Bulky Barbiturates as Non-Toxic Ionic Dye Insulators for Enhanced Emission in Polymeric Nanoparticles

Article abstract of DOI:10.1002/chem.202101986

Bulky hydrophobic counterions (weakly coordinating anions) can insulate ionic dyes against aggregation-caused quenching (ACQ) and enable preparation of highly fluorescent dye-loaded nanoparticles (NPs) for bioimaging, biosensing and light harvesting. Here, we introduce a family of hydrophobic anions based on fluorinated C-acyl barbiturates with delocalized negative charge and bulky non-polar groups. Similarly to fluorinated tetraphenylborates, these barbiturates prevent ACQ of cationic dye alkyl rhodamine B inside polymer NPs made of biodegradable poly(lactic-co-glycolic acid) (PLGA). Their efficiency to prevent ACQ increases for analogues with higher acidity and bulkiness. Their structure controls dye-dye communication, yielding bright NPs with on/off switching or stable emission. They enhance dye encapsulation inside NPs, allowing intracellular imaging without dye leakage. Compared to fluorinated tetraphenylborates known as cytotoxic transmembrane ion transporters, the barbiturates display a significantly lower cytotoxicity. These chemically available and versatile barbiturate derivatives are promising counterion scaffolds for preparation of bright non-toxic fluorescent nanomaterials.

Full text of DOI:10.1002/chem.202101986

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
Bulky Barbiturates as Non-Toxic Ionic Dye Insulators for
Enhanced Emission in Polymeric Nanoparticles
[a]
[a]
[a]
[a]
Bohdan Andreiuk, Ilya O. Aparin, Andreas Reisch, and Andrey S. Klymchenko*
Abstract: Bulky hydrophobic counterions (weakly coordinat-
ing anions) can insulate ionic dyes against aggregation-
caused quenching (ACQ) and enable preparation of highly
fluorescent dye-loaded nanoparticles (NPs) for bioimaging,
biosensing and light harvesting. Here, we introduce a family
of hydrophobic anions based on fluorinated C-acyl barbitu-
rates with delocalized negative charge and bulky non-polar
groups. Similarly to fluorinated tetraphenylborates, these
barbiturates prevent ACQ of cationic dye alkyl rhodamine B
inside polymer NPs made of biodegradable poly(lactic-co-
glycolic acid) (PLGA). Their efficiency to prevent ACQ
increases for analogues with higher acidity and bulkiness.
Their structure controls dye-dye communication, yielding
bright NPs with on/off switching or stable emission. They
enhance dye encapsulation inside NPs, allowing intracellular
imaging without dye leakage. Compared to fluorinated
tetraphenylborates known as cytotoxic transmembrane ion
transporters, the barbiturates display a significantly lower
cytotoxicity. These chemically available and versatile barbitu-
rate derivatives are promising counterion scaffolds for
preparation of bright non-toxic fluorescent nanomaterials.
Introduction
of optical properties of NPs, such as absorption and emission
color, brightness, photostability, etc.
[8]
[9]
[10]
[5a]
Fluorescent nanoparticles (NPs) gained importance in recent
years due to their unique optical properties and numerous
potential applications, including biological imaging and
However, preparation of bright dye-loaded NPs requires
addressing two key challenges. First, fluorescent dyes should be
efficiently encapsulated into the polymer matrix without dye
[1]
[1b,2]
[11]
biosensing as well as light harvesting.
Organic fluorescent
leakage (leaching) in biological media. Second, at high dye
nanoparticles are of particular interest as the highly versatile
nature of organic materials and fluorescent dyes enable
preparation of materials of programmed optical properties, easy
functionalization with targeting ligands, antibodies, DNA,
stealth groups, etc. On the other hand, bio-applications of
organic nanoparticles impose strict requirements such as
biocompatibility, biodegradability, low toxicity and eco-friend-
liness, which remains a challenge even for organic nano-
loading aggregation-caused quenching (ACQ) is commonly
[
5a]
observed, which complicates preparation of bright NPs. Dye
self-quenching is generally caused by the flat aromatic structure
of fluorophores, which favors pi-stacking of dyes into so-called
[12]
H-aggregates with face-to-face assembly.
To address the
problem of ACQ,
proposed, which include aggregation-induced emission,
a
number of approaches have been
[5a]
[3a,13]
[10,14]
the use of bulky side groups,
and bulky hydrophobic
[3]
[15]
materials. Important examples of organic NPs include AIE NPs,
counterions. Our previous works showed that bulky hydro-
phobic counterions are of particular interest for dye-loaded
polymeric NPs, as they can address both the problems of ACQ
[4]
[3b,5]
conjugated polymer NPs, dye-loaded polymeric NPs
and
[6]
lipid NPs. Dye-loaded polymeric NPs are particles of choice in
view of bio-applications, because they present several unique
features. On the one hand, by selecting appropriate polymer
matrix, NPs with desired biocompatibility, biodegradability and
stability can be obtained. For example, poly(lactic-co-glycolic
acid) (PLGA), approved for human use by the United States
Food and Drug Administration and by the European Medicine
Agency, can be used to prepare biodegradable NPs, which are
[
8,11b,15a]
and dye leakage from NPs for different cationic dyes.
