Based Dynamic Fluorescent Materials - Wiley Online Library

FULL PAPER Phototunable Fluorescence

www.afm-journal.de

Phototunable Full-Color Emission of Cellulose-Based Dynamic Fluorescent Materials Weiguo Tian, Jinming Zhang,* Jian Yu, Jin Wu, Jun Zhang,* Jiasong He, and Fosong Wang multicolor materials such as inorganic semiconductor quantum dots,[5] silicon dots,[6] carbon dots,[7,8] perovskite dots,[9] semiconductor polymer dots,[10,11] organic molecular nanoparticles,[12] fluorescent dyes,[13–15] and upconversion nanocrystals.[16] The fluorescence color of these materials can be regulated in the fullcolor gamut by controlling their microstructures, including the particle size,[5–9] aggregation,[10–14] crystalline states,[10–15] or otherwise adjusting their chemical structures, including the elemental constitution,[5,7–9,16] substituted groups,[12–15] conjugated structure,[10–12] degree of polymerization,[10,11] and so forth. These color-regulating measures are based on the mechanisms such as the quantum size effect,[5–8] surface effect,[9,12,16] intramolecular charge transfer,[10–15] etc. However, it should be noted that the fluorescence color of these ready-made materials cannot change and thus they exhibit static fluorescence intrinsically. Different from these static fluorescent materials, some species can dynamically change their fluorescence properties under different stimuli. For instance, the reversible supramolecular assembly[17,18] can control the colors of the fluorescent dyes. Upconversion nanoparticles emit trichromic fluorescence when excited with different laser pulses.[19] Some photochromic photoswitches, e.g., spiropyran, spirooxazine, and diarylethene, are commonly used to regulate the fluorescence of the hybrid materials.[20] These dynamic fluorescent materials have exhibited some competitive advantages over static fluorescence in practical areas such as multiplex analysis, display, and antifalsification. Nevertheless, it is still challenging to obtain tunable fluorescent materials that meet all the following criteria at the same time: (i) wide-spectrum emission, even full color; (ii) dynamic properties of color and intensity that can be finely tuned by external stimuli; (iii) outstanding reversibility and stability; (iv) versatility in material processing. Herein, we describe a phototunable full-color solid fluorescent material based on trichromacy and the dynamically tunable fluorescence resonance energy transfer (FRET) process. By taking advantage of the structure of cellulose, the most abundant biopolymer on earth, the trichromic materials CA-SP (red), CAFITC (green), and CA-Pyr (blue) are obtained by coupling the spiropyran (SP), fluorescein (FITC), and pyrene (Pyr) groups, respectively, with cellulose skeletons. In particular, CA-SP shows a unique photoinduced fluorochromic phenomenon in

An iridescent chameleon-like material that can change its colors under different circumstances is always desired in color-on-demand applications. Herein, a strategy based on trichromacy and the dynamically tunable fluorescence resonance energy transfer (FRET) process to design and prepare these chameleon-like fluorescent materials is proposed. A set of trichromic (red, green, and blue), solid fluorescent materials are synthesized by covalently attaching spiropyran, fluorescein, and pyrene onto cellulose chains independently. After simply mixing them together, a full range of color is realized. The chameleon-like nature of these materials is based on the dynamic tunable FRET process between donors (green and blue) and acceptors (red) in which the energy transfer efficiency can be finely tuned by irradiation. Ultimately, the reversible and nonlinear regulation of fluorescence properties, including color and intensity, is achieved on a timescale recognizable by the naked eye. Benefited by the excellent processability inherited from the cellulose derivatives, the as-prepared materials are feasibly transformed into different forms. Particularly, a fluorescent ink with the complicated fluorescent input–output dependence suggests more than a proof-of-concept; indeed, it suggests a unique method of information encryption, security printing, and dynamic anticounterfeiting.

