Dual-fluorescence carbon dots have great potential as nanosensors in life and materials sciences. Such carbon dots can be obtained via a solvothermal synthesis route with glutathione and formamide. In this work, we show that the dual-fluorescence emission of the synthesis products does not originate from a single carbon dot emitter, but rather from a mixture of physically separate compounds. We characterized the synthesis products with UV-vis, Raman, infrared, and fluorescence spectroscopy, and identified blue-emissive carbon dots and red-emissive porphyrin. We demonstrate an easy way to separate the two compounds without the need for time-consuming dialysis. Understanding the nature of the system, we can now steer the synthesis toward the desired product, which paves the way for a cheap and environmentally friendly synthesis route toward carbon dots, water-soluble porphyrin, and mixed systems.
Dual-fluorescence carbon dots have great potential as nanosensors in life and materials sciences. Such carbon dots can be obtained via a solvothermal synthesis route with glutathione and formamide. In this work, we show that the dual-fluorescence emission of the synthesis products does not originate from a single carbon dot emitter, but rather from a mixture of physically separate compounds. We characterized the synthesis products with UV-vis, Raman, infrared, and fluorescence spectroscopy, and identified blue-emissive carbon dots and red-emissive porphyrin. We demonstrate an easy way to separate the two compounds without the need for time-consuming dialysis. Understanding the nature of the system, we can now steer the synthesis toward the desired product, which paves the way for a cheap and environmentally friendly synthesis route toward carbon dots, water-soluble porphyrin, and mixed systems.
Carbon dots (CDs) are
fluorescent carbon-based nanoparticles with
tunable and functional photoluminescence.[1,2] Since
their serendipitous discovery during carbon nanotube purification
in 2004,[3] they have gained significant
interest for application as a new generation light source,[4−6] solution-based chemo-sensors for detecting chemicals and metal cations,[2,7−11] and nanoprobes for mapping biological systems.[2,7−11] Moreover, they have shown great potential to be applied directly
or in hybrid materials for photocatalytic and electrocatalytic applications.[12−16] CDs are generally regarded as environmentally friendly, nontoxic
material because the synthesis precursors are relatively abundant,
harmless, and environmentally friendly and the synthesis does not
involve extreme reaction conditions nor produces harmful waste.[17−21] This motivates their use outside the academic research lab in industrial
and consumer applications. An example being their use as sensors for
the detection of dangerous chemicals, for the detection of reaction
products in high-throughput catalyst testing, or even to map the heterogeneity
in the chemical composition of porous materials. All these applications
revolve around the control over their tunable[22] and functional photoluminescence (PL) properties.A thorough
understanding of CD synthesis and its effect on the
PL properties are essential to obtain the desired sensor behavior.
Briefly, CDs have been synthesized via both top-down and bottom-up
approaches. In the top-down synthesis approach, larger carbon structures,
such as graphite, are broken down into smaller CDs, while in the bottom-up
synthesis approach, small carbonaceous precursors are heated to obtain
CDs through carbonization and graphitization.[23] CDs are on the order of ten nanometers in size and consist of both
amorphous and crystalline carbon structures with sp2–sp3 hybridization. The carbon structures contain primarily oxygen
and nitrogen heteroatoms and the exact composition is heavily dependent
on the synthesis precursor and synthesis conditions.[23]The PL properties of CDs are of paramount importance
for their
application. Four origins of PL emission in CDs have been identified:
the carbon core state luminescence from a conjugated π-system
with quantum confinement, the surface state luminescence from interaction
between the core state and surface groups, embedded fluorophores in
the carbon matrix, and the cross-link enhanced emission effect.[24−26] It has traditionally been reported that the core and surface states
