Most of the current fluorescence sensing materials belong to the turn-off type, which make it hard to detect toxic substances such as benzene, toluene, and xylene (BTX) due to the lack of active chemical sites, thereby limiting their development and practical use. Herein, we show a guest-host mechanism stemming from the confined emitter's self-trapped exciton (STE) states or electron-phonon coupling to achieve turn-on fluorescence. We designed a luminescent guest@metal-organic framework (LG@MOF) composite material, termed perylene@MIL-68(In), and established its E-type excimeric emission properties in the solid state. Upon exposure to BTX, especially xylene, we show that the E-excimer readily converts into the Y-excimer due to nanoconfinement of the MOF structure. Such a transformation elevates the fluorescence intensity, thus realizing a turn-on type fluorescent sensor for detecting BTX solvents. Our results further demonstrate that controlling the STE states of perylene at room temperature (vs the previous report of <50 K) is possible via nanoscale confinement, paving the way to enabling turn-on type luminescent sensors for engineering practical applications.
Most of the current fluorescence sensing materials belong to the turn-off type, which make it hard to detect toxic substances such as benzene, toluene, and xylene (BTX) due to the lack of active chemical sites, thereby limiting their development and practical use. Herein, we show a guest-host mechanism stemming from the confined emitter's self-trapped exciton (STE) states or electron-phonon coupling to achieve turn-on fluorescence. We designed a luminescent guest@metal-organic framework (LG@MOF) composite material, termed perylene@MIL-68(In), and established its E-type excimeric emission properties in the solid state. Upon exposure to BTX, especially xylene, we show that the E-excimer readily converts into the Y-excimer due to nanoconfinement of the MOF structure. Such a transformation elevates the fluorescence intensity, thus realizing a turn-on type fluorescent sensor for detecting BTX solvents. Our results further demonstrate that controlling the STE states of perylene at room temperature (vs the previous report of <50 K) is possible via nanoscale confinement, paving the way to enabling turn-on type luminescent sensors for engineering practical applications.
Fluorescence sensing has recently
become a popular research direction because of its simplicity, portability,
rapid response, high selectivity, and high sensitivity.[1−4] It works mainly through weakening of the fluorescence intensity
(i.e., quenching mechanism) of the materials by affecting the photoinduced
electron transfer,[5] Förster resonance
energy transfer,[6,7] and/or charge transfer[8] to achieve “turn-off” type fluorescence
sensing.Although the turn-off type sensing by fluorescence
quenching has
its own advantages,[9] such as the ease of
practical implementations,[2] this approach
poses two major limitations which are worth addressing. The first
is that the side effects of environmental interference cannot be efficiently
ruled out. For example, molecules in the environment, such as water
moisture, may also cause a decrease in emission intensity.[10,11] The second limitation is that it is hard to produce a fluorescent
sensor to detect molecules that do not have active chemical sites,[12] such as “BTX”, namely, benzene,
toluene, and xylene. However, the capability for BTX sensing is urgently
needed because exposure to these kinds of substances presents the
risk of reproductive toxicity, and it is prevalent in both industrial
sectors and in our daily life.[13−17]To address the foregoing problems, herein, we demonstrate
a novel
“turn-on” type fluorescence sensing for the detection
of BTX molecules without active chemical sites; we accomplished this
by controlling the self-trapped exciton (STE) states of the emitters.
Theoretically, the analyte may affect the coupling effect between
excitons and phonons, thereby resulting in different fluorescent responses
that can be harnessed for sensor applications.In terms of the
STE, perylene is a good starting point to discuss.
