Gram-Scale Synthesis of the 1,1,n,n-Tetramethyl[n](2,11)teropyrenophanes

Article abstract of DOI:10.1002/chem.202003828

A gram-scale synthesis of a series of 1,1,n,n-tetramethyl[n](2,11)teropyrenophanes (n=7–9) has been accomplished as well as the first synthesis of the next higher homologue 1,1,10,10-tetramethyl[10](2,11)teropyrenophane. The scale-up of the original small

Full text of DOI:10.1002/chem.202003828

Accepted Article
Accepted Article
Title: Gram-Scale Synthesis of the 1,1,n,n-Tetramethyl[n]
(2,11)teropyrenophanes
Authors: Kiran Sagar Unikela, Parisa Ghoda Ghasemabadi, Vaclav
Houska, Louise Dawe, Yuming Zhao, and Graham J. Bodwell
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202003828
Link to VoR: https://doi.org/10.1002/chem.202003828
01/2020

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
Gram-Scale Synthesis of the 1,1,n,n-Tetramethyl[n](2,11)teropyre-
nophanes
Kiran Sagar Unikela,[a] Parisa Ghods Ghasemabadi,[a] Václav Houska,[b] Louise N. Dawe,[c] Yuming
Zhao,[a] Graham J. Bodwell*[a]
[a]
Dr. K. S. Unikela, Dr. P. Ghods Ghasemabadi, Dr. Y. Zhao, Dr. G. J. Bodwell
Department of Chemistry
Memorial University of Newfoundland
283 Prince Philip Drive, St. John’s, NL, Canada, A1B 3X7
E-mail:
[b]
[c]
Dr. V. Houska
Institute of Organic Chemistry and Biochemistry
Academy of Sciences of the Czech Republic
Flemingovo nám. 2, 166 10 Prague 6, Czech Republic
Dr. L. N. Dawe
Department of Chemistry and Biochemistry
Wilfrid Laurier University
75 University Avenue West, Waterloo, ON, ,Canada, N2L 3C5
Supporting information for this article is given via a link at the end of the document.
Chart 1. Structures of superphane (1), buckminsterfullerene (C60
corannulene (3) and the CPPs (4).
) (2),
Abstract:
A
gram-scale synthesis of
a
series of 1,1,n,n-
tetramethyl[n](2,11)teropyrenophanes (n
=
7–9) has been
accomplished as well as the first synthesis of the next higher
homologue 1,1,10,10-tetramethyl[10](2,11)teropyrenophane. The
The original report of C60 (2) by Kroto, Smalley and Curl[3]
produced only minute quantities of the new carbon allotrope.
Currently, C60 is being produced commercially on a metric ton
scale.[4] The first synthesis of corannulene by Barth and Lawton[5]
was achieved in 0.4% overall yield using a 17-step synthesis.
Later, more efficient syntheses were developed (9-12 steps),
which improved the overall yield to up to 85%.[6] More recently, a
kilogram-scale synthesis of corannulene (3) was disclosed.[7] In
the same vein, Bertozzi et al.[8] reported the first synthesis of
CPPs ([9]- (4, x=4), [12]- (4, x=7) and [18]CPP (4, x=13) with just
a few milligrams of each compound bring produced. A few years
later, Jasti et al.[9] reported gram-scale syntheses of [8]CPP (4,
x=3) and [10]CPP (4, x=5), and Yamago et al.[10] communicated
gram-scale syntheses of [10]CPP (4, x=5), [8]CPP (4, x=3) and
[6]CPP (4, x=1). Several CPPs are commercially available.[11]
The ability to produce tons, kilograms and grams, respectively, of
these fascinating molecules has enabled the comprehensive
investigation of their physical and chemical properties as well as
their use in synthesis and as materials in ways that reflect the
scale of their availability. More recent breakthrough syntheses of
structurally novel and interesting π systems on a mg scale, e.g.
carbon nanobelts and related systems,[12] highlight the need for
the continued development of superior, larger scale synthetic
approaches.
scale-up of the original small-scale synthesis required the
development of several heavily modified synthetic methods, including
a chlorination / Friedel-Crafts alkylation protocol and an iodination /
Wurtz coupling protocol, which were performed on 25-30 g and 30-60
g scales, respectively. Two separate sets of conditions for the key
teropyrene-forming cyclodehydrogenation reaction at the end of the
synthetic pathway were developed, an acid-promoted one for the two
less strained congeners and an acid-free method for the two more
strained homologues.
Introduction
The design and efficient synthesis of p-electronic systems
continues to be an area of considerable interest due to the very
broad range of conceivable structures and properties, especially
their optoelectronic properties, which may lead to applications in
the field of materials science. Often, the first synthesis of a new
designed p-system is a tour de force, which establishes access to
the target compound on just a very small (milligram) scale. In
some cases, e.g. superphane (1) (Chart 1),[1] the original
synthesis (syntheses) ends up never being repeated/improved,
but lives on as a classic piece of synthetic work in textbooks.[2]
The lack of material limits the extent to which the molecule can be
studied, let alone be used as a building block for more elaborate
systems or to be incorporated into other interesting systems. In
other cases, further generations of synthetic routes are developed
and a molecule moves from being novel to being an abundant
building block, e.g. buckminsterfullerene (C60) (2), corannulene
(3) and the [n]cycloparaphenylenes (CPP) (4) (Chart 1).