1. Introduction If a picture is worth a thousand words, a color picture should be worth a million, and a dynamic color picture should be worth even more, similar to comparing video with images and text. This is especially so because color always catches the eye, and the visual perception of human beings is much more effective and conveys more information than other senses.[1–4] Therefore, tunable multicolor fluorescent materials have great application in the color-on-demand areas, including displaying, imaging, sensing, visualized indicating, and anticounterfeiting, etc. Recently, much effort has been devoted to preparing tunable

Dr. W. G. Tian, Dr. J. M. Zhang, Dr. J. Yu, Dr. J. Wu, Prof. J. Zhang, Prof. J. S. He, Prof. F. S. Wang Beijing National Laboratory for Molecular Sciences CAS Key Laboratory of Engineering Plastics Institute of Chemistry Chinese Academy of Sciences (CAS) Beijing 100190, China E-mail: [email protected]; [email protected] W. G. Tian, Prof. J. Zhang University of Chinese Academy of Sciences Beijing 100049, China

DOI: 10.1002/adfm.201703548

Adv. Funct. Mater. 2017, 1703548

1703548 (1 of 8)

© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com

www.afm-journal.de

addition to the photochromism. These trichromic materials inherit the excellent solubility and processability of the original cellulose derivatives, so that they can be readily converted into printing ink. After simply blending the trichromic materials together, the dynamic FRET systems are established without any sophisticated molecular design and integrated synthesis. The ratio of FRET donor and acceptor components can be precisely controlled by a common desktop inkjet printer, resulting in full-color fluorescent materials that change their colors (RGB values) along with the irradiation time. The fluorescence outputs, emission wavelength, and intensity crucially depend on the input variables, including the RGB ratio, excitation time, and intensity. Such a property is perfectly suited for anticounterfeiting applications because the extraordinary complexity of the dynamic nonlinear dependence between the fluorescent inputs and outputs makes the fluorescent printing patterns extremely difficult to duplicate, counterfeit, and reverse engineer. Additionally, the broad compatibility with surface of different substrates also improves the availability of the full-color cellulose-based fluorescent materials in information encryption, security printing, dynamic anticounterfeiting, etc.

trichromic (RGB) fluorescent materials, which are the elementary blocks for full color; (ii) photoinduced fluorochromism, which occurs in at least one of the trichromic fluorescent materials, i.e., the emission and absorption (or excitation) spectra change along with the irradiation time, as the R(t) in Figure 1. Additionally, effective FRET channels have to be established between the fluorochromic components and the rest counterparts. Eventually, when the FRET donors and acceptors are within the 1–10 nm range, the fluorescent intensity of R, G, and B components varies with the irradiation time due to the change in FRET efficiency originating from the spontaneous change of the fluorochromic component concentration over the irradiation time. Hence, the RGB values are described as functions of time (t), R(t), G(t), and B(t). Moreover, the FRET efficiency (spectroscopic overlap) between different color components is distinctly different, which determines the rate and extent of the color shift. The parameters R∞, G0, and B0 in the functions are controlled by the original concentration of the trichromic materials. Theoretically, phototunable dynamic fluorescent materials can be realized in the full-color space on the basis of the above mechanism.

2. Results and Discussion

2.2. Trichromic Materials

2.1. Principle Design

In our recent work,[21] the common aggregation-caused quenching luminogens were successfully converted into excellent solid fluorescent materials through covalently attaching the luminogens onto cellulose skeletons. As a result, the synergy of

To achieve phototunable full-color dynamic fluorescence based on trichromacy and FRET, there are two prerequisites: (i)

Figure 1. Schematic mechanism of the phototunable full-color dynamic fluorescent materials. R(t), the dynamic fluorescent component; G0 and B0, the static fluorescent counterparts.


