediynes or in their solutions in 10.6 M 1,4-cyclohexadiene (1,4-CHD)) strongly overestimate the reactivity of 1,2-diethynylbenzene, suggesting that DSC cannot be taken as a reliable indicator of enediyne reactivity. This discrepancy is likely to stem from the presence of side reactions with low activation barriers, especially important when the reaction is conducted in neat enediyne. On the other hand, kinetic measurements based on monitoring the concentrations of enediyne reactants and naphthalene products provide reliable general trends that include the parent benzannelated enediyne. These measurements confirm that both ortho-NO₂ and ortho-CHO substituents substantially decrease activation energies for the Bergman cyclization, supporting earlier computational predictions. A comparison of theory and experiment suggests that computations at the Moeller-Plesset second-order perturbation theory (MP2)/6-31G** level provide an excellent alternative to DFT when an accurate description of the contribution of noncovalent interactions to the activation energy is needed. Activation energies derived from k_eff, the effective rate constant under the pseudo-first-order approximation, depend on the 1,4-CHD concentrations. The true rate constant, k_-1, for the cyclization step and the ratio of constants for the retro-Bergman ring opening, k_-1, and the intermolecular H-atom abstraction, k₂, were determined from the dependence of cycloaromatization kinetics of ortho- and para-NO₂ substituted enediynes on the concentration of 1,4-CHD.

Full text of DOI:10.1021/jo0520801

Ortho Effect in the Bergman Cyclization: Comparison of
Experimental Approaches and Dissection of Cycloaromatization
Kinetics
Tarek A. Zeidan, Serguei V. Kovalenko, Mariappan Manoharan, and Igor V. Alabugin*
Department of Chemistry and Biochemistry, Florida State UniVersity, Tallahassee, Florida 32306-4390
alabugin@chem.fsu.edu
ReceiVed October 4, 2005
Four different experimental sources of kinetic information were combined to study the effect of ortho
substituents on the rate of Bergman cycloaromatization. All methods confirm that the cyclization barrier
is highly sensitive to the nature of the ortho substituents. However, the measured activation energies
strongly depend on the choice of experimental technique: even the relative trends provided by the different
methods agree with each other only in the case of acceptor substituents. Both the onset peaks and the
activation energies determined by differential scanning calorimetry (DSC; either in neat enediynes or in
their solutions in 10.6 M 1,4-cyclohexadiene (1,4-CHD)) strongly overestimate the reactivity of 1,2-
diethynylbenzene, suggesting that DSC cannot be taken as a reliable indicator of enediyne reactivity.
This discrepancy is likely to stem from the presence of side reactions with low activation barriers, especially
important when the reaction is conducted in neat enediyne. On the other hand, kinetic measurements
based on monitoring the concentrations of enediyne reactants and naphthalene products provide reliable
general trends that include the parent benzannelated enediyne. These measurements confirm that both
ortho-NO2 and ortho-CHO substituents substantially decrease activation energies for the Bergman
cyclization, supporting earlier computational predictions. A comparison of theory and experiment suggests
that computations at the Moeller-Plesset second-order perturbation theory (MP2)/6-31G** level provide
an excellent alternative to DFT when an accurate description of the contribution of noncovalent interactions
to the activation energy is needed. Activation energies derived from keff, the effective rate constant under
the pseudo-first-order approximation, depend on the 1,4-CHD concentrations. The true rate constant, k1,
for the cyclization step and the ratio of constants for the retro-Bergman ring opening, k-1, and the
intermolecular H-atom abstraction, k2, were determined from the dependence of cycloaromatization kinetics
of ortho- and para-NO2 substituted enediynes on the concentration of 1,4-CHD.
Introduction
DNA cleavage and self-programmed cell death, or apoptosis.
Bergman cyclization was also applied to the development of
polymeric materials with valuable thermal properties7 and to
the synthesis of polycyclic compounds.8 Furthermore, this
reaction continues to play an important role in calibrating and
developing practical theoretical methods for studies of chemical
reactivity.9-11
The Bergman cycloaromatization1 transforms enediynes into
highly reactive 1,4-dehydrobenzene (or para-benzyne) di-
radicals2-4 capable of abstracting two hydrogen atoms from
the sugar backbone of two strands of DNA. In the case of
natural enediyne antibiotics,5,6 this process leads to double strand
10.1021/jo0520801 CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/04/2006
962
J. Org. Chem. 2006, 71, 962-975

Ortho Effect in the Bergman Cyclization
In the general sense, control over enediyne reactivity can be
achieved through either strain12-15 or electronic effects.16 The
seminal hypothesis rationalizing the influence of electronic
factors was advanced by Koga and Morokuma17 who attributed
the high activation energy18 of the Bergman cyclization to a
strong electron repulsion between the in-plane occupied acety-
lene π orbitals. The dominating role of the in-plane π orbitals
in the cyclization kinetics was particularly well-illustrated in a
valence bond study of Galbraith and co-workers.19 These
findings received further support in a recent natural bond orbital
(NBO) study by our group20 that also found that the role of
repulsion between the in-plane filled orbitals is accentuated by
a parallel decrease in the attractive two-electron interaction
between the π orbital and the π* orbital. At the so-called
“Nicolaou’s threshold” (ca. 3.2 Å),10 the attractive interaction
vanishes because the in-plane π orbitals become parallel,
mimicking the symmetry forbidden transition state (TS) for
thermal [2 + 2] cycloaddition.20 The above studies imply that
electron-withdrawing substituents should promote the cyclization
only when they are capable of an interaction with the in-plane
π system.21,22 As a result, the position of the acceptor substituent
becomes important. For example, Schreiner et al. demonstrated
that the cyclization is accelerated by σ acceptors at the terminal
acetylene atoms,23 whereas Zaleski et al. found that such effects
can be accentuated by coordination with Lewis acids in the
vicinity of alkyne termini.24 In contrast, Jones and Plourde
showed that the progressive substitution of hydrogen atoms by
halogens at the vinyl position decreases the reaction rate,25 while
Chen et al. found that protonation of the nitrogen atom in 3-aza-
enediynes increases the cyclization barrier.26 Because developing
radical centers do not communicate with the orthogonal π
system, neither the remote substitution of the central double
bond by a benzene ring (benzannelation) nor the presence of a
(1) Jones, R. R.; Bergman, R. G. J. Am. Chem. Soc. 1972, 94, 660.
Bergman, R. G. Acc. Chem. Res. 1973, 6, 25. Other cycloaromatization
reactions: (a) Nagata, R.; Yamanaka, H.; Okazaki, E.; Saito, I. Tetrahedron
Lett. 1989, 30, 4995. (b) Myers, A. G.; Dragovich, P. S.; Kuo, E. Y. J. Am.
Chem. Soc. 1992, 114, 9369. (c) Nakatani, K.; Isoe, S.; Maekawa, S.; Saito,
I. Tetrahedron Lett. 1994, 35, 605. (d) Sullivan, R. W.; Coghlan, V. M.;
Munk, S. A.; Reed, M. W.; Moore, H. W. J. Org. Chem. 1994, 59, 2276.
(e) Toda, F.; Tanaka, K.; Sano, I.; Isozaki, T. Angew. Chem., Int. Ed. Engl.
1994, 33, 1757. (f) Nicolaou, K. C.; Skokotas, G.; Maligres, P.; Zuccarello,
G.; Schweiger, E. J.; Toshima, K.; Wendeborn, S. Angew. Chem., Int. Ed.
Engl. 1989, 28, 1272. (g) Engels, B.; Lennartz, C.; Hanrath, M.; Schmittel,
M.; Strittmatter, M. Angew. Chem., Int. Ed. 1998, 37, 1960. (h) Schmittel,
M.; Steffen, J. P.; Angel, M. A. W.; Engels, B.; Lennartz, C.; Hanrath, M.
Angew. Chem., Int. Ed. 1998, 37, 1562. (i) Kawatkar, S. P.; Schreiner, P.
R. Org. Lett. 2002, 4, 3643.
(2) (a) Nicolaou, K. C.; Smith, A. L. Acc. Chem. Res. 1992, 25, 497. (b)
Maier, M. E.; Bosse, F.; Niestroj, A. J. Eur. J. Org. Chem. 1936, 1, 1. (c)
Grissom, J. W.; Gunawardena, G. U.; Klingberg, D.; Huang, D. H.
Tetrahedron 1996, 52, 6453. (d) Fallis, A. G. Can. J. Chem. 1999, 77,
159. (e) Caddick, S.; Delisser, V. M.; Doyle, V. E.; Khan, S.; Avent, A.
G.; Vile, S. Tetrahedron 1999, 55, 2737. (f) Wang, K. K. Chem. ReV. 1996,
96, 207.
(11) Cramer, C. J.; Nash, J. J.; Squires, R. R. Chem. Phys. Lett. 1997,
277, 311. Cramer, C. J.; Squires, R. R. J. Phys. Chem. A 1997, 101, 9191.
Wierschke, S. G.; Nash, J. J.; Squires, R. R. J. Am. Chem. Soc. 1993, 115,
11958. Johnson, W. T. G.; Sullivan, M. B.; Cramer, C. J. Int. J. Quantum
Chem. 2001, 85, 492. Johnson, W. T. G.; Cramer, C. J. J. Phys. Org. Chem.
2001, 14, 597. Cramer, C. J.; Thompson, J. J. Phys. Chem. A 2001, 105,
2091. Kraka, E.; Cremer, D. J. Comput. Chem. 2001, 22, 216.
(12) Incorporation of the enediyne unit into a strained 9- or 10-membered
ring was shown to facilitate the reaction through ground-state destabilization.
For the experimental work, see: (a) Nicolaou, K. C.; Zuccarello, G.; Ogawa,
Y.; Schweiger, E. J.; Kumazawa, T. J. Am. Chem. Soc. 1988, 110, 4866.
(b) Iida, K.; Hirama, M. J. Am. Chem. Soc. 1995, 117, 8875.
(13) For a theoretical analysis of the strain effects on the cyclization,
see: (a) Snyder, J. P. J. Am. Chem. Soc. 1989, 111, 7630. (b) Snyder, J. P.;
Tipsword, G. E. J. Am. Chem. Soc. 1990, 112, 4040. (c) Snyder, J. P. J.
Am. Chem. Soc. 1990, 112, 5367.
(14) Kraka, E.; Cremer, D. J. Am. Chem. Soc. 1994, 116, 4929.
