Effect of chromophore-charge distance on the energy transfer properties of water-soluble conjugated oligomers

Article abstract of DOI:10.1021/ja028961w

The synthesis of 1,4-bis(9,9′-bis(3″-(N,N,N-trimethylammonium)-propyl)-2 ′-fluorenyl)benzene tetrabromide (C₃), 1,4-bis(9,9′-bis(4″-(N,N,N-trimethylammonium)-butyl)-2 ′-fluorenyl)benzene tetrabromide (C₄), 1,4-bis(9,9′-bis(6″-(N,N,N-

Full text of DOI:10.1021/ja028961w

Published on Web 05/10/2003
Effect of Chromophore-Charge Distance on the Energy
Transfer Properties of Water-Soluble Conjugated Oligomers
Bin Liu, Brent S. Gaylord, Shu Wang, and Guillermo C. Bazan*
Contribution from the Institute for Polymers and Organic Solids, Departments of Chemistry and
Materials, UniVersity of California, Santa Barbara, California 93106
Received October 15, 2002; E-mail: bazan@chem.ucsb.edu
Abstract: The synthesis of 1,4-bis(9,9′-bis(3′′-(N,N,N-trimethylammonium)-propyl)-2′-fluorenyl)benzene
tetrabromide (C3), 1,4-bis(9,9′-bis(4′′-(N,N,N-trimethylammonium)-butyl)-2′-fluorenyl)benzene tetrabromide
(C4), 1,4-bis(9,9′-bis(6′′-(N,N,N-trimethylammonium)-hexyl)-2′-fluorenyl)benzene tetrabromide (C6), and 1,4-
bis(9,9′-bis(8′′-(N,N,N-trimethylammonium)-octyl)-2′-fluorenyl)benzene tetrabromide (C8) is reported. Fluo-
rescence energy transfer experiments between C3-C8 and the acceptors pentasodium 1,4-bis(4′(2′′,4′′-
bis(butoxysulfonate)-styryl)styryl)-2-(butoxysulfonate)-5-methoxybenzene (3), fluorescein labeled single-
stranded DNA and fluorescein labeled double-stranded DNA in water, buffer, and methanol reveal the
importance of hydrophobic and electrostatic forces in determining chromophore-chromophore close
proximity. In water, the oligomers with longer side chain length show better energy transfer, as well as
higher Stern-Volmer quenching constants (Ksv), largely due to a stronger hydrophobic attraction between
the optically active components. In methanol, the differences in energy transfer are leveled, and the oligomers
with shorter side chain lengths show higher Ksv values. Compounds C3, C4, C6, and C8 were also used to
dissect the different contributors to DNA hybridization assays based on cationic conjugated polymers.
simplified protocols.12 Homogeneous methods, in particular,
Introduction
require less manipulation after exposure to the target DNA and
avoid difficulties associated with attaching the probe to a
surface.13 Assays based on Fo¨rster energy transfer, also known
as fluorescence resonance energy transfer (FRET), in which the
distance between a donor and an acceptor changes when the
target strand is present, have been used for some time.12
Traditional energy/electron transfer pairs for strand specific
assays require dual modification of the same DNA strand, which
can result in singly labeled impurities, lower yields, and more
difficult probe purification protocols.14,15 These considerations
make the development of simple and economic methods that
utilize the ease of homogeneous fluorescence assays with
minimal DNA modification highly desirable.
Novel techniques to determine DNA hybridization and
sequence characterization are under intense investigation.1 They
find use in many applications, including medical diagnostics,
examination of gene mutations, and drug delivery monitoring.2
DNA detection methods are expected to increase in importance,
as samples from patients will need to be quickly screened and
compared to genomic databases. New techniques under devel-
opment include homogeneous fluorescence-based DNA hybrid-
ization assays,3 DNA microarray technology which relies on
the hybridization between DNA sequences on a microarray
surface,4 the use of semiconductor crystals or quantum dots as
fluorescent probes,5-7 nanoparticle-amplified surface plasmon
resonance (SPR) techniques,8-10 and the use of redox-active
nucleic acid replica for amplified bioelectrocatalytic DNA
detection.11
Work in the field of gene delivery has elucidated the strong
interaction between cationic polyelectrolytes and negatively
charged DNA strands.16 The spontaneous formation of these
interpolymer complexes is thought to be primarily the result of
cooperative electrostatic forces; however, the importance of
Fluorescence-based detection methods are well established
but continue to be optimized for enhanced sensitivity and for
(1) (a) Dubertret, B.; Calame, M.; Libchaber, A. Nat. Biotechnol. 2001, 19,
365. (b) Wang, J. Nucleic Acids Res. 2000, 28, 3011. (c) Cardullo, R. A.;
Agrawal, S.; Flores, C.; Zamecnik, P. C.; Wolf, D. E. Proc. Natl. Acad.
Sci. U.S.A. 1988, 85, 8790. (d) Umek, R. M.; Lin, S. W.; Vielmetter, J.;
Terbrueggen, R. H.; Irvine, B.; Yu, C. J.; Kayyem, J. F.; Yowanto, H.;
Blackburn, G. F.; Farkas, D. H.; Chen, Y. P. J. Mol. Diagn. 2001, 3, 74.
(e) Schork, N. J.; Fallin, D.; Lanchbury, J. S. Clin. Genet. 2000, 58, 250.
(2) Balakin, K. V.; Horshun, V. A.; Mikhalev, I. I.; Maleev, G. V.; Malakhov,
A. D.; Prokhorenko, J. A.; Berlin, Yu. A. Biosens. Bioelectron. 1998, 13,
771.
(3) Sueda, S.; Yuan, J.; Matsumoto, K. Bioconjugate Chem. 2002, 13, 200
and references therein.
(4) Niemeyer, C. M.; Blohm, D. Angew. Chem., Int. Ed. 1999, 38, 2865.
(5) Gerion, D.; Parak, W. J.; Williams, S. C.; Zanchet, D.; Micheel, C. M.;
Alivisatos, A. P. J. Am. Chem. Soc. 2002, 124, 7070.
(8) He, L.; Musick, M. D.; Nicewarner, S. R.; Salinas, F. G.; Benkovic, S. J.;
Natan, M. J.; Keating, C. D. J. Am. Chem. Soc. 2000, 122, 9071.
(9) Nelson, B. P.; Grimsrud, T. E.; Liles, M. R.; Goodman, R. M.; Corn, R.
M. Anal. Chem. 2001, 73, 1.
(10) Tombelli, S.; Minunni, M.; Mascini, M. Anal. Lett. 2002, 35, 599.
(11) (a) Patolsky, F.; Weizmann, Y.; Willner, I. J. Am. Chem. Soc. 2002, 124,
770. (b) Caruana, D. J.; Hellerr, A. J. Am. Chem. Soc. 1999, 121, 769.
(12) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed.; Kluwer
Academic/Plenum Publisher: New York, 1999.
(13) Paris, P. L.; Langenhan, J. M.; Kool, E. T. Nucleic Acids Res. 1998, 26,
3789.
(14) Knemeyer, J.; Marm, N.; Sauer, M. Anal. Chem. 2000, 72, 3717.
(15) Castro, A.; Williams, J. G. K. Anal. Chem. 1997, 69, 3915.
(16) Kabanov, A. V., Felgner, P., Seymour, L. W., Eds. Self-Assembling
Complexes for Gene DeliVery. From Laborotary to Clinical Trial; John
Wiley: Chichester, 1998.
(6) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science
1998, 281, 2013.
(7) Chan, W. C. W.; Nie, S. M. Science 1998, 281, 2016.
9
10.1021/ja028961w CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 6705-6714
6705

A R T I C L E S
Liu et al.