Our group has reported two synthetic approaches to a small
series of 1,1,n,n-tetramethyl[n](2,11)teropyrenophanes (5a-c),[13]
which feature a severely bent 36-carbon polycyclic aromatic
system (teropyrene), a sizeable cavity and good solubility in
common organic solvents (unlike the parent teropyrene,[14] which
has extremely low solubility). The two synthetic routes are closely
related, differing only in the method used for the installation of the
first two-carbon bridge in the triply bridged pyrenophanes 6a-c
and 7a-c, which served as direct synthetic precursors to the
teropyrenophanes 5a-c (Scheme 1). Dipyren-2-ylalkanes 8a-c
were employed as common starting points for the synthesis of 6a-
c and 7a-c. The original syntheses delivered just 5-20 mg of
material, which allowed for the determination of their crystal
structures and the measurement of various physical and
spectroscopic properties. However, insufficient amounts were
available for the detailed investigation of their covalent and
supramolecular chemistry.[15a]
We recently reported the
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
development of
a
gram-scale synthesis of 5c,[15b] which
significant amounts of chloroalkene and diene side products
arising from the elimination of one or two equivalents of HCl
(Supporting Information, Sec. 3.1), the use of which in the
subsequent Friedel-Crafts alkylation reactions resulted in a
substantial lowering of the yield (10-25% for 8a-c, cf. 40-45% on
a 2 g scale).
underpinned the discovery of a highly regioselective four-fold
bromination reaction. We now report the full details of the
development of gram-scale syntheses of all three members of the
series 5a-c as well as the first synthesis of the next higher
homolog 5d.
The Friedel-Crafts alkylation reactions had always given rise to
several side products (TLC analysis) and the scale-up work
enabled the isolation and identification of some of them. For
example, Friedel-Crafts alkylation of pyrene with dichloride 11b
(containing some chloroalkenes and dienes) on a 10 g scale led
to the isolation of 8b (Rf = 0.22, 16% from diol 10b) and linear
oligomers 12 (22%, Rf = 0.15), 13 (13%, Rf = 0.10) and 14 (6%,
Rf = 0.05) (Scheme 3, Supporting Information, Sec. 3.2). The
41% combined yield of these products and the fact that 5.0
equivalents of pyrene were used suggest that monoalkylated
pyrene systems are reacting significantly faster than unalkylated
pyrene. The presence of alkenylpyrene side products 15 and 15‘,
which would arise from the alkylation of pyrene by choloralkenes
S1, was indicated by LCMS analysis of the crude product, but
none of these compounds was isolated in pure form.
Scheme 1. Existing synthetic approaches to the teropyrenophanes 5a-c.
Results and Discussion
Initial work aimed at the synthesis of synthetically useful amounts
of the teropyrenophanes 5a-c involved direct scale-up of the
existing Wurtz/McMurry route (via 7a),[13b,c] but this proved to be
problematic from the outset. Ultimately, almost every step of the
synthesis had to be modified to accommodate scale-up. At the
same time, a new teropyrenophane, 5d, was added to the series.
Grignard Reactions. The first step of the existing route, Grignard
reaction between methylmagnesium bromide and diesters 9a-c
(Scheme 2), had always been limited to a <10 g scale (vigorous
exotherm). For larger scale reactions, the commercially available
solution of the Grignard reagent (3.0 M in diethyl ether) was
diluted (ca. 1:1) with THF prior to the addition of the diesters 9a-
d, more efficient stirring was employed, and the rate of the
addition of 9a-d was carefully controlled. Using the new
procedure, the Grignard reactions could be performed
comfortably on a 30-100g scale of the diesters 9a-d to afford diols
10a-d in slightly improved yield (92-95%). It was also found that
either drum THF or recovered THF from previous Grignard
reactions could be used instead of dry THF, but an additional 0.5-
1.0 equiv of the Grignard reagent was required to obtain
equivalent yields.
Scheme 3. Initial attempts to scale up the synthesis of dipyren-2-ylalkanes 8a-
c and side products obtained from them. Reaction conditions: a) conc. HCl, rt,
3 h. b) pyrene (5.0 equiv.), AlCl3, CH2Cl2, rt, 4 h.
Scheme 2. Synthesis of diols 10a-d. Reaction conditions: a) MeMgBr (3.0 M
in Et2O), THF, -15 °C to reflux, 17-24 h.
Chlorination and Friedel-Crafts Alkylation. The second step
in the synthesis was the conversion of diols 10a-d into the
corresponding dichlorides 11a-d using concentrated aqueous
HCl. These reactions were performed previously on up to a 2 g
scale. Moving to a 10 g scale brought with it the formation of
Various ways of minimizing the formation of linear oligomers were
explored, including the use of other Lewis acids in conjunction
with dichloride 11c, the use of Brønsted acids in conjunction with
diol 10c and the introduction of substituents at one end of the
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
pyrene system (Supporting Information, Sec. 4).
These
in concentration. The recovered THF could be reused without any
adverse effect on the reaction. The diols 17a-d exhibited good
solubility in CH2Cl2, THF and ethyl acetate. Attempted purification
by column chromatography or crystallization resulted in
substantial losses of material. 1H NMR analysis of the crude
products indicated that they were easily of >90% purity, so they
were used without purification in the following step.
approaches were largely unsuccessful and attention was returned
to the original dichloride/AlCl3 system. The main challenge was
to minimize the amount of chloroalkene and diene side products
in the dichlorides 11a-d and this problem was solved by replacing
the original chlorinating agent (37% aqueous HCl solution) with
HCl gas. Bubbling gaseous HCl into a solution of diol 10c in dry
dichloromethane resulted in complete consumption of the starting
diol 10c in 4-5 h and the exclusive formation of 11c (TLC analysis).
No trace of chloroalkene or diene side products was observed.
The resulting solution of 11c was purged with nitrogen gas and
then pyrene (5 equiv.) and AlCl3 were added. This led to the
formation of dipyren-2-ylalkane 8c in 38% yield on a 2 g scale and
42% on a 10 g scale. In both cases, oligomers 12, 13 and 14
were still present (TLC analysis), but there were clearly fewer
additional spots than before (Scheme 3).
Bromination / Wurtz coupling reaction. In the original small-
scale synthesis, diols 17a-c were brominated using PBr3 to
produce the corresponding dibromides 18a-c, which were then
used as substrates in intramolecular Wurtz coupling reactions.[13b]
During initial attempts to increase the scale of the bromination of
diol 17c, it was found that hydrolysis of the resulting dibromide
18c (to give back diol 17c) during both workup and (to a greater
extent) column chromatography on silica gel became increasingly
prevalent as the scale of the reaction increased. For example, on
a 10 g scale, 18c was obtained in only 37% yield (2 steps from
dialdehyde 16c) after chromatography. The stability of the 1-
pyrenylmethyl cation[17] is most likely the cause for the ease with
which 18c undergoes hydrolysis.
