Phenothiazine based co-crystals with enhanced luminescence

Article abstract of DOI:10.1016/j.dyepig.2019.108164

A series of binary phenothiazine based co-crystals were designed with the aim to reach materials with improved quantum efficiency in solid state. To this end, a host dibromine functionalized phenothiazine was mixed with a guest formyl based phenothiazine

Full text of DOI:10.1016/j.dyepig.2019.108164

Dyes and Pigments 175 (2020) 108164
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: http://www.elsevier.com/locate/dyepig
Phenothiazine based co-crystals with enhanced luminescence
Luminita Marin*, Andrei Bejan, Sergiu Shova
“Petru Poni” Institute of Macromolecular Chemistry of Romanian Academy, Iasi, Romania
A R T I C L E I N F O
A B S T R A C T
Keywords:
A series of binary phenothiazine based co-crystals were designed with the aim to reach materials with improved
quantum efficiency in solid state. To this end, a host dibromine functionalized phenothiazine was mixed with a
guest formyl based phenothiazine in various molar ratios, from 1:1 up to 1:9, to form single crystals and thin
films as well. Single crystal X-ray diffraction, polarized light microscopy, 1H NMR and FTIR spectroscopies
indicated the successful formation of supramolecular synthons via halogen bonds. Polarized light microscopy and
atomic force microscopy indicated the formation of submicrometric crystals into the co-crystal films. The pho-
tophysical behaviour was investigated by UV–vis and photoluminescence spectroscopies, and showed a
remarkable improvement of the quantum yield, reaching values of 27% for single co-crystals and of 42% for thin
films of nano-co-crystals.
Phenothiazine
Co-crystals
Halogen bonds
Quantum efficiency
Miscibility
1. Introduction
a nucleophilic site promotes the spin-orbit coupling produced by the
heavy atom effect, which in turn led to an amazing increase of the
Co-crystals are hybrid materials designed in order to achieve com-
bined properties of two or more compounds and additional/synergic
properties residing from their interaction. Discovered by Wohler in 1844
[1], the organic co-crystallization emerged as an excellent strategy for
better understanding of the biologic processes, e.g. protein-DNA
recognition [2], lipid organization [3], protein-protein interactions
which are effective in muscle contraction [4] and for improving the
materials properties in view of their practical application, e.g. solubility
and stability of pharmaceutics [5–7], mechanical properties of the
crystalline solids [8,9], quantum efficiency of the optical materials
[10–12], charge mobility of the semiconductors [13–15], or lasing [16].
In the field of optoelectronic materials, organic co-crystallization proved
to be an attractive approach for improving the luminescence and tuning
the colour of emitted light. Many chromophore building blocks were
used to this aim, aromatic or heteroaromatic ones, e.g. phenylene,
naphthalene, anthracene, pyrene, phenantrene, fluorene, carbazole,
oxadiazole, dibenzofuran, dibenzothiophene, naphthalene, phenantro-
line, phenantridine, benzoquinoline [16–20]. Their combination in bi-
nary or even ternary co-crystals gave promising results which
challenged the scientific community to investigate more novel mixed
crystalline systems to further enrich the luminescent material family.
The most spectacular results were obtained when the driving force for
crystallization was the formation of halogen bonds [21,22]. It was
demonstrated that the non-covalent forces between a halogen atom and
quantum efficiency by activating the phosphorescence phenomenon.
The effect is so effective, that outstanding results were obtained even for
simple molecules based on a single aromatic ring, functionalized by
bromine halogen and formyl nucleophile [10].
Phenothiazine is a fused heterocycle with strong electron-donating
ability, employed in building donor-acceptor systems for electronic
and opto-electronic devices, such as solar cells, field effect transistors or
organic light emitting diodes [23–26]. Our group research in the syn-
thesis and investigation of new phenothiazine derivatives revealed its
promising potential for preparing materials with excellent emission in
solid state, a primary necessity in building OLEDs [27–32]. Inspired by
the great luminescence improvement yielded by the halogen bond
driven co-crystals, we designed and prepared co-crystals based on the
phenothiazine chromophore functionalized with bromine and formyl
unit, respectively. The rational design was as follows.
2. Rational design
A series of binary co-crystals of two phenothiazine derivatives, a
guest containing bromine and formyl units (7-Bromo-10-(-4-hexyloxy-
phenyl)-10H-phenothiazine-3-carbaldehyde (noted A)) doped into a di-
bromine containing host (2,7-Dibromo-10-(4-hexyloxy-phenyl)-10H-
phenothiazine (noted B)), in different molar ratios, has been prepared in
order to reach new materials with improved photophysical properties. In
* Corresponding author.
E-mail address: lmarin@icmpp.ro (L. Marin).
https://doi.org/10.1016/j.dyepig.2019.108164
Received 18 November 2019; Received in revised form 19 December 2019; Accepted 20 December 2019
Available online 24 December 2019
0143-7208/© 2019 Elsevier Ltd. All rights reserved.

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
this regard, their rational design took into consideration the structural
peculiarities of the two components, as follows. (i) Both compounds
have the ability to grow as crystals, moreover, the A compound easily
grows as single crystals and nanocrystals, from a large variety of solvents
[29,31]; (ii) they are isostructural, fact which should be kinetically and
thermodynamically favourable for co-crystallization, i.e. they contain a
phenothiazine core and an aliphatic tail, which should facilitate their
co-crystallization by rigid-flexible segregation; (iii) the two compounds
have functional groups (CHO and Br) favourable for the occurrence of
supramolecular synthons by halogen bonding, offering thus the pre-
dictability of co-crystals growing and good luminescence properties as
well [10,22].
bromosuccinimide dissolved into 5.2 mL chloroform and 3.48 mL acetic
acid. The product was recrystallized from ethyl acetate to give yellow
greenish needle single crystals.