Scheme 1. Molecular Structures of 1-3
Scheme 2. DNA Hybridization Assay Based on FRET from a
Water-Soluble Cationic Conjugated Polymer to a
Fluorophore-Tagged ssDNA Probe
hydrophobic interactions between aromatic polymer units and
DNA bases has also been recognized.17 The overall free energy
of these interactions is controlled by the molecular structure of
the participating species, in conjunction with variables such as
pH, ionic strength, and temperature.18-20 Despite the complexity
of these interactions, synthetic polycations have recently been
shown to recognize the tertiary structure of plasmid DNA.18
Cationic polyelectrolytes have also been studied because of their
effect on the fluorescence of oppositely charged acceptor
complexes.21,22 After complexation with DNA, positively charged
synthetic polyelectrolytes change the fluorescence emission and
energy transfer processes when the polyelectrolyte is tagged with
a chromophore such as pyrene.2
Conjugated polymers provide highly responsive optical
platforms for chemical and biological detection.23,24 Water-
soluble conjugated polymers25 are of particular interest for
reporting biological recognition events. Polymers such as the
anionic MPS-PPV (1, in Scheme 1) and the cationic poly(9,9-
bis(6′-N,N,N-trimethylammonium)hexyl)-fluorene phenylene)
(2) have shown extremely large quenching efficiencies in the
presence of oppositely charged acceptors.23,26,27
Cationic polymers such as 2 form part of a DNA-sensor
technique which utilizes the optical amplification of conjugated
polymers and the complexation of oppositely charged polyelec-
trolytes (Scheme 2).28,29 In this method, one excites the
conjugated polymer and compares the emission of a dye (C*)
attached to the probe single-stranded DNA (shown in red) to
that of the conjugated polymer (shown in black). The optical
properties of the components are chosen so that only the polymer
absorbs at the excitation frequency and the absorption of C*
overlaps the emission from the polymer (FRET condition).
Hybridization of the probe strand to its complementary strand
(shown in blue) results in a more efficient FRET ratio than
when a noncomplementary strand (shown in green) is added to
the solution. Furthermore, that the emission of the dye is stronger
when exciting the conjugated polymer than upon direct dye
excitation demonstrates the light-harvesting properties of the
conjugated polymer.24,30,31
In this contribution, we probe how the distance between the
positively charged trimethylalkylammonium group and the
conjugated chromophore influences the FRET to fluorophore-
tagged DNA. To eliminate inherent problems associated with
controlling polymer chain length and molecular weight distribu-
tions, we focus on well-defined conjugated oligomers with the
bis(fluorenyl)benzene structure. The general structure of these
molecules is shown below.
(17) Ganachaud, F.; Ela¨ıssari, A.; Pichot, C.; Laayoun, A.; Cros, P. Langmuir
1997, 13, 701.
(18) Bronich, T. K.; Nguyen, H. K.; Eisenberg, A.; Kabanov, A. V. J. Am. Chem.
Soc. 2000, 122, 8339.
(19) Kuhn, P. S.; Barbosa, M. C.; Yan, L. Physica A 2000, 283, 113.
(20) Wolfert, M. A.; Dash, P. R.; Nazarova, O.; Oupicky, D.; Seymour, L. W.;
Smart, S.; Strohalm, J.; Ulbrich, K. Bioconjugate Chem. 1999, 10, 993.
(21) Song, X. D.; Wang, H. L.; Shi, J.; Park, J. W.; Swanson, B. I. Chem. Mater.
2002, 14, 2342.
(22) Wang, D. L.; Gong, X.; Heeger, P. S.; Rininsland, F.; Bazan, G. C.; Heeger,
A. J. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 49.
We will examine the energy transfer process to a pentaanionic
chromophore (3 in Scheme 1), to establish the role of hydro-
(23) Chen, L.; Mcbranch, D. W.; Wang, H. L.; Helgeson, R.; Wudl, F.; Whitten,
D. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12287.
(24) McQuade, D. T.; Pullen, A. E.; Swager, T. M. Chem. ReV. 2000, 100, 2537.
(25) Pinto, M. R.; Schanze, K. S. Synthesis-Stuttgart 2002, 9, 1293.
(26) Wang, J.; Wang, D.; Miller, E. K.; Moses, D.; Bazan, G. C.; Heeger, A. J.
Macromolecules 2000, 33, 5153.
(29) For an alternate use of conjugated polymers in DNA detection, see: Ho,
H. A.; Boissinot, M.; Bergeron, M. G.; Corbeil, G.; Dore, K.; Boudreau,
D.; Leclerc, M. Angew. Chem., Int. Ed. 2002, 41, 1548.
(30) McQuade, D. T.; Hegedus, A. H.; Swager, T. M. J. Am. Chem. Soc. 2000,
122, 12389.
(27) Stork, M. S.; Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. AdV. Mater. 2002,
14, 361.
(28) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc. 2003, 125,
896.
(31) Yang, J. S.; Swager, T. M. J. Am. Chem. Soc. 1998, 120, 11864.
9
6706 J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003

Properties of Water-Soluble Conjugated Oligomers
A R T I C L E S
to aggregation at concentrations useful for good signal-to-noise
ratios (10-2-10-4 M). Therefore, the spectra for the oligomers
C3-C8 were measured at elevated temperatures in D2O.
1
Integration of the H NMR signals of C3 at 90 °C indicates
>99% quaternization, by comparison against the integrated
values for -CH2N and -NCH3 peaks at 3.52 and 3.30 ppm,
respectively. In the case of C4, C6, and C8, because the signals
for the methylene-N and the methylamino protons overlap, the
degree of quaternization was calculated by the ratio of -CH2N
and -NCH3 (∼3.40 ppm) to the characteristic methylene groups
(∼1.90 ppm) at the 9-position of fluorene. Further evidence of
the identity of C3-C8 was obtained by electrospray ionization
mass spectrometry (ESI-MS). Peaks corresponding to the correct
molecular structures were detected for the four oligomers.
Table 1 shows the absorption and emission data for oligomers
3 and C3-C8. For the neutral precursors 10-13, there is no
difference in the absorption maxima in chloroform (λmax ) 330
nm). In water, the absorption maxima range from 329 nm for
C3 to 336 nm for C8. The blue shift with decreasing distance to
the positive charge is likely due to a combination of electronic
effects and perhaps a decreased level of aggregation for oligomer
C3. Tether length, however, does not influence the fluorescence
properties of the chromophores; the emission of C3-C8 in water
occurs at ∼396 nm, with a calculated quantum efficiency of 65
( 5%.
phobic forces and electrostatic interactions. We next probe how
chromophore-tagged single-stranded DNA (ssDNA) and double-
stranded DNA (dsDNA) behave as FRET acceptors. Finally,
we compare situations where complementary and noncomple-
mentary strands are added to the ssDNA carrying the acceptor
chromophore to dissect the different forces operating in bio-
sensor protocols such as those shown in Scheme 2.
