Letter Cite This: Nano Lett. 2017, 17, 7710−7716
pubs.acs.org/NanoLett
Tracking the Source of Carbon Dot Photoluminescence: Aromatic Domains versus Molecular Fluorophores Florian Ehrat,†,‡ Santanu Bhattacharyya,†,‡ Julian Schneider,§ Achim Löf,# Regina Wyrwich,∥ Andrey L. Rogach,§ Jacek K. Stolarczyk,*,†,‡ Alexander S. Urban,*,†,‡ and Jochen Feldmann†,‡
Downloaded via UNIV OF CONNECTICUT on June 15, 2018 at 10:00:58 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Chair for Photonics and Optoelectronics, Department of Physics and Center for NanoScience (CeNS), Ludwig-Maximilians-Universität München, Amalienstraße 54, 80799 Munich, Germany ‡ Nanosystems Initiative Munich (NIM), Schellingstraße 4, 80799 Munich, Germany § Department of Materials Science and Engineering and Center for Functional Photonics (CFP), City University of Hong Kong, Hong Kong, Hong Kong SAR # Chair of Applied Physics, Department of Physics and Center for NanoScience (CeNS), Ludwig-Maximilians-Universität München, Amalienstraße 54, 80799 Munich, Germany ∥ Department of Chemistry, Ludwig-Maximilians-Universität München, Butenandtstr. 5-13 (E), 81377 Munich, Germany S Supporting Information *
ABSTRACT: Carbon dots (CDs) are an intriguing fluorescent material; however, due to a plethora of synthesis techniques and precursor materials, there is still significant debate on their structure and the origin of their optical properties. The two most prevalent mechanisms to explain them are based on polycyclic aromatic hydrocarbon domains and small molecular fluorophores, for instance, citrazinic acid. Yet, how these form and whether they can exist simultaneously is still under study. To address this, we vary the hydrothermal synthesis time of CDs obtained from citric acid and ethylenediamine and show that in the initial phase molecular fluorophores, likely 2-pyridone derivatives, account for the blue luminescence of the dots. However, over time, while the overall size of the CDs does not change, aromatic domains form and grow, resulting in a second, faster decay channel at similar wavelengths and also creating additional lower energetic states. Electrophoresis provides further evidence that the ensemble of CDs consists of several subsets with different internal structure and surface charge. The understanding of the formation mechanism enables a control of the chemical origin of these emitters and the ensuing optical properties of the CDs through synthetic means. KEYWORDS: Carbon dots, photoluminescence, fluorophore, polycyclic aromatic hydrocarbons, optical spectroscopy, electrophoresis
W
ithin a short time since their discovery in 2004,1 carbon dots (CDs) have found widespread applications in light emission,2 photocatalysis,3−6 bioimaging,7 and sensing.8,9 This rapid rise was enabled by a number of advantageous properties, in particular high photoluminescence quantum yield (PL QY), facile and flexible synthesis, biocompatibility, and photo- and water stability.10−13 Setting them apart from inorganic quantum dots, CDs exhibit an unusually large Stokes shift, while the peak PL emission often, but not always, varies with excitation wavelength.14,15 These properties depend strongly on the precursors, preparative conditions, and postsynthetic treatment, indicating a complex internal and surface structure of the CDs.16 This inherent complexity is the reason why a straightforward explanation of the origin of the optical properties of CDs has to date proven elusive. Nonetheless, several models have emerged aiming to account for the commonly observed features. One of the first approaches envisioned energy transfer between distinct states © 2017 American Chemical Society
associated with the core and the surface of the CD as the likely source of the observed optical properties.17,18 While functional groups occupy the surface, the carbogenic core itself is usually considered to contain sp2-hybridized domains immersed in an amorphous sp3-hybridized matrix.7,8 In this context, it was shown that a combination of just three small polycyclic aromatic hydrocarbons (PAHs) could reproduce the Stokes shift and excitation dependence of the emission wavelength of the CDs.14 Specifically, as the absorption and PL spectra of the PAHs differ, but partially overlap, the excitation-dependent PL can be understood in terms of direct emission and energy transfer. Moreover, exciton self-trapping on π-stacked molecular dimers can fully explain the observed Stokes shift. Another model has emerged in recent studies, which have proposed that Received: September 8, 2017 Revised: November 29, 2017 Published: November 30, 2017 7710
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Letter
