Near-Infrared BODIPY-Acridine Dyads Acting as Heavy-Atom-Free Dual-Functioning Photosensitizers

Article abstract of DOI:10.1002/chem.202002549

Boron dipyrromethene (BODIPY) dyes represent a particular class within the broad array of potential photosensitizers. Their highly fluorescent nature opens the door for theragnostic applications, combining imaging and therapy using a single, easily synthesized chromophore. However, near-infrared absorption is strongly desired for photodynamic therapy to enhance tissue penetration. Furthermore, singlet oxygen should preferentially be generated without the incorporation of heavy atoms, as these often require additional synthetic efforts and/or afford dark cytotoxicity. Solutions for both problems are known, but have never been successfully combined in one simple BODIPY material. Here, we present a series of compact BODIPY-acridine dyads, active in the phototherapeutic window and showing balanced brightness and phototoxic power. Although the donor–acceptor design was envisioned to introduce a charge transfer state to assist in intersystem crossing, quantum-chemical calculations refute this. Further photophysical investigations suggest the presence of exciplex states and their involvement in singlet oxygen formation.

Full text of DOI:10.1002/chem.202002549

Accepted Article
Accepted Article
Title: Near-Infrared BODIPY-Acridine Dyads Acting as Heavy-Atom-
Free Dual-Functioning Photosensitizers
Authors: Jasper Deckers, Tom Cardeynaels, Huguette Penxten, Anitha
Ethirajan, Marcel Ameloot, Mikalai Kruk, Benoît Champagne,
and Wouter Maes
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202002549
Link to VoR: https://doi.org/10.1002/chem.202002549
01/2020

10.1002/chem.202002549
Chemistry - A European Journal
Near-Infrared BODIPY-Acridine Dyads Acting as Heavy-Atom-Free Dual-Functioning
Photosensitizers
Jasper Deckers,[a,b] Tom Cardeynaels,[a,b,c] Huguette Penxten,[a] Anitha Ethirajan,[b,d] Marcel
Ameloot,[e] Mikalai Kruk,[f] Benoît Champagne,[c], Wouter Maes*[a,b]
[a] UHasselt – Hasselt University, Institute for Materials Research (IMO), Design & Synthesis of Organic
Semiconductors (DSOS), Agoralaan, 3590 Diepenbeek, Belgium
[b] IMEC, Associated Lab IMOMEC, Wetenschapspark 1, 3590 Diepenbeek, Belgium
[c] UNamur – University of Namur, Laboratory of Theoretical Chemistry (LTC), Theoretical and Structural
Physical Chemistry Unit, Namur Institute of Structured Matter, Rue de Bruxelles 61, 5000 Namur, Belgium
[d] UHasselt – Hasselt University, Institute for Materials Research (IMO), Nano-Biophysics and Soft Matter
Interfaces (NSI), Wetenschapspark 1, 3590 Diepenbeek, Belgium
[e] UHasselt – Hasselt University, Biomedical Research Institute (BIOMED), Agoralaan, 3590 Diepenbeek,
Belgium
[f] Belarusian State Technological University, Sverdlov Str., 13а, 220006, Minsk, Belarus
Abstract
Boron dipyrromethene (BODIPY) dyes represent a particular class within the broad array of potential
photosensitizers. Their highly fluorescent nature opens the door for theragnostic applications,
combining imaging and therapy using a single, easily synthesized chromophore. However, near-
infrared absorption is strongly desired for photodynamic therapy to enhance tissue penetration.
Furthermore, singlet oxygen should preferentially be generated without the incorporation of heavy
atoms, as these often require additional synthetic efforts and/or afford dark cytotoxicity. Solutions
for both problems are known, but have never been successfully combined in one simple BODIPY
material. Here, we present a series of compact BODIPY-acridine dyads, active in the
phototherapeutic window and showing balanced brightness and phototoxic power. Although the
donor-acceptor design was envisioned to introduce a charge transfer state to assist in intersystem
crossing, quantum-chemical calculations refute this. Further photophysical investigations suggest the
presence of exciplex states and their involvement in singlet oxygen formation.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
Introduction
To date, thera(g)nostics represent an auspicious concept toward personalized cancer treatment.[1]
The concept implies the combination of diagnosis and a therapeutic approach of the cancerous
disease in order to localize the target, monitor the therapeutic outcome as well as to improve drug
dosimetry.[2] One such a theragnostic platform is the combination of fluorescence bioimaging and
photodynamic therapy (PDT).[3] Established already in 1903, PDT is nowadays a well-recognized and
successful method, which has been clinically approved by the FDA more than two decades ago.[4]
Herein, light of a specific wavelength is absorbed by a photosensitizer (PS) to achieve an
electronically excited singlet state (Sn) (Figure 1). Intersystem crossing (ISC) to a long-living triplet
state (Tn) enables to produce different reactive oxygen species (ROS) via two possible pathways.[5]
Whereas the type I mechanism considers the interaction of the excited PS with the substrate or
solvent, thereby producing a variety of ROS, the type II mechanism involves transfer of energy of the
excited triplet state directly to molecular oxygen (3O2) to produce singlet oxygen (1O2). The ratio of
both mechanisms depends on the characteristics of the PS, the substrate and the amount of available
1
oxygen, but O2 is assumed to be the main cytotoxic agent.[6] It attacks different cellular targets,
including nuclei, plasma membranes, mitochondria, and lysosomes, resulting in cell death by necrosis
or apoptosis, immune responses, or vascular shutdown.[7] PDT has several advantages, such as being
noninvasive, repeatable, and affording low systemic toxicity and minimal side effects.[8] Even more,
the combination of careful illumination of the cancerous cells and the high reactivity, short lifetime,
1
and slow diffusion rate of O2 in biological media opens up the possibility for a very selective
treatment.[9] Thanks to these advantages, PDT is used for the treatment of a wide scope of tumors.[10]
Besides, the technique is frequently applied for treatment of dermatological and ophthalmic
diseases, inactivation of viruses, and to treat bacterial or fungal infections.[4a, 10b, 11]
The ideal PS is stable, has a high molar extinction coefficient (ε > 20,000‒30,000 M-1cm-1), a
sufficiently high singlet oxygen quantum yield (ФΔ), a low dark cytotoxicity, and is easily
synthesized.[8] The activation energy of the PS is important as well. Before the applied light reaches
the target tumor, it will be scattered and absorbed by the surrounding tissue.[12] Shorter wavelengths
are absorbed by different tissue chromophores and often give rise to skin photosensitivity.[6] To
minimize these effects, longer wavelengths are desirable for PS activation. In this way, light
scattering is reduced, the background signal from autofluorescence of biomolecules is minimized,
and the penetration depth is strongly enhanced.[13] As water absorption becomes significant above
1300 nm, the area between 600 and 1200 nm is referred to as the ‘optical window’ of biological
1
tissue.[14] However, wavelengths above 800 nm lack energy to generate O2, since the associated
triplet state would be too low in energy.[15] In this respect, the phototherapeutic area in the near-
infrared (NIR) region (~600‒800 nm) has triggered a lot of interest.[13] Most synthetic and naturally
occurring PSs (whether or not satisfying the aforementioned ideal characteristics) are oligopyrrolic
macrocycles, but a broad scope of other structures are widely studied as well.[6, 16] The majority of
the (pre)clinical stage PSs are derivatives of porphyrins, such as Photofrin, Tookad, and NPe6.[16g]
Although showing several advantages, these compounds are often correlated to dark cytotoxicity,
prolonged skin photosensitivity, and other side effects, explaining the increased interest in
nonporphyrinic PSs.[17]
Recently, 4,4-difluoro-4-bora-3a,4a-diaza-s-indacenes or boron dipyrromethenes (BODIPYs) have
been added to the arsenal of cancer theragnostic materials.[18] Often referred to as porphyrin’s little
sister, BODIPY dyes possess many characteristics to be suitable PSs. These compounds have high
molar extinction coefficients, a high chemical and photochemical stability, and are easily modified,
allowing to fine-tune their photophysical properties.[19] Even more, due to its restricted flexibility, the
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
BODIPY core inherently affords high fluorescent quantum yields (Фf) and a narrow emission profile,
interesting characteristics for the development of a dual-functioning PS. However, there are two
main shortcomings. First, unmodified BODIPY dyes have their absorption and emission peaks around
500 nm. Luckily, a broad scope of structural modifications to the BODIPY core are known to afford
1
NIR activity.[20] The second, even more important, problem is the lack of ability to generate O2 from
BODIPY dyes due to the very poor triplet formation yields. The most common solution is the
introduction of bromine, iodine, or transition metal complexes.[18, 21] Thanks to spin-orbit coupling
(SOC), this induces increased ISC through the heavy-atom effect.[22] In this regard, a vast amount of
NIR photoactive aza-, styryl-, and other BODIPY PSs are known.[18, 21] However, this approach turned
out to be far from ideal as these heavy atoms are correlated to a shortening of the triplet state
lifetimes, increased dark cytotoxicity, and additional synthetic efforts and related costs.[23] Moreover,
fluorescence signals are quenched almost completely when a high ФΔ is reached, since ISC and
fluorescence are competing pathways for the excited state. Aiming at dual-functioning PSs with both
1
a significant O2 production and fluorescence quantum yield, these problems necessitate the search
for alternative methods to achieve a balanced (molecular) brightness (BT, product of ε and Фf) and
phototoxic power (PP, product of ε and ФΔ).
