A. M. Craciun†
a,
A. Diac†b,
M. Focsana,
C. Socacibc,
K. Magyarid,
D. Maniuae,
I. Mihalachef,
L. M. Veca*f,
S. Astilean*ae and
A. Terec*b
aNanobiophotonics and Laser Microspectroscopy Center, Interdisciplinary Research Institute in Bio-Nano-Sciences, Babes-Bolyai University, 42 T. Laurian, 400271, Cluj-Napoca, Romania. E-mail: simion.astilean@phys.ubbcluj.ro
bFaculty of Chemistry and Chemical Engineering, Babes-Bolyai University, 1 M. Kogalniceanu, Cluj-Napoca 400084, Romania. E-mail: asuciu@chem.ubbcluj.ro
cNational Institute of Research and Development for Isotopic and Molecular Technologies, Donat 67-103, RO-400293 Cluj-Napoca, Romania
dNanostructured Materials and Bio-Nano-Interfaces Center, Interdisciplinary Research Institute in Bio-Nano-Sciences, Babes-Bolyai University, 42 T. Laurian, 400271, Cluj-Napoca, Romania
eBiomolecular Physics Department, Faculty of Physics, Babes-Bolyai University, 1 M. Kogalniceanu, 400084, Cluj-Napoca, Romania
fNational Institute for Research and Development in Microtechnologies, IMT Bucharest, Bucharest 077190, Romania. E-mail: monica.veca@imt.ro
First published on 6th June 2016
The versatility of the surface chemistry of carbon dots (CDs) can generate high flexibility in engineering their surface functionality for a wide and precise array of applications. In respect of this, we perform for the first time the surface passivation of carboxylated carbon nanoparticles with p-phenylenediamine to prepare water soluble photoluminescent CDs exhibiting intriguing physicochemical properties. The new strategy used to engineer the surface functionality allows us to synthesize CDs with a stable excitation independent green photoluminescence with a quantum yield of 14% and lifetime of 11.8 ns, suitable for application in bioimaging, sensing and optoelectronics.
Most of the research has focused on improving the quantum yield (QY) of CDs, a quantitative measure of the efficiency of the PL process in the CDs. While CDs with QY up to 93%6 were reported, there is still a growing need to understand better the physicochemical properties which boost their QY, an important aspect for more successful applications of CDs. Surface passivation via covalent bonding of amine-containing agents and/or doping has been demonstrated in various studies to improve significantly the PL of carbonaceous nanomaterials.7,8 For example, polymers like polyethylene glycol (PEG), amine terminated polyethylene glycols (PEG-1500N),9,10 poly(ethylenimide)-co-poly(ethyleneglycol)-co-poly(ethyl-enimide) (PPEI),11 4,7,10-trioxa-1,13-tridecanediamine (TTDDA)10 and polyethyleneimine (PEI)12 are currently the most effective passivating agents for enhancing the PL properties of CDs. One of their major disadvantages lies in the tedious workup needed for the complete removal of the unreacted polymer from the reaction mixture.
Surface functionalization with small aliphatic diamine molecules, such as 1,2-ethylenediamine,8,13 1,4-butanediamine,8 2,2′-(ethylenedioxy)bis(ethylamine),14 diethylamine15 has been demonstrated to increase the PL efficiency and the QY of the functionalized CDs. Zhai et al. confirmed that the primary amine molecules serve as both N-doping precursors and surface passivation agents, which considerably enhance the PL of CDs.8 On the contrary, functionalization with hydroxyl or thiol end-capped organic molecules showed no improvement in the PL intensity, suggesting thus that the N-doping is responsible for this enhancement.