are the origin of PL emission in CDs. However, recent studies have
demonstrated that embedded fluorophores are responsible for PL emission
with high quantum efficiency. These embedded fluorophores are commonly
found in CDs prepared by bottom-up approaches.[27] In some systems, the interaction of surface states with
the CD’s local environment results in a measurable response
in the fluorescence intensity and/or wavelength. Such sensing capacities
are especially powerful when present for one of the two bands of dual-fluorescence
CDs. These nanoparticles have two distinct PL emission bands that
both originate from the same CD, which allows the extraction of sensory
information from one band as well as a local CD concentration from
the other. Such a ratiometric approach is advantageous, because the
acquired sensory data is independent of the CD concentration. Macairan
et al. and Wang et al. recently reported a procedure to synthesize
dual-fluorescence CDs based on glutathione (GSH) solved in formamide
(FA).[7,28,29] When excited
at 405 nm, they emit a broad blue PL emission band with a maximum
emission around 460–490 nm and a narrow red emission band at
650–680 nm. The system was reported to be suitable for mapping
cells and intracellular parameters of the environment, such as the
local pH, and consequently, it may be suitable for mapping of heterogeneous
catalysts and other functional porous materials.To correctly
use the GSH/FA CD system as a nanoprobe, we need to
know the mechanism behind the blue and red PL emission to understand
their response to the environment. In this work, we show that the
blue- and red-emitting species are two physically separable compounds,
and we reveal their chemical identity. Moreover, we demonstrate an
easy and environmentally friendly separation method to purify the
obtained synthesis products. After the components of the GSH/FA CD
system have been identified, we demonstrate control over the synthesis
via acid and base catalysis, which we employ to obtain the desired
products in increased yield. Altogether, this paves the way for a
successful use of this CD system as a nanosensor in porous solids
and other sensing applications.
Methods
Synthesis and
Dialysis
The carbon dots (CDs) were synthesized
following an adaptation from ref (28). Reduced L- glutathione (GSH) was dissolved
in a formamide (FA) at room temperature. The mixture was poured in
a Teflon-lined autoclave (Parr 4749) and was heated in solvothermal
conditions to a temperature of 180 °C for 3 or 18 h. The obtained
solution was filtered through a 0.22 μm filter and centrifuged
at 5000 rpm for 5 min to remove large aggregates. The supernatant
solution was dialyzed using a 3.5 kDa dialyzing membrane (D-Tube Dialyzer
Maxi, MWCO 3.5 kDa, Merck) for 1 week, while the water was replaced
daily. The progress of the separation was followed via the luminescence
of the dialysate excited with a 405 nm laser pointer (5 mW). Approximately
7 days of dialysis turned out to be sufficient to reduce the concentration
of luminescent species in the dialysate below a level observable by
the eye. Additional samples with other solvents (H2O and N,N-dimethylformamide) or in acidic and
basic environments have been prepared (Table S1).
Purification by Kaolinite
Red-emissive products were
separated from blue-emissive ones via purification with kaolinite.
Ten milliliters synthesis product solution (before dialysis) was mixed
with 1 g kaolinite and 0.1 mL HCl (37%). The mixture was left for
10 min, and the kaolinite particles were sedimented by centrifugation
at 5000 rpm for 5 min. The supernatant containing the blue-emissive
product was taken. To remove blue-emissive product impurities from
the sediment, it was washed with 20 mL methanol and separated from
kaolinite by centrifugation (5000 rpm, 5 min). The methanol washing
step was repeated once more. To release the red-emissive products
from the sediment, it was dispersed in a solution of 10 mL of 7 M
ammonia in methanol for 10 min before the kaolinite particles were
centrifuged down (5000 rpm, 5 min) and the supernatant containing
the red-emissive product was taken. Finally, this solution was dried
at room temperature under N2 flow (20 mL/min), and a dark
green powder was obtained and stored in the freezer. The obtained
powder can be easily solved in water for further use.
