J.J. Dressler et al.
Tetrahedron Letters 61 (2020) 152429
Table 2
Electronic absorption and emission data for 2, 3, 5, and 15.
dibromoarene 9 [9b] resulted in terphenyl 13. Terphenyl 13, con-
taining two TEG units, was very challenging to separate from the
excess 12 (and its deborylated side product) employed in the
Suzuki coupling because of its highly polar nature and strong inter-
actions between the ethylene glycol units of the product and side
products. However, after numerous trials, we found that column
chromatography with THF in hexane immediately followed by a
subsequent column with EtOAc in CH2Cl2 was essential to achieve
this separation. All other eluent conditions tested resulted in co-
elution. In previous studies to achieve crankshafts 2, 3, and 5 we
found that direct demethylation of terphenyls analogous to 13
with BBr3 conveniently achieved dilactone-bridged final products
in a single step. In this current study, attempts to selectively cleave
the methyl ethers on 13 using BBr3 unfortunately resulted in
decomposition, likely due to competing cleavage of the TEG
groups, even with careful stoichiometric addition at low tempera-
tures. This necessitated an alternative route consisting of a ceric
ammonium nitrate (CAN) oxidative cleavage of the p-dimethoxy
groups to obtain quinone 14. Quinone 14 was subsequently
reduced with NaBH4 and then underwent acid-promoted lac-
tonization in a single pot to achieve solid TEG solubilized crank-
shaft dilactone 15 in good overall yields.
Unlike previously studied lactone-bridged oligophenyls 1–7
(Fig. 2) [9b,9c,11a,11b], TEG-containing 15 displays solubility in a
wide variety of solvents (Table 1). Previously reported crankshaft
compounds containing the ‘‘bridge-flipped” lactone orientation
(unsubstituted 4 [11b] and tert-butyl solubilized 5 [9b]) show
slightly better solubility than isomeric crankshafts 1-3 [9b,11a]
or triskelions 6 and 7 [9c]. Compound 1 [11a] has only been shown
to dissolve in concentrated H2SO4 and compounds 2 [9b], 3 [9b], 6
[9c], and 7 [9c] are sparingly soluble in THF, CH2Cl2, CHCl3, and
compd
lowest E abs kmax[nm] (
e
[MÀ1cmÀ1])a
em kbmax[nm]
UFc
2
3
5
15
364 (36832)
384 (65080)
335 (37983)
374 (51400)
385
394
439
405
0.23
0.92
0.24
0.33
aUV–vis spectral data from 2, 3, and 5 dissolved in CH2Cl2 and 15 dissolved in
CH3CN. bEmission spectral data from 2, 3, and 5 dissolved in CH2Cl2 and 15 dis-
solved in CH3CN and excited at 350 nm. cCalculated relative to a quinine sulfate
standard dissolved in 0.5 M H2SO4 and excited at 350 nm (values corrected for
analyte and standard solvents) [21].
blue light with Stokes shifts ranging from 10 to 104 nm and rela-
tive UF between 0.23 and 0.92. However, these studies were again
limited by the poor solubility of these compounds and all spectro-
scopic studies were required to be done in CH2Cl2. Indeed, lactone-
bridged terphenyl crankshaft 15 was found to have very favorable
optical properties (Table 2), similar to the previously studied 2, 3,
and 5. The primary advantage of TEG-solubilized 15 over 2/3/5 is
that spectroscopic studies could now be performed in a variety
of solvents, but most importantly, a polar solvent. UV–vis studies
revealed that 15 had a kmax = 374 nm with very strong molar
absorptivity (e
= 51400 MÀ1cmÀ1) in acetonitrile. Crankshaft 15
also displayed significant blue-violet fluorescence at kmax = 405 nm
in acetonitrile with a UF = 0.33.
