Novel fluorescence resonance energy transfer optical sensors for vitamin B₁₂ detection using thermally reduced carbon dots

Abstract

In this paper, a novel thermally-reduced carbon dot (t-CD) based fluorescence resonance energy transfer (FRET) sensor for the determination of vitamin B₁₂ (VB₁₂) in aqueous solutions is reported. Carbon dots (CDs) have attracted great attention due to their excellent tunable optical properties, low cost, easy fabrication and low toxicity, which make them ideal candidates for optical sensors. Through esterification reactions, blue luminescent t-CDs were prepared by the carbonization of citric acid and thermally reduced by a thermogravimetric analyzer. After thermal reduction, the quantum yield of the t-CDs demonstrated a 5-fold increase, which makes t-CDs excellent donors in the FRET process. The t-CDs were used to detect VB₁₂ with concentrations ranging from 1 to 12 μg ml⁻¹ and their limit of detection (LOD) was as low as 0.1 μg ml⁻¹. The as-synthesized t-CD based optical probing technique is demonstrated to be simple, cost-effective, sensitive and selective for the detection of biologically significant VB₁₂.

---

1. Introduction

Vitamins are known as essential organic compounds needed for healthy humans to function. Vitamin B₁₂ (VB₁₂) is one of a series of cobalt tetraazamacrocyclic complexes, which plays a vital role in red blood cell formation and nerve cell maintenance.^1–3 VB₁₂ deficiency causes severe and irreversible damage, such as anemia, metabolic abnormalities and psychosis. Excessive absorption of VB₁₂ may also lead to unexpected adverse effects. Therefore, the determination of VB₁₂ has received tremendous attention in the last few decades. Traditional approaches like high-performance liquid chromatography (HPLC), chemiluminescence and atomic absorption spectrometry are widely applied to detect VB₁₂ with high sensitivity.^4–8 Unfortunately, the above-mentioned detection techniques are significantly hindered by time-consuming procedures, high equipment costs, or complex pre-separation. Thus, it is necessary to develop a cost-effective and highly sensitive method to determine the level of VB₁₂.

Due to simple instruments and easy operation properties, fluorescence resonance energy transfer (FRET) techniques have received tremendous interest for the detection of ions and small organic molecules,^9–12 and utilize the energy transfer between two chromophores (one is a donor and another is an acceptor). Briefly, the donor is initially excited to a high energy state and transfers energy to the acceptor via nonradiative dipole–dipole coupling.¹³ The FRET system should satisfy two conditions: (a) the emission spectrum of the donor and the absorption spectrum of the acceptor are appreciably overlapped to an extent; (b) the distance between the donor and the acceptor is in the nanometer scale (1 to 10 nm).

Carbon dots (CDs) have received tremendous attention due to their excellent tunable optical properties, better photostability, easy fabrication, low cost and better biocompatibility.^14,15 In contrast to conventional fluorescent materials like organic dyes and rare earth quantum dots, the emission peak of CDs can be easily controlled in a big wavelength range with a relatively small FWHM (full wavelength at half maximum), which makes CDs appropriate to detect diverse bio-organic molecules but only if the emission peak of the CDs (donors) and the absorbance peak of the bio-molecules (acceptors) are highly matched. Moreover, the surface structure of CDs can be easily engineered to demonstrate strong bonding with specific detected acceptors, which allows for a short distance between the CDs and the acceptors. These excellent properties satisfy the requirements of the FRET system and make CDs excellent potential donors in FRET optical systems with high selectivity and sensitivity.

In this paper, a thermal reduction strategy was applied to obtain highly luminescent CDs via heating at 300 °C (as shown in Scheme 1(a)). Compared to other modification methods, the CDs would be less contaminated because no chemical reagents are used in this method.^16–23 Therefore, the thermally reduced CDs are suitable in biological and medical applications. The t-CD based optical sensor was developed to determine the level of VB₁₂ (as shown in Scheme 1(b)). The limit of detection (LOD) was decreased from 0.4 μg ml⁻¹ (LOD for CD based optical sensor) to 0.1 μg ml⁻¹ (100 ppb), which was improved in comparison with a CdTe QD based optical sensor (LOD = 150 ppb)²⁴ and was close to conventional approaches like HPLC and chemiluminescence.