1703548 (2 of 8)



www.afm-journal.de

the anchoring and diluting effects of cellulose chains and the electrostatic repulsion could effectively mitigate the aggregationcaused quenching phenomenon. By a similar strategy, the cellulose-based solid fluorescent materials, CA-SP, CA-FITC, and CA-Pyr, which correspondingly emit red, green, and blue fluorescence, are synthesized independently. Their chemical structures and fluorescent images (under 365 nm) are shown in Figure 2a. From the normalized emission spectra in Figure 2b, it is clear that the emission peaks of these trichromic materials are separated: CA-SP, λmax = 633 nm and FWHM (full width at half maxima) = 72 nm; CA-FITC, λmax = 525 nm and FWHM = 65 nm; and CA-Pyr, λmax = 387 nm (shoulder peak at 403 nm relating to the aggregation state) and FWHM = 40 nm approximately. In Figure 2c, the UV–vis absorption spectrum of CA-SP has two characteristic absorption bands: the range 500–700 nm (λmax = 568 nm) is the absorption of the ring-open form of spiropyran (merocyanine, MC; the

excitation spectrum of CA-SP, the dash line in Figure 2c, almost coincides with the absorption band of MC), and the band below 500 nm, the ring-closed form of spiropyran (SP). These absorption bands exactly overlap with the emission peaks of CA-FITC and CA-Pyr. These results indicate that effective FRET channels can be established between CA-FITC, CA-Pyr, and CA-SP. The fluorescent properties over different UV irradiation time are shown in Figure 2a,d,e. Apparently, the fluorescent intensity of CA-SP increases with prolonging the UV irradiation time. The kinetic linear fitting plot inserted in Figure 2e suggests that the fluorescence variation of CA-SP follows first-order kinetics (Supporting Information) as a result of the isomerization of spiropyran in CA-SP, which generally is a typical first-order kinetic reaction.[22] In contrast, CA-FITC and CA-Pyr almost fully retain their initial fluorescence intensity (images in Figure 2a, the spectra in Figure 2e; Figure S7 in the Supporting Information) over a long UV irradiation time. Consequently, the transitional

Figure 2. Photophysical study of the trichromic components. a) Chemical composition of the cellulose-based trichromatic solid fluorescent materials (CA-SP, CA-FITC, and CA-Pyr), and their fluorescent images change with the length of UV irradiation (365 nm); b) the emission spectra of the trichromatic materials (CA-SP, CA-FITC, and CA-Pyr); c) the overlap between the emission spectra of CA-FITC and CA-Pyr, and the absorption spectrum or the excitation spectrum (dash line) of CA-SP; d) the gradual change in the emission spectra of CA-SP 20 with the prolongation of UV irradiation time (365 nm, the light source in spectrometer); e) the emission intensity alternation at λmax of CA-SP 20 (λmax = 635 nm), CA-FITC (λmax = 545 nm), and CA-Pyr (λmax = 409 nm), the insert plot is the first-order kinetic fitting result of the emission intensity change of CA-SP 20 at λmax = 635 nm.


1703548 (3 of 8)



www.afm-journal.de

FRET processes are accomplished with the dynamic component, CA-SP and the static counterparts, CA-FITC, and CA-Pyr, which act as the elementary constituents of the phototunable full-color dynamic fluorescent materials.

