A push-pull macrocycle with both linearly conjugated and cross-conjugated bridges

Article abstract of DOI:10.1021/ol401697d

A series of shape-persistent macrocycles featuring both m-phenylene and 2,5-thiophene linkers has been synthesized, including an example where they bridge electron-rich (veratrole) and electron-poor (phthalimide) units. Charge transfer in this push-pull macrocycle has been investigated by UV-vis and fluorescence spectroscopies and DFT calculations. The effect of pairing structurally distinct conjugated bridges is discussed in the context of acyclic and symmetrical macrocyclic analogs.

Full text of DOI:10.1021/ol401697d

ORGANIC
LETTERS
2013
Vol. 15, No. 14
3762–3765
A PushꢀPull Macrocycle With
Both Linearly Conjugated and
Cross-Conjugated Bridges
Wade C. W. Leu and C. Scott Hartley*
Department of Chemistry & Biochemistry, Miami University, Oxford, Ohio 45056,
United States
scott.hartley@miamioh.edu
Received June 14, 2013
ABSTRACT
A series of shape-persistent macrocycles featuring both m-phenylene and 2,5-thiophene linkers has been synthesized, including an example
where they bridge electron-rich (veratrole) and electron-poor (phthalimide) units. Charge transfer in this “pushꢀpull macrocycle” has been
investigated by UVꢀvis and fluorescence spectroscopies and DFT calculations. The effect of pairing structurally distinct conjugated bridges is
discussed in the context of acyclic and symmetrical macrocyclic analogs.
Despite long-standing interest in one-dimensional (wire-
like) pushꢀpull compounds, only relatively recently have
analogous systems with two-dimensional conjugation1,2
begun to attract attention. Accordingly, compounds have
been reported with multiple donor and acceptor groups
and multiple conjugated paths between them.3ꢀ7 The two-
dimensional nature of these structures is integral to their
design and resultant properties. For example, spatial separa-
tion of the frontier orbitals in cruciform structures gives rise
to particularly useful photophysical properties for sensing.6
Of course, there are many possible fundamental con-
figurations of two-dimensional π-systems. We recently
reported the synthesis and characterization of a set of
“pushꢀpull macrocycles”, D(mP)2A and D(Th)2A, which
feature electron-rich and -poor units linked by two
conjugated bridges.8ꢀ10 Pairing of the bridges was found
to have a large effect on the photophysical properties of the
compounds when compared to acyclic analogs D(mP)A
and D(Th)A. Notably, the rate of radiative charge recom-
bination for the linearly conjugated 2,5-thiophene-based
system was dramatically reduced, an effect we attributed to
a reduction in electronic coupling because of the symmetry
of the macrocycle.11,12
Here, we report a new set of macrocycles with both 2,5-
thiophene and m-phenylene bridges. Included in this series
is compound D(mP/Th)A, which allows us to explore the
properties of chromophores with donor and acceptor
groups linked by different bridges with very different behav-
ior: the 2,5-thiophene bridge is linearly conjugated, and the
m-phenylene bridge is cross-conjugated.13 Taken separately,
charge transfer mediated by the linearly conjugated bridge
should be efficient, whereas cross-conjugated moieties are
(1) Opsitnick, E.; Lee, D. Chem.;Eur. J. 2007, 13, 7040–7049.
(2) Diederich, F.; Kivala, M. Adv. Mater. 2010, 22, 803–812.
(3) Hilger, A.; Gisselbrecht, J.-P.; Tykwinski, R. R.; Boudon, C.;
€
Schreiber, M.; Martin, R. E.; Luthi, H. P.; Gross, M.; Diederich, F.
(8) Leu, W. C. W.; Fritz, A. E.; Digianantonio, K. M.; Hartley, C. S.
J. Org. Chem. 2012, 77, 2285–2298.
J. Am. Chem. Soc. 1997, 119, 2069–2078.
(4) Spitler, E. L.; Monson, J. M.; Haley, M. M. J. Org. Chem. 2008,
73, 2211–2223.