(15) The cyclization can also be controlled by the release of strain in
the TS: (a) Magnus, P.; Carter, P.; Elliott, J.; Lewis, R.; Harling, J.; Pitterna,
T.; Bauta, W. E.; Fortt, S. J. Am. Chem. Soc. 1992, 114, 2544.
(16) For the most recent review, see: Rawat, D. S.; Zaleski, J. M. Synlett
2004, 293.
(17) Koga, N.; Morokuma, K. J. Am. Chem. Soc. 1991, 113, 1907.
(18) (a) Roth, W. R.; Hopf, H.; Horn, C. Chem. Ber. 1994, 127, 1765.
(b) Wenthold, P. G.; Squires, R. R. J. Am. Chem. Soc. 1994, 116, 6401.
(19) Galbraith, J. M.; Schreiner, P. R.; Harris, N.; Wei, W.; Wittkopp,
A.; Shaik, S. Chem.sEur. J. 2000, 6, 1446.
(3) Nicolaou, K. C.; Zuccarello, G.; Riemer, C.; Estevez, V. A.; Dai,
W. M. J. Am. Chem. Soc. 1992, 114, 7360.
(4) Myers, A. G.; Dragovich, P. S. J. Am. Chem. Soc. 1992, 114, 5859.
(5) (a) Enediyne Antibiotics as Antitumor Agents; Borders, D. B., Doyle,
T. W., Eds.; Marcel Dekker: New York, 1995. (b) Neocarzinostatin: The
Past, Present, and Future of an Anticancer Drug; Maeda, H., Edo, K.,
Ishida, N., Eds.; Springer: New York, 1997.
(6) (a) Smith, A. L.; Nicolaou, K. C. J. Med. Chem. 1996, 39, 2103. (b)
Paloma, L. G.; Smith, J. A.; Chazin, W. J.; Nicolaou, K. C. J. Am. Chem.
Soc. 1994, 116, 3697. (c) Kappen, L. S.; Goldberg, I. H. Biochemistry 1983,
22, 4872. (d) Zein, N.; McGahren, W. J.; Morton, G. O.; Ashcroft, J.;
Ellestead, G. A. J. Am. Chem. Soc. 1989, 111, 6888. (e) De Voss, J. J.;
Townsend, C. A.; Ding, W. D.; Morton, G. O.; Ellestad, G. A.; Zein, N.;
Tabor, A. B.; Schreiber, S. L. J. Am. Chem. Soc. 1990, 112, 9669. (f)
Hangeland, J. J.; Voss, J. J.; Heath, J. A.; Townsend, C. A.; Ding, W. D.;
Ashcroft, J. S.; Ellestad, G. A. J. Am. Chem. Soc. 1992, 114, 9200. (g)
Dedon, P. C.; Salzberg, A. A.; Xu, J. H. Biochemistry 1993, 32, 3617. (h)
Christner, D. F.; Frank, B. L.; Kozarich, J. W.; Stubbe, J.; Golik, J.; Doyle,
T. W.; Rosenberg, I. E.; Krishnan, B. J. Am. Chem. Soc. 1992, 114, 8763.
(i) Sugiura, Y.; Shiraki, T.; Konishi, M.; Oki, T. Proc. Natl. Acad. Sci.
U.S.A. 1990, 87, 3831. (j) Shiraki, T.; Uesugi, M.; Sugiura, Y. Biochem.
Biophys. Res. Commun. 1992, 188, 584. (k) Shiraki, T.; Sugiura, Y.
Biochemistry 1990, 29, 9795. (l) Sugiura, Y.; Arakawa, T.; Uesugi, M.;
Shiraki, T.; Ohkuma, H.; Konishi, M. Biochemistry 1991, 30, 2989. (m)
Matsumoto, T.; Okuno, Y.; Sugiura, Y. Biochem. Biophys. Res. Commun.
1993, 195, 659. (n) Xu, Y.; Zhen, Y. S.; Goldberg, I. H. Biochemistry 1994,
33, 5947.
(7) Chen, X.; Tolbert, L. M.; Hess, D. W.; Henderson, C. Macromolecules
2001, 34, 4104. Shah, H. V.; Babb, D. A.; Smith, D. W., Jr. Polymer 2000,
41, 4415. John, J. A.; Tour, J. M. J. Am. Chem. Soc. 1994, 116, 5011.
(8) Bowles, D. M.; Palmer, G. J.; Landis, C. A.; Scott, J. L.; Anthony,
J. E. Tetrahedron 2001, 57, 3753. Bowles, D. M.; Anthony, J. E. Org. Lett.
2000, 2, 85.
(9) (a) Scheiner, A. C.; Schaefer, H. F., III; Liu, B. J. Am. Chem. Soc.
1989, 111, 3118. (b) Nicolaides, A.; Borden, W. T. J. Am. Chem. Soc. 1993,
115, 11951. (c) Lindh, R.; Persson, B. J. J. Am. Chem. Soc. 1994, 116,
4963. (d) Lindh, R.; Lee, T. J.; Bernhardsson, A.; Persson, B. J.; Karlstro¨m,
G. J. Am. Chem. Soc. 1995, 117, 7186. (e) Kraka, E.; Cremer, D.; Bucher,
G.; Wandel, H.; Sander, W. Chem. Phys. Lett. 1997, 268, 313. (f) Lindh,
R.; Ryde, U.; Schu¨tz, M. Theor. Chem. Acc. 1997, 97, 203. (g) Chen, W.
C.; Chang, N. W.; Yu, C. H. J. Phys. Chem. A 1998, 102, 2584. (h)
McMahon, R. J.; Halter, R. J.; Fimmen, R. L.; Wilson, R. J.; Peebles, S.
A.; Kuczkowski, R. L.; Stanton, J. F. J. Am. Chem. Soc. 2000, 122, 939.
(10) Schreiner, P. R. J. Am. Chem. Soc. 1998, 120, 4184. Schreiner, P.
R. J. Chem. Soc., Chem. Commun. 1998, 483.
(20) Alabugin, I. V.; Manoharan, M. J. Phys. Chem. A 2003, 107, 3363.
(21) Schmittel, M.; Kiau, S. Chem. Lett. 1995, 953.
(22) Maier, M. E.; Greiner, B. Liebigs Ann. Chem. 1992, 855.
(23) Jones, G. B.; Warner, P. M. J. Am. Chem. Soc. 2001, 123, 2134.
Konig, B.; Pitsch, W.; Klein, M.; Vasold, R.; Prall, M.; Schreiner, P. R. J.
Org. Chem. 2001, 66, 1742. Prall, M.; Wittkopp, A.; Fokin, A. A.; Schreiner,
P. R. J. Comput. Chem. 2001, 22, 1605. Chemistry of metalloenediynes:
Benites, P. J.; Rawat, D. W.; Zaleski, J. M. J. Am. Chem. Soc. 2000, 122,
7208. Rawat, D. S.; Zaleski, J. M. J. Am. Chem. Soc. 2001, 123, 9675.
Coalter, N. L.; Concolino, T. E.; Streib, W. E.; Hughes, C. G.; Rheingold,
A. L.; Zaleski, J. M. J. Am. Chem. Soc. 2000, 122, 3112. Chandra, T.;
Pink, M.; Zaleski, J. M. Inorg. Chem. 2001, 40, 5878. Rawat, D. S.; Zaleski,
J. M. Synth. Commun. 2002, 32, 1489. Rawat, D. S.; Benites, P. J.; Incarvito,
C. D.; Rheingold, A. L.; Zaleski, J. M. Inorg. Chem. 2001, 40, 1846.
Schmitt, E. W.; Huffman, J. C.; Zaleski, J. M. Chem. Commun. 2001, 167.
Rawat, D. S.; Zaleski, J. M. Chem. Commun. 2000, 2493. Benites, P. J.;
Lato, S. M.; Ellington, A. D.; Zaleski, J. M. J. Inorg. Biochem. 1999, 74,
75.
(24) Bhattacharyya, S.; Pink, M.; Baik, M. H.; Zaleski, J. M. Angew.
Chem., Int. Ed. 2005, 44, 592.
(25) Jones, G. B.; Plourde, G. W. Org. Lett. 2000, 2, 1757.
(26) Hoffner, J.; Schottelius, M. J.; Feichtinger, D.; Chen, P. J. Am. Chem.
Soc. 1998, 120, 376.
J. Org. Chem, Vol. 71, No. 3, 2006 963

Zeidan et al.
SCHEME 1. Bergman Cyclization of Ortho-Substituted
Benzannelated Enediynes
cyclization by destabilizing reactants through steric repulsion
between the ortho substituent and the in-plane acetylenic orbitals
(Figure 1). This interaction becomes less significant in the TS
when the acetylene moiety is bent away from the ortho
functional group. As a result, the TS and the product are
destabilized to a lesser extent than the starting material and the
activation barrier is decreased through the classic steric-
assistance mechanism.33 The steric nature of this effect is clear
from the comparison with the respective para isomers. For
example, the ortho-nitroenediyne is less stable than the para
isomer by 5.8 kcal/mol at the B3LYP/6-31G** level, as
illustrated in Figure 1. The difference in stability decreases to
2.8 kcal/mol in the TS and to 2.0 kcal/mol in the product,
accounting for the 3 kcal/mol decrease in the cyclization barrier
and the 3.8 kcal/mol decrease in the reaction endothermicity.
In contrast, such substituents as CH3, NH2, NH3+, and syn-
OH are predicted to stabilize respective ortho-enediynes relative
to their para isomers because of a hydrogen-bond formation
between the X-H moiety and the in-plane π bond.34 Interest-
ingly, both B3LYP computations and BLYP computations
predict that this stabilization will decrease in the TS for all of
these neutral substrates and increase for the positively charged
ammonium group, accounting for the barrier decrease in the
latter case.31 Although this difference in the dynamics of such
noncovalent interactions between the neutral and the charged
moieties can be rationalized through a larger electrostatic
contribution to H bonding of cationic species, such subtle
variations along with the known deficiencies of DFT in the
description of noncovalent interactions35 do suggest that such
DFT predictions should be taken with caution until they are
verified experimentally.