These counterions are generally weakly coordinating anions,
[16]
most commonly tetraphenylborate (TPB) derivatives. They are
characterized by large size, high hydrophobicity and strong
delocalization of the negative charge over the large volume of
the anion. Weakly coordinating anions are often constituents of
ionic liquids and they are commonly used for stabilization of
[7]
[17]
generally known in the field of drug delivery. On the other
hand, selection of the dye and its loading level allow fine tuning
reactive cations and for preparation of ion-selective electro-
[
16,18]
des and electric batteries.
Earlier works showed that they
[19]
can minimize self-quenching in pure dye salts, and the most
efficient were those characterized by large size and high level
of fluorination. As we showed later on, these counterions can
form highly hydrophobic ion pairs with a cationic dye, which
ensures its nearly quantitative encapsulation inside the hydro-
phobic polymeric matrix even at high loading during nano-
[a] Dr. B. Andreiuk, Dr. I. O. Aparin, Dr. A. Reisch, Dr. A. S. Klymchenko
[20]
Laboratoire de Bioimagerie et Pathologies
UMR 7021 CNRS, ITI Chimie des Systèmes Complexes
Université de Strasbourg
74 route du Rhin, 67401, Illkirch (France)
E-mail: andrey.klymchenko@unistra.fr
[11b,21]
particle synthesis by nanoprecipitation.
Moreover, these
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/chem.202101986
bulky anions serve as spacers (or insulators) between encapsu-
Chem. Eur. J. 2021, 27, 1–8
1
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
lated cationic dyes, thus preventing their self-quenching
hydrophobic, similarly to weakly coordinating anions based on
borates and aluminates, but without their limitations. Classical
examples of hydrophobic anions are oxonol dyes (Figure 1A), in
which a single negative charge is delocalized over a long
conjugation system. However, oxonol dyes are colored, which
would alter or quench the emission of a cationic dye in the
counterion-based approach. Therefore, an oxonol analogue
with a much shorter conjugation should be developed.
[11b,15a, 22]
through pi-stacking into H-aggregates.
Therefore,
despite very high loading of cationic dyes with these bulky
hydrophobic anions (up to 50 wt% vs. polymer), the obtained
NPs exhibited unprecedentedly high fluorescence quantum
[30]
[23]
yields (~50%). In this way it was possible to obtain NPs 6–
1
00 fold brighter than quantum dots of similar size and
[9,15a,24]
emission color.
Owing to the short dye-dye distances
inside these highly loaded NPs, the loaded dyes exhibited
unique collective behavior due to coupling by ultrafast
In the present work, we synthesized a family of organic
anions based on a barbiturate scaffold and studied their
capacity to prevent ACQ in dye-loaded polymeric NPs. Their
weakly acidic methylene group was C-acylated in order to
enable formation of the anion with delocalized negative charge
after deprotonation (Figure 1A,B), similarly to oxonols, but
without absorption of light in the visible region. Our results
showed that acylated barbiturates with perfluorobenzoyl group
can form stable ion pairs with a cationic rhodamine dye and
ensure its efficient encapsulation and excellent optical proper-
ties inside polymer NPs. This work constitutes the first
demonstration of an anion with π-delocalized negative charge
capable to replace conventional weakly coordinating anions in
preparation of dye-loaded polymer NPs. Importantly, the
obtained anions are much less toxic than fluorinated tetraphe-
nylborate, which enables preparation of much safer nano-
materials for bioimaging applications and increases their
potential for clinical translation. Finally, a relatively simple
chemistry of these anions opens the way to their further
functionalization and scale up.
[15a]
excitation energy transfer,
which enabled preparation of a
giant light-harvesting nanoantenna for single-molecule detec-
[25]
tion at ambient light.
Moreover, the concept of bulky
counterions works with both rhodamine and cyanine dyes,
which enabled preparation of differently colored ultrabright
NPs and development of a cell barcoding system for in vitro
[8]
and in vivo applications. Such NPs have also been used for the
assembly of biosensors for detection of nucleic acids with
[
23–24]
single-molecule sensitivity
and live tracking inside mouse
[26]
brain. However, so far, the counterion approach was mainly
limited to fluorinated tetraphenylborate derivatives
an aluminate of perfluorinated tert-butanol. These ions can
transport cations, including protons, through biological mem-
branes, such as cell plasma membranes and mitochondria,
[8,11b,15a]
and
[21]
[27]
which leads to cytotoxicity at micromolar concentrations.