Results and Discussion
Synthesis and Characterization. The synthetic approach to
oligomers, 1,4-bis(9,9′-bis(3′′-(N,N,N-trimethylammonium)-pro-
pyl)-2′-fluorenyl)benzene tetrabromide (C3), 1,4-bis(9,9′-bis(4′′-
(N,N,N-trimethylammonium)-butyl)-2′-fluorenyl)benzene tetra-
bromide (C4), 1,4-bis(9,9′-bis(6′′-(N,N,N-trimethylammonium)-
hexyl)-2′-fluorenyl)benzene tetrabromide (C6), and 1,4-bis(9,9′-
bis(8′′-(N,N,N-trimethylammonium)-octyl)-2′-fluorenyl)ben-
zene tetrabromide (C8), is shown in Scheme 3. Treatment of
2-bromofluorene with 50% KOH, followed by reaction with
excess R,ω-dibromoalkane, provides compounds 2-bromo-9,9′-
bis(3′′-bromopropyl)fluorene (4), 2-bromo-9,9′-bis(6′′-bromo-
hexyl)fluorene (5), and 2-bromo-9,9′-bis(8′′-bromooctyl)fluorene
(6) in yields of 35-80%. Instead of forming the desired
2-bromo-9,9′-bis(4′-bromobutyl)fluorene, the direct treatment
of 2-bromofluorene with KOH and 1,4-dibromobutane produces
2-bromo-9,9-gem-cyclopentyl cleanly. To circumvent this com-
plication, 1,4-dibromobutane was first treated with 1 equiv of
4-methylphenol to give 4-bromo-1-(4′-methylphenoxy)butane
(7) in approximately 80% yield.32 An excess of 7 was
subsequently reacted with deprotonated 2-bromofluorene to form
intermediate 8, which can be subsequently transformed into
2-bromo-9,9′-bis(4′′-bromobutyl)fluorene (9) by treatment with
hydrobromic acid in acetic acid under refluxing conditions over
a period of 2 days.32
Table 1 also contains the data for absorption and emission
maxima of the oligomers in methanol. The emission occurs at
the same frequency for the four oligomers. We note that the
fluorescence efficiency of oligomers C3-C8 in methanol (70
( 5% using 9,10-diphenylanthracene in cyclohexane as the
standard) is slightly higher than that measured in water (Table
1), consistent with less aggregation in methanol, relative to
water.
The synthesis of pentasodium 1,4-bis(4′(2′′,4′′-bis(butoxysul-
fonate)-styryl)styryl)-2-(butoxysulfonate)-5-methoxybenzene (3
in Scheme 1) has been reported previously.33
Fo1rster Energy Transfer between C3-C8 and 3. The
absorption spectrum of 3 in water overlaps the emission spectra
of oligomers C3-C8 (Table 1); thus efficient Fo¨rster energy
transfer should be expected between these molecules.12,27 Indeed,
when 3 is added to the oligomer solutions, one observes a
decrease in the emission of C3-C8, with a concomitant increase
in the emission of 3. Figure 1 shows the fluorescence changes
upon addition of 3 to C8 ([C8] ) 2.0 × 10-6 M), [3] varies
from 3.0 × 10-8 to 1.0 × 10-6 M). Similar changes in emission
spectra were obtained with solutions of C3-C6. In these
mixtures, 3 acts as the acceptor, and the 1,4-bis(fluorenyl)-
benzene chromophore acts as the donor.
Coupling of 2 equiv of 4, 5, 6, or 9 with 1,4-phenyldiboronic
acid under Suzuki coupling conditions using Pd(PPh3)4 in
refluxing THF/H2O (2:1) over 36 h gives oligomers 1,4-bis-
(9,9′-bis(3′′-bromopropyl)-2′-fluorenyl)benzene (10), 1,4-bis-
(9,9′-bis(4′′-bromobutyl)-2′-fluorenyl)benzene (11), 1,4-bis(9,9′-
bis(6′′-bromohexyl)-2′-fluorenyl)benzene (12), and 1,4-bis(9,9′-
bis(8′′-bromooctyl)-2′-fluorenyl)benzene (13) in 45-60% yields.
Formation of C3, C4, C6, and C8 from 10, 11, 12, and 13,
respectively, was achieved by using condensed trimethylamine
in THF/H2O (2:1) mixture. As compared to our previous report,
which involved addition of a methylhalide, such as MeI, to 1,4-
bis(9,9′-bis(6′′-(N,N-dimethylamino)hexyl)-2′-fluorenyl)ben-
zene, the new procedure bypasses the need to prepare trialkyl-
amino-substituted oligomers.27 The alkylamino intermediates are
highly polar and adsorb strongly to chromatographic supports,
which makes their purification considerably more complicated
and reduces reaction yields.
Stern-Volmer analysis gives insight into the efficiency of
fluorescence quenching or energy transfer according to the
equation:
PL0/PL ) 1 + Ksv[Q]
where PL0 and PL correspond to the integrated emission of the
bis(fluorenyl)benzene chromophore in the absence and presence
of 3, respectively.12 A plot of PL0/PL versus [3] in water, with
[C3-C8] ) 2.0 × 10-6 M, is shown in Figure 2.
The 1H NMR spectroscopy signals of C3-C8 at room
temperature are broad and indistinguishable, presumably due
(33) Gaylord, B. S.; Wang, S.; Heeger, A. J.; Bazan, G. C. J. Am. Chem. Soc.
2001, 123, 6417.
(32) Rau, I. U.; Rehahn, M. Acta Polym. 1994, 45, 3.
9
J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003 6707

A R T I C L E S
Liu et al.
a
Scheme 3. General Synthetic Scheme for the Preparation of Compounds C3-C8
a (i) KOH/H2O/Br(CH2)nBr; (ii) K2CO3/18-crown-6/Br(CH2)4Br; (iii) KOH/H2O/7; (iv) HBr/CH3COOH; (v) 1,4-(B(OH)2)2-C6H4/Pd(PPh3)4/Na2CO3/
THF/H2O; (vi) NMe3/THF/H2O.
Table 1. Absorption and Fluorescence Data
c
λ
max (nm)a
λ
max (nm)b
QE (±5%)
K (M-1
)
sv
compound water methanol water methanol water methanol
water
methanol
3
410
329
330
334
336
413
332
332
332
332
504 477, 507 10% 56%
C3
C4
C6
C8
394 373, 393 64% 76% 9.02 × 105 1.81 × 106
395 373, 393 66% 75% 9.26 × 105 1.75 × 106
396 373, 393 66% 75% 1.52 × 106 1.54 × 106
396 373, 393 65% 74% 2.07 × 106 1.46 × 106
a Absorption. b Emission. c Using 9,10-diphenylanthracene as the stan-
dard.
In the Stern-Volmer plots, the fluorescence intensity was
corrected for any changes in absorbance. A linear region is
observed at low quencher concentrations, followed by an
upsloping nonlinear region where the energy transfer may be
described by a “sphere of action” mechanism.26 From the slope
of the linear region ([3] ) 0 to 3 × 10-7 M) of the plots, one
obtains the Stern-Volmer constant Ksv, which provides a
measure of the rate for quenching or energy transfer. In water,
oligomer C8 exhibits the highest Ksv (2.07 × 106 M-1, Table
1). The next highest value corresponds to C6 (1.52 × 106 M-1),
while the difference between C3 and C4 falls within experimental
uncertainty (9.02 × 105 and 9.26 × 105 M-1). At the same
concentration of 3, C8 gives the highest emission intensity from
the acceptor. These differences are illustrated in Figure 3, which
shows the emission spectra from solutions containing [3] ) 7.9
× 10-7 M and [C3-C8] ) 2.0 × 10-6 M. These studies show
that in water the molecules with the longest tether give rise to
the most efficient energy transfer process.
Figure 1. Emission spectra for mixtures of C8 with 3 in water. [C8] ) 2.0
× 10-6 M, [3] ) 3.0 × 10-8 M to 1.0 × 10-6 M.
As reported earlier for 3 and 1,4-bis(9,9′-bis(6′′-(N,N,N-
trimethylammonium)hexyl)-2′-fluorenyl)benzene tetraiodide, that
is, oligomer C6 with iodide counteranions, the fluorescence
quenching mechanism is dominated by the formation of a
ground-state complex.27 This interaction is favored by the ion
pairing between the positively charged donor and the negative
acceptor. Oligomer 3, however, aggregates in water, giving rise
to large (∼190 nm) clusters, even at concentrations as low as
10-5 M.33 The formation of these aggregates is driven by the
hydrophobic nature of the conjugated chromophore framework.
It is reasonable to expect a similar contribution from hydro-
9
6708 J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003

Properties of Water-Soluble Conjugated Oligomers
A R T I C L E S
Scheme 4. Schematic Diagram Which Shows the Preferred
Average Intermolecular Associations in Water and in Methanol
size, electronic communication between subunits in the ag-
gregate, and the proximity between donor and acceptor.