Compounds 2, 3, and 5 were originally envisioned as possible
pH-driven molecular switches [9b] and TEG-solubilized 15 is also
a potential pH-driven switch with switch states similar to those
proposed for 2 and 3 (Fig. 4). Unlike simple unsubstituted p-ter-
phenyl, the lactone-bridges engineered in terphenyls 2, 3, and 15
allows for these compounds to adopt a rigid conformation with
toluene, often requiring heat to dissolve. The strong
p- p inter-
molecular interactions of crankshafts 1–3 make it difficult to solu-
bilize these compounds, despite 2 and 3 containing tert-butyl and
3,5-di-tert-butylphenyl solubilizing groups. The strength of their
intermolecular interactions can be probed indirectly by simple
melting point analysis. Compounds 1 [11a], 2 [9b], and 3 [9b] have
very high reported melting points of 420–422 °C, 366 °C, and
440 °C, respectively. Triskelions 6 [9c] and 7 [9c] also display
restricted rotation, resulting in a high degree of
p-orbital overlap
between the individual arenes. Although X-ray quality crystals of
these compounds were elusive in our hands, evidence of the rigid-
ity of 2, 3 and 15 can be observed in the significant downfield
chemical shifts of the sterically-compressed aromatic H’s [17] in
the 1H NMR spectra of these compounds relative to their non-
bridged precursors (see previous studies [9b] and supplementary
data). This rigidity leads to these compounds being very efficient
at absorbing near-UV/visible light and emitting blue-violet light.
The potential of 2, 3, and 15 as molecular switches lies in the abil-
ity of their lactone bridges being readily cleaved by excess tetra-
butylammonium hydroxide (TBAOH), which results in non-planar
strong intermolecular
p- p interactions with melting points in
excess of 340 °C. In contrast, TEG-solubilized 15 has a much lower
melting point of 160–161 °C, despite having a much greater molar
mass than 1 and 2, and a similar molar mass to 3. This suggests that
the conformational flexibility of the TEG-groups can better over-
come
p-p interactions and allow for solute–solvent interactions
anions 2a, 3a, and 15a with weakened
p-overlap. The individual
to compete more favorably compared to 1–3. The higher solubility
of 15 makes it more tractable to characterize, process, and poten-
tially more useful in optical applications.
The spectroscopic properties of lactone-bridged terphenyls 2, 3,
and 5 were studied previously by us (Table 2) [9b]. All three com-
pounds were found to absorb strongly in the near UV to visible
arenes are now no longer rotationally restricted and both absorp-
tion and emission of light should be greatly attenuated. The lactone
can then be re-formed by addition of trifluoroacetic acid (TFA), and
thus the desirable optical properties should return. However, it is
known that hydroquinone dianions are highly susceptible to oxida-
tion and, under strongly basic conditions like those required to
generate 2a, 3a, and 15a, the resulting benzoquinones could read-
ily decompose [18]. Thus, the switching of these compounds could
be limited. In our previous studies [9b] the drastic difference in sol-
ubility between 2/3 and 2a/3a meant that studies probing reversi-
ble switching could not be done, since 2a/3a were found to
precipitate upon addition of TBAOH. However, newly synthesized
TEG-solubilized 15 can overcome these challenges because both
15 and 15a should each be at least partially soluble in polar organic
solvents.
region with large molar extinction coefficients (e) all greater than
35,000 MÀ1cmÀ1. The restricted rotational motion caused by the
lactone bridges between the individual arene units of 2, 3, and 5
also resulted in significant visible fluorescence, emitting violet-
Table 1
General solubility of 15 in various solvents.
A. Readily
B. Dissolves with heat
C. Dissolves with
D.
dissolves at rt
and remains in solution heat but ppt after Insoluble
Since 15 was soluble in acetonitrile, studies could be performed
to probe the capabilities of 15 as a molecular switch. We found that
the intensity of absorption and emission were drastically reduced
when a drop of TBAOH was added to samples of 15 (Fig. 5), sug-
gesting that lactone cleavage was occurring resulting in non-rigid
without heating after cooling to rt
cooling to rt
at all T
THF, CH2Cl2,
CHCl3
dioxane, DMSO,
acetone, acetonitrile,
EtOAc, DMF, toluene
iPrOH, EtOH,
MeOH
Hexane,
H2O,
diethyl
ether
4