Moreover, bulkier and more fluorinated anions were found to
be more cytotoxic because of their stronger capacity to
[27]
transport cations.
encapsulating these counterions were not cytotoxic,
tions can be raised about potential toxicity of these anions after
Even though dye-loaded polymeric NPs
[8,21]
ques-
biodegradation of the polymer matrix. Moreover, due to the Results and Discussion
symmetric nature of these anions, it is synthetically challenging
[
28]
to mono-functionalize them. Therefore, there is a need to
look for alternative scaffolds for a hydrophobic ion. One
approach, called small-molecule ionic isolation lattices (SMILES)
proposed to generate bulky hydrophobic anions by complexing
small inorganic anions with a large macrocycle (cyanostar),
which allows to significantly decrease ACQ of cationic dyes in
Primarily, we synthesized barbiturates with (R) groups being
cyclohexyl and R’ being undecyl (1) or phenyl (2) (Figure 1B,
Scheme 1). These molecules were readily obtained from the
barbituric acid derivative and the corresponding carboxylic
acids using EDC as coupling agent in the presence of DMAP.
The obtained C-acyl barbiturates were mixed with octadecylr-
hodamine B (R18) perchlorate in DCM with excess of triethyl-
amine for deprotonation of barbiturates and ion exchange.
However, no stable enough ion pairs of compounds 1 and 2
[
15b]
[29]
bulk polymeric materials
and nanoparticles.
Another
attractive approach is the design of an organic ion with high
charge delocalization (“soft” ion), which would be also highly
Figure 1. (A) Chemical structures of the oxonol dye scaffold and C-acyl barbiturates scaffolds with their resonance structures. (B–D) Schematic presentation of
a fluorescent nanoparticle (D) encapsulated with ion pairs of a cationic dye R18 (C) and bulky hydrophobic counterions (B).
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
2
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
(Figure 1, Supporting Information), similarly to the ion pair R18/
F5-TPB. We suppose that lower pKa values (estimated value
3
.68) help to generate more stable ion pairs with the R18 dye.
Further, we replaced the cyclohexyl (R) groups with
diphenylmethyl to obtain the even bulkier perfluorophenyl
analogue BarPh2 (Figure 1B, Scheme 1). The anionic form of
BarPh2 also appeared to form stable ion pairs with R18
(Figure 1, Supporting Information), in line with similarly low
calculated pKa values (3.98). Moreover, the BarPh2 salt migrated
faster than that of BarC6, indicating a higher hydrophobicity of
the former. Moreover, in contrast to the salt of BarC6, the
BarPh2 salt did not show any trace of the R18 at the start of the
TLC, indicating that it is more stable on silica.
Both ion pairs of novel barbiturate anions with the R18 dye,
namely R18/BarC6 and R18/BarPh2, alongside with the already
known R18/F5-TPB and R18/ClO as reference compounds, were
4
encapsulated inside biodegradable PLGA (poly(lactide-co-glyco-
lide)) nanoparticles (Figure 1) using a simple method based on
nanoprecipitation
from
the
water-miscible
solvent
[15a,31]
acetonitrile.
The dyes were loaded at 5, 20 and 50 mM
concentration with respect to polymer.
The size of the obtained nanoparticles was measured by
TEM (Figure 2) and DLS (Figure 3A). For both new barbiturates
(
BarC6 and BarPh2) at 50 mM dye loading, TEM images revealed
spherical shape of the obtained NPs, similarly to F5-TPB
Figure 2). The measured size values were similar for all three
Scheme 1. Synthesis of target barbiturates (in blue) bearing N-cyclohexyl (A)
and N-diphenylmethyl (B) groups: (a) malonic acid, THF; (b) dodecanoic acid,
EDC×HCl, DMAP, DCM; (c) benzoic acid, EDC×HCl, DMAP, DCM; (d)
perfluorobenzoic acid, EDC×HCl, DMAP, DCM; (e) CDI, THF (f) malonyl
chloride, DCM; (g) perfluorobenzoic acid, EDC×HCl, DMAP, DCM.