Characterization
UV–vis spectra were recorded
using a UV-Cary 200 spectrophotometer. Photoluminescence emission
spectra in Figure b were recorded with a Jasco spectrofluorometer (FP 8300) at 200
nm/min. The photoluminescence spectra in all other figures were recorded
using AvaSpec-ULS2048CL-EVO as a spectrophotometer and AvaLight-HPLED-405
as a light source with a wavelength of 405 nm. All emission spectra
were recorded at 405 nm excitation wavelength unless otherwise noted.
Fourier-transform infrared (IR) spectra have been measured in attenuated
total reflectance mode by a PerkinElmer 2000 instrument on a droplet
of CDs with a concentration of 1 g/L. X-ray diffractograms (XRD) have
been recorded by a D2 Bruker diffractometer in the 2θ range
of 10–40° using a Co Kα source and slit size of
1 mm. For XRD measurements, 1 mL aqueous solution of CDs with a concentration
of 1 g/L was dried on a glass substrate, which was repeated until
a layer had formed. The same dried samples were used for Raman spectroscopy
measurements. Raman spectra were acquired using a Horiba XPlora microscope
equipped with 638 and 532 nm lasers. The power at the sample was 0.22
mW for the 638 nm laser and 0.53 mW for the 532 nm laser. All measurements
were done with a 1200 l/mm grating and an exposure time of 100 s.
Because porphyrin absorbs light at 638 nm, the signal from the porphyrin
is enhanced as a result of resonance Raman scattering. For atomic
force microscopy (AFM) measurements, a droplet of diluted solution
(C ≈ 10 mg/L) of CDs was dried on a circular glass coverslip
and the measurements were performed on a Bruker Multimode 8 in peakforce/Scanasyst
mode with ScanAsyst-air probes. The scanned area was about 2 ×
2 μm for the recorded AFM images. The NanoScope Analysis V1.80
software was used to construct a topography map from the AFM images.
Transmission Electron Microscopy (TEM) images were acquired on a FEI
Talos L120C electron microscope. Five microliters of as-prepared sample
solution drop-casted on a copper Formvar grid and were left to dry
in the air. The optical images of dried sample droplet on glass coverslip
were measured using the ZEISS AXIO Zoom.V16 microscope using 2.3×
objective.
Figure 1
(a) Schematic of the solvothermal synthesis of carbon dots and
porphyrin from glutathione in formamide. After peptide hydrolysis
and cyclization, small molecular fluorophores are formed in solution.
These fluorophores react further to carbon dots and porphyrin. X represents
polar functional groups such as COOH, OH, SH, phenol, and glutamate
and YH is S or NH. Please note that the structures of the fluorophores
are examples, and many more variations are possible. (b) Photoluminescence
(PL) emission spectra as a function of the excitation wavelength after
3 h of solvothermal synthesis and dialysis of the product with a 3.5
kDa membrane (GSH-FA-3h). The carbon dots have an excitation-dependent
emission (400–600 nm) and the porphyrin has an excitation-independent
emission in the red (600–750 nm). (c) Thin-layer chromatography
of the GSH-FA-3h sample using 50/50 (v/v) water/THF mixture as the
mobile phase. Three bands with separable compounds are observed (top),
which were identified as carbon dots with blue emission and porphyrin
with red emission. (d) PL emission of GSH-FA-3h dissolved dried powder
in different solvents excited with a 405 nm laser pointer. In water,
the emission is red, indicating that the porphyrin is water-soluble
and hydrophilic, while the emission is blue in THF, demonstrating
amphiphilic nature of the carbon dots. (e) Raman spectrum of the GSH-FA-3h
sample recorded with 538 and 638 nm excitation wavelength is similar
to that of hematoporphyrin. (f) Infrared transmission spectrum of
the GSH-FA-3h sample indicates the formation of conjugated carbon
structures with oxygen and nitrogen functional groups.
(a) Schematic of the solvothermal synthesis of carbon dots and
porphyrin from glutathione in formamide. After peptide hydrolysis
and cyclization, small molecular fluorophores are formed in solution.
