Synthesis of bent [4]phenylene (cyclobuta[1,2-a:3,4-b′]bisbiphenylene) and structure of a bis(trimethylsilyl) derivative: The last [4]phenylene isomer

Article abstract of DOI:10.1039/b109789j

The syntheses of the title compounds were accomplished by cobalt-catalyzed alkyne cyclotrimerizations using two strategies; the properties of the bent phenylene frame reflect the combined effects of benzocyclobutadienofusion of the component [3]phenylene

Full text of DOI:10.1039/b109789j

Synthesis of bent [4]phenylene (cyclobuta[1,2-a:3,4-bA]bisbiphenylene)
and structure of a bis(trimethylsilyl) derivative: the last [4]phenylene
isomer†
Dennis T.-Y. Bong, Lionel Gentric, Daniel Holmes, Adam J. Matzger, Frank Scherhag and K. Peter C.
Vollhardt*
Center for New Directions in Organic Synthesis, Department of Chemistry, University of California at
Berkeley, and the Chemical Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California
94720-1460, USA. E-mail: vollhard@cchem.berkeley.edu
Received (in Corvallis, OR, USA) 25th October 2001, Accepted 3rd December 2001
First published as an Advance Article on the web 16th January 2002
The syntheses of the title compounds were accomplished by
cobalt-catalyzed alkyne cyclotrimerizations using two strat-
egies; the properties of the bent phenylene frame reflect the
combined effects of benzocyclobutadienofusion of the com-
ponent [3]phenylene substructures.
lyzed coupling with a protected o-diethynylbenzene synthon
51a,d to give 6. Subsequent (trimethylsilyl)ethynylation and
deprotection furnished triyne 7, which underwent cobalt-
catalyzed cyclization to give orange, air-sensitive 1 in 8 steps
(from o-diethynylbenzene) and 1.7% overall yield.8 A more
convergent double cyclization approach is depicted in Scheme
2 and features the remarkably regioselective desymmetrization
of 1,2,4,5-tetrabromobenzene to the 5-iodotribromo analogue,
allowing for its selective alkynylation to eventually furnish
pentayne 8.9 Simultaneous intra- and intermolecular [2 + 2 +
2]cycloaddition in bis(trimethylsilyl)acetylene produced the
relatively (with respect to 1) less sensitive 9 in 5 steps and 7%
overall yield.8 While this derivative decomposed on attempted
conversion to 1, it proved valuable as a source of material
suitable for X-ray crystal analysis (Fig. 2)10 and preliminary
experiments.
Among the simple phenylene topologies hitherto assembled—
linear, angular, zig-zag, and branched1—is a glaring omission,
bent [4]phenylene 1. This compound is of interest because it
constitutes the fifth and last [4]phenylene isomer,2 and its
preparation allows the completion of an experimental analysis
of the comparative properties of this group of topomers,
including data based on calculations.3 Hydrocarbon 1 is also the
first and smallest member of a subclass of phenylenes in which
the linear and angular topologies are juxtaposed. Infinite arrays
of either (but not mixed) type have been analyzed theoretically
as potential novel electronic and magnetic materials.4
It is instructive to view 1 as the result of benzocyclobutadie-
nofusion of 2 and 3, respectively (Fig. 1). Such alteration of 2 is
expected to cause increased bond alternation in ring C, relieving
the antiaromaticity of the neighboring cyclobutadiene fragment,
in turn ‘relaxing’ the bisallyl frame of B to become more
benzenoid, i.e. diatropic and perhaps less reactive. In turn, the
effect of the corresponding fusion on 3 is anticipated to be bond
alternation in ring B in the direction indicated in structure 1 and
thus, by relay, even more pronounced cylohexatriene character6
in C, rendering this ring even less diatropic and possibly more
reactive. One notes (Fig. 1) that the trends in NICS values3a
corroborate these expectations. Does this simple picture hold?
Scheme 1 depicts the strategy used to assemble 1, starting
with 2,3-diiodobiphenylene7 and its (non-selective) Pd-cata-
On the basis of the NMR criterion of aromaticity,11 the
answer to the above question is affirmative. Thus (Fig. 1), the
protons of ring C in 1 are shielded relative to those correspond-
ing in 3 (and 4), and H6 more so than H5. On the other hand, the
Scheme 1 Reagents and conditions: i) [Pd(MeCN)2Cl2], CuI, PPh3,
piperidine, 90 °C, 40 h. ii) Me3SiC2H, [Pd(PPh3)2Cl2], CuI, piperidine, 44
h. iii) Bu4NF, THF, 20 min. iv) CpCo(CO)2, m-xylene, hn, D, 15 h.
Fig. 1 1H NMR chemical shifts (CDCl3, d, ppm) and NICS values (inside
the respective rings) of 1–4, and labeling of the benzene rings in 1 (A–
D).
Scheme 2 Reagents and conditions: i) a. BuLi, Et2O, 278 °C; b. I2, Et2O,
278 °C. ii) 5, [Pd(PPh3)2Cl2], CuI, Et3N, 15 h. iii) Me3SiC2H,
[Pd(PPh3)2Cl2], CuI, Et3N, 120 °C, 2.5 d. iv) Bu4NF, THF, 2 h. v)
CpCo(CO)2, Me3SiC2SiMe3, hn, D, 16 h. vi) H2, 1 atm, Pd/C, Et2O, 10
min.
† Electronic supplementary information (ESI) available: selected bond
distances and angles for 9, spectral and analytical information. See http:
//www.rsc.org/suppdata/cc/b1/b109789j/
278
CHEM. COMMUN., 2002, 278–279
This journal is © The Royal Society of Chemistry 2002