Some alternative methods to increase triplet populations without relying on the heavy-atom effect
have emerged, such as radical enhanced ISC, the use of a spin converter, and charge recombination
induced ISC.[24] In this regard, halogen-free BODIPY PSs were designed over the past few years.[21, 25]
However, radical enhanced ISC and C60 spin converters quench the fluorescence, while the synthetic
effort and related costs increase.[26] Another example are thienopyrrole-fused BODIPYs, but the
triplet formation seems intrinsic to this system.[27] Recently, photoinduced electron transfer (PET)
was applied to donor-acceptor BODIPY dyads and dimers to successfully obtain controllable triplet
states and Фf.[25] In PET, electron transfer or charge separation after light absorption results in the
occupation of a highly polar excited state, i.e. a charge transfer (CT) or charge separated state
1
(CSS).[28] This CT state can undergo charge recombination into a triplet state, thereby rendering
heavy atoms unnecessary for 1O2 generation.[29] Depending on the molecular structure, two different
charge recombination induced ISC mechanisms can be distinguished: radical pair ISC (RP-ISC) and
spin-orbit charge transfer ISC (SOCT-ISC).[24c, 25] The former is more pronounced when there is a large
separation distance between the donor and acceptor parts of the PS, resulting in a weak electronic
coupling. This process involves the formation of an intermediate triplet CT state (3CT) via hyperfine
interaction, followed by fast charge recombination into the localized triplet (3LE) state. The second
mechanism is found in dyads and dimers with a small spacer or no spacer at all between both
1
1
moieties. SOCT-ISC becomes possible if the CT-3LE energy difference is small and the CT state is
3
converted into a LE state through back-electron transfer and spin conversion. To visualize the
difference between traditional PDT based on SO-ISC and the heavy-atom-free RP-ISC and SOCT-ISC
mechanisms, simplified energy diagrams are provided in Figure 1. There are some examples of
1
compact BODIPY dimers and dyads that successfully combine fluorescence and O2 production
through this mechanism.[24] However, to the best of our knowledge, there are no previous examples
of BODIPY PSs wherein 1O2 is obtained efficiently together with a significant Фf when irradiated in the
phototherapeutic window. There is one particular example reported by Ortiz and coworkers, who
synthesized an orthogonal BODIPY dimer whereby one of the two BODIPY units is modified with two
styryl groups at the α-positions.[30] Although having absorption maxima at 509 and 658 nm, and a
fluorescence maximum at 685 nm (Фf = 0.56), 1O2 generation (ФΔ = 0.11) was only observed when the
system was excited using green light. This indicates that triplet formation is occurring in the non-
altered BODIPY unit, whereas this is hampered in the π-extended BODIPY unit. It seems that
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
enlarging the delocalized area enhances the radiative processes from the CT state, thereby hindering
the ISC for triplet formation.
Figure 1: Energy diagrams presenting the principle of spin-orbit intersystem crossing (SO-ISC, left) and radical
pair intersystem crossing or spin-orbit exciplex intersystem crossing (RP-ISC, SOCT-ISC, right). The different
photophysical processes are indicated with red octagons: 1) localized absorption, 2) localized emission, 3)
1
1
1
intersystem crossing (SO-ISC: LE3LE, RP-ISC: CT3CT, SOCT-ISC: CT3LE), 4) energy transfer to molecular
oxygen, 5) charge transfer absorption, 6) charge transfer emission, 7) photoelectron transfer, and 8) internal
conversion.
In this manuscript, we present a series of novel dual-functioning BODIPY PSs showing a reasonably
high 1O2 production, while maintaining satisfying fluorescent abilities, all within the phototherapeutic
window. We started from different highly fluorescent 1,3,5,7-tetramethyl-BODIPY cores and shifted
the absorption and emission to the NIR region through a simple Knoevenagel-type condensation of
the α-methyl groups. In accordance with the idea of an SOCT-ISC mechanism, acridine moieties were
attached to provide an orthogonal donor-acceptor design. Density functional theory (DFT) was used
to investigate the BODIPY dye geometries and frontier molecular orbitals. Extensive photophysical
characterization of the dyads was performed in different solvents and the influence of the
incorporated dimethylacridine end groups was examined. Although these photophysical data at first
sight seemed to match with the idea of a CT state being involved, time-dependent (TD) DFT
calculations starting from the ground state geometries did not confirm this mechanism. Further
investigations unveiled the probable presence of ‘exciplexes’[31] in these systems. As their nature is
similar to the highly polar CT state, they could also play an important role in singlet oxygen
formation.[32] Since exciplex singlet and triplet states are close to each other in energy, they serve the
same purpose as the CT energy levels depicted in Figure 1, thereby assisting triplet formation.
Results and discussion
Structural design and synthesis
The design of the PSs started from the highly fluorescent 1,3,5,7-tetramethyl-BODIPYs 7a-c, as
depicted in Scheme 1. The methyl groups are acidic enough to enable Knoevenagel-type
condensations to extend the π-conjugated system.[33] The introduction of vinyl groups provides a
direct and easy way to obtain bathochromically shifted absorption and emission features. We chose
for a strongly electron donating acridine end group, as often applied for materials displaying
thermally activated delayed fluorescence (TADF).[34] Because the BODIPY core is relatively electron
deficient, the introduction of an electron donating group via Knoevenagel condensation generates a
push-pull type structure, further increasing the bathochromic shift. This donor-acceptor dyad design
was supposed to introduce the desired CT state required for efficient ISC.[24c] Although the exact
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
structural requirements for SOCT-ISC are still not very well understood, orthogonality between the
BODIPY core and the rest of the structure is suggested to support this process as the change in
molecular orbital angular momentum will compensate for the change in electron spin angular
momentum during ISC, hence slowing down the charge recombination process to the ground state
and enhancing triplet formation.[29a] In our design, the C-N induced torsion between the donor and
acceptor part is hence an essential aspect.
Scheme 1: Synthesis of the BODIPY PSs and their building blocks: i) thionyl chloride, methanol, 0 °C; 16 h at 60
°C (77%); ii) methylmagnesium bromide, dry THF, -78 °C; 12 h at RT (92%); iii) H3PO4, 2 h at 35 °C (75%); iv) 4-
bromobenzaldehyde, Pd2(dba)3, Cs2CO3, dppf, dry toluene, inert atm, 16 h at 100 °C (91%); v) 5-
bromothiophene-2-carbaldehyde, Pd(OAc)2, NatBuO, XPhos, dry toluene, inert atm, 16 h at 100 °C (61%); vi)
triethoxymethane, trifluoroacetic acid, dry CH2Cl2, inert atm, 2 h at RT; Et3N, BF3·OEt2, dry toluene, 1 h at RT
(47%); vii) 2,4,6-trimethylbenzaldehyde, trifluoroacetic acid, dry CH2Cl2, inert atm, 3 h at 0 °C; DDQ, 20 min at 0
°C, 1 h at RT; Et3N, BF3·OEt2, 12 h at RT (72%); viii) 2,6-dichlorobenzaldehyde, trifluoroacetic acid, dry CH2Cl2,
inert atm, 1h at RT; DDQ, 10 min at 0 °C, 1 h at RT; Et3N, BF3·OEt2, 2 h at RT (23%); ix) glacial acetic acid,
piperidine, dry DMF, inert atm, 5 min at 150 °C, microwave irradiation (40-66%).
Three different BODIPY cores with varying meso-groups – hydrogen (7a), mesityl (7b), and 2,6-
dichlorophenyl (7c) – were synthesized according to literature procedures starting from 2,4-
dimethylpyrrole (6).[35] Dimethylacridine 4 was obtained via a three-step synthesis from N-
phenylanthranilic acid (1) with a good overall yield of 53%. As an aldehyde functionality is needed for
the Knoevenagel condensation, a spacer should be introduced between the donor an acceptor
moiety. To screen two spacer types in the target PSs, dimethylacridine 4 was subjected to a
Buchwald-Hartwig amination with either 4-bromobenzaldehyde or 5-bromothiophene-2-
carbaldehyde. The thienyl spacer was envisioned to provide an additional small red-shift. First tests
were performed using Pd2(dba)3 and BINAP, but without success. When using Pd(OAc)2 and P(t-
Bu)3HBF4, aldehydes 5a and 5b were obtained in moderate yields (33 and 43%, respectively). Further
optimization was performed using Pd(OAc)2 in combination with XPhos, increasing the yields to 54
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
and 61%, respectively. Eventually, we were able to maximize the yield of 5a to 91% using Pd2(dba)3
and 1,1′-bis(diphenylphosphino)ferrocene (dppf) as the catalytic system.
The Knoevenagel condensation itself was initially optimized for BODPY 8b, as summarized in Table 1.