These fluorophores react further to carbon dots and porphyrin. X represents
polar functional groups such as COOH, OH, SH, phenol, and glutamate
and YH is S or NH. Please note that the structures of the fluorophores
are examples, and many more variations are possible. (b) Photoluminescence
(PL) emission spectra as a function of the excitation wavelength after
3 h of solvothermal synthesis and dialysis of the product with a 3.5
kDa membrane (GSH-FA-3h). The carbon dots have an excitation-dependent
emission (400–600 nm) and the porphyrin has an excitation-independent
emission in the red (600–750 nm). (c) Thin-layer chromatography
of the GSH-FA-3h sample using 50/50 (v/v) water/THF mixture as the
mobile phase. Three bands with separable compounds are observed (top),
which were identified as carbon dots with blue emission and porphyrin
with red emission. (d) PL emission of GSH-FA-3h dissolved dried powder
in different solvents excited with a 405 nm laser pointer. In water,
the emission is red, indicating that the porphyrin is water-soluble
and hydrophilic, while the emission is blue in THF, demonstrating
amphiphilic nature of the carbon dots. (e) Raman spectrum of the GSH-FA-3h
sample recorded with 538 and 638 nm excitation wavelength is similar
to that of hematoporphyrin. (f) Infrared transmission spectrum of
the GSH-FA-3h sample indicates the formation of conjugated carbon
structures with oxygen and nitrogen functional groups.
Results and Discussion
Solvothermal heating of GSH
in FA results in the generation of
fluorescent synthesis products. GSH is a peptide comprised of three
amino acids that hydrolyses and cyclizes at elevated temperatures
forming fluorescent reaction intermediates (Figure a). These intermediates were observed in
the dialysate through a 3.5 kDa membrane, confirming their size in
the range of small molecules (Figure S1a). These molecular fluorophores can be detected during the initial
steps of heating (T = 160 °C, Figure S1b). Over time, these species react further to the
CD products. We have observed that the synthesis products after dialysis
have two distinct emission bands, which were ascribed by Macairan
et al. to dual emissive CDs (Figure b).[28] These authors have
attributed the 400–600 nm PL emission to a core state and the
600–750 nm emission to a molecular surface state but did not
find evidence for energy transfer between the states. Our findings
show that these emissive states originate from a blue-emitting CD
and a red-emitting molecular fluorophore called porphyrin (400 nm
excitation, Figure a–b), revealing that the two states are not just physically
distinct but rather physically separated.We have observed that
the blue PL emission is excitation-dependent,
while the red emission is excitation-independent (Figure b). Excitation-dependent emission
is typical for CDs, which shows that the blue-emitting compounds are
CDs. The origin of excitation-dependent emission is a matter of conflict
in literature, but it is generally accepted that the behavior can
be explained by a heterogeneous composition of fluorescent centers
or complex energy levels.[25,30−32] Here, only the subpopulation of excited states that are efficiently
populated at the excitation wavelength contribute to the PL emission,
which results in a shift of the emission with excitation wavelength.
This mechanism has been attributed to embedded molecular fluorophores
in the CDs, aggregation of CDs, and the presence of different types
of CDs.[25,30−32] It was found that the
red emission is excitation independent, which indicates that it is
related to a specific molecular fluorophore. This fluorophore is not
removed during dialysis with a 3.5 kDa membrane (Figure S1a). Therefore, it can be concluded that the red-emissive
molecule either is embedded in the CDs or is a macromolecule. To address
this question, we performed thin-layer chromatography (TLC) using
a 50/50 (v/v) water/THF mixture as a mobile phase. The separated analytes
are visible as a dark band upon excitation of the TLC plate with 254
nm ultraviolet (UV) light (Figure c). We used 365 nm excitation to record the PL color
and identify the separated species. We find three bands: the first
broad band is from analytes running with the THF mobile phase and
contains mainly blue-emitting CDs; the other two narrow bands originate
from the water mobile phase and contain blue-emitting CDs and red-emitting
macromolecules. TLC shows that red emissive products are not embedded
in the CDs as they can be separated from the blue-emitting species
and must be a macromolecule. The CDs are amphiphilic and dissolve
in both water and THF, with a larger affinity for the THF phase, while
the macromolecule is hydrophilic and only dissolves well in the water
mobile phase. Indeed, solvent extraction with water and THF results
in the expected separation (Figure d). Red emission was observed in water, while blue
emission was observed in THF when excited with 405 nm laser light.
