Coherent random lasing from nano-scale aggregates of hybrid molecules by enhanced near zone scattering

Abstract

Nano-scale aggregates of hybrid molecules, such as N,N′-di-[3-(isobutyl polyhedral oligomeric silsesquioxanes)propyl] perylene diimide (DPP) which is composed of perylene diimide (PDI) as gain group and polyhedral oligomeric silsesquioxanes (POSS) as scattering group at a mole ratio of 1 : 2, are formed in carbon disulfide (CS₂) solution within a concentration range from 3 × 10⁻⁵ M to 10⁻³ M. The aggregation of DPP in the solution was characterized using ultraviolet visible (UV-vis) and fluorescence spectroscopy. Under pumping of a 532 nm pulse laser, coherent random laser (RL) is produced from the aggregates in CS₂ solution with a threshold of 17.4 μJ and a Q value of 2440 at a DPP concentration of 10⁻⁴ M. Enhanced near zone scattering by aggregation is thought to boost coherent RL. This research not only extends the RL system to the extremely weak scattering regime but also provides a new direction for research on nano-scale materials for lasers using molecular design.

---

Introduction

In photonics, disorder and aperiodicity are often considered to be particularly undesirable because they spoil optical performance and quality. However, disorder created by a naturally occurring or artificially introduced event can in fact make a contribution which enhances the functionality of the photonic devices.^1,2 Therefore, disordered materials have become a developing research topic in the development of photonic devices. When the disorder is introduced into the gain media, a called random laser (RL) can be obtained by multiple light scattering through the forming resonant random cavities.³ Considering the mechanism of boosting RL, there are two necessary elements in a random lasing system. One is the gain medium such as a laser dye, quantum dots (QDs) and so on. The other one is the scatterer and this can be classified into two types: passive scatterers such as polyhedral oligomeric silsesquioxanes (POSS) nanoparticles (NPs),^4,5 and so on, which only act as scatterers, and active scatterers, which work simultaneously as scatterers and gain media.⁶ A semiconductor is a good candidate as an active scatterer because of its wide bandgap. Zinc oxide (ZnO),⁷ gallium arsenide (GaAs),⁸ and so on, have been used in RL systems in the form of powders, nanowires, nanorods, and so on, as active scatterers. Other active scatterers that have been used are cold atoms,⁹ QDs,¹⁰ π-conjugated polymer matrix,¹¹ submicron size powders prepared from rare earth doped crystals,¹² and so on. Normally, laser dyes are large organic molecules and only used as gain media. Dye-based random lasers have been produced with the help of active scatterers such as ZnO,¹³ or passive scatterers such as titanium dioxide powders,¹⁴ nematic liquid crystals,^15,16 biological tissues,¹⁷ and so on.

Light scattering by groups of particles is currently being researched because of important multiple scattering effects, for example, enhanced backscattering^18,19 and the negative polarization branch²⁰ in the field of light detection and ranging (LiDAR) experiments²¹ and astrophysics.²² In groups, multiple scattering has been classified into three categories according to distance between particles (d),^18,23 including scattering in the near zone (d ≪ λ), the transitional zone and the far zone (d ≫ λ). Fundamental research on RL from grouped media should produce important progress in how to deal with the light scattering of particles. However, in most papers about traditional RL media, only far zone scattering has been considered to account for RL. Research on how near zone scattering influences RL still remains a great challenge in the design and fabrication of RL materials.