η
¼ 77%; m.p. ¼ 147.1–150.9 �C.
1H NMR (400.13 MHz, DMSO‑d6, ppm) δ ¼ 9.74 (s, 1H), 7.54 (s, 1H),
7.47 (d, 1H), 7.39 (d, 2H), 7.32 (s, 1H), 7.25 (d, 2H), 7.13 (d, 1H), 6.22
(d, 1H), 6.04 (d, 1H), 4.09 (t, 2H), 1.83–1.76 (m, 2H), 1.53–1.34 (su-
perposed bands, 6H), 0.93 (t, 3H).
3.3. Co-crystal growth
3. Experimental
3.1. Materials
7-Bromo-10-(4-hexyloxy-phenyl)-10H-phenothiazine-3-carbalde-
hyde (coded A) and 2,7-Dibromo-10-(4-hexyloxy-phenyl)-10H-pheno-
thiazine (coded B) were synthesised in our laboratory, accordingly to a
published procedure [29]. Their synthesis is shortly described below.
Ethyl acetate and hexane were purchased from Aldrich and dried on
molecular sieves before use.
3.2. Synthesis
The single crystal technique provides unequivocally information
related to the intermolecular forces between molecules. This is of great
interest in the understanding of the structure-properties relationship
which guides further studies in the design of high performance mate-
rials. This was the reason why, one of the important tasks of this work,
was to grow single co-crystals suitable for X-ray analysis. To do this, we
designed growing conditions considering the structure of the two com-
ponents, as follows. (i) A 1/1 (v/v) mixture of polar/nonpolar solvents
has been chosen as growing medium, in line with the aromatic polar part
and the aliphatic nonpolar residue of A and B molecules (ii) The solution
was kept in sealed vials in order to decrease the evaporation rate and
thus to assure a sufficient time for co-crystal growth. This decision was
done considering the weak intermolecular forces between pure A and B
molecules, respectively [29], which enabled us to estimate weak inter-
molecular forces between them. (iii) To limit the possibility of crystal-
lization of the pure components, and to favour the co-crystallization, the
vials were kept at a relative constant temperature around 27 �C, in the
darkness, in order to avoid the light radiations. After many attempts, the
successful pathway towards single co-crystals was the following. A and
B, in different molar ratios (Table 1), were dissolved into a 1/1 mixture
of ethyl acetate/hexane to give a 1% clear solution. The solution vial was
kept 1 h at 45 �C, to assure a good mixing of the two components, and
then the vial was sealed and kept into a black closed box. The
co-crystallization occurred at different times (Table 1), related to the
content of B, i.e. longer time was necessary for higher percentage of B.
The co-crystals were yellow-greenish in colour, with a more pronounced
green tone as the content of B increased (Scheme 1). The single crystal
X-ray diffraction confirmed the presence of the A and B components in
3.2.1. 2,7-Dibromo-10-(4-hexyloxy-phenyl)-10H-phenothiazine (B)
In a round bottom flask were introduced sequentially: 0.5 g of 10-(4-
hexyloxyphenyl)-10H-phenothiazine, 3 mL chloroform, 4.5 mL acetic
acid and 0.495 g of N-bromosuccinimide, and then the flask was
immersed in an ice bath. The reaction mixture has been vigorous stirred
for 2 h, at 0 �C, under nitrogen atmosphere. After that, the organic phase
was neutralized by adding 10 mL of 1% NaOH, washed three times with
brine solution in a separating funnel, extracted three times with chlo-
roform, dried on anhydrous MgSO4, filtered, and concentrated by rotary
evaporation to give viscous oil. The product was purified by recrystal-
lization from cold ethyl acetate (in a freezer) when crystals in shape of
white foils, proper for diffraction, were obtained.
η ¼ 81%; m.p. ¼
106.3–108.1 �C
1H NMR (400.13 MHz, DMSO‑d6, ppm) δ ¼ 7.33 (d, 2H), 7.27 (s, 2H),
7.20 (d, 2H), 7.10 (d, 2H), 6.02 (d, 2H), 4.06 (t, 2H), 1.80–1.83 (m, 2H),
1.48–1.33 (superposed bands, 6H), 0.90 (t, 3H).
3.2.2. 7-Bromo-10-(-4-hexyloxy-phenyl)-10H-phenothiazine-3-
carbaldehyde (A)
Table 1
Single co-crystal growing conditions.
Code
A
C5
C4
C3
C2
C1
B
A/B molar ratio
100/
0
50/
50
40/
60
30/
70
20/
80
10/
90
0/
100
0
A (mg)
100
0
50
40
30
20
10
B (mg)
50
60
70
80
90
100
EtAc/Hexane
(mL)
5/5
time (months)a
1/30b
3
6
6
6
8
10
a
b
The similar procedure described above was applied for 0.3 g 10-(-4-
Co-crystallization time.
hexyloxy-phenyl)-10H-phenothiazine-3-carbaldehyde and 0.132 g N-
The crystallization occurred in 1 day.
2

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
the C1–C5 co-crystals, but only for C5 the crystal quality was enough
satisfactory to reach a complete fitting of the structure and thus to
provide unequivocal information. This is why, the X-ray data for C1–C4
were used for discussions, but not exhaustive presented in the paper.
