A Macrocyclic 1,4-Diketone Enables the Synthesis of a p-Phenylene Ring That Is More Strained than a Monomer Unit of [4]Cycloparaphenylene

Article abstract of DOI:10.1021/acs.orglett.6b01536

The synthesis of a p-terphenyl-based macrocycle, containing a p-phenylene unit with 42.6 kcal/mol of strain energy (SE), is reported. The conversion of a macrocyclic 1,4-diketone to a highly strained arene system takes place over five synthetic steps, fea

Full text of DOI:10.1021/acs.orglett.6b01536

Letter
pubs.acs.org/OrgLett
A Macrocyclic 1,4-Diketone Enables the Synthesis of a p‑Phenylene
Ring That Is More Strained than a Monomer Unit of
[4]Cycloparaphenylene
Nirmal K. Mitra, Hector H. Corzo, and Bradley L. Merner*
Department of Chemistry and Biochemistry, Auburn University, Auburn, Alabama 36849, United States
S
* Supporting Information
ABSTRACT: The synthesis of a p-terphenyl-based macrocycle, containing
a p-phenylene unit with 42.6 kcal/mol of strain energy (SE), is reported.
The conversion of a macrocyclic 1,4-diketone to a highly strained arene
system takes place over five synthetic steps, featuring iterative dehydrative
reactions in the aromatization protocol. Spectroscopic data of the deformed
benzenoid macrocycle are in excellent agreement with other homologues
that have been reported, indicating that the central p-phenylene ring of 9 is
aromatic.
he synthesis of increasingly strained aromatic systems has
workers demonstrated that [5]CPP was in fact benzenoid.7 The
Trequired increasingly high levels of synthetic innovation next smallest member of this class of carbon nanohoops is
predicted to have 144 kcal/mol of SE (36 kcal/mol per
benzene ring) contained within the macrocyclic structure.8 This
high degree of SE has led many to believe that [4]CPP (1) may
not be composed of benzen(oid)e rings and may possibly
prefer a quinoidal-type structure (2, Figure 1), which is in part
throughout the rich history of nonplanar arene synthesis. For
example, the synthesis of the most strained [n]paracyclophanes
featured the conversion of [n.2.2]propelladiene-based (Dewar
benzene) precursors to benzenoid systems. In the case of the
most strained [n]paracyclophane, [4]paracyclophane, strategi-
cally placed tetrasubstituted alkenyl groups were incorporated
in the bridging unit to prevent ipso-protonation, and
subsequent strain relief, from occurring at the para-carbon
atoms.1 An equally clever molecular design was employed by
Tsuji and co-workers in their synthesis of a [1.1]-
paracyclophane derivative, which still stands as the most
distorted para-phenylene ring to be characterized by X-ray
crystallography.2 For nonbenzenoid systems, such as the
3
fullerene allotrope of carbon C60 and a short [5,5] armchair
carbon nanotube (CNT) end-cap fragment,4 strategically
placed pentagons within polycyclic aromatic hydrocarbon
(PAH) frameworks and chlorine atoms at specific arene
vertices enabled Scott and co-workers to synthesize both of
these molecules. The recent emergence of the [n]cyclopara-
phenylenes ([n]CPPs) and their potential application in the
controlled bottom-up chemical synthesis of (n,n) armchair
CNTs has renewed interest in the development of synthetic
strategies for accessing highly strained benzenoid systems.5
While several groups have entered this field over the past 8
years, only two have reported on the synthesis of the most
strained CPP molecules.