Raising the number of equiv. of pyrene from 5.0 to 10.0 improved
the yield of 8c to 53% on a 10 g scale. With a much-improved
procedure in hand, the diols 10a-d were then converted to the
corresponding dipyren-2-ylalkanes 8a-d on a 25-30 g scale[16]
with no reduction in yield (51-54%) (Scheme 4). It was therefore
possible to produce 35-40 g of 8a-d in a single synthetic operation
from diols 10a-d. Furthermore, it found that replacing dry CH2Cl2
with drum grade CH2Cl2 had no detrimental effect on the yield.
The odd-tethered compounds 8a and 8c were obtained as fluffy
solids and were noticeably more soluble in CH2Cl2 than the even-
tethered compounds 8b and 8d, which were obtained as
crystalline solids.
Since the majority of the hydrolysis appeared to occur during
chromatography, it was decided to use the crude dibromide 18c
in the subsequent intramolecular Wurtz coupling reaction leading
to [n.2]pyrenophane 19c (Scheme 5). When this reaction was
performed on a 0.5 g scale, cyclophane 27c was obtained in 45%
overall yield from dialdehyde 16c. However, the yield for the
three-step sequence dropped to 22% when the scale was
increased to 5 g. Several side products were formed (TLC
analysis) and the most abundant one 20c (12%) was isolated in
pure form.
Scheme 4. Reagents and conditions: Large scale synthesis of diols 17a-d. a)
i) HCl (gas), 4-5 h; ii) N2 (gas), 20 min; iii) pyrene, AlCl3, CH2Cl2, rt, 1 h; b)
Cl2CHOCH3, TiCl4, CH2Cl2, rt, 2 h; c) NaBH4, THF, rt, 2 h.
Rieche Formylation and Reduction. Rieche formylation of the
dipyren-2-ylalkanes 8a-d required only minor changes to the
original protocol. On a 0.5 g scale, 2.5 equiv. of both TiCl4 and
Cl2CHOCH3 were used to convert 8a-c to the corresponding
dialdehydes 16a-c in 82-90% yield. Direct scale-up of these
reactions to several grams of 8a-c resulted in incomplete reaction
and lower yields (40-44%). Therefore, 5.0 equiv. of TiCl4 and
Cl2CHOCH3 were employed and the temperature was changed
from 0 °C to room temperature. With these modifications, the
Rieche formylations of 8a-d could be performed conveniently on
a 30-59 g scale to produce the corresponding dialdehydes 16a-d
in 88-90% yields (Scheme 4). As before, it was found that drum
grade CH2Cl2 could be used, but this required the addition of 0.5-
1.0 additional equiv. of both TiCl4 and Cl2CHOCH3. All of the
dialdehydes 16a-d exhibited good solubility in CH2Cl2, THF and
ethyl acetate.
Scheme 5. Initial attempts to scale up the Wurtz coupling of 18c. Reagents and
conditions: a) PBr3, CH2Cl2, 0 °C rt, 1 h; b) n-BuLi, THF, −15 °C, 10 min.
To avoid hydrolysis during both the aqueous workup and column
chromatography, a further modification to the procedure was
made. Treatment of diol 17c (50 mg) with PBr3 as before cleanly
generated dibromide 18c (TLC analysis). Switching the solvent
to THF followed by addition of n-BuLi gave rise to a complex
mixture of products from which cyclophane 19c was isolated in
10% yield (from diol 17c) (Table 1, Entry 1).
Iodination / Wurtz coupling reaction. Wurtz couplings are
known to work better for alkyl iodides than the corresponding alkyl
bromides,[18] so the possibility of replacing dibromide 18c with
diiodide 21c was investigated. Attempted conversion of diol 17c
(50 mg) into diiodide 21c using the Appel reaction (PPh3/I2/DMAP)
gave a mixture of products (TLC analysis). Addition of n-BuLi to
The next step in the synthetic pathway was the reduction of
dialdehydes 16a-d with NaBH4, which proceeded efficiently to
afford diols 17a-d on a 30-55 g scale (Scheme 4). The only
modifications that were made to the small-scale procedure were
the use of drum THF instead of dry THF and a ten-fold increase
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
this mixture gave an even more complex mixture (Table 1, Entry
2).
provides maximum separation of the two aromatic systems
(Supporting Information, Sec. 12).
Das et al. reported the iodination of benzylic alcohols using
PMHS/I2 in CHCl3.[19] Applying these conditions to diol 17c (200
mg) resulted in clean conversion to what was presumed to be
diiodide 21c within 20 min at room temperature (TLC analysis).
Replacement of the solvent with dry THF followed by the addition
of 6.5 equiv. of n-BuLi at –15 °C led to the formation of 17c in 61%
yield (Table 1, Entry 3). Using THF for the iodination step
removed the need for a change in solvent and gave an identical
result (Table 1, Entry 4). More importantly, an increase in yield
was observed as the reaction scale was increased to 2 g (61%),
9 g (66%) and then 30 g (75%) (Table 1, Entries 5 to 7). For the
30 g scale reaction, the concentration was increased (ca. two-
fold) and it was found that drum THF could be used with no effect
on the yield, but an additional 0.5-1.0 equiv. of n-BuLi was
required to consume the starting material. An additional benefit
of using diiodide 21c instead of dibromide 18c was that none of
the side product 20c was formed (TLC analysis).
Scheme 6. Optimized halogenation / Wurtz coupling reactions of 17a-d.
Reagents and conditions: a) 1. I2/PMHS, THF, rt, 20 min; 2. n-BuLi, THF, −15
°C, 10 min.
Table 2. Rieche formylation of 19c.
Table 1. Scale up of the halogenation / Wurtz coupling reactions of 17c.