The condensation of an aldehyde with a methyl group at the BODIPY 3- and 5-positions is well-
known.[20] However, refluxing in benzene overnight in the presence of 3 Å molecular sieves did not
result in the target product. A switch was made to microwave irradiation, yielding 36% of BODIPY 8b
in only 5 minutes. Further optimization pointed out that a higher concentration is preferred (entries
3 and 5), a longer reaction time results in more degradation (entry 4), more equivalents of
benzaldehyde push the reaction to the right (entry 6), and a temperature of 150 °C is ideal to
increase the conversion (entries 6 to 8). Finally, the optimal reaction conditions were used to
combine aldehydes 5a and 5b with BODIPYs 7a-c to obtain six different π-extended BODIPY dyes.
Three of them, BODIPYs 8a-c, bear a phenyl spacer between the BODIPY core and the acridine
moiety, while dyads 9a-c are equipped with a thienyl spacer. Reactions with BODIPY 7a were
performed with only three equivalents of aldehydes 5a-b as tri- and tetra-substitution was observed
to a larger extent, likely due to the lack of a sterically hindering group at the meso-position. In the
end, good yields ranging from 40 to 66% were obtained for all final products. Dyads 8a-c have a blue
color in solution, while BODIPYs 9a-c give a green solution, thereby already indicating a variation in
spectroscopic properties due to the different spacer. For a detailed description of the synthesis
protocols, we refer to the supporting information.
Table 1: Optimization of the Knoevenagel-type condensation toward BODIPY 8b.[a,b]
Reaction
time
(min)
960
5
Solvent
volume
(mL/mmol)
200
Yield[c]
8b-
Equiv Equiv Temperature
Yield[c] Remnant[c]
Entry
7b
5a
(°C)
8b
7b
mono[d]
0%
[e]
1
2
3
4
5
6
7
8
1
1
1
1
1
1
1
1
3
3
3
3
3
4
4
4
80
0%
-
[e]
150
150
150
150
150
175
125
240
120
120
60
60
60
60
7%
7%
9%
6%
9%
11%
17%
36%
46%
46%
53%
66%
51%
45%
-
5
15
5
5
5
44%
27%
34%
23%
10%
35%
5
[a]
All reactions were done on a 0.1 mmol scale, except for entries 1 and 2, which were done on a 0.05 mmol
[b]
scale. Experimental conditions for entry 1: 0.05 mL glacial acetic acid, 0.05 mL piperidine, 3 Å molecular
sieves, benzene, inert atmosphere, reflux. Experimental conditions for entries 2 to 8: 0.1 mL glacial acetic acid,
[c]
0.1 mL piperidine, dry DMF, inert atmosphere, microwave irradiation, 1 min pre-stirring. All reported yields
are isolated yields. [d] Monostyryl side product of distyryl-BODIPY 8b. [e] Not isolated.
Structural analysis
To confirm our design objectives, DFT was used to optimize the ground state geometries of BODIPYs
8a-c and 9a-c in three different solvents with varying polarity (toluene, chloroform, and dimethyl
sulfoxide (DMSO)) using the polarizable continuum solvation model (PCM).[36] All vibrational
frequencies are real, demonstrating the optimized geometries correspond to minima on the
potential energy surface. From the optimized geometries in chloroform (Figure 2 and Figure 3), a
large dihedral angle (θ > 80°) between the donor (acridine) and acceptor (styryl/thienylvinyl-BODIPY)
units can be seen. The phenyl and thienyl spacers are coplanar with the BODIPY core, thus extending
the conjugation. For all BODIPYs, the highest occupied molecular orbital (HOMO) and lowest
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
unoccupied molecular orbital (LUMO) reside on the styryl/thienylvinyl BODIPY system. The HOMO-1
and HOMO-2 can be found on the acridine units, while the HOMO-3 is again found on the
styryl/thienylvinyl BODIPY core (Figure S1 and S2). The mesityl and 2,6-dichlorophenyl meso-groups
in 8b/9b and 8c/9c, respectively, are nearly perpendicular to the rest of the BODIPY core and almost
completely electronically decoupled. Therefore, these groups are not expected to influence the
experimental and calculated energy levels to a large extent. The geometries and frontier orbital
topologies are similar in toluene and DMSO.
8a
8b
8c
HOMO
LUMO
Figure 2: Optimized geometries and frontier orbitals for BODIPYs 8a-c in chloroform solution obtained using
DFT calculations with M06-2X/6-311G(d) and the PCM model. Isosurface values of 0.02 (a.u.) were used for all
orbitals.
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
9a
9b
9c
HOMO
LUMO
Figure 3: Optimized geometries and frontier orbitals for BODIPYs 9a-c in chloroform solution obtained using
DFT calculations with M06-2X/6-311G(d) and the PCM model. Isosurface values of 0.02 (a.u.) were used for all
orbitals.
Photophysical characterization
The six heavy-atom-free potential BODIPY PSs were subjected to a detailed photophysical
characterization to examine their intended use as theragnostic agents. Initial photophysical
characterization of BODIPY dyes 8a-c and 9a-c was performed in chloroform solution. Absorption and
emission spectra afforded the spectral maxima (λabs(max), λem(max)), Stokes shifts (Δ휈̅), and the full-
width-at-half-maximum (fwhmabs, fwhmem) of both the emission and absorption bands. Fluorescence
quantum yields (Фf) were obtained relative to a suitable reference dye and using the corresponding
excitation wavelengths. Singlet oxygen quantum yields (ФΔ) were collected by monitoring the
absorbance of 1,3-diphenylisobenzofuran (1,3-DPBF) as singlet oxygen scavenger during excitation at
639 nm (Figure S4).[37] In combination with the molar attenuation coefficient (ε), the brightness (BT)
and phototoxic power (PP) have been determined. All data represented in Figure 4 and Table 2 are
mean values which originate from three independent measurements for each compound. Only data
from the wavelength region of interest are displayed here. For the full absorption spectra, we refer
to Figure S3 and Table S2. Furthermore, all systems were found to be photostable for more than
three hours under 639 nm illumination.
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
Figure 4: Normalized absorption spectra of BODIPYs 8a-c and 9a-c (top) and their corresponding normalized
fluorescence emission spectra (bottom) in chloroform solution.
Table 2: Spectroscopic data for BODIPY dyads 8a-c and 9a-c as obtained in chloroform solution.[a]
λabs
λem
Δ휈
̅
fwhmabs fwhmem
ε
BT
PP
[i]
BODIPY
Фf[h]
ФΔ
(nm)[b] (nm)[c] (cm-1)[d] (cm-1)[e] (cm-1)[f] (M-1cm-1)[g]
(M-1cm-1)[j] (M-1cm-1)[k]
8a
643
633
648
674
664
681
654
645
664
270
294
379
316
311
344
748
777
768
767
805
781
883
917
122,300
119,900
100,100
107,300
98,200
0.47
0.29
57,600
75,200
38,200
7,400
33,100
27,800
30,400
2,600
±0.02 ±0.04
0.63 0.23
±0.03 ±0.02
0.38 0.31
±0.02 ±0.07
0.069 0.03
±0.002 ±0.01
0.091 0.05
±0.002 ±0.03
0.052 0.06
±0.002 ±0.04
8b
8c
9a
9b
9c
1431
689,
764
678,
751
697,
782
793,
1154
768,
1190
743,
1111
9,000
6,000
98,400
5,100
5,200
[a]
Spectrograde chloroform was used for all measurements. All values are averages from three independent
[b]
[c]
[d]
measurements. Absorption maximum. Fluorescence emission maximum/maxima. Stokes shift between
the localized absorption and emission maxima. Full-width-at-half-maximum of the absorption band. Full-
width-at-half-maximum of the emission band(s).
[e]
[f]
[g]
[h]
Molar attenuation coefficient.
Fluorescence quantum
yield determined vs nile blue (Фf = 0.27, λexc = 605 nm) (for 8a-c) or vs aluminum phthalocyanine chloride (Фf =
[i]
0.41, λexc = 645 nm) (for 9a-c) in spectrograde ethanol as a reference. Standard deviations are reported.
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
Singlet oxygen quantum yield determined vs methylene blue (ФΔ = 0.52, λexc = 639 nm) in spectroscopic grade
ethanol as a reference by monitoring the absorbance of 1,3-DPBF at 414 nm. Standard deviations are reported.
[j] Fluorescence brightness. [k] Phototoxic power.
For styryl-BODIPYs 8a-c, introduction of the acridine-styryl groups resulted in a bathochromic shift of
around 140 nm for the absorption maximum with respect to the starting BODIPY cores 7a-c. The
fluorescence maxima change in the same manner, retaining the small Stokes shift characteristic of
BODIPY dyes. As such, these three materials are active in the targeted phototherapeutic window. The
Фf values decrease, as expected for red-shifted dyes due to the energy-gap law, but are still
significant (47, 63, and 38% for 8a, 8b, and 8c, respectively). Both the absorption and emission
profiles resemble the typical narrow BODIPY shape with a smaller shoulder. To our delight, the
BODIPY dyads with a phenyl spacer between the BODIPY and acridine subunits were able to generate
1O2 when irradiated using a 639 nm LED. Relative ФΔ measurements were performed, resulting in
good values ranging from 23% for the meso-mesityl variant 8b to 31% for the meso-2,6-
1
dichlorophenyl variant 8c. We can exclude that the O2 generation capability originates from the
orthogonality between the meso-group and the BODIPY core, since meso-unsubstituted BODIPY 8a
also affords a significant ФΔ value of 29%. Moreover, the high molar attenuation coefficients (above
100,000) indicate their excellent light harvesting abilities. This results in high and balanced BT and PP
values, enabling the use of these dyes for combined bio-imaging and photodynamic therapy.