Perylene diimide (PDI) is a type of fluorescent dye that has a high fluorescence quantum yield (near 100%) in dilute solutions.²⁴ It is widely used in the areas of liquid crystal display,²⁵ fluorescent sensors,²⁶ photoelectrical and solar cell materials²⁷ because of its unique properties such as thermal, photo and chemical stability,²⁸ wide absorption of visible light²⁹ and the previously mentioned highly efficient photoluminescence. However, this type of dye has seldom been used as a laser gain medium because its luminescence is highly quenched in highly concentrated solutions³⁰ and especially in the solid state³¹ because of the attraction of dipole–dipole interactions and π–π stacking between molecules.^32–38

In this work, PDI is firstly introduced into the RL system through an organic–inorganic hybrid molecule, N,N′-di-[3-(isobutyl polyhedral oligomeric silsesquioxanes)propyl] perylene diimide (DPP) and is composed of PDI as a gain group and POSS as scattering group at a mole ratio of 1 : 2. DPP is synthesized with two POSS NPs covalently attached to both side N atoms of PDI and could work as an active scatterer as PDI works as a gain group while the two POSS NPs work as scattering groups. Also, such a new type of active scatterer is also an ideal model to study the effect of near zone scattering on RL because the distance of POSS NPs is known to be 1.2 nm by simple calculation from its chemical structure, which is far smaller than the emission wavelength λ = 584 nm (d ≪ λ). In order to compare them with each other, the RL of a reference solution of DCP and 2 × 10⁻⁴ M POSS (2POSS) was also measured in this research. Interestingly, a high quality of coherent RL can only be observed in concentrated DPP solutions, which cannot be obtained in dilute solutions of DPP as well as in DCP and 2POSS solutions. These phenomena are experimentally investigated in detail, and a physical model is presented for the enhancement of near zone scattering by aggregates of DPP in a solution of carbon disulfide (CS₂) for boosting RL, which presents a route for designing new nano-scale active scatterers and to explore their near zone scattering with “d ≪ λ”.

Experimental

Materials and measurements

Molecular structures of the DPP, DCP and POSS are shown in Scheme 1; DPP and DCP were prepared according to the previously reported procedures.^24,39 The commercial aminopropylisobutyl POSS was obtained from Hybrid Plastics. CS₂ was obtained from a commercial source and used as solvent in all samples without further purification. A Q-switched neodymium-doped yttrium aluminium garnet (Nd:YAG) laser (pulse duration 10 ns, repetition rate 10 Hz) with an output of 532 nm was used to pump the sample with a focus lens of f = 10 cm. Pump pulse energy was controlled using a Glan prism. Emitted light was collected using a QE65 Pro fiber spectrometer (Ocean Optics, FL, USA; resolution 0.4 nm, integration time 100 ms). Ultraviolet visible (UV-vis) absorption spectra were measured on a UV-2550 PC spectrophotometer (Shimadzu). Fluorescence spectra were measured on a RF-5301PC spectrofluorophotometer (Shimadzu).

Scheme 1 Molecular structures of DPP, DCP and POSS.

Sample preparation

In this research, two types of samples were prepared for RL measurement; one was an active scatterer DPP solution in CS₂ with a series of concentrations (10⁻³ M to 10⁻⁷ M), the other was a reference solution in CS₂, in which DCP was used as gain media and POSS was used as a passive scatterer with a series of concentrations (DCP: 10⁻³ M to 10⁻⁷ M) at a 1 : 2 mole ratio of DCP : POSS. All of the sample solutions were encapsulated in a glass tube with an inner (outer) diameter of 200 μm (300 μm).