2.3. FRET Processes Due to the considerable overlap between the absorption spectrum of CA-SP and the emission spectra of CA-FITC and CA-Pyr and their excellent compatibility with each other, efficient FRET processes can be established by simply blending trichromic components together. Before blending, the fluorescence intensity of CA-FITC and CA-Pyr remain unchanged over a long period of UV irradiation, labeled G0 and B0. After blending CA-SP with CA-Pyr or CA-FITC in the mass ratio of 1:1, the spectra in Figure 3a,c show the typical emission spectroscopic changes of donor and acceptor during FRET process. By extending the UV irradiation time, as observed in Figure 3 b,d, the emission bands of CA-Pyr (λ = 338 nm) and CA-FITC (λ = 543 nm) decrease gradually because of the energy transfer toward CA-SP. In contrast, the emission of CA-SP (λ = 642 or 618 nm) grows stronger by degrees owing to the population of the MC form. The kinetic fitting plots of the fluorescent intensity rise and fall (λ = 642 nm and λ = 338 nm; λ = 618 nm and λ = 543 nm) present in the insets of Figure 3b,d. The linear results reveal that the emission of both the donor and acceptor varies by first-order kinetics and their kinetic rates (slopes) are comparable to that of CA-SP before blending (more kinetic results are in Figure S15 and Table S1 in the Supporting Information). Additionally, the isosbestic points (Figure 3a, 570 nm; Figure 3c, 580 nm) in the FRET spectra indicate that the ratedetermining step in the FRET process occurs in a 1:1 stoichiometric proportion. Collectively, these two evidences point toward the isomerization of spiropyran in CA-SP determining the dynamic emission of the trichromic mixtures. In such scenarios, the fluorescence intensity of the trichromic components is formulated as R(t), G = G0 (1 − EG→R), B = B0(1 − EB→R), respectively, where E is the FRET efficiency. According to the kinetics analysis (details in the Supporting Information), E ∝ ekt, E ∝ c 02, E ∝ x2, the efficiency E and the related fluorescent intensity (G, B) are predominated by the excitation time t, concentration of the acceptor (c0, CA-SP), and the blending ratio (molar or mass ratio) of acceptor to donor (xG and xB, the blending ratio of R/G and R/B, respectively). The influence of the irradiation time (t) and blending ratio (xG, xB) on the fluorescence properties of the trichromic mixtures is clearly manifested in the colorful ribbon in Figure 3e. The ribbon printed with the trichromic inks in a gradient of ratio (xG, xB) exhibits reversible full-color changes under different irradiation conditions (wavelength and time). Figure 3f presents the CIE (1931) coordinate drifts calculated from the spectra of the mixtures with different blending ratios. By extending the UV irradiation time, the color of the mixtures transits from the beginning states to the red zone of CA-SP. This phenomenon shows again that the trend of color change is determined by the dynamic component CA-SP. In addition, the excitation intensity is another important factor that influences the fluorescent intensity and the RGB values of


the trichromic mixtures. The formula[23] Ie = 2.3I0εϕfb × c (Ie, the intensity of emitted light; I0, the irradiation intensity; ε, the extinction coefficient; ϕf, the quantum yield; b, path length of the cell; c, the concentration of the luminogen or fluorophore) describes how the initial fluorescence intensity of CA-FITC, and CA-Pyr (G0, and B0) and the final emission intensity of CA-SP at equilibrium state (R∞) directly depend the excitation intensity (I0) and the concentration of CA-SP (c0). Conclusively, the fluorescence intensity, i.e., RGB values of the trichromic mixtures is the function of excitation intensity (I), irradiation time (t), and the blending ratios (xG, xB), i.e., RGB = [R(I, t, xG, xB)][G(I, t, xG, xB)][B(I, t, xG, xB)]. 2.4. Dynamic Full-Color Emission The dynamic solid fluorescent materials originating from the trichromic CA-SP, CA-FITC, and CA-Pyr have a wide emission color gamut (full color) and excellent reversibility and stability. The color palette in Figure 4a exhibits the full-color fluorescent images of the disk-like samples prepared by simply mixing the trichromic materials at different ratios. It is more evident in the CIE coordinates calculated from the spectra of these fluorescent disks. The triangular zone [(0.17, 0.07), (0.29, 0.51), (0.55, 0.30)] in Figure 4b presents the color space that can be achieved by the trichromic materials. Under the irradiation cycles of UV 365 nm and visible light, as presented in Figure 4c, the emission peaks (CA-Pyr, 390 nm; CA-FITC, 540 nm; CA-SP, 640 or 618 nm) of the different components in the mixing samples fluctuate reversibly. After several cyclic irradiations, the peak and valley values of fluorescent intensity have not been remarkably attenuated within the extent of experimental error. This stability reveals the excellent reversibility and antifatigue property of the cellulose-based dynamic fluorescent materials. For better displaying such unique performance, we print the painting, The Starry Night (Vincent van Gogh), with the fluorescent inks made of the trichromic materials. For better displaying such unique performance, we have printed a fluorescent rainbow with the inks made of the trichromic materials. In Fig. 4d, the printed pattern is utterly invisible under natural visible light, but a rainbow emerges after UV (365 nm) irradiation. Especially, with the prolongation of UV irradiation time, the color tone of the fluorescent pattern transits from blue to full color gradually and it returns to its original blue tone again under visible light irradiation (the dynamic change refers to the video No. 1 in supplementary information). Such unique optical performance suggests that the cellulose-based fluorescent inks might be an ideal material for security printing and anticounterfeiting. Additionally, the outstanding reversibility and stability also satisfy the basic requirements in practical applications.