Although π-conjugated polymers have played a central role in the development of novel organic electronics their covalent linkage onto the CDs’ surface has yet to be explored. Polyanilines have been comprehensively investigated16,17 due to their various optical, electrochemical and chemical properties. Similar to aniline, phenylene diamines, can also lead to oligomers and polymers by oxidation, and the properties of the thus obtained materials are comparable to those of polyaniline, surpassing it in some cases (e.g. a lower toxicity is observed for the former).18 Taking into account the well-known susceptibility of p-phenylenediamine (pPD) to oxidative polymerization, as well as its capability to form linear polyaminoanilines which possess multifunctional properties, we have turned our attention to this particular aromatic diamine. Moreover, chemical oxidative polymerization of a pPD monomer was achieved in the presence of graphene sheets leading to nanocomposites with interesting electrochemical properties.19,20
In this work, we have developed a novel and cost-efficient strategy to prepare water soluble photoluminescent CDs by covalent grafting of pPD on carboxylated carbon nanoparticles (CNPs) with the subsequent oligomerization on the surface of the carbon dots. The structural composition of the obtained oligomeric poly-pPD (PpPD) decorated CDs (hereafter PpPD-CDs) was investigated through NMR, FT-IR and XPS measurements. Steady-state and time-resolved fluorescence spectroscopy revealed their excitation-independent PL with a lifetime of 11.8 ns and QY of 14%. We emphasize that the functionalization of CNPs with aromatic amines has not been studied so far, making our study of particular interest. We therefore believe that our findings regarding the formation of photoluminescent CDs functionalized with PpPD will open the door for the development of such optical carbon-based nanomaterials with controllable PL properties and applicability in the field of multicolor light-emitting nanodevices, optical and electrochemical detection systems or supercapacitors.
Fourier transform infrared absorption (FT-IR) measurements were performed with a Bruker Equinox 55 Fourier transform infrared spectrometer on solid samples. Each spectrum was measured in attenuated total reflectance (ATR) mode with 60 scans and 2 cm−1 resolution.
Nuclear Magnetic Resonance (NMR) spectra were recorded at room temperature, in D2O as solvent, on instruments operating at 600 MHz for 1H and 150 MHz for 13C. Chemical shifts (δ) are reported in parts per million (ppm) using residual solvent peak as internal reference.
X-ray photoelectron spectroscopy (XPS) spectra were recorded with a SPECS PHOIBOS 150 MCD system employing a monochromatic Al-Kα source (1486.6 eV), a hemispherical analyser and charge neutralization device. Samples were fixed on double-sided carbon tape and care was taken to ensure that the sample particles covered the tape. Experiments were performed by operating the X-ray source with a power of 200 W, while the pressure in the analysis chamber was in the range of 10−9 to 10−10 mbar. The binding energy scale was charge referenced to the C1s at 284.6 eV. The elemental composition was determined from survey spectra acquired at a pass energy of 60 eV. High resolution spectra were obtained using an analyzer pass energy of 20 eV. Analysis of the data was carried out with Casa XPS software. A Shirley background was used for all curve-fitting along with the Gaussian/Lorentzian product form (70% Gaussian and 30% Lorentzian).
The steady-state PL spectra were collected using an Edinburgh F920 spectrofluorimeter equipped with a 450 W xenon lamp as a light source. In all cases, samples were analyzed in solution and PL spectra were corrected for the intensity of the excitation lamp and the wavelength dependence of the detection system. The relative QY of the synthesized PpPD-CDs was measured by using quinine sulfate in 0.5 M sulfuric acid as fluorescence standard.21 PL lifetime measurements were performed on a PicoQuant MicroTime 200 time-resolved confocal fluorescence microscope system based on an inverted microscope (IX 71, Olympus) equipped with a UPLSAPO 60×/NA 1.2 water immersion objective. The excitation beam was provided by 8 μW picosecond diode laser heads (LDH-D-C-375 and LDH-D-C-510, PicoQuant) operating at 375 nm and 510 nm, pulsed at 40 MHz repetition rate. The samples were dropped on microscope cover glasses and analyzed in solution. The signal collected through the objective was spatially and spectrally filtered by a 50 μm pinhole and longpass filters (HQ405LP and HQ519LP, Chroma, Brattleboro, USA), respectively, before being focused on a PDM Single Photon Avalanche Diode (SPAD) from MPD. The detector signals were processed by the PicoHarp 300 Time-Correlated Single Photon Counting (TCSPC) data acquisition unit, from PicoQuant. Time-resolved fluorescence decay curves were recorded and analyzed using the SymPhoTime software (PicoQuant). The PL lifetimes were obtained through nonlinear iterative deconvolution algorithm. The instrument response function (IRF) was recorded from the laser light backscattered from plain cover glass working in similar experimental conditions. The quality of the fits was judged by analyzing the chi-square (χ2) values and the distribution of the residuals.