Crystal data for C5: C24.5H22.5Br1.5NO1.5S, Mr ¼ 508.97 g molÀ 1, size
0.30 � 0.20 � 0.10 mm3, trilinic, space group P-1, a ¼ 8.1116 (16) Å, b
Table 2
Combination ratios of B and A in co-crystals, calculated from 1H NMR.
Code
C5
C4
C3
C2
C1
A/B ratio (expected)
A/B ratio (calculated)
1/1
1/1.5
1/1.6
1/2.3
1/2.3
1/4
1/9
1/0.97
1/4.4
1/9.4
¼ 8.9559 (4) Å, c ¼ 15.9838 (9) Å,
α
¼ 88.067 (4)�, β ¼ 76.471 (5)�, γ ¼
79.330 (5)�, V ¼ 1109.36 (11) Å3, Z ¼ 2, ρcalcd ¼ 1.524 g cmÀ 3
, μ(MoKα)
and Origin 8 software.
¼ 2.920 mmÀ 1, F (000) ¼ 516, 9367 reflections in h (-9/9), k (-10/10), l
(-19/19), measured in the range 4.628 � 2Θ � 50.05, T ¼ 180 K,
Crystallographic measurements for compound C5 were carried out
with an Oxford-Diffraction XCALIBUR E CCD diffractometer equipped
with graphite-monochromated Mo-Kα radiation. The single crystal was
completeness Θmax ¼ 99.92%, 39082 independent reflections, Rint
¼
positioned at 46 mm from the detector and 506 frames were measured
each for 250 s over 1� scan width. The unit cell determination and data
integration were carried out using the CrysAlis package of Oxford
Diffraction [33]. The structure was solved by direct methods using
Olex2 [34] and refined by full-matrix least-squares on F2 with
SHELXL-2015 [35]. Atomic displacements for non-hydrogen atoms were
refined using an anisotropic model. All H atoms were introduced in
idealized positions using the riding model with their isotropic
displacement parameters fixed at 120% of their riding atom. The posi-
tional parameters of disordered atoms were refined in combination with
PART restraints using anisotropic model for non-H atoms. When the
occupancy factors for disordered bromide and aldehyde were allowed to
refine they confirmed their equivalent distribution in the crystal. The
molecular plots were obtained using the Olex2 program.
0.0241, 241 parameters, 0 restraints, R1obs ¼ 0.0698, wR2obs ¼ 0.0875,
R1all ¼ 0.1852, wR2all ¼ 0.2003, GoF ¼ 1.041, largest difference peak
and hole: 0.94/-1.42 e AÀ 3. CCDC 1965164 contain the supplementary
crystallographic data for this contribution. These data can be obtained
free of charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
In order to obtain thin films of co-crystals, the 1% clear solutions
were casted on glass lamella and kept into a black closed box for two
weeks, when thin depositions were obtained. The samples were noted
C1f-C5f.
3.4. Equipment and techniques
1H NMR spectra of the A and B pure compounds and the five C1–C5
co-crystals were registered on Bruker equipment, in DMSO‑d6 solvent. In
order to determine the ratio of the two components into the co-crystals,
their 1H NMR spectra were recorded in triplicate, on three different
single crystals, and the average molar ratio was taken as result. The ratio
of the A guest to B host was calculated as IB(CH2)/IA(CH2), where IB(CH2)
represents the integral of the CH2 protons neighbour to oxygen in B and
Atomic force microscopy images on the surface of co-crystal thin
films were acquired with a Solver PRO-M, NT-MDT, Russia instrument,
in semi-contact mode. Nova v.1443 software was used to record and to
analyse the AFM topographic and phase contrast images.
Miscibility test, transition temperatures and textures of the co-
crystals were monitored on an Axioscop 40 Zeiss polarizing optical
microscope equipped with Linkam heating stage, Linksys 32 tempera-
ture control system, and QImaging/Retiga-1000R camera. For all mea-
I
A(CH2) represents the integral of the CH2 protons neighbour to oxygen in
A. The obtained values were verified with the equation: (IAr-H – IA)/IA,
where IAr-H is the integral of the total number of the aromatic protons in
co-crystal and IA is the integral of the aromatic protons of A. For both
equations, IA(CH2) and IA was found setting the integral of the aldehyde
proton ¼ 1. The results were given in Table 2. They were in good
agreement for both cases.
surements
a 40x objective was used. The eyepiece has 10�
magnifications. For miscibility test, equimolar amounts of A and B were
deposited in contact on a glass lamella.
The absolute values of the quantum yield (Φfl) were measured on a
FluoroMax-4 spectrofluorometer in conjunction with a Quanta-phi
integrating sphere accessory Horiba Jobin Yvon, at room temperature.
The samples were excited with the light corresponding to the absorption
maxima of A and B molecules in THF solution. In order to obtain a good
FTIR spectra were recorded on KBr pellets, using a FTIR Bruker
Vertex 70 Spectrometer, at room temperature, with a resolution of 2
cmÀ 1 and accumulation of 32 scans. They were processed using Opus 6.5
Scheme 1. Representation of the co-crystals preparation and their images under UV lamp.
3

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
optical luminescence signal-to-noise ratio, the slit width and detector
parameters were optimized to maximize but not saturate the excitation
Rayleigh peak.