With a reliable large-scale procedure in hand, the other diols 17a,
17b and 17d were then subjected to the iodination/ Wurtz
coupling sequence. The reaction of 17a proceeded smoothly on
a 40 g scale and cyclophane 19a was obtained in 62% yield
(Scheme 6). On the other hand, the reactions of the substrates
with even-numbered tethers 17b and 17d gave much lower yields
of 19b (15%) and 19d (8%). In contrast to the other reactions, a
precipitate started to form as soon as the addition of n-BuLi
commenced. The solids, which were found to be insoluble in a
range of common organic solvents, are presumably linear
oligomers arising from intermolecular Wurtz coupling reactions.
The difference in the outcome between the odd and even series
is likely a consequence of conformational preferences in the
tether. Although the conformational behavior in these systems is
surely complex, it is worth noting that the odd series has easy
access to a fully staggered conformation that places the two
pyrene units in a parallel, face-to-face orientation. The same is
not true for the even series, where an anti-like conformation
Second Rieche formylation. To prepare for the final bridge-
forming event, Rieche formylation was performed on the
[n.2](7,1)pyrenophanes 19a-d. As before, the system with the 9-
membered bridge (19c) was selected for scale-up work. In the
original synthesis of 5c, Rieche formylation of 19c on a 50 mg
scale afforded the corresponding dialdehyde 22c in 75% yield
(Table 2, Entry 1). When the reaction scale was increased to 250
mg (Table 2, Entry 2), the reaction was sluggish and the yield of
22c (30%) dropped significantly. Furthermore, in addition to a
monoformylated species, two previously unobserved products
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
were isolated using careful chromatography (Table 2). These
compounds were identified as regioisomeric dialdehydes 24c and
25c (Supporting Information, Sec. 3.3). To force the formylation
of 19c to completion, 6.0 equiv. of both TiCl4 and CHCl2OCH3
were employed (Table 2, Entry 3) and the yield of 22c improved
to 63% (250 mg scale). The yield of 22c dropped significantly
with increasing temperature (Table 2, Entries 4 to 6) or with an
increase in the scale of reaction up to 4.5 g (Table 2, Entries 7
and 8).
The order of addition of the reagents (TiCl4 and Cl2CHOCH3) to
19c was then changed. Instead of the direct addition of
Cl2CHOCH3 and then TiCl4 to a solution of cyclophane 19c in
CH2Cl2 at 0 °C, a freshly-prepared mixture of the two reagents in
CH2Cl2 at 0 °C was added dropwise to a 0 °C solution of 19c (500
mg scale) in CH2Cl2. The resulting reaction mixture was stirred at
room temperature for a period of 17 h, which afforded 22c in 50%
yield (Table 2, Entry 9). Increasing the number of equiv. of TiCl4
and Cl2CHOCH3 to 15 resulted in an increase in the yield to 58%
(Table 2, Entry 10). Only a moderate reduction in yields was
observed upon increasing the scale of the reaction to 5 g (Table
2, Entries 10 to 12). On this scale, significant amounts of the two
regioisomeric side products 24c (19%) and 25c (16%) were still
formed (Table 2, Entry 12).
Scheme 8. McMurry reaction of cyclophanedialdehydes 22a-c. Reagents and
conditions: a) TiCl4, Zn, pyridine, THF, reflux, 4 h.
When the largest cyclophanedialdehyde 22d (3 g scale) was
subjected to McMurry reaction under the conditions used for other
members of the series, the reduction product 26d (3%) and the
intended intramolecular coupling product 7d were again formed,
but the latter was accompanied by the formation of several side
products with very similar Rf values (Rf = 0.36-0.37, 30%
dichloromethane / hexanes) (Scheme 9). This compound mixture
was then subjected to ca. 15 further column chromatographic
separations, which led to the isolation of five compounds with
varying levels of purity.
With improved conditions in hand, all of the cyclophanes 19a-d
were converted to the corresponding dialdehydes 22a-c on 8-25
g scales (Scheme 7). The best result was obtained for the
smallest member of the series 19a, where the desired dialdehyde
22a was isolated in 76% yield along with 24a (10%) and 25a (6%).
The results for the higher homologues were less favorable (22b-
d
(41-44%), 24b-d (19-21%) and 25b-d (17-20%)), but
nevertheless afforded multigram quantities of material. As before,
drum grade CH2Cl2 could be used with no detrimental effect on
the yield. All of the dialdehydes 22a-d, 24a-d and 25a-d are
readily soluble in CH2Cl2.
Hy
Hx
a)
22d
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
(CH2)4
(CH2)4
Me
26d, 3%
Me
7d, 12%
3 g
28d
Me
Me
(CH2)4
Me
Me
29d, 5%
30d
Scheme 7. Modified Rieche formylation reactions of 19a-d. Reagents and
conditions: a) TiCl4, Cl2CHOCH3, CH2Cl2, 0 °C rt, 17 h.
Me
Me
Me
Me
(CH2)4
5d, 0.1%
McMurry reaction. The scale-up of the intramolecular McMurry
reactions of the three smallest cyclophanedialdehydes 22a-c to
afford [n.2.2](7,1,3)pyrenophanes 7a-c was accomplished on a 6-
12 g scale without changing the existing conditions. The reaction
of 22d to give 7d was not as straightforward and is addressed
separately below. The products 7a-c were isolated in 30-32%
yields, which are a little lower than those obtained previously on
a 0.20 g scale (36-44%) (Scheme 8). As in the previously
reported small-scale reactions, fully reduced compounds 26a-c
were formed (2-5%). Hydrocarbon 26b was the only one that
could not be isolated in pure form. Bright green fluorescent side
products 27a-b (E/Z mixtures, Supporting Information, Sec. 3.4),
which evidently arose through intermolecular McMurry reaction
and reduction (in either order), were also isolated.
Scheme 9. McMurry reaction of dialdehyde 22d. Reagents and conditions: a)
TiCl4, Zn, pyridine, THF, reflux, 4 h.