The morphology of CDs was analyzed with conventional transmission electron microscopy (TEM) using a JEOL 100 U type TEM microscope operated at 100 kV accelerating voltage.
On the contrary, when the reaction was stopped after 3 days, the 1H-NMR spectrum of PpPD-CDs (Fig. 1(a)) shows in the aromatic region signals only traces of unreacted pPD along with the presence of at least two sets of signals with approx. 2.5:
1 relative intensities. The first set includes a pair of doublets at 6.99 and 7.35 ppm, respectively, a singlet at 7.47 ppm along with two singlets corresponding to imine-type protons at 8.19 and 8.23 ppm, respectively. The second set of signals includes three multiplets at 6.98–7.02, 7.21 and 7.47 ppm, respectively and a doublet at 7.12 ppm. Imine protons appear slightly deshielded at 8.47 and 8.58 ppm, respectively. These data indicate that more than one type of oligomeric entity has been grafted on the surface of CDs.
13C APT-NMR spectrum (Fig. 1(b)) supports these observations, of more than one set of signals. For example, the spectrum exhibits two sets of signals of different intensities for imine-type C atoms (162.5 and 162.6 ppm vs. 165.4 and 165.6 ppm).
Both the reaction yield and the properties of the thus obtained material were optimal in these reaction conditions and our attention was focused on the investigation of its properties. Therefore our experimental data indicate the oligomerization of pPD in the presence of oxidized CNPs, with the formation of different-sized oligomers at the surface, in agreement with the already mentioned data published on the polymerization of pPD in the presence of graphene oxide through a redox process.19,20 According to the literature and fitting well with our experimental results, the most probable structure of oligomers is a linear one, similar to that of polyaniline.22
The FT-IR technique employed to identify the organic functional groups on the surface of the amine-functionalized CDs brought interesting aspects regarding the structural composition of PpPD-CDs. The FT-IR spectra of PpPD0 and PpPD-CD do not show significant differences except for relative intensities.
Specifically, as shown in Fig. 2, the intense and sharp peaks in the 3200–3400 cm−1 region of PpPD0 spectrum, due to N–H stretching vibrations, appear as small shoulders on a broad band in the IR spectrum of PpPD-CDs, certifying a decrease in the number of NH2 groups which is a direct indication of oligomerization. The presence in the PpPD-CD IR spectrum of specific amide/imine C–N stretching vibrations at 1677 cm−1 and 1302 cm−1, indicate the formation of PpPD-CDs. In addition, the spectrum of PpPD-CDs reveals the presence of aromatic stretching vibrations characteristic to quinoid (1555 cm−1) and benzenoid (1520 cm−1) units, and aromatic C–H out-of-plane deformation vibration of substituted benzene ring (826 cm−1).
