Activated phenacenes from phenylenes by nickel-catalyzed alkyne cycloadditions

Article abstract of DOI:10.1002/anie.201103428

Going zigzag, helical, or in-between: Angular phenylenes show a propensity for adding alkynes from the bay region in processes catalyzed by [Ni(cod)(PMe₃)₂] (see scheme). These transformations generate novel strain- and electronicall

Full text of DOI:10.1002/anie.201103428

DOI: 10.1002/anie.201103428
Polycyclic Aromatics
Activated Phenacenes from Phenylenes by Nickel-Catalyzed Alkyne
Cycloadditions**
Zhenhua Gu, Gregory B. Boursalian, Vincent Gandon, Robin Padilla, Hao Shen,
Tatiana V. Timofeeva, Paul Tongwa, K. Peter C. Vollhardt,* and Andrey A. Yakovenko
Polycyclic aromatic hydrocarbons have emerged as key
components of future (opto)electronic and other nanodevi-
ces.[1] Functional constraints, such as air-sensitivity, instability,
unfavorable band-gap, insolubility, and encumbered process-
ability require tunable synthetic approaches to specific
targets.[2] We report the Ni-catalyzed cycloaddition of alkynes
to the angular phenylene motif,[3,4] which engenders novel
sterically and electronically activated extended phenacenes[5]
with extensive selectivity. The potential of phenacenes as
components for light-emitting diodes, field-effect transistors,
and superconductors has been discovered only recently.[6]
In principle, the angular phenylene frame could undergo
attack by alkynes at either the bay (full arrows) or the non-
bay region (open arrows), which, if complete and regiochemi-
cally pristine, would furnish only one of the two extremes,
phenacenes or helicenes, respectively (Figure 1). In the
absence of such selectivity, the number of possible products
is substantial: 5 for parent system 1, 17 for angular
[4]phenylene 2, and 6 for C3-symmetric [4]phenylene 3. We
report a more optimistic experimental picture that is asso-
ciated with an unexpected bifurcation in mechanism, as
pinpointed by DFT computations.
The results of a comparative study of the [Ni(cod)-
(PMe3)2]-catalyzed cycloaddition (cod = 1,5-cyclooctadiene)
of diphenylacetylene (dpa) to equimolar amounts of 1, 2, and
3, respectively, under identical reaction conditions are shown
in Table 1.[7] Inspection of the product structures reveals
remarkable selectivity toward the formation of phenacene
(sub)units derived from multiple insertions, even though only
one equivalent of alkyne reagent is present. Consequently,
varying amounts of starting phenylenes are recovered, but the
mass balances are good to excellent.
Each one of the ensuing topologies (six of which were
detailed by X-ray analysis; Table 1)[7] exhibits unique features
as a result of p-activation through benzocyclobutadienofusion
and/or extreme s-distortion from planarity caused by the
crowded 4,5-diphenylphenanthrene[8] substructures. Thus,
yellow 4 presents the unknown benzo[3,4]cyclobuta[1,2-a]-
phenanthrene connectivity.[9] Its phenyl groups are rotated
extensively relative to the attached p-core (as also seen for
the other structures), the center of one of which (at C11) is
located directly above H10 of the biphenylene fragment
(distance 2.496 ꢀ), causing extraordinary shielding of this
nucleus (d = 4.01 ppm!). This phenomenon is also observed
for 6, 7, and 9.[7,10] The activation of the phenanthrene nucleus
is structurally evident in the change in the sequence of bond
alternation in the annelated terminal ring[7] and in the
electronic spectrum [e.g., highest lmax = 420 nm; cf. phenan-
threne: 345 nm, or the “hexagonal squeeze”[11] [4]phenacene
(chrysene): 360 nm].[12] These effects are even more pro-
nounced in the [a,i] and [a,c] doubly fused and topologically
new[13] red phenanthrenes (air sensitive) 6 [lmax = 484 nm; cf.
[5]phenacene (picene): 376 nm)[12] and 9 (lmax = 503 nm; cf.
benzo[g]chrysene: 371 nm).[7,14] Turning to the [5]phenacene
structures, colorless derivative 5 (lmax = 405 nm) suffers the
consequences of severe twisting caused by the 5,6,7,8-
tetraphenyl substitution, most dramatically illustrated by the
Figure 1. The two extremes of angular phenylene reactivity in cyclo-
additions to alkynes.
