Molecular tuning of phenylene-vinylene derivatives for two-photon photosensitized singlet oxygen production

Article abstract of DOI:10.1021/jo9020216

(Chemical Equation Presented) Substituent-dependent features and properties of the sensitizer play an important role in the photosensitized production of singlet oxygen, O₂(a¹Δ_g). In this work, we systematically examine the effect of molecular changes in the sensitizer on the efficiency of singlet oxygen production using, as the sensitizer, oligophenylene-vinylene derivatives designed to optimally absorb light in a nonlinear two-photon process. We demonstrate that one cannot always rely on rule-of-thumb guidelines when attempting to construct efficient two-photon singlet oxygen sensitizers. Rather, as a consequence of behavior that can deviate from the norm, a full investigation of the photophysical properties of the system is generally required. For example, it is acknowledged that the introduction of a ketone moiety to the sensitizer chromophore often results in more efficient production of singlet oxygen. However, we show here that the introduction of a carbonyl into a given phenylene-vinylene can, rather, have adverse effects on the yield of singlet oxygen produced. Using these molecules, we show that care must also be exercised when using qualitative symmetry-derived arguments to predict the relationship between one-and two-photon absorption spectra.

Full text of DOI:10.1021/jo9020216

pubs.acs.org/joc
Molecular Tuning of Phenylene-Vinylene Derivatives for Two-Photon
Photosensitized Singlet Oxygen Production
Christian B. Nielsen,†,‡ Jacob Arnbjerg,† Mette Johnsen,† Mikkel Jorgensen,‡ and
Peter R. Ogilby*,†
†
˚
Center for Oxygen Microscopy and Imaging, Department of Chemistry, Aarhus University, Arhus DK-8000,
Denmark, and ‡Polymer Department, Riso National Laboratory, DK-4000 Roskilde, Denmark
progilby@chem.au.dk
Received September 21, 2009
Substituent-dependent features and properties of the sensitizer play an important role in the photo-
sensitized production of singlet oxygen, O2(a1Δg). In this work, we systematically examine the effect of
molecular changes in the sensitizer on the efficiency of singlet oxygen production using, as the sensitizer,
oligophenylene-vinylene derivatives designed to optimally absorb light in a nonlinear two-photon process.
We demonstrate that one cannot always rely on rule-of-thumb guidelines when attempting to construct
efficient two-photon singlet oxygen sensitizers. Rather, as a consequence of behavior that can deviate from
the norm, a full investigation of the photophysical properties of the system is generally required. For
example, it is acknowledged that the introduction of a ketone moiety to the sensitizer chromophore often
results in more efficient production of singlet oxygen. However, we show here that the introduction of a
carbonyl into a given phenylene-vinylene can, rather, have adverse effects on the yield of singlet oxygen
produced. Using these molecules, we show that care must also be exercised when using qualitative
symmetry-derived arguments to predict the relationship between one-and two-photon absorption spectra.
Introduction
Over the years, a great deal of information has been accumu-
lated regarding features of a sensitizer that result in the efficient
production of singlet oxygen upon the absorption of a single
photon (i.e., experiments performed in the domain where light
absorption scales linearly with the incident power).3 It has been
shown, however, that singlet oxygen can also be formed upon
two-photon absorption of a sensitizer (Figure 1)4-6 and that the
near-IR singlet oxygen phosphorescence thus produced can be
used to accurately quantify the two-photon absorption cross
section, δ, of the sensitizer.7 In turn, this has initiated a flurry of
The photosensitized production of singlet molecular
oxygen, O2(a1Δg), is a topic that has seen a great deal of
attention over the last 40 years. Interest in singlet oxygen is
broad-based, reflecting its importance in disciplines that
range from synthetic organic chemistry1 to photoinduced
cell death.2 Singlet oxygen is conveniently produced in a
photosensitized process wherein a light-absorbing molecule
(the so-called sensitizer) transfers its energy of excitation to
ground state oxygen, O2(X3Σg-), (Figure 1). This process of
energy transfer occurs most efficiently from the compara-
tively long-lived, lowest energy triplet state (T1) of the
sensitizer.3
(4) Frederiksen, P. K.; Jorgensen, M.; Ogilby, P. R. J. Am. Chem. Soc.
2001, 123, 1215–1221.
(5) Frederiksen, P. K.; McIlroy, S. P.; Nielsen, C. B.; Nikolajsen, L.;
Skovsen, E.; Jorgensen, M.; Mikkelsen, K. V.; Ogilby, P. R. J. Am. Chem.
Soc. 2005, 127, 255–269.
(1) Clennan, E. L.; Pace, A. Tetrahedron 2005, 61, 6665–6691.
(2) Redmond, R. W.; Kochevar, I. E. Photochem. Photobiol. 2006, 82,
1178–1186.
(6) Karotki, A.; Kruk, M.; Drobizhev, M.; Rebane, A.; Nickel, E.;
Spangler, C. W. IEEE J. Selected Top. Quantum Electron. 2001, 7, 971–975.
(7) Arnbjerg, J.; Johnsen, M.; Frederiksen, P. K.; Braslavsky, S. E.;
Ogilby, P. R. J. Phys. Chem. A 2006, 110, 7375–7385.
(3) Schweitzer, C.; Schmidt, R. Chem. Rev. 2003, 103, 1685–1757.
9094 J. Org. Chem. 2009, 74, 9094–9104
Published on Web 11/11/2009
DOI: 10.1021/jo9020216
r
2009 American Chemical Society

Nielsen et al.
JOCArticle
of δ.18-22 Much of the discussion has focused on the
architecture conducive to the facile intramolecular redistri-
bution of electron density. The key underlying principle here
is that, with the redistribution of large amounts of charge
over a comparatively large intramolecular distance, one can
increase the magnitude of the transition dipole moment and,
in turn, increase the transition probability as manifested in a
larger value of δ.18,21,23,24 Molecular features that can sig-
nificantly influence the intramolecular distribution of charge
include the judicious placement of electron donor and ac-
ceptor groups on a conjugated framework of extended
length. As one might expect, molecular properties that
depend on the transfer and localization of charge should
likewise be quite susceptible to the effect of the surrounding
solvent. In this regard, it is somewhat surprising that only
few systematic attempts have been made to correlate ob-
served δ-values to changes in solvent.25-27 We have shown
that the OPV scaffold can be used in the development of both
hydrophobic and hydrophilic two-photon singlet oxygen
sensitizers.5,7,8,28
FIGURE 1. Illustration of one- and two-photon triplet-state
photosensitized production of singlet oxygen, O2(a1Δg). Depending
on the sensitizer, the simultaneous absorption of two photons may
or may not populate the same state as that created upon the
absorption of a single higher energy photon. ISC denotes inter-
system crossing, and IC denotes internal conversion. The label “a-X
emission” refers to the 1270 nm phosphorescence of singlet oxygen.
Principles of molecular design known to be beneficial for
producing large two-photon transition probabilities are
often undesired in terms of optimizing the photosensitized
production of singlet oxygen (i.e., features that result in large
values of δ often result in small values of the quantum yield
of singlet oxygen production, ΦΔ).5,8,28 In particular, the
extent of charge-transfer (CT), both within the sensitizer
itself (i.e., intramolecular CT) as well as in the sensitizer-
oxygen complex (i.e., intermolecular CT), can have a large
adverse effect on the efficiency with which singlet oxygen is
generated.3,29-33 In general, CT character in a sensitizer and/
or sensitizer-oxygen complex facilitates nonradiative deac-
tivation of the excited state at the expense of energy transfer
from the sensitizer to produce singlet oxygen. However, as
mentioned previously, a high degree of intramolecular CT is
desirable for obtaining large δ values. Therefore, developing
a sensitizer that is both a good two-photon absorber as well
as an efficient singlet oxygen generator requires delicate
balancing of the photophysical consequences of a change
in the molecular architecture. In this argument, we assume
that the extent of intermolecular CT in the sensitizer-oxygen
complex will be larger for a sensitizer with a proclivity for
studies designed to elucidate features of a given molecule that
optimize singlet oxygen production in this process where light
absorption scales quadratically with the incident power.8-15 One
attribute of the latter nonlinear phenomenon is that the volume
of excitation, and hence volume of singlet oxygen production,
can be localized to a small spatial domain which can be
advantageous in applications ranging from imaging to spatially
selective chemistry.16,17
The nonlinear optical properties of oligo phenylene-vinylenes
(OPVs) have been widely studied, and it is acknowledged
that many molecules from this class of compounds can have
large two-photon absorption cross sections.18-20 Over
the years, many groups have focused on elucidating key
molecular features in OPVs that result in large values
(8) Nielsen, C. B.; Johnsen, M.; Arnbjerg, J.; Pittelkow, M.; McIlroy,
S. P.; Ogilby, P. R.; Jorgensen, M. J. Org. Chem. 2005, 70, 7065–7079.
(9) Arnbjerg, J.; Paterson, M. J.; Nielsen, C. B.; Jorgensen, M.;
Christiansen, O.; Ogilby, P. R. J. Phys. Chem. A 2007, 111, 5756–5767.