The aldehyde case is even more complicated. Depending on
the orientation of the carbonyl group, this enediyne may exist
as two conformers that have different barriers for the transfor-
mation. B3LYP, BLYP, and Moeller-Plesset second-order
perturbation theory (MP2) calculations with the 6-31G** basis
set uniformly predict the anti conformer of the starting material
to be 2.4-2.8 kcal/mol more stable than the syn isomer. Both
conformers are predicted to be more reactive than para-NO2
substituted enediyne by DFT: the syn conformer by about 3
kcal/mol and the anti conformer by about 2 kcal/mol. The
Curtin-Hammett analysis suggests that the cyclization should
mostly proceed through the more stable anti conformer, with
the overall decrease in the barrier on the order of 2 kcal/mol.
Interestingly, computational comparison with the respective para
isomer suggests that a decrease in the cyclization barrier in the
case of the anti isomer stems both from reactant destabilization
and from TS stabilization, whereas the Bergman cyclization of
the syn isomer is accelerated by the steric assistance (similar to
the effect of the NO2 moiety).
substituent at the aromatic ring strongly affects the cyclization
barrier in benzannelated enediynes.27
Nevertherless, the rates of cyclizations of benzannelated
enediynes were reported to be slower and more sensitive to
hydrogen-donor (HD) concentration than in the case of com-
parable acyclic enediynes.28 This effect was attributed to the
faster retro-Bergman ring opening accentuated by the decreased
reactivity of p-benzynes toward hydrogen abstraction.29,30 The
retro-Bergman ring opening of benzannelated enediynes is fast
because the loss of aromatic stabilization upon the transforma-
tion of naphthalene to benzene is less than in the process of the
opening of the benzene ring in the parent p-benzyne. Although
the importance of the out-of-plane interactions increases pro-
gressively after the TS, the low sensitivity of reaction kinetics
to substituent effects is one of the largest challenges to efficient
control and rational design of cycloaromatization reactions.
Russell and co-workers have shown that the influence of para
substituents on the rates of disappearance of benzannelated
enediynes is mainly transmitted through the field effect and,
thus, is relatively small.27 Recently, our group confirmed these
findings computationally and suggested that ortho substituents
that are spatially close to the enediyne moiety can control the
rate of Bergman cyclization more efficiently (Scheme 1).31
Depending on the substituent, their interaction with the enediyne
moiety and adjacent radical center in the p-benzyne can be either
attractive (H bonding) or repulsive (steric). The third mechanism
for the ortho effect is different and is based on the interception
of the p-benzyne diradical by intramolecular H abstraction,
which renders the cyclization step irreversible.32
In particular, the NO2, CF3, syn-CHO, and syn-OMe groups
were predicted to decrease the activation energy for the Bergman
(27) Choy, N.; Kim, C. S.; Ballestero, C.; Artigas, L., Diez, C.;
Lichtenberg, F.; Shapiro, J.; Russell, K. C. Tetrahedron Lett. 2000, 41, 6955.
(28) Semmelhack, M. F.; Neu, T.; Foubelo, F. J. Org. Chem. 1994, 59,
5038. Kaneko, T.; Takahashi, M.; Hirama, M. Tetrahedron Lett. 1999, 40,
2015. For a recent discussion, see: Semmelhack, M. F.; Sarpong, R. J.
Phys. Org. Chem. 2004, 17, 807.
(29) For the decrease in retro-Bergman barrier, see: Haberhauer, G.;
Gleiter, R. J. Am. Chem. Soc. 1999, 121, 4664. Koseki, S.; Fujimura, Y.;
Hirama, M. J. Phys. Chem. A 1999, 103, 7672. Stahl, F.; Moran, D.;
Schleyer, P. v. R.; Prall, M.; Schreiner, P. R. J. Org. Chem. 2002, 67, 1453.
(30) Diradicals are less reactive than the respective monoradicals because
of the loss of TB interaction between the two radical centers in the first
hydrogen abstraction step. For p-benzyne, the difference in the barriers of
the first and the second H abstractions from methanol is about 1.6 kcal/
mol at the CASPT2N/6-31G**//CAS/3-21G level: Logan, C. F.; Chen, P.
J. Am. Chem. Soc. 1996, 118, 2113. See also: Schottelius, M. J.; Chen, P.
J. Am. Chem. Soc. 1996, 118, 4896. For a more detailed description of TB
interactions, see: Hoffman, R.; Imamura, A.; Hehre, W. J. J. Am. Chem.
Soc. 1968, 90, 1499. Hoffman, R. Acc. Chem. Res. 1971, 4, 1. Squires, R.
R.; Cramer, C. J. J. Phys. Chem. A 1998, 102, 9072. For the theoretical
analysis of the TB interaction in cyclohexyl cations, see: Alabugin, I. V.;
Manoharan, M. J. Org. Chem. 2004, 69, 9011.
Steric compression of the enediyne moiety by ortho substit-
uents is further illustrated in Figure 2, which clearly shows that
steric destabilization is accompanied by a decrease in the
distance between the terminal acetylenic carbons (C1C6).
In this paper, we test our computational predictions with
experimental studies designed to determine whether ortho
substituents are capable of decreasing the activation energy of
the Bergman cyclization. To obtain reliable results, we will
(31) Alabugin, I. V.; Manoharan, M.; Kovalenko, S. V. Org. Lett. 2002,
4, 1119.
(32) Zeidan, T. A.; Manoharan, M.; Alabugin, I. V. J. Org. Chem.
Preceding paper in this issue.
(33) Eliel, E. L.; Wilen, S. H.; Doyle, M. P. Basic Organic Stereochem-
istry; Wiley-Interscience: New York, 2001; p 459.
(34) For NBO analysis of these H bonds, see ref 31.
(35) Mu¨ller-Dethlefs, K.; Hobza, P. Chem. ReV. 2000, 100, 143.
964 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
FIGURE 1. Steric-assistance (a, c) and H-bonding (b, d) mechanisms for the “ortho” effect. Energy profile for the para isomer is given in solid
blue lines, whereas data for the ortho isomer are shown in solid red lines. Calculations were performed at the BLYP/6-31G** (in bold) and B3LYP/
6-31G** levels. P stands for products, and R stands for reactants.
to compare the accuracy of commonly used kinetic methods
and to show that the complex thermal reactivity of enediynes
requires more detailed kinetic analysis of other steps in the
cycloaromatization cascade. As a result, both experiment and
theory are refined, and a more balanced description of such
processes is achieved.
Results and Discussions
Synthesis of Enediynes. The model enediynes were synthe-
sized using standard literature procedures outlined in Scheme
2. The aromatic dihalides or ditriflates36 were coupled with
trimethylsilylacetylene under Sonogashira conditions.37 In ad-
dition to the ortho-substituted substrates, we prepared two
reference compounds: the parent benzannelated enediyne 1 and
the para-NO2 enediyne 3. Treatment of the trimethylsilyl
(TMS)-protected enediynes with a base (typically 1 N NaOH)
afforded ortho-substituted benzannelated enediynes in good
yields. Bergman cyclizations of enediynes on a preparative scale
FIGURE 2. Correlation between the C1C6 distance and the relative
energies of ortho and para isomers of substituted benzannelated
enediynes.
critically evaluate several common experimental approaches.
Such an evaluation is necessary because the available body of
literature on the Bergman cyclization kinetics is highly hetero-
geneous and, thus, a direct comparison of related results from
different groups is often problematic. To avoid complications
that exist when other reactions compete with the Bergman
cyclization (vide infra), we will concentrate on the Bergman
cycloaromatization of enediynes with acceptor substituents,
where side reactions are minimized. This choice will allow us
(36) Powell, N. A.; Rychnovsky, S. D. Tetrahedron Lett. 1996, 37, 7901.
(37) (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975,
4467. (b) Sonogashira, K. In Metal-Catalyzed Reactions; Diederich, F.,
Stang, P. J., Eds.; Wiley-VCH: New York, 1998. (c) Stang, P. J. In Modern
Acetylene Chemistry; Diederich, F., Ed.; VCH: Weinheim, Germany, 1995.
(d) Tykwinski, R. R. Angew. Chem., Int. Ed. 2003, 42, 1566.
J. Org. Chem, Vol. 71, No. 3, 2006 965

Zeidan et al.
FIGURE 3. DSC scans of neat benzannelated enediynes at a heating rate of 5, 10, 15, and 20 °C/min and the nonrefined activation energies45
(insets) for (a) 1,2-diethynylbenzene, (b) 2,3-diethynylnitrobenzene, (c) 3,4-diethynylnitrobenzene, and (d) 2,3-diethynyl-1-formylbenzene.
SCHEME 2. Synthesis of Ortho-Substituted Benzannelated
Enediynes
Differential Scanning Calorimetry (DSC) Analysis. An
initial comparison of the thermal reactivities of the ortho-
substituted enediynes was carried out using DSC, the method
which has rapidly gained popularity as a way to scan enediynes
for increased reactivity. The quickest but also the crudest way
to use DSC is based on the onset of the exothermic peak.38
Results of these measurements are given in Figure 3. Contrary
to the expectations based on the computational analysis, the
onset temperature for 1,2-diethynylbenzene is only 9-10 °C
higher than that for the ortho-NO2 and ortho-CHO substituted
enediynes and 60 °C lower than for the para-NO2 enediyne.39
The latter observation is in direct conflict with the experimental
trend determined by Russell and co-workers in solution.27
Interestingly, the ortho-CHO and para-NO2 enediynes react
immediately after melting. These observations suggest that the
solid state may impose constraints on reactivity.40 Considering
all of the above, it is clear that trends in the onset DSC
temperatures of pure enediynes cannot be used as a universal
criterion of the relative reactivity.
(38) For reviews, see ref 16 and Basak, A.; Mandal, S.; Bag, S. S. Chem.
ReV. 2003, 103, 4077. Perera, K. P. U.; Shah, H. V.; Foulger, S. H.; Smith,
D. W., Jr. Thermochim. Acta 2002, 388, 371.
(39) Onsets of the exothermic peaks for enediynes at different heating
rates (5, 10, 15, and 20 °C) are 132, 142, 149, and 151 °C (1), 114, 123,
133, and 129 °C (2), 193, 203, 210, and 217 °C (3), and 127, 137, 142, and
147 °C (4).
(40) For an example of the crystal control of photochemical reactivity,
see: Zimmerman, H. E.; Alabugin, I. V., Smolenskaya, V. N. Tetrahedron
2000, 56, 6821 and references therein.
were performed in Pyrex glass tubes containing a solution of
the enediyne and 1,4-cyclohexadiene (1,4-CHD) in chloro-
benzene at 0.10 and 10.0 M concentration of enediynes and
1,4-CHD, respectively.