However, we propose a simple picture that accounts for the
trends by focusing on the two extreme cases of C3 and C8. The
larger hydrophobic content of C8, relative to C3, increases the
aggregate size in water and also the hydrophobic interactions
between the optically active parts of 3 and C8. We show the
relationship in Scheme 4, with the charges pairing up and the
organic portions coming together in aqueous media. In methanol,
the hydrophobic contacts between the organic units are con-
siderably reduced, and the electrostatic interactions dominate
the order of rank of Ksv magnitudes. The more efficient energy
transfer from C3 results from its shorter tether length, which
brings the two chromophores closer together, and the 1/r6
dependence of Fo¨rster energy transfer.12
Figure 2. Stern-Volmer plot in water for [C3-C8] ) 2.0 × 10-6 M with
[3] ) 0 to 5.0 × 10-6 M. The emission spectra were measured by exciting
at the absorption maximum of C3-C8.
Comparison of Figure 4 to Figure 1 shows more emission
from 3 in methanol, relative to that in water. These observations
are consistent with the higher quantum efficiency of 3 in
methanol (56 ( 5%), relative to that in water (10 ( 5%).
Energy Transfer Studies between Oligomers and ssDNA-
C*. As discussed in the Introduction, our interest in probing
the basic energy transfer characteristics of C3-C8 stems from
the use of cationic conjugated polymers in DNA hybridization
assays.28 We now examine the effect of the tether length on
fluorescence quenching with a ssDNA labeled with an acceptor
chromophore.35 The oligonucleotide used is 5′-Fl-CCA ATC
AGT CCA GTG ATA CG (ssDNA-C*), with fluorescein (C*)
at the 5′ position. The spectra in Figure 5 show the overlap
between the emission of C3-C8 and the absorption of ssDNA-
C* in the 350-530 nm range. While this spectral overlap is
not as extensive as that between C3-C8 and 3, it is sufficient
to examine the dependence of FRET on the molecular structure
and solvent medium.
Only the C6/ssDNA-C* and C8/ssDNA-C* combinations
show sufficient C* emission to be detected using a standard
fluorometer (Figure 6). Measurements were carried out in buffer
(0.1 M sodium chloride + 0.01 M sodium citrate) and at a fixed
ssDNA-C* concentration (2.0 × 10-8 M), with the donor
concentration varying from 2.0 × 10-8 to 2.0 × 10-7 M. The
emission maxima of the C* in the C6/ssDNA-C* and C8/
ssDNA-C* mixtures are red-shifted (λ ) 527 nm) relative to
its emission when excited directly in the absence of the donor
oligomers (λ ) 517 nm). This emission red shift indicates
Figure 3. FRET comparison from C3-C8 to 3 in water. [C3-C8] ) 2.0
× 10-6 M and [3] ) 7.9 × 10-7 M. The emission spectra were measured
by exciting at the absorption maximum of C3-C8.
phobic forces in the fluorescence quenching of C3-C8 by 3.
Furthermore, the larger molecular volume is expected to provide
diminished charge density in the case of C8.34 The enhanced
energy transfer observed for C8 strongly suggests that the larger
hydrophobic component in C8 is responsible for closer contacts
with 3, when compared to C3.
To further examine the importance of hydrophobic forces and
the role of the medium, we evaluated the Ksv constants in
methanol. This solvent is sufficiently polar to dissolve the
charged chromophores, but, because of its organic nature, it
should decrease the aggregation of the organic fragments. Table
1 shows that the Ksv values are of the same magnitude as those
obtained in water. However, the order of Ksv values is reversed;
C3 is quenched more readily, whereas C8 shows the smallest
Ksv.
That the sequence of Ksv magnitude for the oligomers follows
opposite structural trends in water and methanol highlights the
interplay between hydrophobic and electrostatic forces. It is
reasonable that, with increased side chain length, the oligomer
becomes more hydrophobic. The differences in Ksv between
methanol and water as a function of tether length can be
rationalized in terms of several phenomena such as aggregation
(35) Fluorescein was chosen because it is one of the most ubiquitous dyes used
in FRET experiment and because it can be easily attached to DNA. Its
fluorescence and absorbance properties are known to be pH dependent,
and the highest quantum yield is obtained for the fluorescein dianion. The
pH of the buffer (0.1 M sodium chloride + 0.01 M sodium citrate) used
for this study is about 8.3. In our experiment, the quantum yield of ssDNA-
C* was determined to be 94%, using 9,10-diphenylanthracene as the
standard.
(34) Israelachvili, J. Intermolecular & Surface Forces, 2nd ed.; Harcourt Brace
& Company, Publisher: Orlando, FL, 1992.
9
J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003 6709

A R T I C L E S
Liu et al.
Figure 4. Emission spectra for mixtures of C8 with 3 in methanol, [C8] )
2.0 × 10-6 M, [3] ) 3.0 × 10-8 to 1.2 × 10-6 M.
Figure 7. FRET comparison for solutions containing ssDNA-C* and C6
or C8 in buffer ([ssDNA-C*] ) 2.0 × 10-8 M, [C6] ) [C8] ) 8.0 × 10-8
M). The fluorescence spectra were measured by exciting at 345 nm.
Figure 5. Spectral overlap between the photoluminescence of C3 (a), C6
(b), and C8 (c) and the absorbance of ssDNA-C* (d).
Figure 8. Effect of ionic strength on the FRET between C3 and ssDNA-
C*. The emission of C3 is normalized for comparison. The concentration
of NaCl ranges from 0 to 0.1 M, with a ) 0 M, b ) 0.02 M, and c ) 0.1
M.
Figure 7 shows a direct comparison of the emission in buffer
upon excitation of C6 or C8 in the presence of ssDNA-C*. The
C* emission is 6 times more intense for C8/ssDNA-C* than
for C6/ssDNA-C*. Under similar [C8] or [C6], there is no
fluorescein emission when free fluorescein is used. This
demonstrates that the ssDNA structure plays an important role
in bringing the donor and C* together.
Energy transfer is observed from the four oligomers to the
ssDNA-C* in pure water. The absence of buffer ions under
these conditions allows electrostatic forces to bring the two
opposite charged molecules into close proximity.26 Addition of
NaCl to C3/ssDNA-C* solutions results in a decrease of C*
emission and an increase of C3 emission (Figure 8). Note that
the data in Figure 8 are normalized relative to oligomer emission
to highlight the decrease in C* emission. At 0.1 M NaCl, C*
emission is absent, indicating ionic screening of the C3/DNA-
C* attraction. The C* emission under these conditions, obtained
by direct excitation, occurs at 517 nm, giving further evidence
of the absence of complexation. For the C8/ssDNA-C* mixture,
when the NaCl concentration reaches 0.2 M, only 30% of the
emission from fluorescein is lost, and it is still detectable even
when the NaCl concentration reaches 1.07 M (Figure 9).
Figure 6. Emission for the C3/ssDNA-C* (a and b) or C8/ssDNA-C* (c
and d) solutions obtained by exciting at two different wavelengths, 345 nm
(a and c) and 480 nm (b and d). The conditions for this comparison are
[ssDNA-C*] ) 2.0 × 10-8 M, [C3] ) [C8] ) 8.0 × 10-8 M. The emission
for ssDNA-C* (e) is also shown for comparison (excited at 480 nm). Note
that the spectra in (b) and (e) overlap each other exactly.
complex formation and the close proximity of the cationic
structure, which changes the polarity near fluorescein. Direct
excitation at 480 nm in the presence of C3 or C4 results in
emission at 517 nm. Thus, in buffer, the C3/ssDNA-C* and C4/
ssDNA-C* combinations do not lead to complex formation.