ꢀ
(10) Arnbjerg, J.; Jimenez-Banzo, A.; Paterson, M. J.; Nonell, S.; Borrell,
J. I.; Christiansen, O.; Ogilby, P. R. J. Am. Chem. Soc. 2007, 129, 5188–5199.
ꢀ
(11) McIlroy, S. P.; Clo, E.; Nikolajsen, L.; Frederiksen, P. K.; Nielsen,
C. B.; Mikkelsen, K. V.; Gothelf, K. V.; Ogilby, P. R. J. Org. Chem. 2005, 70,
1134–1146.
ꢀ
(12) Oar, M. A.; Serin, J. M.; Dichtel, W. R.; Frechet, J. M. J.;
Ohulchanskyy, T. Y.; Prasad, P. N. Chem. Mater. 2005, 17, 2267–2275.
(13) Drobizhev, M.; Stepanenko, Y.; Dzenis, Y.; Karotki, A.; Rebane,
A.; Taylor, P. N.; Anderson, H. L. J. Phys. Chem. B 2005, 109, 7223–7236.
(14) Andrasik, S. J.; Belfield, K. D.; Bondar, M. V.; Hernandez, F. E.;
Morales, A. R.; Przhonska, O. V.; Yao, S. ChemPhysChem 2007, 8, 399–404.
(15) Belfield, K. D.; Bondar, M. V.; Hernandez, F. E.; Masunov, A. E.;
Mikhailov, I. A.; Morales, A. R.; Przhonska, O. V.; Yao, S. J. Phys. Chem. C
2009, 113, 4706–4711.
(16) Skovsen, E.; Snyder, J. W.; Ogilby, P. R. Photochem. Photobiol.
2006, 82, 1187–1197.
(17) Collins, H. A.; Khurana, M.; Moriyama, E. H.; Mariampillai, A.;
Dahlstedt, E.; Balaz, M.; Kuimova, M. K.; Drobizhev, M.; Yang, V. X. D.;
Phillips, D.; Rebane, A.; Wilson, B. C.; Anderson, H. L. Nat. Photonics 2008,
2, 420–424.
(22) Terenziani, F.; Katan, C.; Badaeva, E.; Tretiak, S.; Blanchard-
Desce, M. Adv. Mater. 2008, 20, 4641–4678.
(23) Mongin, O.; Porres, L.; Moreaux, L.; Mertz, J.; Blanchard-Desce,
M. Org. Lett. 2002, 4, 719–722.
(24) Bischel, W. K.; Kelly, P. J.; Rhodes, C. K. Phys. Rev. A. 1976, 13,
1817–1828.
(25) Johnsen, M.; Ogilby, P. R. J. Phys. Chem. A. 2008, 112, 7831–7839.
(26) Woo, H.-Y.; Liu, B.; Kohler, B.; Korystov, D.; Mikhailovsky, A.;
Bazan, G. C. J. Am. Chem. Soc. 2005, 127, 14721–14729.
(27) Toro, C.; De Boni, L.; Yao, S.; Belfield, K. D.; Hernandez, F. E. J.
Phys. Chem. B 2008, 112, 12185–12190.
(28) Johnsen, M.; Paterson, M. J.; Arnbjerg, J.; Christiansen, O.; Nielsen,
C. B.; Jorgensen, M.; Ogilby, P. R. Phys. Chem. Chem. Phys. 2008, 10, 1177–
1191.
ꢀ
(18) Albota, M.; Beljonne, D.; Bredas, J.-L.; Ehrlich, J. E.; Fu, J.-Y.;
Heikal, A. A.; Hess, S. E.; Kogej, T.; Levin, M. D.; Marder, S. R.; McCord-
(29) Kristiansen, M.; Scurlock, R. D.; Iu, K.-K.; Ogilby, P. R. J. Phys.
Chem. 1991, 95, 5190–5197.
€
Maughon, D.; Perry, J. W.; Rockel, H.; Rumi, M.; Subramaniam, G.; Webb,
W. W.; Wu, X.-L.; Xu, C. Science 1998, 281, 1653–1656.
(30) McGarvey, D. J.; Wilkinson, F.; Worrall, D. R.; Hobley, J.; Shaikh,
W. Chem. Phys. Lett. 1993, 202, 528–534.
(19) Pond, S. J. K.; Rumi, M.; Levin, M. D.; Parker, T. C.; Beljonne, D.;
ꢀ
Day, M. W.; Bredas, J.-L.; Marder, S. R.; Perry, J. W. J. Phys. Chem. A 2002,
106, 11470–11480.
(31) Flors, C.; Ogilby, P. R.; Luis, J. G.; Grillo, T. A.; Izquierdo, L. R.;
Gentili, P.-L.; Bussotti, L.; Nonell, S. Photochem. Photobiol. 2006, 82, 95–
103.
(20) Strehmel, B.; Sarker, A. M.; Detert, H. ChemPhysChem 2003, 4, 249–
259.
(32) Arnbjerg, J.; Johnsen, M.; Nielsen, C. B.; Jorgensen, M.; Ogilby, P.
R. J. Phys. Chem. A 2007, 111, 4573–4583.
(21) He, G. S.; Tan, L.-S.; Zheng, Q.; Prasad, P. N. Chem. Rev. 2008, 108,
1245–1330.
(33) Jensen, P.-G.; Arnbjerg, J.; Tolbod, L. P.; Toftegaard, R.; Ogilby, P.
R. J. Phys. Chem. A 2009, 113, 9965–9973.
J. Org. Chem. Vol. 74, No. 23, 2009 9095

Nielsen et al.
JOCArticle
FIGURE 2. General OPV structure where X denotes the moiety in
which systematic molecular changes were made.
intramolecular CT (i.e., intermolecular CT to yield a Sens-O2
state with Sensþ.O2-. character will more readily occur for a
sensitizer with an electron-donating moiety).29-33
Although general features of a molecule that give rise to
large values of δ do not often give rise to large values of ΦΔ,
exceptions to this “rule” exist.10,13,17 With this in mind, it is
now necessary to embark on a series of more methodical and
systematic studies to elucidate molecular features of a chro-
mophore that can concomitantly result in large values of δ
and ΦΔ. In the final analysis, it is the magnitude of the
product δΦΔ that carries meaning in characterizing the
efficiency of a two-photon singlet oxygen sensitizer. To this
end, we set out to examine the effects of systematically
adding/deleting specific functional groups in a given sensiti-
zer system. To establish continuity with published data, we
felt it would be beneficial to work with a specific OPV
framework with diphenylamino groups at both ends of the
chromophore (Figure 2).
FIGURE 3. Structures of the parent dibromo OPV, 2, and of the
corresponding carbonyl-containing analogue, 1, used to examine
one aspect of the two-photon-sensitized production of singlet
oxygen.