The presence of sediments confirmed the poor solubility of the red-emitting
macromolecule in THF.We identified the macromolecule as a water-soluble
porphyrin using
UV–vis, Raman, and infrared (IR) spectroscopy. These synthesis
products cannot be identified from the X-ray diffraction pattern since
they tend to form an amorphous layer upon drying (Figure S2). However, the product obtained has clear Soret
and Q bands in the UV–vis absorption spectrum, which are characteristic
for porphyrin molecules. Moreover, the recorded Raman spectra (Figure e) are to a large
extent similar to the spectra of hematoporphyrin complexes, and their
band assignment can be found in Berjot et al.[33] We found many narrow bands, which indicate the presence of a molecular
species. We did not observe the D and G bands in the Raman spectrum,
which are related to the degree of graphitization in CDs, and the
features of porphyrin dominated.[28,34] The narrow
band at 1600 cm–1 could be falsely assigned to the
carbon G band but is actually associated with the intensity of the
bands around 600–800 cm–1, which are due
to ring deformations of the porphyrin. Finally, IR spectroscopy confirmed
the formation of conjugated carbon structures with oxygen and nitrogen
functional groups.[35−37] The assigned functional groups are listed in Figure f and Table S2. The most characteristic functional
groups are the N–H at 3320 cm–1, C=N
or C=O in amide around 1673 cm–1, and C=C
in the aromatic ring at 1591 cm–1.[35−37] On top of that, the IR spectra have high similarity to water-soluble
porphyrin molecules and building blocks (e.g., proline).[38] Hence, all these spectroscopic features are
very similar to CDs synthesized from natural carbon sources rich in
chlorophyll-based porphyrin, which further supports our assignment.[39,40]Next, we have developed an alternative purification method
to dialysis
by using kaolinite as a sorbent, which is a faster and more efficient
way to separate porphyrin from CDs and most of the small molecules.
Separation facilitates detailed analysis of the individual components
as well as the use in applications requiring either CDs or porphyrin.
Porphyrin molecules, especially the water-soluble derivatives, have
a high affinity to be adsorbed in the layered structure of kaolinite.[41] For separation, we stirred the synthesis product
mixture with kaolinite in acidic conditions, which promotes the sorption
of porphyrin (Figure a). Then, we sedimented the kaolinite particles by centrifugation:
the porphyrin partitioned in the kaolinite sediment, while the CDs
remained in solution. This allowed us to harvest the CDs from the
supernatant solution. We then removed the remaining blue-emitting
impurities from the kaolinite by washing it two times with methanol.
Finally, we released the porphyrin from the sediment by addition of
a basic methanol solution. A dark green powder was obtained after
drying under nitrogen flow. The last step can also be performed with
an aqueous solution; however, we used methanol as the main solvent,
since it is easier to dry the obtained solution.
Figure 2
(a) Schematic representation
of the kaolinite purification method.
The synthesis reaction mixture containing carbon dots, porphyrin,
and residual molecular fluorophores (1) were separated by adsorption
of the porphyrin and residual fluorophores on the kaolinite (GSH-FA-3h).
The supernatant containing the carbon dots (2) was taken and the porphyrin
and residual fluorophores were released in basic ammonia solution
(3). (b–c) The normalized photoluminescence (PL) emission spectra
prior to (b) and after (c) purification. The numerals (1–3)
indicate the respective solution as indicated in (a). (d–e)
UV–vis absorption spectra prior to (b) and after (c) purification.