4. Results and discussions
A series of five binary co-crystals was prepared by using a mixed
crystal design consisting in the combination in various molar ratios of
two phenothiazine based compounds: one functionalized in 3,7 posi-
tions with bromine (B), and the other functionalized with bromine and
formyl units (A) (Scheme 1). The co-crystals, abbreviated with C1–C5
codes (Table 1), were prepared as single crystals and thin films. Under
an UV lamp, they emitted yellow-green light of various intensities,
pointing for the encapsulation of different amounts of A guest in the B
host. 1H NMR spectra of the co-crystals displayed the chemical shifting
characteristic for the two components (Fig. S1), and allowed the
calculation of their combination ratio (Table 2). The initial molar ratio
and that calculated from NMR spectra presented a good concordance,
confirming the successful encapsulation by co-crystallization of A guest
into the B host. Further, FTIR spectra of the co-crystals revealed a new
vibration band at 1741 cmÀ 1, attributed to the occurrence of Br⋯O
halogen bonds [36]. Moreover, the vibration band of the formyl group
was shifted to higher wavenumbers, from 1676 to 1683 cmÀ 1, in
accordance with the alteration of the conjugation caused by the occur-
rence of halogen bonding (Fig. S2).
4.1. Supramolecular architecting of the co-crystals
Fig. 1. View of the 3D supramolecular architecture in the C5 co-
crystal structure.
To understand the ability of the supramolecular aggregation of co-
crystals to tune their photophysical properties, the crystalline struc-
ture of the C1–C5 samples was investigated by single crystal X-ray
diffraction. A previous reported structure analysis of the A and B com-
pounds revealed specific features [29], as follows. (i) The phenothiazine
core adopted an almost planar conformation (dihedral angle of 3.4�) for
A and a butterfly one, folded along the S–N vector (dihedral angle of
23.8�) for B; (ii) The supramolecular packing of A and B molecules was
4.2. Morphology and thermotropic behaviour
The structural compatibility of the two compounds used for co-
crystallization was investigated by a miscibility test monitored by hot
stage polarized optical microscopy [39,40]. Equal amounts of B and A
were placed in contact on a lamella, heated up to isotropic state and
further cooled (The obtained mixed sample was noted A-B). During the
heating, the B compound started to melt at a temperature much lower
than its Tm (72 �C) and diffused towards the A crystals, surrounding
them (Fig. 2c). The A crystals started to melt slowly at 79 �C and reached
a completely isotropic phase at the temperature corresponding to the
melting point of A (150 �C). No phase segregation was observed. Cooling
down the isotropic liquid, the crystallization occurred immediately, at
149 �C, displaying a fast growing of board-like shapes (Fig. 2d), a
crystallization pattern completely different compared to that of the pure
components; B formed by cooling an amorphous glass, and A crystalized
forming a continuous strong birefringent texture (Fig. 2a and b) [29].
This behaviour confirmed the structural compatibility of the two com-
ponents, which led to a good miscibility, and consequently to the ability
to co-crystalize together forming a new phase.
governed by the π-π stacking, with centroid-to-centroid distances of
3.632 for A, and 3.820 Å for B. (iii) The closest distance between the
bromine atoms in B architecture was of 5.08 Å, indicating no cohesive
–
–
interactions between halogens [37]. (iv) The distance between the C
O
group and bromine atoms into A architecture was of 3.973 Å.
By comparison, in the C5 structure significant configuration alter-
ations appeared. First, A and B molecules are crystallographic asym-
metric but chemical and geometric very similar. (i) They exhibited a
similar planarity of the phenothiazine core, with a dihedral angle along
S–N vector of 5.3977�. This indicates that the intermolecular forces
which governed their supramolecular packing inflicted the planariza-
tion of B molecules, while A molecules become slightly less planar. (ii)
Consequently, the centroid-centroid distance was of 3.747 Å, an average
value between those of pure A and B crystals. (iii) Contrary to the pure A
and B crystals, the C5 co-crystal displayed quite short Br–Br contacts of
3.646 Å, indicating them as guiding forces in the packing process. (iv)
Melting point is a well-known parameter used for identification and
characterization of the organic compounds, its value reflecting the in-
tensity of intermolecular forces among the molecules. In this light, the
thermotropic behaviour of the C1–C5 co-crystals has been monitored by
hot stage polarized light microscopy on a heating/cooling cycle, in order
to have additional information on their formation, and besides to esti-
mate the intensity of the new forces developed between the two com-
ponents. As a general remark, the co-crystals melted on a large
temperature range compared to the pure components (10–20 �C
compared to 2–3 �C), reflecting their discontinue nature (Table 3). The
onset of melting occurred at higher temperatures compared to that of the
dibromine host, attributable to the stronger intermolecular forces
developed between the two components (see Supramolecular architecting
of the co-crystals). Moreover, the C2 melting onset was even higher than
that of the A component, while for C2 and C3 co-crystals even the end of
melting occurred at higher temperatures. During the cooling, the co-
Moreover, in the C1–C4 co-crystals, a main distinctive feature was the
–
occurrence of the Br⋯O CH- halogen bonds, of 2.98 and 2.89 Å
–
(Fig. S3), placing them among the shortest Br⋯O halogen bonds re-
ported in literature, indicative of electronic interactions between the
two atoms, i.e. a directed heavy atom effect which may be responsible
for enhancing the luminescence of co-crystals [10,38]. The occurrence
of the halogen bonds appear to be responsible for the formation of su-
pramolecular synthons which characterize the crystal packing motif. In
the crystal they are further extended through π-π stacking to form su-
pramolecular ribbons, which in combination with the rigid-flexible
self-assembling guided the formation of the three-dimensional supra-
molecular network (Fig. 1).