Figure 1. [4]CPP: Benzenoid or quinoid?.
due to the known [4]CPP derivative “picotube” having been
shown to be quinoidal. Herges and co-workers reported the
synthesis of picotube (3) in 1996,10 which proceeded through a
photochemically induced ring expanding metathesis-based
strategy. The precursor hydrocarbon of picotube contained
2
2
Csp −Csp π-bonds bonds between adjacent “pre-arene”
subunits and, therefore, may not represent the best model for
1. A potentially better suited macrocyclic precursor to [4]CPP
2
2
In 2014, Yamago6 and Jasti7 individually reported the
synthesis of the smallest CPP, [5]CPP. This macrocyclic
compound, containing five para-linked benzene rings, is
predicted to have a strain energy (SE) of 119 kcal/mol (23.8
kcal/mol per benzene ring).8 Prior to its synthesis and
subsequent characterization by X-ray crystallography, the
benzenoid nature of its structure was called to question.9
Indeed, the solid state structure obtained by Jasti and co-
is one that contains the requisite Csp −Csp σ-bonds between
adjacent arene and prearene units. However, the synthesis of
such a macrocyclic precursor is not trivial (Figure 2).
The synthetic approaches used to construct the macrocyclic
precursors of [5] and [6]CPP involved a common intermediate
Received: May 27, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b01536
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Recently, we reported the synthesis of a homologous series
of p-terphenyl-based macrocyclic systems, containing highly
deformed p-phenylene units using a mild dehydrative
aromatization reaction that proceeded without rearrangement
of the central phenylene ring.13 Following this strategy,
dialdehyde 11 was synthesized using a known procedure.14
However, the best yield that could be obtained for this reaction,
in our hands, was 13%.15 Nonetheless, gram-scale quantities of
11 could be prepared.16 Applying a streamlined four-step
protocol that has been developed in our laboratory13 afforded
macrocyclic 1,4-diketone 10 in 15% overall yield. Similar to
what was observed during the synthesis of a homologous 15-
membered 1,4-diketo macrocycle, the ring-closing metathesis
(RCM) reaction of the intermediate diene gave a significant
amount of a higher molecular weight metathesis byproduct.13
Treatment of 10 with vinylmagnesium chloride gave the
corresponding syn-allylic diol (17, Scheme SI-1)17 as a single
diastereomer, which was directly subjected to a second RCM
reaction to afford macrocycle 12 in 59% yield over two steps.
Previously, direct aromatization of cyclohex-2-ene-1,4-diol
precursors, such as 12, was accomplished using a Burgess
reagent-mediated dehydrative aromatization reaction to afford
the central p-phenylene unit. In the case of the most strained
system that had been prepared using this strategy, 37.0 kcal/
mol of SE was introduced into the macrocyclic benzenoid
target upon elimination of two molecules of water, affording a
p-phenylene ring containing 28.4 kcal/mol of SE.13 Subjecting
diol 12 to the same reaction conditions gave the mono-
dehydration product 13 in 19% yield, with a trace amount
(<5%) of the desired 1,4-dioxa[4](3,3″)p-terphenylophane (9,
entry 1, Scheme 1a).18
Figure 2. (a) The smallest CPPs via a common precursor, (b)
Potential precursors of [4]CPP, (c) A macrocylic 1,4-diketone-based
approach to a p-phenylene ring containing 42.6 kcal/mol of SE.
Scheme 1. (a) Synthesis of Macrocycle 13, (b) Proposed
Mechanism for the Formation of 13 and 14
(Figure 2a).11 Directed arylation startegies, via Ni (Yamago)
2
and Pd-mediated (Jasti) reactions, led to the formation of Csp −
2
Csp bonds and macrocycle synthesis. Employing the same type
of macrocyclization strategy to a smaller precursor molecule
poses a significant synthetic challenge. The introduction of
boat-shaped, prearene subunits in the common precursor
(Figure 2a) and related systems is critical to the success of
macrocyclization reactions,5 as the syn-1,4-diol units provide the
necessary curvature or kink in the linear acyclic precursors to
bring the reacting arene vertices into proximity and facilitate
C−C bond formation. The sheer distance between the terminal
para-vertices of an acyclic precursor containing two arene and
two (boat-shaped) prearene units will be difficult to overcome,
as will competing intermolecular reactions between an
aryllithium or Grignard-type reagent (e.g., 6, Figure 2b).