[*] Dr. Z. Gu, G. B. Boursalian, Dr. R. Padilla, Dr. H. Shen,
Prof. Dr. K. P. C. Vollhardt
Department of Chemistry, University of California at Berkeley
Berkeley, CA 94720-1460 (USA)
E-mail: kpcv@berkeley.edu
Prof. Dr. V. Gandon
ICMMO, UMR CNRS 8182, Univ Paris-Sud 11
91405 Orsay cedex (France)
Prof. T. V. Timofeeva, P. Tongwa, A. A. Yakovenko
Department of Natural Sciences, New Mexico Highlands University
Las Vegas, NM 87701 (USA)
[**] This work was supported by the NSF (CHE-0907800; K.P.C.V.; STC
MDITR, DMR-0120967, PREM DMR-0934212; T.V.T.), and CRIHAN
(project 2006-013; V.G.). We acknowledge illuminating comments
from Prof. R. G. Bergman.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201103428.
Angew. Chem. Int. Ed. 2011, 50, 9413 –9417
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9413

Communications
Table 1: Phenylated phenacenes by the Ni-catalyzed cycloaddition of dpa
dihedral angle of the bonds to the attached 6,7-phenyls of
76.28, and bond elongation (relative to [5]phenacene] along
the path C4a–C8a. As a consequence, the molecule is chiral
and configurationally rigid, as evidenced by the NMR spectra,
which reveal dissymmetry of both aryl groups at ꢀ608C.[7]
Those at C5/8 exhibited coalescence by rotation on warming
to 1–3.[a]
Phenylene
Products[b]
Yield
[%][c]
¼
to room temperature, DG287 = (13.0ꢁ0.1) kcalmolꢀ1. The
ortho and meta CH signals of the bay phenyls remained
anisochronous up to 708C, when slight line broadening
occurred, suggesting a barrier to rotation and/or enantiomer-
ization[15] of > 20 kcalmolꢀ1. An attempt to model a meso
diastereomer in silico led to a highly distorted array,
54 kcalmolꢀ1 more energetic than 5.[7] Phenylenoactivation
is again evident in the air-sensitive red 7 [lmax = 456 nm; cf.
[6]phenacene (fulminene): 384 nm][16] and 10 [lmax = 510 nm
(sh); cf. benzo[s][5]phenacene: 350 nm).[17] Both show encum-
brance of the phenyl groups in the NMR experiment similar
to that of 5, quantified for the relatively faster phenyl rotation
1[d]
4
20
5
6
59
33
¼
at C5 and C8 in 10: DG287 = (13.2ꢁ0.1) kcalmolꢀ1.[7] The
[5]phenacene warping is similar to that in 5. As in 6, the
atypically[3] short phenacene bond of benzofusion C16b C12c
ꢀ
in 10 (1.379 ꢀ)[7] heralds increased benzocyclobutadienoid
character and hence reactivity toward oxygen. Such could be
verified by deliberate oxygenolysis of this bond to the
corresponding ring-opened dione (72%).[7,18] Finally, color-
less [7]phenacene 8 [lmax = 427 nm; p-band 388 nm; cf.
[7]phenacene: p-band 344 nm)[19] is even more twisted than
5, with averaged dihedral angles C6(8)-C6a(8a)-C6b(8b)-
C7(9) (bay carbons) = 35.48, Ph-C6(8)-C7(9)-Ph = 88.28, and
the associated deplanarization of the internal [5]phenacene
unit (dihedral angles in the individual benzene rings up to
268). It adopts approximate meso symmetry, the correspond-
ing chiral isomer computed to lie 4.1 kcalmolꢀ1 higher in
energy.[7] Configurational rigidity is manifest in the NMR
spectra (not quantified), with locked central phenyl groups at
room temperature, whereas the outside phenyl signals begin
to decoalesce < 08C.[7]
2[e]
7
8
27
28
A simple (if counterintuitive) rationale for the dispropor-
tionate generation of multiple adducts would be increasing
reactivity of the phenylene frame with successive cycloaddi-
tions. However, subjection of 4 and 9, respectively, to renewed
reaction not only proceeded about an order of magnitude
more slowly, but the former gave colorless isomer 11
3[f]
9
33
(calculated to be more stable by 4.4 kcalmolꢀ1; lmax
=
406 nm; cf. benzo[c]chrysene: 386 nm)[14] instead of 5, the
latter the air-sensitive, red 12 (lmax = 496 nm) and not 10 as
the main products.[7] There must be alternative pathways to
these respective assemblies.