(9) For another example of a pushꢀpull macrocycle, see: Wu, Y.-L.;
ꢀ
Bures, F.; Jarowski, P. D.; Schweizer, W. B.; Boudon, C.; Gisselbrecht,
J.-P.; Diederich, F. Chem.;Eur. J. 2010, 16, 9592–9605.
(10) Dickson, N. M.; Leu, W. C. W.; Hartley, C. S. Submitted.
(11) Zeng, Y.; Zimmt, M. B. J. Am. Chem. Soc. 1991, 113, 5107–5109.
(12) Oliver, A. M.; Paddon-Row, M. N.; Kroon, J.; Verhoeven, J. W.
Chem. Phys. Lett. 1992, 191, 371–377.
(5) Marsden, J. A.; Miller, J. J.; Shirtcliff, L. D.; Haley, M. M. J. Am.
Chem. Soc. 2005, 127, 2464–2476.
(6) Zucchero, A. J.; McGrier, P. L.; Bunz, U. H. F. Acc. Chem. Res.
2010, 43, 397–408.
(7) Baxter, P. N. W. J. Org. Chem. 2004, 69, 1813–1821.
r
10.1021/ol401697d
Published on Web 07/10/2013
2013 American Chemical Society

used for our original series of macrocycles,8 as shown in
Scheme 1. Beginning with 1-bromo-2-iodoarenes, Sonogashira
coupling was used to add first the 2,5-thiophene bridges
followed by the m-phenylene bridges, giving compounds 1
and then 2 in good yields. The tert-butyl groups were
included on the m-phenylene units to ensure good solubi-
lity and hinder aggregation, which would complicate
spectroscopic characterization. Unfortunately, deprotec-
tion of the TIPS groups to give compounds 3 gave only
modest isolated yields. Although this is typically a highly
efficient reaction, we observed decomposition to unidenti-
fied byproducts; further, chromatographic isolation of
the target compounds from nearly coeluting impurities
required us to sacrifice yield for purity. The key macro-
cyclization step using copper-free Sonogashira coupling
conditions with Pd(PtBu3)2 as the catalyst18 gave the target
compounds in modest but workable yields, similar to those
previously obtained for bis(2,5-thiophene) macrocycles
(e.g., 34% for D(Th)2A). These low yields are readily
explained by the relatively poor fit of the 2,5-thiophene
unit within the macrocyclic framework (see below).
known to facilitate charge transfer but hinder recombina-
tion;14,15 cross-conjugated systems have also been proposed
as the basis for single-molecule transistors.16,17 This pushꢀ
pull macrocycle allows us to address whether one bridge
dominates the photophysical properties (i.e., whether the
properties are predicted by acyclic analogs such as D(Th)A)
or whether these multiple conjugation paths can only be
understood as a combined unit. The control over excited-
state properties and charge-transfer rates afforded by this
two-dimensional π-system design may ultimately have ap-
plications in photovoltaics and single-molecule devices.
Scheme 1
Figure 1. UVꢀvis (left) and fluorescence (right) spectra of
H(mP/Th)H, D(mP/Th)D, and A(mP/Th)A (top) and of
D(mP/Th)A (bottom). The spectra correspond to cyclohexane
solutions except for the fluorescence spectra of D(mP/Th)A.
To the best of our knowledge, macrocycles with this
combination of 2,5-thiophene and m-phenylene units are
unknown. The synthesis of pushꢀpull, donorꢀdonor,
acceptorꢀacceptor, and unsubstituted compounds D-
(mP/Th)A, D(mP/Th)D, A(mP/Th)A, and H(mP/Th)H is
a straightforward modification of the synthetic strategy we
(13) Gholami, M.; Tykwinski, R. R. Chem. Rev. 2006, 106, 4997–5027.
(14) van Walree, C. A.; Kaats-Richters, V. E. M.; Veen, S. J.;
Wieczorek, B.; van der Wiel, J. H.; van der Wiel, B. C. Eur. J. Org.
Chem. 2004, 2004, 3046–3056.