TABLE 1. Photophysical Properties of 1 and 2 in Air-Saturated
Toluene
a
b
c
O2a
fT
a
ΦF
ΦT
Φic
ΕΤ/kJ/mol
ΦΔ
δ(λ)d/GM
1
2
0.33
0.44
0.33
0.44
0.34
0.12
0.84
0.90
160
148
0.28
0.38e
1740
1310
a10% uncertainty. b∼20% uncertainty. cObtained from the relation-
ship Φic = 1 - ΦT - ΦF. d15% uncertainty. Determined at λmax of the
two-photon spectrum, which is 845 nm for 1 and 770 nm for 2 (see Figure
e
4). We have reported a value of 0.45 ( 0.05 for 2,4,7 and the present
number, although smaller, is still within our error limits.
Results and Discussion
does not result in the expected increased yield of singlet
oxygen (Table 1). Keeping our overall mechanistic perspec-
tive in mind, we performed a series of complementary
experiments to help elucidate these observations.
Our starting point in this study was the dibromo OPV 2
(Figure 3). This compound produces singlet oxygen in
moderate yield and has a reasonably large δ value.7 Relative
to the corresponding OPV that lacks the Br substituents, the
presence of these heavy atoms in 2 boosts the efficiency of
S1 f T1 intersystem crossing which, in turn, results in the
more efficient production of singlet oxygen (Figure 1).11
In addition to the above-mentioned internal heavy atom
effect, another well-known approach to facilitate S1-T1
intersystem crossing is to introduce carbonyl moieties into
the chromophore (El-Sayed’s Rules).34 As such, some of the
best known and most efficient singlet oxygen sensitizers are
ketone-containing molecules.35 We expected that the sym-
metrical addition of two carbonyl moieties to the chromo-
phore in 2 might be advantageous in two ways: (1) we could
increase the efficiency of intersystem crossing to the triplet
state and, as such, increase the singlet oxygen yield, and (2)
we could increase the extent of intramolecular charge redis-
tribution which would be manifested in a larger value of δ.
Thus, we first chose to examine the behavior of compound 1
(Figure 3) and compare this to compound 2.
In conjunction with the detailed study of 1 and 2, we set
out to use our OPV system to systematically examine the
effects of (1) changing the substituents on the central aro-
matic ring and (2) changing the central aromatic ring itself.
The intent here was to establish a broader database upon
which substantive and compelling general conclusions could
be drawn.
1. Comparing the Behavior of 1 and 2. Although the
introduction of the carbonyl groups into the chromophore
of this OPV indeed gives rise to a slightly larger value of δ, it
Quantum Yields. We first note that, for the triplet state
photosensitized production of singlet oxygen, ΦΔ can be ex-
pressed as the product of three parameters: the triplet state quan-
tum yield, ΦT, the fraction of triplet states quenched by ground
O2
state oxygen, fT , and the fraction of these latter quenching
events that result in singlet oxygen production, SΔ (eq 1).3
O2
ΦΔ ¼ ΦTf T SΔ
ð1Þ
Like ΦΔ, values of ΦT were measured in independent
experiments. Using [O2]-dependent changes in the T1 lifetime,
O2
we obtained fT
Information. Values of SΔ could then be obtained using eq 1.
. Details are provided in the Supporting
O2
Although fT for 1 is indeed slightly smaller than that for
2 (Table 1), this difference is not enough to account for the
observed difference in ΦΔ. Moreover, the data indicate that,
for each molecule, SΔ ∼1. (Using independently established
arguments,3,36 these values of SΔ are consistent with the ET
values of 160 and 148 kJ/mol for 1 and 2, respectively.)
Indeed, carrying this point further, with SΔ values of ∼1 for
both 1 and 2, we hesitate to suggest that the respective T1
states would have a different amount of CT character. Thus,
on the basis solely of the model described in the Introduction,
our data already reveal an unexpected result.
Going further, the explanation for the difference in ΦΔ is
readily seen in the corresponding values of ΦT; the yield of
the singlet oxygen precursor, the sensitizer triplet state, is
smaller in 1 than in 2. Simply put, in contrast to our
expectation, addition of the carbonyl to our chromophore
does not result in an increased triplet yield.
(34) Lower, S. K.; El-Sayed, M. A. Chem. Rev. 1966, 66, 199–241.
(35) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data
1993, 22, 113–262.
(36) Schmidt, R. Photochem. Photobiol. 2006, 82, 1161–1177.
9096 J. Org. Chem. Vol. 74, No. 23, 2009

Nielsen et al.
JOCArticle
FIGURE 5. Solvent-dependent absorption and emission spectra
for compounds 1 and 2. Solvents used: (red) benzonitrile; (magenta)
acetonitrile; (green) THF; (black) cyclohexane; (blue) toluene.
FIGURE 4. One-photon absorption (O), two-photon excitation
(b), and fluorescence (0) spectra of 1 and 2, all recorded in toluene.
The lower x-axis refers to the total transition wavelength, while the
upper x-axis shows the wavelength used in the two-photon experi-
ments.
does not occur in 1 and 2 under our conditions (see pertinent
data in the Supporting Information).
One-Photon Spectra. One-photon absorption and fluor-
escence spectra for 1 and 2 dissolved in toluene are shown in
Figure 4. The respective absorption spectra are similar,
whereas the fluorescence spectra are distinctly different: the
emission profile of 2 has pronounced vibronic structure,
whereas the profile of 1 is broadened and structureless.
Moreover, the Stokes shift observed for 1 is significantly
greater than that for 2. Although these data are consistent
with our suggestion that 1 is more conformationally flexible
than 2, they could also reflect the fact that the emitting state
in 1 has more CT character than that of 2.25,41
In an attempt to comment further on the relative amounts
of CT character, at least in the emitting state of these
molecules, we examined the fluorescence spectra of 1 and 2
in solvents of varying polarity (Figure 5). The results clearly
show that, for both molecules, the absorption spectrum is
not strongly dependent on the solvent, whereas the emission
spectrum red-shifts as the polarity of the solvent is increased.
This latter effect is more pronounced for 1, indicating that
this molecule may indeed have more CT character in its
solvent-stabilized emitting state than does 2.
Using the relation Φic = 1 - ΦT - ΦF, we ascertain that 1
has a much larger quantum yield of internal conversion, Φic,
than does 2 (Table 1). Thus, the lower triplet yield in 1 is
principally a consequence of events that better facilitate
nonradiative S1 f S0 coupling in 1. Among other things,
such coupling could arise as a consequence of (1) a greater
probability for CT-mediated interaction between the S0 and
S1 states in 1 than in 2 and/or (2) a larger molecular flexibility
of 1 relative to 2 which promotes the transfer of electronic
energy to vibrational/rotational degrees of freedom.
In light of this notion of conformational flexibility, it is
important to demonstrate that the molecules do not undergo
irreversible photoisomerization under the conditions of our
experiments. Specifically, photoinduced cis-trans geometric
isomerization is known to occur in stilbene and cinnamic
acid derivatives37-40 and, as such, may influence our data.
1
However, in independent absorption, fluorescence, and H
NMR experiments, we ascertained that photoisomerization
Although the solvent-dependent phenomenon shown in
Figure 5 is often described simply in terms of an increase in
(37) Waldeck, D. H. Chem. Rev. 1991, 91, 415–436.
(38) Lub, J.; Ezquerro, M. P.; Malo, B. Mol. Cryst. Liq. Cryst. 2006, 457,
161–180.
the amount of CT character in the emitting state,25,41
a
(39) Saltiel, J. J. Am. Chem. Soc. 1968, 90, 6394–6400.
(40) Brey, L. A.; Schuster, G. B.; Drickamer, H. G. J. Am. Chem. Soc.
1979, 101, 129–134.
(41) Suppan, P.; Ghoneim, N. Solvatochromism; The Royal Society of
Chemistry: Cambridge, 1997.
J. Org. Chem. Vol. 74, No. 23, 2009 9097

Nielsen et al.
JOCArticle
related model which uses slightly different language can
likewise be applied. Specifically, and given the apparent
symmetry of these molecules, one could invoke the solvent-
dependent “symmetry-breaking” arguments of Terenziani
et al.42 Here, the charge distribution in the excited state is
seen to preferentially lose its initial high degree of symmetry
due to a dynamic coupling with the solvent (e.g., coupling
between the vibrational modes of the excited-state solute and
the solvation coordinates). In any event, the net result would
be the solvent-dependent introduction of electronic asym-
metry in the excited state which, for our purposes, is equiva-
lent to a solvent-dependent “increase in the amount of CT
character”.
