difference in the effect of angular versus linear annelation.
1
NICS(1.0) data13,14 corroborate the structural and H NMR
Scheme 2 a
trends: (proceeding along one “arm” from the inside to the
outside, i.e., form ring A to E, Scheme 1) 1a, -3.6, -0.3,
-5.3, 4.9, -8.8; 5, -1.4, -2.3, -7.5, 7.6, -7.6 ppm. Thus,
the central cyclohexatriene A in the former is more diatropic
than that in the latter (and in 8, -1.1 ppm),13a the two
respective adjacent cyclobutadienes B are not paratropic (as
in 8), but the remote ones D are, and more so in 5 than in
1a. Ring C in 1a is also more diatropic than the analogous
moieties in 6 (-3.3 ppm), 9 (-4.7 ppm), 10 (-4.7 ppm),
and 12 (-2.9 ppm).13a
In line with previous results attesting to the flexibility of
the phenylenes,3,13 we find a weak out-of-plane vibrational
mode at 20.4 cm-1. Initially puzzling were the estimates of
the heats of formation, established through the homodesmotic
reaction 1a/5 + 3 × biphenylene f 3 × 6 + 8, indicating
that 1a (∆H°f ) 484.6 kcal mol-1) is less stable than 5 (∆H°f
) 481.9 kcal mol-1) by 2.7 kcal mol-1. This result is at odds
with the general expectation2,13c that linear phenylene frag-
ments should be destabilized relative to their angular
counterparts. For example, the ∆∆H°f on going from 6 to 7
is +2.4 kcal mol-1,15 and the five [4]phenylene isomers
become more energetic along the series 8, 9/10, 12, 11 (total
∆∆H°f ) 8.6 kcal mol-1; BLYP/6-31G*).13c The rationale
for these trends is the destabilizing antiaromaticity associated
with the linear fusion, largely removed in the angular
topology. Indeed, angular dicylobutadienobenzene is sub-
stantially more stable than its linear isomer (17.6 kcal mol-1,
SCF/3-21G).16 On the other hand, there are indications that
without the π electronic component, σ strain inverts this order
of stability.16b In the higher phenylenes of mixed topology,
the energy differences between isomers are small enough
that this effect may become significant. Thus, the incremental
linear destabilization effect in going from 9/10 to 12 is much
smaller than that calculated on converting 6 to 7, and in the
family of the 12 [5]phenylenes, a number of inversions of
the expected stability ordering are noted; e.g., 13 is found
to be more stable than 14 (∼1.0 kcal mol-1; see the
Supporting Information).13b To address this issue unambigu-
ously, a calculational experiment was designed in which the
additional σ strain on linear and/or angular annelation on
the biphenylene nucleus to give 6, 7, 9-11, and 12 was
a Reaction conditions: (a) (CH3)3SiC2H, PdCl2(PPh3)2, CuI, NEt3,
THF, 97%; (b) CBr4, Zn, PPh3, CH2Cl2, 99%; (c) (i) LDA, THF,
-78 °C; (ii) aq NH4Cl, 95%; for 1a: (d) 1-iodo-2-(trimethylsilyl-
ethynyl)benzene, PdCl2(PPh3)2, CuI, i-Pr2NH, THF, 77%; (e)
K2CO3, MeOH, THF, 61%; (f) CpCo(CO)2, m-xylene, hV, ∆, 2%;
for 1b: (d) 1-(1-octynyl)-2-iodobenzene, PdCl2(PPh3)2, CuI, NEt3,
THF, 68%; (e) K2CO3, MeOH, THF, 86%; (f) CpCo(CO)2,
m-xylene, hV, ∆, 0.24%.
of 5, which extends to 527 nm.9 Most informative, in the 1H
NMR spectrum of 1a there are the two doublets for the
internal hydrogens at δ ) 6.49 and 6.42 ppm (for 1b, a
singlet at δ ) 6.25 ppm), close, but slightly shielded relative
to the corresponding hydrogens in (hexakistrimethylsilylated)
5 (δ ) 6.56 ppm),9 a trend identical to that observed in
angular (6; δ ) 6.13 ppm) versus linear [3]phenylene (7; δ
) 6.24 ppm).2 However, both of the former experience lesser
shielding compared to the latter, but for different reasons:
in 1, bond fixation in the central cyclohexatriene ring6
relieves some of the same in the penultimate benzenes,
diminishing the decrease in diatropism observed in the central
ring of 6; in 5, central bond fixation decreases the para-
tropism of the adjacent cyclobutadienes relative to that in 7.
To obtain further information about the pertinent physical
properties of 1a and 5, DFT calculations were performed at
the B3LYP/6-31G* level (see the Supporting Information),
a method that has provided remarkably accurate structural
information on these systems.3,6-10,13 Indeed, the calculated
structure of 5 is essentially within experimental error identical
with that of that determined for a hexasilyl derivative9 (except
for the substituted rings). Thus, by the criterion established
previously,6 its central ring is 92% bond fixed (cf. 8, 97%),
the terminal ones 29% (cf. 7, 32%). In contrast, 1a reveals
only 76% alternation in the central ring, 50% in the
penultimate (cf. 6, 62%), and 24% in the terminal benzenes
(cf. 6, 22%), exactly as predicted taking into account the
(12) (a) 1-Iodo-2-(trimethylsilylethynyl)benzene: Baxter, P. N. J. Org.
Chem. 2001, 66, 4170. (b) 1-Iodo-2-(1-octynyl)benzene was made from
1-bromo-2-iodobenzene; see the Supporting Information.
(13) (a) Schulman, J. M.; Disch, R. L.; Jiao, H.; Schleyer, P. v. R. J.
Phys. Chem. A 1998, 102, 8051. (b) Schulman, J. M.; Disch, R. L. J. Am.
Chem. Soc. 1996, 118, 8470. (c) Schulman, J. M.; Disch, R. L. J. Phys.
Chem. A 1997, 101, 5596.
(14) Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H.; Van E.
Hommes, N. J. R. J. Am. Chem. Soc. 1996, 118, 6317.
(15) The problem of the relative stability of these two isomers has a
controversial history; see, inter alia: (a) Rajca, A.; Safronov, A.; Rajca,
S.; Ross, C. R., II; Stezowski, J. J. J. Am. Chem. Soc. 1996, 118, 7272. (b)
Trinajstic´, N.; Schmalz, T. G.; ZÄ ivkovic´, T. P.; Hite, G. E.; Klein, D. J.;
Seitz, W. A. New J. Chem. 1991, 15, 27. (c) Diercks, R.; Vollhardt, K. P.
C. Angew. Chem., Int. Ed. Engl. 1986, 25, 266. (d) Berris, B.; Hovakeemian,
G.; Lai, Y.-H.; Mestdagh, H.; Vollhardt, K. P. C. J. Am. Chem. Soc. 1985,
107, 5670 and references therein.
(16) (a) Ma, J.; Li, S.; Jiang, Y. J. Phys. Chem. 1996, 100, 15068. (b)
Faust, R.; Glendening, E. D.; Streitwieser, A.; Vollhardt, K. P. C. J. Am.
Chem. Soc. 1992, 114, 8263.
Org. Lett., Vol. 5, No. 4, 2003
551