Org. Lett., Vol. 15, No. 14, 2013
3763

UVꢀvis and fluorescence spectra of the macrocycles are
shown in Figure 1. In all cases, the shapes of the spectra are
independent of concentration (10ꢀ5ꢀ10ꢀ7 M), suggesting
that ground-state aggregation of the macrocycles is insig-
nificant under these conditions and that emission does not
occur from excimer states. The shapes of the UVꢀvis and
fluorescence spectra of the symmetrical macrocycles
H(mP/Th)H, D(mP/Th)D, and A(mP/Th)A show little
dependence on solvent polarity (see Figure S1). The
UVꢀvis spectra of D(mP/Th)A are also solvent-independent
(see Figure S2), suggesting that excitation of this pushꢀ
pull compound occurs to a locally excited state even in
polar media. In contrast, its fluorescence spectra red shift
with increasing solvent polarity, with the loss of vibronic
structure, suggesting that emission occurs from a charge-
transfer state in polar solvents. Fluorescence properties of
the macrocycles are compiled in Table 1; data for the
previously reported D(Th)2A and D(Th)A are included
for comparison.8 In general, the mixed-bridge macrocycles
are highly fluorescent (Φf ≈ 0.2). The excited-state dipole
moment of the charge-transfer state of D(mP/Th)A
(μe = 34 D) was estimated by LippertꢀMataga analysis
of the fluorescence solvatochromism, as described in the
Supporting Information. Fluorescence lifetimes (τf) were
measured by time-correlated single photon counting which
gave good monoexponential fits in all cases. Radiative (kr)
and nonradiative (knr) rate constants were calculated from
Φf and τr (kr = Φfτfꢀ1, knr = (1 ꢀ Φf)τfꢀ1).
hydrogen atoms (giving D(mP/Th)A0, etc.). The geome-
tries were optimized at the B3LYP/6-31þG(d,p) level
and are shown in Figures 2 and S3. The rather shallow
angle of the 2,5-thiophendiyl moiety (∼140°) is not well-
accommodated within the macrocyclic framework. Con-
sequently, intraannular steric interactions lead to a
∼23°ꢀ25° twist of thiophene rings from the planes of the
macrocycles; this out-of-plane twist is about half that
previously determined for D(Th)2A at the same level of
theory (∼42°). In contrast, the m-phenylene unit is very
close to coplanarity with the macrocycle.
To model the spectroscopic properties, the optimized geom-
etries were further probed by TD-DFT using the CAM-
B3LYP functional,19 which should better handle charge-
transfer states20 (see FMOs in Figures 2 and S3). Key
calculated properties of all compounds are summarized in
Table 2, and predicted spectra are shown in Figure S4. The
variation in the energies of the first excited states, extracted
from the low-energy absorption bands of the UVꢀvis spectra,
are well-reproduced by the calculations, with a slight over-
estimation of ∼0.1 eV in each case. The shifts in the spectra of
the symmetrical compounds, H(mP/Th)H, D(mP/Th)D,
and A(mP/Th)A (Figure 1, top), are easily explained by
their HOMOꢀLUMO gaps, as shown in Figure S5.
As expected, for the pushꢀpull compound D(mP/Th)A0
the electron-rich veratrole moiety is the principal contri-
butor to the HOMO, whereas the electron-deficient phtha-
limide moiety is the principal contributor to the LUMO.
Delocalization through the linearly conjugated thiophene
bridge is more pronounced than through the cross-
conjugated m-phenylene bridge. The lowest energy transi-
tion of D(mP/Th)A0 is predicted to be primarily a direct
HOMOꢀLUMO transition (60% contribution), corre-
sponding to a net change in dipole of Δμ = 8.40 D. This
value is much smaller than that estimated by Lippertꢀ
Mataga analysis of the experimental fluorescence spectra
and is consistent with the minimal solvatochromism
observed in the UVꢀvis spectra.