To further confirm the oligomerization of pPD on CNPs, we recorded the XPS spectra of CNPs and PpPD-CDs. The XPS survey scans (not presented here) indicate the presence of strong C and O signals in both samples. They also indicate a significant increase of the N peak in PpPD-CDs, compared to CNPs, where a very weak band was detected. Low levels of chloride, calcium, sodium and sulphur are also identified in the samples, as a consequence of chemical procedures performed during synthesis. The relative contributions of C, N and O in the samples is 70.2, 1.8 and 18.1%, respectively, for CNPs, and 70.4, 7.5 and 17.7%, respectively, in the case of PpPD-CDs. As shown in Fig. 3(a), the C1s XPS spectrum of CNPs can be resolved into five distinct carbon states at 284.4, 285.4, 286.6, 288.1, 289 and 290.4 eV which are attributed to sp2 bonding, sp3 bonding, C–O, CO, O–C
O and π–π*, respectively.23 In the C1s XPS spectrum of PpPD-CDs (Fig. 3(b)) the peaks at 286.3, 287.4 and 288.7 eV can be attributed to C–O/C–N, C
O/C
N and O
C–NH bonds, respectively, due to the formation of amide bonds and oligomerization of pPD.24,25 The N1s XPS spectrum of PpPD-CDs (Fig. 3(d)) reveals three distinct peaks at 398.6, 399.6 and 401.2 eV which can be attributed to pyridinic N, O
C–N–H and C
N bonds, respectively.8,26 The extremely weak bands at 400.1 and 406.1 eV from the N1s XPS spectrum of CNPs (Fig. 3(c)) could arise, as previously discussed, from the chemical compounds used during preparation procedures.
Furthermore, the fitting analysis of O1s XPS spectra (Fig. 3(e) and (f)) reveals an increase of the peak assigned to CO bonds detected in PpPD-CDs compared to CNPs, from 50.3% to 61.8%. The increased signals from the C
O, C–N and N–H bonds (Fig. 3(b), (d) and (f)) observed in PpPD-CDs XPS spectra, compared with CNPs XPS spectra, suggest the formation of amide bonds due to surface-passivation, while the appearance of C
N bond suggests the oligomerization of pPD on CNPs, in good agreement with the FT-IR results.
![]() | ||
Fig. 4 (a) Absorption spectra of PpPD-CDs (black) and PpPD0 (red); inset: zoom in on the 360–600 nm range; (b) illustrative TEM images of PpPD-CDs. Scale bar: 50 nm (20 nm in inset). |
The absorption spectrum of PpPD-CDs is dominated by an intense band located at 257 nm and displays also a broad, weak band around 520 nm. In turn, along with the absorption bands observed in the PpPD-CDs, the PpPD0 exhibits two additional transitions at 392 and 437 nm. While the absorption in the UV region is associated with the transitions in the benzene ring, the bands at 392, 437, 520 nm could be assigned to the electronic transitions between conjugated adjacent benzenoid rings, π–π* transition of substituted phenazine conjugated to the lone electron pairs on the adjacent amine group, and to the electron transition from the benzenoid ring to the quinoid rings, respectively.22 A shoulder at 360 nm is also visible in the spectrum of PpPD-CDs from the n–π* transition of the CO/C
N bonds (see arrow in inset). These results, in concordance with the NMR and FT-IR studies, are suggesting the oligomerization of the pPD.
The illustrative TEM images of PpPD-CDs, presented in Fig. 4(b), reveal a pattern of uniform dark dots with almost spherical morphology and a particle size centered at 6 ± 1 nm.
Further, we evaluated the luminescent properties of both PpPD-CDs and PpPD0 under excitation wavelengths in the 300–560 nm range (see Fig. 5).
As shown in Fig. 5(a), PpPD-CDs exhibit an intense excitation-independent PL band centred at 505 nm obtained for excitation wavelengths between 320 nm and 420 nm. The highest intensity of this band is obtained at 360 nm excitation, well-corroborated by the presence of the absorption shoulder at the same wavelength (see inset Fig. 4(a)).