4

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
out for a similar ordering pattern. This enables us to believe that under
the pressure of the fast solvent removal, the crystallization yielded
micro/nano-crystals. To have a complementary confirmation of this
POM observation, the morphology of the C1f-C5f films was investigated
at nano-level by atomic force microscopy. Topographic images of their
surface showed a rough, granular morphology, resulted by the coales-
cence of sub-micrometric grains, reaching nanometric size in the case of
C1 (Fig. 3, Fig. S4). Except C5f, the films showed low values of the
roughness, from 15 to 24 nm, when scanning area was of 1 � 1
μ
m2, in
line with a low size of the crystals forming the films, certainly at
sub-micrometric level (Table S1). Moreover the values of the roughness
exponent extracted from the graphical representation of the average
roughness versus scan size gave values which increased as the content of
the A increased into the samples [42,43]. This suggested that the size of
the crystals was correlated to the presence of A with high crystallization
ability, i.e. larger crystals were obtained for higher amount of A into
co-crystal. A similar trend has been showed by the average roughness
measured on 30 � 30
μ
m2 (Table S1).
The data correlated well with the photophysical measurements, the
best quantum yield being obtained for the C2f sample, which displayed
the lowest roughness attributable to the smallest crystals. They should
assure enough high surface-to-volume ratio, while having a reasonable
amount of A to give a high percent of Br⋯O synthon fluorophores.
5. Photophysical properties
The main goal of the study was to create luminescent materials by
the co-crystallization strategy, starting from phenothiazine based com-
pounds with high quantum efficiency in solution but lacking solid-state
luminescence. The effective implementation of this approach was
investigated by measuring the quantum yield, in single crystal and thin
film, for the C1–C5, C1f-C5f samples and their A and B precursors. The
quantum yield has been measured when exciting with light of two
different wavelengths, (i) 395 nm corresponding to the absorption
maximum of intramolecular charge transfer (ICT) transition of A and (ii)
Fig. 2. Optical polarized microphotographs, crossed polarizers: pure A and B
precursors (a,b) and their mixture (c, d) and co-crystals (e,f) (1H: first heating;
1C: first cooling).
342 nm corresponding to the π-π* transition of the conjugated skeleton
Table 3
of A, which was close to that of B (339 nm) (Fig. S5). The results were
given in Table 4.
Thermotropic behaviour of the co-crystals and pure components.
Code
Heating
Tomnset
Cooling
It should be noted that the high quantum yield of 66%, measured for
A in solution (when excited with 342 nm) [29] was abruptly suppressed
around 2% in solid state. B showed a quantum yield of 8% in solution
and of 0.7% in solid state. The drastic luminescence quenching was
Temnd
A
147
110
157
150
120
173
The crystallization process occurred at 121 �
C
C1
C2
The sample froze forming an amorphous glass.
The sample froze forming an amorphous glass, which
subsequently crystallized.
attributed to the
π-π* stacking in solid state which favoured the
non-radiative deexcitation through converting the energy to the excimer
formation [29,44].
C3
C4
C5
B
135
134
133
106
155
148
140
108
The sample froze forming an amorphous glass, which
subsequently crystallized.
The sample freeze forming an amorphous glass, which
subsequently crystallized.
Nevertheless, the co-crystals showed a significant enhancement of
the quantum yield compared to their precursor, both for single crystals
and thin films (Table 4). The quantum yield reached values of 7% (C1)
and 19% (C2f) when the samples were excited with light of higher en-
ergy (λ ¼ 342 nm). A spectacular enhancement of the quantum yield
values was noted when the samples were excited with the light of lower
energy (λ ¼ 395 nm), yielding values of 27% (C1) and 42% (C2f). This
should be surprising considering that the ICT transition induced by
The sample freeze forming an amorphous glass, which
subsequently crystallized.
The sample freeze forming an amorphous glass.
Tomnset: the melting starts at the crystals border; Temnd: the sample reached a
completely isotropic state.
crystals initially froze in isotropic state forming an amorphous glass, in a
similar way to B. However, after one week C2–C5 samples showed
strong birefringent textures characteristic to the crystalline state (Fig. 2e
and f). This should be correlated with the preponderance of B host,
which hampered the occurrence of the specific intermolecular forces in
viscous molten state, i.e. halogen bonds, delaying the crystallization.
Nevertheless, the thermotropic behaviour confirmed the formation of
the co-crystals by developing new strong intermolecular forces, i.e.
halogen bonds.
exciting with light of 395 nm is less probably compared to the
transition stimulated by exciting with light of 342 nm [29]. It appears
that exciting with the energy necessary for the * transition, an
π-π*
π-π
important loss of energy by non-radiative relaxation occurred, a phe-
nomenon generally met for conjugated compounds in solid state [44,
45]. On the other hand, the remarkable improvement of the quantum
yield when exciting with energy characteristic to the ICT transition can
be correlated with the halogen bonding between the carbonyl oxygen of
A molecules and bromine of A or B ones. It could be envisaged that the
more localized charge of the ICT complex of A directed a better delo-
calization of the excited state electrons from the carbonyl oxygen to the
outer orbitals of the heavy bromine atoms facilitating the spin–orbit
coupling, and therefore amplifying the radiative deactivation by
Further, polarized light microscopy has been employed to investigate
the phase nature of the C1f-C5f thin films. They showed a birefringent
granular texture, signature of a short-range ordered phase [41]. Their
heating revealed a similar behaviour to the C1–C5 co-crystals, pointing
5

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
Fig. 3. Representative AFM images of the co-crystal films.