966 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
FIGURE 4. Correlation between activation energies calculated using MP2, B3LYP, and BLYP levels79 with the 6-31G(d,p) basis set and activation
energies determined using the DSC method of benzannelated enediynes in 1,4-CHD bearing electron acceptor groups in the ortho position: (a) neat
1,4-CHD solution and (b) neat benzannelated enediynes. 1,2-Diethynylbenzene is not included in these correlations.
To evaluate the effect of crystalline restraints, we performed
the DSC analysis in neat 1,4-CHD (see Supporting Information
for the plots). In these experiments, 50 µL portions of the stock
solutions of enediynes in 1,4-CHD were sealed in “high volume
pans”41 under argon in a glovebox. DSC scans were carried
out under similar conditions as described above but with a
maximum temperature of 250 °C (see Supporting Information).42
In all cases, reaction exotherms were observed after the initial
endothermic response caused by the partial solvent vaporization.
After each DSC run, the samples were tested by TLC and found
to contain the corresponding naphthalenes as the major products.
However, the trends for the onset peaks were similar to those
observed for the neat enediynes.
More detailed kinetic information can be obtained from the
combination of the DSC analysis with the ASTM E 698 thermal
stability method (insets).43 The ASTM E 698 analysis requires
three or more experiments at different heating rates and, under
the assumption of a first-order kinetic model, provides activation
energies for the reaction.44 We carried out the nonisothermal
DSC scans for neat enediynes and their solutions in 1,4-CHD
at four different heating rates (5, 10, 15, and 20 °C). As
expected, an increase in the heating rate shifts the peaks of the
exotherms to higher temperatures. The activation energies are
then obtained by plotting ln(ꢀ) versus 1/RT, where ꢀ is the peak
maximum of the exotherm.45
reactions with low activation barriers that become especially
important when the reaction is conducted in neat enediyne. The
cumulative heat release does not provide individual information
about the evolution of the chemical species involved in specific
processes which may coexist competing with the Bergman
cyclization. Chain radical cyclizations involving the addition
of p-benzyne or other radical intermediates to the acetylene
moieties or to 1,4-CHD are the most likely candidates for such
processes.47-49 Moreover, the shape of the 1,2-diethynylbenzene
exotherm is indicative of an autocatalytic reaction, which agrees
very well with the chain radical process.50 These observations
also explain the relatively low (35%) yield for the naphthalene
product, even in the presence of a large excess of 1,4-CHD.
Similar to the first set of DSC experiments, higher heating rates
resulted in the sharpening of the exothermic peak and the
shifting of its maximum to higher temperatures. As a result of
the lower reactivity of the para-NO2 enediyne, lower heating
rates were used to ensure the completion of the exothermic peak
before the maximum operational temperature of 250 °C of the
high volume pans was reached.
The DSC activation energies for the cyclization of three
enediynes with acceptor substituents are in excellent agreement
with the earlier DFT estimates. Correlations given in Figure 4
suggest that BLYP values are closer to the experimental DSC
data. However, the difference is small and B3LYP is at least as
good as BLYP (or better) in predicting relatiVe trends in the
reactivity of this family of substituted enediynes. However, it
is clear that the case of X ) H is an outlier. One can attribute
Again, the anomalously low DSC activation energy for 1,2-
diethynylbenzene disagrees with the experimental results of
Grissom et al.46 and Russell et al.27 and with the computationally
predicted trends. This discrepancy is likely to stem from side
(47) (a) Ko¨nig, B.; Pitsch, W.; Klein, M.; Vasold, R.; Prall, M.; Schreiner,
P. R. J. Org. Chem. 2001, 66, 1742. (b) Kovalenko, S. V.; Peabody, S.;
Manoharan, M.; Clark, R. J.; Alabugin, I. V. Org. Lett. 2004, 6, 2457.
Peabody, S. W.; Breiner, B.; Kovalenko, S. V.; Alabugin, I. V. Org. Biomol.
Chem. 2005, 3, 218. (c) Johnson, J. P.; Bringley, D. A.; Wilson, E. E.;
Lewis, K. D.; Beck, L. W.; Matzger, A. J. J. Am. Chem. Soc. 2003, 125,
14708.
(48) Rule, J. D.; Wilson, S. R.; Moore, J. S. J. Am. Chem. Soc. 2003,
125, 12992.
(49) For theoretical analysis, see: (a) Alabugin, I. V.; Manoharan, M.
J. Am. Chem. Soc. 2005, 127, 9534. (b) Alabugin, I. V.; Manoharan, M. J.
Am. Chem. Soc. 2005, 127, 12583.
(41) High volume pans are aluminum pans usually used for liquid
samples. They were purchased from TA Instrument, Inc.
(42) As a result of high pressure produced inside the DSC high volume
pans caused by heating the 1,4-CHD, the maximum temperature reached
was 250 °C.
(43) Standard Test Method for Arrhenius Kinetic Constants for Thermally
Unstable Materials (ANSI/ASTM E 698-99) ASTM Book of Standards;
ASTM: Philadelphia, PA, 2000; Vol. 14.02, pp 299-305.
(44) Ozawa, T. J. Thermal Anal. 1970, 2, 301.
(45) Activation energies determined from the plots in Figure 3 are refined
according to the ASTM E 698-99 protocol (see ref 42). The refined
activation energies are reported in Table 4.
(46) Grissom, J. W.; Calkins, T. L.; McMillen, H. A.; Jiang, Y. H. J.
Org. Chem. 1994, 59, 5833.
(50) Jacobs, P. W.; Tompkins, F. C. In Chemistry of the Solid State;
Garner, W. E., Ed.; Academic Press: New York, 1955; p 184. Also see:
Bou-Diab, L.; Fierz, H. J. Hazard. Mater. 2002, 93, 137.
J. Org. Chem, Vol. 71, No. 3, 2006 967

Zeidan et al.
FIGURE 5. Arrhenius plots for the disappearance of ortho-substituted enediynes (solid line) and the appearance of the corresponding naphthalenes
(dashed lines). (a) 1,2-diethynylbenzene, (b) 2,3-diethynylnitrobenzene, (c) 3,4-diethynylnitrobenzene, and (d) 2,3-diethynyl-1-formylbenzene.
this difference to the greater thermal stability of acceptor
enediynes and to their lesser propensity to participate in radical
processes.47 Interestingly, the presence of 1,4-CHD does not
improve the quality of the correlation, and in both cases, the
slope of the correlation between experimental and theoretical
barriers is significantly different from 1.
in Supporting Information). Similarly, the appearance of naph-
thalene products followed a pseudo-first-order kinetic behavior.
The effective rate constants, keff, for the formation of the
Bergman products were determined by fitting the data to eq 2,
which normalizes the formation of the product to the observed
reaction yield,52
The anomalous position of the unsubstituted enediyne sug-
gests that DSC is not universally reliable for kinetic studies of
the Bergman cyclization. Only in the cases of kinetically simple
and chemically efficient processes within a family of related
compounds (e.g., enediynes with acceptor substituents) can such
comparisons be justifiable. Thus, we carried out a more thorough
kinetic analysis on the basis of following the concentrations of
the reactants and the products during the cycloaromatization
process.
Effective Rate Constants and Activation Energies for the
Cyclization of Ortho-Substituted Enediynes. The conditions
described by Grissom et al. (freeze/pump/thaw degassed solu-
tions, sealed capillaries, and chlorobenzene as a solvent, see
Experimental Section)51 were used to allow a comparison with
the earlier literature. Effective rate constants for the disappear-
ance of ortho-substituted enediynes and the appearance of the
corresponding naphthalene products were measured at different
temperatures. The consumption of enediynes followed the
pseudo-first-order exponential dependence of eq 1 (Figure S1
efft)
[A] ) [A]0e(-k
(1)
(2)
[P∞]
ln(X) ) ln
) kefft
(
)
[P∞] - [C]
where [A]0 is the initial concentration of the enediynes, [P∞] is
the yield of the product, and [A] and [C] are the concentrations
of the enediynes and the naphthalenes at time t, respectively.53
Arrhenius plots based on the pseudo-first-order effective rate
constants at greater than or equal to four different temperatures
provided the activation energies given in Figure 5.
Unlike the DSC data, all of the above results are consistent
with the theoretically predicted trends. Although the rates of
the CHO-naphthalene formation have the largest margin of
error as a result of the increased formation of byproducts,54,55
(52) The use of [P∞] instead of [A]0 in eq 2 eliminates the nonlinearity
of the fit at higher conversions and has an effect on the values of the
measured Arrhenius barriers.
(53) Some of the data points at long reaction times had to be discarded
because the reaction becomes less efficient at the low concentration levels
of the starting enediynes and errors in integrating the areas of the naphthalene
products increase substantially. See Supporting Information for complete
details of kinetic experiments.
(51) (a) Grissom, J. W.; Calkins, T. L J. Org. Chem. 1993, 58, 5422.
(b) Grissom, J. W.; Calkins, T. L.; McMillen, H. A.; Jiang, Y. H. J. Org.
Chem. 1994, 59, 5833. (c) Grissom, J. W.; Gunawardena, G. U. Tetrahedron
Lett. 1995, 36, 4951.
968 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
TABLE 1. Effective Rate Constants for the Disappearance of
Nitroenediynes (kEefDf ) and the Appearance of Nitronaphthalenes
(knefafphth) at 140 °Ca
SCHEME 3. Kinetically Important Steps Involved in the
Cycloaromatization Cascade
ortho- NO2 enediynesb
para-NO2 enediynesc
[1,4-CHD]
(M)
ED
eff
naphth
eff
ED
naphth
eff
k
× 104
k
× 104
k
× 104
k
× 104
eff
0.73
0.58
0.44
0.29
0.15
0.07
0.03
1.85 ( 0.02 1.64 ( 0.01
1.68 ( 0.03 1.57 ( 0.01
1.60 ( 0.03 1.53 ( 0.01
1.48 ( 0.02 1.46 ( 0.01
1.22 ( 0.02 1.15 ( 0.01
0.99 ( 0.02 0.92 ( 0.02
0.73 ( 0.01 0.63 ( 0.01
6.98 ( 0.07 6.63 ( 0.06
6.44 ( 0.10 6.44 ( 0.03
5.74 ( 0.06 5.48 ( 0.01
4.31 ( 0.02 4.28 ( 0.02
3.06 ( 0.02 2.66 ( 0.01
2.23 ( 0.02 1.06 ( 0.02
a Pseudo-first-order effective rate constants (keff) are in s-1 b Initial
.
it is clear that both ortho-CHO and ortho-NO2 substrates have
increased reactivity relative to para-NO2 enediyne and 1,2-
diethynylbenzene (effective rate constants are given in Sup-
porting Information). Among the two reference compounds, the
para-nitroenediyne is slightly more reactive, as shown earlier
by Russell and co-workers.27 When eq 2 is used, activation
energies for the disappearance of all the enediynes are in
excellent agreement with the activation energies derived from
the appearance of the naphthalene products.