Scheme 1 (a) Thermal reduction process from CDs to t-CDs and (b) fluorescence resonance energy transfer process from t-CDs to vitamin B₁₂.

2. Experimental section

2.1 Materials

Citric acid, sodium hydrate, quinine sulfate and vitamin B₁₂ were purchased from Sigma-Aldrich. Dialysis bags (molecular weight cut off = 2000 Da) were also ordered from Sigma-Aldrich. All reagents were used as received without any further purification.

2.2 Sample preparation

The CDs were prepared as described previously.²⁵ Briefly, 2 g of CA were put into a 10 mL beaker and heated at 200 °C on a hot plate. About 5 minutes later, the CA was liquated. Subsequently, the color of the liquid changed from colorless to pale yellow, and then orange after 30 minutes, implying the formation of the CDs. The obtained orange liquid for preparing the CDs was neutralized to pH 7.0 with a 10 mg ml⁻¹ NaOH solution.

The mixture was further dialyzed in a dialysis bag (retained molecular weight: 2000 Da) to obtain greenish yellow fluorescent CDs. The thermal reduction reactions were carried out in a thermogravimetric analysis (TGA) instrument (Q50 TA). 3 mg of the CDs were put onto a platinum pan and heated at 300 °C for 2 hours. After that, brightly blue luminescent t-CDs were obtained.

The CDs and t-CDs were dispersed in DI water and were both sonicated for 30 minutes in an ice bath, using a bath ultrasonicator (60 W, frequency 40 KHz, Model 2510, Branson) to achieve uniform and stable solutions (25 μg ml⁻¹). A VB₁₂ solution was prepared by dissolving the required amount of VB₁₂ powder in DI water. Then 200 μl of the CD solution or t-CD solution were firstly added into a 5 ml volumetric flask. After that, various specific amounts of the VB₁₂ solution were also added and finally the mixture was mixed with DI water to obtain 5 ml of solution. The concentration of CDs or t-CDs in the mixed solution was kept at 10 μg ml⁻¹ and the concentration of VB₁₂ was varied from 0.04 μg ml⁻¹ to 12 μg ml⁻¹.

2.3 Characterization

UV absorbance measurements were carried out on a JASCO V-550 UV-vis spectrophotometer, equipped with a Peltier temperature control accessory. Fluorescence spectra were measured on a FluoroMax-3 spectrofluorometer. All spectra were recorded in a 1.0 cm path length cell. FTIR characterization was carried out on a Nicolet IS10 FTIR spectrometer by the KBr pellet method. The samples were thoroughly ground with exhaustively dried KBr. XPS spectra were obtained on an X-ray photoelectron spectroscope (XPS, PHI5000 Versa Probe). TEM images were recorded using a HITACHI 8100 transmission electron microscope operating at 75 kV. The zeta potential was characterized by a Malvern Zetasizer Nano ZS90 using dynamic light scattering. Quantum Yield (QY) measurements were used to calculate the quantum yields of the CDs and t-CDs. Quinine sulfate in 0.1 M H₂SO₄ (QY = 0.54) was chosen as a standard. The quantum yields of the CDs and t-CDs in water were calculated according to the formula: Φ = Φ_s(I/I_s)(A/A_s)(n_s/n),²⁶ where Φ is the quantum yield, I is the measured integrated emission intensity, n is the refractive index of the solvent (1.33 for water), and A is the optical density. The subscript “s” refers to the reference standard with a known quantum yield.

