Ionic liquid-assisted thermal decomposition synthesis of carbon dots and graphene-like carbon sheets for optoelectronic application

Abstract

For academic research and practical applications, exploration of a facile, low cost, and easy to operate synthesis strategy has become a primarily important work in the development of low dimensional carbon-based nanomaterials. As new members in the family of carbon nanomaterials, zero-dimensional carbon dots and two-dimensional graphene sheets offer many possibilities for extensive applications, especially in the optoelectronic field, but their synthesis processes are somewhat complicated. More challenging is the design and synthesis of carbon dots and graphene sheets that can be directly used in photovoltaics with simple device fabrication processes and low cost. In this work, a novel and easy process route for synthesis of zero-dimensional carbon dots and two-dimensional graphene-like carbon sheets is reported, which relies on direct carbonization of small organic molecules in a liquid-phase by employing ionic liquid as solvent. It is found that the as-prepared carbon material exhibits excellent dispersibility in common solvents, easy film-formation ability and evident photoelectrochemical activity. Meanwhile, the synergistic effect of the carbon dots and carbon sheets for electronic charge generation and transport on their enhanced photovoltaic performance is also explored. It is anticipated that our work may open a new window to facilely prepare novel carbon nanostructures for a wide range of applications.

---

1. Introduction

In recent years, low dimensional carbon-based nanomaterials, such as fullerene,¹ carbon dots,² carbon nanotubes,³ and graphene,⁴ have received great research attention and triggered a scientific “gold rush” because of their superior and interesting physical and chemical properties. Among all the carbon-based nanomaterials, graphene, an atomically thin film composed of a singer layer of carbon atoms, has attracted intense attention and regard as one of the highly promising optoelectronic materials owing to its many merits including large specific surface area, high electrical conductivity and so on.⁵ In addition, many studies have revealed that carbon dots possess a series of unique features such as large optical absorptivity, size-tunable optical response, efficient multiple carrier generation, cost-effective, and low cytotoxicity.^6–8 As a result of these characteristics, carbon dots have been viewed as one of highly promising photoelectric functional materials compared with their counterparts inorganic semiconductor quantum dots,^9,10 and it can be exploited as a novel leading nanomaterial for construction of all carbon-based ‘‘dot/sheet” hybrid nanostructures in optoelectronic-related applications where conventional semiconductor quantum dots and graphene are currently used.^11,12

On the synthetic front, as newly developed “bottom-up” approaches, such as hydrothermal,^13,14 microwave-assisted polymerization,¹⁵ solvent-free decomposition,^16,17 and high-temperature pyrolysis,¹⁸ preparation of both small fragments of carbon dots and large flakes of graphene-like materials from small organic molecules has considered to be an available avenues because of these methods have advantages of simple synthetic procedures, low cost and wide precursor tolerance.^19,20 However, most of these approaches suffer from tedious procedure and difficulty of scaling up, which make them less attractive for practical production. More challenging is the design and synthesis of carbon dots and graphene those can be directly used in application with simple post-purification procedures, easy fabrication processes and low cost. To move forward, exploration of a new route for synthesis of carbon dots and graphene-like structure with merits of facile synthetic processes and easy device fabrication processing for optoelectronic application is still highly desirable.

In this paper, a simple strategy to facile preparation of carbon dots (CD) and graphene-like carbon sheets (CS) is reported, which relies on thermal decomposition of small organic molecule using ionic liquid (IL) of 1-butyl 3-methyl imidazolium bromide ([bmim]Br) as solvent by taking advantage of the excellent features of IL, including low vapor pressure, high thermal and chemical stability, good electronic conductivity and low melting point,^21–23 This study reveals that the [bmim]Br plays crucial roles in optimization of the synthetic operation process, improvement of the sample's dispersibility in common solvents and simplifying the processes for fabrication of photovoltaic device for the IL blended products (noted as IL + CD/CS). Meanwhile, the synergistic effects of the CD and CS for electronic charge generation and transport on their enhanced photocurrent response activity and photoelectric conversion efficiency are also explored. It is anticipated that our work may open a new window to facilely prepare novel carbon nanomaterials with high performance for a wild range of applications.

2. Results and discussion

As illustrated in Scheme 1a, the thermal decomposition synthetic strategy for preparation of CD and graphene-like CS starting from L-cysteine (L-cys) using [bmim]Br as solvent is described herein as an example. The thermogravimetric analysis (TGA) revealed that the decomposition temperature of [bmim]Br and L-cys are 250 °C and 210 °C, respectively (see ESI, Fig. S1†). Furthermore, the nuclear magnetic resonance (¹H NMR) spectra reveal that the chemical shift signals of hydrogen atoms in the [bmim]Br are not clearly changed even heating treatment of [bmim]Br at 240 °C for 120 min (ESI, Fig. S2†), suggesting that [bmim]Br possesses excellent thermal and chemical stability. According to the above results, the temperature for the thermal decomposition was set to be 240 °C. For the IL-assisted thermal decomposition approach, the solid L-cys powder was quickly dissolved in [bmim]Br and formed a liquid state mixture, which was immediately changed its color from pale yellow to dark brown with vigorous stirring, as illustrated in Scheme 1b. Scheme 1c revealed that a gel-like product was finally generated after heating treatment of [bmim]Br and L-cys at 240 °C with 30 min thermal duration. It's worth pointing out that the as-prepared product can be easily transferred to glass substrate and further facilely processed to a uniform film with high transparent (Scheme 1d) through dip/disperse-coating techniques. Meanwhile, the as-prepared product possesses good dispersibility in common solvents, such as water, chloroform, and especially for ethanol (EtOH). It is observed from Scheme 1e that the as-prepared product dispersed in EtOH with high dispersibility even without ultrasonic treatment.