The absorption profiles of BODIPYs 9a-c resemble these of their styryl counterparts, but are red-
shifted to a greater extent, as expected due to the more electron donating character of the thienyl
linkers. Bathochromic shifts up to 170 nm with respect to BODIPYs 7a-c are observed, bringing the
absorption deeper into the NIR. An additional, small absorption band is seen around 740 nm.
Remarkably, the thienyl spacer results in a very different fluorescence emission profile. The first
emission maxima are found between 680 and 700 nm, corresponding to small Stokes shifts.
However, a second, more intense and broad emission band with maxima ranging from 750 up to 780
nm appears. Although high molar attenuation coefficients are obtained for these systems, Фf and ФΔ
values are found to be very low. Clearly, dyads 9a-c behave very differently with respect to the
distyryl analogues 8a-c, making them unsuitable as dual PSs.
Elucidation of the spectral differences
To get an idea on the origin of the observed spectral differences, the nature of the second emission
band observed for molecules 9a-c was investigated in more detail. Since the dyads were designed to
form an intramolecular CT state, this was examined first. A possible explanation for the lack of
fluorescence and phototoxicity for 9a-c could be the (non-)radiative charge recombination from the
CT state to the ground state. The combination of non-radiative losses and restricted triplet formation
would result in the observed low Фf and ФΔ values.
To verify the CT character of the lowest energy fluorescence band in BODIPY dyads 9a-c, and the
possible involvement of a CT band in the efficient ISC for 8a-c, the effect of solvent polarity on the
absorption and emission profiles was investigated. When increasing solvent polarity, the highly polar
CT state is stabilized, resulting in a bathochromic shift and broadening of the CT band, accompanied
with a decrease in Фf.[38] However, the CT state is not always visible through its emission, as it could
also undergo non-radiative charge recombination to the ground state or recombination into a triplet
state, depending on the polarity of the medium.[29] BODIPYs 8a-c and 9a-c were hence subjected to
the same photophysical characterization as performed in chloroform (ET(30) = 39.1), but now in the
more apolar solvent toluene (ET(30) = 33.9) and the more polar solvent DMSO (ET(30) = 45.1).[39]
Normalized absorption and emission spectra are given in Figure 5. Here, the profiles are grouped by
10
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
BODIPY PS in different solvents to more clearly visualize the influence of polarity changes. The
extracted data from three independent measurements in each solvent are summarized in Table S3.
Only data from the wavelength region of interest are displayed here. For the complete absorption
spectra in each solvent, we again refer to Figure S3 and Table S2. The monitored decrease of 1,3-
DPBF absorption during 1O2 measurements in each solvent is also illustrated in Figure S4.
Figure 5: Normalized absorption spectra (solid lines) of BODIPYs 8a-c and 9a-c and their corresponding
normalized fluorescence emission spectra (dashed lines) in toluene, chloroform, and DMSO.
The absorption maxima and profiles of PSs 8a-c are not influenced to a large extent by solvent
polarity, indicating an unaffected localized singlet (1LE) energy in different solvents. However, the
fluorescence signals are distinctly different. When going from chloroform to toluene, the localized
emission peak, originating from the radiative 1LES0 transition, is almost identical, as expected from
1
the rigid LE state. The same can be seen in the more polar solvent DMSO, but with the arise of a
second band. This red-shifted emission peak is clearly observed for BODIPY 8a at 728 nm. It is also
11
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
present in dyads 8b and 8c, although partially overlapping with the localized emission peak. The
variable nature depending on solvent polarity suggests a CT character for these emission bands. The
difference in the position of this band between the three BODIPY dyes bearing a phenyl spacer is
remarkable. Whereas both emission bands are visually separated for 8a, these bands are closer to
each other in BODIPY 8c, whereas for 8b it is only seen as an additional broadening of the localized
emission peak. This would suggest a close encounter of the 1LE and 1CT energy levels in BODIPY 8b, in
contrast to the larger energy difference in 8a. The change in solvent polarity also affects the Фf and
ФΔ values (Table S3). The fluorescence capability in toluene is higher as compared to chloroform for
1
all styryl-BODIPYs. This occurs at the expense of the O2 generation capabilities. Surprisingly, the
opposite is not the case in DMSO. In this polar solvent, both Фf and ФΔ are close to zero.
Different trends are observed when comparing the photophysical characteristics of BODIPYs 9a-c in
toluene, chloroform, and DMSO. Concerning photon absorption, comparable absorption profiles and
maxima are observed. As in chloroform, a second, although relatively weak, low-energy absorption
band can also be observed in toluene and DMSO, which is not present in the 8a-c series. Regarding
the fluorescence signals, the same influence of solvent polarity was noticed for 9a, 9b, and 9c. In all
1
solvents, localized emission from the LE state is very similar. The second peaks again change in
accordance to the polarity of the surrounding medium, as expected for CT emission bands. The
emission maxima decrease by approximately 13 nm when going from chloroform to toluene, while a
16 nm red-shift is obtained in DMSO. The bands also broaden with increasing solvent polarity. In all
1
solvents, fluorescence and O2 production remain relatively low, which could be due to extensive
non-radiative losses by the presence of the mobile thienyl spacer. Фf values seem to increase slightly
in apolar media, whereas ФΔ values increase a bit in more polar environments.
From these results, it looks like CT states could be involved in all systems, although its presence is not
always observed spectroscopically. To get additional confirmation, TDDFT calculations were run to
verify the position and CT character of this additional energy level. Despite known issues in
accurately describing the singlet and triplet excited states for BODIPY type compounds related to CT
transitions, multireference character states and double excitations,[40] Chibani et al. showed that
using M06-2X as the exchange correlation functional in TDDFT calculations on BODIPY compounds
can give fairly accurate representations of the excitation energies.[41] Furthermore, M06-2X has been
found to give good estimates for the singlet and triplet excitation energies for various small
molecules and donor-acceptor compounds when benchmarked versus MS-CASPT2, CC2 and CC3
calculations.[42] Unfortunately, due to the size of the systems under investigation here, such
wavefunction methods are not feasible as the computational demands are too high. Therefore,
TDDFT calculations were performed at the M06-2X/6-311G(d) level to calculate the vertical excitation
energies of the singlet and triplet states in toluene, chloroform, and DMSO with the PCM model
(Table 3, visualized in Figure S5). The singlet excitation energies for BODIPYs 8a-c show that 8b has
the highest first excitation energy, followed by 8a and 8c. This is expected given the electron
donating nature of the mesityl group, whereas the 2,6-dichlorophenyl group is slightly electron
withdrawing. BODIPYs 9a-c have lower excitation energies, in correspondence to the
bathochromically shifted absorption and emission spectra. Furthermore, these compounds follow
the same trend concerning variations of the meso-groups. In all cases, the energy differences
between the different solvents are minimal, suggesting similar absorption maxima. These
observations correspond nicely to the trends that were also found experimentally (Figure 5, Table
S3). The oscillator strength of the S0S1 transition is large (> 0.9) for all compounds, hinting to a large
extinction coefficient for this transition. The second singlet energy level is much higher in energy (>
0.7 eV) in all cases, with the corresponding oscillator strength being negligible, except for dyad 9a in
DMSO and 9c in all three solvents, where it is very small. There are two triplet states of interest in
12
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
these BODIPY compounds. The lowest triplet state has an excitation energy of around 1.00 eV for all
molecules. The second triplet state is around 2.49 eV for BODIPYs 8a-c and around 2.24 eV for
BODIPYs 9a-c. According to Kasha’s rule, energy transfer from the BODIPY triplet state to the triplet
ground state of molecular oxygen should occur from the lowest triplet state.[43] With an energy of
3
around 1.00 eV, there is still enough energy to overcome the excitation barrier for O2, which is
around 0.98 eV (1270 nm). Based on the calculations, and keeping the possible error on the obtained
triplet excitation energies in mind, the lowest triplet state in 9a and 9c might be too low in energy for
1
efficient O2 formation, as observed experimentally (Table S3). The dominant nature of the one-
particle transitions for the first singlet states is of HOMOLUMO (localized) character. For the
second singlet excited state, the character varies between HOMO-1LUMO (8a, 8b chloroform, 9a,
9b, and 9c DMSO) and HOMO-2LUMO (8b toluene and DMSO, 8c and 9c toluene and chloroform)
character, depending on the compound and the solvent in which it was calculated, but the nature
remains the same, since both HOMO-1 and HOMO-2 are localized on the acridine moiety, leading to
a CT-type excitation. For the first and second triplet excited states, the dominant nature is of
HOMOLUMO and HOMO-3LUMO character, respectively.