Nano Letters molecular fluorophores - derivatives of citrazinic acid - are also formed during the synthesis of the CDs and significantly contribute to their emission.19−26 It has thus been shown that blue-emitting fluorophores, such as citrazinic acid and other 2pyridone-based molecules, can be produced in a common “bottom up” synthesis route, starting from citric acid and an amine-based nitrogen precursor.27,28 Interestingly, citrazinic acid also features a strong Stokes shift, with absorption and emission maxima similar to those seen in CDs. Whether the molecules are free in solution, bound to the CD surface, or incorporated into their structure remains a subject of debate.21,26,29 Irrespective of the mode, these molecules could be responsible for the main blue emission peak of CDs,23,24 although they cannot explain their excitation-dependent luminescence. It appears that there may be several contributions to the PL, originating from PAH domains and/or molecular fluorophores, depending on the preparative procedure. For instance, at high synthesis temperatures, aromatic domains dominate the PL.21,25 At the same time, it is not clear whether the overall PL is a result of population averaging or all individual dots contain different kinds of emitters. In this Letter we readdress the preparative procedure and systematically vary one synthesis parameter, namely, the synthesis time of CDs made from two of the most common precursors, citric acid and ethylenediamine, in order to study the evolution and chemical nature of the emitters during the formation process (cf. Figure 1). We show that the molecular
3, 5, and 10 h. The UV−vis absorption spectra, taken from 1000-fold diluted solutions of the CDs and shown in Figure 2a, feature only a small shoulder at 250 nm for the early stage samples (15 and 30 min). Another peak at 350 nm arises for the samples from 1 h onward, maximizing in intensity at 3 h and decreasing slightly for longer synthesis times. In addition, a broad shoulder emerges at 450 nm after 3 h but remains more or less constant in intensity for the 5 and 10 h samples. The optical densities (ODs) of the two bands most commonly observed in CDs, at 350 and 450 nm, are plotted in dependence of the synthesis time in Figure 2b. The PL spectra, acquired under 350 nm excitation and at an OD of 0.1 cm−1, feature a broad asymmetric peak with a maximum at 450 nm that looks virtually identical in all samples (see Figure 2c and Figure S1). The PL intensity, nearly nonexistent at 15 min, increases to a maximum for the 1 h sample and decreases for the samples grown for longer times. Consequently, the PL QY reaches its maximum of 66% for the 1 h sample and then decreases but retains relatively high values in the range of 40 to 50% (Figure 2d). Only for the initial sample is the QY negligible with a value of 1%, suggesting that the component responsible for the high PL QY forms during the first 30 min of the synthesis. Exciting the sample at the absorption shoulder at 450 nm results in a considerably lower QY of 7% and 13% for the 30 min and 1 h samples, respectively, followed by a gradual increase to reach a plateau at 16−17% for longer times (cf. Figure S2). Given the differences in PL QY, it is intriguing that the positions of the PL maxima shift with excitation in the same manner for all but the 15 min sample (see Figure 2e). In line with earlier observations, there is an excitation-independent region below 380 nm and a nearly continuous red-shift for longer excitation wavelengths.14 However, PL excitation (PLE) spectra taken at 450, 510, and 550 nm reveal further differences (see Figure S3). For the 1 h sample, the spectra are virtually identical, regardless of emission wavelength, with a single feature at 350 nm coinciding with the absorption peak. In contrast, for the 5 h sample there are three features at 350, 450, and 520 nm for the emission at 550 nm, similar to our previous study.14 This suggests the formation of additional lower energetic states later in the synthesis. To further analyze the PL origin, we used time-correlated single photon counting (TCSPC) to determine the PL lifetimes for 350 nm excitation (Figure 2f). PL was found to decay exponentially for the 30 min and 1 h samples, with a lifetime of 15.3 ns. For longer synthesis times, the decay becomes faster but also more complex and can no longer be reproduced with a simple exponential function. Together with a decrease in QY, this implies an increase in the nonradiative recombination rate. Interestingly, in a biexponential fit, these samples still show a long time decay component of 15.3 ns along with a faster decay constant of 5.1 ns. This suggests that while only one emissive species is present after 1 h of the synthesis, a second one forms later on. These two lifetimes could stem either from different types of noninteracting substructures within a single CD or from two subsets of CDs in the ensemble. The synthesis time is often used to control the size of colloidal nanocrystals, which in turn determines the quantum confinement and thereby their optical properties.30 To investigate this factor, we attempted to compare the CD size by high-resolution transmission electron microscopy (HRTEM). We were able to observe some single nanoparticles as well as larger clusters of material, likely of CDs, as depicted for the 5 h sample in Figure 3a. Magnifying these images, we were
Figure 1. Schematic illustration of the possible formation of sp2hybridized aromatic domains or molecular fluorophores during the synthesis of CDs from citric acid and ethylenediamine.