Scheme 1 Schematic representation of [bmim]Br assisted thermal decomposition of L-cys for synthesis of carbon dots (CD) and graphene-like carbon sheets (CS).

For insight into the morphology and chemical structure of the L-cys generated product, the as-prepared sample was dispersed in deionized (DI) water and then filtrated through a filter membrane (with pore diameter of 450 nm). The filtrate was further purified through a column chromatography separation (silica gel, EtOH/chloroform = 1/3, v/v) to get small fragments of CD. As shown in Fig. 1a and b, the photographs of CD in aqueous solution under visible (a) and 365 nm UV (b) light reveal a pale yellow and blue fluorescence emitting, respectively. In Fig. 1c, excitation-dependent photoluminescence (PL) behavior was observed, which is similar to previous report.¹⁴ The morphology of CD was characterized by transmission electron microscope (TEM) and atomic force microscope (AFM). The TEM images of the CD shown in Fig. 1d and S3† reveal good dispersibility with uniform size distribution (ca. 91.5 ± 5 nm). Compared to previous reported work, the particle size of the as-prepared CD is much larger than that of pervious reported CD,¹³ which is likely due to the longer reaction time as well as the high polymerization and condensation reaction efficiency under the homogeneous phase reaction condition provided by [bmim]Br. Meanwhile, AFM image in Fig. 1e shows that the topographic heights of the obtained CD are mostly distributed in the range from 1.0 to 3.5 nm, according to the height profiles along the line (Fig. 1f and g), which suggests that the thickness of CD is similar to those of single layer and few layers (less than four layers) of graphene. Moreover, it is observed that the XRD pattern of CD displays a wide reflection peak centered at around 23.53° (Fig. 1h), suggesting the interlayer spacing of (002) diffraction peak is 0.38 nm. The relatively larger d-spacing of the as-prepared CD than that of pristine graphite (0.34 nm) is due to the intercalation of IL and the formation of oxygen-containing functional groups between the layers of CD.^24,25

Fig. 1 (a–b) Digital photos of CD in daylight (a) and under 365 nm UV light (b); (c) PL spectrum of CD in water; (d) TEM image of the prepared CD; (e) AFM image and the height profile along the line of the as-prepared CD. (f–g) Height profiles along the line (1) and (2), respectively. (h) XRD pattern of the thermal decomposition prepared CD. The CD was prepared from L-cys at 240 °C for 30 min.

The chemical compositions of the CD were investigated by X-ray photoelectron spectroscopy (XPS) analysis (Fig. S4†). The XPS spectrum indicates that the CD is mainly composed of carbon, oxygen, nitrogen, and sulfur elements, which give the atomic compositions of C, O, N, S are 70.04%, 9.64%, 7.74%, and 12.58%, respectively (Fig. S4a†). The higher resolution spectrum of C_1s (Fig. S4b†) revealed mainly four individual ones assignable to C–C (∼284.5 eV), C–N/C–S (∼285.5 eV), C–O (∼286.5 eV), and C=O (∼288.0 eV) bonds.¹³ The presence of nitrogen and sulfur atoms and the fact that they are incorporated into the CD lattices were also ascertained by XPS measurement. The N_1s spectrum (Fig. S4c†) reveals that the nitrogen atom mainly exists as the pyridine-like sp²-hybridization form (399.7 eV), pyrrole-like sp³ hybridized form (401.2 eV) as well as lactam and imide-like sp³ hybridized form (400.7 eV).²⁶ It is speculated that the sp³ hybridized form was belonged to the incompletely carbonized L-cys which existed at the framework of CD. The high-resolution spectrum of the S_2p bands reveals the presence of C–S–C units (Fig. S4d†).¹³

Following the chemical and structural characterization for the small fragments of CD, the focus was shifted toward the morphology and structural characterization of the as-prepared large flake of graphene-like CS. Upon the completion of filtration treatment, the remaining CS film collected on the filter membrane (Fig. 2a) was dissolved in DI water after sonication treatment, as shown in Fig. 2b. AFM in Fig. 2c revealed a micro-sized flake, and the height distribution analysis inset of Fig. 2c illustrated that the as-prepared CS with thickness of 3.0 ± 0.5 nm. We inferred that the CS has larger thickness than that of the pristine graphene mostly due to the incompletely carbonized L-cys existing at their frameworks. Raman spectrum under 633 nm laser excitation gave a D-band at 1337 cm⁻¹, G-band at 1568 cm⁻¹, correspond to the defects in the CS basal plane and the sp² lattice structures, respectively, thus indicating that the prepared CS has a high quality as similar to graphene oxide (Fig. 2d).^27,28 Meanwhile, the TEM images in Fig. 2e and S5† illustrated graphene oxide-like morphology with hundreds of nanometers to micrometers in size, in which few layer sheets with corrugated flake-like sharp were clearly observed. From the high resolution TEM (HRTEM) image (Fig. 2f), the lattice spacing of 0.21 nm agrees with that of in-plane lattice spacing of graphene (100 facet).²⁹

Fig. 2 (a) Digital photo of the CS film coated on a millipore filtration by vacuum filtration. (b) Photograph of the CS in water. (c) AFM image of the prepared CS. Inset is the height profile along the line. (d) Raman spectrum of the CS under 633 nm laser excitation. (e) TEM and (f) HRTEM images of the CS. Typical CS with lattice parameter of 0.21 nm. The sample was prepared by thermal decomposition of L-cys at 240 °C for 30 min and then purified by filtration.