Finally, it is important to comment on the appearance, or
lack thereof, of vibronic structure on the emission profiles.
Specifically, in the nonpolar solvents cyclohexane and toluene,
where CT character is not expected to play a large role, we see
that the emission profile of 1 is structureless whereas that of 2 is
not. Under these conditions, the data may now reflect the fact
that 1 is more conformationally flexible than 2.
Two-Photon Spectra. In discussing the two-photon spec-
tra shown in Figure 4, we first focus on the marked
increase in δ observed for both molecules as the transi-
tion energy increases (i.e., as the two-photon irradiation
wavelength is decreased below ∼700 nm). This likely
reflects the phenomenon of “resonance-enhancement”,
which becomes significant as the two-photon excitation
wavelength approaches the onset of the one-photon
absorption profile.10,43
in the one-photon spectrum (transition at 425 nm). Selection
rules established for two-photon transitions appear to be
pertinent in this case. Specifically, in a centrosymmetric
molecule, transitions between states with the same parity
are allowed for two-photon processes, whereas an excited
state with a parity opposite to that of the initial state (e.g., a
gerade-ungerade transition) is accessible only through a
one-photon transition.44 Thus, the absorption data obtained
for 2 conform to the Jablonski diagram shown in Figure 1 in
which the states populated in the one- and two-photon
processes are different, and indicate that 2 is a centrosym-
metric molecule.
In contrast, the one- and two-photon absorption bands of
1 closely resemble each other, with a transition λmax of ∼420
nm. Within the framework of the established selection rules,
such spectral coincidence indicates that, unlike 2, molecule 1
appears to be noncentrosymmetric. Thus, our analysis must
clearly go beyond simple static representations of a given
molecule, as shown in Figure 3.
In previous reports, we established that, when working
with large molecules that have multiple degrees of conforma-
tional freedom, care should indeed be exercised when using
qualitative symmetry arguments to interpret one- and two-
photon spectra.10 Rather, it is more appropriate to perform a
rigorous calculation of one- and two-photon transition
probabilities as a function of thermally accessible ground-
state geometric conformers. That is, one needs to consider a
Boltzmann-weighted average of the energetically accessible
conformers.
In the spectral range 700-900 nm, both 1 and 2 have a two-
photon absorption band with a distinct maximum. At this
maximum, in the nominally nonpolar solvent toluene, 1 indeed
has a larger value of δ than does 2 (Table 1), as we had hoped in
the design of our chromophores (vide supra). However, this
difference in δ values is small and barely exceeds our error
limits. The fact that the respective δ values are similar implies
that there is not an appreciable difference in the extent of CT
character in 1 and 2, at least in toluene. In turn, this implies that
the difference in ΦΔ values for 1 and 2 is not due to CT effects
but, rather, reflects another phenomenon.
Note that, as plotted in Figure 4, these two-photon
absorption bands of 1 and 2 appear narrower than the
corresponding lowest energy one-photon absorption bands.
This is a consequence of the different wavelength scales used
to present the one- and two-photon data, respectively. Any
further discussion of the relative band widths, although
interesting, becomes potentially complicated because one
must consider vibronic contributions to the virtual state,
among other things.21 It is also interesting to note the small
peak/shoulder at ∼890 nm in both two-photon spectra.
Although it is possible that this could be an artifact of our
experiment, we rather believe that it is a real, but presently
unassigned, feature of the spectra. In any event, these aspects
of the data do not preclude us from making substantive
statements.
Using an ab initio computational approach based on
response theory,9,10,28 we have calculated one- and two-
photon transition probabilities for both 1 and 2. The results
obtained, shown in the Supporting Information, are consis-
tent with the experimental data shown in Figure 4. We find
that when multiple ground-state geometric conformers of 1
are accessible, the one- and two-photon absorption spectra
are coincident. The computational results specifically sug-
gest that a noncentrosymmetric ground-state conformer of 1
is readily populated (it is, in fact, calculated to be the lowest
energy conformer), and that it contributes to the spectral
data. For compound 2, where the lowest energy ground-state
conformer is centrosymmetric, the first allowed two-photon
transition always occurs at a higher energy than the first
allowed one-photon transition (e.g., see the Jablonski dia-
gram in Figure 1).
Thus, our one- and two-photon spectral data recorded in
the nominally nonpolar solvent toluene, along with the
associated computations, are consistent with our proposed
model in which the differences in the oxygen-dependent
photophysics of 1 and 2 derive principally from the fact that
molecule 1 appears to be more conformationally flexible
than molecule 2.
2. Systematic Changes in the Central Moiety of the OPV
System. Given the results in the preceding section, we
thought it appropriate to investigate the effects of putting
the carbonyl group directly on the central aromatic ring of
our OPV system. We also wanted to assess the effects of
changing the central aromatic moiety through (1) the addi-
tion of functional groups other than a carbonyl and a heavy
atom and (2) replacement of the phenyl ring with a larger
For molecule 2, the distinct two-photon absorption band
has λmax = 770 nm (i.e., transition at 385 nm). This band
is clearly blue-shifted relative to the most dominant band
(42) Terenziani, F.; Painelli, A.; Katan, C.; Charlot, M.; Blanchard-
Desce, M. J. Am. Chem. Soc. 2006, 128, 15742–15755.
(43) Drobizhev, M.; Karotki, A.; Kruk, M.; Rebane, A. Chem. Phys.
Lett. 2002, 355, 175–182.
(44) McClain, W. M. Acc. Chem. Res. 1974, 7, 129–135.
9098 J. Org. Chem. Vol. 74, No. 23, 2009

Nielsen et al.
JOCArticle
TABLE 2. OPV Derivatives Used to Examine the Effects of Systematic
Structural Changes on Selected Photophysical Propertiesa
FIGURE 6. One-photon absorption (solid line) and two-photon
excitation (9) spectra for molecule 12. The upper x-axis shows the
wavelengths used in the two-photon experiment.
parameters, we could readily establish a meaningful picture
of features important in the design of OPV-based two-
photon singlet oxygen sensitizers. The compounds chosen
for this work and the pertinent photophysical parameters
obtained are shown in Table 2.
We start by noting that all values of ΦΔ are quite small and
are significantly less than those obtained for compounds 1
and 2. This is particularly of interest for compounds 3-7
where a carbonyl group was directly attached to the central
aromatic ring. Thus, the data indicate that, for these OPVs,
including carbonyl functionalities in this way does not
appear to be as successful a strategy as using a heavy atom
to induce intersystem crossing.
With regard to heavy atoms, we note that, in the case of
the naphthalene-based sensitizers 14 and 15, the addition
of Br atoms to the central aromatic system indeed appears
to have an appreciable influence on the probability of
intersystem crossing (i.e., ΦF for 15 is much smaller than
that for 14). However, this does not translate into an
increased yield of singlet oxygen in 15. This is in contrast
to the single-ring phenyl-based sensitizers, where ΦΔ for
the dibromo-substituted molecule 2 is appreciably larger
than that for the unsubstituted compound 10. Moreover,
the ΦΔ value for the phenyl-based dibromo compound 2 is
noticeably larger than that for the naphthyl-based dibromo
compound 15. This latter result may reflect an increased
amount of CT character in the napthyl-based system that
adversely affects SΔ. In any event, the key conclusion here
is that, for these OPVs, replacing the central phenyl-based
aromatic moiety with a more extensive naphthyl-based
system does not contribute to the desired effect of increas-
ing ΦΔ.
aAll data reported in this table were recorded in air-saturated toluene.
b
c
d
∼20% uncertainty. ∼10% uncertainty. A value of 0.91 reported by
Kauffman and Moyna45 is consistent with our value. eIn these cases, we
did not record a clear and distinct two-photon absorption band max-
imum (see the Supporting Information). Thus, the absorption cross
f
section reported here likely underestimates the true value. This com-
Two-photon excitation spectra were recorded for all com-
pounds. Although the spectra were, in general, not recorded
with as high a resolution and/or over as large a wavelength
range as those recorded for 1 and 2, the data are sufficient to
yield useful information (see the representative spectrum in
Figure 6; all other spectra can be found in the Supporting
Information.). For most of the molecules listed in Table 2, we
pound does not fluoresce, nor does it produce singlet oxygen. Thus, a
probe based on an optical approach could not easily be used to quantify
two-photon absorption.
naphthyl group. The intent was to survey a comparatively
large number of compounds and, as such, it was sufficient to
only determine values of δ, ΦΔ, and ΦF. On the basis of these
J. Org. Chem. Vol. 74, No. 23, 2009 9099

Nielsen et al.