fluorophores form early within and around the CDs and initially dominate the PL, but over time the carbonization leads to growth of aromatic domains with the molecular fluorophores likely acting as seeds for the aromatic domain formation. Electrophoretic and fluorescence lifetime measurements reveal that the CDs comprise several different subspecies with different surface charges, albeit with similar absorption and fluorescence spectra. These measurements along with additional X-ray photoelectron spectroscopy (XPS) data further reveal how these different subpopulations develop over time. The results imply that it is possible to control the origin of the emitters and thereby selectively tune the optical properties of the CDs through simple synthetic means. The hydrothermal synthesis of the CDs was carried out with citric acid/ethylenediamine as previously described,8,14 and the reaction was stopped after different times: 15 and 30 min and 1, 7711
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Letter
Nano Letters
Figure 2. (a) Absorption spectra of six CD samples prepared by varying the duration of the synthesis and (b) the corresponding ODs at 350 nm (blue) and 450 nm (red). (c) PL spectra of CDs excited at 350 nm, where the ODs of all samples were adjusted to 0.1 cm−1 and (d) the corresponding QY. The inset shows a photograph of the samples under UV illumination. (e) Spectral dependence of the PL emission maximum as a function of the excitation wavelength. (f) 450 nm PL decay dynamics after excitation at 350 nm. The dotted lines serve as guides to the eye to show the different exponential decay components. The color-coding in panels (c), (e), and (f) is identical to that in panel (a); while in panel (b) it refers to the left and right axes.
(which also contains aromatic nitrogen) is much lower (Figure 3e, lower panel). This suggests that the cores of the CDs predominantly comprise sp2 and sp3 carbon. As shown in Figure 3f, the total amount of carbon−carbon bonds increases with synthesis time, as could be expected for a carbonization process. Interestingly, the CC bonds only start to appear for the 3 h sample, while the C−C ratio increases from the start showing values of more than 10% already in the 15 min sample. Another dominant feature in the C 1s spectra, the CN peak, reaches a maximum for the 1 h sample and significantly decreases for the later samples. This trend correlates well with the change in PL QY during synthesis (cf. Figure 2d), suggesting that CN plays a role in the high PL emission of the CDs. At the same time, the CC content trend coincides with the appearance of the absorption shoulder at 450 nm (Figure 2a) and the long wavelength peaks in the PLE spectra (Figure S3). We also performed Fourier-transform infrared (FTIR) spectroscopy on all of the CD samples dispersed in heavy water to confirm the conclusions drawn from the XPS measurements (cf. Figure S8). In general, there are three regimes, which can be monitored since they are not affected by the absorption of D2O. These are between 900 and 1150 cm−1, 1300 and 2200 cm−1, and 2800 and 4000 cm−1. In the latter regime, stretching vibrations of O−H and N−H bonds lead to a wide peak around 3400 cm−1. A second, smaller one is present around 2950 cm−1 and can be attributed to C−H bonds for sp3 hybridized carbon atoms. Both peaks reduce in intensity over the whole synthesis time with a significant drop between 30 min and 1 h. These vibrations can be attributed to surface groups and are also consistent with the presence of the remaining precursor material (citric acid and ethylenediamine)
able to observe clear lattice fringes, suggesting crystalline domains roughly a couple of nanometers in size (Figure 3b). Interestingly, there were nearly no crystalline dots in the 30 min and 1 h samples, while we found many crystalline particles of a few nanometers size in different locations, either within a cluster or isolated on the grid in the 5 h sample. This strongly suggests that crystallinity in the dots increases with synthesis time. While we could give a rough estimate of the CD size, unfortunately, the poor contrast in these images, resulting from the fact that we are imaging predominantly carbon structures on carbon grids, does not allow for a precise determination of the actual CD sizes. To this end, we applied atomic force microscopy (AFM) after dispersing them on mica substrates (see Figure 3c). Although the CDs tend to aggregate on substrates (cf. Figure S4), these aggregates are easily distinguishable from individual CDs, so that reliable height profiles of the CDs could be obtained by AFM. The resulting histograms (Figures 3d and S5) clearly show no difference in particle height between the samples. XPS measurements on C 1s, N 1s, and O 1s signals (Figure S6) corroborate the argument that the composition, rather than the size, of the CDs changes during the synthesis. While the pristine samples all show strong contributions from C−O/C− N, CO, and CN bonded carbon (Figure S7), as expected for the surface functional groups, sputtering the samples with Ar+ reveals differences in the exposed cores. Individual contributions from C−C, CC, CN, C−N/C−O, and CO bonded carbon (Figure 3e, top panel) can be identified for the 1 h sample, albeit with a very weak CC signal. The 5 h sample has the same components, but the contribution from C−C and CC is significantly higher, while that from CN 7712