Table 3: Calculated vertical singlet (S1 and S2) and triplet (T1 and T2) excitation energies and their corresponding
oscillator strengths for BODIPYs 8a-c and 9a-c.
S1
S2
Osc.
T1
Nature[d]
T2
Nature[d]
BODIPY
Solvent[a]
Toluene
E
Osc.
(eV)[b] (nm)[b] Str.[c]
2.17
E
E
E
λ
Nature[d]
Nature[d]
(eV)[b] Str.[c]
2.96 0.00
(eV)[b]
1.05
(eV)[b]
2.51
8a
1.25
1.24
1.23
1.16
1.15
1.14
1.19
1.18
1.17
1.08
1.06
0.90
1.00
0.99
HL
(98%)
HL
(98%)
HL
(97%)
HL
(95%)
HL
(96%)
HL
(97%)
HL
(97%)
HL
(97%)
HL
(97%)
HL
(97%)
HL
(97%)
HL
(96%)
HL
H-1L
HL
(94%)
HL
(94%)
HL
(94%)
HL
(91%)
HL
(92%)
HL
(93%)
HL
(93%)
HL
(94%)
HL
(94%)
HL
(93%)
HL
(93%)
HL
(91%)
HL
(92%)
HL
H-3L
572
569
567
560
558
557
573
571
569
598
596
591
585
583
(91%)
H-1L
(87%)
H-1L
(85%)
H-2L
(69%)
H-1L
(88%)
H-2L
(88%)
H-2L
(91%)
H-2L
(84%)
H-2L
(83%)
H-1L
(76%)
H-1L
(88%)
H-1L
(82%)
H-1L
(76%)
H-1L
(78%)
(49%)
H-3L
(48%)
H-3L
(48%)
H-3L
(51%)
H-3L
(51%)
H-3L
(50%)
H-3L
(52%)
H-3L
(52%)
H-3L
(51%)
H-3L
(48%)
H-3L
(47%)
H-3L
(47%)
H-3L
(49%)
H-3L
(49%)
Chloroform 2.18
3.00 0.00
3.04 0.00
3.03 0.00
3.05 0.00
3.08 0.00
2.93 0.00
2.95 0.00
2.96 0.00
3.02 0.00
3.05 0.00
3.06 0.01
3.08 0.00
3.09 0.00
1.06
1.07
1.16
1.16
1.17
1.06
1.06
1.07
0.99
1.00
1.03
1.10
1.10
2.51
2.51
2.50
2.49
2.50
2.49
2.49
2.489
2.26
2.25
2.27
2.25
2.25
DMSO
2.19
2.21
8b
8c
9a
9b
Toluene
Chloroform 2.22
DMSO
2.23
2.16
Toluene
Chloroform 2.17
DMSO
2.18
2.07
Toluene
Chloroform 2.08
DMSO
2.10
2.12
Toluene
(97%)
HL
(97%)
Chloroform 2.13
(92%)
13
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
DMSO
2.13
2.07
583
599
597
596
0.94
1.06
1.05
1.03
HL
(97%)
HL
(97%)
HL
(97%)
HL
(97%)
3.10 0.00
3.11 0.01
3.12 0.01
3.13 0.01
H-1L
(87%)
H-2L
(88%)
H-2L
(88%)
H-1L
(88%)
1.10
1.00
1.00
1.00
HL
(92%)
HL
(93%)
HL
(93%)
HL
(93%)
2.24
2.24
2.24
2.23
H-3L
(48%)
H-3L
(49%)
H-3L
(49%)
H-3L
(49%)
9c
Toluene
Chloroform 2.08
DMSO 2.08
[a] Solvents are listed from top to bottom according to increasing polarity ET(30). [b] Vertical excitation energy. [c]
Oscillator strength. [d] H = HOMO, L = LUMO.
The CT character of the various transitions was investigated and the distance over which charge is
transferred (denoted as dCT) and the change in dipole moment upon excitation were calculated
(Table 4). The charge-transfer distance was calculated according to the work by Le Bahers et al.[44]
From the dCT values (< 1.0 Å), it is clear that the S0S1, S0T1, and S0T2 transitions are of local
character. On the other hand, the S0S2 transition is of CT character (dCT >> 1.0 Å). The change in
dipole moment (Δµ) follows the same trend: a large increase in dipole moment is observed for S0S2,
whereas the other transitions show a slight decrease in dipole moment upon excitation. These
findings are visualized from the excited state ‒ ground state electron density differences (Figure S6
and S7) and are in line with the picture drawn by the molecular orbitals and natures of the various
transitions.
Table 4: Amount of charge-transfer character (dCT) and change in dipole moment (Δµ, excited state dipole -
ground state dipole) accompanying the S0->Sn and S0->Tn transitions in chloroform.
S0S1
S0S2
S0T1
S0T2
BODIPY
dCT (Å)[a] Δµ (D)[b] dCT (Å)[a] Δµ (D)[b] dCT (Å)[a] Δµ (D)[b] dCT (Å)[a] Δµ (D)[b]
8a
8b
8c
9a
9b
9c
0.57
0.67
0.79
0.82
0.94
0.99
-1.1
-1.4
-1.6
-1.7
-2.0
-2.1
4.86
5.00
5.13
4.53
4.79
4.93
17.8
18.1
17.8
18.1
19.3
20.8
0.48
0.65
0.71
0.52
0.71
0.71
-1.2
-1.8
-2.0
-1.5
-2.1
-2.2
0.47
0.38
0.62
0.64
0.62
0.74
-0.7
-0.6
-1.0
-1.2
-1.1
-1.4
[a]
[b]
Distance over which charge is transferred between the indicated states upon excitation. Dipole moment
change upon excitation.
The TDDFT calculations correspond well with the observed localized absorption and emission
spectra, in which very little solvatochromism is observed (Figure 5, Table S3). However, from Table 4,
the supposed CT emission as found in dyads 8a and 8c in DMSO, and 9a-c in all solvents, could not be
confirmed at the theoretical level. There is indeed an energy level with CT character, but it is much
higher in energy than the first singlet state, and not slightly lower as would be expected from the
photophysical data, thereby contradicting the observed bathochromically shifted broad band as
being a CT band. Furthermore, oscillator strengths are negligible for this transition in all cases.
Therefore, another process is expected to be involved in these BODIPY systems.
For SOCT-ISC processes, the Gibbs free-energy change (ΔGCS) can be estimated using the Rehm-
Weller equation to check the feasibility of a possible charge separation to form a CT state.[45]
However, for our systems we came across several problems when trying to determine ΔGCS. Although
the structural prerequisites seem to be fulfilled in the BODIPY-acridine dyads, TDDFT calculations
14
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
suggest that no CT state is involved in triplet formation. Furthermore, a definition of the true donor
and acceptor parts is troublesome in these systems as the spacer itself also has a donating character
and is coplanar with the BODIPY acceptor. Even more, the double bonds can play an important role in
terms of the distance between donor and acceptor. Efforts were made to calculate ΔGCS for BODIPYs
8a and 9a (see supporting information). To this extent, distyryl-BODIPY 8x and thienylvinyl-BODIPY 9x
were synthesized and denoted as the electron acceptor components,[46] whereas 9,9-dimethyl-9,10-
dihydroacridine (4) was applied as the electron donor. The results seem to point to a possible lower-
lying CT state, but can hardly be taken as a solid proof because of the above considerations.
To demonstrate the crucial influence of the acridine end groups on the fluorescence properties and
phototoxicity, we compared the photophysical characteristics of dyads 8a and 9a with their
constituent parts (BODIPY core 7a and acridine 4) and the analogous dyes without dimethylacridine
moieties (BODIPYs 8x and 9x). The absorption and emission profiles, and the extracted data, are
given in Figure 6 and Table 5, respectively. Although only two out of the six dyads are discussed here,
these results can safely be extrapolated to the other systems.
Figure 6: Normalized absorption spectra (top) of 4, 7a, 8x, 9x, 8a, and 9a, and their corresponding normalized
fluorescence emission spectra (bottom) in chloroform solution.
Table 5: Spectroscopic data for 4, 7a, 8x, and 9x as obtained in chloroform solution.[a]
λabs
λem
Δ휈
̅
fwhmabs fwhmem
[h]
Compound
Фf[g] ФΔ
(nm)[b] (nm)[c] (cm-1)[d] (cm-1)[e] (cm-1)[f]
[i]
4
7a
285
509
457
513
13206
153
5209
599
2036
1134
0.01
0.96
-
-
[i]
15
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
8x
9x
634,
346
661,
371
644
671
245
225
1123,
1706
743,
524
600
0.79 0.01
0.50 0.02
2336
[a]
[c]
Spectrograde chloroform was used for all measurements. [b] Absorption maximum. Fluorescence emission
[d]
[e]
maximum/maxima.
Stokes shift between the localized absorption and emission maxima.