2.5. Security Printing for Encryption The fluorescence color or RGB values can be described by a function of several variables including excitation intensity (I), irradiation time (t), and blending ratio (xG, xB) of the trichromic components, RGB = [R(I, t, xG, xB)][G(I, t, xG, xB)][B(I, t, xG, xB)].

1703548 (4 of 8)



www.afm-journal.de

Figure 3. FRET processes in color shifts. With the prolongation of the UV irradiation time (365 nm, the light source in the spectrometer): a) the emission spectra and b) the emission intensity change at λmax = 388 nm and λmax = 642 nm of the sample R/B 1:1 (a blend of trichromatic material CA-SP 5 (R) and CA-pyr (B), the plot in the inset shows the first-order kinetic fitting results of the emission intensity change of R/B 1:1 at λmax = 388 nm and λmax = 642 nm); c) emission spectra and d) the emission intensity variation at λmax = 543 nm and λmax = 618 nm of the sample R/G 1:1 (a blend of trichromatic material CA-SP 20 (R) and CA-FITC (G), the plot in the inset shows the first-order kinetic fitting results of the emission intensity change of R/B 1:1 at λmax = 543 nm and λmax = 618 nm); e) the reversible emission change of the gradient-color ribbon printed with the trichromatic materials (excited with UV 365 nm, 200 µW cm−2, and visible light, respectively); f) the CIE coordinates of different R/B (CA-SP 5 and CA-Pyr) or R/G (CA-SP 20 and CA-FITC) blends during the FRET process.

Such a dynamic nonlinear relationship between fluorescent inputs and outputs can be used as an encryption algorithm in the concept of a dynamic full-color fluorescent anticounterfeiting method within a timescale (30 s) that can be recognized by the naked eye. Figure 5a presents a fluorescent QR code (website: www. iccas.ac.cn) printed with the trichromic fluorescent inks. Under 365 nm irradiation (365 channel), the fluorescent images gradually shift from the blue-green pattern (0 s) to the full color QR


(30 s). Meanwhile, in the images under natural light (visible channel), a red-violet part of the QR codes emerges from the positions of the invisible printed pattern by degrees. When resolving the fluorescent QR code into RGB channels, the patterns in the G and B channels have no evident changes, and a transitional pattern transformation coinciding with the patterns under natural light (the visible channel) occurs in the R channel. Therefore, the variation in the R channel, relating to isomerization of spiropyran in CA-SP, defines the entire

1703548 (5 of 8)



www.afm-journal.de

Figure 4. Color gamut and reversibility of the dynamic fluorescent materials. a) Color palette obtained from the cellulose-based trichromatic fluorescent materials; b) the CIE coordinate gamut calculated according to the emission spectra of the trichromatic materials and their diverse blend samples; c) the reversible emission intensity change at λmax of R/B 3:2 (CA-SP 5 and CA-Pyr) and R/G 2:1 (CA-SP 20 and CA-FITC) excited at 365 nm, 200 µW cm−2, and visible light for 1 min, respectively; d) the color-change loop of the fluorescent rainbow with the cellulose-based trichromatic materials, then excited sequentially at 365 nm, 200 µW cm−2, and visible light.