3. Results and discussion

3.1 Structural characterization

The CDs which emit green luminescence were synthesized as previously reported,²⁵ and then they were thermally reduced at 300 °C for 2 hours. The obtained thermally reduced CDs (t-CDs) showed an improvement in their photoluminescence intensity and an aqueous solution of the t-CDs was stable for at least six months at room temperature. As shown in Fig. 1(a) and (b), the transmission electron microscopy (TEM) images showed that both the CDs and t-CDs were well dispersed and their size distributions were similar. It demonstrated that the thermal reduction made no major difference in the size of the CDs. The dynamic light scattering results also presented a similar size distribution (4.8–9 nm) and a similar average size (6 nm) for the CDs and t-CDs, as shown in Fig. 1(d) and (e), which were consistent with the results of the TEM. It indicated that the luminescent blue shift of the t-CDs could result from the change in the surface chemical structure and the decrease in the surface defect after thermal reduction, rather than size distribution. A high-resolution TEM (HRTEM) image (Fig. 1(c)) reveals the crystalline structure of the CDs, which is similar to that of many other reported CDs.^27–29

Fig. 1 TEM images of (a) the CDs and (b) the t-CDs, high-resolution TEM (HRTEM) image of the CDs (c) and size distribution of (d) the CDs and (e) the t-CDs via dynamic light scattering (DLS).

Fourier transform infrared (FT-IR) spectroscopy was used to further study the surface functional groups and chemical components of the CDs and t-CDs. As shown in Fig. 2(a), the stretching vibrations of C–OH at 3430 cm⁻¹, –C=C– at 1595 cm⁻¹ and –C–O–C– at 876 cm⁻¹ were found for both the CDs and t-CDs. However, the stretching vibration absorption bands of –COO at 1715 cm⁻¹ and –C–O–C– at 1250 cm⁻¹ were only observed for the CDs. The vibrations of C–OH, –C–O–C– and –COO for the CDs indicated the existence of hydroxyl, carboxyl and epoxy functional groups on the surface of the CDs. After thermal reduction, the vibration absorption bands of –COO at 1715 cm⁻¹ and –C–O–C– at 1250 cm⁻¹ disappeared. Meanwhile, vibrations of –CH at 2920 cm⁻¹ were observed in the t-CDs, demonstrating that the –COO functional groups are thermally reduced on the surface of the t-CDs.³⁰ In addition, X-ray photoelectron spectroscopy (XPS) results (Fig. S1, ESI†) indicate that the CDs and t-CDs are mainly composed of carbon and oxygen. The results demonstrate that the oxygen content is reduced after the thermal reduction. The high resolution spectrum of C1s exhibits four main peaks (Fig. 2(b) and (c)). The binding energy peak at 284.5 eV confirms the graphitic structure (sp² C–C) of the CDs and t-CDs. The peak at about 286.7 eV suggests the presence of C–O, and the peak around 288.1 eV and 288.9 could be assigned to C=O and –COOH, respectively. After thermal reduction, the amounts of C=O and –COOH are decreased. These results are consistent with the FT-IR results.

Fig. 2 FT-IR spectra of the t-CDs and CDs (a) and deconvoluted XPS C1s core level spectra of the CDs (b) and t-CDs (c).

The average zeta potentials of the CDs and t-CDs are −21.36 ± 1.38 mV and −15.16 ± 2.01 mV (pH = 7.0), respectively. This indicates that the thermal reduction decreased the surface charges of the CDs due to the decrease in epoxy and carboxyl groups. These results are consistent with the FT-IR and XPS results.

3.2 Optical characterization

The optical properties of the CDs and t-CDs were studied using UV-vis absorption and photoluminescence spectroscopy. As shown in Fig. 3(a), due to the n–π* transition, an obvious peak at 350 nm was observed in the UV-vis absorption spectra of the CDs. In comparison, no obvious peak was observed in the UV-vis spectra of the t-CDs, which was due to the decrease in carboxyl and epoxy groups during the thermal reduction process. As shown in Fig. 3(b) and (c), the CDs showed maximum excitation and emission wavelengths near 370 nm and 460 nm, whereas the maximum excitation and emission wavelengths of the t-CDs were near 330 nm and 420 nm.