In general, the para-nitroenediyne is much better suited to
be the reference point than 1,2-diethynylbenzene because it
forms 2-nitronaphthalene cleanly and in a high yield27 and
because the presence of the para-NO2 moiety at the para position
approximates the electronic effect of the ortho-nitro group on
the out-of-plane π system. A comparison of para- and ortho-
nitroenediynes clearly confirms that it is not the presence of an
acceptor substituent per se but its location that is of primary
importance.
Generalized Kinetic Scheme for the Cycloaromatization
Cascade. Although the usual focus of kinetic studies of the
thermal reactivity of enediynes is on the Bergman cyclization,
a number of further steps are needed to complete the cycloaro-
matization cascade, as illustrated in Scheme 3. The earlier
observation that the cycloaromatization rate of benzannelated
enediynes is sensitive to the 1,4-CHD concentration strongly
suggests that these steps also contribute to the overall reaction
kinetics.28 Thus, we have chosen para-nitro and ortho-nitro
benzannelated enediynes (2,3- and 3,4-diethynylnitrobenzenes)
for a more detailed kinetic study aimed at the better understand-
ing of the other steps in the cycloaromatization cascade.
The two steps of particular significance are the retro-Bergman
cyclization, which is much faster in benzannelated enediynes29
than in their acyclic analogues, and the first hydrogen abstraction
step, which is decelerated by the loss of a through-bond (TB)
interaction between the two radical centers.30
When it is assumed that the major path for the consumption
of nitroenediynes and the formation of nitronaphthalenes is the
one depicted in Scheme 3 and that the formation of byproducts
D is independent of the HD presence (e.g., the major byproducts
are polymers from the reaction of p-benzynes with enediynes),
the effective rate constant for the consumption of A and the
formation of C can be described by eqs 3 and 4, respectively.56
concentration of enediynes ) 2.7 × 10-3 M. c Initial concentration of
enediynes ) 3.8 × 10-3 M.
eqs 3 and 4 to eq 5. The latter can be rearranged to eq 6 which
can be used to obtain k1 graphically as the intercept of the plot
of 1/keff versus 1/[HD]:
k1k2[HD]
keff
)
)
(5)
(6)
k2[HD] + k-1
k-1
1
keff
1
k1
+
k1k2[HD]
Effect of HD Concentration on the Effective Rate Con-
stants. The above kinetic scheme has several verifiable con-
sequences. First, it explains the effects of HD concentration on
the rates of the Bergman cycloaromatization on a quantitative
basis. Second, eq 6 allows one to recover the true first-order
rate constants for the cyclization step using a simple graphical
method. Most importantly, this scheme allows one to develop
a general method for deconvoluting kinetic contributions from
the individual steps in the overall kinetic scheme of the
cycloaromatization cascade (k1, k-1, k2, etc.). The key element
in this development involved the analysis of the HD concentra-
tion effect on the overall reaction kinetics.
For a detailed mechanistic investigation of the Bergman
cyclization, we have studied the effect of the concentration of
the hydrogen-atom donor, 1,4-CHD, on the effective rate
constant for the disappearance of enediynes and for the
appearance of the corresponding naphthalene. 2,3-Diethynyl-
nitrobenzene (ortho-NO2) and 3,4-diethynylnitrobenzene (para-
NO2) were chosen as model compounds.
Rate constants for the disappearance of enediynes and the
appearance of nitronaphthalenes at 140 °C are given in Table
1. Single-exponential decay (eq 1) was used to model the
disappearance of enediynes with R2 > 0.9, while linear
correlation based on eq 2 was applied to the appearance of
naphthalene. The effective rate constants (keff), determined from
monitoring the disappearance of enediynes, include the rates
of the Bergman cyclization, the retro-Bergman ring opening,
and the hydrogen abstraction steps in addition to other reactions
that do not lead to the naphthalene products. On the other hand,
the effective rate constants (keff) determined from monitoring
k1k2[HD] + k1k3
kCeff
kAeff
)
) kCeff + kDeff
)
(3)
(4)
yield
k2[HD] + (k-1 + k3)
(54) According to the 1H NMR spectra, the major byproducts do not
have aldehyde functionality but incorporate the modified 1,4-CHD moiety.
(55) Several of these rates even become slightly larger than the rates of
the enediyne disappearance, and the result would be physically impossible
if it were outside the error margins.
k1k2[HD]
k2[HD] + (k-1 + k3)
kCeff
)
When the hydrogen abstraction step is much faster than the
step leading to the byproducts D, k2[HD] . k3, one can reduce
(56) Derivations of the rate equations are given in Supporting Informa-
tion.
J. Org. Chem, Vol. 71, No. 3, 2006 969

Zeidan et al.
FIGURE 6. Dependence of effective rate constant, keff, from 1,4-CHD concentration at 140 °C; (a) disappearance of enediynes and (b) appearance
of nitronaphthalenes. Filled red circles are data for ortho-nitroenediyne and hollow red circles are data for para-nitroenediyne; eq 5 was used to fit
the data (filled black squares and hollow squares) when k1 ) 8.3 × 10-4 s-1 and k-1/k2 ) 0.125 M for ortho-nitro and k1 ) 1.8 × 10-4 s-1 and
k-1/k2 ) 0.067 M for para-nitro.
TABLE 2. Activation Energies Ea Determined from Effective Rate Constants for the Consumption of NO2 Enediynes and the Appearance of
Nitronaphthalenes at Different Concentrations of [1,4-CHD]a
Ea for ortho-NO2 enediyneb
Ea for para-NO2 enediynec
[1,4-CHD]
(M)
disappearance of
appearance of
naphthalene
disappearance of
appearance of
naphthalene
enediyne
enediyne
0.73
0.58
0.44
0.29
0.15
0.07
0.03
24.8 ( 0.3
25.0 ( 0.4
24.0 ( 0.8
23.9 ( 0.3
22.9 ( 0.4
22.1 ( 0.5
22.2 ( 0.6
24.9 ( 0.8
25.2 ( 0.6
23.4 ( 0.9
22.8 ( 0.9
23.5 ( 0.5
22.9 ( 2.3
23.2 ( 4.8
23.2 ( 0.5
21.7 ( 0.1
21.5 ( 0.1
20.5 ( 0.1
19.6 ( 0.9
19.1 ( 0.4
24.0 ( 0.3
22.8 ( 1.3
21.0 ( 0.9
19.6 ( 0.7
20.9 ( 2.8
19.8 ( 2.0
a Activation energies in kcal/mol. b Initial concentration of enediyne ) 2.7 × 10-3 M. c Initial concentration of enediyne ) 3.8 × 10-3 M.
the appearance of naphthalene products comprise only the steps
en route to the naphthalene product.
scattered when the concentration of 1,4-CHD is low. Conse-
quently, activation energies for the appearance of nitronaph-
thalenes at 0.07 and 0.03 M 1,4-CHD have error bars in the
order of (2.0-4.8 kcal/mol. These results are consistent with
the change in the main reaction path from the Bergman
cycloaromatization to the polymerization and formation of
byproducts D at the lower concentrations. At a low concentration
of 1,4-CHD, the effective rate constant in eq 3 transforms to
that of eq 7 (see Supporting Information), which is based mainly
on the rate constants of the path leading to the formation of
byproducts D. Table 2 suggests that the activation energies for
the formation of byproducts D are less than 19.1 and 22.1 kcal/
mol for ortho-NO2 and para-NO2 enediynes, respectively.
Because these values are lower than the activation barriers for
the Bergman cyclization, the side reactions are unlikely to
involve p-benzynes. Instead, chain polymerization is a reason-
able possibility.47
Both the effective rate constants for the consumption of
enediynes and the formation of nitronaphthalene showed a
dependence on the concentration of the hydrogen-atom donor
that is qualitatively similar to that of the literature dependence
for 1,2-diethynylbenzene.28 A plot of the effective rate constants
versus the 1,4-CHD concentration is shown in Figure 6. At low
concentrations of 1,4-CHD, the effective rate constants for the
consumption of enediynes are highly dependent on the concen-
tration of 1,4-CHD. However, this dependence is less pro-
nounced at higher concentrations of the HD, where the reaction
rate finally reaches a plateau, as reported earlier for other
enediynes.28 A similar behavior is observed for the effective
rate constants for the formation of nitronaphthalene.
Effective rate constants at different temperatures allowed us
to determine Arrhenius activation energies at different concen-
trations of 1,4-CHD. The measured effective activation barriers
are given in Table 2. An increase in the concentration of 1,4-
CHD increases the activation energies for both nitroenediynes,
but the ortho isomer is more sensitive to the 1,4-CHD
concentration effect. At all concentrations of 1,4-CHD, cor-
relations for effective rate constants for the consumption of
enediynes are good, with less than (1 kcal/mol error in respect
to the Arrhenius activation energies. On the other hand, effective
rate constants for the formation of naphthalene products are
Dissection of Kinetics. Because the HD concentration effects
on the activation barriers derived from the pseudo-first-order
approximation can lead to noticeable discrepancies between the
results from different sources and methods, it is beneficial to
determine the true rate of the first-order cyclization constants,
k1. In the following section, we will outline a simple way to
determine these constants, which provide the real activation
energies for the Bergman cyclization.
970 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
TABLE 3. Rate Constants at 140 °C Determined Using Equations 5 and 6a
ortho-NO2
para-NO2
disappearance of
enediyne
appearance of
naphthalene
disappearance of
enediyne
appearance of
naphthalene
Equation 5
7.97 ( 2.22
13.63 ( 2.40
k1 × 104 (s-1
)
8.52 ( 1.33
1.88 ( 1.11b
6.15 ( 1.56b
1.79 ( 1.00
7.25 ( 2.01
k-1/k2 × 102 (M)
13.78 ( 2.12
R2
0.997
0.984
0.951
0.990
Equation 6
8.27 ( 0.49
15.11 ( 0.10
k1 × 104 (s-1
)
8.26 ( 0.23
1.87 ( 0.07
6.82 ( 0.06
1.76 ( 0.04
6.86 ( 0.41
k-1/k2 × 102 (M)
12.58 ( 0.02
R2
0.997
0.991
0.974
0.989
a The lowest 1,4-CHD concentration (0.03 M) was not included in the fit, unless otherwise noted. b Data for 0.03 M 1,4-CHD was included.