SCHEME 1. General Procedure Used To Prepare 1
JOCArticle
have clear evidence that, as in the case of molecule 2, the two-
photon spectrum is blue-shifted relative to the one-photon
spectrum. Thus, and as we have discussed with respect to 2 in
the previous section, the spectroscopic data for these com-
pounds point to ground state populations dominated by a
centrosymmetric conformation.
Unlike the values of ΦΔ which are small and essentially
substituent-independent, the δ values are generally large and
respond to the nature of the substituent. This not only
confirms that the OPV class of molecules is indeed well-
suited for two-photon absorption but that the two-photon
transition probability can indeed be sensitive to the func-
tional groups attached to the chromophore. With respect to
the latter, it certainly appears that the largest values of δ are
obtained from molecules in which the central aromatic ring
is functionalized with an electron-withdrawing substituent;
an observation which is consistent with established electron
push-pull arguments.18,20 As an illustration of this, we find
a reasonably good correlation between our δ values and the
NMR chemical shift of the two aromatic protons on
the central ring of our OPVs; the latter parameter reflects
the ground-state electron density in this part of the molecule
which, in turn, is reflected in the extent to which the protons
are shielded (see the Supporting Information for the data).
Given the relative insensitivity of the ΦΔ values to changes in
these particular substituents, it is difficult to generalize and
quantify the extent of correlation between the two-photon
absorption properties and singlet oxygen quantum yields. Never-
theless, it is still useful to look at selected examples where inter-
esting changes of substance are observed. For example, the
effects of replacing a formyl with an acetyl group (i.e., molecules
3 and 6) are arguably larger than what one might expect,
certainly in light of the changes observed with other substituents.
In any event, the data clearly show that, as we established
in our more detailed study of molecules 1 and 2, one should
be cautious when trying to look for and apply rule-of-thumb
arguments in the development of optimized two-photon
singlet oxygen sensitizers. Even for as “simple” a series as
the OPVs in Tables 1 and 2, there apparently is still no
substitute for an in-depth investigation of the pertinent
photophysical properties for a given molecule of interest.
Arguably, one of the more useful conclusions we can make
is that none of the molecules in Table 2 are likely to be
successful candidates for future applications relying on two-
photon induced singlet oxygen production. The ΦΔ values
are simply too low, and despite promising δ values, the
overall product δΦΔ is rather small. This is certainly true
in comparison with δΦΔ values reported, for example, for
various porphyrin derivatives and porphycenes.10,17
Although additional efforts may improve values of δ in such
OPVs, one must be less optimistic about the chance of
significantly enhancing the singlet oxygen yield at the same
time. On the other hand, the majority of these OPV mole-
cules are good fluorophores and would thus be well-suited
for applications based on two-photon induced fluorescence.
Although a great deal is indeed already known about factors
that independently (i) govern singlet oxygen production
in a sensitized process and (ii) influence the probability of
two-photon absorption, the complexity of the combined
two-photon singlet oxygen photosystem still makes simple
qualitative predictions untenable.
We have shown that the well-studied OPV structural
motif examined herein can provide useful information about
structure-property relationships pertinent for a better under-
standing of two-photon-sensitized singlet oxygen production.
However well-suited to nonlinear light absorption, the OPV
system is not ideal for the development of molecules optimized
for the production of singlet oxygen following two-photon
absorption. Although exceptional and suitable compounds
may still be found from this class of molecules, a systematic
approach to successfully develop two-photon singlet oxygen
sensitizers will more likely come from a focus on other classes
of organic molecules.
Experimental Section
Photophysical Characterization. Singlet oxygen quantum
yields were determined by comparing the intensity of singlet
oxygen phosphorescence produced by the molecule under study
to that obtained from a standard.7 In these experiments, a Nd:
YAG nanosecond laser was used as the excitation source, and a
liquid nitrogen cooled Ge-detector was used to record the time-
resolved singlet oxygen phosphorescence decay. Fluorescence
and triplet-state quantum yields, ΦF and ΦT, respectively,
triplet-state energies, ET, and triplet-state quenching parameters
were likewise determined using established procedures (see the
Suppporting Information).
Two-photon excitation spectra, and associated δ values, were
quantified using methods7 wherein either the near-IR phosphor-
escence of singlet oxygen or the sample fluorescence were used as
spectroscopic probes (see the Supporting Information). In these
approaches, recorded data are compared to those obtained from
a reference standard for which the pertinent photophysical
parameters have been independently determined.
General Synthetic Strategy. Important precursors in the synth-
esis of all compounds were 4-(N,N-diphenylamino)-benzaldehyde,
DPAB, and N,N-diphenyl-4-vinylbenzenamine; the latter was ob-
tained from DPAB using a published procedure.46 These com-
pounds were used in either a Horner-Wadsworth-Emmons
(HWE) or Heck reaction to yield the “wings” attached to the central
aromatic moiety of our OPVs (Figure 2).
Synthesis of 1 was achieved using a Knoevenagel condensation
reaction between DPAB and 1-(4-acetyl-2,5-dibromophenyl)etha-
none with potasium tert-butoxide as the base in ethanol (Scheme 1).
A general reason for choosing the diphenylamino end-group is that,
for OPVs and related compounds, this moiety has been shown to
confer better photostability in comparison with dialkylamino
Conclusion
The concious and methodical design of viable and efficient
two-photon singlet oxygen sensitizers is a nontrivial exercise.
(45) Kauffman, J. M.; Moyna, G. J. Org. Chem. 2003, 68, 839–853.
9100 J. Org. Chem. Vol. 74, No. 23, 2009
(46) Sengupta, S. Tetrahedron Lett. 2003, 44, 307–310.

Nielsen et al.
JOCArticle
SCHEME 2. OPV Derivatives Prepared Using the HWE Reaction
SCHEME 3. Naphthalene-Based OPV Molecules Prepared by Using the HWE Reaction
SCHEME 4. Synthesis of the Dibromonaphthalene-Based Diphosphonate Ester Required for the Preparation of Compound 15
groups.45 Furthermore, a diarylamino group is a less efficient
quencher of singlet oxygen than a dialkyl amino group,47,48
phenomenon which can be significant in systems where the local
concentration of the sensitizer is high. The synthesis of compounds 2
and 10, both of which have been published,4,49 usedthesamegeneral
approach as that shown in Scheme 1.
The HWE reaction was used to prepare compounds 11-13
and is outlined in Scheme 2. The details with respect to com-
pound 13 have been published.5,19
The naphthyl-containing OPV compounds were likewise pre-
pared using HWE methodology (Scheme 3). For compound 14,
we used [6-(diethoxyphosphorylmethyl)naphthalen-2-ylmethyl]-
phosphonic acid diethyl ester in this procedure. However, for
compound 15, we needed to make a dibromonaphthalene-based
phosphonate ester. The synthesis of this compound is shown in
Scheme 4. Here, we first prepared 2,6-dimethylnaphthalene in a
condensation reaction with p-methylbenzylmagnesium chloride
and 4,4-dimethoxybutan-2-one. This reaction has previously
been described by Kochetkov et al.,50 where it was claimed that
a 3:2 mixture of sulfuric and phosphoric acid was the best choice
for final hydrolysis and that the use of HBr resulted in signifi-
cantly lower yields. However, the 24% yield of 2,6-dimethyl-
naphthalene obtained in the present work using HBr does not
deviate significantly from the 27% yield reported by Kochetkov
usingH2SO4/H3PO4. Treatmentofthe dimethylnaphthalenewith
Br2 yielded 1,5-dibromo-2,6-dimethylnaphthalene. Subsequent
bromination of the methyl groups with NBS and conversion
using the Michaelis-Arbuzov reaction gave the required phos-
phonate ester needed for the HWE reaction.
a
The HWE reaction works well with reactants insensitive to
basic conditions. However, to incorporate ketones, for example,
into the OPV motif, the carbonyl groups need to be protected if a
HWE strategy is to be applied. Rather, we decided to use a
different approach, the Heck reaction,51 to accomplish the key
coupling step. To this end, we prepared several 2,5-dibromo-
substituted benzenes, which were then reacted with N,N-diphenyl-
4-vinylbenzenamine under Heck conditions as shown in Scheme 5.