Fig. 3 (a) UV-vis absorption and normalized photoluminescence spectra of the CDs and t-CDs. Inset: photograph of the CD (left) and t-CD (right) aqueous solutions under visible light and 365 nm UV light, respectively. (b) Emission spectra of the CDs with excitation at different wavelengths, (c) emission spectra of the t-CDs with excitation at different wavelengths.

In contrast to the CDs, a blue shift of the emission peak was observed in the photoluminescence spectra of the t-CDs. This increase in the band gap was due to less carboxyl groups existing on the surface of the t-CDs. The CDs and t-CDs both exhibited excitation-dependent properties. The maximum emission wavelength of the CDs was dependent on the excitation wavelength, shifting from 430–520 nm when excited from 280–480 nm. In comparison, the t-CDs had a maximum emission wavelength which ranged from 420–470 nm, while the excitation wavelength ranged from 280–465 nm. The broad and excitation-dependent emission of the CDs and t-CDs presented the effects of the different emission sites of each sp² cluster.^31,32 The imperfect absorption peak of the CDs and broad absorption below 600 nm of the t-CDs both demonstrated that the sp² clusters contained in the CDs and t-CDs were not uniform in size.

The QYs of the CDs and t-CDs at 340 nm were 3.27% and 16.28%, respectively (selecting quinine sulfate as the standard), which indicated that the thermal reduction could significantly improve the QY of the CDs (as shown in Table 1). Combining the FT-IR and XPS results and the QY measurements, the decrease in carboxyl groups in the t-CDs significantly enhanced the photoluminescence of the CDs.³³ As shown in Fig. S2 (ESI†), the results show a good photostability of the t-CDs in an aqueous solution.

Table 1 Quantum yields of the as-synthesized CDs and t-CDs

Sample	Integrated emission intensity (a.u.)	Absorbance at 340 nm (a.u.)	Refractive index of solvent (n)	Quantum yield (%)
Quinine sulfate	2 504 394 589	0.0418	1.33	54
CDs	211 713 269	0.0583	1.33	3.27
t-CDs	988 546 174	0.0525	1.33	16.28

---

3.3 CDs and t-CDs for sensing VB₁₂

In order to detect VB₁₂ in an aqueous solution, CD and t-CD based optical sensors were developed. As shown in Fig. 4(a), VB₁₂ has an obvious peak at 550 nm in the range from 400 to 800 nm, which is attributed to the π–π* transition. The photoluminescence spectra of the CDs and t-CDs at pH = 7 are shown in Fig. 4(b), where the t-CDs presented a luminescence intensity 4 times higher than that of the CDs, which was consistent with the quantum yield results of the CDs (3.27%) and t-CDs (16.28%).

Fig. 4 (a) UV-vis absorbance spectrum of vitamin B₁₂, (b) 0.01 mg ml⁻¹ CD and t-CD PL intensity via 360 nm excitation, (c) the spectral overlap between the fluorescence of the CDs (360 nm excitation) and the absorbance spectrum of VB₁₂, (d) the spectral overlap between the fluorescence of the t-CDs (360 nm excitation) and the absorbance spectrum of VB₁₂, and (e) Stern–Volmer plot of the CDs quenched by a VB₁₂ aqueous solution.

The integral overlap spectra of the UV-vis absorbance spectrum of VB₁₂ and the emission spectra of the CDs or t-CDs are shown in Fig. 4(c) and (d). These spectra exhibited an evident overlap between the luminescence spectra of the CDs or t-CDs and the absorption spectrum of VB₁₂. This indicated the possibility of a fluorescence resonance energy transfer process from the CDs or t-CDs to VB₁₂. As shown in Fig. 4(e), the fluorescence intensity vs. the VB₁₂ plot can be curve-fitted into (I₀/I) = K_SV[VB₁₂] + 0.9769, close to the Stern–Volmer equation with a correlation coefficient (R²) of 0.998 and K_SV of 38.337 mg ml⁻¹. Fig. S3 (ESI†) shows the kinetic behavior of the fluorescence in the t-CD–VB₁₂ system, which has a long time stability.