Full-width-at-
[f]
[g]
half-maximum of the absorption band. Full-width-at-half-maximum of the emission band(s). Fluorescence
quantum yield determined vs 1,4-bis(5-phenyl-2-oxazolyl)benzene (λexc = 300 nm, Фf = 0.97 in cyclohexane; for
4), rhodamine 6G (λexc = 488 nm, Фf = 0.94 in ethanol; for 7a), cresyl violet (λexc = 580 nm, Фf = 0.56 in ethanol;
[h]
for 8x), or vs nile blue (λexc = 605 nm, Фf = 0.27 in ethanol; for 9x) as a reference. Singlet oxygen quantum
yield determined vs methylene blue (ФΔ = 0.52, λexc = 639 nm) in spectroscopic grade ethanol as a reference by
monitoring the absorbance of 1,3-DPBF at 414 nm. [i] Not determined.
The bathochromic shift observed in dyads 8a and 9a originates from the increased delocalized
system by introducing the slightly donating styryl and thienylvinyl parts, respectively. This can be
seen in BODIPYs 8x and 9x, which have typical BODIPY absorption and emission profiles, and maxima
red-shifted up to 150 nm as compared to the unmodified BODIPY core 7a. Introduction of the
acridine end groups has only minor effects on these values (~10 nm shift), indicating a disconnection
between the acridine and BODIPY parts due to their orthogonality. The additional absorption bands
around 350 and 375 nm in 8x/8a and 9x/9a can be ascribed to the styryl and vinylthienyl units,
respectively. The absorption of the dimethylacridine moiety in 8a and 9a is practically unaltered,
which is line with the twisted conformation. The fluorescence intensity is expected to decrease
compared to BODIPY core 7a, but is still high in 8x and moderate in 9x (Фf = 0.79 and 0.50,
respectively). The ФΔ values remain close to zero in both cases. The reduced fluorescence and low ФΔ
in 9x indicate an increase in non-radiative losses for materials with a thienyl linker. When acridine 4
1
is attached, O2 production becomes viable for dyad 8a. The thirty-fold increase in ФΔ occurs at the
expense of the fluorescence capability, which decreases by 40%. For 9a, phototoxicity remains low
despite the 80% drop in fluorescence emission. Moreover, a second emission band appears. These
data clearly indicate the substantial effects of the acridine moieties in our design. Without it, NIR
absorption and emission is still possible, but singlet oxygen generation is directly linked to the
presence of this additional unit. The absence of the second broad emission band in both 8x and 9x is
remarkable, indicating it originates from the incorporation of the dimethylacridine structure.
To complete our precursor investigation, DFT and TDDFT calculations were also performed on
BODIPYs 8x and 9x (Figure S9 and Table S4). In both cases, the first singlet and triplet, and second
triplet vertical excitation energies are not altered to a great extent as compared to the excitation
energies of 8a and 9a (Table 3), which is consistent with the local nature of these transitions. From
the positioning of the energy levels, a newly formed singlet state with CT character in 8a and 9a
arises between the S1 and S2 states of 8x and 9x when adding the acridine moiety.
Finally, the emission of BODIPYs 8a and 9a was investigated at different concentrations. A stock
solution was made (ca. 1 mg in 5 mL chloroform), from which a dilution series was derived and
emission spectra were recorded (Figure 7). Similar results were obtained in DMSO (Figure S10). Non-
normalized fluorescence spectra can be found in Figure S11 and S12. Although only two out of the six
dyads are discussed here, these results can likely be extrapolated to the other systems.
16
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
Figure 7: Fluorescence emission spectra, normalized to concentration, for a dilution series of 8a (top) and 9a
(bottom) in chloroform (λexc = 605 nm, slit width = 2 nm for 8a; λexc = 645 nm, slit width = 5 nm for 9a). The
stock solution contained ca. 1 mg BODIPY dissolved in 5 mL chloroform.
In our initial investigation of dyad 9a, a second broad fluorescence band was observed in chloroform
(Figure 4). The same profile is seen at different concentrations, although the relative intensity of the
localized emission and the red-shifted band varies. In the most diluted solutions, the fluorescence
1
signal originating from the decay of the LE state is most pronounced. Upon increasing the BODIPY
concentration, the broad low-energy band starts to increase in intensity until the point is reached
where this band surpasses the LE one. Further increase in concentration decreases the fluorescence
intensity of both signals, but the localized emission decreases to a greater extent. This concentration
dependency suggests the formation of excited state complexes or exciplexes[31]. Series 9a-c seems to
be very sensitive to exciplex formation as it is perceived in every solvent in diluted solutions. In the
1H NMR, broad peaks for some of the aromatic protons also confirm the idea of molecular stacking,
despite having a seemingly ‘good’ solubility.
BODIPY 8a exhibits a standard, sharp BODIPY-like emission profile in chloroform (Figure 4). However,
looking at different concentrations, it appears there is a second emission band in these systems,
originating from a populated exciplex state. Upon increasing concentration, the relative fluorescence
intensity of the supposed shoulder increases. At 1/8th of the stock concentration, the red-shifted
band becomes more intense than the LE one. In the most concentrated samples, fluorescence is
quenched, and the localized emission peak becomes insignificant.
Now the question raises what is the influence of these exciplex states in our systems and, more
importantly, in their brightness and phototoxicity? Although the original design was based on a
17
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
charge recombination induced mechanism, the involvement of a CT state was disputed. As direct ISC
from singlet to triplet is unlikely, exciplex formation could be a key element in the ISC abilities.
Exciplex states can live relatively long as there is no ground state counterpart, and they have a strong
polar nature, as seen for CT states.[47] Hence, exciplexes can serve the same purpose as CT states (see
energy level scheme in Figure 1), thereby assisting in triplet formation. In the BODIPY-acridine dyads,
1
3
ISC from a Exciplex state to a LE is questionable. From the TDDFT data (Table 3, Figure S5), it
appears that the second 3LE is even higher in energy than the first LE, and the first LE is too low to
1
3
be reached directly from the Exciplex state. As an alternative, a 3Exciplex state should be present to
1
provide ISC, immediately followed by a fast internal conversion to the lowest 3LE state. In this way, a
RP-ISC like mechanism would be obtained. This is a reconcilable statement. The structural double
bond results in an increased donor-acceptor distance, thereby decreasing their electronic coupling
1
3
and reducing the energy splitting between Excipled and Exciplex states, a feature frequently
ascribed to exciplex systems.[25] Even more, similar ISC behavior into an exciplex triplet state from an
excited state complex has already been observed in some BODIPY dyes, although the examples are
scarce.[31,47,48] To fit in the series of other ISC mechanisms, we depicted this alternative as ‘exciplex
intersystem crossing’ (EX-ISC).
The photophysical observations from Figure 5 and Table S3 can now be rationalized using the EX-ISC
idea. The broad emission observed (8a-c in DMSO, 9a-c in all solvents), was the result of poorly
emissive exciplex emission, which explains the low fluorescence quantum yields. The increased back
electron transfer hinders triplet formation and thus explains the lack of 1O2 formation. In toluene and
chloroform, only a localized emission peak can be seen in dyads 8a-c at low concentrations. Due to
the polar nature of these exciplex states, its energy will increase in more apolar solvents due to its
destabilization. In toluene, the energy is too high to allow efficient charge separation from the first
singlet excited state, resulting in higher Фf values, but low ФΔ. Eventually, in chloroform, it has the
ideal position in between the singlet and triplet energy levels. In this way, charge separation can take
1
3
place in a subtle way to retain localized fluorescence. Furthermore, the Exciplex and Exciplex state
are close enough in energy to enable ISC and prevent charge recombination to the ground state. In
DMSO, exciplex formation is enhanced as the energy of the exciplex state is lowered, leading to
increased charge recombination to the ground state, which is detrimental for the fluorescence and
singlet oxygen quantum yields.
Conclusions
For the first time, we have successfully designed and synthesized near-infrared photoactive BODIPY
photosensitizers with a high and balanced brightness and phototoxic power for potential use as
thera(g)nostic[1] agents in photodynamic therapy. A large bathochromic shift in the absorption and
emission profiles was realized by extending the conjugated system through a fast and simple
Knoevenagel condensation strategy. Dimethylacridine moieties were chosen to provide an
orthogonal donor-acceptor design in the BODIPY dyads. The structural constraints were confirmed
and visualized using density functional theory computational analysis. The acridine moiety was
proven to be essential in this design to achieve enhanced triplet population, necessary for singlet
oxygen production without the use of heavy atoms. A combination of photophysical studies in
solvents of varying polarity at different concentrations and time dependent density functional theory
calculations revealed the involvement of exciplex states, which play a crucial role in so-called exciplex
intersystem crossing (EX-ISC). Despite their activity deeper into the near-infrared, the BODIPY-
acridine dyads connected through a thienyl spacer showed poor fluorescence and singlet oxygen
quantum yields. Fortunately, when using a phenyl spacer, fluorescence quantum yields ranging from
38 to 63% were observed in combination with singlet oxygen quantum yields of 23 to 31% (in
18
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
chloroform solution). Further detailed investigations are scheduled to fully understand the excited
state deactivation mechanisms taking place. Nevertheless, in combination with their high molar
attenuation coefficients, these BODIPY-acridine dyads are very promising for the further
development of heavy-atom-free BODIPY photosensitizers active in the phototherapeutic region,
combining imaging and the therapeutic process itself for detection and eradication of cancerous
cells. Anti-tumor effects and fluorescence optical imaging of our compounds, incorporated in suitable
(tumor-targeting) nanoparticle systems, will be pursued in near future as a next step to validate the
true potential of this system for theragnostic applications.[49]
Supporting information
Detailed PS synthesis procedures, absorption and emission spectra, singlet oxygen generation
1
13
experiments, detailed (TD) DFT data, and H and C NMR spectra can be found in the supporting
information.