dynamic fluorescent pattern. Notably, compared with the conventional QR code, the full-color fluorescent QR code is much more complex, because the patterns in different channels (visible, 365 nm, Ch. R, Ch. G, Ch. B, overlay) are utterly different at every node during the UV irradiation period (0–30 s). Further, if these patterns in different channels are arranged in distinct combinations and sequences, e.g. [visible, 365 nm, Ch. R, Ch. G, Ch. B, overlay] or [Ch. G, Ch. R, Ch. B, overlay], such complicated combinations and sequences can be applied to convey the encrypted information far beyond the conventional black-white QR code. Correspondingly, the encryption algorithm is described as [R(I, t, xG, xB)][G(I, t, xG, xB)][B(I, t, xG, xB)][Sequence]. The encryption and decryption procedure is illustrated in Figure 5b. First, according to the previous algorithm, the encrypted information is translated into the [C, M, Y, K] protocol that controls the printer to print the security patterns. In the decryption step, massive data would be recorded under uncertain


irradiation conditions (intensity and time). However, if the cypher were provided, i.e., the excitation intensity, irradiation time, and the specific sequence [I, t, Sequence], the plain text can be interpreted from the security pattern. For example, the cypher in Figure 5a is defined as [(365 nm, 200 µW cm−2), 30 s, QR code by the overlay of RGB channels], and the website www.iccas.ac.cn is obtained by graphic recognition of the QR code converted from the fluorescent image after 30 s of UV irradiation. Unlike the conventional fluorescent QDs, polymer dots and dyes, etc., the cellulose-based dynamic fluorescent materials exhibit great advantages in anticounterfeiting and security printing: (1) the phototunable reversibility of the material ensures that the authentication is both noncontact and nondestructive, which is superior to other methods such as the chemical authentication method currently in practice. (2) The fluorescent inks can be easily prepared with the trichromic materials because of the excellent processability inherited from

1703548 (6 of 8)



www.afm-journal.de

Figure 5. Design and concept of dynamic full-color fluorescent anticounterfeiting. a) Dynamic changes in the QR code printed with the cellulose-based trichromatic materials over UV irradiation (365 nm, 200 µW cm−2) time: from top to bottom are the images under natural visible light, fluorescent images under UV field, red channel (R), green channel (G), and blue channel (B) of the fluorescent images, the overlay pattern of the RGB channel, respectively; b) dynamic encryption and authentication process using the cellulose-based trichromatic materials.


1703548 (7 of 8)



www.afm-journal.de

cellulose materials. Their outstanding film-forming ability also makes the fluorescent inks compatible with surface of different substrates, such as plastics, glass, ceramics, and even metals. (3) FRET is accomplished by a simple blending method, so an inkjet printer can finely control the ratio of the trichromic components. (4) The dynamic fluorescence in full-color space is more distinguishable than the monochrome and therefore conveys more information. Additionally, if the multiresponsive properties of spiropyran and fluorescein in the trichromic materials, e.g., thermochromism, halochromism, ionochromism, etc., except photochromism were introduced into the authentication step, it would greatly increase the difficulty in counterfeiting or reverse engineering.

3. Conclusion In summary, we have described a feasible and economical strategy for preparing the full-color phototunable fluorescent materials by trichromacy and a dynamic FRET mechanism. A set of trichromic materials, CA-SP, CA-FITC, and CA-Pyr, are synthesized from cellulose, the most abundant natural biopolymer on earth. The FRET between photochromic and fluorochromic CA-SP and its counterparts, CA-FITC and CA-Pyr, is accomplished by simply blending them together. The energy transfer efficiency between FRET donors and acceptors is finely tuned by the excitation intensity, irradiation time, and blending ratio of the trichromic components to regulate the fluorescent properties (color and intensity) within a timescale recognizable by the naked eye. Herein, we propose a concept for dynamic full-color fluorescent anticounterfeiting stemming from a complicated dynamic nonlinear fluorescent input–output relationship, which is extremely difficult to duplicate and reverse engineer. The convenient preparation and broad compatibility of the fluorescent inks also enhances the utility of this noncontact and nondestructive dynamic authentication method when applied to information encryption and security printing. Furthermore, the dynamic fluorescent properties could be used to develop more applications of cellulose-based fluorescent material, including bioimaging, fluorescent tracing, and chemical detection and monitoring.