9
6710 J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003

Properties of Water-Soluble Conjugated Oligomers
A R T I C L E S
Figure 10. FRET comparison for solutions containing dsDNA2-C* and
C3, C6, or C8 in buffer. [dsDNA2-C*] ) 2.0 × 10-8 M, [Cn] ) 1.6 × 10-7
M. The excitation wavelength was 345 nm for all spectra.
Figure 9. Effect of ionic strength on the FRET between C8 and ssDNA-
C*. The emission of C8 is normalized for comparison. The concentration
of NaCl ranges from 0 to 1.07 M, with a ) 0 M, b ) 0.20 M, c ) 0.46 M,
and d ) 1.07 M.
a new single-stranded DNA probe sequence, 5′-Fl-ATC TTG
ACT ATG TGG GTG CT, which would not quench the
emission of the fluorescein upon hybridization. Indeed, the direct
excitation of the new double-stranded DNA (dsDNA2-C*) shows
a fluorescence efficiency for fluorescein which is only 10%
lower than the ssDNA2-C*. We therefore focused our energy
transfer studies on dsDNA2-C*.
Figure 10 shows that, in buffer, the ability to energy transfer
to dsDNA2-C* follows the order C8 > C6, with no FRET
observed with C3 (or C4, not shown). This dependence on
molecular structure is similar to that observed with the ssDNA
experiments. Upon excitation at 345 nm, the emission maximum
of C8/dsDNA2-C* is 18 nm red-shifted relative to the direct
excitation of the dsDNA2-C* in the absence of the donor. At
concentrations of [dsDNA2-C*] ) 2.0 × 10-8 M and [C8] )
3.0 × 10-7 M, the integrated C* emission is 3-fold greater than
that obtained by directly exciting the chromophore at its
maximum absorption in the absence of the donor.
We now dissect the DNA biosensor process shown in Scheme
2 into its specific components. Figure 11 shows the emission
from solutions of [C8] ) 1.2 × 10-7 M when in the presence
of [ssDNA2-C*] ) 2.0 × 10-8 M or [dsDNA2-C*] ) 2.0 ×
10-8 M. A higher FRET ratio is observed for ssDNA2-C* (10.6)
than for dsDNA2-C* (1.4). These data are consistent with a
stronger attraction between C8 and ssDNA2-C*. When [C8]
increases, the FRET ratios for dsDNA-C* and ssDNA-C*
become more equal. As shown by Figure 12, by the time that
[C8] ) 2.3 × 10-7 M and [ssDNA2-C*] ) 2.0 × 10-8 M or
[dsDNA2-C*] ) 2.0 × 10-8 M, the FRET ratios are nearly the
same.
Structural differences between dsDNA and ssDNA account
for the concentration effect on FRET. At low concentrations,
there is a greater affinity between ssDNA and C8 as a result of
electrostatic interactions and the exposed hydrophobic portion
of ssDNA. At higher concentration regimes, a larger number
of donor molecules can approach the vicinity of dsDNA because
it has a higher number of charges. Consistent with this idea is
the fact that after the ratio of C8 to C* is 6:1 (i.e., approximately
a 1:1 charge ratio for ssDNA), the C* emission intensity ceases
In pure water or in buffer (0.1 M NaCl + 0.01 M sodium
citrate), the maximum emission intensity from C* obtained by
energy transfer from C8 was achieved when the ratio of cationic
to anionic charges was slightly larger than 1 (1.1-1.4). In these
experiments, the ssDNA-C* concentration was kept constant,
and C8 was added incrementally. Once the charge ratio increases
beyond ∼1.2, the C8 emission increases sharply, and only a
small enhancement occurs for the C* emisson. Indeed, the
emission of C8 in buffer was essentially quenched until the
charge ratio reaches 1.2. Such a charge ratio dependence on
dye emission intensity should be expected because when [C8]/
[ssDNA-C*] < 1, not all of the ssDNA-C* strands can be in
close proximity to the donor oligomer units. Conversely, in the
[C8]/[ssDNA-C*] > 1 regime, not all of the photons harnessed
by 1 (the donor) can be transferred to the ssDNA-C* acceptor.
A similar dependence of charge ratio on C* intensity was
observed when examining the efficiency of energy transfer from
the cationic conjugated polymer 2 and DNA-C* or PNA-C*
(PNA ) peptide nucleic acid, C* ) fluorescein) hybridized with
DNA.28,36
Energy Transfer between C8 and Double-Stranded DNA
(dsDNA). When ssDNA-C* is hybridized with a complementary
strand, the resulting double-stranded DNA (dsDNA) has a more
rigid and more compact structure and exhibits an increased
charge density.37 In double-stranded DNA, the hydrophobic
bases are packed inside the helix, which minimizes external
hydrophobic interactions. The exposed bases in single-stranded
DNA, however, more readily enable hydrophobic interactions.38
Initial hybridization of the ssDNA-C* oligonucleotide probe
described in the previous section (5′-Fl-CCA ATC AGT CCA
GTG ATA CG) with the complementary strand leads to a
dsDNA-C*, in which the C* emission, upon direct excitation
at 480 nm, is 55% of the emission observed from ssDNA-C*.
This efficiency loss is due to the quenching by the guanine in
the complementary strand, which is spatially adjacent to C* after
hybridization.39,40 To examine the role of dsDNA, we designed
(36) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 10954.
(37) Pullman, B.; Lavery, R.; Pullman, A. Eur. J. Biochem. 1982, 124, 229.
(38) Diogo, M. M.; Queiroz, J. A.; Monteiro, G. A.; Martins, S. A. M.; Ferreira,
G. N. M.; Prazeres, D. M. F. Biotechnol. Bioeng. 2000, 68, 576.
(39) Knemeyer, J. P.; Marme´, N.; Sauer, M. Anal. Chem. 2000, 72, 3717.
(40) Torimura, M.; Kurata, S.; Yamada, K.; Yokomaku, T.; Kamagata, Y.;
Kanagawa, T.; Kurane, R. Anal. Sci. 2001, 17, 155.
9
J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003 6711

A R T I C L E S
Liu et al.
Figure 11. Emission spectra from solutions containing C8 and ssDNA2-
C* (a) and dsDNA2-C* probes (b) in buffer ([ssDNA2-C*] or [dsDNA2-
C*] ) 2.0 × 10-8 M, [C8] ) 1.2 × 10-7 M). The excitation wavelength
was 345 nm for all spectra.
Figure 13. Emission spectra from solutions containing C8 and (a) a
hybridized (dsDNA2-C*) and (b) a nonhybridized (ssDNA2-C* + 5′-GAC
TCA ATG GCG TTA GAC TG) ssDNA pair in buffer (([ssDNA2-C*] or
[dsDNA2-C*] ) 2.0 × 10-8 M, [C8] ) 2.3 × 10-7 M). The spectra are
normalized with respect to the C8 emission.
fragments.27 Additionally, direct quaternization of the side chains
of the fully assembled chromophores simplifies the overall
synthetic scheme by one step. The compounds C3, C4, C6, and
C8 provide a series of chromophores that modulates the
separation of charged side groups from a common conjugated
framework, resulting in varying levels of hydrophobic content.
FRET experiments in water between the tetracationic C3-
C8 and the pentaanionic chromophore 3 reveal that the most
efficient energy transfer occurs with C8. This result seems at
first counterintuitive, if one considers the distance dependence
of FRET, together with the longer tether between the chro-
mophore and the charged quaternary nitrogen, and highlights
the importance of hydrophobic interactions in this class of
compounds. Polar organic compounds, such as methanol, reduce
the attraction between the hydrophobic fragments, and one
observes a FRET efficiency that increases with shorter thethers;
C3 is most effective.
Figure 12. Emission spectra from solutions containing C8 and (a) ssDNA2-
C* and (b) dsDNA2-C* in buffer ([ssDNA2-C*] or [dsDNA2-C*] ) 2.0 ×
10-8 M, [C8] ) 2.3 × 10-7 M). The excitation wavelength used was 345
nm.