6

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
structure of the co-crystals, it should be regarded that the positive effect
of the halogen bonding on the quantum efficiency overcame the
Table 4
Quantum yield (QY) values of the studied samples.
self-quenching favoured by the π-π* stacking.
Code
QYSC
Code
QYfilm
Code
QYphysical
To highlight the importance of co-crystallization on the lumines-
cence efficiency, the quantum yield has been measured on powder of the
two components grinded together in the corresponding molar ratio
(Table 4). The quantum yield presented values even lower than those of
the pure crystals, supporting the hypothesis of the positive influence of
halogen bonding on the quantum efficiency.
mixture
λex
:
λex
:
λex
:
λex
:
λex
:
λex:
395
342
395
342
395
342
A
4.2
2.12
3.7
Af
2.33
3.3
1.72
3.7
A
–
1.2
–
0.39
C5
11.2
C5f
A-B
(1:1)
A-B
C4
C3
C2
C1
B
13.6
17.2
17.3
27.1
–
5.3
6.07
8.8
7.8
0.7
C4f
C3f
C2f
C1f
Bf
12.9
33.4
42
8.2
2.53
2.35
4.04
4.74
–
1.54
0.72
0.68
0.71
–
Contrary to other reported data, the C1f-C5f thin films showed
higher quantum yield than the single co-crystals [10]. This was attrib-
uted to the larger surface-to volume ratio raised by the multitude of
small co-crystals compared to that of a unique co-crystal. The better
quality of the co-crystals favoured by the great crystallization potential
of the phenothiazine based derivatives should play also an important
role to this.
(1:1.5)
A-B
14.2
19.9
12.02
2.18
(1:2.3)
A-B
(1:4)
A-B
37.8
–
(1:9)
B
Comparing the values of the quantum yield, it can be remarked an
increasing trend, as the amount of A guest into B host decreased. This
was associated with the formation of smaller co-crystals into the crystal
of B host [11], inducing a solid solution –like effect [31,46,47].
Looking to the emission spectra, significant differences of the emis-
sion maximum were observed between the C1–C5 single co-crystals and
C1f-C5f co-crystal films compared to those of the pure components
(Fig. 4). Thus, while the A crystal has an emission maximum at 550 nm
and B crystal at 480 nm, the emission maximum of the C1–C5 single co-
crystals and C1f-C5f thin films lays at intermediate values, around 507
and 518 nm, respectively. This indicates a new fluorophore responsible
for emission, i.e. the new formed supramolecular synthons. The shifting
QYSC: quantum yield determined on the C1–C5 single crystal samples; QYfilm
:
quantum yield determined on the C1f-C5f film samples. QYp: quantum yield
determined on the physical mixture of the A and B components grinded together,
in the corresponding molar ratio given in brackets; For both samples, the QY was
measured when excited with light of 395 nm - characteristic to the intra-
molecular charge transfer transition in A, and 342 nm – characteristic to the π-π*
transition of the conjugated system of A and B.
phosphorescence. Complementary, the non-radiative deactivation
channels could be limited, due to the lower transition energy compared
to that of the conjugated skeleton. Considering the supramolecular
Fig. 4. Emission spectra of the single C1–C5 co-crystals when excited at (a) 342 nm and (b) 395 nm and of the C1f-C5f co-crystal films when excited at (c) 342 nm
and (d) 395 nm.
7

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
Fig. 5. Chromaticity diagram of the studied samples when excited with 420 nm as (a) co-crystals and (b) co-crystal films.
of the emission maximum was correspondingly accompanied by the
change of the colour of the emitted light. In the chromaticity diagrams,
the colour of light emitted by C1–C5 co-crystals positioned in the green
visible region, spanning between the yellowish-green of A crystals and
the bluish-green of B crystals (Fig. 5a). In the film state, the C1f-C5f
emission colour shifted more to the green-yellow domain, phenome-
non encountered for nano-crystalline materials, which show a close
correlation between size and colour of emitted light [31,48,49].
European Union’s Horizon 2020 research and innovation programme
under grant agreement 667387.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.dyepig.2019.108164.
References
6. Conclusions
[1] Wohler F. Untersuchungen über das Chinon. Justus Liebigs Ann Chem 1844;51(2):
145–63.
The study demonstrated the co-crystallization guided by halogen
bonding as an excellent strategy for improving the quantum efficiency in
solid state of the isostructural phenothiazine derivatives. This main
conclusion was drawn from the correlation of the structural, supramo-
lecular, thermotropic and photophysical data obtained for a series of
binary co-crystals prepared from bromine and formyl functionalized
phenothiazine derivatives in various molar ratios. It was demonstrated
that combination of the phenothiazine donor with a formyl acceptor
promotes the charge delocalization, which is beneficial for halogen
bonding and thus luminescence improving. The study is an important
step towards solid state materials with high luminescence efficiency for
organic light emitting diodes.
[2] Jordan SR, Whitcombe TV, Berg JM, Pabo CO. Systematic variation in DNA length
yields highly ordered repressor-operator cocrystals. Science 1985;230:1383–5.
[3] Short KW, Wallace BA, Myers RA, Fodor SPA, Dunker AK. Comparison of lipid/
gramicidin dispersions and cocrystals by Raman-scattering. Biochemistry 1987;26:
557–62.