(
types of NPs in the range of 33–36 nm. By contrast, PLGA NPs
loaded at 50 mM with R18 and small hydrophilic perchlorate
counterion (R18/ClO4) yielded large aggregates according to
[21]
our earlier TEM data. DLS confirmed small size (40–50 nm) of
the formulated PLGA NPs, which was independent of dye
loading (Figure 3A). By contrast, ion pair R18/ClO4 with the
hydrophilic perchlorate counterion induced formation of much
larger (>500 nm) and colloidally unstable nanoparticles at
50 mM dye loading. The latter observation, in line with our
with cationic dye R18 were formed to be purified by silica
chromatography (Figure 1, Supporting Information). This result
could be explained by insufficient electron-withdrawing effect
of the used acyl groups, which made possible re-protonation of
the formed enolate anion on silica, in line with relatively high
calculated pKa of 6.51 and 6.04 for compounds 1 and 2,
respectively. Therefore, to increase the acidity of the barbitu-
rates, the R’ group was changed to perfluorophenyl in molecule
BarC6 (Figure 1B), which was obtained following the same
methodology (Scheme 1). The anion of the obtained molecule
could form ion pairs with the R18 cation that could be detected
on silica thin layer chromatography as a faster moving spot
[11b,15a]
previous reports,
can be explained by cationic dye
adsorption at the anionic nanoparticle surface, which leads to
decrease of surface charge and further NP aggregation. Zeta
potential measurements showed strong negative values for
PLGA NPs with all three studied bulky counterions at 50 mM
dye loading (Figure 2, Supporting Information). In contrast, the
absolute value of the zeta potential was significantly lower in
Figure 2. TEM images of NPs loaded with R18 dye and bulky counterions: F5-TPB (A), BarC6 (B) and BarPh2 (C). Graphs on the right of the TEM images
correspond to size distribution histograms. The provided diameters are mean values of NPs based on analysis of >200 particles per condition. The error is the
half-width at the half-maximum.
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
3
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
Figure 3. Properties of the developed NPs, loaded with ion pairs of R18 dye with different counterions. (A) Hydrodynamic diameter according to DLS, (B)
fluorescence quantum yield, (C) fluorescence anisotropy and (D) encapsulation efficiency at 50 mM dye loading. (E) Absorption and fluorescence spectra of
NPs at 50 mM dye loading and solution of R18/ClO in methanol. Excitation wavelength was 520 nm.
4
case of R18/ClO4, confirming neutralization of the negative
surface charge of NPs by cationic R18.
According to our results, already at 5 mM dye loading the
fluorescence anisotropy values were very low (>0.01) for the
[15a]
To evaluate quantitatively the dye encapsulation efficiency
in NPs, we used our previously developed protocol – dialysis for
F5-TPB counterion, in line with earlier data.
In case of the
BarPh2 counterion, the anisotropy was also relatively low at this
loading (Figure 3C). It further decreased with dye loading
reaching nearly zero at 50 mM for both counterions, suggesting
fast excitation energy transfer in these NPs. In sharp contrast,
for the BarC6 counterion, the anisotropy values were relatively
high at all loadings, similarly to those for the perchlorate
counterion, implying that dye-dye communication inside NPs
was weak. Thus, among the two barbiturates only the bulkier
and more hydrophobic one (BarPh2) favors dye-dye communi-
cation, similarly to F5-TPB.
[11b,21]
2
4 h of NPs solutions versus 1 mM beta-cyclodextrin.
During this procedure all the non-encapsulated dyes and those
adsorbed on the particle surface are washed off. Then, the
comparison of dye concentration before and after dialysis gives
us the encapsulation efficiency. The obtained results (Figure 3D)
showed poor encapsulation efficiency of R18/ClO4 (51�2%)
and almost quantitative encapsulation of R18/F5-TPB (92�4%)
at 50 mM (corresponding to ~7.5%w/w) loading. For the
barbiturate BarC6 anion, the dye encapsulation efficiency was
7
4�1%, whereas for the bulkier and more hydrophobic BarPh2
the highest dye encapsulation efficiency was achieved (96�
%). This efficient encapsulation of R18/BarPh2 can be
The structure of the counterion also affected spectroscopic
properties of the encapsulated R18 dye. Although at 5 mM
loading the effects in the absorption spectra were minor, at
50 mM dye loading they became much more pronounced
(Figure 3E). In line with previous studies, NPs loaded with R18/
F5-TPB showed a blue-shifted absorption maximum compared
1
explained by higher hydrophobicity of BarPh2 compared to its
cyclohexyl-substituted rival BarC6.
At 5 mM dye loading, independently of the counterion, the
quantum yields (QY) of all NPs were high (>70%, Figure 3B).
However, upon increase of loading to 50 mM, perchlorate
demonstrated its inability to prevent dye self-quenching (QY=
to R18/ClO -loaded NPs. The absorption maxima of R18 with
4
barbiturate counterions were intermediate between R18/ClO4
and R18/F5-TPB, though the BarPh2 was slightly red shifted
compared to BarC6. Similar tendencies were observed in the
fluorescence spectra. The red shifted absorption and emission
of the BarPh2 is unexpected, given its similar performance with
F5-TPB. One could speculate that, unlike the case of F5-TPB,
where negative charge is mostly localized at weakly polarizable
fluorine atoms, BarPh2, due to its higher electronic polarizability
decreases the energy of the excited state of R18.