Although CDs with excitation-independent PL have been successfully reported in the literature, the conclusions regarding the PL mechanisms are not yet clear. According to previous studies, an excitation-independent PL behaviour indicates that the PL centres are uniform due to a surface rich in amino-groups.27,28 In other words, as our structural characterization suggested, PpPD-CDs have a uniform structure, with few types of functional groups determining a single, dominant radiative transition mode. Several studies demonstrated that excitation-independent PL indicates a homogeneous surface state of sp2 clusters.29,30 A similar mechanism has been used to explain excitation-independent PL in graphene quantum dots.31 Similarly, Wang et al. propose that the excitation-independent in N-doped CDs may originate from the bandgap-like transition of the amide/imine groups in CDs.32 Furthermore, Zhai et al. suggest that excitation-independent PL luminescent in CDs synthesized through microwave-assisted pyrolysis originates from relatively uniform and well-passivated CDs surface.8
We therefore believe that the PL centred at 505 nm corresponds to the CDs that are uniformly passivated with oligo PpPD. The relative quantum yield of the synthesized PpPD-CDs, measured at λex = 375 nm using quinine sulfate was found to be 14%, comparable to or even higher than the values reported in the literature for passivation of CDs with small diamines.13,33 It is worth noting that the PL spectra acquired three months after their synthesis showed no changes, except that the relative quantum yield decreased to 7%. An additional evidence of their long time stability is the insignificant decrease in the PL intensity after one hour of excitation with a 6 mW UV source (375 nm).
Further, when we gradually shift the excitation wavelength above 440 nm, we observe that the PL signal decreases significantly and red-shifts towards 600 nm. To identify the origin of this signal we further investigated the PL properties of the control PpPD0 sample. As shown in Fig. 5(b), PpPD0 exhibits an excitation-dependent composite PL signal in the 400–550 nm region and a more intense excitation-independent PL band located at 595 nm, obtained for excitation in the range of 320–420 nm, and 440–520 nm, respectively. The optimal emissions are obtained for excitations at 340 nm and 520 nm, respectively, close to the absorption bands from the absorption spectrum in Fig. 4(a). We presume that the composite PL signal in the 400–550 nm region arises from different types of oligomers obtained during the oxidation process. This phenomenon might be due to the superposition effect of different emitter entities. As regards the second PL band, located in the red region, we recall that PpPD-CDs also exhibit a weak PL band at around 600 nm that has the optimal intensity for 520 nm excitation (inset Fig. 5(a)). Interestingly, Jiang et al. reported photoluminescent CDs obtained from solvothermal treatment of pPD exhibiting PL at 604 nm at the optimal excitation wavelength of 510 nm.34 Although our method is quite different, we assume that the excitation-independent PL signal recorded at around 600 nm in both samples could arise from surface-passivated CDs that might have been obtained during the thermal treatment of pPD. Nevertheless, a more complex investigation of these species would be beyond the purpose of our study. More importantly, the PL signal at 600 nm is significantly lower than that of functionalized CD, thus the contribution of these secondary species to the final solution is minimal.
Subsequently, we investigated the PL excitation spectra of both PpPD-CDs and PpPD0. The obtained spectra are comparatively presented in Fig. 6 along with the PL spectra for excitation close to the maximum of the excitation bands. The PL excitation spectrum of PpPD-CD obtained for λem = 505 nm (solid black line), presents a single band at 353 nm which can be correlated with the n–π* transition previously observed at 360 nm in the absorption spectrum (see inset Fig. 4(a)) and has mirror symmetry to the PL band obtained at 360 nm excitation (dashed black line). Along with the excitation-independent behaviour of the PL, the emergence of a single band in the excitation spectrum of PpPD-CD proves their high optical uniformity.
![]() | ||
Fig. 6 Comparison between the PL excitation (solid lines) and PL emission (dashed lines) spectra of PpPD-CDs (black) and PpPD0 (red). |
On the contrary, PpPD0 shows an excitation maximum at 525 nm, a smaller band at 365 nm and a shoulder at 450 nm, matching the profile of the absorption spectrum from Fig. 4(a). Also, the excitation band is mirror symmetric with the maximum emission band for excitation at 520 nm (dashed red line).
To support our results from steady-state PL measurements presented above, we further performed PL lifetime measurements. The time-resolved PL curve of PpPD-CDs obtained at 375 nm and 510 nm excitation is presented in Fig. 7 along with the curves obtained for PpPD0, while the obtained PL lifetime values are listed in Table 1.