The chemical compositions of the flake-structural CS were also investigated by XPS analysis (Fig. S6†). The XPS data of CS gave the atomic compositions of C, O, N and S at 70.19%, 18.73%, 7.55% and 3.52%, respectively. The appeared Si signal was ascribed to the typical peaks of the silicon substrate. The deconvoluted C_1s peaks (Fig. S6b†) consist of mainly two individual ones assignable to C=C (284.5 eV), C–N/C–S (285.5 eV) bonds.¹³ The appearance of another two binding energy peaks with weak intensity respective at 286.5 eV and 288.0 eV confirms there are few sp³ hybridized C–O and C=O in the framework of CS. The higher resolution spectrum of N_1s (Fig. S6c†) revealed that the nitrogen atom mainly exists as the pyridine-like sp²-hybridization form (398.7 eV), pyrrole-like sp³ hybridized form (399.7 eV), as well as lactam and imide-like sp³ hybridized form (400.7 eV).²⁶ The high-resolution spectrum of the S_2p bands (Fig. S6d†) displayed the presence of C–S–C units.²⁵ Besides, the high-resolution spectra of C_1s, N_1s and S_2p reveal that the hybridization manner of the CS is similar to that in the CD. Additionally, it is not observed the Br_3d5/2 signal (70.4 eV)³⁰ in the XPS spectrum of L-cys (C₃H₇O₂NS) generated CS sample, indicating the [bmim]Br (C₈H₁₅N₂Br) was not physically adsorbed on the surface of CS. In order to confirm the [bmim]Br (C₈H₁₅N₂Br) was only used as solvent but not acted as precursor to participate the carbonization reaction processes, another type carbon sheet generated from citric acid (C₆H₈O₇, without N and Br elements) was also prepared through IL-assisted thermal decomposition method, and its chemical components were also investigated by XPS (Fig. S7†). It is not observed the N_1s and Br_3d5/2 signals in the XPS spectrum of the citric acid generated graphene-like flake sample (TEM images were shown in Fig. S8†), suggesting the [bmim]Br indeed was not joining the carbonization reaction processes.

According to the above results, we proposed that L-cys molecules undergo decarboxylation and produce low dimensional CD and CS at their decomposition temperature. As illustrated in Scheme 2, we inferred that the L-cys molecules were firstly dissolved in [bmim]Br and then decomposited to form the nucleus of sp² domain that is composed of C=C bonds. Subsequently, the growth of sp² domain occurred at the nucleus surfaces or edges with consuming the source molecules of L-cys. Owing to the homogeneous phase reaction condition created by [bmim]Br, the source molecules could quickly reach the surface of the sp² domain in the formed CD and generate new sp² domain via the thermal polymerization process. Meanwhile, the freshly size enlarged sp² domains are fused together and further assist the increase of CD size by repeating the above-mentioned process.^31,32 With extension of reaction duration, large size CD is crosslinking and stacking to form large size CS until most of the L-cys are consumed.

Scheme 2 Schematic illustration of probable reaction process for preparation of CD and CS using [bmim]Br as solvent. (a) The L-cys molecules are well-dissolved in [bmim]Br; (b) L-cys molecules were decomposited to form the nucleus of CD; (c) the diameter of the CD can be increased by prolonging the heating time; (d) with extension of reaction duration, large size CD crosslinking and stacking to form CS until most of the L-cys are consumed.