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Letter
Nano Letters
Figure 3. (a) HRTEM image of CDs from the 5 h sample and (b) magnification of the image in (a) corresponding to the dashed box. Lattice fringes can be seen clearly, suggesting crystalline domains. (c) AFM image of CDs synthesized for 3 h and (d) height distributions of CDs determined from AFM images. (e) XPS C 1s spectra of the 1 and 5 h samples, sputtered with Ar+ ions. (f) Evolution of the relative amount of C−C, CC, and CN bonded C atoms inside the CDs during the synthesis obtained from XPS measurements.
presented in Figure 4a.31 Size effects could be excluded in the case of the CDs, as these were shown to be of the same size by AFM (Figure 3b). After separation, brown traces are visible under white light illumination, indicating the position of the CDs. Unexpectedly, we found drift in both directions for all samples. This means that species with either positively or negatively charged surfaces are simultaneously present in the ensemble. Significant differences are clearly visible under UV irradiation (Figure 4b). Specific bands appear in the samples toward the positive and negative sides, with the positions identical for the different synthesis times. The only exception is the band at the most negative position (which will be denoted −4 later on), which is present in all but the 1 h sample. Interestingly, the bands visible under UV do not overlap exactly with the traces seen under white light illumination, suggesting that the PL QY varies between them. The gels were subsequently cut into several pieces, which were labeled by numbers ranging from +4 to −4, indicating the position and representing the relative charge and its sign. The pieces were freeze-dried to extract the CDs with spin columns in order to enable a separate analysis using spectroscopic methods (see Figure 4c for the 5 h sample and Figure S9 for comparison with the 1 h sample). The absorption peak around 350 nm is present for all samples, with a weak red-shift for the most positively charged species (denoted +4), shown for the 5 h sample. However, the shoulder around 450 nm varies in intensity relative to the main absorption peak for different gel positions. It is highest around the starting point (denoted “0”),
in the solution. This would explain the observed decrease in intensity with synthesis time. Similar arguments can be put forward for the C−N and C−O bond stretching vibration modes observed around 1080 cm−1. The peaks at 1413 and 1456 cm−1, which also decrease in the initial stages of the synthesis, can be potentially assigned to C−H bending vibrations in the citric acid and ethylenediamine. The strong carbonyl stretch at 1706 cm−1, which increases and then decreases, is in agreement with the formation of 2-pyridone derivatives (such as citrazinic acid) in the early stages of the synthesis and their subsequent decomposition. This matches also the presence of the 1583 cm−1 peak, typical for Ncontaining aromatic rings, in particular 2-pyridone/2-pyridinol and its derivatives, such as citrazinic acid. Therefore, these peaks are consistent with the formation of 2-pyridone-based fluorescent molecules in the synthesis. There are also two C C stretching peaks that increase with synthesis time and indicate protracted formation of isolated and conjugated CC bonds (1652 cm−1) and conjugated aromatic rings (1546 cm−1). The latter peak agrees very well with the position of CC stretching vibration, for example, in pyrene. Together, these peaks are consistent with the formation of the aromatic domains during the synthesis. In order to elucidate whether there are subsets of CDs with varying optical properties present in the ensembles, we performed gel electrophoresis on the 1, 3, and 5 h CD samples. Generally, in this method, particles are separated in the gel according to their surface charge and their size, as 7713
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Letter
Nano Letters