Single-particle brightness was studied by wide-field micro-
scopy. NPs loaded at 50 mM with R18/BarPh2, R18/BarC6 and
R18/F5-TPB as well as commercial quantum dots of similar
emission range, QD585, were immobilized on glass and excited
by a LED source at 550 nm (Figure 4AÀ D). All of the tested dye-
loaded NPs appeared much brighter than quantum dots
QD585: R18/F5-TPB, R18/BarC6 and R18/BarPh2-loaded NPs
4
5
%), in contrast to R18/F5-TPB loaded NPs, displaying a QY of
5% at this loading. Remarkably, for the bulky barbiturate-
based ion pair R18/BarPh2, the QY value was almost as high
52%), suggesting that BarPh2 is as effective in preventing ACQ
(
as the reference counterion F5-TPB. The ion pair R18/BarC6
displayed a QY of about 20% at 50 mM loading, which was 5-
fold higher compared to perchlorate, but still not as high as for
F5-TPB. These results show that quantum yield and encapsula-
tion efficiency of dye-loaded NPs correlate directly with
bulkiness and hydrophobicity of the barbiturates, respectively,
in line with our previous data.
Fluorescence anisotropy measurement is one of the ways to
study dye-dye communication inside NPs, as an energy transfer
[32]
[33]
from dye to dye leads to loss of fluorescence anisotropy.
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
4
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
Figure 4. Single-particle fluorescence properties of the tested dye-loaded PLGA NPs. (A–D) Wide-field fluorescence images of QD585 (A), R18/F5-TPB (B), R18/ 2
BarC6 (C) and R18/BarPh2 (D) NPs at 50 mM dye loading. Identical imaging conditions were applied for dye-loaded NPs (excitation power density of 4.4 W/cm
at 550 nm), ten times higher illumination power was used for QDs. Image size was 15×15 μm. (E–G) Typical single-particle fluorescence traces of NPs loaded
with R18/F5-TPB (E), R18/BarC6 (F) and R18/BarPh2 (G), recorded at the same conditions.
were, respectively, 33, 29 and 28-fold brighter. Surprisingly, the
brightness of R18/BarC6 NPs (Figure 4C) was close to those of
R18/F5-TPB and R18/BarPh2 NPs (Figure 4B,D), despite having
more than two-fold lower quantum yield and lower encapsula-
tion efficiency. This peculiar behavior of these NPs can be
explained based on single-particle fluorescence intensity traces
high loading originates from rather homogeneous distribution
of fluorophores within the polymer matrix of NPs, whereas
bulkier and more hydrophobic R18/BarPh2 tends to cluster. In
our previous work, we observed similar effects in the
fluorescence anisotropy and particle blinking when bulky
fluorinated TPBs were compared to the much smaller anion
[11b]
(
Figure 4E–G). R18/BarPh2 NPs (Figure 4G) underwent ON/OFF
fluorescence switching (blinking), similarly to R18/F5-TPB NPs
Figure 4E). Statistical analysis based on the mean of three
B(CF ) .
Together with the present data, we have now a solid
3
4
background to conclude that size and hydrophobicity of the
counterion is a tool to control dye clustering in the polymer
matrix of NPs. We expect that larger and more hydrophobic
counterions favor dye clustering, because their salts with R18
are significantly more hydrophobic than the PLGA polymer, so
that during nanoprecipitation they precipitate before the
polymer, forming the core with clustered dye. In contrast, R18
salts with counterions of intermediate hydrophobicity (like
BarC6 and B(CF ) ) would co-precipitate together with the
(
measurements (>400 particles of each type) suggested that
blinking took place for 95+/À 0.5% and 78+/À 2% of all
recorded R18/F5-TPB and R18/BarPh2 NPs, respectively. Accord-
ing to earlier studies of R18/F5-TPB NPs, such behavior is only
possible when the dyes are in close proximity inside the
nanoparticle, making ultrafast energy transfer possible and in
consequence effective quenching of the whole particle by
transient dark species (such as triplet/radical states of
3
4
polymer, thus favoring more homogeneous dye distribution in
the polymer with longer dye-dye distances, leading to higher
fluorescence anisotropy and stable emission without blinking.
To test the obtained NPs in biological media, they were
incubated with KB cells for 3 h and then the cells were washed
and imaged by confocal microscopy. All four counterions at the
highest (50 mM) and lowest (5 mM) dye loading were tested
(Figure 5). R18/F5-TPB NPs at 5 mM loading (Figure 5B) ap-
peared as bright dots, being localized in the perinuclear
regions, and without clear background signal from other parts
of the cells. This result is in line with previous data, confirming
that after endocytosis of R18/F5-TPB NPs, no significant dye
leakage is observed. R18/BarPh2 and R18/BarC6 NPs at this
loading showed similar behavior (Figure 5D and 4 C, respec-
[9,15a]
rhodamine).