![]() | ||
Fig. 7 Comparison between the PL decay curves of PpPD-CDs and PpPD0 at λexc = 375 nm (a) and λexc = 510 nm (b). Cyan lines represent the fitting curves. Intensity is on a logarithmic scale. |
Sample | λexc (nm) | 〈τ1〉 (ns) | 〈A1〉 (%) | 〈τ2〉 (ns) | 〈A2〉 (%) | τavg (ns) |
---|---|---|---|---|---|---|
a τn and An are the lifetime and amplitude of the nth component; τavg is the amplitude-weighted average lifetime calculated with:![]() |
||||||
PpPD-CDs | 375 | 11.8 | 100 | — | — | 11.8 |
PpPD0 | 375 | 7.2 | 22.1 | 2.3 | 77.9 | 3.4 |
PpPD-CDs | 510 | 1.1 | 22 | 3.1 | 78 | 1.6 |
PpPD0 | 510 | 1.2 | 51 | 2.2 | 49 | 1.7 |
The PL decay profile of PpPD-CDs obtained at 375 nm excitation was fitted with a single exponential function, while for PpPD0 a two-component exponential function was needed for fitting the decay profile at the same excitation wavelength (see Fig. 7(a)).
As shown in Table 1, the PL lifetime of PpPD-CDs was found to be 11.8 ns. Our results are in good agreement with the results of Ding et al. on CDs exhibiting a surface-state controlled PL characterized by a monoexponential lifetime and a single emission band.23 One of the generally accepted mechanisms for the monoexponential behaviour is the uniformity of the PL centres passivation which enables a single emission.11,23,35 According to the study by Zhang et al., long lived fluorescence lifetime arises from their deprotonation and formation of amide bonds, being assigned to an n–π* transition.36
Moreover, Xu et al. demonstrated that the increase of fluorescent QY and the lifetime of CDs is correlated with the increase of nitrogen amount,35 which infers that the long lived fluorescence lifetime exhibited by PpPD-CDs is well correlated with the amount of nitrogen attained after passivation, as FT-IR and XPS results show. Furthermore, after passivation, the –COOH groups are transformed into –CONHR which reduces the non-radiative recombination induced by –COOH and protects CDs from quenching reagents and environmental factors, which also contributes to the observed long fluorescence lifetime.37,38
It is noteworthy that the single lifetime component obtained for PpPD-CDs is well corroborated with their excitation-independent PL activity observed in the steady-state PL experiments for excitation between 340 nm and 420 nm (see Fig. 5(a)). The rather long PL lifetime of synthesized PpPD-CDs, significantly longer than that of most organic dyes (1–5 ns) and cell auto fluorescence (2–3 ns), enables their use in biological applications, considering, in addition, the solubility in water of the synthesized nanomaterial. According to previous studies, CDs with a long PL lifetime could be used as reliable labels for long-term cell imaging,39 or have potential applicability in fluorescence lifetime imaging microscopy (FLIM)40 or solar cells.41
On the contrary, the average fluorescence lifetime of PpPD0 at 375 nm excitation is 3.4 ns, significantly lower than the lifetime of PpPD-CDs. The lifetime components obtained after fitting operations (see Table 1) could be assigned to different-sized PpPD oligomers obtained during synthesis, as the asymmetry of the PL band obtained at 380 nm excitation indicates (see Fig. 5(b)).
Therefore, the time-resolved PL measurements further indicate the successful passivation of CDs and the formation of completely new fluorescent products, i.e. PpPD-CDs, with well established PL features. Finally, in order to verify our previous hypothesis regarding the existence of a similar photoluminescent source with PL at ∼600 nm in both PpPD-CDs and PpPD0, we recorded the PL decay curves of both samples at 510 nm excitation. Our fitting procedures revealed an important resemblance between the two lifetime components extracted as well as between the corresponding average lifetimes, as shown in Table 1. We therefore assume that the PL at 600 nm has the same origin in both samples and more interestingly, although not the subject of our study, could be related to the presence of a small amount of short lived photoluminescent CDs obtained directly from pPD.
Footnote |
† These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2016 |