(denoted −4). This shows that the CD ensemble spectra, shown previously in Figure 2a, contain contributions from all the components now separated in the gel. There is no particular variation in the PL peak position and shape after the electrophoresis (cf. Figure S10), but their QYs differ substantially, especially for the excitation at 350 nm (see Figure S11). The PL QY, estimated by the ratio of integrated PL intensity over the optical density at the excitation wavelength, is very high for the bands at the extreme parts of the gel, but three to five-fold lower in the middle of the gel. Consequently, the particles located at these most negative/ positive positions seem to have the largest influence on the PL properties of the entire sample. In addition, it cannot be excluded that there are aggregated CDs present close to the zero positions, leading to a quenching of the fluorescence. To further explore these different regions, we acquired timeresolved PL of material procured from individual positions (between +4 and −4) for the 1, 3, and 5 h samples (Figures 4d and S10), except for the absent −4 feature in the 1 h sample. The PL decay for the +4 positions is nearly identical for the three samples, with an exponential behavior and a decay time of 15.3 ns. On the extreme negative side (position −4) of the 3 and 5 h samples, the decay was also exponential with a faster characteristic time of roughly 8 ns (cf. Figure S12a). However, the decay profiles for all other intermediate positions could only be fitted with a biexponential function. The long decay constant in all of these was again close to 15 ns, while the second one was much faster and tended to be around 5 ns with varying contributions from the two components (further details in Figure S13 and Table S1). Crucially, the 15.3 ns component is the same in the exponential decay observed for the early time 30 min and 1 h ensemble samples (Figure 2f) and for the +4 positions of the 1, 3, and 5 h (Figure S12b) samples extracted after the gel electrophoresis. The ensembles grown for longer times all exhibit a biexponential decay with a long component of 15.3 ns and a short component of 5.1 ns, but it is clear that the long decay component becomes less prominent, while the faster channel rises in contribution during the synthesis. This is consistent with the results obtained for all of the other positions extracted from the gel electrophoresis. Interestingly, the long lifetime component lies close to that previously observed for citrazinic acid derivative fluorophores, such as IPCA.26 Consequently, there appear to be two different types of emissive species absorbing at 350 nm and emitting around 450 nm present in the CDs. These could be molecular fluorophores and PAH domains (Figure 5). The former, likely 2-pyridone derivatives such as citrazinic acid or IPCA,21,23,27 form early during synthesis and lead to a rapid rise in the PL emission and high PL QY values of the CDs. Electrophoresis does not separate these molecules from the CDs, suggesting that they are either embedded in the amorphous matrix or bound to the surface. Later, these molecules give way to the growth of aromatic domains comprising carbon and nitrogen. Initially, these domains show similar spectral characteristics, but faster decay times. As they grow, a shoulder at longer wavelengths appears in the absorption spectra. This agrees well with an increase in the contribution of both sp3- and sp2-hybridized carbon−carbon bond formation in the overall C 1s signal derived from XPS measurements. At later stages, the carbonization renders the cores nearly void of nitrogen and oxygen, which can be found only at the surface in the functional groups. Protonation or deprotonation of these groups can imbue the CDs with an overall positive or negative surface charge,
Figure 4. Postelectrophoresis photographs of the gel containing 1, 3, and 5 h samples taken under (a) ambient and (b) UV illumination and compared to that of the 5 h ensemble spectrum (gray dashed line). (c) Absorption spectra of the 5 h sample obtained for the most positive, most negative, and neutral positions in the gel, respectively, marked with squares in (a) and (b). (d) Time-resolved PL decay for the samples marked in (a) and (b) with squares of corresponding colors. The gray data points denote the 5 h sample before electrophoresis.
decreases for migrated particles in both directions, and is completely absent for the most negatively charged components 7714
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Nano Letters
■
Letter
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected] Tel.: +49 89 2180-2039 Fax: +49 89 2180-3441. *E-mail:
[email protected] Tel.: +49 89 2180-1356. ORCID
Andrey L. Rogach: 0000-0002-8263-8141 Alexander S. Urban: 0000-0001-6168-2509 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors would like to thank Karl-Heinz Mantel for helping with the FTIR measurements and Markus Döblinger for his help with the TEM measurements and analysis. This work was supported by the Bavarian State Ministry of Science, Research, and Arts through the grant “Solar Technologies go Hybrid (SolTech)”, by the German Academic Exchange Service (DAAD) through a PPP Hong Kong 2017 grant (Project No: 57320254) and the related grant from the Germany/Hong Kong Joint Research Scheme sponsored by the Research Grants Council of Hong Kong and the German Academic Exchange Service (Reference No.: G-CityU103/16), the LMUexcellent Junior Researcher Fund (to A.S.U), and by the Alexander von Humboldt foundation (to S.B.).