This material is widely used in light-emitting diodes,[18,19] photovoltaics,[20] and organic field effect
transistors.[21] More importantly, perylene
possesses two different STE states (Y- and E-states) with different
fluorescence properties.[18,22−24] Nevertheless, it is hard to use perylene itself as a sensor. The
reasons are that perylene is susceptible to the aggregation-caused
quenching (ACQ) effect; it thus has no fluorescence in the solid-state
form, and the solubility of perylene is not high in many solvents,
making it challenging to form a dimeric structure or STE states.[25]Because of their porous, ordered, and
highly adjustable crystalline
structures, metal–organic frameworks (MOFs) are believed to
be one of the most promising materials to combine with perylene to
yield tunable luminescent sensing properties.[26−30] In principle, the pores/channels of the MOF “host”
can be used to encapsulate and isolate perylene “guest”
molecules to overcome the ACQ effect, and when confined within an
MOF structure, the solubility of perylene is no longer a concern.[31] Of note, the emerging concept of the confinement
of a luminescent guest (LG) in an MOF host, conferring an “LG@MOF”
composite system, has huge potential for designing and engineering
unconventional turn-on type luminescent sensors and lighting devices.[9]In this work, we demonstrate a fluorescent
sensing LG@MOF material
by encapsulating perylene into the easy-to-synthesize and stable Materials
of Institut Lavoisier-68(In) [MIL-68(In)]. For this perylene@MIL-68(In)
composite, we show that perylene exhibits the E-state excimer fluorescence
in the solid-state powder form, and it changes to the Y-state excimer
emission when exposed to BTX, resulting in a prominent turn-on type
sensing response. To the best of our knowledge, this kind of turn-on
sensing from the transformation of STE states is the first example
realized in the research field of LG@MOFs.[9] Significantly, it is also the first time the transformation of the
perylene E-state to the Y-state is evidenced under the room-temperature
condition.
Synthesis and Structure of the Perylene@MIL-68(In) Composite
System
The perylene@MIL-68(In) system was synthesized by
using the simple one-pot high-concentration reaction (HCR) method,
a facile approach first described by Chaudhari et al.[32,33] Full details of the synthetic steps for perylene@MIL-68(In) are
given in the Experimental Section. It is worth
mentioning that due to the deprotonation of triethylamine (TEA), this
is the first method to produce MIL-68(In) at room temperature. The
resulting composite materials were subsequently characterized by powder
X-ray diffraction (PXRD), attenuated total reflection Fourier transform
infrared spectroscopy (ATR–FTIR), synchrotron radiation infrared
(SR-IR) spectroscopy, nanoindentation, and scanning electron microscopy
(SEM).From Figure a, it can be seen that the PXRD patterns of the prepared MIL-68(In)
MOF host and perylene@MIL-68(In) are consistent with the simulations,
which indicate that MIL-68(In)’s crystal structure is successfully
generated by the HCR method, and the introduction of perylene molecules
does not hinder the crystal formation of the MOF host.[34] The same conclusion can also be drawn from the
FTIR results (Figure c) due to the high similarity between the spectra of MIL-68(In) and
perylene@MIL-68(In). The nanoindentation results show that the hardness
of the crystal increased by ∼20% after the encapsulation of
perylene (Figure S1 and Table S1), suggesting that the formation of perylene@MIL-68(In)
may have given rise to an interstitial hardening effect. SEM images
(Figures b and S2) show that the synthesized MIL-68(In) and
perylene@MIL-68(In) possess nanosized columnar crystals, which coincide
with the typical crystal morphology of MIL-68 produced with a long
reaction time (hours) at a high temperature (>55 °C).[35] Together, our results demonstrate that the HCR
method can produce MIL-68(In) immediately under a significantly milder
reaction condition, which is potentially useful for future commercialization.
Figure 1
(a) Comparison
of the XRD patterns of the simulated vs synthesized
MIL-68(In) and perylene@MIL-68(In) featuring three perylene concentrations
(1P, 05P, and 025P show that the amount of perylene used in the synthesis
is 1, 0.5, and 0.25 mmol, respectively). (b) Field-emission SEM (FESEM)
images of MIL-68(In), where the upper panel shows the rod-like nanocrystals
in the axial direction, while the lower panel shows the morphology
in the transverse direction. (c) FTIR results of MIL-68(In) and perylene@MIL-68(In)
(left: synchrotron-radiation-FTIR; right: ATR–FTIR).
(a) Comparison
of the XRD patterns of the simulated vs synthesized
MIL-68(In) and perylene@MIL-68(In) featuring three perylene concentrations
(1P, 05P, and 025P show that the amount of perylene used in the synthesis
is 1, 0.5, and 0.25 mmol, respectively). (b) Field-emission SEM (FESEM)
images of MIL-68(In), where the upper panel shows the rod-like nanocrystals
in the axial direction, while the lower panel shows the morphology
in the transverse direction. (c) FTIR results of MIL-68(In) and perylene@MIL-68(In)
(left: synchrotron-radiation-FTIR; right: ATR–FTIR).