The first compound to be eluted (ca. 95% purity, 1H NMR analysis)
was identified as 28d by X-ray crystallography (Figure 1). This
side product is the result of intramolecular McMurry coupling of
dialdehyde 25d. The starting material 22d must therefore have
been contaminated with a small amount of its regioisomer 25d. In
the crystal, the two pyrene systems of 28d are essentially planar
and the angle between the two best planes is 51.3°. The distance
between the two internal carbon atoms is quite short (2.78 Å)
(Supporting Information, Sec. 10).
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
been completely consumed, the spot for 5c diminished
significantly in intensity (TLC analysis).
Table 3. Scale-up of the conversion of 7c to 5c.
Figure 1. Two views of a single molecule in the crystal structure 28d, with 50%
displacement ellipsoids. A second identical molecule in the asymmetric unit and
lattice solvent molecules omitted for clarity.
The second and third compounds to elute were dihydroteropyre-
nophane 29d and teropyrenophane 5d, which were isolated as
the major components of ca. 80:20 and 70:30 mixtures with 28d,
respectively. The pathway to these two cyclophanes involves
closure of the central C–C bond and this can be explained by the
initial formation of cyclophanemonoene 7d, followed either by a
valence isomerization and then dehydrogenation, or a Scholl-type
cycloaromatization. Either way, formal oxidations took place
Curiously, the reaction of cyclophanemonoene 7d to give the less
strained teropyrenophane 5d required somewhat more forcing
conditions (6.0 equiv. of CH3SO3H and 3.0 equiv. of DDQ). Under
these conditions, teropyrenophane 5d was isolated in 90% yield
on both a 10 mg and 200 mg scale (Scheme 10). Dihydroteropy-
renophane 29d, which was obtained during the McMurry reaction
of 7d, was also smoothly converted to 5d using 3.0 equiv. of
CH3SO3H and 1.5 equiv. of DDQ (Scheme 10). A very small
amount of a side product was isolated from the 10 mg scale
reaction. 1H NMR (e.g. d 3.25 ppm) and HRMS analysis (m/z =
738.3222) were consistent with structure 31 (2%).
under reductive conditions.
Similar unusual results were
described previously in the synthesis of a porphyrin[20] and a
(1,6)pyrenophane under McMurry conditions.[21] The closure of
the central C–C bond was not observed in the reactions leading
to 7a-c, which is likely a consequence of the lower level of strain
in teropyrenophane 5d (SEcalc = 38.7 kcal/mol) than in its lower
homologues 5a-c (SEcalc = 44.7-61.8 kcal/mol).[22] This is
reminiscent of what has been observed in the syntheses of
[n](2,7)pyrenophanes, where precursors to the less strained
members of the series undergo similar C–C bond-forming
reactions under the conditions of their formation.[23]
The fourth compound to elute was the desired cyclophanemono-
ene 7d, which was isolated in pure form in 12% yield. Considering
how much chromatography was performed, this is very likely a
substantial under-representation of the actual yield of this
compound. The fifth and final compound to elute from the cluster
of compounds was (1,6)(1,8)[2.2]pyrenophane-1-monoene 30d.
This compound clearly did not originate from 7d. A discussion of
its origin, structure and spectroscopic properties will be published
elsewhere.
Conversion of 7a-d to 5a-d. The teropyrenophanes become
increasingly strained as the bridge becomes shorter (SEcalc
=
Scheme 10. Conversion of 7d to 5d. Reagents and conditions: a) CH3SO3H
(6.0 equiv.), DDQ (3.0 equiv.), CH2Cl2, 40 min, rt. b) CH3SO3H (3.0 equiv.), DDQ
(1.5 equiv.), CH2Cl2, 30 min, rt, 90%.
38.7, 44.7, 54.2, 73.6 kcal/mol for 5d-a, respectively (Supporting
Information, Sec. 11), so they present quite different levels of
synthetic challenge. The original small-scale syntheses of 5a-c
were performed on a 20-23 mg scale under quite harsh conditions
(DDQ, m-xylene, 125-145 °C)[13b] and required a large excess of
DDQ, presumably due to the reaction of DDQ with m-xylene.[24] It
was immediately apparent that directly scaling up the reaction
was not only impractical, but also tied to a sharp decrease in the
yield. For example, 7c afforded 5c in just 21% yield on a 0.5 g
scale (Supporting Information, Sec. 5).
The use of Rathore’s conditions[25] for intramolecular Scholl
reactions (DDQ, 9:1 CH2Cl2 / CH3SO3H, 0 °C) resulted in
complete consumption of 7c in less than 2 min with only 2.0 equiv.
of DDQ, but the product 5c appeared to be reactive under the
conditions of its formation and was isolated in just 5% yield. After
some optimization (Table 3, Entries 1-3), it was found that the use
of just 2.0 equiv. of CH3SO3H along with 2.2 equiv. of DDQ in
dichloromethane at ambient temperature afforded 5c in very good
yield (83-88%), in less than 10 min and on as much as a 2.2 g
scale (Table 3, Entries 4-8). Even with the stoichiometric amount
of acid, 5c still appeared to be reactive under the conditions of its
formation. If the reaction was allowed to proceed after 7c had
Crystal structure of [10](2,11)teropyrenophane
Crystals suitable for the X-ray crystallographic studies of 5d were
obtained from an ethyl acetate / hexanes solution (Figure 2,
Supporting Information, Sec. 10). The end-to-end bend angle
(qtot) for the teropyrene system in 5d is 145.2º, which is (as
expected) significantly smaller than that of its next smaller
homologue 5c (154.3°). The ß angles[26] are 4.0° and 4.5°. As
seen in the other teropyrenophanes 5a-c, the bend angle of the
central pyrene system (q2 = 81.0º) in 5d is larger than those of the
two flanking pyrene systems (q1 = 57.6º and q3 = 63.6º), but the
differences in magnitude between them are smaller than those in
5a-c.[13a-c] Nevertheless, the teropyrene system of 5d still has a
semi-elliptical profile rather than a semi-circular one. As in the
case of 5a-c, the bent teropyrene system has the effect of
stretching the bridge. The C–C–C bond angles in the bridge of
5d lie in the range of 113.1–115.9° (avg. = 114.5°).