Thus, similarly to F5-TPB, the BarPh2 anion may
favor clustering of R18 dyes inside the polymer matrix, which is
in agreement with the observed low fluorescence anisotropy
values (Figure 3C). By contrast, the majority (76�6%) of R18/
BarC6 NPs particles showed constant emission without blinking
(Figure 4F). This very different behavior is probably related to
inefficiency of dye-dye energy transfer, as suggested by
relatively high values of the fluorescence anisotropy. Thus,
higher than expected relative brightness of R18/BarC6 NPs
could be explained by poor dye-dye communication, which
prevents quenching of the encapsulated dyes by a single dark
state or by singlet-singlet annihilation. We can speculate that
poor dye-dye communication for R18/BarC6 even at relatively
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
5
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
Figure 5. Confocal images of KB cells incubated for 3 h with dye-loaded PLGA NPs. NPs were loaded with R18/ClO4, R18/F5-TPB, R18/BarC6 and R18/BarPh2 at
and 50 mM concentrations. Excitation wavelength was 561 nm, emission was collected from 570 to 650 nm. Image size was 58.4×58.4 μm.
5
tively). A significant degree of leakage was observed for R18/
ClO NPs at 5 mM loading (Figure 5A) as a diffuse fluorescence
4
inside the cells, so that the characteristic perinuclear dots
(endosomes/lysosomes) were practically not visible. The same
tendency was observed at 50 mM loading, where R18/ClO NPs
4
showed strong fluorescence all over the cells, while for NPs
with other ion pairs, a clear dotted fluorescence was still
observed (Figure 5E–H). Nevertheless, for both R18/F5-TPB and
R18/BarC6 NPs at 50 mM loading some diffuse intracellular
fluorescence started to be visible (Figures 5F and 5G, respec-
tively), indicating imperfect encapsulation of these ion pairs at
the highest loading. Remarkably, for 50 mM R18/BarPh2 NPs,
diffuse fluorescence was negligible (Figure 5H), suggesting
absence of dye leakage in biological media. The observed
cellular data correlated well with the encapsulation studies,
showing the highest encapsulation efficiency for R18/BarPh2
and the lowest for R18/ClO4 (Figure 3D). Thus, the novel
counterion BarPh2 outperforms our “golden standard” F5-TPB
for bioimaging purposes. In this respect, it is close to the
recently reported F9-Al, a strongly fluorinated aluminum-based
counterion, showing superior encapsulation of the R18 dye.
Finally, we compared the cytotoxicity of the counterions vs.
our standard F5-TPB. To this end, we prepared salts of BarC6
and BarPh2 with 1.2 eq. of triethylamine and studied them
together with the F5-TPB lithium salt in HeLa cells using
standard MTT assay. After 24 h incubation, the F5-TPB counter-
ion was found to be the most cytotoxic, showing <50% of cell
viability at 3 μM concentration (Figure 6). In sharp contrast,
both barbiturates demonstrated virtually no cytotoxicity at this
concentration. Moreover, BarPh2 showed low cytotoxicity with
cell viability >50% for the studied concentration range up to
Figure 6. Viability of HeLa cells exposed for 24 h to different concentrations
of the three counterion salts: F5-TPB lithium salt, BarC6 and BarPh2 salts
with triethylamine (1.2 eq.). Cell viability was tested in quadruplicate and
repeated at least two times using MTT. The error bars correspond to the
standard deviation.
connected to the lower capacity of barbiturates to transport
protons and other cations through biological membranes. We
can speculate that barbiturates due to their high hydro-
phobicity and relatively weak acidity (3.68 and 3.98) are bound
to lipid membranes in molecular form which prevents them
from deprotonation. The fact that the more hydrophobic
BarPh2 is much less toxic than BarC6 supports this hypothesis
because the former should have even stronger partitioning into
lipids, further decreasing the probability of its deprotonation.
By contrast, F5-TPB can preserve the anionic form when bound
to the lipid membrane and therefore can operate as an efficient
shuttle of protons and other cations.
[21]
3
0 μM. BarC6, having lower hydrophobicity and slightly lower Conclusions
pKa than BarPh2, displayed intermediate cytotoxicity between
F5-TPB and BarPh2, with cell viability <50% at 10 μM
concentration. Thus, both barbiturates, especially BarPh2, are
less cytotoxic than our standard F5-TPB, which is probably
In conclusion, we describe a new family of hydrophobic anions
based on acylated barbiturates with strongly delocalized
negative charge, which, similarly to weakly coordinating anions
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
6
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

Full Paper
doi.org/10.1002/chem.202101986
Chemistry—A European Journal
based on boron or aluminum, can enhance encapsulation and
emission of cationic dyes inside polymer NPs. Among the
studied acyl barbiturates, the bulkiest and the most hydro-
phobic anion BarPh2 prevents self-quenching of a cationic
rhodamine B derivative as efficiently as the “golden standard”
tetrakis(pentafluorophenyl)borate and at the same time enables
more efficient dye encapsulation with minimal leakage in living
cells. Moreover, the chemical structure (hydrophobicity) of the
barbiturate counterion can control the dye-dye communication,
as evidenced by fluorescence anisotropy and, thus, tune the
emission mode of nanoparticles from complete on/off switching
[4] a) L. Feng, C. Zhu, H. Yuan, L. Liu, F. Lv, S. Wang, Chem. Soc. Rev. 2013,
2, 6620; b) C. Wu, D. T. Chiu, Angew. Chem. Int. Ed. 2013, 52, 3086.