Luminescent Properties of Perylene@MIL-68(In)
We then
performed photophysical characterization of perylene@MIL-68(In) with
different amounts of perylene guest loading. The excitation and emission
spectra are shown in Figure a,b, respectively. For control, a physical mixture of perylene
+ MIL-68(In) was also prepared; notably, this sample exhibits emission
(Figure S3) completely different from the
perylene@MIL-68(In) samples obtained by the HCR encapsulation method.
It is apparent that the fluorescence properties of the perylene@MIL-68(In)
systems are derived from perylene molecules because pure MIL-68(In)
is virtually nonemissive at the selected excitation wavelength of
440 nm. For perylene@MIL-68(In) at all three concentrations, their
emission peaks can be divided into two parts: (1) a small peak at
around 475 nm and (2) a broad and intense peak at around 600 nm. On
the basis of previous studies on perylene,[18,22−25] we reasoned that the first peak originates from the emission of
free excitons, and the second peak is attributed to the STE emission.
These photophysical characteristics indicate that the α-perylene-like
structure is generated inside the MIL-68(In) channel and emits as
the E-state excimer.[36] Furthermore, the
lifetime data of the system obtained by using the time-correlated
single photon counting (TCSPC) technique (Table S2 and Figure S4) show a lifetime
component (τ4) that is around 18 ns, further supporting
the existence of the E-state excimer. It is worth mentioning that
the fluorescence performance of perylene solutions with different
concentrations verified our theoretical hypothesis in the introduction
part: perylene is challenging to form a dimeric structure and induce
excimer emission due to the solubility problem, but MOFs can help
to overcome this limitation through confinement in nanoscale pores/channels.
Figure 2
Normalized
(a) excitation spectra (measured under em@600 nm) and
(b) emission spectra (measured under ex@440 nm) of perylene solutions,
MIL-68(In) powders, and perylene@MIL-68(In) in the solid state. Note:
MIL-68(In) is not normalized due to its weak signal.
Normalized
(a) excitation spectra (measured under em@600 nm) and
(b) emission spectra (measured under ex@440 nm) of perylene solutions,
MIL-68(In) powders, and perylene@MIL-68(In) in the solid state. Note:
MIL-68(In) is not normalized due to its weak signal.Considering the E-state emission characteristic,
the channel size
of MIL-68(In) (16 and 6 Å), and the previous research outcome
in ref (22), it is
reasonable to infer that in the channel of MIL-68(In), perylene molecules
exist in a disordered dimeric structure. This means that the longest
axis of perylene molecules tilts in the radial direction of MIL-68(In)
channels, as shown in Scheme . This kind of structural alignment of confined guests also
explains the incremental red shift evidenced for the 025P, 05P, and
1P@MIL-68(In) samples (Figure b) because the dimeric perylene can interact with the adjacent
perylenes, and the interaction will be enhanced with an increase of
perylene amount trapped within the MOF channel. The decrease in quantum
yields (QYs, Table S3) and excimer lifetime
(τ4, Table S2 and Figure S4) when the perylene amount increases
also supports the proposed structure as the enhanced interactions
will increase the non-radiative decay.[26,30]
Scheme 1
(a,b) LG@MOF
Crystal Structure of Perylene@MIL-68(In) and Its Initial
Molecular Configuration under Confinement of One-Dimensional Channels
and (c) Proposed Sensing Mechanism of Perylene@MIL-68(In) Subject
to BTX Molecules. Color Scheme: Indium in Dark Red, Carbon in Gray,
and Oxygen in Red
BTX Sensing Performance of Perylene@MIL-68(In)
Subsequently,
we tested the BTX sensing properties of the perylene@MIL-68(In) system.
In order to better observe the subtle changes, we chose 025P@MIL-68(In)
with the least amount of perylene and the highest QY for testing.
As shown in Figure , perylene@MIL-68(In) delivers a prominent turn-on sensing response
with a slight blue shift and splitting of the emission peak (∼600
nm) when exposed to the BTX molecules in solution. On the contrary,
when exposed to DMF/acetone, polar aprotic electron-deficient solvents,
the peak at 600 nm exhibits a red shift with a decline in its intensity.
Figure 3
(a) Excitation
(measured under em@600 nm) and (b) emission spectra
(measured under ex@440 nm) of 025P@MIL-68(In) in different solvents.
The inset of (b) shows the emission spectra of 025P@MIL-68(In) in p-, o-, and m-xylene.