10
66
ꢂ
[a] Ph-C C-Ph (1 equiv), [Ni(cod)(PMe3)2] (10 mol%), THF, 758C, 23 h
(1), 5 h (2), 48 h (3). [b] For fully characterized minor other isomers, see
reference [7]. [c] Based on the alkyne as the limiting reagent. [d] 10%
recovered. [e] 37% recovered. [f]<1% recovered.
bay region twist C6-C6a-C6b-C7 = 31.18, the associated
deplanarization of the internal phenanthrene unit (dihedral
angles in the individual benzene rings up to 238), the extreme
9414
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9413 –9417

Figure 2. Computed relative free energies, DG298 (kcalmolꢀ1) for the conversion of 1 with [Ni(dpa)(PMe3)2] and dpa (0 kcalmolꢀ1). Free energies
of activation are indicated in italics.
A series of semi-quantitative experiments, focused on 1
and its products and monitored by 1H NMR spectroscopy, was
carried out to shed some light on this mechanistic problem.
Thus, to rule out a competitive dinuclear catalytic species,[4e]
the catalyst concentration was varied from 7 to 35 to 50 mol%
with 0.003 molar 1 and dpa in THF. The rate of disappearance
of 1 increased proportionally. Surprisingly, the initial ratios of
4:5 with different nickel catalyst loadings were identical and
high (10:1), and constant up to 20% conversion. On further
reaction, the concentration of 4 leveled, whereas that of 5
increased to furnish the eventually isolated proportion of 2:3
(Table 1). No intermediates were detectable by 1H NMR
spectroscopy. That this change in relative rates was due to the
amount of dpa present in the mixture was ascertained by
varying the latter: higher starting concentrations of dpa
slowed the overall reaction, but the rate of formation of 5
more than that of 4 (e.g., 50 mol% catalyst, 4 equiv dpa,
408C; 4:5 = 20:1; to 20% conversion). This effect of added
dpa was reproduced by other ligands, such as PMe3 or cod.
Conversely, the production of 5 maximized when keeping the
level of dpa low. This finding could be exploited preparatively
by applying high dilution conditions (10 mol% catalyst,
syringe pump addition of 2 equiv dpa, boiling THF, 13 h)
converting 1 quantitatively (NMR) to 5 (87%) and 4 (13%).[7]
Turning to the behavior of 4, an analogous change in
product ratios 11:5 with alkyne concentration was recorded:
2:1 with 1 equivalent, 13:1 with 3 equivalents. Similarly,
following the conversion of 4 with 1 equivalent of alkyne by
1H NMR spectroscopy (50 mol% catalyst, 608C, THF)
revealed an initial ratio of 5:1 (0.7 h), which changed
gradually to the final number of 2:1 (72 h), in the absence
of any other species. Finally, 1 appears inert to the catalyst
(1:1) (20–608C) on its own,[21] but addition of the alkyne
(1 equiv) caused immediate quantitative formation of [Ni-
(PMe3)2(dpa)][20] (and cod), cycloaddition commencing slowly
only on warming to 408C.
Armed with these experimental data, further mechanistic
input was sought by DFT calculations.[7] This task was
complicated by the appearance of several energetically
close-lying Ni species in the starting mixture, the following
transition state (TS) for the first insertion, and the resulting
products.[7] Gratifyingly, however, [Ni(PMe3)2(dpa)] emerged
as the lowest-energy initial species (Figure 2; 0 kcalmolꢀ1), in
consonance with observation. Similarly, TS B was computed
as the most favorable structure for the ring opening step, and
C for its product, replicating the experimental selectivity for
bay region attack. The regioselectivity of this step may be due
to relative stabilization of the polarized Ni-C(a-biphenylene)
bond in C by the electron-withdrawing neighboring cyclo-
butadienoid ring,[3] a notion also supported by the finding that
in the reaction of 2 (50 mol% catalyst, 1 equiv dpa, 408C) the
first new species is 6 exclusively, before the gradual appear-
ance of 7 and then 8. The precursor nickelacyclopentadiene
would enjoy double adjacent strained ring stabilization. From
C, dpa adds and inserts regioselectively to eventually render
ligated 4 in the form of D. We postulate this species to be the
crucial bifurcating point from which, possibly favored by
excess dpa (or other ligand), 4 escapes in an exothermic, but
reversible, step. Alternatively, at low dpa loading, the
attached metal moiety can undergo rapid haptotropism,[22]
thus positioning itself for the second insertion from E. Again
in accord with experiment, renewed bay region addition is
favored kinetically (F), as well as thermodynamically, enjoy-
ing anchimeric assistance of the neighboring phenyl group in
producing G,[22] which cascades on to 5.[7]
To rationalize the finding that 4 proceeds preferentially to
11 at high dpa concentrations, we propose a different Ni
species, e.g., H (Figure 3), for which computation suggests a
Angew. Chem. Int. Ed. 2011, 50, 9413 –9417
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9415

Communications
two different catalysts, whose viability appears alkyne con-
centration dependent. It is reasonable to assume that a similar
mechanistic bifurcation characterizes the transformations of 2
and 3. Our results suggest that with variation of catalyst,[23]
substrate, and reaction conditions extensive, if not complete,
designed control of the outcome of these processes might be
attainable.