Acknowledgements
The authors thank Hasselt University and the University of Namur for continuing financial support
(PhD scholarships JD and TC). MK, BC, and WM thank the Research Foundation – Flanders (FWO) for
support through project G087718N, Hercules project GOH3816NAUHL, and the Scientific Research
Community (WOG) ‘Supramolecular Chemistry and Materials’ (W000620N). The calculations were
performed on the computers of the « Consortium des équipements de Calcul Intensif (CÉCI) »
(http://www.ceci-hpc.be), including those of the ‘UNamur Technological Platform of High-
Performance Computing (PTCI)’ (http://www.ptci.unamur.be), for which we gratefully acknowledge
the financial support from the FNRS-FRFC, the Walloon Region, and the University of Namur
(Conventions No. 2.5020.11, GEQ U.G006.15, U.G018.19, 1610468, and RW/GEQ2016).
References
[1]
[2]
S. Frangos, J. R. Buscombe, Eur. J. Nucl. Med. Mol. Imaging 2019, 46, 519.
a) X. Chen, S. T. C. Wong, Cancer Theranostics, Academic Press, Oxford, 2014; b) S. S. Kelkar,
T. M. Reineke, Bioconjug. Chem. 2011, 22, 1879-1903; c) J. P. Celli, B. Q. Spring, I. Rizvi, C. L.
Evans, K. S. Samkoe, S. Verma, B. W. Pogue, T. Hasan, Chem. Rev. 2010, 110, 2795-2838.
a) X. Li, S. Kolemen, J. Yoon, E. U. Akkaya, Adv. Funct. Mater. 2017, 27; b) B. del Rosal, B. Jia,
D. Jaque, Adv. Funct. Mater. 2018, 28, 1803733; c) C.-N. Ko, G. Li, C.-H. Leung, D.-L. Ma,
Coord. Chem. Rev. 2019, 381, 79-103.
a) T. J. Dougherty, J. Clin. Laser Med. Surg. 2002, 20, 3-7; b) D. E. Dolmans, D. Fukumura, R. K.
Jain, Nat. Rev. Cancer 2003, 3, 380-387; c) H. v. Tappeiner, Ergebnisse der Physiologie 1909,
8, 698-741.
[3]
[4]
[5]
a) C. S. Foote, Photochem. Photobiol. 1991, 54, 659-659; b) S. Kwiatkowski, B. Knap, D.
Przystupski, J. Saczko, E. Kedzierska, K. Knap-Czop, J. Kotlinska, O. Michel, K. Kotowski, J.
Kulbacka, Biomed. Pharmacother. 2018, 106, 1098-1107.
[6]
[7]
A. P. Castano, T. N. Demidova, M. R. Hamblin, Photodiagnosis and Photodynamic Therapy
2004, 1, 279-293.
a) J. F. Lovell, T. W. Liu, J. Chen, G. Zheng, Chem. Rev. 2010, 110, 2839-2857; b) T. J.
Dougherty, C. J. Gomer, B. W. Henderson, G. Jori, D. Kessel, M. Korbelik, J. Moan, Q. Peng, J.
Natl. Cancer Inst. 1998, 90, 889-905; c) N. L. Oleinick, R. L. Morris, I. Belichenko, Photochem.
Photobiol. Sci. 2002, 1, 1-21; d) J. M. Dabrowski, L. G. Arnaut, Photochem. Photobiol. Sci.
2015, 14, 1765-1780.
[8]
B. C. Wilson, M. S. Patterson, Phys. Med. Biol. 2008, 53, R61-109.
19
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
[9]
a) J. Moan, K. Berg, Photochem. Photobiol. 1991, 53, 549-553; b) J. Moan, J. Photochem.
Photobiol. B: Biol. 1990, 6, 343-344.
[10]
[11]
a) W. M. Sharman, C. M. Allen, J. E. van Lier, Drug Discovery Today 1999, 4, 507-517; b) Z.
Huang, Technol. Cancer Res. Treat. 2005, 4, 283-293.
a) M. R. Hamblin, T. Hasan, Photochem. Photobiol. Sci. 2004, 3, 436-450; b) T. Dai, B. B. Fuchs,
J. J. Coleman, R. A. Prates, C. Astrakas, T. G. St Denis, M. S. Ribeiro, E. Mylonakis, M. R.
Hamblin, G. P. Tegos, Front. Microbiol. 2012, 3, 120; c) P. Babilas, S. Schreml, M. Landthaler,
R. M. Szeimies, Photodermatol. Photoimmunol. Photomed. 2010, 26, 118-132; d) M. A.
MacCormack, Semin. Cutan. Med. Surg. 2008, 27, 52-62; e) H. I. Pass, J. Natl. Cancer Inst.
1993, 85, 443-456.
[12]
a) T. J. Dougherty, S. L. Marcus, Eur. J. Cancer 1992, 28, 1734-1742; b) K. Plaetzer, B.
Krammer, J. Berlanda, F. Berr, T. Kiesslich, Lasers Med. Sci. 2009, 24, 259-268.
K. Deng, C. Li, S. Huang, B. Xing, D. Jin, Q. Zeng, Z. Hou, J. Lin, Small 2017, 13, 1702299.
P. Agostinis, K. Berg, K. A. Cengel, T. H. Foster, A. W. Girotti, S. O. Gollnick, S. M. Hahn, M. R.
Hamblin, A. Juzeniene, D. Kessel, M. Korbelik, J. Moan, P. Mroz, D. Nowis, J. Piette, B. C.
Wilson, J. Golab, CA Cancer J. Clin. 2011, 61, 250-281.
[13]
[14]
[15]
[16]
A. Juzeniene, K. P. Nielsen, J. Moan, J. Environ. Pathol. Toxicol. Oncol. 2006, 25, 7-28.
a) R. R. Allison, G. H. Downie, R. Cuenca, X.-H. Hu, C. J. H. Childs, C. H. Sibata, Photodiagnosis
and Photodynamic Therapy 2004, 1, 27-42; b) M. R. Detty, S. L. Gibson, S. J. Wagner, J. Med.
Chem. 2004, 47, 3897-3915; c) V. Kral, J. Davis, A. Andrievsky, J. Kralova, A. Synytsya, P.
Pouckova, J. L. Sessler, J. Med. Chem. 2002, 45, 1073-1078; d) R. D. Teo, J. Y. Hwang, J.
Termini, Z. Gross, H. B. Gray, Chem. Rev. 2017, 117, 2711-2729; e) S. Swavey, M. Tr, in Recent
Advances in the Biology, Therapy and Management of Melanoma, 2013; f) S. J. Wagner,
Transfus. Med. Rev. 2002, 16, 61-66; g) A. E. O'Connor, W. M. Gallagher, A. T. Byrne,
Photochem. Photobiol. 2009, 85, 1053-1074.
[17]
[18]
R. K. Pandey, J. Porphyrins Phthalocyanines 2000, 04, 368-373.
a) S. G. Awuah, Y. You, RSC Adv. 2012, 2, 11169-11183; b) A. Turksoy, D. Yildiz, E. U. Akkaya,
Coord. Chem. Rev. 2019, 379, 47-64; c) L. Yao, S. Xiao, F. Dan, J. Chem. 2013, 2013, 1-10; d) A.
Kamkaew, S. H. Lim, H. B. Lee, L. V. Kiew, L. Y. Chung, K. Burgess, Chem. Soc. Rev. 2013, 42,
77-88; e) C. S. Kue, S. Y. Ng, S. H. Voon, A. Kamkaew, L. Y. Chung, L. V. Kiew, H. B. Lee,
Photochem. Photobiol. Sci. 2018, 17, 1691-1708; f) L. Huang, G. Han, Small Methods 2018, 2,
1700370; g) W. Sun, X. Zhao, J. Fan, J. Du, X. Peng, Small 2019, 15, 1804927; h) M. L. Agazzi,
M. B. Ballatore, A. M. Durantini, E. N. Durantini, A. C. Tomé, J. Photochem. Photobiol. C:
Photochem. Rev. 2019, 40, 21-48; i) D. Chen, Z. Zhong, Q. Ma, J. Shao, W. Huang, X. Dong,
ACS Appl. Mater. Interfaces 2020, 12, 26914-26925; j) P. Chinna Ayya Swamy, G. Sivaraman,
R. N. Priyanka, S. O. Raja, K. Ponnuvel, J. Shanmugpriya, A. Gulyani, Coord. Chem. Rev. 2020,
411, 213233; k) W. Lin, D. Colombani-Garay, L. Huang, C. Duan, G. Han, WIREs Nanomed.