(c) Lifetimes (excimer component) and (d) QYs of 025P@MIL-68(In) powders
and when subject to BTX (the QY of each sample was tested three times
to determine the average and standard deviations). (e) Normalized
excitation (measured under em@600 nm) and (f) emission spectra (measured
under ex@380 nm) of 025P@MIL-68(In) and its pellets subjected to two
different nominal pressures.
(a) Excitation
(measured under em@600 nm) and (b) emission spectra
(measured under ex@440 nm) of 025P@MIL-68(In) in different solvents.
The inset of (b) shows the emission spectra of 025P@MIL-68(In) in p-, o-, and m-xylene.
(c) Lifetimes (excimer component) and (d) QYs of 025P@MIL-68(In) powders
and when subject to BTX (the QY of each sample was tested three times
to determine the average and standard deviations). (e) Normalized
excitation (measured under em@600 nm) and (f) emission spectra (measured
under ex@380 nm) of 025P@MIL-68(In) and its pellets subjected to two
different nominal pressures.The rising intensity and splitting of the peak
at ∼600 nm
(Figure b) can be
explained by the perylene in the MIL-68(In) channels undergoing a
transformation from the E-state to Y-state.[24] This kind of transformation can be attributed to the formation of
the π–π interaction bridge shown in Scheme . When the perylene inside
encounters BTX, electron-rich solvents, the organic linker of MIL-68(In)
benzenedicarboxylate (BDC), BTX, and dimeric perylene form a relatively
strong π–π interaction “chain”, thereby
inducing an effect similar to the lattice confinement effect.[24,26,37,38] In this situation, the electron–phonon coupling of the E-state
will be affected to generate the Y-state. Compared with the single
emission characteristic of the E-state, Y-emission itself possesses
multiple luminescent peaks, a higher peak intensity, and a shorter
peak wavelength.[24] Therefore, when exposed
to BTX, the perylene@MIL-68(In) peak showed an intensity enhancement,
splitting, and blue shift. The deconvolution result of the emission
peak (Figure S5) reveals the typical feature
of the Y-excimer, in which the fitted peaks denoted as 2, 3, and 4
may correspond to the emission of the Y-excimer reported in ref (24).Furthermore, based
on this, the occurrence of a higher peak intensity
and more pronounced peak splitting upon contact with xylene (compared
to that with benzene and toluene) can also be understood. As the electron-donating
ability of BTX becomes stronger (i.e., benzene < toluene < xylene),
the π–π interaction bridge will become stronger
correspondingly, resulting in a more substantial confinement effect
for perylene in the channel.[39] In addition,
xylene was then selected to test the sensitivity, selectivity, and
reversibility of the system (see Figure S6, including the limit of detection value). It is worth mentioning
that the Y-emission can only be observed at temperatures below 50
K for pure perylene.[24] Remarkably, herein,
we show for the first time that, thanks to the LG@MOF assembly, the
transformation between the E-state and Y-state is realized at room
temperature.To further validate the proposed theory, we performed
the lifetime
and QY measurements on perylene@MIL-68(In) (Figure c,d). The lifetime data shown in Figure c, Table S4, and Figure S7 suggest
that the 18 ns time component (belonging to the E-state) increases
when contacting BTX solvents.[25] Meanwhile,
the QY also shows an increasing trend. More importantly, this rise
can be associated with BTX’s electron-donating ability (Figure b–d): the
higher the electron-donating ability, the larger the lifetime and
QY. These phenomena indicate that the non-radiative decay of perylene
in MIL-68(In) is reduced due to the π–π interaction
bridge, thus inducing a strong confinement effect and proving the
transformation from the E- to Y-emissions.The fluorescence
performance of perylene@MIL-68(In) subject to
different compressive pressures also supports our theory. According
to the previous studies,[26,38,39] it can be deduced that when the pressure increases, the π–π
interaction between the linker (BDC) and perylene will increase, thus
causing perylene to become more tightly constrained. The excitation
and emission spectra of perylene@MIL-68(In) under pressure (Figure e,f) illustrate that
when the interaction is enlarged, the peak intensity corresponding
to the perylene excimer increases. Furthermore, the τ4 excimer lifetime also increases sharply when pressure rises (Table S5 and Figure S8), compared with that of the τ1 to τ3 lifetime components which are relatively unchanged. Therefore, it
is shown that enhancing the π–π interaction in
this LG@MOF system does enhance the confinement effect and hence results
in a stronger excimer emission. The pressure-dependent luminescence
data further substantiate the rational of our proposed theory from
another standpoint.By way of comparison, the behavior of perylene@MIL-68(In)
when
exposed to electron-deficient solvents can be explained by the proposed
theory as well. When the system is in DMF/acetone solvents, the emission
at ∼600 nm shows a relative decline and red shift (Figure b). In Table S4 and Figure S7, the perylene@MIL-68(In) lifetime is smaller in DMF and acetone
than in BTX. Based on the theory, it can be interpreted that electron-deficient
solvents reduce the π–π interaction between the
BDC linkers and perylene, causing the decline in intensity and redshift
of the emission peak.