Supporting Information Supporting information is available from the Wiley Online Library or from the author.

Acknowledgements This work was supported by the National Science Foundation of China (Nos. 51425307 and 51573196) and the Program of Taishan Industry Leading Talents (Shandong Province).


Conflict of Interest The authors declare no conflict of interest.

Keywords cellulose, dynamic fluorescence, full-color emission, phototunable fluorescence, time-contrast Received: June 28, 2017 Revised: August 6, 2017 Published online:

[1] D. Ackerman, A Natural History of the Senses, Vintage, New York, USA 1991. [2] M. Antona, C. Stephanidis, Universal Access in Human-Computer Interaction, Springer, Berlin Heidelberg, German 2016. [3] L. R. Gregory, Nature 1965, 207, 16. [4] W. Zhou, Y. Jiang, S. He, D. Chen, Curr. Biol. 2010, 20, 1356. [5] J. Zhou, Y. Yang, C. Zhang, Chem. Rev. 2015, 115, 11669. [6] X. Cheng, S. B. Lowe, P. J. Reece, J. J. Gooding, Chem. Soc. Rev. 2014, 43, 2680. [7] S. Y. Lim, W. Shen, Z. Gao, Chem. Soc. Rev. 2015, 44, 362. [8] K. Jiang, S. Sun, L. Zhang, Y. Lu, A. Wu, C. Cai, H. Lin, Angew. Chem., Int. Ed. 2015, 54, 5360. [9] F. Zhang, H. Zhong, C. Chen, X. Wu, X. Hu, H. Huang, J. Han, B. Zou, Y. Dong, ACS Nano 2015, 9, 4533. [10] Y. Rong, C. Wu, J. Yu, X. Zhang, F. Ye, M. Zeigler, M. E. Gallina, I. Wu, Y. Zhang, Y. Chan, W. Sun, K. Uvdal, D. T. Chiu, ACS Nano 2013, 7, 376. [11] C. Wu, B. Bull, C. Szymanski, K. Christensen, J. McNeill, ACS Nano 2008, 2, 2415. [12] Y. S. Zhao, H. Fu, A. Peng, Y. Ma, D. Xiao, J. Yao, Adv. Mater. 2008, 20, 2859. [13] J. N. Zhang, H. Kang, N. Li, S. M. Zhou, H. M. Sun, S. W. Yin, N. Zhao, B. Z. Tang, Chem. Sci. 2017, 8, 577. [14] K. Umezawa, Y. Nakamura, H. Makino, D. Citterio, K. Suzuki, J. Am. Chem. Soc. 2008, 130, 1550. [15] E. Kim, M. Koh, B. J. Lim, S. B. Park, J. Am. Chem. Soc. 2011, 133, 6642. [16] F. Wang, X. Liu, Chem. Soc. Rev. 2009, 38, 976. [17] Q. Zhang, D. Li, X. Li, P. B. White, J. Mecinovic´, X. Ma, H. Ågren, R. J. M. Nolte, H. Tian, J. Am. Chem. Soc. 2016, 138, 13541. [18] X. Hou, C. Ke, C. J. Bruns, P. R. McGonigal, R. B. Pettman, J. F. Stoddart, Nat. Commun. 2015, 6, 6884. [19] R. Deng, F. Qin, R. Chen, W. Huang, M. Hong, X. Liu, Nat. Nanotechnol. 2015, 10, 237. [20] J. Chen, W. Zhong, Y. Tang, Z. Wu, Y. Li, P. Yi, J. Jiang, Macromolecules 2015, 48, 3500. [21] W. G. Tian, J. M. Zhang, J. Yu, J. Wu, H. Nawaz, J. Zhang, J. S. He, F. S. Wang, Adv. Opt. Mater. 2016, 4, 2044. [22] W. G. Tian, J. Tian, Dyes Pigm. 2014, 105, 66. [23] E. V. Anslyn, D. A. Dougherty, Modern Physical Organic Chemistry, University Science Books, Sausalito, California, USA 2006.

1703548 (8 of 8)