Results and discussion

Random lasing from aggregates by pumping of a 532 nm pulse laser

RL measurements were performed by using the experimental set-up illustrated schematically in Fig. 1a. A 532 nm pulse laser source was used to pump samples (solution) in a glass tube with the laser direction perpendicular to the direction of the length of the tube and a fiber probe of a fiber spectrometer was kept close to the pump region on the tube and perpendicular to both the direction of the laser and direction of the length of the tube. The emission spectra of all samples were measured (see Fig. S1 (ESI†) and Table 1) and the emission spectra of 10⁻⁴ M DPP, 10⁻⁴ M DCP and 2 × 10⁻⁴ M POSS in CS₂ solution at a pump energy of 120 μJ are shown in Fig. 1b. Multiple sharp peaks can be observed in the spectra of the DPP solution and the main peak is centered at λ = 584.0 nm whereas only fluorescent emission can be detected in the DCP and 2POSS solutions. Typically, such a set of sharp spectral features as shown in Fig. 1b indicates coherent lasing resonances.^40,41 Fig. 1c shows the relationships between pump energy, main peak intensities (square) and half peak widths (triangle) for DPP solution with a concentration of 10⁻⁴ M, from which a threshold for RL can be determined as 17.4 μJ, and the smallest line width of 0.24 nm, thus a coherent RL with a high quality value (Q = 2440) and low threshold (17.4 μJ) was demonstrated experimentally in this RL solution system.

Fig. 1 Schematic illustration of the experimental setup for the detection of R (a); emission spectra of 10⁻⁴ M DPP and 10⁻⁴ M DCP and 2 × 10⁻⁴ M POSS in CS₂ solution at a pump energy of 120 μJ (b); dependence of the intensity (black square) and line width (blue triangle) of the main peak of the emission spectra of 10⁻⁴ M DPP in CS₂ solution at different pump energies (c).

Table 1 Dependence of RL action on concentration of solutes

Sample	Concentration (mol L^−1)	ε0-1/ε0-0	RL (Y/N)	Threshold (μJ)
a DCP concentration in CS2 solution is 10^−7 M, 10^−6 M, 10^−5 M, 10^−4 M, 10^−3 M.b POSS concentration in the solution is twice that of the DCP concentration and is to 2 × 10^−7 M, 2 × 10^−6 M, 2 × 10^−5 M, 2 × 10^−4 M, 2 × 10^−3 M, respectively.
DPP-1	10^−7	0.61	N	—
DPP-2	10^−6	0.61	N	—
DPP-3	10^−5	0.61	N	—
DPP-4	3 × 10^−5	0.63	Y	185.5
DPP-5	5 × 10^−5	0.67	Y	62.8
DPP-6	7 × 10^−5	0.83	Y	59.2
DPP-7	10^−4	1.00	Y	17.4
DPP-8	10^−3	1.02	Y	7.8
DCP and 2POSS	10^−7 to 10^−3^a, and 2 × 10^−7 to 2 × 10^−3^b	0.63–1.02, and —	N	—

---

For a random gain system, one criterion for generating RL is formation of sufficiently strong stimulated emission before amplified light escapes the random system because of the light scattering, that is l_s > l_g, where l_s is the scattering mean free path defined as the average distance that light travels between two consecutive scattering events, and l_g is the gain length defined as the path length over which the intensity is amplified by a factor of e⁺¹.⁴² The value of l_g can be estimated by l_g = 1/ρ_gσ_em, where ρ_g and σ_em are the number density and emission cross section of gain materials, respectively. Obviously, in a weak scattering regime, the criterion is satisfied as l_s is always very large; however, it was found in a dye solution with very weak scattering that the only role of the dye (also scatterers) was to scramble the directionality of the amplified spontaneous emission, in which the concentration of dye in the gain medium is 2.5 mM.⁴³ Therefore, a smaller l_s, but still larger than l_g, becomes necessary for generating RL. Usually there are three methods to achieve this requirement: one is to dope passive scatterers into the solution,⁴³ the second is to use closely packed (high concentration) active scatterers⁴⁴ and the third is to use a high pumping energy for generating RL.¹³

In previous research, POSS was found to be a passive scatterer for generating coherent RL in a dye solution, in which the concentration of POSS was varied from 24.5 wt% to 1.0 wt%, corresponding to concentration range of 3.7 × 10⁻¹ M to 1.5 × 10⁻² M, respectively.⁵ The high POSS concentration adopted in the research is partly responsible for the coherent RL according to the principle of the first method reported previously. For comparison, a low POSS concentration was explored in a CS₂ solution that contains POSS as scatterers and DCP as dyes. The emission spectrum obtained which is shown in Fig. 1b is normal fluorescence from the solution with a DCP concentration of 10⁻⁴ M and POSS concentration of 2 × 10⁻⁴ M. However, with a DPP concentration of 10⁻⁴ M, coherent RL emission is clearly produced from DPP solution in CS₂ under the same experimental conditions as for the DCP solution in CS₂ as shown in Fig. 1. This phenomenon taking place at such a dye concentration has not been observed before, to our knowledge, and is worth being investigated in detail.