One of the most efficient reagents used for the Heck reation
is P(tBu)3,52 but this compound is pyroforic and must be
handled in a glovebox. However, P(tBu)3 can be generated
in situ from the [(tBu)3PH]BF4 salt and a suitable base. We
used the method reported by Netherton and Fu52 where
P(tBu)3 is generated in situ using N-cyclohexyl-N-methylcy-
clohexanamine (Cy2NMe) as the base. The yields for 3-6
(31-59%) are noticeably larger than those for 7-8 (3-
10%) and may reflect the effects of electron withdrawing
(methanesulfonyl, nitro) or sterically hindered (tert-butyl
ketone) groups.
(47) Monroe, B. M. J. Phys. Chem. 1977, 81, 1861–1864.
(48) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data
1995, 24, 663–1021.
(49) Poulsen, T. D.; Frederiksen, P. K.; Jorgensen, M.; Mikkelsen, K. V.;
Ogilby, P. R. J. Phys. Chem. A. 2001, 105, 11488–11495.
(50) Kochetkov, N. K.; Nifantev, E. E.; Nesmeyanov, A. N. Dokl. Akad.
Nauk SSSR 1955, 104, 422–426.
(51) Heck, R. F.; Nolley, J. P. J. Org. Chem. 1972, 37, 2320–2322.
(52) Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295–4298.
J. Org. Chem. Vol. 74, No. 23, 2009 9101

Nielsen et al.
JOCArticle
SCHEME 5. Preparation of OPV Derivatives Using the Heck Reaction
Synthesis Details and Characterization of Compounds
CHCl3) and recrystallization from a 1:1 CH2Cl2/petroleum
ether mixture at 60-80 °C yielded 0.22 g (15%) of 1 as a dark
yellow powder: mp 217-219 °C; 1H NMR(CDCl3, 300 K, 250
MHz) 7.62 (s, 2H), 7.42 (d, 4H, J = 9 Hz), 7.37 (d, 2H, J = 10
Hz), 7.34-7.27 (m, 8H), 7.19-7.07 (m, 12H), 7.00 (d, 4H, J =
9 Hz), 6.89 (d, 2H, J = 16 Hz); 13C NMR(CDCl3, 300 K, 62.5
MHz) 192.9, 151.0, 148.2, 146.5, 143.8, 133.3, 130.2, 129.6,
126.6, 125.8, 124.5, 122.5, 120.9, 118.5. Anal. Calcd for
C48H34Br2N2O2: C, 69.41; H, 4.13; N, 3.37. Found: C,
68.85; H, 4.44; N, 3.22.
(E,E)-2,5-Diformyl-1,4-bis[2-(40-diphenylaminophenyl)vinyl]-
benzene (3). Reaction of Pd2(dba)3 (175 mg, 0.19 mmol), [(t-
Bu)3PH]BF4 (84 mg, 0.29 mmol), Cy2NMe (1.24 g, 6.35 mmol),
1,4-dioxane (50 mL), 2,5-dibromoterephthalaldehyde (0.77 g,
2.64 mmol), diphenyl(4-vinylphenyl)amine (1.56 g, 5.75 mmol),
and DCVC (n-heptane/1,2-C2H4Cl2) yielded 0.90 g (51%) of the
title compound as dark red crystals: mp 238-240 °C; 1H NMR
(CDCl3, 300 K) 10.43 (s, 2H), 8.15 (s, 2H), 7.88 (d, 2H,
J = 16 Hz), 7.45 (d, 4H, J = 9 Hz), 7.34-7.26 (m, 6H),
7.19-7.01 (m, 20H); 13C NMR (CDCl3, 300 K) 192.0, 148.4,
147.3, 138.3, 135.3, 134.3, 130.6, 130.3, 129.4, 128.1, 124.9, 123.5,
2,5-Dibromoterephthaloyl Dichloride (16, See Scheme 1).53
To a stirred solution of 2,5-dibromoterephthalic acid (7.15 g,
22.1 mmol) in toluene (100 mL) was added (COCl)2 (6.05 g,
47.7 mmol) along with a few drops of DMF as catalyst. The
reaction mixture was stirred overnight at 40 °C, and the
solvent was removed under reduced pressure. The isolated
crude product was redissolved in toluene (20 mL), and a
small amount of CaH2 was added. The mixture was stirred
for 1 h at rt and filtered. Removal of the solvent under
reduced pressure left 7.62 g (96%) of the title compound as
off-white flakes: 1H NMR(CDCl3, 300 K) 8.20 (s, 2H); 13
NMR(CDCl3, 300 K) 164.3, 139.8, 137.0, 119.3.
C
1-(4-Acetyl-2,5-dibromophenyl)ethanone(17,SeeScheme1).
Methyllithium (1. 4 M) in Et2O (49 mL, 97 mmol) was added
to a slurry of CuCN (7.00 g, 78.2 mmol) in THF (100 mL) at
-78 °C under an Ar atmosphere, and the mixture was stirred
at this temperature for 1.5 h. Freshly prepared 2,5-dibromo-
terephthaloyl dichloride (7.17 g, 19.9 mmol) was then added,
and thesolution was stirredfor40minat -78°C. The reaction
was quenched at -78 °C by addition of water (150 mL), and
the mixture was allowed to warm to rt and then filtered
through a pad of Celite. The aqueous layer was then extracted
three times with Et2O, and the combined organic phases were
washed with brine and then dried with MgSO4. The solvent
was removed under reduced pressure leaving the crude pro-
duct which was recrystallized from EtOH yielding 2.93 g
(46%) of the title compound as white needles: mp 142-144
122.8, 120.8. Anal. Calcd for C48H36N2O2 0.125C2H4Cl2: C,
3
84.58; H, 5.37; N, 4.09. Found: C, 84.64; H, 5.34; N, 4.15.
(E,E)-2,5-Diethoxycarbonyl-1,4-bis[2-(40-diphenylamino-
phenyl)vinyl]benzene (4). Reaction of Pd2(dba)3 (97 mg,
0.11 mmol), [(t-Bu)3PH]BF4 (47 mg, 0.16 mmol), Cy2NMe
(0.63 g, 3.22 mmol), 1,4-dioxane (50 mL), 2,5-dibromoter-
ephthalic acid diethyl ester (0.40 g, 1.05 mmol), diphenyl(4-
vinylphenyl)amine (0.63 g, 2.32 mmol), and DCVC (1,2-
C2H4Cl2/n-heptane) followed by recrystallization of the
crude product from a 1:1 mixture CH2Cl2 and EtOH
yielded 0.47 g (59%) of the title compound as red needles:
mp 223-224 °C; 1H NMR (CDCl3, 300 K) 8.20 (s, 2H), 7.79
(d, 2H, J = 16 Hz), 7.43 (d, 4H, J = 8 Hz), 7.33-7.21 (m,
8H), 7.16-6.98 (m, 18H), 4.44 (q, 4H, J = 7 Hz), 1.44 (t, 6H,
J = 7 Hz); 13C NMR (CDCl3, 300 K) 167.2, 147.8, 147.5,
137.0, 131.6, 131.4, 129.3, 128.7, 127.8, 124.6, 124.4, 123.4,
°C; 1H NMR (CDCl3, 300 K) 7.66 (s, 2H), 2.63 (s, 6H); 13
C
NMR(CDCl3, 300 K)198.9, 144.0, 133.7, 117.8, 30.1;HRMS-
(EIþ) m/z calcd for C10H8Br2O2þ• 317.8891, found 317.8883.