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
CF3COOH, CHCl2COOH, CH3COOH and p-TsOH; pKa range –
2.0 to 4.76), solvents, temperatures and reactions times were
screened, all of which ended in no improvement in the yield of 5b
(Table S2). Additionally, replacing DDQ with other oxidizing
agents delivered either moderate or no conversion of 7b to 5b or
produced a complex mixture (Table S2).
From the reactions described in Table S2, Entries 8-13, three
types of side products 32, 33 and 34 were consistently observed,
all of which could be isolated using column chromatography
(Scheme 11, Supporting Information, Sec. 3.5). The formation of
32 involves protonation at the bridgehead carbon atom to give
cation 35, followed by the addition of trifluoroacetate anion to give
36 and then dehydrogenation by DDQ (Scheme 12). Several
other sites of protonation of 7b could also lead to 32. Similarly,
when dichloroacetic acid was used (Table S2, Entries 16-19) two
side products containing at least one dichloroacetoxy group were
observed (Supporting Information, Sec. 3.6). The isolation and
characterization of these side products shed some light on the
reactivity of the teropyrene system, which in turn helped with the
development of new reaction conditions.
Figure 2. Two views of a single molecule in the crystal structure of 5d, with
50% displacement ellipsoids. Disordered lattice solvent molecules omitted for
clarity.
The scale-up of the conversion of monoene 7a using DDQ /
CH3SO3H (1.0 equiv.) at 40 °C for 2 h afforded the most highly
strained teropyrenophane 5a in only 20% yield, along with some
unreacted starting material 7a (15%) (Scheme 13). Cyclophane
7a was not completely consumed, even after increasing the
amount of CH3SO3H to 3.5 equiv. As with 7b, it was evident that
5a was reactive under the conditions of its formation (TLC
analysis).
The aromatic signals of 5d appear at d 8.90 (Ha), 8.64 (Hb), 7.88
(Hc) and 7.62 ppm (Hd). These signals are all at lower field (Dd =
0.08-0.13 ppm) than those of the corresponding protons in 5c,[13a-
c]
which is in line with the previously observed trend that the
aromatic signals all move to higher field with increasing bend in
the teropyrene system. The highest field protons in the bridge
resonate at d –0.46 ppm, which compares to d –1.00, –0.66 and
–1.15 ppm for 5c, 5b and 5a, respectively.[13a-c]
The lowest energy absorption envelope of 5d has vibronic
structure with lmax = 499, 469 and 438 nm (Figure 3). These
values are at slightly lower energy than those of 5c (lmax = 496,
464 and 435 nm) and fit with the trend of a small blue shift with
increasing bend in the teropyrene system.[13b] By comparison, the
longest wavelength absorptions of 5b and 5a are 489 and 484 nm,
respectively. The emission spectrum of 5d shows mirror image
symmetry with the absorption spectrum (lmax = 516, 543 and
585(sh) nm, Figure 3). Again, these values are at slightly lower
energy than those of 5c (lmax = 512, and 533 nm).[13b] The results
of a much more detailed study of the photophysical properties of
5a-d will be published elsewhere.
Scheme 11. Side products formed in the conversion of 7b to 5b in the presence
of CF3CO2H.
absorption ---
max (5d) = 499 nm
emission
¾
1
0.5
0
l
l
max (5d) = 516 nm
543 nm
585 nm
210
260
310
360
410
460
510
560
610
660
l (nm)
Scheme 12. Proposed mechanism for the formation of compound 32.
Figure 3. Normalized absorption and emission spectra of 5d in acetonitrile.
Irradiation wavelength = 350 nm.
Using the previously established reaction conditions for the
conversion of 7c to 5c, 7b was consumed completely in 5 min,
but 5b was obtained in only 5% yield (Table S2), which suggested
that 5b is more reactive under the conditions of its formation than
its less strained congener 5c. In an attempt to find more suitable
reaction conditions, the use of a variety of acids (CH3SO3H,
Scheme 13. Conversion of 7a to 5a. Reagents and conditions: a) DDQ (2.2
equiv.), CH3SO3H (1.0−3.5 equiv.), CH2Cl2, 30 min, 40 °C.
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
DDQ (2.7 equiv.)
7b
5b
CH3CN
Acid-free formation of 5a and 5b. The observation of side
products that contained structural units originating from the acids
that were used to promote the formation of 5a and 5b prompted
the reinvestigation of acid-free conditions. A parameter that had
only received scant attention to this point was the solvent. Ten
common organic solvents were screened for the conversion of 7a
to 5a on a 5 mg scale with 2.2 equiv. of DDQ in 2 mL solvent
under microwave irradiation (Table 4). Good conversion of 7a into
5a occurred only when DCE, CH3CN and CHCl3 were used (Table
4, Entries 1-3) and no side products were observed when using
CH3CN (TLC analysis). Modest conversion occurred in EtOAc,
m-xylene and benzene (Table 4, Entries 4-6) and no reaction took
place in hexanes, 1,4-dioxane, THF and toluene (Table 4, Entries
7-10). Due to the scale limitation under microwave conditions,
attention was turned to the use of thermal conditions. Of the 4
different solvents (DCE, CHCl3, EtOAc and CH3CN) used on a 5
mg scale, almost all of the starting material was consumed in DCE
and CH3CN (Table 4, Entries11-14). Increasing the scale in these
two solvents to 25 mg and increasing the amount of DDQ to 2.7
equiv. led to the complete consumption of the starting material in
3 h (Table 4, Entries 15-18). The yields in CH3CN were superior,
so further scale-up was performed using this solvent. Ultimately,
7a was converted into 5a on a 1.25 g scale in 81% yield (Table 4,
Entries 19-20). The conversion of 7b to 5b also proceeded
smoothly using these conditions on 5 mg, 25 mg and 1.29 g
scales in 85-90% yields (Table 5). The reaction times (2-2.5 h)
were significantly shorter than those leading to 5a (3-4 h), which
is presumably a reflection of the lower strain energy of 5b (54.2
kcal/mol vs. 61.8 kcal/mol, Supporting Information, Sec. 11).