Similar to 3, single-stranded DNA provides sites for elec-
trostatic and hydrophobic interactions. In the buffer conditions
required for DNA-DNA hybridization, we find that only C6
and C8 show efficient FRET to the ssDNA tagged with
fluorescein (ssDNA-C*). Increasing electrolyte strength reduces
the Coulombic attraction between donor and acceptor molecules,
in agreement with standard Debye-Hu¨ckel Theory.34 That in
water one observes energy transfer from C3-C8 to ssDNA-C*
demonstrates that the buffer ions screen electrostatic forces, to
the point that they are insufficient for bringing the less
hydrophobic chromophores C3 and C4 within FRET distance.
The effect of salt addition on electrostatic screening in Figures
8 and 9 shows the difference in the binding strength for C3/
ssDNA-C* and C8/ssDNA-C*. This dependence is in agreement
with Bloomfield’s calculation to determine the roles of elec-
trostatics and hydrophobicity in the binding of cationic lipids
to DNA. From their calculations, every additional methylene
group in an aliphatic lipid chain increases the association
constant about 4-fold by increasing the binding cooperativity
with DNA.41 It is also noteworthy that double-stranded DNA
to grow for ssDNA, but continues to increase in the case of
dsDNA. Additional charge units remain with the dsDNA
structure, which are capable of attracting more C8 units.
However, even after a 12:1 ratio of C8 to C*, the FRET to
ssDNA is slightly higher.
We can now examine how these solutions behave as sensors
by comparing the FRET of C8 to dsDNA2-C* against the FRET
to ssDNA2-C* in the presence of a noncomplementary single-
stranded DNA (5′-GAC TCA ATG GCG TTA GAC TG). As
shown in Figure 13, more efficient energy transfer occurs in
the case of dsDNA2-C*. The lower FRET for the noncomple-
mentary strand thus must arise from a competition between
ssDNA-C* and the unhybridized ssDNA, rather than a stronger
attraction between C8 and dsDNA-C*, relatiVe to ssDNA-C*.
Summary and Conclusion
In summary, Scheme 3 provides an improved synthetic entry
into cationic water-soluble conjugated oligomers. Suzuki cou-
pling of 2 equiv of the bromide derivatives 4, 5, 6, and 9 with
1,4-phenyldiboronic acid circumvents the difficult chromato-
graphic separation associated with amine-containing fluorene
(41) Matulis, D.; Rouzina, I.; Bloomfield, V. A. J. Am. Chem. Soc. 2002, 124,
7331.
9
6712 J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003

Properties of Water-Soluble Conjugated Oligomers
A R T I C L E S
Scheme 5. Molecular Structures of 2-C3 and 2-C8
temperature absorbance spectroscopy. Fluorescence intensities were
determined from the integrated areas under emission spectra of both
the donor and the acceptor fluorescein. The differences in energy
transfer were compared by measuring the fluorescence intensity of the
acceptor in the presence of the same concentration of the donor.
General Procedure for 2-Bromo-9,9′-bis(ω′′-bromoalkyl)fluorene
Compounds. 2-Bromofluorene (1.23 g, 5 mmol) was added to a mixture
of aqueous potassium hydroxide (100 mL, 50%), tetrabutylammonium
bromide (0.33 g, 1 mmol), and dibromoalkane (50 mmol) at 75 °C.
After 15 min, the mixture was cooled to room temperature. After
extraction with CH2Cl2, the combined organic layers were washed
successively with water, aqueous HCl (1 M), water, and brine and then
dried over MgSO4. After removal of the solvent and the excess 1,6-
dibromoalkane, the residue was purified by silica gel column chroma-
tography using hexane and chloroform (9:1) as the solvent.
“hides” the hydrophobic component within its helical structure.
However, the hydrophobic interactions remain important in
determining the FRET efficiency between dsDNA-C* and the
donors because only C6 and C8 result in measurable FRET when
in the presence of buffer.
We have also determined that in these solutions the energy
transfer between C8 and ssDNA is more efficient than that
between C8 and dsDNA (Figure 11), perhaps as a result of
ssDNA’s more exposed hydrophobic component and more
flexible structure. With this information, we have shown that
the success of Scheme 2 as a sensor method does not rely on
stronger interactions between cationic conjugated polymers and
ssDNA or dsDNA. Rather, more efficient FRET occurs with
dsDNA-C* because in situation B the nonhybridized strands
remain in solution and screen the interaction between the
conjugated polymer and the tagged ssDNA-C*.
These findings also open up possibilities and guidelines for
the design of new biosensor materials. For example, it would
seem reasonable that a polymer such as 2-C3 (Scheme 5) would
be more sensitive to variations in analyte charge than 2-C8, for
which hydrophobic interactions would be less discriminating.
If so, one could generate homogeneous sensory optical reporters
that specifically respond to changes in the charge of the target
analyte. These studies are underway in our laboratories.
2-Bromo-9,9′-bis(3′′-bromopropyl)fluorene (4). Following the
procedure described above, we obtained compound 4 as a white solid
1
in a yield of 35%. H NMR (200 MHz, CDCl3) δ: 1.09-1.13 (br,
4H), 2.05-2.18 (m, 4H), 3.07-3.14 (t, 4H), 7.35-7.37 (m, 3H), 7.49-
7.51 (m, 2H), 7.53-7.58 (m, 1H), 7.60-7.72 (m, 1H). 13C NMR (50
MHz, CDCl3) δ: 27.15, 34.15, 38.74, 54.36, 120.24, 121.52, 122.98,
126.21, 127.85, 128.18, 130.81, 140.18, 140.29, 148.55, 151.23. Anal.
Calcd for C19H19Br3: C, 46.85; H, 3.93. Found: C, 47.07; H, 3.99.
2-Bromo-9,9′-bis(6′′-bromohexyl)fluorene (5). Following the pro-
cedure described above, we obtained compound 5 as a light yellow oil
1
in a yield of 75%. H NMR (200 MHz, CDCl3) δ: 0.60-0.65 (m,
4H), 1.07-1.21 (m, 8H), 1.54-1.71 (m, 4H), 1.91-1.99 (m, 4H), 3.26-
3.28 (t, 4H), 7.33-7.36 (m, 3H), 7.45-7.48 (m, 2H), 7.52-7.60 (m,
1H), 7.64-7.72 (m, 1H). 13C NMR (50 MHz, CDCl3) δ: 23.64, 27.94,
29.17, 32.79, 34.13, 40.30, 55.42, 120.03, 121.31, 122.97, 126.21,
127.27, 127.78, 130.23, 140.19, 140.32, 150.08, 152.76. Anal. Calcd
for C25H31Br3: C, 52.57; H, 5.47. Found: C, 52.32; H, 5.31.
2-Bromo-9,9′-bis(8′′-bromooctyl)fluorene (6). Following the pro-
cedure described above, we obtained compound 6 as a light yellow oil
1
in 80% yield. H NMR (200 MHz, CDCl3) δ: 0.50-0.62 (m, 4H),
1.01-1.07 (m, 12H), 1.21-1.35 (br, 4H), 1.68-1.80 (m, 4H), 1.90-
1.98 (m, 4H), 3.30-3.32 (t, 4H), 7.32-7.34 (br, 3H), 7.42-7.46 (m,
2H), 7.52-7.58 (m, 1H), 7.62-7.70 (m, 1H). 13C NMR (50 MHz,
CDCl3) δ: 23.76, 28.20, 28.73, 29.13, 29.92, 32.90, 34.23, 40.39, 55.49,
119.97, 121.25, 123.00, 126.25, 127.16, 127.70, 130.11, 140.15, 140.30,
150.29, 152.99. Anal. Calcd for C29H39Br3: C, 55.52; H, 6.27. Found:
C, 55.78; H, 6.51.
Experimental Section
General Details. 1H and 13C NMR spectra were collected on Varian
Inova 400 MHz or Varian ASM-100 200 MHz spectrometers. The UV-
vis absorption spectra were recorded on a Shimadzu UV-2401 PC diode
array spectrometer. Photoluminescence spectra were obtained on a Spex
Fluorolog 2 spectrometer, using 90° angle detection for solution
samples. Mass spectrometry and elemental analysis were performed
by the UC Santa Barbara Mass Spectrometry Lab and elemental analysis
center. Reagents were obtained from Aldrich Co. and used as received.