Introduction of three boat-shaped units into an acyclic
precursor (e.g., 7, Figure 2b) may bring the desired reacting
termini closer together; however, controlling the relative
stereochemical relationship between vicinal tertiary diol units
will be a considerable synthetic challenge. To address both of
these (potential) shortcomings, it is proposed that a cyclo-
phane-based approach where a 1,4-diketo-bridging group
spanning two arene units, which are further bridged by a
single boat-shaped unit, may represent a viable macrocyclic
precursor to [4]CPP (e.g., 8, Scheme 2b). To validate this
approach we set our sights on the synthesis of a p-terphenyl-
based macrocyclic system containing a p-phenylene unit that is
more strained than a monomer unit of [4]CPP (9, Figure 2c).
The key feature involves the conversion of a macrocyclic 1,4-
diketone unit into a highly strained p-phenylene system.12
B
DOI: 10.1021/acs.orglett.6b01536
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
In recent synthetic studies, we have screened different
dehydrative aromatization reaction conditions and found that
THF can be successfully employed as solvent for this reaction.
Moreover, dehydrative aromatization reactions in THF can be
run at lower temperatures than the toluene variation, making
the reaction conditions potentially better suited for the
formation of increasingly strained p-phenylene units. Treatment
of 12 with 2.0 equiv of the Burgess reagent in THF at 50 °C for
1 h, led to the formation of monodehydration product 13 in
68% yield (entry 2, Scheme 1a). Prolonged heating of this
reaction and increasing the temperature to 70 °C did not afford
any of the desired aromatized macrocycle 9. Increasing the
number of equivalents (2.0 to 3.0 equiv) of Burgess reagent
used led to the formation of 13 (42%) and an allylic alkylation
product 14 in 31% yield (entry 3, Scheme 1a). In fact, the
addition of 4.0 equiv of Burgess reagent afforded 14 as the sole
product of this reaction in 37% yield (entry 4, Scheme 1a). It
appears that the second dehydration reaction (Path A, Scheme
1b) under these conditions is less favorable than the
corresponding allylic alkylation reaction (Path B, Scheme 1b).
However, the observation that the highly strained p-phenylene
ring was formed, albeit in trace amount, under dehydrative
conditions in toluene led us to pursue an alternative
aromatization protocol.
Close monitoring of the dehydration reaction of 12, to afford
a trace amount of 9 (entry 1, Scheme 1a), gave no indication
that the desired compound was prone to decomposition or
rearrangement reactions. However, the low yield of these two
products, when compared to previous results and the THF
variation of this reaction, was cause for concern. During the
final stages of Jasti and co-workers’ synthesis of [5]CPP,7 it was
necessary to modify their aromatization protocol to furnish the
remaining two p-phenylene rings of the desired nanohoop. In
the case of 13, we envisioned a β-elimination of AcOH could be
employed to furnish the highly strained aromatized product 9
(Scheme 2). Thus, 13, SE = 16.7 kcal/mol, was acetylated to
1
Figure 3. (a) H NMR spectra of [4]PTPP (9) and [5]PTPP (16);
(b) optimized geometry of 9, deformation angles and aromatic proton
resonances for 9 and 16.
deformation angles α and β derived from the X-ray crystal
structure of 16 (Figure 3a),13 with the DFT-derived values for
16, shows excellent agreement (Figure 3b). In fact, it appears
that the computationally derived value of α for 16 is slightly
lower than the experimentally determined value. Thus, in the
case of 9, the computational value of α = 19.1° is likely very
close to the predicted value of α in [4]CPP (19.4°).19 The β
angle of 9 at 30.2° is considerably larger than the average value
of β reported (24.9°) for the most distorted p-phenylene ring
to be characterized by X-ray crystallography.2 One of the major
questions surrounding the structure of [4]CPP is whether or
not the highly strained p-phenylene rings will retain their
benzenoid form. In the case of 9, the central arene unit of the p-
terphenyl system represents a good model for a monomer unit
of [4]CPP and can provide valuable insight into addressing this
question. Its computed SE of 42.6 kcal/mol is greater than that
of a monomer unit of 1, and the boat deformation angle (α) of
19.1° is in good agreement with the computational predicted
value of the bent benzene rings of [4]CPP. Furthermore, the
biaryl bonds that connect the central p-phenylene ring simulate
the structure of a monomer unit of 1 and represent a good
model of the arene units contained within the macrocycle.