The luminescence spectra of the CDs or t-CDs in the presence of different concentrations of VB₁₂ are presented in Fig. 5. The emission intensity of the CDs or t-CDs decreased with an increasing concentration of VB₁₂, which indicated that the CD– or t-CD–VB₁₂ systems were efficient FRET systems. Under the specified optimal reaction conditions, linear calibration graphs were constructed for the energy transfer efficiency of the CDs or t-CDs at different concentrations of the vitamin B₁₂ samples. Each concentration was analyzed three times. The efficiency of the fluorescence energy transfer process has been calculated from the following equation

where F is the donor fluorescence intensity in the absence (F₀) and presence (F) of the acceptor. For a 0.4 μg ml⁻¹ addition of VB₁₂ to the t-CD aqueous solution only a 2% energy transfer occurs, whereas for a 12 μg ml⁻¹ addition the energy transfer is 31%, as shown in Fig. 5(c). This demonstrated that the efficiency of the energy transfer increases with the gradual addition of VB₁₂. The same tendency is obtained with the CDs and the energy transfer efficiency is 32% for the 12 μg ml⁻¹ VB₁₂ addition, which is almost equal to that of the t-CDs (31%), as shown in Fig. 5(a). This confirmed that the CDs and t-CDs are both excellent donors in the FRET process, and demonstrated that the reduced epoxy and carboxyl groups did not diminish the efficiency of the FRET process and that the t-CDs still had a strong affinity to VB₁₂. The graphs of the CDs and t-CDs in Fig. 5(b) and (d) were linear in the concentration range 1–12 μg ml⁻¹ with correlation coefficients, R² = 0.98623 and 0.98962, respectively. The high correlation coefficients indicated that CD or t-CD based optical sensors had a high detection accuracy.

Fig. 5 (a) Fluorescence quenching of the CDs in the absence (black line) and presence (other colors) of concentrations of vitamin B₁₂ from 1 to 12 μg ml⁻¹. (b) The calibration graph for the efficiency of the FRET process vs. concentration. (c) Fluorescence quenching of t-CDs in the absence (black line) and presence (other colors) of concentrations of vitamin B₁₂ from 1 to 12 μg ml⁻¹. (d) The calibration graph for the efficiency of the FRET process vs. concentration.

According to an IUPAC criterion,³¹ the limit of detection (LOD) is calculated as the concentration of VB₁₂ which produces an analytical signal that is three times larger than the standard deviation (n = 4) of the blank signal. From this criterion, the standard deviation of the blank signal was an essential parameter to affect the LOD of the optical sensor. For the CD based optical sensor, the standard deviation of the photoluminescence intensity was 1.87%, whereas that of the t-CD based optical sensor was 0.36%, which was much smaller. It exhibited that the as-synthesized t-CD aqueous solution acquired more uniform and stable optical properties and a higher photostability, due to decreased carboxyl groups after the thermal reduction. The carboxyl group had a significant impact on the nonradiative recombination during a change in the surroundings like pH.³⁴ Repeated experiments were performed to obtain the LOD with various concentrations of VB₁₂ in a 10 μg ml⁻¹ CD or t-CD aqueous solution. The LOD of the CD based optical sensor was 2 μg ml⁻¹, however the LOD of the t-CD based optical sensor was 0.1 μg ml⁻¹ (as shown in Fig. S4 and S5, ESI†). This demonstrated that the t-CD based optical sensor obtained a significantly higher sensitivity to detect VB₁₂ in contrast to the CD based optical sensor. As shown in Fig. S6 (ESI†), 100 μg ml⁻¹ of glycine, tyrosine and melamine was added to a 25 μg ml⁻¹ t-CD solution, respectively. It is obvious that there was no evident change in PL intensity in the presence of glycine, tyrosine and melamine, which showed that the t-CD-based sensor obtained good selectivity.