The numerals (1–3) indicate the respective solution as indicated
in (a). The Soret and Q-band(s) of the porphyrin are indicated after
purification.
(a) Schematic representation
of the kaolinite purification method.
The synthesis reaction mixture containing carbon dots, porphyrin,
and residual molecular fluorophores (1) were separated by adsorption
of the porphyrin and residual fluorophores on the kaolinite (GSH-FA-3h).
The supernatant containing the carbon dots (2) was taken and the porphyrin
and residual fluorophores were released in basic ammonia solution
(3). (b–c) The normalized photoluminescence (PL) emission spectra
prior to (b) and after (c) purification. The numerals (1–3)
indicate the respective solution as indicated in (a). (d–e)
UV–vis absorption spectra prior to (b) and after (c) purification.
The numerals (1–3) indicate the respective solution as indicated
in (a). The Soret and Q-band(s) of the porphyrin are indicated after
purification.We could demonstrate the quality
of separation of the kaolinite
purification method via fluorescence and UV–vis spectroscopy
(Figure b–e).
The PL spectral features we ascribed to CDs and porphyrin were well
separated after kaolinite purification (Figure c). However, we find low intensity PL around
400–600 nm in the porphyrin fraction, which disappears after
further purification with THF, indicating that this signal stems from
residual molecular fluorophores. In the UV–vis spectrum of
the porphyrin fraction, absorption bands at 400 nm and a series of
absorption bands around 620 nm are present (Figure e). The band at 400 nm is the well-known
Soret band, which is insensitive to the functional groups of the porphyrin
and is absent only when porphyrin macrocyclic conjugation is disrupted.[42] The series of bands around 620 nm are the Q-band
absorptions. They are often observed between 500–650 nm but
usually below 600 nm. A change in the functional groups of porphyrin,
e.g., due to a change in pH, can cause shifts in the Q-bands, while
the Soret band remains unaffected.[42] This
is illustrated by the change in the Q-bands prior to (Figure c) and after purification (Figure e). Altogether, we
can conclude that the observed red-emissive macromolecule is indeed
porphyrin.During the GSH/FA solvothermal synthesis, two main
synthesis pathways
are in competition (Figure a). The molecular fluorophores formed from GSH can either
react with formamide to form porphyrin or carbonize to CDs. During
the carbonization, the molecular fluorophores are embedded into amorphous
carbon, giving the CDs their fluorescent properties. Indeed, the PL
emission of the molecular fluorophores is almost identical to that
of the CDs (Figure c). Moreover, the UV–vis absorption spectrum of the CD fraction
solely comprises absorption in the UV range as would be expected for
small molecular fluorophores (Figure e). These molecular fluorophores are formed in an early
stage of the reaction (Figure S3).[24] We believe that a major part of these species
are pyrrole derivatives, which are known to form by heating some of
the amino acids[43,44] and have a PL emission spectrum
similar to pyrrole. It is these pyrrole derivatives together with
aldehydes that can condense into porphyrin molecules in the Rothemund
condensation.[45] Formamide plays a crucial
role in the synthesis, because we did not obtain porphyrin in the
GSH solvothermal synthesis in water and DMF (Figure S4). These results suggest that formamide can take up the role
of an aldehyde due to its similar chemical structure and condense
with pyrrole derivatives to porphyrin. Remarkably, we found that the
individual amino acids in GSH do not form porphyrin (Figure S5). Moreover, a mixture of the amino acids including
glycine, cysteine, and glutamic acid do not yield porphyrin either.