Aggregation of DPP molecules characterized by UV-vis and fluorescence spectroscopy

Optical properties of DPP and DCP in CS₂ were investigated using UV-vis and fluorescence spectroscopy. Measurements with a variable concentration (10⁻⁷ M to 10⁻³ M) were carried out for both DPP and DCP solutions and results are given in Fig. 2a and b, respectively. It is clearly seen that three main absorption peaks of DPP are located at 530 nm, 493 nm and 461 nm shown in Fig. 2a whereas three peaks of DCP are located at 531 nm, 494 nm and 462 nm as shown in Fig. 2b, corresponding to 0-0, 0-1, 0-2 electronic transitions, respectively.^45,46 With an increase of concentration, an obvious change of the ratio of ε_0-1/ε_0-0 (ref. 47) of both types of samples is shown in Fig. 2c and d, respectively. In detail, when the concentration was increased from 10⁻⁷ M to 10⁻⁵ M, the ε_0-1/ε_0-0 remained unchanged at 0.61, which showed that DPP or DCP is dispersed in the solution as a single molecule, and no aggregation had taken place. When the concentration was above 3 × 10⁻⁵ M, the ε_0-1/ε_0-0 increased with the concentration increasing for both of the samples, which showed an evident aggregation of solute molecules.⁴⁷ When the concentration reached 10⁻³ M, both ε_0-1/ε_0-0 increased to about 1.02, which showed that the extent of aggregation of both solute molecules, DPP and DCP, had increased with the concentration increase.

Fig. 2 UV-vis absorption spectra of DPP (a) and DCP (b) in CS₂ solution with variable concentrations, ε_0-1/ε_0-0 of DPP (c) and DCP (d) in CS₂ solution with variable concentrations.

Fluorescence measurements of both solutions shown in Fig. 3, could further confirm the previous conclusion. From Fig. 3c, intensities of the three characteristic emission peaks located at 540 nm, 580 nm and 630 nm, corresponding to the 0-0, 0-1, 0-2 electronic transitions, respectively, firstly increase with the increase in concentration from 10⁻⁷ M to 10⁻⁵ M and then decrease as the concentration is further increased from 10⁻⁵ M to 10⁻³ M. The former results from an increase of dye concentration in the dilute solution whereas the latter is because of the distance between dye molecules which shortens with the increase of concentration, which will lead to a decrease of fluorescent intensity,⁴⁸ namely fluorescence quenching. The red shift of the emission peak at 540 nm corresponding to 0-0 electronic transitions with increase of concentration (Fig. 3c) results from self-absorption^49,50 and also aggregation.^34,37

Fig. 3 Fluorescence spectra of DPP (a) and DCP (b) in CS₂ solution with variable concentrations. Wavelength (black square) and intensity (blue triangle) of emission peak at about 540 nm corresponding to the 0-0 electronic transition of DPP (c) and DCP (d) in CS₂ solution with variable concentrations.

Random lasing and aggregation

To find the relationship between RL action and the aggregation behavior discussed previously, the dependence of the RL action on different concentrations of all of the two series of solutions is illustrated in Table 1, from which it can be found that there is no phenomenon of RL from a dilute solution of DPP (DPP-1 to DPP-3) with an ε_0-1/ε_0-0 which remains at 0.61. With an increase of the concentration (DPP-4 to DPP-8), aggregation of DPP occurs as ε_0-1/ε_0-0 increases from 0.61 to 1.02, and at the same time, RL phenomena are observed. This result shows that the RL action of DPP solution is closely related to aggregation of the DPP molecule in the solution.