FIGURE 7. Correlation between activation energies calculated using UMP2 with the 6-31G(d,p) basis set and activation energies determined for
(a) the consumption of enediynes and (b) the formation of the corresponding naphthalene products. Activation energies determined from the Bergman
cyclization rate constants, k1, are shown in filled red circles.
Both k1 and the k-1/k2 ratio can be determined either by
multicomponent fitting of the data to eq 5 or by the graphical
method based on eq 6. Table 3 compares the k1 and the k-1/k2
values obtained by these two methods for the two nitroene-
diynes.
the activation energies between ortho-NO2 and para-NO2
enediynes unequivocally supports the earlier computational
prediction of the ortho effect in the Bergman cyclization.31
Discussion of Substituent Effects. First, the DSC data should
be used with care and only for the most general qualitative
assessments of reactivity. The perfect fit of DSC results for
acceptor substituents discussed earlier in this paper is likely to
be fortuitous because the DSC data are inconsistent with the
results obtained from the classic kinetic approach either under
pseudo-first-order conditions or using multicomponent kinetic
dissection that recovers k1. At this stage, results based on
multicomponent analysis have significant error margins. There-
fore, our discussion will concentrate on results from the pseudo-
first-order analysis. Encouragingly, the activation energies for
the disappearance of enediynes with acceptor substituents and
the appearance of the corresponding naphthalenes generally
agree with each other (Figure 7). This observation suggests
kinetically well-defined processes, further justifying the pseudo-
first-order approximation.
The above results suggest that only at a 1,4-CHD concentra-
tion of 0.14 M do the rates of the hydrogen abstraction reach
that of the retro-Bergman opening and, thus, half of the ortho-
nitro-p-benzyne intermediate is intercepted before reverting to
the starting material. Such trapping is more efficient in the case
of the para-nitro substrate, where a 50/50 threshold is achieved
at a twofold smaller concentration of the H donor. One can
attribute this difference to the steric effect of the nitro group in
ortho-enediyne, which hinders the approach of 1,4-CHD from
one side of the p-benzyne (syn to the nitro substituent). The
difference in the rates of the retro-Bergman processes for the
two p-benzynes may also be important. In both cases, a change
to the rate-limiting step occurs at the steep part of the slope at
1
approximately /2 of the plateau height, as expected from the
qualitative reasoning.
Because cycloaromatization reactions provide a popular
testing ground for the development of computational methods,
in particular, different variations of DFT functionals,57 it is
interesting to comment on the performance of the two most
commonly used functionals (BLYP vs B3LYP) for these
reactions and investigate whether better alternatives are available
for the computational description of such processes.
The ortho-NO2 enediyne is indeed more reactive than the
two reference enediynes unsubstituted at the ortho position. The
3.6-3.9 kcal/mol difference in the barriers for X ) H agrees
The final Arrhenius activation energies determined from k1
(the true rate constant of the Bergman cyclization step) for the
consumption of ortho- and para-NO2 enediynes are 23.7 ( 1.1
and 25.4 ( 0.2 kcal/mol, whereas the activation energies
determined from the formation of the nitronaphthalene products
are 23.4 ( 3.2 and 24.3 ( 1.2 kcal/mol, respectively. These
activation energies compare favorably with the activation
energies determined from the effective rate constants, keff, at
high 1,4-CHD concentrations. Although these values should be
more accurate than all of the previous results, accuracy here is
achieved at the expense of precision and, thus, the uncertainty
of the measurements increases. Nevertheless, the difference in
(57) Prall, M.; Wittkopp, A.; Schreiner, P. R. J. Phys. Chem. A 2001,
105, 9265.
J. Org. Chem, Vol. 71, No. 3, 2006 971

Zeidan et al.
TABLE 4. Comparison of Activation Parameters from DFT Computations and from DSC Experimentsa
activation energies
(kcal/mol)
pre-exponential factor
log[A] (s-1
)
X
BLYP (B3LYP)b
neat ED
in 1,4-CHD
neat ED
in 1,4-CHD
H
24.5 (31.3)
20.9 (27.9)
24.1 (30.9)
20.6 (27.7)
22.1 (29.2)
21.7 ( 0.3
23.2 ( 0.1
25.8 ( 0.3
23.1 ( 0.1
22.0 ( 0.6
25.9 ( 0.3
10.8 ( 0.3
11.9 ( 0.1
12.6 ( 0.2
10.0 ( 0.1
10.3 ( 0.6
10.8 ( 0.3
ortho-NO2
para-NO2
ortho-CHO (syn)
ortho-CHO (anti)
24.4 ( 0.2
23.2 ( 0.3
12.6 ( 0.2
10.8 ( 0.3
a Errors are standard deviations of the Arrhenius fits with 90% confidence limits. b 6-31G** basis set.
TABLE 5. MP2/6-31G** Activation Energies, Experimental Activation Parameters Derived from Pseudo-First-Order Experiments, and
Experimental Yields of 1-Substituted Naphthalenesa
activation energies
(kcal/mol)
pre-exponential factor
log[A] (s-1
)
enediyne
consumption
naphthalene
formation
enediyne
consumption
naphthalene
formation
yield
(%)
X
MP2b
H
25.3
22.1
24.9
22.1
23.6
25.9 ( 1.1
21.5 ( 0.1
23.9 ( 0.3
26.0 ( 1.7
21.0 ( 0.9
24.5 ( 0.4
9.4 ( 0.6
8.2 ( 0.1
8.8 ( 0.2
7.8 ( 0.8
7.9 ( 0.5
9.0 ( 0.2
35
75
70
ortho-NO2
para-NO2
ortho-CHO (syn)
ortho-CHO (anti)
22.1 ( 0.4
22.4 ( 1.0
7.9 ( 0.5
8.4 ( 0.5
48
a Errors are standard deviations of the Arrhenius fits with 90% confidence limits. b 6-31G** basis set.
TABLE 6. Comparison of Substituent Effects on the Stability of Reactants, TS, and Products of the Bergman Cyclization and Related
Naphthalene Monoradicals at the BLYP (B3LYP) [MP2]/6-31G** Levelsa
δER
δETS
δEP
δEMR
NO2
CHO (syn)
CHO (anti)
5.40 (5.83) [4.91]
3.85 (3.88) [3.32]
1.19 (1.25) [0.39]
2.19 (2.84) [2.22]
0.19 (0.54) [0.47]
-0.91 (-0.60) [-1.02]
1.20 (2.06)
2.17 (0.50)
1.26 (-0.40)
(2.08) [2.49]
(0.01) [1.06]
(0.62) [1.38]
a Para-substituted isomers are taken as the reference points; positive values indicate destabilization by the ortho substituent.
well with the 3.4-3.6 kcal/mol DFT predictions (Table 4). The
2 kcal/mol barrier decrease for the para-NO2 substrate is slightly
lower than the approximately 3 kcal/mol DFT prediction, but
this difference is well within the experimental error.
determine the activation energies and may have an edge over
DFT when noncovalent interactions contribute to the barrier
variations. Table 6 illustrated the difference between DFT and
MP2 methods in the description of the substituent effects where
such direct through-space interactions are important. The steric
destabilization of the reactants is noticeably reduced at the MP2
level, while the TSs have similar absolute energies (Table 5).
We did not attempt a comparison of the two methods toward
p-benzyne products as a result of the known difficulties in
describing multiconfigurational wave functions of singlet diradi-
cals. On the other hand, the comparison of the respective
monoradicals is possible and suggests a lesser decrease in the
steric destabilization at the MP2 level relative to the unrestricted
DFT data. Together, lesser destabilization of the reactant and
greater destabilization of the product result in a 0.9-1.6 kcal/
mol increase in the reaction endothermicity and the attenuation
of the ortho effect. Such differences between DFT and MP2
may become important for the accurate computational predic-
tions of the activation barriers of the Bergman cyclizations of
designed enediynes.
These results parallel the thermal behavior of ortho-CHO
substituted enediyne, where the activation barriers determined
by the three methods (Tables 4 and 5) are again within the
experimental error. One has to mention that the rates were less
accurate in this case because the data are more scattered as a
result of the presence of side reactions and a lower yield of the
naphthalene product. In addition, we did not study the effect of
1,4-CHD on the cyclization kinetics in this case. As a result,
the activation barriers may be underestimated, explaining why
the barrier lowering relative to X ) H is more than expected
from the DFT computations. Nevertheless, the Arrhenius fits
at 0.27 M 1,4-CHD are acceptable and the 1.8-2 kcal/mol
difference relative to the para-NO2 enediyne agrees with the
DFT prediction (Table 4) rather well.
From a practical point of view, an interesting observation is
that, in some cases, the best fit to the experimentally determined
activation energies is provided not by DFT but by MP2
calculations. The advantage of MP2 in such cases is likely to
be a better description of the noncovalent interactions respon-
sible for the ortho effect. Although unrestricted Moeller-Plesset
second-order perturbation theory (UMP2) performs poorly in
the description of p-benzynes, it has been shown before that
the TS of the Bergman cyclization is reactant-like with little
diradical character.19,20 As a result, MP2 can be used to
Conclusion
The experimental results presented in this work confirm
earlier computational predictions of the ortho effect in benz-
annelated enediynes. Both ortho-NO2 and ortho-CHO substit-
uents substantially accelerate the Bergman cyclization. In
particular, the difference in the activation energies for ortho-
972 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
1H NMR (300 MHz, CDCl3): δ 10.2 (s, 1H), 8.0 (dd, 1H, J ) 7.5,
1.5 Hz), 7.8 (dd, 1H, J ) 8.1, 1.2 Hz), 7.7 (dd, 1H, J ) 8.1, 7.8
Hz). 13C NMR (75.5 MHz, CDCl3): δ 185.0, 141.1, 140.2, 131.0,
130.2, 129.8, 128.5, 118.3 (q, J ) 321.2 Hz), 118.2 (q, J ) 321.8
Hz). UV/vis (CH3CN): λmax (log ꢀ) 280 (3.18), 238 (3.85), 203
(4.54) nm. IR (neat): 3401, 3094, 2888, 2758, 1709, 1432, 1223,
1210, 1129, 958, 873 cm-1. HRMS (EI+): calcd for C9H4O7F6S2,
401.930 25; found, 401.929 98.
and para-nitroenediynes is in excellent agreement with the
difference in the activation barriers calculated at the BLYP/6-
31G** and BLYP/6-31G* levels. The absolute values of the
BLYP activation energies are closer to the respective experi-
mental barriers. Interestingly, the MP2/6-31G** level that is
not commonly used for theoretical studies on the Bergman
cyclization provides results that generally rival and, in some
cases, have a slight edge over the BLYP performance. We
suggest that MP2 is a reasonable alternative to DFT in those
cases where an accurate description of the contribution of the
noncovalent interactions to the activation energy is needed.