This suggests that the full GSH peptide is required to form the pyrrole
porphyrin precursors, and we speculate that cyclization occurs in
the center of the GSH leading to pyrrole derivatives with two functional
groups. This process appears to be universal as we have found a report
of CDs synthesized from natural pulp-free lemon juice containing GSH,
which have porphyrin spectroscopic signatures in both absorbance and
PL emission.[46]
Figure 3
(a) Simplified scheme
of the two competing reactions during the
solvothermal synthesis of porphyrin and carbon dots from glutathione-condensation
and -cyclization products. (b) Table of the favored product (right)
as a function of increased factor during synthesis (left) following
the reaction scheme in (a). Other factors were held constant with
respect to the reported synthesis conditions in the methods. The presence
of the favored product was derived from the photoluminescence (PL)
emission spectra, where emission in the 400–600 nm range indicates
the presence of carbon dots and unreacted molecular fluorophores,
while emission in the 600–750 nm range marks the presence of
porphyrin. (c–d) Normalized PL emission spectra (c) and atomic
force microscopy (AFM) topography histogram (d) of the nonpurified
synthesis products as a function of increasing reaction time (200×
diluted in pH = 7.3 phosphate buffer; GSH-FA-3h and GSH-FA-18h). (e)
Normalized PL emission spectra of the nonpurified synthesis products
as a function of increasing glutathione concentration during synthesis
(200× diluted in pH = 7.3 phosphate buffer; GSH-l-FA-3h, GSH-FA-3h,
and GSH-FA-3h). (f) Normalized PL emission spectra of the nonpurified
synthesis products as a function of synthesis pH (200× diluted
in pH = 7.3 phosphate buffer; GSH-FA-5 min-basic, GSH-FA-5 min, and
GSH-FA-5 min-acidic).
(a) Simplified scheme
of the two competing reactions during the
solvothermal synthesis of porphyrin and carbon dots from glutathione-condensation
and -cyclization products. (b) Table of the favored product (right)
as a function of increased factor during synthesis (left) following
the reaction scheme in (a). Other factors were held constant with
respect to the reported synthesis conditions in the methods. The presence
of the favored product was derived from the photoluminescence (PL)
emission spectra, where emission in the 400–600 nm range indicates
the presence of carbon dots and unreacted molecular fluorophores,
while emission in the 600–750 nm range marks the presence of
porphyrin. (c–d) Normalized PL emission spectra (c) and atomic
force microscopy (AFM) topography histogram (d) of the nonpurified
synthesis products as a function of increasing reaction time (200×
diluted in pH = 7.3 phosphate buffer; GSH-FA-3h and GSH-FA-18h). (e)
Normalized PL emission spectra of the nonpurified synthesis products
as a function of increasing glutathione concentration during synthesis
(200× diluted in pH = 7.3 phosphate buffer; GSH-l-FA-3h, GSH-FA-3h,
and GSH-FA-3h). (f) Normalized PL emission spectra of the nonpurified
synthesis products as a function of synthesis pH (200× diluted
in pH = 7.3 phosphate buffer; GSH-FA-5 min-basic, GSH-FA-5 min, and
GSH-FA-5 min-acidic).We investigated the effect
of various synthesis conditions, as
summarized in Figure b. A longer reaction time favored carbonization, which was marked
by the relative decrease in porphyrin PL emission in the 600–750
nm range (Figure c).