protons of ring B are deshielded compared to those correspond-
ing in 2, the slightly larger chemical shift of H12 being typical
of ‘bay region’ hydrogens in the phenylenes.1a The most
notable effect in the 13C NMR spectra is the shielding ( ~ 2–6
ppm) of the carbons of ring C relative to those corresponding in
3 (and 4), again signalling increased cyclohexatrienic charac-
ter.1
which allowed the detection of two distinct, interconverting
conformers (1+1)13 containing diagnostic doublets assignable to
H7, H8, H11, and H12.
In summary, the synthesis of the last isomer of the
[4]phenylenes has been accomplished. Its properties constitute
a blend of those of the component linear and angular
[3]phenylene substructures, and further delineate the effect of
topology on the interplay between antiaromatic cyclobutadie-
noid and aromatic cyclohexatrienic-benzenoid circuits.
This work was funded by the NSF (CHE-0071887). F. S. was
the recipient of a DFG postdoctoral fellowship. A. J. M. was an
ACS Division of Organic Chemistry Graduate Fellow (enabled
by Rohm and Haas Co.). We are grateful to Dr K. Oertle (at
formerly Ciba-Geigy AG) for a gift of chlorodimethyl(1,1,2-
trimethylpropyl)silane. The Center for New Directions in
Organic Synthesis is supported by Bristol-Myers Squibb as
sponsoring member.
The calculated (HF/6-31G*)3b energies (kcal mol21) within
the series of [4]phenylene isomers decrease with the number of
angular fusions in the order linear (relative energy +11.0) > 1
(+5.9) > 4 (+4.5) = zigzag (+4.3) > C3-symmetric [4]phenyl-
1
ene (0). Comparison of the H NMR data1a,d reveals that this
trend is (roughly) paralleled by net increased deshielding of all
the hydrogens (daverage
= 6.33, 6.58, 6.70, 6.66, 7.19,
respectively), as expected on the basis of GIAO and NICS
calculations3b which predict overall decreasing paratropic and
correspondingly increasing diatropic character of the cyclobuta-
diene and benzene rings, respectively. Interestingly, in as much
as these trends may be reflected in increasing HOMO–LUMO
gaps along the series and, in turn, in the electronic spectra, 1
displays a lowest energy lmax at 486 nm, almost identical to that
of the topomeric linear framework (lmax at 488 nm), whereas
the other isomers show relative hypsochromic shifts.1a,d It
appears that the presence of linear substructures has a strong
effect on the phenylene chromophore, as observed previously in
the branched series.1b
Notes and references
1 (a) D. L. Mohler and K. P. C. Vollhardt, in Advances in Strain in
Organic Chemistry, Vol. 1, ed. B. Halton, JAI, London, 1996, p. 121; (b)
R. Boese, A. J. Matzger, D. L. Mohler and K. P. C. Vollhardt, Angew.
Chem., Int. Ed. Engl., 1995, 34, 1478; (c) C. Eickmeier, H. Junga, A. J.
Matzger, F. Scherhag, M. Shim and K. P. C. Vollhardt, Angew. Chem.,
Int. Ed. Engl., 1997, 36, 2103; (d) C. Eickmeier, D. Holmes, H. Junga,
A. J. Matzger, F. Scherhag, M. Shim and K. P. C. Vollhardt, Angew.
Chem., Int. Ed., 1999, 38, 800.
2 I. Gutman, S. J. Cyvin and J. Brunvoll, Monatsh. Chem., 1994, 125,
887.
3 (a) J. M. Schulman, R. L. Disch, H. Jiao and P. von R. Schleyer, J. Phys.
Chem. A, 1998, 102, 8051; (b) J. M. Schulman and R. L. Disch, J. Phys.
Chem. A, 1997, 101, 5596; (c) J. M. Schulman and R. L. Disch, J. Am.
Chem. Soc., 1996, 118, 8470.
The crystal structure of 9 reveals the distinctive patterns of
bond length and angle distortions observed for the substructures
2 and 3 (identical within the range of standard deviations),5