[4] Carr HJ, O’Brien EJ, Morris EP. Structure of tropomyosin-troponin T cocrystals.
J Muscle Res Cell Motil 1988;9:384–92.
[5] Williams HD, Trevaskis NL, Charman SA, Shanker RM, Charman WN, Pouton CW,
Porter CJH. Strategies to address low drug solubility in discovery and development.
Pharmacol Rev 2013;65:315–499.
[6] Trask AV, Motherwell WDS, Jones W. Physical stability enhancement of
theophylline via cocrystallization. Int J Pharm 2006;320:114–23.
[7] Schultheiss N, Newman A. Pharmaceutical cocrystals and their physicochemical
properties. Cryst Growth Des 2009;9:2950–67.
[8] Mukherjee A, Tothadi S, Desiraju GR. Halogen bonds in crystal engineering: like
hydrogen bonds yet different. Acc Chem Res 2014;47:2514–24.
[9] Karki S, Friscic T, Fabian L, Laity PR, Day GM, Jones W. Improving mechanical
properties of crystalline solids by cocrystal formation: new compressible forms of
Paracetamol. Adv Mater 2009;21:3905–9.
Declaration of competing interest
[10] Bolton O, Lee K, Kim HJ, Lin KY, Kim J. Activating efficient phosphorescence from
purely organic materials by crystal design. Nat Chem 2011;3:205–10.
[11] Bolton O, Lee D, Jung J, Kim J. Tuning the photophysical properties of metal-free
room temperature organic phosphors via compositional variations in
bromobenzaldehyde/dibromobenzene mixed crystals. Chem Mater 2014;26:
6644–9.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
CRediT authorship contribution statement
[12] Usman R, Khan A, Wang ML, Luo Y, Sun WW, Sun H, et al. Investigation of charge-
transfer interaction in mixed stack donor-acceptor cocrystals toward tunable solid-
state emission characteristics. Cryst Growth Des 2018;18:6001–8.
[13] Ferraris J, Cowan DO, Walatka Jr V, Perlstein JH. Electron transfer in a new highly
conducting donor-acceptor complex. J Am Chem Soc 1973;95:948–9.
[14] Arkalekha M, Anwesha C, Parameswar KI, Prasenjit M. Charge transfer versus
arene–perfluoroarene interactions in modulation of optical and conductivity
properties in cocrystals of 2,7-di-tert-butylpyrene. J Phys Chem C 2019;123:
18198–206.
Luminita Marin: Conceptualization, Data curation, Funding acqui-
sition, Methodology, Project administration, Supervision, Writing -
original draft, Writing - review & editing. Andrei Bejan: Formal anal-
ysis, Investigation, Software. Sergiu Shova: Formal analysis, Investi-
gation, Software, Writing - original draft.
[15] Huang YJ, Wang ZR, Chen Z, Zhang QC. Organic cocrystals: beyond electrical
conductivities and field-effect transistors (FETs). Angew Chem Int Ed 2019;58:
9696–711.
Acknowledgments
[16] Yu Y, Li Z-Z, Wu J-J, Wei G-Q, Tao Y-C, Pan M-L, et al. Transformation from
nonlasing to lasing in organic solid-state through the cocrystal engineering. ACS
Photonics 2019;6:1798–803.
The paper is part of a project financed through a Romanian National
Authority for Scientific Research MEN – UEFISCDI grant, project num-
ber PN-III-P1-1.2-PCCDI2017-0917, 21PCCDI/2018 and through the
8

L. Marin et al.
Dyes and Pigments 175 (2020) 108164
[17] Sun H, Peng J, Zhao K, Usman R, Khan A, Wang M. Efficient luminescent
microtubes of charge-transfer organic cocrystals involving TCNB, carbazole
derivatives and pyrene derivatives. Cryst Growth Des 2017;17:6684–91.
[18] Khan A, Usman R, Sayed SM, Li R, Chen H, He N. Supramolecular design of highly
efficient two-component molecular hybrids toward structure and emission
properties tailoring. Cryst Growth Des 2019;19:2772–8.
[33] CrysAlis RED. Oxford Diffraction Ltd; 2003. Version 1.171.36.32.
[34] Dolomanov OV, Bourhis LJ, Gildea RJ, Howard JAK, Puschmann H. OLEX2: a
complete structure solution, refinement and analysis program. J Appl Crystallogr
2009;42:339–41.
[35] Sheldrick GM. Crystal structure refinement with SHELXL. Acta Crystallogr C 2015;
71:3–8.
�� �
[36] Yan D, Delori A, Lloyd GO, Friscic T, Day GM, Jones W, et al. A cocrystal strategy to
[19] Liu JJ, Liu T, Xia SB, He CX, Cheng FX, Lin MJ, et al. Luminescence cocrystals of
naphthalene diimide with naphthalene derivatives: a facile approach to tune the
luminescent properties. Dyes Pigments 2018;149:59–64.
tune the luminescent properties of stilbene-type organic solid-state materials.
Angew Chem Int Ed 2011;50:12483–6.
[20] Zhang HC, Guo EQ, Zhang YL, Ren PH, Yang WJ. DonorÀ Acceptor-Substituted
anthracene-centered cruciforms: synthesis, enhanced two-photon absorptions, and
spatially separated frontier molecular orbitals. Chem Mater 2009;21:5125–35.
[21] Goswami A, Garai M, Biradha K. Interplay of halogen bonding and hydrogen
bonding in the cocrystals and salts of dihalogens and trihalides with N,N0-bis(3-
pyridylacrylamido) derivatives: phosphorescent organic salts. Cryst Growth Des
2019;19:2175–88.