Macairan et al. concluded that a longer reaction time in the GSH/FA
CD synthesis resulted in more C–N and C=N bonds measured
with X-ray photoelectron spectroscopy (XPS), which indicate increased
carbonization of the CDs.[28] Also for other
synthesis methods, a longer reaction time has been linked to a higher
degree of carbonization.[25] We observed
with atomic force microscopy (AFM) that the size of the CDs increases
from roughly 4 to 8 nm with reaction times from 3 to 18 h (Figure d). The size distribution
of the CDs is narrow and well-defined, which is in line with a nucleation
and growth mechanism reported previously.[25] Porphyrin is below the detection limit of AFM as the planar molecule
is known to arrange flat on the substrate surface. Transmission electron
microscopy confirms the CD size and, in combination with optical microscopy,
shows that the porphyrin self-assembles into dendritic structures
at high concentration (Figure S6). The
formation of porphyrin occurs mostly in the first 10 min of the reaction,
monitored via the PL emission at various synthesis times (Figure S3a–b). As the reaction progresses,
the relative abundance of porphyrin decreases, and the carbonization
pathway dominates. A higher GSH concentration during the solvothermal
synthesis also favors CD formation, again marked by the decrease of
the porphyrin PL emission in the 600–750 nm domain (Figure e). Altogether, this
results in a higher probability for nucleation, which leads to a higher
CD concentration.To further investigate the mechanism of porphyrin
and CD formation,
the synthesis was performed in acidic and basic conditions. Carbonization
occurs rapidly in the acidic environment and no porphyrin PL emission
was observed in the reaction mixture (Figure f). Conversely, basic reaction conditions
favor porphyrin as main synthesis products because of the low carbonization
rate. As discussed before, we believe that pyrrole derivatives are
the precursor for porphyrin in this reaction. In accordance with the
literature, the aromatization,[47] condensation,[48] and pyrrole formation[49] are all favored in alkaline environments, while in acidic condition
the dehydration and degradation of organic compounds are favored.[47,50] In particular, chemical dehydration carbonizes the products and
decreases the H/C and O/C ratio.[47,50] We conclude
that in an acidic environment side reactions are dominant, such as
dehydration, which inhibits small molecule condensation and porphyrin
formation.In addition to blue-emitting fluorophores, which
have a maximum
PL emission around 460 nm, an additional emission band around 550
nm is intensified in a neutral and basic environment (Figure f). In accordance with literature,
fluorophores with single aromatic rings (e.g., citrazinic acid) usually
have blue emission, while those with two or more rings have green
emission (e.g., hydroxy-1H-pyrrolo[3,4-c]pyridine-1,3,6(2H,5H)-trione (HPPT)).[27,51−53] Therefore, we conclude that in neutral and basic
reaction conditions, in addition to condensation reaction forming
porphyrin molecules, the formation of polycyclic aromatic fluorophores
is also possible. The high abundance of pyrrole species in basic synthesis
conditions increases the probability for further reaction to polycyclic
aromatic dyes, marked by the 550 nm PL emission (Figure f). Both the lower carbonization
rate as well as the higher formation rate of pyrrole derivatives result
in higher selectivity toward porphyrin. Thus, by steering the competition
between carbonization and porphyrin formation, the desired product(s)
can be obtained at increased yields.
Conclusion
It
was found that dual emission of GSH/FA-based products does not
originate from a single CD emitter but rather from a mixture of physically
separate blue-emissive CDs and red-emissive porphyrin compounds. We
demonstrate an easy way to separate the two synthesis compounds without
the need for time-consuming dialysis. Systematic variations in synthesis
parameters resulted in dramatically different product distributions,
which can be explained via a simple model in which a porphyrin-forming
reaction pathway is in direct competition with a carbonization pathway
forming CDs. By understanding the nature of the system, we are now
able to steer the synthesis process toward the desired product, which
opens up a cheaper and more environmentally friendly synthesis route
to CDs, water-soluble porphyrin, and mixed systems. With such control
over the GSH/FA-based synthesis, the obtained CD system is another
step closer to its application. The large shift between the Soret
and Q absorption band(s) of the water-soluble porphyrin makes it,
for example, interesting for applications in dye-sensitized solar
cells.[54] Finally, our findings motivate
the detailed characterization and purification of other dual-fluorescence
CD systems both from synthetic and natural carbon sources.[40]
Authors: Ming Fu; Florian Ehrat; Yu Wang; Karolina Z Milowska; Claas Reckmeier; Andrey L Rogach; Jacek K Stolarczyk; Alexander S Urban; Jochen Feldmann Journal: Nano Lett Date: 2015-08-17 Impact factor: 11.189
Authors: Musa Yahaya Pudza; Zurina Zainal Abidin; Suraya Abdul Rashid; Faizah Md Yasin; A S M Noor; Mohammed A Issa Journal: Nanomaterials (Basel) Date: 2020-02-12 Impact factor: 5.076