Anal. Calcd for C10H8Br2O2: C, 37.54; H, 2.52. Found: C,
37.96; H, 2.41.
(E,E)-2,5-Dibromo-1,4-bis[3-(4-diphenylaminophenyl)acryl-
oyl]benzene (1). KOtBu (0.08 g, 0.7 mmol) was added to a
solution of 1-(4-acetyl-2,5-dibromophenyl)ethanone (0.55 g,
1.7 mmol) and 4-diphenylaminobenzaldehyde54 (1.04 g, 3.80
mmol) in EtOH (100 mL), and the reaction mixture was
refluxed for 2 h. The reaction mixture was filtered after
cooling to rt, and Celite was added. The solvent was removed
in vacuo, and the remaining solid was loaded onto a column.
Purification by DCVC55 (petroleum ether (60-80 °C)/
123.2, 61.4, 14.4. Anal. Calcd. for C52H44N2O4 0.125-
CH2Cl2: C, 81.15; H, 5.78; N, 3.63. Found: C, 81.52; H,
5.65; N, 3.65.
3
(E,E)-2,5-Diethylaminocarbonyl-1,4-bis[2-(40-diphenylamino-
phenyl)vinyl]benzene (5). The general procedure for the Heck
reaction was followed, with 5 days of reflux instead of over-
night: Pd2(dba)3 (132 mg, 0.15 mmol), [(t-Bu)3PH]BF4 (71
mg, 0.24 mmol), Cy2NMe (0.80 g, 4.09 mmol), 1,4-dioxane
(50 mL), 2,5-dibromo-N,N,N0,N0-tetraethylterephthalamide
(0.95 g, 2.19 mmol), and diphenyl(4-vinylphenyl)amine (1.21
(53) Lamba, J. J. S.; Tour, J. M. J. Am. Chem. Soc. 1994, 116, 11723–
11736.
(54) Wang, X.-M.; Zhou, Y.-F.; Yu, W.-T.; Wang, C.; Fang, Q.; Jiang,
M.-H.; Lei, H.; Wang, H.-Z. J. Mater. Chem. 2000, 10, 2698–2703.
(55) Pedersen, D. S.; Rosenbohm, C. Synthesis 2001, 16, 2431–2434.
9102 J. Org. Chem. Vol. 74, No. 23, 2009

Nielsen et al.
JOCArticle
g, 4.43 mmol). After the quenched reaction mixture was filtered
through a plug of silica, it was allowed to stand at rt for 2 days.
The precipitate was filtered yielding 0.34 g (44%) of the title
compound as yellow needles: mp 257-258 °C; 1H NMR
(CDCl3, 300 K) 7.53 (s, 2H), 7.36-7.21 (m, 12H), 7.15-6.99
(m, 18H), 6.94 (d, 2H, J = 16 Hz), 4.00-3.74 (bm, 2H),
3.51-3.27 (bm, 2H), 3.15 (q, 4H, J = 7 Hz), 1.32 (t, 6H, J =
7 Hz), 1.02 (t, 6H, J = 7 Hz); 13C NMR (CDCl3, 300 K) 170.1,
147.9, 147.4, 136.5, 133.0, 130.8, 129.3, 127.6, 124.7, 123.3,
123.1, 122.9, 122.1, 42.9, 39.0, 14.0, 13.0. Anal. Calcd for
(4-vinylphenyl)amine (0.74 g, 2.7 mmol), DCVC (n-heptane/
1,2-C2H4Cl2) (2 times) and recrystallization of the crude
product from a mixture of CH2Cl2 and n-heptane yielded
1
0.10 g (10%) of the title compound: mp 221-222 °C; H
NMR (CDCl3, 300 K) 8.26 (s, 2H), 7.41 (d, 4H, J=9 Hz),
7.35-7.24 (m, 10H), 7.21-7.01 (m, 18H); 13C NMR (CDCl3,
300 K) 149.2, 149.1, 147.1, 135.4, 131.7, 129.4, 129.2, 128.4,
125.2, 123.8, 123.6, 122.4, 118.1. Anal. Calcd. for
C46H34N4O4 0.25CH2Cl2: C, 76.30; H, 4.78; N, 7.70. Found:
3
C, 76.15; H, 4.65; N, 8.02.
C56H54N4O2 0.5C4H8O2: C, 81.09; H, 6.80; N, 6.52. Found:
3
[4-(Diethoxyphosphorylmethyl)-2,5-dimethyoxybenzyl]phos-
phonic Acid Diethyl Ester56 (18, See Scheme 2). 1,4-Dimeth-
oxy-2,5-bis-bromomethylbenzene (3.67 g, 11.3 mmol) was
refluxed in P(OEt)3 (100 mL) for 2 h. The reaction mixture
was concentrated by distilling off excess P(OEt)3. Filtering of
the product from the cooled concentrated reaction mixture
yielded 4.22 g (85%) of white crystals: 1H NMR (CDCl3, 300
K) 6.90 (s, 2H), 4.01 (p, 8H, J= 7 Hz), 3.78 (s, 6H), 3.20 (d, 4H,
J = 20 Hz), 1.22 (t, 12H, J = 7 Hz); 13C NMR (CDCl3, 300 K)
151.0 (d, JPC = 4 Hz), 119.4 (d, JPC = 5 Hz), 114.0, 61.7 (d, JPC
= 7 Hz), 56.1, 26.4 (d, JPC = 140 Hz), 16.2 (d, JPC = 6 Hz).
(E,E)-2,5-Dimethoxy-1,4-bis[2-(40-diphenylaminophenyl)-
vinyl]benzene (11). Potassium tert-butoxide (0.55 g, 4.9
mmol) was added to a stirred solution of 1,4-dimethoxy-
2,5-bis-bromomethylbenzene (0.57 g, 1.3 mmol) and 4-
diphenylaminobenzaldehyde (0.73 g, 2.67 mmol) in THF
(100 mL) under an argon atmosphere. The solution was
stirred at rt for 45 min and quenched with water. The
product precipitated from the quenched reaction mixture
and was isolated by filtration: yield 0.45 g (51%); yellow
powder; mp 233-234 °C; 1H NMR (DMSO, 300 K) 7.51 (d,
4H, J = 9 Hz), 7.41-7.21 (m, 14H), 7.15-7.03 (m, 12H),
6.99 (d, 4H, J = 9 Hz), 3.90 (s, 6H); 13C NMR (CDCl3, 300
K) 151.5, 147.6, 147.2, 132.3, 129.3, 128.3, 127.4, 126.7,
124.4, 123.7, 123.0, 121.8, 109.1, 56.4. Anal. Calcd for
C, 80.71; H, 6.46; N, 6.69.
(E,E)-2,5-Diacetyl-1,4-bis[2-(40-diphenylaminophenyl)vinyl]-
benzene (6). Reaction of Pd2(dba)3 (112 mg, 0.12 mmol), [(t-
Bu)3PH]BF4 (63 mg, 0.22 mmol), Cy2NMe (0.74 g, 3.8 mmol),
1,4-dioxane (50 mL), 1-(4-acetyl-2,5-dibromophenyl)-
ethanone (0.37 g, 1.2 mmol), diphenyl(4-vinylphenyl)amine
(0.70 g, 2.6 mmol), and DCVC (1,2-C2H4Cl2/n-heptane) (2
times) followed by recrystallization of the crude product from
a 1:1 mixture of CH2Cl2 and EtOH yielded 0.25 g (31%) of the
title compound as a red powder: mp 249-251 °C; 1H NMR
(CDCl3, 300 K): 7.86 (s, 2H), 7.49-7.36 (m, 6H), 7.32-7.25
(m, 8H), 7.15-7.09 (m, 8H), 7.09-7.01 (m, 8H), 6.98 (d, 2H,
J = 16 Hz), 2.65 (s, 6H); 13C NMR (CDCl3, 300 K) 202.1,
148.0, 147.4, 139.8, 135.4, 132.0, 130.9, 129.3, 127.8, 127.0,
124.7, 123.9, 123.3, 123.2, 30.3. Anal. Calcd for C50H40N2O2
0.33CH2Cl2: C, 82.91; H, 5.62; N, 3.84. Found: C, 82.70; H,
5.58; N, 3.77.