[37] Marin L, Harabagiu V, van der Lee A, Arvinte A, Barboiu M. Structure-directed
functional properties of symmetrical and unsymmetrical Br-substituted Schiff-
bases. J Mol Struct 2013;1049:377–85.
[38] Metrangolo P, Resnati G. Halogen bonding: a paradigm in supramolecular
chemistry. Chem Eur J 2001;7:2511–9.
[39] Mitsuo F, Kenichi T, Kato T. Preliminary communication - miscibility of a
hydrogen-bonded mesogenic complex with normal liquid crystals. Liq Cryst 1998;
24:325–7.
[22] Priimagi A, Cavallo G, Metrangolo P, Resnati G. The halogen bond in the design of
functional supramolecular materials: recent advances. Acc Chem Res 2013;46(11):
2686–95.
[40] Zabulica A, Perju E, Bruma M, Marin L. Novel luminescent liquid crystalline
polyazomethines. Synthesis and study of thermotropic and photoluminescent
properties. Liq Cryst 2014;41:252–62.
[23] Huang ZS, Meier H, Cao D. Phenothiazine-based dyes for efficient dye-sensitized
solar cells. J Mater Chem C 2016;4:2404–26.
[41] Marin L, Zabulica A, Sava M. New symmetric azomethinic dimer: the influence of
structural heterogeneity on the liquid crystalline behaviour. Liq Cryst 2011;38:
433–40.
[24] Okazaki M, Takeda Y, Data P, Pander P, Higginbotham H, Monkman AP, et al.
Thermally activated delayed fluorescent phenothiazine-dibenzo[a,j]phenazine-
phenothiazine triads exhibiting tricolor-changing mechanochromic luminescence.
Chem Sci 2017;8:2677–86.
[42] Macanas J, Palacio L, Pradanos P, Hernandez A, Munoz M. Atomic force
microscopy as a suitable technique for surface characterization of activated
composite membranes for metal ion facilitated transport. Appl Phys A 2006;84:
277–84.
[25] Hwang DH, Kim SK, Park MJ, Lee JH, Koo BW, Kang IN, et al. Conjugated polymers
based on phenothiazine and fluorene in light-emitting diodes and field effect
transistors. Chem Mater 2004;16:1298–303.
[43] Marin L, Timpu D, Cozan V, Rusu GI, Airinei A. Solid state properties of thin films
of new copoly(azomethine-sulfone)s. J Appl Polym Sci 2011;120:1720–8.
[44] Chen Y, Lam JWY, Kwok RTK, Liu B, Tang BZ. Aggregation-induced emission:
fundamental understanding and future developments. Mater Horiz 2019;6:428–33.
[45] Albagli D, Bazan GC, Schrock RR, Wrighton MS. New functional polymers prepared
by ring-opening metathesis polymerization: study of the quenching of
luminescence of pyrene end groups by ferrocene or phenothiazine centers in the
polymers. J Phys Chem 1993;97:10211–6.
[26] Shi J, Xu L, Chen C, Lv X, Ding Q, Li W, et al. Efficient and color-purity blue
electroluminescence by manipulating the coupling forms of D‒A hybrids with
phenothiazine as the strong donor. Dyes Pigments 2019;160:962–70.
[27] Zabulica A, Balan M, Belei D, Sava M, Simionescu BC, Marin L. Novel luminescent
phenothiazine-based Schiff bases with tuned morphology. Synthesis, structure,
photophysical and thermotropic characterization. Dyes Pigments 2013;96:686–92.
[28] Belei D, Dumea C, Bicu E, Marin L. Phenothiazine and pyridine-N-oxide-based AIE-
active triazoles: synthesis, morphology and photophysical properties. RSC Adv
2015;5:8849–58.
[46] Malina I, Kampars V, Belyakov S. Luminescence properties of 2-benzoyl-1,3-
indandione based Eu3þ ternary and tetrakis complexes and their polymer films.
Dyes Pigments 2018;159:655–65.
[29] Bejan A, Shova S, Damaceanu MD, Simionescu BC, Marin L. Structure-directed
functional properties of phenothiazine brominated dyes: morphology and
photophysical and electrochemical properties. Cryst Growth Des 2016;16:
3716–30.
[47] Sun Q, Zhang K, Zhang Z, Tang L, Xie Z, Chi Z, et al. A simple and versatile strategy
for realizing bright multicolor mechanoluminescence. Chem Commun 2018;54:
8206–9.
[30] Marin L, Bejan A, Ailincai D, Belei D. Poly(azomethine-phenothiazine)s with
efficient emission in solid state. Eur Polym J 2017;95:127–37.
[48] Zrazhevskiy P, Sena M, Gao X. Designing multifunctional quantum dots for
bioimaging, detection, and drug delivery. Chem Soc Rev 2010;39:4326–54.
[49] Marin L, Zabulica A, Moleavin I-A. Luminescent guest-host composite films based
on an azomethine dye in different matrix polymers. Opt Mater 2014;38:290–6.
[31] Bejan A, Marin L. Phenothiazine based nanocrystals with enhanced solid state
emission. J Mol Liq 2018;265:299–306.
[32] Bejan A, Ailincai D, Simionescu BC, Marin L. Chitosan hydrogelation with a
phenothiazine based aldehyde: a synthetic approach toward highly luminescent
biomaterials. Polym Chem 2018;9:2359–69.
9