Received: May 18, 2011
Published online: August 26, 2011
Keywords: aromatic compounds · cycloaddition · nickel ·
.
phenacenes · reaction mechanisms
Figure 3. A computed route to 11 through Ni species H.
[1] For selected monographs, see: a) M. Sharon, M. Sharon, Carbon
Nanoforms and Applications, McGraw-Hill, New York, 2010;
b) Carbon-Rich Compounds: From Molecules to Materials
(Eds.: M. M. Haley, R. R. Tykwinski), Wiley-VCH, Weinheim,
2006; c) R. G. Harvey, Polycyclic Aromatic Hydrocarbons,
Wiley-VCH, New York, 1997.
sterically prohibited bay region attack by dpa, but a kineti-
cally feasible route to product. H might arise directly from 4
by [Ni(PMe3)(dpa)] ligation or by interception of the DQE
manifold.
A brief excursion was undertaken to probe the generality
of this chemistry by variation of the alkyne (Scheme 1). The
selectivity of addition was less pronounced than that occur-
ring with dpa, notably some non-bay first addition product 14,
albeit minor, now detectable.
[2] For recent reviews, see: a) K. B. Jørgensen, Molecules 2010, 15,
4334; b) X. L. Feng, W. Pisula, K. Mꢁllen, Pure Appl. Chem.
2009, 81, 2203; c) R. Sanz, Org. Prep. Proced. Int. 2008, 40, 215;
d) J. S. Wu, Curr. Org. Chem. 2007, 11, 1220; e) R. G. Harvey,
Curr. Org. Chem. 2004, 8, 303.
ˇ
´
[3] For a review of the [N]phenylenes, see: O. S. Miljanic, K. P. C.
Vollhardt in Carbon-Rich Compounds: From Molecules to
Materials (Eds.: M. M. Haley, R. R. Tykwinski), Wiley-VCH,
Weinheim, 2006, pp. 140 – 197.
[4] For metal-catalyzed additions of alkynes to biphenylene, see,
inter alia: a) T. Shibata, G. Nishizawa, K. Endo, Synlett 2008,
765; b) T. Schaub, M. Backes, U. Radius, Organometallics 2006,
25, 4196; c) T. Schaub, U. Radius, Chem. Eur. J. 2005, 11, 5024;
d) C. Mꢁller, R. J. Lachicotte, W. D. Jones, Organometallics
2002, 21, 1975; e) C. Perthuisot, B. L. Edelbach, D. L. Zubris, N.
Simhai, C. N. Iverson, C. Mꢁller, T. Satoh, W. D. Jones, J. Mol.
Catal. A 2002, 189, 157, and references therein.
[5] For complementary approaches to phenacenes, by metal-cata-
lyzed intramolecular hydroarylation of alkynes or stilbene
photocyclization, see: a) T.-A. Chen, T.-J. Lee, M.-Y. Lin,
S. M. A. Sohel, E. W.-G. Diau, S.-F. Lush, R.-S. Liu, Chem.
Eur. J. 2010, 16, 1826; b) F. B. Mallory, K. E. Butler, A. Bꢂrubꢂ,
E. D. Luzik, Jr., C. W. Mallory, E. J. Brondyke, R. Hiremath, P.
Ngo, P. J. Caroll, Tetrahedron 2001, 57, 3715, and references
therein.
[6] a) Y. Sugawara, Y. Kaji, K. Ogawa, R. Eguchi, S. Oikawa, H.
Gohda, A. Fujiwara, Y. Kubozono, Appl. Phys. Lett. 2011, 98,
013303; b) R. Mitsuhashi, Y. Suzuki, Y. Yamanari, H. Mitamura,
T. Kambe, N. Ikeda, H. Okamoto, A. Fujiwara, M. Yamaji, N.
Kawasaki, Y. Maniwa, Y. Kubozono, Nature 2010, 464, 76; c) X.