The LOD of the t-CDs is 100 ppb, which is similar to that of the CdTe quantum dot based FRET optical sensors (150 ppb²⁴). However, the photoluminescence behaviors of the t-CDs are tunable and their simple synthesis allows for mass production, which is significantly better than the CdTe quantum dot based FRET optical sensor. In addition, the as-synthesized t-CD based FRET optical sensors possess more advantages like a low cost, environmental friendly fabrication process. Moreover, the lower cytotoxicity and higher biocompatibility make it possible to detect VB₁₂in vivo. The advantages discussed above demonstrate that t-CDs are superior in the FRET process for the detection of VB₁₂ in comparison with CdTe quantum dots.

4. Conclusion

In summary, an environmentally benign method was firstly used to improve the quantum yield of carbon dots via thermal reduction. Compared to the neat carbon dots, the thermally-reduced carbon dots demonstrated more uniform and stable optical properties with a significantly enhanced quantum yield. This is mainly due to a reduced number of carboxyl groups. Thermally-reduced carbon dot based fluorescence resonance energy transfer optical sensors were demonstrated to be very effective in the detection of VB₁₂ in an aqueous solution with the concentration ranging from 1 to 12 μg ml⁻¹ and their limit of detection (LOD) was as low as 0.1 μg ml⁻¹. The as-synthesized thermally-reduced carbon dot based optical probing technique was demonstrated to be simple, cost-effective, sensitive and selective for detecting biologically significant VB₁₂. The present work may open a pathway for a facile and fast method to detect other small biomolecules through the utilization of thermally-reduced carbon dots as the novel donor in FRET optical probes.

Acknowledgements

The authors would like to acknowledge the support from NSF grant #1228127. The authors thank Dr Brandon Weeks for the UV-vis spectra characterization and Dr Hope Weeks for their fluorescence spectra characterization.