The acceptor substitution alleviates complications from the
competing radical reactions, allowing for a more accurate
analysis of kinetic trends. The plots of the effective rate constant
keff for the consumption of enediynes and for the formation of
the naphthalene products versus changes in the 1,4-CHD
concentration initially show a steep increase in the rates, which
is followed by a plateau at the higher concentrations of the HD.
At the lower HD concentrations, other reaction pathways
successfully compete with cycloaromatization. The cyclization
becomes the rate-limiting step at a threshold 1,4-CHD concen-
tration that is unique for each enediyne. Thus, a thorough
understanding of H-atom donor effects is important for accurate
kinetic studies of reversible Bergman cyclizations where the
dissection of reaction kinetics on the basis of a simple model
allows one to recover the true first-order rate constants.
General Procedure for Sonogashira Coupling. A suspension
of aryl dihalide or ditriflate (5.8 mmol), PdCl2(PPh3)2 (0.29 mmol),
and Cu(I) iodide (0.29 mmol) in 40 mL of amine solvent was
degassed three times with the freeze/pump/thaw technique in a
flame-dried, screw-cap, pressure tube. Trimethylsilylacetylene (14.5
mmol) was added using a syringe. The tube was capped, and the
mixture was heated in an oil bath equipped with a thermostat. The
reaction progress was monitored by TLC. After the total consump-
tion of the aryl halide, the reaction mixture was filtered through
Celite and washed with methylene chloride (3 × 30 mL). The
organic layer was washed with a saturated solution of ammonium
chloride (2 × 30 mL) and water (2 × 30 mL) and dried over
anhydrous Na2SO4. The solvent was removed in vacuo. The reaction
mixture was purified by flash chromatography on silica gel.
Synthesis of 2,3-Bis(trimethylsilylethynyl)nitrobenzene, 6. The
title compound was prepared by the Sonogashira coupling of 2,3-
dibromonitrobenzene (2.2 g, 7.8 mmol) with trimethylsilylacetylene
(1.92 g, 19.6 mmol) in 60 mL of (i-Pr)2NH at 95 °C for 24 h. The
reaction mixture was purified by flash chromatography on silica
gel (hexanes) to afford 1.85 g (72%) of the desired product as a
brown solid: mp 67-68 °C. 1H NMR (300 MHz, CDCl3): δ 7.86
(d, 1H, J ) 7.5 Hz), 7.68 (dd, 1H, J ) 8.1 Hz), 7.36 (dd, 1H, J )
8.1, 7.5 Hz), 0.28 (s, 9H), 0.27 (s, 9H). 13C NMR (75.5 MHz,
CDCl3): δ 150.9, 136.2, 128.8, 127.7, 123.7, 120.0, 108.6, 101.4,
101.1, 97.0, -0.5, -0.7. UV/vis (CH3CN): λmax (log ꢀ) 328 (4.28),
264 (4.20), 250 (4.48), 214 (4.49) nm. IR (neat): 2962, 2900, 2164,
1532, 1353, 1248, 846, 761 cm-1. HRMS (EI+): calcd for C16H21-
NO2Si2, 315.111 09; found, 315.111 04.
Computational Details
All reactant, product, and TS geometries involved in Bergman
cyclizations were optimized at the UB3LYP/6-31G**, UBL-
YP,58 UMP2, and MP2 levels using Gaussian 98 and Gaussian
03 programs.59 The TS wave functions were identical at UMP2
and MP2 levels. From frequency computations, we confirmed
that all transition structures had a single imaginary frequency.
Synthesis of 3,4-Bis(trimethylsilylethynyl)nitrobenzene, 7. The
title compound was prepared by the Sonogashira coupling of 3,4-
dibromonitrobenzene (3.1 g, 9.8 mmol) with trimethylsilylacetylene
(2.41 g, 246 mmol) in 60 mL of (i-Pr)2NH at 110 °C for 24 h. The
reaction mixture was purified by flash chromatography on silica
gel (hexanes) to afford 2.78 g (80%) of the desired product as an
off-white solid: mp 118-121 °C. 1H NMR (300 MHz, CDCl3): δ
8.3 (d, 1H, J ) 2.4 Hz), 8.0 (dd, 1H, J ) 9.0, 2.7 Hz), 0.29 (s,
18H). 13C NMR (75.5 MHz, CDCl3): δ 146.9, 133.2, 132.1, 127.4,
127.3, 122.8, 105.1, 102.0, 101.6, 101.1, 0.03, -0.01. UV/vis (CH3-
CN): λmax (log ꢀ) ) 307 (4.17), 261 (4.34), 229 (4.27), 220 (4.28)
Experimental Section
3-Nitrophenol, 1,2-dibromobenzene, trifluoromethanesulfonic
anhydride, trimethylsilylacetylene, and 2,3-dihydroxy-1-formyl-
benzene are commercially available and were used as received. 1,4-
CHD was distilled before use. 1,2-Diethynylbenzene46 and 2,3-
diethynylanisole31 were prepared according to the literature
procedures. Capillaries (300-mm length, 1.5-1.8-mm outer diam-
eter, 0.2-mm wall) used in kinetic studies were purchased from a
commercial supplier.
nm. IR (neat): 3280, 3106, 2110, 1573, 1522, 1354, 909, 661 cm-1
.
HRMS (EI+): calcd for C16H21NO2Si2, 315.111 09; found,
315.110 16.
Synthetic Procedures. Synthesis of 3-Formyl-phenyl-1,2-bis-
(trifluoromethanesulfonate), 5. 2,3-Dihydroxy -1-formylbenzene
(0.50 g, 3.62 mmol) and pyridine (1.5 mL, 18.1 mmol) were
dissolved in CH2Cl2 (25 mL). The mixture was cooled to 0 °C under
argon, and trifluoromethanesulfonic anhydride (1.4 mL, 8.0 mmol)
was added dropwise. The reaction was slowly warmed to room
temperature and stirred for 6 h. The mixture was diluted with CH2-
Cl2, washed with water, and dried over Na2SO4. The solvent was
removed under vacuo. The reaction mixture was purified by
chromatography on silica gel (10% ethyl acetate/hexanes) to afford
1.82 g (98%) of the desired product as a white solid: mp 38 °C.
Synthesis of 2,3-Bis(trimethylsilylethynyl)-1-formylbenzene,
8. The title compound was prepared by the Sonogashira coupling
of 3-formyl-phenyl-1,2-bis(trifluoromethanesulfonate) (0.70 g, 1.74
mmol) with trimethylsilylacetylene (0.41 g, 4.18 mmol) in 20 mL
of dry THF and with K2CO3 (1.2 g, 8.71 mmol) as the base at 110
°C for 8 h. The reaction mixture was purified by flash chromatog-
raphy on silica gel (1% ethyl acetate/hexanes) to afford 0.28 g (54%)
of the desired product: 1H NMR (300 MHz, CDCl3) δ 10.5 (s,
1H), 7.8 (d, 1H, J ) 7.5 Hz), 7.7 (d, 1H, J ) 7.5 Hz), 7.3 (dd, 1H,
J ) 8.1, 7.2 Hz), 0.28 (s, 9H), 0.27 (s, 9H); 13C NMR (75.5 MHz,
CDCl3) δ 191.4, 137.2, 136.2, 128.8, 127.9, 127.4, 126.4, 106.9,
101.7, 100.1, 97.9, -0.4, -0.5. UV/vis (CH3CN): λmax (log ꢀ) 322
(3.07), 284 (3.79), 253 (4.31) nm; IR (neat) 2956, 2846, 2361, 2336,
1699, 1559, 1250, 844 cm-1. HRMS (EI+) calcd for C17H22OSi2,
298.120 93; found, 298.120 17.
(58) (a) Becke, A. D. Phys. ReV. A: At., Mol., Opt. Phys. 1988, 38,
3098. (b) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B: Condens.
Matter 1988, 37, 785. (c) Stephens, P. J.; Devlin, F. J.; Chabalowski, C.
F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623.
(59) (a) Frisch, M. J.; Et al. Gaussian 98, revision A.9; Gaussian, Inc.:
Pittsburgh, PA, 1998. (b) Frisch, M. J.; Et al. Gaussian 03, revision C.01;
Gaussian, Inc.: Wallingford, CT, 2004 (see Supporting Information for the
complete references).
General Procedure for TMS Deprotection. An aliquot of 1 N
NaOH (5 mL) was added to a methanol solution (50 mL) of
enediyne (3.2 mmol). The mixture was stirred at room temperature
for 15 min. The progress of the reaction was monitored by TLC.
J. Org. Chem, Vol. 71, No. 3, 2006 973

Zeidan et al.
After the total transformation of the TMS acetylenes, the solvent
was removed in vacuo. Aqueous HCl (1 N, 10 mL) was added to
the crude mixture. The acidic mixture was extracted by dichlo-
romethane (3 × 20 mL). The organic layer was washed with water
(2 × 30 mL) and dried over anhydrous Na2SO4. The solvent was
removed in vacuo. The reaction mixture was purified by flash
chromatography on silica gel.
Conventional DSC was used to determine the reaction kinetics
using ASTM E 698 thermal stability protocol.43 This protocol is
used to determine Arrhenius activation energies and pre-exponential
factors.
Kinetic Studies. Kinetic Studies of 2,3-Diethynylnitrobenzene
(ortho-NO2). A 10 mL volumetric flask was charged with 2,3-
diethynylnitrobenzene (3.8 mg), 1,2,3,4-tetraphenylnaphthalene (7.8
mg), and 1,4-CHD (0.22 M). The solution was mixed and analyzed
with HPLC to determine the initial concentration of the enediynes
(2.2 × 10-3 M). Kinetic experiments were performed at 140, 148,
156, and 164 °C. The following conditions were used for kinetic
analysis using HPLC: A/B, 9:1 solvent system (A, hexanes; B,
hexanes/ethyl acetate, 30:1); 1 mL/min flow rate; and 265 nm
detector wavelength. Retention times were as follows: tstd ) 6.9
min, tnaphth ) 11.3, and tED ) 21.5 min.