Conclusions
In conclusion, we have shown the facile
synthesis of perylene@MIL-68(In)
under ambient conditions by harnessing the HCR method. Of note, perylene@MIL-68(In)
exhibits the E-excimer emission characteristics in the solid state.
When exposed to BTX, the perylene molecules present in the MIL-68(In)
channels will receive a strong confinement effect and affect the STE
states. This guest–host confinement effect switches the E-state
to Y-state emission, resulting in the turn-on fluorescent response
when subject to the electron-rich BTX solvents. The sensing mechanism
proposed here using the transformation between different STE states
is the first exemplar in the field of LG@MOF research.[9] The simple synthesis method and the uncommon turn-ontype
sensing behavior have opened up a new approach for developing highly
selective fluorescent sensors.
Experimental Section
Synthesis of Perylene@MIL-68(In) and MIL-68(In) by Using the
HCR Method
The synthesis was accomplished by leveraging the
HCR method.[32,33] 15 mL of a dimethylformamide
(DMF) solution of 4.8 mmol BDC and TEA (9.6 mmol) was combined with
50 mL of a dichloromethane (DCM) solution of 0.25/0.5/1 mmol perylene.
After the combination, 15 mL of a DMF solution of 4.8 mmol indium
nitrate was immediately added into the mixture. Then, the product
was formed instantly and washed thoroughly five times (two times with
DCM, two times with DMF, and one time with methanol) to remove the
excess guests adhered to the external MOF surfaces. The nanocrystals
of perylene@MIL-68(In) were separated from the suspension by centrifugation
at 8000 rpm for 10 min.
Sample Preparation for Fluorescence Characterization
The pellets for the mechanofluorochromic study were made using a
manual hydraulic press (Specac Atlas) with a 1.2 cm diameter die under
a uniaxial compressive force of 4 and 8 tones. Perylene@MIL-68(In)
suspensions for the solvatochromic study were prepared by diluting
5 mg of perylene@MIL-68(In) in 8 mL of the solvent. For sensitivity
and selectivity research, 5 mg of perylene@MIL-68(In) was first dispersed
in 8 mL of cyclohexane, and then, 2 mL of the solution was used as
the base solution for the subsequent measurements.
Materials Characterization
The crystal morphologies
and structures were examined using a field-emission scanning electron
microscope (FESEM LYRA3 GM TESCAN). PXRD patterns were
recorded using a Rigaku MiniFlex with a Cu Kα source (1.541
Å). The nanoindentation tests were conducted using an iMicro
nanoindenter (KLA-Tencor). Steady-state fluorescence spectra, lifetimes,
and QYs were recorded using the FS-5 spectrofluorometer (Edinburgh
Instruments). For TCSPC lifetime measurement, a 445 nm laser was used;
the stop condition was set to be at 10,000 counts, and a filter was
used to avoid the interference of scattering light. ATR–FTIR
spectra were recorded using a Nicolet iS10 FTIR spectrometer.
SR-IR Spectroscopy
High-resolution SR-IR vibrational
spectra of all compounds were recorded at the multimode IR imaging
and microspectroscopy (MIRIAM) Beamline B22 at the Diamond Light Source
synchrotron. Measurements were performed in vacuum via a Bruker Vertex 80v FTIR spectrometer with an ATR accessory (Bruker
Optics, Germany). For the far-IR spectral range below 700 cm–1, a bolometer cooled by liquid helium was used for the detection
of terahertz signals. All spectra were acquired with a resolution
of 4 cm–1 and a scanner velocity of 20 kHz.
Authors: Brooks A Jones; Michael J Ahrens; Myung-Han Yoon; Antonio Facchetti; Tobin J Marks; Michael R Wasielewski Journal: Angew Chem Int Ed Engl Date: 2004-11-26 Impact factor: 15.336