3
(E,E)-2,5-Di-tert-butylcarbonyl-1,4-bis[2-(40-diphenylamino-
phenyl)vinyl]benzene (7). The general procedure for the Heck
reaction was followed, with 5 days of reflux instead of over-
night: Pd2(dba)3 (118 mg, 0.13 mmol), [(t-Bu)3PH]BF4 (79 mg,
0.28 mmol), Cy2NMe (0.73 g, 3.7 mmol), 1,4-dioxane (50 mL),
1-[2,5-dibromo-4-(2,2-dimethylpropionyl)phenyl]-2,2-dime-
thylpropan-1-one (0.70 g, 1.7 mmol), diphenyl(4-vinyl-
phenyl)amine (0.95 g, 3.5 mmol), and DCVC (n-heptane/
1,2-C2H4Cl2) (2 times) and recrystallization of the crude
product from a mixture of CH2Cl2 and n-heptane yielded
0.04 g (3%) of the title compound as a yellow powder: mp
276-277 °C;1H NMR (CDCl3, 300 K) 7.37 (s, 2H), 7.35-7.21
(m, 14H), 7.16-6.99 (m, 18H), 6.93 (d, 2H, J = 16 Hz), 6.78
(d, 2H, J = 16 Hz), 1.28 (s, 18H).
C48H40N2O2 0.5H2O: C, 84.06; H, 6.03; N, 4.08. Found:
3
C, 84.18; H, 5.91; N, 4.10.
[4-(Diethoxyphosphorylmethyl)-2,3,5,6-tetrafluoroben-
zyl]phosphonic Acid Diethyl Ester (19, See Scheme 2). This
compound was a gift from Dr. Frederik C. Krebs (Riso)
and was prepared according to a published procedure.57
(E,E)-2,3,5,6-Tetraflouro-1,4-bis[2-(40-diphenylaminophenyl)-
vinyl]benzene (12). Potassium tert-butoxide (0.50 g, 4.7 mmol)
was added to a stirred solution of [4-(diethoxyphosphoryl-
methyl)-2,3,5,6-tetrafluorobenzyl]phosphonic acid diethyl ester
(0.44 g, 0.89 mmol) and 4-diphenylaminobenzaldehyde (0.55 g,
2.01 mmol) in DMF (80 mL) under an argon atmosphere. The
solution was stirred under reflux for 30 min and cooled to rt
followed by quenching with water. The product precipitated
from the quenched reaction mixture and was filtered, recrystal-
lized from toluene, purified by DCVC (n-heptane/CHCl3), and
then recrystallized from benzene: yield 0.26 g (42%); mp >250
°C; 1HNMR(CDCl3, 300 K) 7.49-7.35 (m, 6H), 7.32-7.22 (m,
10H), 7.17-7.00 (m, 14H), 6.95 (d, 2H, J = 17 Hz); 13C NMR
(CDCl3, 300 K) 148.7, 147.5, 130.9, 129.4, 127.9, 125.0, 123.6,
123.0; 19F NMR (CDCl3, 330 K) -141.7. Anal. Calcd. for
(E,E)-2,5-Dimethylsulfonyl-1,4-bis[2-(40-diphenylaminophenyl)-
vinyl]benzene (8). The general procedure for the Heck-reaction
was followed, with 3 days of reflux instead of overnight: Pd2(dba)3
(103 mg, 0.11 mmol), [(t-Bu)3PH]BF4 (59 mg, 0.20 mmol),
Cy2NMe (0.69 g, 3.5 mmol), 1,4-dioxane (50 mL), 1,4-dibro-
mo-2,5-bis-methanesulfonylbenzene (0.49 g, 1.3 mmol), diphenyl-
(4-vinylphenyl)amine (0.70 g, 2.6 mmol), and DCVC (n-heptane/
1,2-C2H4Cl2) (2 times) and recrystallization of the crude product
from a 5:1 mixture CH2Cl2 and EtOH yielded 0.03 g (3%) of the
title compound: mp >300 °C; 1H NMR (CDCl3, 300 K) 8.47 (s,
2H), 7.88 (d, 2H, J = 16 Hz), 7.45 (d, 4H, J = 8 Hz), 7.35-7.23
(m, 10H), 7.19-7.02 (m, 16H), 3.13 (s, 6H). Anal. Calcd for
C48H40N2O4S2 0.25H2O: C, 74.15; H, 5.25; N, 3.60. Found: C,
3
73.98; H, 5.31; N, 3.59.
(E,E)-2,5-Dinitro-1,4-bis[2-(40-diphenylaminophenyl)vinyl]-
benzene (9). The general procedure for the Heck-reaction was
followed, with 3 days of reflux instead of overnight: Pd2(dba)3
(105 mg, 0.11 mmol), [(t-Bu)3PH]BF4 (67 mg, 0.23 mmol),
Cy2NMe (0.67 g, 3.4 mmol), 1,4-dioxane (50 mL), 1,4-
dibromo-2,5-bisnitrobenzene (0.45 g, 1.4 mmol), diphenyl-
C46H32F4N2 0.5H2O: C, 79.18; H, 4.77; N, 4.01. Found: C,
3
79.18; H, 4.55; N, 3.88.
(56) Brehm, I.; Hinneschiedt, S.; Meier, H. Eur. J. Org. Chem. 2002,
3162–3170.
(57) Krebs, F. C.; Jensen, T. J. Fluorine Chem. 2003, 120, 77–84.
J. Org. Chem. Vol. 74, No. 23, 2009 9103

Nielsen et al.
JOCArticle
(E,E)-2,6-Bis[2-(40-diphenylaminophenyl)vinyl]naphthalene
(14). Prepared as outlined above for 11 and 12. The product
precipitated from the quenched reaction mixture and was iso-
lated by filtration: yield 2.40 g (86%); yellow powder; mp
C50H36N2Br2 0.33CHCl3: C, 69.94; H, 4.24; N, 3.24. Found:
3
C, 69.93; H, 4.09; N, 3.14.
Acknowledgment. This work was funded by the Danish
National Research Foundation through a block grant for
The Center for Oxygen Microscopy and Imaging.
1
226-228 °C; H NMR (CDCl3, 300 K) 7.83-7.66 (m, 6H),
7.44 (d, 4H, J = 9 Hz), 7.37-7.33 (m, 10H), 7.20-7.00 (m,
18H); 13C NMR (CDCl3, 300 K) 147.6, 147.4, 135.1, 133.2,
131.6, 129.3, 128.4, 128.3, 127.4, 127.1, 126.0, 124.6, 124.0, 123.6,
Supporting Information Available: Determination of ΦF,
ΦT, ET, and triplet statequenchingparameters; determination
of two-photon excitation spectra and absorption cross sec-
tions, δ; absorption and fluorescence spectra of 1 and 2 before
and after irradiation; 1H NMR spectra of 1 and 2 before and
after irradiation; computational methods and results; one-
andtwo-photon spectra for compounds3-15;plotofthe two-
photon absorption cross section against the NMR chemical
shift of selected protons; pertinent 1H and 13C NMR spectra
for compounds 1-17. This material is available free of charge
via the Internet at http://pubs.acs.org.
123.1. Anal. Calcd. for C50H38N2 0.5H2O: C, 88.86; H, 5.82; N,
3
4.14. Found: C, 89.02; H, 5.77; N, 4.00.
(E,E)-1,5-Dibromo-2,6-bis[2-(40-diphenylaminophenyl)vinyl]-
naphthalene (15). Prepared as outlined above for 11 and 12. The
product was recrystallized from toluene and then from CHCl3:
yield 0.06 g (21%); yellow powder; mp 269-271-115 °C; 1H
NMR (CDCl3, 300 K) 8.32 (d, 2H, J = 9 Hz), 7.85 (d, 2H, J =
9 Hz), 7.66 (d, 2H, J = 16 Hz), 7.49 (d, 4H, J = 9 Hz),
7.35-7.26 (m, 6H), 7.19-7.02 (m, 20H). Anal. Calcd for
9104 J. Org. Chem. Vol. 74, No. 23, 2009