t, reflux
entry
7b (g)
t (h)
5b (%)
1
2
3
0.005
0.025
1.298
2.0
2.5
2.5
90
88
85
The observation that the best result was observed in the solvent
with the highest dielectric constant (e = 36.6) suggests that there
is developing charge at the transition state of the reaction. It has
been noted previously that some (2,7)pyrenophanes have
relatively large calculated dipole moments that increase with the
degree of bend in the pyrene system (µ = 0.87-1.74 D)[27] and an
adamantano(2,7)pyrenophane was calculated to have a dipole
moment 2.3 D.[28] Therefore, the dipole moments of 5a-d were
calculated at the B3LYP-D3/6-31G(d) level of theory and found to
be almost identical (µ = 2.31-2.39 D), despite the range of bend
in the teropyrene system (Supporting Information, Sec. 11). It
thus seems that developing polarity in 5a-b indeed plays a role in
the success of the acid-free reaction in acetonitrile.
Conclusion
A scaled-up, 8-step synthesis of the 1,1,n,n-tetramethyl[n](2,11)-
teropyrenophanes 5a-c leading to gram quantities of material has
been developed and
a first synthesis of 5d has been
Table 4. Optimized conditions for the conversion of 7a to 5a.
accomplished on a 0.2 g scale. On the surface, the synthetic
route closely resembles the previously reported synthetic
pathway, but substantial modifications had to be made at almost
every stage. The most important advances that enabled progress
on a large scale were 1) the development of a chlorination /
Friedel-Crafts reaction leading to 8a-d on a 35-40 g scale, 2) the
development of an iodination / Wurtz coupling reaction leading to
27a-d on a 2-25 g scale and 3) the discovery of two sets of
conditions (with and without acid) for the conversion of
cyclophanemonoenes 7a-d into teropyrenophanes 5a-d. The two
highest homologues 5c and 5d were generated at room
temperature via and acid-promoted cyclodehydrogenation
reaction, whereas the two lowest homologues were formed in a
polar aprotic solvent (acetonitrile) at ca. 82 °C without acid. The
need for different conditions is a reflection of the increasing level
of molecular strain as the bridge becomes shorter. The overall
yields for 5a-d were 4.9% (1.0 g), 0.7% (1.3 g), 3.6% (1.9 g), and
0.2% (0.2 g), respectively. The lower yields for the 5b and 5d
(even number of atoms in the long bridge) are a result of low yields
in the intramolecular Wurtz coupling step. The new teropyreno-
phane 5d features a bent teropyrene system with an end-to-end
bend of 145.2°. The availability of synthetically useful amounts of
5a-d will enable the study of the chemistry these systems and
their use as starting points for the construction of cyclophanes
with larger aromatic systems.
Acknowledgements
Table 5. Optimized thermal conditions for the conversion of 7b to 5b.
Financial support of this work (L.N.D., Y.Z., G.J.B.) from the
Natural Sciences and Engineering Research Council (NSERC) of
Canada is gratefully acknowledged.
Keywords: cyclophanes • nonplanar aromatic compounds •
teropyrene • pyrene • cyclodehydrogenation
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
[1] a) Y. Kine, M. Brown, V. Boekelheide, J. Am. Chem. Soc. 1979, 101,
3126-3127; b) Y. Sekine, V. Boekelheide, J. Am. Chem. Soc., 1981,
103, 1777-1785; c) W. D. Rohrbach, R. Sheley, V. Boekelheide,
Tetrahedron 1984, 40, 4823-4828; d) S. El-Tamany, H. Hopf, Chem.
Ber. 1983, 116, 1682–1685.
[21] P. R. Nandaluru, P. Dongare, C. M. Kraml, R. A. Pascal, Jr., L. N. Dawe,
D. W. Thompson and G. J. Bodwell, Chem. Commun. 2012, 48,
7747-7749.
[22] These numbers are different from those reported earlier (ref. [13b]).
The previous calculations did not include a dispersion correction, while
the ones reported here do.
[2] H. Hopf, Classics in Hydrocarbon Chemistry: Synthesis, Concepts,
Perspectives; Wiley-VCH, Weinheim, 2000.
[23] G. J. Bodwell, J. N. Bridson, M. K. Cyranski, J. W. J. Kennedy, T. M.
Krygowski, M. R. Mannion, D. O. Miller, J. Org. Chem. 2003, 68,
2089-2098.
[3] H. W. Kroto, J. R. Heath, S. C. O’Brien, R. F. Curl, R. E. Smalley, Nature
1985, 318, 162-163.
[4] H. Murayama, S. Tomonoh, J. M. Alford, M. E. Karpuk, Fullerenes,
Nanotubes and Carbon Nanostructures 2004, 12, 1-9.
[24] V. S. Batista, R. H. Crabtree, S. J. Konezny, O. R. Luca, J. M.
Praetorius, New J. Chem. 2012, 36, 1141-1144.
[5] W. E. Barth, R. G. Lawton, J. Am. Chem. Soc. 1971, 93, 1730-1745.
[6] a) L. T. Scott, M. M. Hashemi, D. T. Meyer, H. B. Warren, J. Am. Chem.
Soc. 1991, 113, 7082-7084; b) L. T. Scott, P.-C. Cheng, M. M.
Hashemi, M. S. Bratcher, D. T. Meyer, H. B. Warren, J. Am. Chem. Soc.
1997, 119, 10963-10968; c) A. Sygula, P. W. Rabideau, J. Am. Chem.
Soc. 2000, 122, 6323-6324; d) T. J. Seiders, E. L. Elliot, G. H. Grube,
J. S. Siegel, J. Am. Chem. Soc. 1999, 121, 7804-7813.
[25] L. Zhai, R. Shukla, R. Rathore, Org. Lett. 2009, 11, 3474-3477.