The oligonucleotide that was used in the single-stranded DNA study
(ssDNA1) was the 20 base pair sequence 5′-Fl-CCA ATC AGT CCA
GTG ATA CG, containing a fluorescein chromophore at the 5′ position.
The oligonucleotides that were used for complementary and non-
complementary studies (ssDNA2) were 5′-Fl-ATC TTG ACT ATG
TGG GTG CT and its complementary strand 5′-AGC ACC CAC ATA
GTC AAG AT. The noncomplementary strand was the sequence 5′-
GAC TCA ATG GCG TTA GAC TG. The samples were prepared by
initially determining DNA strand concentrations based on 260 nm
absorbance measurements done on 150 µL samples in a 500 µL quartz
cuvette using a Shimadzu UV-vis spectrometer. Once the concentration
of both strains was established, a 1:1 ratio between complementary
single strands was mixed for annealing. The mixtures of complementary
strands were annealed at 57.5 °C (2 °C below its melting point of 59.5
°C) for 25 min and then cooled to room temperature slowly, and this
was done in an identical fashion for the noncomplementary strands.
The absorbance of the hybridized strands was measured to determine
their concentration. The extent of hybridization was checked by variable
4-Bromo-1-(4′-methylphenyloxy)butane (7). A mixture of 1,4-
dibromobutane (50 mL), potassium carbonate (27.6 g, 200 mmol),
4-methylphenol (6.49 g, 60 mmol), and a catalytic amount of 18-
crown-6 in 400 mL of acetone was refluxed for 2 days. After removal
of the acetone, the residue was poured into water and was extracted
with diethyl ether (3 × 100 mL). The combined organic layer was
washed with aqueous HCl (1 M) and then with water. After being dried
and removal of the solvent, the residue was distilled under vacuum to
afford 7 (10.5 g, 75%) as a colorless oil. 1H NMR (200 MHz, CDCl3)
δ: 1.88-2.11 (m, 4H), 2.29 (s, 3H), 3.46-3.53 (t, 2H), 3.94-4.0 (t,
2H), 6.77-6.81 (d, 2H), 7.06-7.11 (d, 2H). 13C NMR (50 MHz, CDCl3)
δ: 20.74, 30.40, 32.68, 65.52, 69.02, 114.61, 130.16, 130.30, 156.80.
2-Bromo-9,9′-bis(4′′-methylphenyloxybutyl)fluorene (8). Follow-
ing the same procedure as described for the 2-bromo-9,9′-bis(ω′′-
bromoalkyl)fluorene compounds, after silica gel chromatography using
hexane and chloroform (3:1) as the eluent, we obtained compound 8
1
as white crystals (65%). H NMR (200 MHz, CDCl3) δ: 0.72-0.83
(m, 4H), 1.46-1.57 (m, 4H), 1.98-2.06 (m, 4H), 2.25 (s, 6H), 3.65-
3.72 (t, 4H), 6.63-6.69 (d, 4H), 6.99-7.04 (d, 4H), 7.31-7.34 (m,
3H), 7.43-7.48 (m, 2H), 7.49-7.53 (m, 1H), 7.58-7.68 (m, 1H). 13
C
NMR (50 MHz, CDCl3) δ: 20.46, 20.61, 29.60, 40.12, 55.41, 67.67,
114.47, 120.07, 121.35, 123.03, 126.25, 127.35, 127.83, 129.80, 129.94,
130.33, 140.18, 140.32, 149.78, 152.44, 156.89. Anal. Calcd for C35H37-
BrO2: C, 73.81; H, 6.55. Found: C, 73.89; H, 6.83.
9
J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003 6713

A R T I C L E S
Liu et al.
2-Bromo-9,9′-bis(4′′-bromobutyl)fluorene (9). A mixture of 2-bromo-
9,9-bis(4-methylphenyloxybutyl)fluorene (91.0 g, 17.5 mmol), 4 mL
of 48% HBr, and 20 mL of acetic acid was refluxed for 2 days. The
reaction mixture was poured into water, and the resulting solution was
extracted with chloroform (3 × 50 mL). The combined organic layers
were washed with dilute aqueous Na2CO3 (1 M, 2 × 50 mL) and water
and were dried over MgSO4. The solvent was removed, and the resulting
oil was purified by silica gel chromatography using hexane and
chloroform (7:1) as the eluent to afford 9 (580 mg, 64%) as a colorless
118.33, 120.23, 120.45, 121.76, 123.27, 126.27, 127.29, 127.53, 127.96,
129.46, 140.00 140.34, 140.64, 141.16, 151.27, 151.75. MS (MALDI-
TOF): 1170.6 (M), 1090.6 (M - Br), 1010.7 (M - 2Br), 930.7 (M -
3Br). Anal. Calcd for C64H82Br4: C, 65.65; H, 7.06. Found: C, 65.06;
H, 6.45.
General Procedure for 1,4-Bis(9,9′-bis(ω′′-(N,N,N-trimethylam-
monium)-alkyl)-2′-fluorenyl)benzene Tetrabromide (C3, C4, C6, and
C8). The precursor oligomer (100 mg) was dissolved in 10 mL of THF
and was then diluted with 5 mL of water. After the mixture was cooled
to - 78 °C, condensed trimethylamine (1 mL) was added dropwise.
The mixture was stirred at - 78 °C for 12 h and was then gradually
warmed to room temperature. After an additional 24 h, the THF and
water were removed, and the residue was recrystallized from ethanol
to yield the target oligomers with yields of 40-75%.
1
oil, which solidified upon standing. H NMR (200 MHz, CDCl3) δ:
0.69-0.81 (m, 4H), 1.55-1.69 (m, 4H), 1.97-2.02 (m, 4H), 3.12-
3.19 (t, 4H), 7.32-7.38 (m, 3H), 7.45-7.54 (m, 2H), 7.56-7.59 (m,
1H), 7.68-7.70 (m, 1H). 13C NMR (50 MHz, CDCl3): 22.69, 33.21,
33.33, 39.41, 55.13, 120.25, 121.53, 123.00, 126.26, 127.61, 128.01,
130.56, 140.18, 140.30, 149.37, 152.07. Anal. Calcd for C21H23Br3:
C, 48.96; H, 4.50. Found: C, 48.83; H, 4.64.
General Procedure for 1,4-Bis(9,9′-bis(ω′′-bromoalkyl)-2′fluo-
renyl)benzene. A mixture of 2-bromo-9,9′-bis(ω′′-bromoalkyl)fluorene
(1.05 mmol), 1,4-phenyldiboronic acid (0.5 mmol), Pd(PPh3)4 (10 mg),
and sodium carbonate (5 mmol) in 10 mL of THF and 2.5 mL of water
was degassed and stirred at 80 °C for 48 h. The mixture was cooled to
room temperature and diluted with 100 mL of chloroform. The organic
layer was collected, washed with water and then brine, dried over
magnesium sulfate, and then the solvent was removed. The residue
was adsorbed on silica gel and purified by chromatography (hexane:
toluene ) 1:1 to 1:3) to give the desired compounds in yields from
35% to 60%.
1,4-Bis(9,9′-bis(3′′-(N,N,N-trimethylammonium)-propyl)-2′-fluo-
renyl)benzene Tetrabromide (C3). Oligomer C3 was obtained as a
1
white powder in a yield of 40%. H NMR (500 MHz, D2O, 90 °C) δ:
1.64-1.70 (m, 8H), 2.82-2.89 (m, 8H), 3.30-3.34 (s, 36H), 3.52-
3.55 (m, 8H), 8.10-8.58 (m, 18H). 13C NMR (125 MHz, D2O, 90 °C)
δ: 18.76, 36.38, 53.77, 55.53, 67.69, 121.53, 121.96, 122.68, 124.83,
128.02, 128.66, 129.38, 129.5, 140.75, 140.84, 141.73, 141.82, 149.74,
150.56. MS (ESI-MS): 1047 (M - Br)+, 483 (M - 2Br)2+, 295 (M -
3Br)3+, 201 (M - 4Br)4+
.