In order to confirm the above hypothesis, the effect of reaction time (5 min, 10 min, 20 min and 60 min) on the particle sizes and optical property of the products were investigated. First, the as-prepared raw products were firstly filtrated through a millipore filter (with pore diameter of 450 nm) to remove large flake structures, and then purified by a flash chromatography (silica gel, EtOH/chloroform = 1/3, v/v) to obtain the CD. TEM is taken to discern the morphology of the small fragments of CD, which revealed that thermal decomposition of L-cys with 5 min, 10 min, 30 min and 60 min produced CD having average sizes of 34 nm, 48 nm, 84 nm and 137 nm, respectively (Fig. S9†), namely, the sizes for these small fragments of CD show a trend of gradual increase with the extension of reaction time. Additionally, the longer duration time, the more filtration residuals were obtained on the millipore filter, indicating an increasing number of large flakes of CS were formed with the extension of reaction time. Furthermore, the effect of reaction time on the optical property of L-cys synthesized products was further studied. The as-prepared raw products in water were directly used for optical property characterization without further purification. The UV-visible absorption (UV-vis) and PL emission spectra of the resultant aqueous solutions vary sensitively with reaction time (5 min, 10 min, 30 min, and 60 min) are presented in Fig. 3. It is noted that the absorption peak for each sample revealed a trend of red-shift when extension the thermal decomposition durations. In particular, as shown in Fig. 3a–c, the samples prepared with reaction duration for 5 min, 10 min and 30 min display absorption peaks at 324 nm, 327 nm, and 334 nm, respectively, which could be attributed to the feature of the large edge/bulk ratio of CD, rending the remarkable absorption peak for CD.³³ Contrastingly, such evident absorption peak is not observed for the sample with 60 min thermal duration. It is observed that the sample synthesized with reaction duration for 60 min shows a broad absorption below 600 nm but without any obvious peak (Fig. 3d), which is very closed to the characteristic absorption feature of graphene and graphene oxide,³⁴ suggesting there exists large flake structures with graphitic framework of graphene in the product with 60 min thermal duration. Apart from the reaction time-dependent absorption behavior, it is also noted that the PL behaviors of these products demonstrated a trend of gradual red-shift with the extension of reaction time. Along with the changed excitation wavelength from 360 to 440 nm, the PL peak of the products prepared for short duration (5 min and 10 min) red-shifted from 447 to 498 nm, 444 to 496 nm, respectively. In contrast, along with the changed excitation wavelength from 400 nm to 480 nm, the photoluminescence peak of the products prepared for long duration (30 min and 60 min) red-shifted from 487 to 526 nm, 508 to 538 nm, respectively. Furthermore, the emission spectra of L-cys generated products with short duration (5 min, 10 min) showed the same maximum PL excitation at 400 nm with the emission wavelength of 469 nm and 467 nm, respectively. For the L-cys generated products with long duration (30 min, 60 min), their PL spectra revealed the emission wavelength of 523 nm and 533 nm at the same maximum excitation wavelength of 460 nm, respectively. The PL mechanism of these products is still not clear. Generally, it is believed that the PL may be mainly derived from intrinsic state emission and defect state emission. The former one is induced by the quantum size effect, whereas the latter one depends on the molecule-like state.^34–36 In this work, it is inferred that the variation of the PL emission was owing to the distinct particle sizes and graphitic degrees of the CD from the gradually extended reaction time, namely, the size effect is a result of quantum dimensions whereas the surface state is analogous to a molecular state, both of which contribute to the complexity of the excited states of CD.^13,37

Fig. 3 UV-vis absorption and PL spectra for samples thermal decomposition prepared at 240 °C for different durations (5 min, 10 min, 30 min and 60 min). All the samples were dispersed in water and without further treatment.

Moreover, preliminary results on the PL spectral shifts of the L-cys synthesized products that occur as a function of solvent polarity was also studied in our work. With EtOH as solvent, the PL behaviors (Fig. S10†) of the samples thermal decomposition prepared at 240 °C for different durations also revealed a trend of gradual red-shift with the extension of reaction time (5 min, 10 min, 30 min and 60 min), similar to Fig. 3. Interestingly, PL spectra for these samples in EtOH revealed evident red-shift with the extension of reaction time. For example, for the sample with short reaction time (5 min and 10 min), their excitation maximum were 460 nm, 520 nm, and their emission maximum were 521 nm and 592 nm, respectively. While the reaction time was 30 min, the excitation maximum was 600 nm and emission maximum was 654 nm, respectively. If further prolong the reduction time to 60 min, the emission maximum was shifted to 675 nm with excitation maximum of 640 nm. This result has important implications for the use of such product for providing multiple colors under visible light excitation.

Currently, it is of great interest and importance to synthesis of low dimensional carbon based hybrid nanomaterials, especially in anchoring CD on graphene surfaces, because the combination and interaction of CD and graphene will promote the generation of unprecedented physical properties and provide a potential pathway for constructing all carbon-based devices with high light energy conversion performance.^11,12 Moreover, exploration of a novel route for preparation of carbon dots and graphene-like structures with merits of facile synthetic procedures and easy for processing in solar cells are still highly desirable.

Taking advantages of the excellent dispersibility and easy film-formation ability of the prepared IL + CD/CS, as a proof-of-concept, the gel-like IL + CD/CS sample was directly used as a newly photoactive material to demonstrate their potential in photovoltaic (PV) application. As illustrated in Fig. 4a, a chemical etched ITO electrode was employed as the photoelectrode for photovoltaic performance investigation (more details were reported in Experimental section). The IL + CD/CS was directly coated onto a chemical etched ITO area to link the two electrodes to form a uniform thin film through a doctor-blending process, as shown inset of Fig. 4b. Shown in Fig. 4b are typical experimental current–voltage characteristics measured from the blank ITO and IL + CD/CS coated ITO photoelectrodes, respectively. From analysis of the current–voltage curves, the etched ITO exhibits a negligible current value, suggesting that the ITO was well etched, which is crucial to photoconductive photocurrent response measurement. Contrastingly, it is seen that the IL + CD/CS electrode respond rapidly to irradiation, and the photocurrent for the IL + CD/CS film is seen to increase gradually with increasing bias voltage. The inset in the right corner of Fig. 4b shows the rapid and consistent photocurrent responses for each switch-on and switch-off event in multiple 50 s on–off cycles under light illumination. The photocurrent density of the IL + CD/CS coated ITO photoelectrode at zero volt bias is only 0.022 μA cm⁻², and reaching up to ca. 0.42 μA cm⁻² at −0.5 V and ca. 0.63 μA cm⁻² at +0.5 V, respectively. The evident photocurrent response activity of the prepared IL + CD/CS film is probably attributed to the synergistic effect of the CD and CS for charge carriers generation and transport as well as [bmim]Br leads to the electronic conductively enhancement, which is very encouraging and promising prospect in novel carbon-based light energy conversion applications.³⁸