Table 1. Excited-State Dipole Moments, Quantum Yields,
Fluorescence Lifetimes, and Radiative and Nonradiative Rate
Constants
μe
τf
kr
knr
(ꢁ 108 sꢀ1
(D) solventa Φf
(ns) (ꢁ 108 sꢀ1
)
)
D(mP/Th)A 34 Cy
Diox
CH2Cl2 0.18 2.87
0.18 1.61
1.1
1.2
0.63
2.2
1.6
2.2
0.05
1.8
5.1
4.5
2.9
6.4
5.0
8.8
10
0.21 1.74
The behavior of this new pushꢀpull macrocycle D(mP/Th)A
is perhaps best considered in the context of the previously
reported acyclic and symmetrically bridged analogs
D(Th)A and D(Th)2A. Direct comparisons of the UVꢀvis
and fluorescence spectra of D(mP/Th)A with these other
compounds are given in the Supporting Information
(Figure S6). Because m-phenylenes are often considered
to “break” conjugation,21 one might expect the photo-
physical properties of this pushꢀpull macrocycle to clo-
sely resemble those of the acyclic, thiophene-bridged
compound D(Th)A, with the cross-conjugated m-phenylene
bridge acting primarily as a passive structural unit analo-
gous to a nonconjugated linker. Indeed, the fluorescence
spectra of D(mP/Th)A are a good match to those of
D(Th)A; further, LippertꢀMataga analysis yields vir-
tually identical excited state dipole moments, and their
fluorescence quantum yields are similar.
D(mP/Th)D
A(mP/Th)A
H(mP/Th)H
D(Th)2Ab
Cy
Cy
Cy
0.26 1.16
0.24 1.53
0.20 0.91
25 CH2Cl2 0.005 1.0c
33 CH2Cl2 0.11 0.62
D(Th)Ab
14
a Cy = cyclohexane, Diox = 1,4-dioxane. b Data from ref 8.
c Average lifetime from a double exponential fit.
The effect of the two bridges on the photophysical pro-
perties can be understood by considering the electronic
structures of the compounds, which were explored by DFT
calculations. Simplified versions of each macrocycle were
considered, with hexyl and tert-butyl groups replaced with
(15) Thompson, A. L.; Ahn, T.-S.; Thomas, K. R. J.; Thayumanavan,
´
S.; Martınez, T. J.; Bardeen, C. J. J. Am. Chem. Soc. 2005, 127, 16348–
16349.
(16) Cardamone, D. M.; Stafford, C. A.; Mazumdar, S. Nano Lett.
2006, 6, 2422–2426.
(17) Andrews, D. Q.; Solomon, G. C.; van Duyne, R. P.; Ratner,
M. A. J. Am. Chem. Soc. 2008, 130, 17309–17319.
(18) Soheili, A.; Albaneze-Walker, J.; Murry, J. A.; Dormer, P. G.;
Hughes, D. L. Org. Lett. 2003, 5, 4191–4194.
(19) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393,
51–57.
(20) Peach, M. J. G.; Benfield, P.; Helgaker, T.; Tozer, D. J. J. Chem.
Phys. 2008, 128, 044118.
3764
Org. Lett., Vol. 15, No. 14, 2013

intensity of this band for macrocycle D(Th)2A by noting
that the direct charge-transfer transition is symmetry
forbidden. Clearly, this is not the case for D(mP/Th)A,
which is not a symmetrical compound. However, the
TD-DFT calculations do predict that the oscillator
strength for the transition to the first excited state of
D(mP/Th)A0 (f = 0.32) is much closer to that of D(Th)2A0
(f = 0.00) than that of D(Th)A0 (f = 1.86). This similarity
can be understood by the observation that the HOMO and
LUMO of D(mP/Th)A0 remain approximately symmetric
and antisymmetric with respect to the axis of charge transfer
(Figure 2), despite the structural differences between the two
bridges.
This effect extends to the fluorescence properties as well.