1,4-Bis(9,9′-bis(4′′-(N,N,N-trimethylammonium)-butyl)-2′-fluore-
nyl)benzene Tetrabromide (C4). Oligomer C4 was obtained as a white
powder in a yield of 65%. 1H NMR (500 MHz, D2O, 90 °C) δ: 1.20-
1.41 (m, 8H), 1.90-1.92 (br, 8H), 2.68-2.72 (m, 8H), 3.36-3.48 (m,
44H), 7.74-8.58 (m, 18H). 13C NMR (125 MHz, D2O, 90 °C) δ: 21.98,
23.51, 39.08, 54.03, 56.11, 67.51, 121.16, 121.75, 122.22, 125.06,
126.82, 128.36, 128.79, 129.10, 139.28, 140.06, 141.17, 141.69, 151.57,
152.42. MS (ESI-MS): 1104 (M - Br)+, 511 (M - 2Br)2+, 315 (M -
1,4-Bis(9,9′-bis(3′′-bromopropyl)-2′-fluorenyl)benzene (10). Pre-
cursor oligomer 10 was obtained as a white powder in a yield of 35%.
1H NMR (200 MHz, CDCl3) δ: 1.20-1.25 (m, 8H), 2.22-2.27 (m,
8H), 3.10-3.17 (t, 8H), 7.37-7.39 (m, 6H), 7.65-7.78 (m, 12H). 13
C
NMR (50 MHz, CDCl3) δ: 27.30, 34.56, 38.88, 54.23, 120.25, 120.49,
121.48, 122.98, 126.71, 127.78, 140.23, 140.49, 140.55, 140.83, 149.24,
149.78. MS (MALDI-TOF): 890.2 (M), 810.2 (M - Br), 730.2 (M -
2Br), 650.2 (M - 3Br). Anal. Calcd for C44H42Br4: C, 59.35; H, 4.75.
Found: C, 59.17; H, 4.75.
1,4-Bis(9,9′-bis(4′′-bromobutyl)-2′-fluorenyl)benzene (11). Precur-
sor oligomer 11 was obtained as a white powder in a yield of 44%. 1H
NMR (200 MHz, CDCl3) δ: 0.79-0.86 (br, 8H), 1.58-1.69 (m, 8H),
2.03-2.10 (m, 8H), 3.16-3.22 (t, 8H), 7.35-7.39 (br, 6H), 7.63-
7.83 (m, 12H). 13C NMR (50 MHz, CDCl3) δ: 22.72, 33.23, 33.36,
39.54, 54.89, 120.17, 120.40, 121.43, 122.93, 126.43, 127.45, 127.53,
127.76, 139.98, 140.50, 140.63, 140.87, 150.03, 150.49. MS (MALDI-
TOF): 946 (M), 866 (M - Br), 786 (M - 2Br). Anal. Calcd for C48H50-
Br4: C, 60.91; H, 5.32. Found: C, 61.05; H, 5.32.
1,4-Bis(9,9′-bis(6′′-bromohexyl)-2′-fluorenyl)benzene (12). Precur-
sor oligomer 12 was obtained as a white powder in a yield of 55%. 1H
NMR (200 MHz, CDCl3) δ: 0.67-0.71 (m, 8H), 1.10-1.26 (m, 16H),
1.60-1.71 (m, 8H), 2.00-2.08 (m, 8H), 3.29 (t, 8H), 7.35-7.39 (m,
6H), 7.61-7.68 (m, 4H), 7.73-7.82 (m, 8H). 13C NMR (50 MHz,
CDCl3) δ: 23.75, 27.92, 29.24, 32.80, 34.19, 40.43, 55.20, 120.03,
120.26, 121.44, 122.97, 126.14, 127.15, 127.37, 127.74, 139.76, 140.57,
140.69, 140.89, 150.76, 151.24. MS (MALDI-TOF): 1058 (M), 978
(M - Br), 898 (M - 2Br), 818 (M - 3Br). Anal. Calcd for C56H66Br4:
C, 63.53; H, 6.28. Found: C, 64.33; H, 6.45.
3Br)3+, 216 (M - 4Br)4+
.
1,4-Bis(9,9′-bis(6′′-(N,N,N-trimethylammonium)-hexyl)-2′-fluore-
nyl)benzene Tetrabromide (C6). Oligomer C6 was obtained as a white
1
powder in a yield of 75%. H NMR (500 MHz, D2O, 90 °C) δ: 1.02
(m, 8H), 1.33-1.35 (br, 16H), 1.76-1.79 (m, 8H), 2.20-2.30 (m, 4H),
2.45 (m, 4H), 3.39-3.50 (m, 44H), 7.95-8.12 (m, 12 H), 8.39-8.54
(m, 6H). 13C NMR (125 MHz, D2O, 90 °C) δ: 23.10, 24.54, 26.13,
29.33, 31.32, 39.88, 54.08, 56.23, 67.64, 120.85, 121.58, 121.69, 124.94,
126.59, 128.24, 128.37, 128.82, 138.99, 140.34, 141.28, 141.60, 152.41,
153.09. MS (ESI-MS): 566 (M - 2Br)2+, 351 (M - 3Br)3+, 293 (M
- 4Br)4+
.
1,4-Bis(9,9′-bis(8′′-(N,N,N-trimethylammonium)-octyl)-2′-fluore-
nyl)benzene Tetrabromide (C8). Oligomer C8 was obtained as a white
powder in a yield of 72%. 1H NMR (500 MHz, D2O, 90 °C) δ: 1.03-
1.05 (br, 8H), 1.20-1.26 (br, 16 H), 1.26-1.30 (br, 16H), 1.45-1.48
(br, 8H), 1.88-1.90 (br, 4H), 2.36 (br, 4H), 3.52-3.60 (br, 44H), 7.90-
8.00 (m, 6H), 8.24-8.28 (br, 8H), 8.71 (br, 4H). 13C NMR (125 MHz,
D2O, 90 °C) δ: 23.48, 25.15, 26.79, 29.40, 29.54, 29.87, 30.50, 40.69,
54.22, 56.11, 67.81, 120.86, 121.65, 124.70, 126.82, 128.36, 128.45,
128.57, 139.63, 141.07, 141.52, 141.63, 152.30, 152.97. MS (ESI-
MS): 1327 (M - Br)+, 624 (M - 2Br)2+, 389 (M - 3Br)3+, 272 (M
- 4Br)4+
.
Acknowledgment. The authors are grateful to the ONR
(N00014-98-1-0759), the NSF (DMR-0097611 ), and the NIH
(GM62958-01) for financial support and to Professors Alan
Heeger and Kevin Plaxco for useful discussions.
1,4-Bis(9,9′-bis(8′′-bromooctyl)-2′-fluorenyl)benzene (13). Precur-
1
sor oligomer 13 was obtained as a colorless oil in a yield of 60%. H
NMR (200 MHz, CDCl3) δ: 0.69 (br, 8H), 1.07 (br, 24H), 1.25-1.32
(m, 8H), 1.61-1.81 (m, 8H), 1.98-2.06 (m, 8H), 3.29-3.36 (m, 8H),
7.33-7.38 (m, 6H), 7.60-7.81 (m, 12H). 13C NMR (50 MHz, CDCl3)
δ: 24.09, 28.44, 29.00, 29.38, 30.24, 33.15, 34.52, 37.36, 40.79, 55.53,
JA028961W
9
6714 J. AM. CHEM. SOC. VOL. 125, NO. 22, 2003