particularly for rings C and D. Not surprisingly,12 the subtle
effects of the added ring fusion on the remote benzene rings are
not clearly evident. Quite noticeable, however, is the curvature
of 9,5a rendering the molecule chiral. The dihedral angles
between the planes of fused rings range from 1.32–6.38°, and
the displacements of other ring carbons from the mean plane
defined by those of ring D range from 0.07 to 1.70 Å.
4 (a) A. Rajca, A. Safronov, S. Rajca, C. R. Ross II and J. J. Stezowski,
J. Am. Chem. Soc., 1996, 118, 7272; (b) M. Baumgarten, F. Dietz, K.
Müllen, S. Karabunarliev and N. Tyutyulkov, Chem. Phys. Lett., 1994,
On the basis of the respective ease of all-cis-hexahydrogen-
ation of 2 (H2, 1 atm, Pd/C)7 and 3 (H2, 15 atm, Pd/C),1a the
central ring in the former may be regarded to be more activated
than that in the latter. In 1, ring B was expected to be relatively
stabilized, ring C destabilized, the relative extent of which was
tested on 9. Thus, 9 underwent smooth hydrogenation (Scheme
2) under mild conditions to furnish 10 completely regiose-
lectively,8 indicating a complete reversal in relative reactivity of
the inside six-membered rings. Proof of the structure of 10 rests
˘
221, 71; (c) N. Trinajstic, T. G. Schmalz, T. P. Zivkovic´, S. Nikolic´, G.
E. Hite, D. J. Klein and W. A. Seitz, New J. Chem., 1991, 15, 27; (d) J.
L. Brédas and R. H. Baughman, J. Polym. Sci., Polym. Lett. Ed., 1983,
21, 475.
5 (a) D. Holmes, S. Kumaraswamy, A. J. Matzger and K. P. C. Vollhardt,
Chem. Eur. J., 1999, 5, 3399; (b) A. Schleifenbaum, N. Feeder and K.
P. C. Vollhardt, Tetrahedron Lett., 2001, 42, 7329.
6 H.-D. Beckhaus, R. Faust, A. J. Matzger, D. L. Mohler, D. W. Rogers,
C. Rüchardt, A. K. Sawhney, S. P. Verevkin, K. P. C. Vollhardt and S.
Wolff, J. Am. Chem. Soc., 2000, 122, 7819.
7 B. C. Berris, G. H. Hovakeemian, Y.-H. Lai, H. Mestdagh and K. P. C.
Vollhardt, J. Am. Chem. Soc., 1985, 107, 5670.
1
on spectral data, especially low temperature H NMR spectra
8 For spectral and analytical data of 1, 9, and 10, see ESI†.
9 An alternative construction of 8, starting from 1,2,4,5-tetrakis(trime-
thylsilylethynyl)benzene (ref. 7), and its desymmetrization by mono-
desilylation (CH3Li), or from 1,2,4,5-tetraethynylbenzene and re-
spective
single
Pd-catalyzed
couplings
to
o-iodo(trimethylsilylethynyl)benzene proceeded in only statistical
fashion.
10 Crystal data. C30H28Si2, M 444.72, monoclinic, a = 9.5453 (2), b =
22.1287 (3), c = 12.0525 (1) Å, b = 103.155 (1)°, V = 2478.99 (9) Å3,
T = 147 K, space group P21 (#4), Z = 4, m (Mo-Ka) 1.58 cm21, 10629
reflections measured, 7220 unique reflections [I > 3.00s (I), Rint
=
0.027] were used in refinement. The final wR(F) was 0.040 (all data).
CCDC 173172. See http://www.rsc.org/suppdata/cc/b1/b109789j/ for
crystallographic files in .cif or other electronic format.
11 (a) J. A. N. F. Gomes and R. B. Mallion, Chem. Rev., 2001, 101, 1349;
(b) P. Lazzeretti, in Progress in Nuclear Magnetic Resonance
Spectroscopy, Vol. 36, ed. J. W. Emsley, J. Feeney and L. H. Sutcliffe,
Elsevier, Amsterdam, 2000, p. 1.
12 (a) A. R. Katritzky, K. Jug and D. C. Oniciu, Chem. Rev., 2001, 101,
1421; (b) T. M. Krygowski and M. K. Cyranski, Chem. Rev., 2001,
1001, 1385.
13 This observation is analogous to that made for hexahydro-3: R. Diercks
and K. P. C. Vollhardt, Angew. Chem., Int. Ed. Engl., 1986, 25, 266.
Fig. 2 Structure of 9 in the crystal: views from above (top) and the side
(bottom). For selected distances (Å) and angles (°), see ESI†
CHEM. COMMUN., 2002, 278–279
279