[26] The ß angles are analogous to those used for paracyclophanes. See:
P. M. Keehn, in Cyclophanes, Vol. 1, P. M. Keehn, S. M. Rosenfeld,
(Eds.), Academic Press: New York, 1983, pp. 69–238.
[27] B. Zhang, R. A. Pascal Jr., Y. Zhao, G. J. Bodwell, Can. J. Chem. 2017,
95, 460–481.
[28] P. Kahl, J. P. Wagner, C. Balestrieri, J. Becker, H. Hausmann, G. J.
Bodwell, P. R. Schreiner, Angew. Chem. Int. Ed. 2016, 55, 9277-9281.
[7] A. M. Butterfield, B. Gilomen, J. S. Siegel, Org. Process Res. Dev. 2012,
16, 664-676.
[8] R. Jasti, J. Bhattacharjee, J. B. Neaton, C. R. Bertozzi, J. Am. Chem.
Soc. 2008, 130, 17646-17647.
[9] J. Xia, J. W. Bacon, R. Jasti, Chem. Sci. 2012, 3, 3018-3021.
[10] a) E. Kayahara, L. Sun, H. Onishi, K. Suzuki, T. Fukushima, A. Sawada,
H. Kaji, S. Yamago, J. Am. Chem. Soc. 2017, 139, 18480-18483; b) E.
Kayahara, V. K. Patel, J. Xia, R. Jasti, S. Yamago, Synlett 2015, 26,
1615-1619.
[11] A range of cycloparaphenylenes can be purchased from TCI Chemicals.
[12] a) G. Povie, Y. Segawa, T. Nishihara, Y. Miyauchi, K. Itami, Science
2017, 356, 172–175; b) G. Povie, Y. Segawa, T. Nishihara, Y. Miyauchi,
K. Itami, J. Am. Chem. Soc. 2018, 140, 10054–10059; c) K. Y. Cheung,
S. Gui, C. Deng, H. Liang, Z. Xia, Z. Liu, L. Chi, Q. Miao, Chem. 2019,
5, 838–847; d) X. Lu, T. Y. Gopalakrishna, Y. Han, Y. Ni, Y. Zou, J. Wu,
J. Am. Chem. Soc. 2019, 141, 5934–5941; e) S. Nishigaki, Y. Shibata,
A. Nakajima, H. Okajima, Y. Masumoto, T. Osawa, A. Muranaka, H.
Sugiyama, A. Horikawa, H. Uekusa, H. Koshino, M. Uchiyama, A.
Sakamoto, K. Tanaka, J. Am. Chem. Soc. 2019, 141, 14955–14960; f)
K. Y. Cheung, K. Watanabe, Y. Segawa K. Itami, ChemRxiv 2020,
doi.org/10.26434/chemrxiv.12324353.v2; g) Y. Li, Y. Segawa, A. Yagi,
K. Itami, J. Am. Chem. Soc. 2020, 142, 12850–12856.
[13] a) B. L. Merner, L. N. Dawe, G. J. Bodwell, Angew. Chem. Int. Ed. 2009,
48, 5487-5491; b) B. L. Merner, K. S. Unikela, L. N. Dawe, D. W.
Thompson, G. J. Bodwell, Chem. Commun. 2013, 49, 5930-5932; c) K.
S. Unikela, B. L. Merner, P. Ghods Ghasemabadi, C. C. Warford, C. S.
Qiu, L. N. Dawe, Y. Zhao, G. J. Bodwell, Eur. J. Org. Chem. 2019,
4546-4560.
[14] T. Umemoto, T. Kawashima, Y. Sakata, S. Misumi, Tetrahedron Lett.
1975, 16, 1005-1006.
[15] a) Y. Chen, Y. Jami-Alahmadi, K. S. Unikela, G. J. Bodwell, T. D.
Fridgen, ChemPhysPhysChem 2018, 19, 2194–2199. b) K. S. Unikela,
T. L. Roemmele, V. Houska, K. E. McGrath. D. M. Tobin, L. N. Dawe,
R. T. Boeré, G. J. Bodwell, Angew. Chem. Int. Ed. 2018, 57,
1707-1711.
[16] The use of 10 equiv. of pyrene on such a large scale meant that a large
column (70 ´ 12.5 cm) was required for the isolation of 8a-d. This
limited the scale of the reaction to 30 g. Both the eluent (8% chloroform
/ hexanes) and the recovered pyrene (ca. 240 g) could be reused. In
fact, the purity and appearance of the recovered pyrene was superior
to the commercial material.
[17] K. K. Laali, P. E. Hansen, J. Org. Chem. 1997, 62, 5804-5810.
[18] a) S. Nunomoto, Y. Kawakami, Y. Yamashita, Bull. Chem. Soc. Jpn.
1981, 54, 2831-2832; b) J. W. H. Oldham and A. R. Ubbelohde, J.
Chem. Soc. 1938, 201-206.
[19] B. Das, H. Holla, Y. Srinivas, N. Chowdhury, B. P. Bandgar,
Tetrahedron Lett. 2007, 48, 3201-3204.
[20] T. D. Lash, S. A. Jones, G. M. Ferrence, J. Am. Chem. Soc. 2010, 132,
12786-12787.
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202003828
Chemistry - A European Journal
RESEARCH ARTICLE
Entry for the Table of Contents
A large-scale synthetic route to a series of four [n](2,11)teropyrenophanes (n=7-10) has been developed, which delivered the three
smallest homologues on a >1 g scale and enabled the first synthesis of the highest homologue. The synthetic pathway culminates
with a teropyrene-forming cyclodehydrogenation, for which two sets of conditions had to be developed as a result of the broad range
of strain (38.7-61.8 kcal/mol) in these bow-shaped cyclophanes.
Institute and/or researcher Twitter usernames: ((optional))
10
This article is protected by copyright. All rights reserved.