Fig. 4 (a) Schematic illustration of photocurrent response device fabrication out of the IL + CD/CS on etched ITO. (b) Chopping photocurrent–voltage curves of blank etched ITO and IL + CD/CS coated ITO electrodes. The inset picture in the lower left corner is IL + CD/CS constructed photoelectrode. The inset in the right corner is the photogenerated current responses versus light on/off cycle time profiles of the IL + CD/CS photoelectrode with −0.5 V, 0 V, and +0.5 V bias, respectively.

Furthermore, the optoelectronic application of IL + CD/CS was further preliminary explored in TiO₂ based solar cells. The device model is illustrated in Fig. 5a and b, and shown in Fig. 5c and d are experimental current–voltage (I–V) characteristics measured from the devices made by the control sample (TiO₂ only, Fig. 5c) and the IL + CD/CS modified TiO₂ composite (noted as IL + CD/CS-TiO₂, see Fig. 5d). For the control sample, the open-circuit photovoltage (V_oc) is 0.55 V, the fill factor (FF) is 0.41, and the generated photocurrent (J_sc) is 0.04 mA cm⁻², while for the as-prepared composite V_oc is about 0.80 V, FF is 0.37, and the J_sc increased from 0.04 mA cm⁻² (for the pure TiO₂) to 0.13 mA cm⁻² (IL + CD/CS modified TiO₂). The J_sc response of the IL + CD/CS-modified device is over 3 times higher than that of the individual TiO₂ particle. Although device performance is low due to many solar cell performance related factors have not been optimized in this work, the concept of using this composite in the solar cell is novel and the device efficiency enhancement achieved here are already exciting. Investigations on the optimization of the IL + CD/CS-TiO₂ device's structure and improvement of the device performance are in progress.

Fig. 5 (a) The sketch showing the nanostructure of IL + CD/CS-modified TiO₂ particles as active layer in solar cell. (b) Schematic illustration of the model of the device. (c–d) Photocurrent versus voltage curves for (c) individual TiO₂ and (d) IL + CD/CS-modified TiO₂ devices in dark and under 300 W Xenon light illumination.

Following the photovoltaic property characterization for the prepared IL + CD/CS, the focus was shifted toward the nature of the charge transfer between CD and CS. In the present work, the purified CD and CS were employed as building blocks for the construction of all carbon-based hybrid photoelectrodes. As comparison, the individual CD and CS were also used in photoelectrochemical (PEC) performance investigation. The photocurrent response activity of the as-prepared CD/CS hybrid photoelectrode was investigated in a PEC cell, in which the CD/CS hybrid films from electrophoretic deposition were employed as photoelectrodes, and their thickness was adjusted through changing the electrophoretic deposition time (Scheme S1†). It is clear to see that the photogenerated current exhibited a trend of increase first and then decrease alone with increasing of the film thickness (Fig. S11†). This phenomenon is consistent with previous research result, which shows that the photogenerated current response value of the photoelectrode has a close relationship to the photoelectrode film thickness because of the competition between photon absorption and charge recombination factors.^39,40 As shown in Fig. 6a, the CD, CS and CD/CS hybrid films subjected to 10 seconds of on/off cycles exhibited stable and evident photocurrent response curves to changing light on/off states at zero-bias voltage, while the maximum photogenerated current value of the CD/CS hybrid photoelectrode (23.9 ± 0.20 μA cm⁻²) is evident higher than those of the independent CD (18.2 ± 0.58 μA cm⁻²) and CS (10.9 ± 0.19 μA cm⁻²). The enhanced photocurrent response performance indicates higher separation efficiency of the photo induced electron–hole pairs and a lower recombination rate in the CD/CS hybrid structures. The charge transfer resistance and separation efficiency of the CD, CS and CD/CS electrodes have been further investigated by electrochemical impedance spectroscopy (EIS).⁴¹ The EIS Nyquist plots of the prepared CD, CS, and CD/CS hybrid photoelectrodes under light irradiation in Fig. 6b show that the CD/CS hybrid photoelectrode hybrid film has a smaller semicircle in the intermediate frequency region as compared to that of the independent CD and CS electrodes, suggesting the CD/CS electrode possess a smaller charge transfer resistance (R_ct), which represents a more facile charge transfer across the host material/electrolyte interface, namely, the low R_ct reveals the excellent electrochemical capacitive property of the CD/CS hybrid material.⁴² We proposed that the photocurrent response activity and EIS of the photoelectrode are strongly dependent on the charge recombination rate. The CD/CS hybrid photoelectrode shows better performance than that of the independent CD and CS is due to the more facile charge transfer across the photoactive layer in CD/CS system. By combining CD and CS, the obtained CD/CS hybrid film exhibits a significant increase in electrical conductivity by taking advantage of the synergistic effect of electronic charge generation and transport. This phenomenon is consistent with previous research result.^5,43,44

Fig. 6 (a) Photogenerated current responses versus light on/off cycle time profiles of the film thickness optimized CD, CS, and CD/CS photoelectrodes, respectively. (b) EIS Nyquist plots of CD, CS, and CD/CS hybrid electrodes under Xe light irradiation. The inset in (b) represents the corresponding equivalent circuit. (c) UV-vis spectra of the prepared CD, CS and the CD/CS hybrid dispersions; (d) PL spectra of CD aqueous solution adding different amount of CS. Excitation wavelength is 400 nm.