4
[
5] a) A. Reisch, A. S. Klymchenko, Small 2016, 12, 1968; b) S. Y. Ong, C.
Zhang, D. Xiao, S. Q. Yao, Angew. Chem. Int. Ed. 2021, https://doi.org/
10.1002/anie.202101964.
[
[
6] A. S. Klymchenko, F. Liu, M. Collot, N. Anton, Adv. Healthcare Mater.
2
021, 10, 2001289.
7] a) F. Danhier, E. Ansorena, J. M. Silva, R. Coco, A. Le Breton, V. Preat, J.
Controlled Release 2012, 161, 505; b) A. Kumari, S. K. Yadav, S. C. Yadav,
Colloids Surf. B 2010, 75, 1.
[
8] B. Andreiuk, A. Reisch, M. Lindecker, G. Follain, N. Peyriéras, J. G. Goetz,
A. S. Klymchenko, Small 2017, 13, 1701582.
[9] A. Reisch, K. Trofymchuk, A. Runser, G. Fleith, M. Rawiso, A. S.
Klymchenko, ACS Appl. Mater. Interfaces 2017, 9, 43030.
10] K. Trofymchuk, A. Reisch, I. Shulov, Y. Mély, A. S. Klymchenko, Nanoscale
[
(
blinking) to steady emission. Finally, the acyl barbiturates
displayed significantly lower cytotoxicity compared to tetrakis
pentafluorophenyl)borate. Moreover, the least toxic acyl barbi-
2
014, 6, 12934.
[11] a) S. Snipstad, et al., Cytometry Part A 2017, 91, 760; b) K. Trofymchuk, J.
Valanciunaite, B. Andreiuk, A. Reisch, M. Collot, A. S. Klymchenko, J.
Mater. Chem. B 2019, 7, 5199.
(
turate BarPh2 was both the more hydrophobic and slightly less
acidic one, which is probably related to its inability to get
deprotonated and thus to transfer the proton and other cations
through biological membranes. These results provide insights
on molecular design of bulky hydrophobic counterions with
minimized cytotoxicity. Overall, barbiturate anions, owing to
these unique properties, open the route to preparation of non-
toxic and eco-friendly dye-loaded polymer NPs, which can find
application in biosensing, bioimaging and materials science.
[
[
[
[
12] F. Wurthner, T. E. Kaiser, C. R. Saha-Moller, Angew. Chem. Int. Ed. 2011,
50, 3376.
13] W. Qin, D. Ding, J. Z. Liu, W. Z. Yuan, Y. Hu, B. Liu, B. Z. Tang, Adv. Funct.
Mater. 2012, 22, 771.
14] B. Zhang, H. Soleimaninejad, D. J. Jones, J. M. White, K. P. Ghiggino, T. A.
Smith, W. W. H. Wong, Chem. Mater. 2017, 29, 8395.
15] a) A. Reisch, P. Didier, L. Richert, S. Oncul, Y. Arntz, Y. Mély, A. S.
Klymchenko, Nat. Commun. 2014, 5, 4089; b) C. R. Benson, et al., Chem.
2
020, 6, 1978.
[16] I. Krossing, I. Raabe, Angew. Chem. Int. Ed. 2004, 43, 2066.
[
17] T. A. Engesser, M. R. Lichtenthaler, M. Schleep, I. Krossing, Chem. Soc.
Rev. 2016, 45, 789.
[
[
18] A. B. A. Rupp, I. Krossing, Acc. Chem. Res. 2015, 48, 2537.
19] a) D. K. Bwambok, B. El-Zahab, S. K. Challa, M. Li, L. Chandler, G. A. Baker,
I. M. Warner, ACS Nano 2009, 3, 3854; b) P. Magut, S. Das, V. Fernand, J.
Losso, K. McDonough, B. Naylor, S. Aggarwal, I. Warner, J. Am. Chem.
Soc. 2013, 135, 15873; c) H. Yao, K. Ashiba, RSC Adv. 2011, 1, 834; d) T.
Funada, T. Hirose, N. Tamai, H. Yao, Phys. Chem. Chem. Phys. 2015, 17,
Experimental Section
Detailed protocols of the synthesis and characterization of
barbiturates, preparation and characterization of NPs, their studies
in cells as well as some data on TLC of the ion pairs and zeta
potential of NPs are provided in the Supporting Information.
1
1006.
[
[
[
[
20] I. Shulov, S. Oncul, A. Reisch, Y. Arntz, M. Collot, Y. Mely, A. S.
Klymchenko, Nanoscale 2015, 7, 18198.
21] B. Andreiuk, A. Reisch, V. G. Pivovarenko, A. S. Klymchenko, Mater.
Chem. Front. 2017, 1, 2309.