Notes and references

1. J. Zhang, Z. G. Tian, L. Liang, M. Subirade and L. Y. Chen, J. Phys. Chem. B, 2013, 117, 14018–14028.
2. E. Alphandery, S. Faure, O. Seksek, F. Guyot and I. Chebbi, ACS Nano, 2011, 5, 6279–6296.
3. E. Alphandery, C. Carvallo, N. Menguy and I. Chebbi, J. Phys. Chem. C, 2011, 115, 11920–11924.
4. P. Tomcik, C. E. Banks, T. J. Davies and R. G. Compton, Anal. Chem., 2004, 76, 161–165.
5. J. H. Chen and S. J. Jiang, J. Agric. Food Chem., 2008, 56, 1210–1215.
6. Y. K. Zhou, H. Li, Y. Liu and G. Y. Liang, Anal. Chim. Acta, 1991, 243, 127–130.
7. S. Wongyai, J. Chromatogr. A, 2000, 870, 217–220.
8. K. Akatsuka and I. Atsuva, Fresenius' Z. Anal. Chem., 1989, 335, 200–204.
9. F. K. Du, Y. H. Ming, F. Zeng, C. M. Yu and S. Z. Wu, Nanotechnology, 2013, 24, 365101.
10. C. M. Yu, X. Z. Li, F. Zeng, F. Y. Zheng and S. Z. Wu, Chem. Commun., 2013, 49, 403–405.
11. L. S. Fan, Y. W. Hu, X. Wang, L. L. Zhang, F. H. Li, D. X. Han, Z. G. Li, Q. X. Zhang, Z. X. Wang and L. Niu, Talanta, 2012, 101, 192–197.
12. S. H. Su, S. R. Wang and J. J. Qiu, Sci. Adv. Mater., 2014, 6, 203–208.
13. F. Aldeek, X. Ji and H. Mattoussi, J. Phys. Chem. C, 2013, 117, 15429–15437.
14. X. M. Sun, Z. Liu, K. Welsher, J. T. Robinson, A. Goodwin, S. Zaric and H. J. Dai, Nano Res., 2008, 1, 203–212.
15. Q. Qu, A. W. Zhu, X. L. Shao, G. Y. Shi and Y. Tian, Chem. Commun., 2012, 48, 5473–5475.
16. J. L. Wang, Z. Zhou, X. Huang, L. W. Zhang, B. T. Hu, S. Moyo, J. Sun and Y. P. Qiu, J. Adhes. Sci. Technol., 2013, 27, 1278–1288.
17. H. J. Sun, L. Wu, N. Gao, J. S. Ren and X. G. Qu, ACS Appl. Mater. Interfaces, 2013, 5, 1174–1179.
18. Z. Zhou, X. C. Liu, B. T. Hu, J. L. Wang, D. W. Xin, Z. J. Wang and Y. P. Qiu, Surf. Coat. Technol., 2011, 205, 4205–4210.
19. L. L. Li, J. Ji, R. Fei, C. Z. Wang, Q. Lu, J. R. Zhang, L. P. Jiang and J. J. Zhu, Adv. Funct. Mater., 2012, 22, 2971–2979.
20. J. H. Shen, Y. H. Zhu, C. Chen, X. L. Yang and C. Z. Li, Chem. Commun., 2011, 47, 2580–2582.
21. J. H. Wei, J. J. Qiu, L. Li, L. Q. Ren, X. W. Zhang, J. Chaudhuri and S. R. Wang, Nanotechnology, 2012, 23, 335707.
22. Z. Zhou, J. L. Wang, X. Huang, L. W. Zhang, S. Moyo, S. Y. Sun and Y. P. Qiu, Appl. Surf. Sci., 2012, 258, 4411–4416.
23. J. Wei, S. Jacob and J. Qiu, Compos. Sci. Technol., 2014, 92, 126–133.
24. E. Vaishnavi and R. Renganathan, Spectrochim. Acta, Part A, 2013, 115, 603–609.
25. Y. Q. Dong, J. W. Shao, C. Q. Chen, H. Li, R. X. Wang, Y. W. Chi, X. M. Lin and G. N. Chen, Carbon, 2012, 50, 4738–4743.
26. Y. Wang, Q. Z. Lu, S. L. Wu, B. L. Karger and W. S. Hancock, Anal. Chem., 2011, 83, 3133–3140.
27. S. N. Baker and G. A. Baker, Angew. Chem., Int. Ed., 2010, 49, 6726–6744.
28. L. B. Tang, R. B. Ji, X. K. Cao, J. Y. Lin, H. X. Jiang, X. M. Li, K. S. Teng, C. M. Luk, S. J. Zeng, J. H. Hao and S. P. Lau, ACS Nano, 2012, 6, 5102–5110.
29. S. J. Zhu, J. H. Zhang, S. J. Tang, C. Y. Qiao, L. Wang, H. Y. Wang, X. Liu, B. Li, Y. F. Li, W. L. Yu, X. F. Wang, H. C. Sun and B. Yang, Adv. Funct. Mater., 2012, 22, 4732–4740.
30. D. Y. Pan, L. Guo, J. C. Zhang, C. Xi, Q. Xue, H. Huang, J. H. Li, Z. W. Zhang, W. J. Yu, Z. W. Chen, Z. Li and M. H. Wu, J. Mater. Chem., 2012, 22, 3314–3318.
31. W. L. Wilson, P. F. Szajowski and L. E. Brus, Science, 1993, 262, 1242–1244.
32. Y. P. Sun, B. Zhou, Y. Lin, W. Wang, K. A. S. Fernando, P. Pathak, M. J. Meziani, B. A. Harruff, X. Wang, H. F. Wang, P. J. G. Luo, H. Yang, M. E. Kose, B. L. Chen, L. M. Veca and S. Y. Xie, J. Am. Chem. Soc., 2006, 128, 7756–7757.
33. G. Eda, Y. Y. Lin, C. Mattevi, H. Yamaguchi, H. A. Chen, I. S. Chen, C. W. Chen and M. Chhowalla, Adv. Mater., 2010, 22, 505–509.
34. A. Kundu, S. Nandi, R. K. Layek and A. K. Nandi, ACS Appl. Mater. Interfaces, 2013, 5, 7392–7399.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4nj00538d

---