To better understand the charge carrier interactions between CD and CS, a detail optical property characterization was also carried out on the independent CD, CS and their hybrid (CD/CS). It can be seen in Fig. 6c that the UV-vis absorption spectrum of the CD aqueous solution shows an absorption band at around 337 nm. In the case of the CS, less light absorption (in the UV and visible regions) behavior similar to the graphene oxide was observed. For the CD/CS aqueous mixture, a slight increase in absorbance versus independent CD is observed. As previous reported, the PL spectra were utilized to study the surface processes involving the electron–hole fate of the as-prepared CD and CS.⁴⁵ The PL spectrum in Fig. 6d shows a dominant emission peak at 458.5 nm for the CD, while the PL intensity of CS could be neglected. Moreover, it is observed that PL intensity of the CD aqueous solution is gradually quenched with increasing amount of CS added in the system, thus indicating the efficient separation of photogenerated electron–hole charge carriers over CD.⁴⁶

3. Conclusions

In summary, we have demonstrated a facile route for the preparation of zero-dimensional CD and two-dimensional graphene-like CS in the manner of thermal decomposition of small organic molecules in the presence of [bmim]Br, in which the [bmim]Br as reaction solvent plays a crucial role in the optimization of synthetic progresses and enhancement of the dispersibility of the prepared carbon nanostructures. It was found that the product from IL-assisted thermal decomposition exhibits the features of favorable dispersibility in solvents, easy film-processing property and evident photocurrent response activity attributed to the [bmim]Br leads to the electronic conductively enhancement as well as the synergistic effect of the carbon dots and carbon sheets for electronic charge generation and transport on their enhanced photoelectric conversion efficiency. The concept of using this composite in the solar cell is novel, and thus, these results will extend the application of carbon dots and graphene composite in the next generation solar cells and other optoelectronic devices.

4. Experimental section

4.1 Materials

The 1-butyl 3-methyl imidazolium bromide, abbreviated as [bmim]Br, L-cysteine (L-cys) and citric acid were purchased from J&K Chemical Company (Shanghai, China). All the chemicals were of analytical grade and used without further purification. The indium tin oxide (ITO) and fluorine doped tin oxide (FTO) glasses were purchased from Zhuhai Kaivo Optoelectronic Technology Co., Ltd. (Zhuhai, China). The micropore filter membrane (pore diameter of 450 nm) was purchased from Tianjin Jinteng Corp (Tianjin, China), respectively.

4.2 Synthesis

Typically, L-cys (0.5 g) and [bmim]Br (0.2 g) were loaded into a beaker and then heated in a sand bath at 240 °C with vigorous stirring. Upon the completion of reaction for desired duration (the reaction time was set to be 5 min, 10 min, 30 min, and 60 min, respectively), the IL blended CD and CS mixtures were obtained. After cooling down to room temperature, the product was dispersed in water, and then with filtration treatment by using micropore filter membrane and resulted in the large sized CS film. Meanwhile, the CD was obtained by flash chromatography (silica gel, EtOH/chloroform = 1/3, v/v) of the filtered product.

4.3 ITO etching

The ITO area used as electrodes was protected by tapes with adhesive tape exposed, and then the tape-protected ITO was coated with a layer of zinc powder and immersed into 1 M hydrochloride solution to etch ITO without tape protection. Second, the IL blended product (noted as IL + CD/CS) was coated onto the etched area to link the two electrodes by a doctor blending process.⁴⁷

4.4 Preparation of TiO₂ photoanodes

FTO glasses (ca. 12 Ω □⁻¹) were washed with detergent, deionized water, acetone, isopropanol and subsequently dried under nitrogen. The TiO₂ layers were prepared from a precursor, in which titanium butyrate (TiC₁₆H₃₆O₄) was dissolved in isopropanol. The precursor was spin-coated on FTO substrates, and the obtained films were annealed at 400 °C for 40 min in air. After annealing and slow-cooling to room temperature, the TiO₂ films were socked in isopropanol over night to remove the carbonized precursor and subsequently dried under nitrogen. Meanwhile, TiO₂ particles (P25, AEROXIDE) were firstly undergoing thermal annealing treatment at 450 °C for 60 min in air before further use. Subsequently, the IL + CD/CS modified TiO₂ particles were prepared from blending the liquid state IL + CD/CS with P25 under stirring at 70 °C for 30 min. After slow-cooling to room temperature, the obtained IL + CD/CS-P25 layerer were prepared by doctor blading on TiO₂ films. For the control device, the independent P25 was directly coated onto the TiO₂ films by doctor blading.

4.5 Preparation of IL + CD/CS modified TiO₂ solar cell

The solar cell device was fabricated with a sandwiches structure by assembling the TiO₂ photoanode and platinum (Pt) coated FTO. The Pt-coated photocathode was prepared by adopting a dipping-pulling method to coat the H₂PtO₆ (5 mM) on the FTO glass and sintered at 390 °C for 20 min.⁴⁸ Then a drop of the iodide-based liquid electrolyte was injected into the sandwiched cells through the drilled holes on the devices. The effective area of the device is 0.1 cm².