Kinetic Studies of 2,3-Diethynyl-1-formylbenzene (ortho-
CHO). A 10 mL volumetric flask was charged with 2,3-diethynyl-
1-formylbenzene (3.7 mg), 1,2,3,4-tetraphenylnaphthalene (5.3 mg),
and 1,4-CHD (0.24 M). The solution was mixed and analyzed with
GC and HPLC to determine the initial concentration of enediynes
(2.4 × 10-3 M). Kinetic experiments were performed at 140, 150,
160, 170, and 180 °C. The GC time program had the following
parameters: initial temperature ) 70 °C for 5 min, 20 °C/min until
220 °C and hold for 5 min, then 50 °C/min until 320 °C and hold
Synthesis of 2,3-Diethynylnitrobenzene (ortho-NO2 Ene-
diynes), 2. The title compound was prepared by the TMS
deprotection of 2,3-bis(trimethylsilylethynyl)nitrobenzene (1.01 g,
3.21 mmol) with 1 N NaOH (5 mL) in 50 mL of MeOH at room
temperature for 15 min. The reaction mixture was purified by flash
chromatography on silica gel (hexanes) to afford 0.49 g (90%) of
1
the desired product as a white solid: mp 84 °C. H NMR (300
MHz, CDCl3): δ 7.9 (d, 1H, J ) 8.1 Hz), 7.7 (d, 1H, J ) 8.1 Hz),
7.4 (dd, 1H, J ) 8.1), 3.8 (s, 1H), 3.4 (s, 1H). 13C NMR (75.5
MHz, CDCl3): δ 151.0, 136.6, 128.5, 128.4, 124.2, 119.6, 89.7,
83.7, 79.9, 76.2. UV/vis (CH3CN): λmax (log ꢀ) 330 (2.98), 282
(3.85), 261 (4.07), 228 (4.05) nm. IR (neat): 3276, 2105, 1517,
1352, 741 cm-1. HRMS (EI+): calcd for C10H5O2N, 171.032 03;
found, 171.032 11.
Synthesis of 3,4-Diethynylnitrobenzene (para-NO2 Enediynes),
3. The title compound was prepared by the TMS deprotection
procedure of 3,4-bis(trimethylsilylethynyl)nitrobenzene (1.50 g, 4.76
mmol) with 1 N NaOH (10 mL) in 150 mL of MeOH at room
temperature for 15 min. The reaction mixture was purified by flash
chromatography on silica gel (hexanes) to afford 0.59 g (72%) of
for 1 min. Retention times were as follows: tED ) 4.7, tnaphth
)
5.2, and tstd ) 8.9 min. The following conditions were used for
kinetic analysis using HPLC: A/B, 8:2 solvent system (A, hexanes;
B, hexanes/ethyl acetate, 100:1); 1 mL/min flow rate; and 265 nm
detector wavelength. Retention times were as follows: tstd ) 8.5,
tnaphth ) 30.8, and tED ) 33.8 min.
1
the desired product as a white solid: mp 92 °C. H NMR (300
MHz, CDCl3): δ 8.3 (d, 1H, J ) 2.4 Hz), 8.1 (dd, 1H, J ) 8.4,
2.4 Hz), 7.6 (d, 1H, J ) 9.0), 3.6 (s, 1H), 3.5 (s, 1H). 13C NMR
(75.5 MHz, CDCl3): δ 147.3, 133.7, 131.4, 127.6, 126.8, 123.4,
86.6, 84.1, 80.4, 79.9. UV/vis (CH3CN): λmax(log ꢀ) ) 289 nm
(413), 238 (4.26), 218 (4.30). IR (neat): 2960, 2899, 2145, 1523,
1346, 1248, 847 cm-1. HRMS (EI+): calcd for C10H5O2N,
171.032 03; found, 171.031 07.
Kinetic Studies of 3,4-Diethynylnitrobenzene (para-NO2). A
10 mL volumetric flask was charged with 3,4-diethynylnitrobenzene
(6.5 mg), anthracene (10.5 mg), and 1,4-CHD (0.38 M). The
solution was mixed and analyzed with HPLC to determine the initial
enediyne concentration (3.80 × 10-3 M). Kinetic experiments were
performed at 140, 150, 160, and 170 °C. The following conditions
were used for kinetic analysis using HPLC (reverse phase): A/B,
3:7 solvent system (A, water; B, acetonitrile); 1 mL/min flow rate;
and 315 nm detector wavelength. Retention times were as follows:
tED ) 5.8, tnaphth ) 6.6, and tstd ) 15.1 min.
Synthesis of 2,3-Diethynyl-1-formylbenzene (ortho-CHO Ene-
diynes) 4. The title compound was prepared by TMS deprotection
of 2,3-bis(trimethylsilylethynyl)-1-formylbenzene (0.50 g, 1.68
mmol) with K2CO3 (1.00 g) in 30 mL of MeOH at room
temperature for 15 min. The reaction mixture was purified by flash
chromatography on silica gel (hexanes) to afford 0.20 g (78%) of
1
the desired product as a white solid: mp 113-114 °C. H NMR
Kinetic Studies under Different Concentrations of 1,4-CHD.
Kinetic Studies of 2,3-Diethynylnitrobenzene (ortho-NO2). A 10
mL volumetric flask was charged with 2,3-diethynylnitrobenzene
(25.0 mg, 14.6 × 10-3 M) and 1,2,3,4-tetraphenylnaphthalene (25.0
mg). Five master solutions (2.9 × 10-3 M each) were prepared
with different concentrations of 1,4-CHD, [1,4-CHD] ) 0.03, 0.07,
0.15, 0.29, 0.44, and 0.58 M. The enediynes initial concentration
for every stock solution was determined by GC and HPLC as 2.7
× 10-3 M. Kinetic experiments were performed at 140, 148, 156,
and 164 °C. The following conditions were employed for kinetic
analysis by HPLC: A/B, 9:1 solvent system (A, hexanes; B,
hexanes/ethyl acetate, 30:1); 1 mL/min flow rate; and 265-nm
detector wavelength. Retention times were as follows: tstd ) 6.9,
tnaphth ) 11.3, and tED ) 21.5 min.
Kinetic Studies of 3,4-Diethynylnitrobenzene (para-NO2). A
10 mL volumetric flask was charged with 3,4-diethynylnitrobenzene
(30.0 mg, 17.5 × 10-3 M) and anthracene (60.0 mg). Six master
solutions (3.5 × 10-3 M each) were prepared with different
concentrations of 1,4-CHD, [1,4-CHD] ) 0.03, 0.07, 0.15, 0.29,
0.44, 0.58, and 0.87 M. The enediynes initial concentration for every
stock solution was determined by GC and HPLC as 3.8 × 10-3 M.
Kinetic experiments were performed at 140, 150, 160, and 170 °C.
The following conditions were used for kinetic analysis using HPLC
(C13 reverse phase column): A/B, 3:7 solvent system (A, water;
B, acetonitrile); 1 mL/min flow rate; and 315 nm detector
(300 MHz, CDCl3): δ 10.6 (s, 1H), 7.9 (dd, 1H, J ) 7.8, 1.2 Hz),
7.7 (dd, 1H, J ) 8.1, 1.2 Hz), 7.4 (dd, 1H, J ) 8.1, 7.5), 3.7 (s,
1H), 3.4 (s, 1H). 13C NMR (75.5 MHz, CDCl3): δ 190.9, 137.5,
136.7, 128.5, 127.9, 127.1, 126.9, 88.5, 82.5, 80.4, 77.0. UV/vis
(CH3CN): λmax (log ꢀ) 319 (3.00), 262 (4.03), 242 (4.54) nm. IR
(neat): 3261, 2853, 2102, 1694, 1437, 1244, 794, 637 cm-1. HRMS
(EI+): calcd for C11H6O, 154.041 87; found, 154.042 08.
Representative Procedure for the Preparative Scale of Berg-
man Cyclizations. Enediyne (0.5 mmol) was dissolved in anhy-
drous chlorobenzene (9 mL). 1,4-CHD (50 mmol) was added, and
the mixture was placed in a Pyrex glass tube equipped with a joint.
The tube was attached through the joint to a vacuum line, and the
mixture was degassed three times by the freeze/pump/thaw
technique. The tube was sealed under argon, placed in an oil bath,
and heated to the corresponding temperature. After cooling,
chlorobenzene was distilled in vacuo, and the products were isolated
and purified by chromatography.
Differential Scanning Calorimetry Studies. Temperature and
enthalpy calibrations were performed using an indium standard.
The heat capacity was calibrated using a 25-mg sapphire standard.
About 8-10 mg of enediyne was used for each experiment. For
neat samples, enediynes were sealed in aluminum hermetic pans.
The samples were equilibrated in the purge gas (argon) for about
15 min prior to each run. For solution samples, stock solutions of
enediynes (5.0 mg) in 1,4-CHD (0.5 mL) were prepared, and 50
µL of the stock solutions were sealed in high volume aluminum
pans with O-rings in a glovebox under argon.
wavelength. Retention times were as follows: tED ) 5.8, tnaphth
6.6, and tstd ) 15.1 min.
)
974 J. Org. Chem., Vol. 71, No. 3, 2006

Ortho Effect in the Bergman Cyclization
Acknowledgment. The authors are grateful to the National
Science Foundation (CHE-0316598) and to Material Research
and Technology (MARTECH) Center at Florida State University
for partial support of this research, to the 3M Company for an
Untenured Faculty Award, to the Florida State University
Supercomputing Center and the National Center for Supercom-
puting Application (NCSA, UIUC) for providing computational
facilities, and to Professor Jack Saltiel for helpful discussions.
The authors are also thankful to Dr. Umesh Goli and Mr. Hank
Henricks of the Biochemical Analysis and Synthesis Services
(BASS lab) of the Department of Chemistry and Biochemistry
at Florida State University.
Supporting Information Available: Additional information
including the general kinetic experimental setup and NMR spec-
tra: 1H and 13C NMR. Computational details including complete
geometries and absolute energies at all levels of theory. Complete
ref 59. This material is available free of charge via the Internet at
http://pubs.acs.org.
JO0520801
J. Org. Chem, Vol. 71, No. 3, 2006 975