Diels-Alder reactivity of polycyclic aromatic hydrocarbon bay regions: Implications for metal-free growth of single-chirality carbon nanotubes

Article abstract of DOI:10.1021/ja907802g

(Chemical Equation Presented) A soluble bisanthene derivative, 4,11-dimesitylbisanthene, has been synthesized in three steps from bianthrone. In hot toluene, this bisanthene undergoes a clean Diels-Alder reaction with diethyl acetylenedicarboxylate to giv

Full text of DOI:10.1021/ja907802g

Published on Web 10/15/2009
Diels-Alder Reactivity of Polycyclic Aromatic Hydrocarbon Bay Regions:
Implications for Metal-Free Growth of Single-Chirality Carbon Nanotubes
Eric H. Fort, Patrick M. Donovan, and Lawrence T. Scott*
Merkert Chemistry Center, Department of Chemistry, Boston College, Chestnut Hill, Massachusetts 02467-3860
Received September 14, 2009; E-mail: lawrence.scott@bc.edu
Single-chirality carbon nanotubes (CNTs) of predefined diameter
and [M,N] index will someday be accessible by rational chemical
synthesis “from the ground up,” in the same way that Lipitor, Nylon,
and countless other man-made substances are routinely synthesized
today by organic chemists. In principle, one might achieve this
objective by “growing” a nanotube using well understood chemical
reactions, starting from a suitably constituted small hydrocarbon
template.1 The Diels-Alder addition of acetylene to a [5,5]nanotube
end-cap (e.g., 1), for example, would transform the “bay regions”
around the armchair rim of the molecule into new benzene rings,
and successive Diels-Alder reactions could continue indefinitely,
because the molecule can never run out of bay regions (Scheme
1). The highly exothermic aromatization of each newly formed six-
membered ring by loss of two hydrogens would be expected to
occur spontaneously, even at moderate temperatures and in the
absence of oxidizing agents.2
reluctance of PAHs to engage in bay region Diels-Alder cycload-
ditions. That cost can be approximated as the difference in the
aromatic stabilization energies (∆ASE)5 between the starting material
and the initially formed cycloadduct. In the example illustrated
(Scheme 2), the initially formed cycloadduct (3) contains a residual
phenanthrene moiety, the ASE of which is quite large but falls
below that of the starting perylene (2). Phenanthrene itself has never
been seen to undergo a bay region Diels-Alder reaction, presum-
ably because the large ASE of the molecule would drop down to
zero in the initially formed cycloadduct.
It occurred to us that the smaller ∆ASE of the perylene
Diels-Alder reaction (Scheme 2), relative to that of the imaginary
phenanthrene Diels-Alder reaction, might point to even smaller
∆ASE values for bay region Diels-Alder reactions of longer PAHs
that begin to resemble strips from the side walls of armchair CNTs.
Simply counting the resonance structures6 of progressively longer
PAHs and the cycloadducts that they could give in this series lends
support to this intuitively reasonable hypothesis. Hu¨ckel molecular
orbital and semiempirical (AM1) MO calculations on the same
starting materials and cycloadducts reinforce this conclusion, but
such theoretical treatments give only crude estimates of activation
energies (Hammond postulate). We have therefore calculated the
transition state energies for cycloadditions of acetylene to the bay
regions of PAHs in the “periacene” series7 phenanthrene f perylene
f bisanthene f tetrabenzocoronene. Figure 1 shows the activation
energies obtained by subtracting the calculated energies of the
starting materials from the calculated energies of the transition states
(B3LYP/6-31G*). Higher levels of theory would be expected to
yield quantitatively more reliable estimates of the activation
energies, but our hypothesis is clearly validated by theory.
Scheme 1. Proposed Application of the Diels-Alder Reaction to a
Chemical Synthesis of Carbon Single-Walled Nanotubes (SWNTs)
Disregarding, for now, the fact that acetylene is a feeble
dienophile in Diels-Alder reactions, even with relatively reactive
dienes,3 most organic chemists would dismiss this strategy as
unrealistic, because the bay regions of polycyclic aromatic hydro-
carbons (PAHs) are notoriously resistant to Diels-Alder addition
reactions, even when confronted with potent dienophiles. To coerce
perylene (2) to react with maleic anhydride, for example, Clar and
Zander resorted to using molten maleic anhydride as the solvent at
reflux (202 °C), with chloranil added to aromatize the new ring
(Scheme 2).4 Notwithstanding the perceived obstacles to imple-
menting the idealized proposal in Scheme 1, we contend that such
a strategy may still be viable.
Scheme 2. A Diels-Alder Addition to the Bay Region of Perylene
Figure 1. Activation energies calculated for Diels-Alder additions of
acetylene to aromatic hydrocarbon bay regions (B3LYP/6-31G*).
To test this hypothesis experimentally, we synthesized 7,14-
dimesitylbisanthene (4), a simple derivative of the third PAH in
this series.8 The methyl groups ortho to the biaryl bonds in 4 were
incorporated to force the appendages out-of-plane, thereby enhanc-
The energetic cost of temporarily and simultaneously disrupting
the aromaticity of two benzene rings undoubtedly accounts for the
9
16006 J. AM. CHEM. SOC. 2009, 131, 16006–16007
10.1021/ja907802g CCC: $40.75  2009 American Chemical Society