The overall out-of-plane deformation of the central arene of
16 is nearly identical to the p-phenylene units of 4, as indicated
by their α angles of 15.7° and 15.6°, respectively. Despite both
of these large deviations from planarity, which impart a great
deal of strain on the p-phenylene rings (28.4 and 23.8 kcal/mol
for 16 and 4, respectively), the solid state structures of both 16
and 4 revealed that the C−C bond lengths of the strained arene
units are between 1.39 and 1.40 Å and 1.38−1.40 Å,
respectively. Indicating that both rings are indeed benzenoid.
The 1H NMR spectrum of 4 shows one signal at δ = 7.86 ppm,
Scheme 2. Synthesis of 1,4-Dioxa[4](3,3″)p-
terphenylophane (9)
give 15 in 74% yield. In order to introduce the remaining 51.3
kcal/mol of SE, an LDA-mediated elimination of AcOH was
designed. Indeed, treatment of acetate 15 with LDA in toluene
at 0 °C gave the desired macrocycle 9. The strain energy
induced in this reaction (51.3 kcal/mol) per p-phenylene unit
generated is greater than that induced during the Jasti synthesis
of [5]CPP (cf. 43.5 kcal/mol of SE per p-phenylene ring
synthesized).7 Incremental increases in SE induction have been
paramount to the successful preparation of increasingly strained
arenes. Similarly, this strategy of employing iterative elimination
protocols can be tailored to the synthesis of increasingly
strained systems such as [4]CPP.
All attempts to grow suitable crystals of 9 for X-ray
diffraction studies were unsuccessful. However, the optimized
geometry of 9 was obtained from density functional theory
(DFT) calculations using the B3LYP functional, in combination
with the 6-31G(d) basis set (Figure 3b). Comparing the
C
DOI: 10.1021/acs.orglett.6b01536
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
which is shifted downfield by 0.22 ppm relative to [6]CPP (5).
The narrowing of the torsional angle between the adjacent
phenyl rings, and not the increase in distortion of the arene(s),
is cause for this increase in chemical shift. In the case of the
[n](3,3″)p-terphenylophanes that have been reported, there is
very little deviation in the aromatic resonances observed in the
1H NMR spectra due to the larger torsional angles between the
arene units. Protons at the 2 and 2″-positions (a, Figure 3a) are
generally shifted to higher field as the number of methylene
groups in the alkoxy bridging unit decreases.12 This can be
attributed to the increased β angle, which directs these protons
toward the shielding cone of the central p-phenylene ring.
Protons b, c, d, e, and f (Figure 2a) show virtually no deviation
in their chemical shift values, and in the case of 9, its 1H NMR
spectrum is nearly identical to that of 16. Coupled with the X-
ray structure of 16 and identical magnetic susceptibility, these
data are suggestive that the highly strained p-phenylene ring
that is part of a teraryl macrocyclic system, containing an SE
greater than 36 kcal/mol and an α angle of 19°, is benzenoid.
Finally, the computationally derived structure of 9 indicates
that the C−C bond lengths of the central arene unit are
between 1.40 and 1.41 Å.