4.6 Electrodeposition for preparation of photoelectrode

Briefly, the ITO glass was firstly pre-cleaned with detergent-cleaning under sonication for 30 min followed by distilled water washing. Then a home-built electro-deposition system including a glass container (25 mL) and two ITO-coated glass electrode (deposition area: 1.0 cm²), which are facing each other to simulate a parallel-plate-like geometry with a separation of 1.0 cm (Scheme S1a†). Electro-deposition was performed using constant voltage potential of 20 V in CD, CS and CD/CS mixture solution. After electrodeposition, the deposited ITO glass was washed with ethanol and left to dry at room temperature.

4.7 Photocurrent performance measurement

A photoelectrochemical (PEC) cell with three-electrode configuration was constructed in this work (Scheme S1b†). The platinum wire and saturated calomel electrode (SCE) were respectively used as the counter electrode and the reference electrode. The PEC performance of the photoelectrode was measure in 0.1 M Na₂S aqueous electrolyte. A 300 W Xe lamp was used as light source. The current response was recorded with a CHI 660B electrochemical workstation. The photocurrent response activities of the IL + CD/CS films as active layers based devices subjected to light on/off cycles were measured.

4.8 Material characterization

UV-vis absorption spectra were recorded with a T6 UV-vis spectrometer. Photoluminescence (PL) measurements were performed using a LS55 fluorescence spectrometer (PerkinElmer, USA). Raman spectrum was obtained by a confocal microscope Raman system (Renishaw in via Raman microscope). X-ray photoelectron spectroscopy (XPS) was collected on a multifunctional XPS/AES system (Kratos Axis Ultra DLD, Japan) by using Al Kα radiation. Atomic Force Microscopy (AFM) measurement was taken using an Agilent 5500 atomic force microscope. Transmission electron microscope (TEM) measurement was taken with TecnaiTMG2F30. X-ray diffraction (XRD) pattern was recorded by an X-ray diffract meter Bruker AXSD8 using Cu Kα radiation (40 kV, 200 mA) with a Ni filter. Electrochemical impedance spectroscopy (EIS) Nyquist plot obtained at an AC voltage with amplitude of 0.01 mV over the frequency range of 1 × 10⁻¹ to 1 × 10⁵ Hz for CD, CS, and CD/CS hybrid electrodes in 0.1 mol L⁻¹ Na₂S aqueous solution under Xe light irradiation.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (NSFC. 51402095) and the Natural Science Foundation of Hubei Province (2014CFB555).