C O M M U N I C A T I O N S
ing solubility, and also to block any cycloadditions at the central
rings of the two anthracene subunits. Both objectives were achieved,
and we were pleased to confirm the relatively easy transformation
of 4 into the Diels-Alder cycloaddition products 5 and 6 when 4
is heated with diethyl acetylenedicarboxylate in toluene for 24 h at
120 °C (all starting material consumed, Figure 2a). The second
cycloaddition occurs somewhat more slowly than the first9 but can
easily be pushed to completion. For comparison, perylene (2) was
heated with diethyl acetylenedicarboxylate at the same concentration
in toluene. In this case, the cycloaddition is clearly more difficult
(<50% conversion, even after 72 h at 150 °C). The 1:1 cycloadduct
(7) is formed, but we see no evidence for the 2:1 adduct (8) (Figure
2b).10
hand (e.g., [10,10]nanotubes), that impediment will be significantly
reduced. Our laboratory is currently working on synthetic ap-
proaches to nanotube end-caps of several different diameters.1b,13
Finally, we note that the Diels-Alder cycloaddition strategy for
growing CNTs is not limited to armchair nanotubes. It will not
work on [N,0]nanotubes (“zig-zag”), which have no bay regions
where a dienophile could add, but it should work on chiral
nanotubes that have at least one zig-zag edge that is interrupted by
one or more bay region steps. In such nanotubes, transformation
of a bay region into a new six-membered ring creates a new bay
region, and the process can be repeated indefinitely (see SI).
Nanotube edges that contain three quaternary carbon atoms in a
row (so-called “cove regions”), however, are not amenable to
growth solely by Diels-Alder cycloaddition chemistry.
The ultimate test of this strategy for metal-free growth of single-
chirality CNTs must await the availability of suitable small
hydrocarbon templates, many of which are being vigorously sought
by synthetic organic chemists worldwide. “Masked acetylenes” that
perform better as dienophiles than molecular acetylene would also
be useful.
Acknowledgment. We thank the National Science Foundation
for financial support of this research and for funds to purchase mass
spectrometers.
Supporting Information Available: Experimental procedures for
the synthesis of 4, Diels-Alder cycloadditions of 2 and 4, and
competition experiment. Spectroscopic data for all new compounds and
the competition experiment. Calculation details for the first and second
Diels-Alder cycloadditions to the hydrocarbons in Figure 1 and XYZ
coordinates of all species. Figures for a [N]cycloparaphenylene,
[N]cyclophenacene, longer armchair template, and chiral template. This
material is available free of charge via the Internet at http://pubs.acs.org.
Figure 2. Diels-Alder additions of diethyl acetylenedicarboxylate to (a)
4,11-dimesitylbisanthene (4) and (b) perylene (2).
As a direct test of the prediction derived from the calculated
activation energies in Figure 1, a competition experiment was
conducted in which equal molar quantities of 2 and 4 were heated
with diethyl acetylenedicarboxylate in toluene for 22 h at 100 °C.
Under these conditions, only the longer periacene (4) is converted
to its 1:1 Diels-Alder adduct (5), whereas the shorter one (2)
survives totally unchanged (NMR spectra shown in the Supporting
Information (SI)).