22] I. O. Aparin, N. Melnychuk, A. S. Klymchenko, Adv. Opt. Mater. 2020, 8,
Acknowledgements
2
000027.
23] N. Melnychuk, S. Egloff, A. Runser, A. Reisch, A. S. Klymchenko, Angew.
Chem. Int. Ed. 2020, 59, 6811.
This work was supported by the European Research Council
ERC Consolidator grant BrightSens 648528. B.A. was supported
by LabEx Chimie des Systèmes Complexes. I.O.A. acknowledges
support from Marie Curie post-doctoral research grant (H2020-
MSCA-IF-2018, DNANanoProbes 846571) by the European
Union.
[
[
24] N. Melnychuk, A. S. Klymchenko, J. Am. Chem. Soc. 2018, 140, 10856.
25] K. Trofymchuk, A. Reisch, P. Didier, F. Fras, P. Gilliot, Y. Mely, A. S.
Klymchenko, Nat. Photonics 2017, 11, 657.
26] I. Khalin, D. Heimburger, N. Melnychuk, M. Collot, B. Groschup, F. Hellal,
A. Reisch, N. Plesnila, A. S. Klymchenko, ACS Nano 2020, 14, 9755.
27] W. Michael Arnold, U. Zimmermann, W. Heiden, J. Ahlers, Biochim.
Biophys. Acta 1988, 942, 96.
[
[
[
[
28] a) S. Fischer, J. Schmidt, P. Strauch, A. Thomas, Angew. Chem. Int. Ed.
2
013, 52, 12174; b) A. Franzke, A. Pfaltz, Chem. Eur. J. 2011, 17, 4131.
29] J. Chen, S. M. A. Fateminia, L. Kacenauskaite, N. Bærentsen, S. G.
Stenspil, J. Bredehoeft, K. L. Martinez, A. H. Flood, B. W. Laursen, Angew.
Chem. Int. Ed. 2021, 60, 9450.
Conflict of Interest
[
[
30] J. E. Gonzalez, R. Y. Tsien, Chem. Biol. 1997, 4, 269.
31] A. Reisch, A. Runser, Y. Arntz, Y. Mély, A. S. Klymchenko, ACS Nano 2015,
The authors declare no conflict of interest.
9
, 5104.
[
[
32] B. Andreiuk, A. Reisch, E. Bernhardt, A. S. Klymchenko, Chem. Asian J.
Keywords: aggregation, bulky hydrophobic anions, dyes/
pigments, fluorescence, nanoparticles
2019, 14, 836.
33] a) K. A. Colby, J. J. Burdett, R. F. Frisbee, L. Zhu, R. J. Dillon, C. J. Bardeen,
J. Phys. Chem. A 2010, 114, 3471; b) C. J. Bardeen, in Annual Review of
Physical Chemistry, Vol 65, Vol. 65 (Eds.: M. A. Johnson, T. J. Martinez),
2
2
014, pp. 127; c) R. Camacho, D. Tauber, I. C. Scheblykin, Adv. Mater.
019, 31, 1805671.
[
1] a) O. S. Wolfbeis, Chem. Soc. Rev. 2015, 44, 4743; b) N. Hildebrandt, C. M.
Spillmann, W. R. Algar, T. Pons, M. H. Stewart, E. Oh, K. Susumu, S. A.
Díaz, J. B. Delehanty, I. L. Medintz, Chem. Rev. 2017, 117, 536; c) P. D.
Howes, R. Chandrawati, M. M. Stevens, Science 2014, 346, 53.
[
[
2] Y. F. Jiang, J. McNeill, Chem. Rev. 2017, 117, 838.
3] a) J. Mei, N. L. C. Leung, R. T. K. Kwok, J. W. Y. Lam, B. Z. Tang, Chem. Rev.
Manuscript received: June 6, 2021
Accepted manuscript online: June 23, 2021
Version of record online: ■■■, ■■■■
2015, 115, 11718; b) K. Li, B. Liu, Chem. Soc. Rev. 2014, 43, 6570.
Chem. Eur. J. 2021, 27, 1–8
www.chemeurj.org
7
© 2021 Wiley-VCH GmbH
��
These are not the final page numbers!

FULL PAPER
Bulky hydrophobic anions based on
fluorinated C-acyl barbiturates with
delocalized negative charge are found
to provide insulation of ionic dyes,
thus preventing their aggregation-
caused quenching in polymeric nano-
particles. Their chemical structure
defines dye encapsulation, particle
brightness and on/off switching.
Compared to currently used anions, C-
acyl barbiturates are non-toxic, while
offering chemical availability and ver-
satility.
Dr. B. Andreiuk, Dr. I. O. Aparin, Dr. A.
Reisch, Dr. A. S. Klymchenko*
1
– 8
Bulky Barbiturates as Non-Toxic
Ionic Dye Insulators for Enhanced
Emission in Polymeric Nanoparticles