Nanobiotechnol. 2020, 12, e1627.
[19]
[20]
a) A. Loudet, K. Burgess, Chem. Rev. 2007, 107, 4891-4932; b) G. Ulrich, R. Ziessel, A.
Harriman, Angew. Chem. Int. Ed. Engl. 2008, 47, 1184-1201; c) N. Boens, B. Verbelen, W.
Dehaen, Eur. J. Org. Chem. 2015, 6577-6595.
a) A. B. Descalzo, H. J. Xu, Z. Shen, K. Rurack, Ann. N. Y. Acad. Sci. 2008, 1130, 164-171; b) Y.
Ni, J. Wu, Org. Biomol. Chem. 2014, 12, 3774-3791; c) H. Lu, J. Mack, Y. Yang, Z. Shen, Chem.
Soc. Rev. 2014, 43, 4778-4823; d) L. Yuan, W. Lin, K. Zheng, L. He, W. Huang, Chem. Soc. Rev.
2013, 42, 622-661.
[21]
[22]
a) J. Zhao, K. Xu, W. Yang, Z. Wang, F. Zhong, Chem. Soc. Rev. 2015, 44, 8904-8939; b) K.
Chen, Y. Dong, X. Zhao, M. Imran, G. Tang, J. Zhao, Q. Liu, Front. Chem. 2019, 7, 821; c) J.
Wang, Q. Gong, L. Wang, E. Hao, L. Jiao, J. Porph. Phthalocyanines 2020, 24, 603-635.
a) J. Killoran, L. Allen, J. F. Gallagher, W. M. Gallagher, D. F. O'Shea, Chem. Commun. 2002,
1862-1863; b) A. Gorman, J. Killoran, C. O'Shea, T. Kenna, W. M. Gallagher, D. F. O'Shea, J.
Am. Chem. Soc. 2004, 126, 10619-10631; c) T. Yogo, Y. Urano, Y. Ishitsuka, F. Maniwa, T.
Nagano, J. Am. Chem. Soc. 2005, 127, 12162-12163.
20
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
[23]
[24]
a) E. l. G. Azenha, A. C. Serra, M. Pineiro, M. M. Pereira, J. Seixas de Melo, L. G. Arnaut, S. J.
Formosinho, A. M. d. A. Rocha Gonsalves, Chem. Phys. 2002, 280, 177-190; b) J. Zou, Z. Yin, K.
Ding, Q. Tang, J. Li, W. Si, J. Shao, Q. Zhang, W. Huang, X. Dong, ACS Appl. Mater. Interfaces
2017, 9, 32475-32481.
a) J. Zhao, W. Wu, J. Sun, S. Guo, Chem. Soc. Rev? 2013, 42, 5323-5351; b) J. Zhao, K. Chen, Y.
Hou, Y. Che, L. Liu, D. Jia, Org. Biomol. Chem. 2018, 16, 3692-3701; c) Y. Hou, X. Zhang, K.
Chen, D. Liu, Z. Wang, Q. Liu, J. Zhao, A. Barbon, J. Mater. Chem. C 2019, 7, 12048-12074.
M. A. Filatov, Org. Biomol. Chem. 2020, 18, 10-27.
L. Huang, X. Yu, W. Wu, J. Zhao, Org. Lett. 2012, 14, 2594-2597.
a) R. L. Watley, S. G. Awuah, M. Bio, R. Cantu, H. B. Gobeze, V. N. Nesterov, S. K. Das, F.
D'Souza, Y. You, Chem. Asian J. 2015, 10, 1335-1343; b) S. Ji, J. Ge, D. Escudero, Z. Wang, J.
Zhao, D. Jacquemin, J. Org. Chem. 2015, 80, 5958-5963.
[25]
[26]
[27]
[28]
[29]
D. Escudero, Acc. Chem. Res. 2016, 49, 1816-1824.
a) Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 2003, 103, 3899-4032; b) J.
Verhoeven, J. Photochem. Photobiol. C: Photochem. Rev. 2006, 7, 40-60.
N. Epelde-Elezcano, E. Palao, H. Manzano, A. Prieto-Castaneda, A. R. Agarrabeitia, A. Tabero,
A. Villanueva, S. de la Moya, I. Lopez-Arbeloa, V. Martinez-Martinez, M. J. Ortiz, Chem. – Eur.
J. 2017, 23, 4837-4848.
[30]
[31]
[32]
[33]
As the exact structures of these ‘exciplexes’ (i.e. excited state complexes) remain unknown
for now, we prefer to use this more general term.
N. Mataga, H. Chosrowjan, S. Taniguchi, J. Photochem. Photobiol. C: Photochem. Rev. 2005, 6,
37-79.
a) K. Rurack, M. Kollmannsberger, J. Daub, Angew. Chem. Int. Ed. 2001, 40, 385-387; b) O.
Buyukcakir, O. A. Bozdemir, S. Kolemen, S. Erbas, E. U. Akkaya, Org. Lett. 2009, 11, 4644-
4647.
[34]
[35]
a) N. A. Kukhta, H. F. Higginbotham, T. Matulaitis, A. Danos, A. N. Bismillah, N. Haase, M. K.
Etherington, D. S. Yufit, P. R. McGonigal, J. V. Gražulevičius, A. P. Monkman, J. Mater. Chem.
C 2019, 7, 9184-9194; b) M. Y. Wong, E. Zysman-Colman, Adv. Mater. 2017, 29, 1605444.
a) N. Ziebart, P. Schroeer, K. Rueck-Braun, Tetrahedron 2018, 74, 5561-5566; b) Y. Rong, C.
Wu, J. Yu, X. Zhang, F. Ye, M. Zeigler, M. E. Gallina, I. C. Wu, Y. Zhang, Y. H. Chan, W. Sun, K.
Uvdal, D. T. Chiu, ACS Nano 2013, 7, 376-384; c) Y. Zhou, Z. Zhou, Y. Li, W. Yang, Catal.
Commun. 2015, 64, 96-100.
[36]
[37]
[38]
a) Y. Zhao, D. G. Truhlar, Theor. Chem. Acc. 2008, 120, 215-241; b) J. Tomasi, B. Mennucci, R.
Cammi, Chem. Rev. 2005, 105, 2999-3093.
a) R. W. Redmond, J. N. Gamlin, Photochem. Photobiol. 1999, 70, 391-475; b) F. Wilkinson,
W. P. Helman, A. B. Ross, J. Phys. Chem. Ref. Data 1993, 22, 113-262.
H. Heitele, P. Finckh, S. Weeren, F. Poellinger, M. E. Michel-Beyerle, J. Phys. Chem. 1989, 93,
5173-5179.
[39]
[40]
[41]
C. Reichardt, Chem. Rev. 1994, 94, 2319-2358.
M. R. Momeni, A. Brown, J. Chem. Theory Comput. 2015, 11, 2619-2632.
S. Chibani, B. Le Guennic, A. Charaf-Eddin, A. D. Laurent, D. Jacquemin, Chem. Sci. 2013, 4,
1950-1963.
[42]
a) D. Jacquemin, E. A. Perpete, I. Ciofini, C. Adamo, J. Chem. Theory Comput. 2010, 6, 1532-
1537; b) C. Brückner, B. Engels, Chem. Phys. 2017, 482, 319-338.
M. Kasha, Discuss. Faraday Soc. 1950, 9, 14-19.
T. Le Bahers, C. Adamo, I. Ciofini, J. Chem. Theory Comput. 2011, 7, 2498-2506.
D. Rehm, A. Weller, Israel J. Chem. 1970, 8, 259-271.
H. Kang, Y. Si, Y. Liu, X. Zhang, W. Zhang, Y. Zhao, B. Yang, Y. Liu, Z. Liu, J. Phys. Chem. A 2018,
122, 5574-5579.
[43]
[44]
[45]
[46]
[47]
A. Nano, R. Ziessel, P. Stachelek, M. A. Alamiry, A. Harriman, Chemphyschem 2014, 15, 177-
186.
21
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
[48]
[49]
A. C. Benniston, G. Copley, H. Lemmetyinen, N. V. Tkachenko, Chemphyschem 2010, 11,
1685-1692.
R. Penjweini, S. Deville, L. D'Olieslaeger, M. Berden, M. Ameloot, A. Ethirajan, J. Control.
Release 2015, 218, 82-93.
22
This article is protected by copyright. All rights reserved.

10.1002/chem.202002549
Chemistry - A European Journal
Table of Contents Graphic
We present a series of compact, heavy-atom-free, BODIPY-acridine photosensitizers, activated by
near-infrared light and showing balanced brightness and phototoxic power. The BODIPY
chromophores are subjected to extensive photophysical characterization and quantum-chemical
calculations, suggesting the involvement of exciplex states in singlet oxygen formation.
Keywords
BODIPY dyes • photosensitizers • near-infrared • singlet oxygen • fluorescence
23
This article is protected by copyright. All rights reserved.