These results proVide unambiguous experimental confirmation
for our hypothesis that Diels-Alder cycloadditions should become
progressiVely easier in bay regions at the ends of PAHs that
represent progressiVely longer strips of armchair nanotube side-
walls.
Our findings have implications for the proposed metal-free
growth of single-chirality CNTs by the strategy illustrated in
Scheme 1. First, it should be noted that this strategy can be applied
not only to nanotube end-cap templates but also to aromatic belts
that represent short sections of nanotubes, open at both ends. For
hydrogen-terminated armchair CNTs, Diels-Alder cycloadditions
will be very difficult if the template is short (e.g., an [N]cyclopar-
aphenylene11 or an [N]cyclophenacene) but should become easier
with longer templates12 and as the tube grows (see SI). Whether
or not acetylene itself can be coaxed to serve as the dienophile
remains to be determined.
References
(1) Steinberg, B. D.; Scott, L. T. Angew. Chem., Int. Ed. 2009, 48, 5400–
5402. (b) Hill, T. J.; Hughes, R. K.; Scott, L. T Tetrahedron 2008, 64,
11360–11369. (c) Herges, R. Nachrichten Chem. 2007, 55, 962–963, 966-
969.
(2) For thermal [4+2] cycloaddition and cycloreversion reactions of dihydrogen,
see: (a) Hayden, A. E.; Houk, K. N. J. Am. Chem. Soc. 2009, 131, 4084–
4089, and references cited therein. (b) Anet, F. A. L.; Levendecker, F.
J. Am. Chem. Soc. 1973, 95, 156–159. (c) Frey, H. M.; Krantz, A.; Stevens,
I. D. R. J. Chem. Soc. A 1969, 1734–1738.
(3) Walsh, R.; Wells, J. M. Int. J. Chem. Kinet. 1975, 7, 319–29.
(4) Clar, E.; Zander, M. J. Chem. Soc. 1957, 4616–4619.
(5) For leading references on aromatic stabilization energies, see: Wu, J. I.;
Dobrowolski, M. A.; Cyranski, M. K.; Merner, B. L.; Bodwell, G. J.; Mo,
Y.; Schleyer, P. v. R. Mol. Phys. 2009, 107, 1177–1186.
(6) Swinborne-Sheldrake, R.; Herndon, W. C.; Gutman, I. Tetrahedron Lett.
1975, 755–8.
(7) For recent calculations on the ground states of periacenes, see: Jiang, D.-
e.; Dai, S. Chem. Phys. Lett. 2008, 466, 72–75.
(8) Modeled after the synthesis of 4,11-diphenylbisanthene by Maulding, D. R.
J. Org. Chem. 1970, 35, 1221–1223. See SI.
(9) Calculations predict a slower second addition. See SI.
(10) Under more forcing conditions, 2:1 Diels-Alder adducts of perylene can
be prepared: (a) Alibert-Fouet, S.; Seguy, I.; Bobo, J.-F.; Destruel, P.; Bock,
H. Chem.sEur. J. 2007, 13, 1746–1753.
(11) (a) Jasti, R.; Bhattacharjee, J.; Neaton, J. B.; Bertozzi, C. R. J. Am. Chem.
Soc. 2008, 130, 17646–17647. (b) Hiroko, T.; Omachi, H.; Yamamoto,
Y.; Bouffard, J.; Itami, K. Angew. Chem., Int. Ed. 2009, 48, 6112–6116.
(12) Merner, B. L.; Dawe, L. N.; Bodwell, G. J. Angew. Chem., Int. Ed. 2009,
48, 5487–5491.
(13) (a) Steinberg, B. D.; Jackson, E. A.; Filatov, A. S.; Wakamiya, A.;
Petrukhina, M. A.; Scott, L. T. J. Am. Chem. Soc. 2009, 131, 10537–10545.
(b) Belanger, A. P.; Scott, L. T. Abstracts of Papers, 236th ACS National
Meeting, August, 2008, abstr ORGN-247. (c) Tsefrikas, V. M.; Rose, J. A.;
Silk, M. T.; Scott, L. T. Abstracts of Papers, 234th ACS National Meeting,
August 2007, abstr ORGN-797.
For small diameter nanotubes (e.g., [5,5]nanotubes), the addition
of each new six-membered ring will require the introduction of
further molecular strain. For larger diameter tubes, on the other
JA907802G
9
J. AM. CHEM. SOC. VOL. 131, NO. 44, 2009 16007