Notes and references

1. Y. Wang, Nature, 1992, 356, 585–587.
2. L. A. Ponomarenko, F. Schedin, M. I. Katsnelson, R. Yang, E. W. Hill and K. S. Novoselov, et al., Science, 2008, 320, 356–358.
3. S. J. Tans, A. R. M. Verschueren and C. Dekker, Nature, 1998, 393, 49–52.
4. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang and S. V. Dubonos, et al., Science, 2004, 306, 666–669.
5. F.-X. Xiao, J. Miao and B. Liu, J. Am. Chem. Soc., 2014, 136, 1559–1569.
6. X. T. Zheng, A. Ananthanarayanan, K. Q. Luo and P. Chen, Small, 2015, 11, 1620–1636.
7. Z. Huang, F. Lin, M. Hu, C. Li, T. Xu and C. Chen, et al., J. Lumin., 2014, 151, 100–105.
8. X. Li, M. Rui, J. Song, Z. Shen and H. Zeng, Adv. Funct. Mater., 2015, 25, 4929–4947.
9. J. K. Kim, M. J. Park, S. J. Kim, D. H. Wang, S. P. Cho, S. Bae, J. H. Park and B. H. Hong, ACS Nano, 2013, 7, 7207–7212.
10. V. Gupta, N. Chaudhary, R. Srivastava, G. D. Sharma, R. Bhardwaj and S. Chand, J. Am. Chem. Soc., 2011, 133, 9960–9963.
11. Z. Yang, J. Xiao, J.-Y. Wan, Z.-G. Liu, T.-T. Cao and W.-J. Zhang, et al., Phys. Chem. Chem. Phys., 2015, 17, 32283–32288.
12. S.-H. Cheng, T.-M. Weng, M.-L. Lu, W.-C. Tan, J.-Y. Chen and Y.-F. Chen, Sci. Rep., 2013, 3, 2694.
13. Y. Dong, H. Pang, H. B. Yang, C. Guo, J. Shao and Y. Chi, et al., Angew. Chem., Int. Ed., 2013, 52, 7800–7804.
14. S. Zhu, Q. Meng, L. Wang, J. Zhang, Y. Song and H. Jin, et al., Angew. Chem., Int. Ed., 2013, 52, 3953–3957.
15. Y. Liu, N. Xiao, N. Gong, H. Wang, X. Shi and W. Gu, et al., Carbon, 2014, 68, 258–264.
16. Y. Dong, J. Shao, C. Chen, H. Li, R. Wang and Y. Chi, et al., Carbon, 2012, 50, 4738–4743.
17. C.-B. Ma, Z.-T. Zhu, H.-X. Wang, X. Huang, X. Zhang and X. Qi, et al., Nanoscale, 2015, 7, 10162–10169.
18. M. Latorre-Sánchez, A. Primo and H. García, Angew. Chem., Int. Ed., 2013, 52, 11813–11816.
19. S. Y. Lim, W. Shen and Z. Gao, Chem. Soc. Rev., 2015, 44, 362–381.
20. X. Wang, G. Sun, N. Li and P. Chen, Chem. Soc. Rev., 2016, 45, 2239–2262.
21. D. Carriazo, M. C. Serrano, M. C. Gutierrez, M. L. Ferrer and F. del Monte, Chem. Soc. Rev., 2012, 41, 4996–5014.
22. N. Liu, F. Luo, H. Wu, Y. Liu, C. Zhang and J. Chen, Adv. Funct. Mater., 2008, 18, 1518–1525.
23. D. Nuvoli, L. Valentini, V. Alzari, S. Scognamillo, S. B. Bon and M. Piccinini, et al., J. Mater. Chem., 2011, 21, 3428–3431.
24. S. Sahu, B. Behera, T. K. Maiti and S. Mohapatra, Chem. Commun., 2012, 48, 8835–8837.
25. Q. Liang, W. Ma, Y. Shi, Z. Li and X. Yang, Carbon, 2013, 60, 421–428.
26. R. J. J. Jansen and B. van Bekkum, Carbon, 1995, 33, 1021–1027.
27. Y. Li, Y. Hu, Y. Zhao, G. Shi, L. Deng and Y. Hou, et al., Adv. Mater., 2011, 23, 776–780.
28. N. Denys, S. Valentinas, S. Boris, A. Sanna and L. Harri, Nanotechnology, 2012, 23, 465703.
29. S. Zhu, J. Zhang, S. Tang, C. Qiao, L. Wang and H. Wang, et al., Adv. Funct. Mater., 2012, 22, 4732–4740.
30. D. G. Belov and D. T. Shieh, J. Power Sources, 2014, 247, 865–875.
31. L. Tang, R. Ji, X. Cao, J. Lin, H. Jiang and X. Li, et al., ACS Nano, 2012, 6, 5102–5110.
32. L. Li, G. Wu, G. Yang, J. Peng, J. Zhao and J.-J. Zhu, Nanoscale, 2013, 5, 4015–4039.
33. X. Zhou, Z. Tian, J. Li, H. Ruan, Y. Ma and Z. Yang, et al., Nanoscale, 2014, 6, 2603–2607.
34. K. P. Loh, Q. Bao, G. Eda and M. Chhowalla, Nat. Chem., 2010, 2, 1015–1024.
35. S. Wang, Z.-G. Chen, I. Cole and Q. Li, Carbon, 2015, 82, 304–313.
36. Y.-P. Sun, B. Zhou, Y. Lin, W. Wang, K. A. S. Fernando and P. Pathak, et al., J. Am. Chem. Soc., 2006, 128, 7756–7757.
37. J. Shang, L. Ma, J. Li, W. Ai, T. Yu and G. G. Gurzadyan, Sci. Rep., 2012, 2, 792.
38. P. V. Kamat, J. Phys. Chem. Lett., 2011, 2, 242–251.
39. C. X. Guo, H. B. Yang, Z. M. Sheng, Z. S. Lu, Q. L. Song and C. M. Li, Angew. Chem., Int. Ed., 2010, 49, 3014–3017.
40. H.-X. Wang, J. Xiao, Z. Yang, H. Tang, Z.-T. Zhu and M. Zhao, et al., J. Mater. Chem. A, 2015, 3, 11287–11293.
41. S. Yang, Y. Gong, J. Zhang, L. Zhan, L. Ma and Z. Fang, et al., Adv. Mater., 2013, 25, 2452–2456.
42. J. Zhang, J. Yu, M. Jaroniec and J. R. Gong, Nano Lett., 2012, 12, 4584–4589.
43. H.-X. Wang, Q. Wang, K.-G. Zhou and H.-L. Zhang, Small, 2013, 9, 1266–1283.
44. I. V. Lightcap and P. V. Kamat, J. Am. Chem. Soc., 2012, 134, 7109–7716.
45. Y. Zhang, R. Ciriminna, G. Palmisano, Y.-J. Xu and M. Pagliaro, RSC Adv., 2014, 4, 18341–18346.
46. X. Fang-Xing, S.-F. Hung, J. Miao, H.-Y. Wang, H. Yang and B. Liu, Small, 2015, 5, 554–567.
47. H.-R. Xia, J. Li, W.-T. Sun and L.-M. Peng, Chem. Commun., 2014, 50, 13695–13697.
48. L. Zhu, Y. H. Qiang, Y. L. Zhao and X. Q. Gu, Appl. Surf. Sci., 2014, 292, 55–62.

---

Footnotes

† Electronic supplementary information (ESI) available: TGA of [bmim]Br and L-cys, ¹H NMR of [bmim]Br, TEM images of the prepared CD and CS, XPS of CD and CS, photogenerated current responses of CD, CS and CD/CS hybrid. See DOI: 10.1039/c6ra14181a

‡ Jia-Yun Wan and Ze Yang contributed equally to this work.

---