Figure 5. Carbon content in the CDs (left) and ratio between long and short decay components of the fluorescence lifetime (right) over the course of the synthesis of the CDs. These parameters could represent the relative content of small molecular fluorophores similar to citrazinic acid (shown in a blue circle) and aromatic domains (represented in a red circle).
explaining the observations made in the gel electrophoresis. The concomitant decrease in the contribution of the slower decay channel in PL, as discussed above, is shown in Figure 5 (blue line). In summary, we have investigated the formation and development of the internal structure and resulting optical properties of CDs synthesized via a hydrothermal reaction from the very common precursors, citric acid and ethylenediamine. Arresting the synthesis at specific times, we find that the CDs grow rapidly within 30 min and retain the size for the rest of the synthesis. Instead of increasing in size, they undergo a substantial change of their internal structure. We show that early on mainly fluorescent molecules form, possibly similar in structure to citrazinic acid and yielding fluorescence in the blue region, with high PL QY and long PL lifetime. With time, aromatic domains appear, possibly with the molecular fluorophores acting as seeds, which initially absorb and emit in the same spectral range, albeit with a faster PL decay. In time, some of the aromatic domains grow in size, leading to additional absorption and PL at longer wavelengths. Importantly, the rapid formation of aromatic domains is not a general feature of carbonization processes but seems to be enhanced by specific precursors. The elucidated mechanism of the formation of CDs shows that it is possible to control the type and optical properties of the intrinsic emitters through a simple modification of the synthetic conditions. This provides a flexible method to tune the emissive properties of CDs for a specific application and suggests possible pathways for a wider range of syntheses.
■
■
REFERENCES
(1) Xu, X.; Ray, R.; Gu, Y.; Ploehn, H. J.; Gearheart, L.; Raker, K.; Scrivens, W. A. J. Am. Chem. Soc. 2004, 126, 12736−12737. (2) Joseph, J.; Anappara, A. A. J. Lumin. 2016, 178, 128−133. (3) Li, H.; He, X.; Kang, Z.; Huang, H.; Liu, Y.; Liu, J.; Lian, S.; Tsang, C. H. A.; Yang, X.; Lee, S.-T. Angew. Chem., Int. Ed. 2010, 49, 4430−4434. (4) Fernando, K. A. S.; Sahu, S.; Liu, Y.; Lewis, W. K.; Guliants, E. A.; Jafariyan, A.; Wang, P.; Bunker, C. E.; Sun, Y.-P. ACS Appl. Mater. Interfaces 2015, 7, 8363−8376. (5) Yang, P.; Zhao, J.; Wang, J.; Cui, H.; Li, L.; Zhu, Z. RSC Adv. 2015, 5, 21332−21335. (6) Bhattacharyya, S.; Ehrat, F.; Urban, P.; Teves, R.; Wyrwich, R.; Döblinger, M.; Feldmann, J.; Urban, A. S.; Stolarczyk, J. K. Nat. Commun. 2017, 8, 1401. (7) Kang, Y.-F.; Li, Y.-H.; Fang, Y.-W.; Xu, Y.; Wei, X.-M.; Yin, X.-B. Sci. Rep. 2015, 5, 11835. (8) Zhu, S.; Meng, Q.; Wang, L.; Zhang, J.; Song, Y.; Jin, H.; Zhang, K.; Sun, H.; Wang, H.; Yang, B. Angew. Chem., Int. Ed. 2013, 52, 3953−3957. (9) Nie, H.; Li, M.; Li, Q.; Liang, S.; Tan, Y.; Sheng, L.; Shi, W.; Zhang, S. X.-A. Chem. Mater. 2014, 26, 3104−3112. (10) Sun, Y.-P.; Zhou, B.; Lin, Y.; Wang, W.; Fernando, K. A. S.; Pathak, P.; Meziani, M. J.; Harruff, B. A.; Wang, X.; Wang, H.; Luo, P. G.; Yang, H.; Kose, M. E.; Chen, B.; Veca, L. M.; Xie, S.-Y. J. Am. Chem. Soc. 2006, 128, 7756−7757. (11) Kozák, O.; Sudolská, M.; Pramanik, G.; Cígler, P.; Otyepka, M.; Zbořil, R. Chem. Mater. 2016, 28, 4085−4128. (12) Baker, S. N.; Baker, G. A. Angew. Chem., Int. Ed. 2010, 49, 6726−6744. (13) Lim, S. Y.; Shen, W.; Gao, Z. Chem. Soc. Rev. 2015, 44, 362− 381. (14) Fu, M.; Ehrat, F.; Wang, Y.; Milowska, K. Z.; Reckmeier, C.; Rogach, A. L.; Stolarczyk, J. K.; Urban, A. S.; Feldmann, J. Nano Lett. 2015, 15, 6030−6035. (15) Dong, Y.; Pang, H.; Yang, H. B.; Guo, C.; Shao, J.; Chi, Y.; Li, C. M.; Yu, T. Angew. Chem., Int. Ed. 2013, 52, 7800−7804.