Lee, N. Kawasaki, M. Suekiyo, Y. Sugawara, A. Fujiwara, Y.
Kubozono, Org. Electron. 2010, 11, 2010; d) Y. Kaji, N.
Kawasaki, X. Lee, H. Okamoto, Y. Sugawara, S. Oikawa, A.
Ito, H. Okazaki, T. Yokoya, A. Fujiwara, Y. Kubozono, Appl.
Phys. Lett. 2009, 95, 183302; e) H. Okamoto, N. Kawasaki, Y.
Kaji, Y. Kubozono, A. Fujiwara, M. Yamaji, J. Am. Chem. Soc.
2008, 130, 10470; f) A. S. Ionkin, W. J. Marshall, B. M. Fish, L. M.
Bryman, Y. Wang, Chem. Commun. 2008, 2319.
Scheme 1. Ni-catalyzed cycloadditions of 3-hexyne and 1,4-dimethoxy-
2-butyne to 1.
Moreover, further reaction to 15 and 16 was much more
sluggish. It appears that reactivity and selectivity is quite
dependent on the structure of the alkyne and that the
presence of a potentially ligating substituent (i.e. MeO in 1,4-
dimethoxybutyne) is beneficial in these respects, as reflected
by G in Figure 2.
In summary, we have shown that the embedded strain in
the phenylenes can be exploited in multiple cycloadditions of
alkynes catalyzed by Ni. These reactions have the potential to
be highly regioselective and tunable, as demonstrated in the
case of dpa as the substrate. For this alkyne and 1, DFT
calculations in conjunction with experiments provide a
plausible mechanistic rationale, pointing to the operation of
[7] For details, see the Supporting Information. CCDC 826179,
826180, 826181, 826182, 826183, 826184, and 826185 for com-
plexes 4, 5, 6, 8, 9, 10, and 11 contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
9416 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9413 –9417

[8] A. H. A. Tinnemans, W. H. Laarhoven, J. Am. Chem. Soc. 1974,
96, 4617.
[9] For the isomeric [b] fused topology, see: P. R. Buckland, N. P.
Hacker, J. F. W. McOmie, J. Chem. Soc. Perkin Trans. 1 1983,
1443.
[10] For similar effects, see, inter alia: a) Y. Ting, Y.-H. Lai, J. Am.
Chem. Soc. 2004, 126, 909, and references therein; b) M.
Deichmann, C. Nꢃther, R. Herges, Org. Lett. 2003, 5, 1269.
[11] I. Gutman, A. R. Ashrafi, MATCH 2008, 60, 135.
[12] W. V. Mayneord, E. M. F. Roe, Proc. R. Soc. London Ser. A 1935,
152, 299.
[13] For the [b,h] and [c,g] motif, see: M. E. Cracknell, R. A. Kabli,
J. F. W. McOmie, D. H. Perry, J. Chem. Soc. Perkin Trans. 1 1985,
115.
[17] X.-Q. Tang, R. G. Harvey, J. Org. Chem. 1995, 60, 3568.
[18] Compare: H. Mestdagh, K. P. C. Vollhardt, J. Chem. Soc. Chem.
Commun. 1986, 281, and references therein.
[19] F. B. Mallory, K. E. Butler, A. C. Evans, E. J. Brondyke, C. W.
Mallory, C. Yang, A. Ellenstein, J. Am. Chem. Soc. 1997, 119,
2119.
[20] K. R. Pçrschke, R. Mynott, K. Angermund, C. Krꢁger, Z.
Naturforsch. B 1985, 40, 199.
[21] Biphenylene(s) dimerize to the corresponding tetrabenzocy-
clooctatetraene(s) under these conditions: H. Schwager, S.
Spyroudis, K. P. C. Vollhardt, J. Organomet. Chem. 1990, 382,
191.
[22] See: T. Li, J. J. Garcꢄa, W. W. Brennessel, W. D. Jones, Organo-
metallics 2010, 29, 2430, and references therein.
[14] E. Clar, D. G. Stewart, J. Am. Chem. Soc. 1952, 74, 6235.
[15] See, e.g.: R. A. Pascal, Jr., Q. Qin, Tetrahedron 2008, 64, 8630.
[16] K. F. Lang, H. Buffleb, J. Kalowski, Chem. Ber. 1964, 97, 494.
[23] For a recent pertinent study, see: A. B. Chaplin, R. Tonner, A. S.
Weller, Organometallics 2010, 29, 2710, and references therein.
Angew. Chem. Int. Ed. 2011, 50, 9413 –9417
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9417