Consider D(mP/Th)A and D(Th)A in CH2Cl2: the fluor-
escence quantum yields are similar;however, the lifetime of
the charge transfer state of the mixed-bridge macrocycle is
roughly 5-fold longer, corresponding to a 3-fold decrease
in the radiative rate constant (kr) and a 5-fold decrease in
the nonradiative rate constant (knr). The decrease in knr is
likely due to reduced vibrational deactivation because of
the structural rigidity provided by the second bridge. The
decrease in kr is substantial, but not as significant as the
22-fold decrease observed for D(Th)2A. Assuming that the
frontier MOs in the relaxed charge-transfer state are
similar to those in Figure 2, the relative kr values can be
explained by the relative strengths of the HOMOꢀLUMO
transitions as discussed above.
The two-dimensional π-systems of these mixed-bridge
compounds are clearly complex, as analogous acyclic or
symmetrical macrocyclic compounds fail to predict their
behavior. The effect on the rate of radiative charge transfer
suggests that this mixed-bridge structural motif may ulti-
mately offer a means to tune the electronic coupling, a key
parameter in determining the lifetimes of charge-transfer
states22 and in the construction of single-molecule devices.23
In summary, we have prepared the first examples of a set
of shape-persistent macrocycles with both 2,5-thiophene
and m-phenylene moieties. The compounds with symmet-
rical substitution patterns are highly fluorescent (Φf ≈ 0.2),
with similar UVꢀvis and fluorescence spectra that system-
atically shift according to substituent effects on their
HOMOꢀLUMO gaps. Excitation of the pushꢀpull com-
pound yields a charge-transfer state with properties inter-
mediate between those of the acyclic and symmetrical
macrocyclic thiophene-bridged analogs.
Figure 2. Optimized geometries of H(mP/Th)H0 and D(mP/
Th)A0 (PCM(cyclohexane)/B3LYP/6-31þG(d,p)), viewed from
above (top left) and from along the molecular long axis
(top right), and representations of the frontier molecular orbi-
tals (PCM/CAM-B3LYP/6-311þG(2d,2p)//PCM/B3LYP/
6-31þG(d,p)).
Table 2. HOMO/LUMO Levels, First-Excited State Energies
and Compositions, and Experimental First Excitation Energies
HOMO LUMO ΔEHꢀL S0fS1 energy (eV) exptl
(eV)
(eV)
(eV)
[composition]a
(eV)
H(mP/Th)H0 ꢀ6.85 ꢀ1.41 5.44 3.10 [H f L (88%)] 3.02
D(mP/Th)D0 ꢀ6.46 ꢀ1.21 5.25 2.99 [H f L (85%)] 2.90
A(mP/Th)A0 ꢀ7.28 ꢀ2.32 4.96 2.82 [H f L (83%)] 2.73
D(mP/Th)A0 ꢀ6.76 ꢀ2.06 4.70 2.81 [H f L (60%)] 2.69
a H = HOMO, L = LUMO.
However, in contrast to these fluorescence properties,
the UVꢀvis spectra of D(mP/Th)A are a much better
match to those of macrocycle D(Th)2A. In particular, they
share a weak low energy transition above 450 nm (ε ≈
5000 Mꢀ1 cmꢀ1). The lower energy of this HOMOꢀLUMO
absorption in the macrocycles appears to be due to the
stabilization of the LUMO through extension of the
conjugation into the second bridge, irrespective of its
precise structure. We originally rationalized the weak
Acknowledgment. This work was supported by the Air
Force Office of Scientific Research (FA9550-10-1-0377).
We thank the reviewers for helpful comments that have
improved the structure of the manuscript.
Supporting Information Available. Supplemental fig-
ures, experimental details and procedures, NMR spectra
of synthesized compounds, and Cartesian coordinates of
optimized geometries. This information is available free
of charge via the Internet at http://pubs.acs.org.
(21) Reddinger,J.L.;Reynolds,J.R.Macromolecules 1997,30, 479–481.
(22) Verhoeven, J. W.; van Ramesdonk, H. J.; Groeneveld, M. M.;
Benniston, A. C.; Harriman, A. ChemPhysChem 2005, 6, 2251–2260.
(23) Patoux, C.; Coudret, C.; Launay, J.-P.; Joachim, C.; Gourdon,
A. Inorg. Chem. 1997, 36, 5037–5049.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 14, 2013
3765