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b03863. Experimental methods, additional UV−vis, PL, PLE, FTIR, and XPS spectra, PL QY and lifetime plots, AFM images, and AFM height profiles (PDF) 7715
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
Letter
Nano Letters (16) Zhu, S. J.; Meng, Q. N.; Wang, L.; Zhang, J. H.; Song, Y. B.; Jin, H.; Zhang, K.; Sun, H. C.; Wang, H. Y.; Yang, B. Angew. Chem., Int. Ed. 2013, 52, 3953−3957. (17) Ding, H.; Yu, S.-B.; Wei, J.-S.; Xiong, H.-M. ACS Nano 2016, 10, 484−491. (18) Yu, P.; Wen, X.; Toh, Y.-R.; Tang, J. J. Phys. Chem. C 2012, 116, 25552−25557. (19) Sharma, A.; Gadly, T.; Neogy, S.; Ghosh, S. K.; Kumbhakar, M. J. Phys. Chem. Lett. 2017, 8, 1044−1052. (20) Shi, L.; Yang, J. H.; Zeng, H. B.; Chen, Y. M.; Yang, S. C.; Wu, C.; Zeng, H.; Yoshihito, O.; Zhang, Q. Nanoscale 2016, 8, 14374− 14378. (21) Wang, W.; Wang, B.; Embrechts, H.; Damm, C.; Cadranel, A.; Strauss, V.; Distaso, M.; Hinterberger, V.; Guldi, D. M.; Peukert, W. RSC Adv. 2017, 7, 24771−24780. (22) Gude, V.; Das, A.; Chatterjee, T.; Mandal, P. K. Phys. Chem. Chem. Phys. 2016, 18, 28274. (23) Schneider, J.; Reckmeier, C. J.; Xiong, Y.; von Seckendorff, M.; Susha, A. S.; Kasák, P.; Rogach, A. L. J. Phys. Chem. C 2017, 121, 2014−2022. (24) Righetto, M.; Privitera, A.; Fortunati, I.; Mosconi, D.; Zerbetto, M.; Curri, M. L.; Corricelli, M.; Moretto, A.; Agnoli, S.; Franco, L.; Bozio, R.; Ferrante, C. J. Phys. Chem. Lett. 2017, 8, 2236−2242. (25) Xiong, Y.; Schneider, J.; Reckmeier, C. J.; Huang, H.; Kasák, P.; Rogach, A. L. Nanoscale 2017, 9, 11730−11738. (26) Song, Y.; Zhu, S.; Zhang, S.; Fu, Y.; Wang, L.; Zhao, X.; Yang, B. J. Mater. Chem. C 2015, 3, 5976−5984. (27) Kasprzyk, W.; Bednarz, S.; Zmudzki, P.; Galica, M.; Bogdal, D. RSC Adv. 2015, 5, 34795−34799. (28) Zhu, S.; Zhao, X.; Song, Y.; Lu, S.; Yang, B. Nano Today 2016, 11, 128−132. (29) Fang, Q.; Dong, Y.; Chen, Y.; Lu, C.-H.; Chi, Y.; Yang, H.-H.; Yu, T. Carbon 2017, 118, 319−326. (30) Alivisatos, A. P. Science 1996, 271, 933. (31) Southern, E. M. J. Mol. Biol. 1975, 98, 503−517.
7716
DOI: 10.1021/acs.nanolett.7b03863 Nano Lett. 2017, 17, 7710−7716