In summary, the synthesis of a macrocyclic benzenoid system
containing a p-phenylene ring that is predicted to be more
strained than the p-phenylene monomer units of [4]CPP has
been developed. The aromatization reaction relies on the
application of iterative elimination reactions of two hydroxyl
groups present in a precursor macrocycle and is capable of
generating 51.3 kcal/mol of SE at 0 °C. The aromatic region of
the proton NMR spectrum of 9 is essentially identical to a
previously reported homologue, 16, for which an X-ray crystal
structure has been obtained. This provides good evidence that a
(teraryl) p-phenylene unit containing an α angle comparable to
that predicted for [4]CPP and more strain energy than a
monomer unit of the yet to be synthesized nanohoop is both
benzenoid and within reach synthetically. The utility of this
(1,4-diketone) macrocyclic-based approach to strained biaryl p-
phenylene units continues to be explored, and the synthesis of
macrocycles akin to 8 (Figure 2b), for the attempted synthesis
of [4]CPP, are currently underway in our laboratory.
REFERENCES
■
(1) Tsuji, T.; Okuyama, M.; Ohkita, M.; Kawai, H.; Suzuki, T. J. Am.
Chem. Soc. 2003, 125, 951−961.
(2) Kawai, H.; Suzuki, T.; Ohkita, M.; Tsuji, T. Angew. Chem., Int. Ed.
1998, 37, 817−819.
(3) Scott, L. T.; Boorum, M. M.; McMahon, B. J.; Hagen, S.; Mack,
J.; Blank, J.; Wegner, H.; de Meijere, A. Science 2002, 295, 1500−1503.
(4) Scott, L. T.; Jackson, E. A.; Zhang, Q.; Steinberg, B. D.; Bancu,
M.; Li, B. J. Am. Chem. Soc. 2012, 134, 107−110.
(5) Lewis, S. E. Chem. Soc. Rev. 2015, 44, 2221−2304.
(6) Kayahara, E.; Patel, V. K.; Yamago, S. J. Am. Chem. Soc. 2014,
136, 2284−2287.
(7) Evans, P. J.; Darzi, E. R.; Jasti, R. Nat. Chem. 2014, 6, 404−408.
(8) Iwamoto, T.; Wantanabe, Y.; Sakamoto, Y.; Suzuki, T.; Yamago,
S. J. Am. Chem. Soc. 2011, 133, 8354−8361.
(9) Jagadeesh, M. N.; Makur, A.; Chandrasekhar, J. J. Mol. Model.
2000, 6, 226−233.
(10) Kammermeier, S.; Jones, P. G.; Herges, R. Angew. Chem., Int. Ed.
Engl. 1996, 35, 2669−2671.
(11) Xia, J.; Jasti, R. Angew. Chem., Int. Ed. 2012, 51, 2474−2476.
(12) Mitra, N. K.; Meudom, R.; Gorden, J. D.; Merner, B. L. Org.
Lett. 2015, 17, 2700−2703.
(13) Mitra, N. K.; Meudom, R.; Corzo, H. H.; Gorden, J. D.; Merner,
B. L. J. Am. Chem. Soc. 2016, 138, 3235−3240.
(14) Yusuf, M.; Kaur, M.; Jain, P.; Solanki, I. Spectrochim. Acta, Part A
2012, 97, 470−478.
(15) See Supporting Information for experimental details.
(16) It should be noted that this compound is available from
commercial sources.
(17) This scheme and additional details on compound 17 can be
found in the Supporting Information.
(18) HRMS and 1H NMR data were obtained for the trace amount 9
produced in this reaction.
(19) Darzi, E. R.; Jasti, R. Chem. Soc. Rev. 2015, 44, 6401−6410.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.6b01536.
Experimental procedures, characterization data, including
1H and 13C NMR spectra for all new compounds, and
Cartesian coordinates (PDF)
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: blm0022@auburn.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to Auburn University, the College of
Sciences and Mathematics, and the Department of Chemistry
and Biochemistry for financial support of this work. Materia
Inc. is gratefully acknowledged for the donation of catalysts.
D
DOI: 10.1021/acs.orglett.6b01536
Org. Lett. XXXX, XXX, XXX−XXX