Angewandte
Chemie
with a typical slip-packing structure in which molecules avoid
excimer formation thereby enhancing the fluorescence of
solid samples. Introducing substituents such as methyl and
methoxyl at the para-position relative to the carbonyl group
does not greatly change the molecular conformation and
packing mode. However, these functional groups form non-
classical H-bonds which slightly alter the molecular micro-
environments. Thus, the shape and fluorescence of the bulk
crystals are affected in a certain degree.
isopropyl group is used in place of dimethylamino group in
compound 7 to disclose the effect of strong D-A structure on
the emission. The emission bands of both 6 and 7 in the solid
state display a large blueshift of more than 80 nm (Figure S9)
compared with 1. This finding strongly demonstrates that the
intramolecular H-bond and the D-A structure are crucially
important for the deep red/NIR emission of crystals 1–3.
Because all those compounds have the same chalcone
skeleton except for some partial changes, the distinctively
different emission behaviors between 1–3 and 4–7 demon-
strate that the structure modification is of great importance
for organic fluorophores. This also provides an easy but
effective strategy when designing new organic fluorescent
materials with high efficiency in the solid state.
Crystals 1–3 exhibit brighter emission on edge than on
their body, indicating that their emissions are highly confined
inside the crystals. These observations reflect that self-wave-
guided emission occurs in the crystals and thus these crystals
are possible candidates as organic deep red/NIR laser media.
To test this possibility, one isolated single crystal of 1 is excited
with a pulsed laser and the photoluminescence (PL) spectra
are subsequently collected from the edge area. At low
excitation power, the PL spectrum is featured as a broad
emission band peaking at about 714 nm, and an amplified
spontaneous emission is observed above a certain power
threshold. Figure 4a shows the power-dependent PL spectra
as well as the energy dependence of the luminescence
intensity and FWHM of the emissions. It shows a nonlinear
gain and a threshold characteristic of ASE. The threshold of
9.2 kWcmÀ2 pulseÀ1 calculated from the slope of peak inten-
sity versus pump energy curve is among the lowest values for
typical organic crystals. Thus, ASE of organic NIR-fluores-
cent organic crystals has been successfully realized, most
importantly, having the extremely low threshold value which
is a crucial parameter to evaluate the ASE property.
The ASE characteristics of the other two crystals 2 and 3
have also been checked. The thresholds of crystals 2 and 3 are
8.2 and 100 kWcmÀ2 pulseÀ1, respectively. Crystal 3 displays
unfavorable ASE characteristics and its PL spectra start to
narrow at high excitation laser powers. This probably
originates from its larger thickness compared with crystals
1 and 2 which should inevitably result in more defects in the
crystal, adverse to the self-waveguided emission. To verify
this, microcrystals with smaller thickness of compound 3
growing on a silicon substrate are prepared (Figure S10). The
laser beam radiates on a pile of microcrystals and ASE is
measured. The threshold of the microcrystals is
22 kWcmÀ2 pulseÀ1, smaller than that of the normal-sized
crystal 3. In addition, the gain coefficients of crystal 1 have
also been carried out by a successively variable pump stripe
method in which the lengths of pump stripe are adjusted by
a slit (Figure S11). The PL spectra become narrower upon
increasing the pump stripe length, and this narrowing occurs
more rapidly at higher pump energy, consistent with the
prediction of ASE theory.[14] The polarized emission of crystal
1 is detected by using a polarizer in front of the optical fiber at
the pump energy of 127 kWcmÀ2 pulseÀ1. A scatterplot is
shown in Figure S12, depicting the PL intensity in change of
the relative angles between the crystal and the polarizer. Its
The solutions under frozen condition and films of these
compounds doped in polymethylmethacrylate (1–10 wt%)
emit very weak fluorescence (Figure S8), demonstrating that
the restriction of intramolecular rotation is likely not the key
role to activate the bright light of the crystals. To further gain
insights into the reason why compounds 1–3 are not
fluorescent in dispersed states but brightly emissive in
crystals, we synthesized compound 4[13a] and obtained red
slab-like crystals whose color and shape are quite similar to
those of crystals 1–3 (Figure 1). Crystals of 4 have similar
edge-to-face packing structures with 1–3 (Figure 3c,d). How-
ever, the individual molecule of 4 takes a rather bent and
twisted molecular conformation. This crystal is weakly
fluorescent (lem = 720 nm; Ff < 0.01) although there are no
negative factors for fluorescence such as intermolecular p-
stacking and dipole–dipole interactions in the crystal packing
structure. This finding strongly indicates the crucial effect of
molecular planarization on the fluorescence quantum yield
for crystals 1–3. Compound 5[13b] is next synthesized to
demonstrate the vital effect of slip-packing mode and edge-
to-face arrangement structure on the high efficiency of
crystals 1–3 (Figure 1). A similar red slab-like crystal of 5 is
achieved. Nevertheless, the crystal is nearly non-emissive
(lem = 699 nm; Ff < 0.01). The molecules in crystal 5 have
a relatively planar conformation (Figure 3e), similar with
those in crystals 1–3. The difference of crystal structures
between 5 and 1–3 is the molecular packing structure. As
shown in Figure 3 f, unlike the edge-to-face packing mode of
molecules in crystals 1–3, every two molecules in crystal 5
take a face-to-face packing structure with strong p–p
interactions; on the other hand, each molecule has strong
dipole–dipole interactions with its neighboring molecule,
which undoubtedly has an adverse effect on the quantum
yield. The different packing structures make crystal 5 nearly
non-emissive, whereas crystals 1–3 are highly fluorescent.
Meanwhile, from the comparison of crystal structures among
crystals 4, 5, and 1–3, we can draw the conclusion that the
molecular conformation and molecular packing structure of
these molecules can be tuned by the position of the
substituent. Considering the much similar molecular struc-
tures and the greatly different emission behaviors of these
compounds, it is easy to find that the substituent modification
has great importance on tuning the solid-state emissions of
organic materials.
For a further understanding of the deep red/NIR emission
of crystals 1–3, 2’-methoxychalcone derivative 6[13c] and
isopropyl-substituted 7[13d] are synthesized (Figure 1). In 6,
the hydroxy group is replaced by a methoxy group. Con-
sequently, there is no intramolecular H-bond in compound 6.
In other words, the ESIPT process cannot occur. The
Angew. Chem. Int. Ed. 2015, 54, 8369 –8373
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8371