From Table 1, it is also found that there is no RL phenomenon in the reference solutions of DCP and 2POSS, whether the concentration of the solution is high or low. To a great extent, scattering by POSS will also be enhanced at high concentrations of the reference solution because two types of solution have the same concentration of both the scattering group and the gain group. From the point of scattering, in a spherical and weakly scattering system (particle–particle interaction vanishes), and a mean scattering free path l_s can be obtained using:⁴³

in which λ is the wavelength of the emitted light, C is the concentration, N_A = 6.02 × 10²³ mol⁻¹, D is diameter of scatterers, m is the complex refractive index of scatters relative to the adjacent medium, and

In solutions considered in this research:

thus
in which C_POSS is the numerical value of POSS concentration, and D is the diameter of POSS scatterers. For DPP solutions, the lowest concentration that boosts RL was 3 × 10⁻⁵ M. Supposing it was a dilute solution, C_POSS = 6 × 10⁻⁵ (M), D = 1.17 (nm), and l_s = 8.4 × 10⁵ cm. Supposing that it was a dimer solution, C_POSS = 3 × 10⁻⁵ (M), D = 2.34 (nm), and l_s = 2.6 × 10⁴ cm. Considering the real system is a mixture of the two states discussed previously, the l_s is between 8.4 × 10⁵ cm and 2.6 × 10⁴ cm. Apparently, the multiple scattering events are not enough to support coherent feedback because l_s is far greater than the sample size L (the tube diameter of 200 μm). For DCP and 2POSS solution, the l_s is also very large, up to 10⁻⁵ M DCP and 2 × 10⁻⁵ M POSS, 10⁻⁴ M DCP and 2 × 10⁻⁴ M POSS, and 10⁻³ M DCP and 2 × 10⁻³ M POSS systems, the l_s are 2.5 × 10⁶ cm, 2.5 × 10⁵ cm and 2.5 × 10⁴ cm, respectively. These calculated results mean that the two solutions are all weakly scattering systems, therefore there must exist other key factors to assist the RL emission observed.

Near zone scattering in a weakly scattering system

Fig. 4 further illustrates the dependence of the threshold of the RL action on the aggregation extent of DPP represented by ε_0-1/ε_0-0, from which it is found that the higher aggregation would induce a lower threshold of the RL action. In other words, a solution of DPP with a higher concentration is more likely to boost RL, which seems reasonable for a higher concentration of active scatterer which means a larger gain and stronger multiple scattering. However, it is well known that aggregation results in quenching of fluorescence and is of low quantum yield compared with the corresponding single molecule.⁵¹ Such a phenomenon is also demonstrated by the fluorescence properties of DPP solution as shown in Fig. 3c, from which DPP is found to have aggregated at a higher concentration, resulting in fluorescence quenching.

Fig. 4 Dependence of threshold on ε_0-1/ε_0-0 of DPP samples.

To explore the hidden factors which boost the coherent RL, the distance of scattering of NPs in DPP solution is considered and this is shown in Fig. 5. The distance of two POSS groups in a single molecule is only 1.2 nm because of the chemical bonding, which is far smaller than λ = 584.0 nm (d ≪ λ). Even when aggregation occurs, the distance of POSS aggregates accompanied by DPP dimer remains at about 1.2 nm inside the dimer.⁴⁷ Furthermore, these POSS groups are just near the gain media PDI, which creates a suitable environment for realization of near zone scattering, by which a RL could be boosted in such a weak far zone scattering system. Also, it is necessary to mention the solutions of DCP and 2POSS, in which the d is 9.4 nm, 20.3 nm, 43.7 nm when the concentrations of DCP are 2 × 10⁻³ M, 2 × 10⁻⁴ M, 2 × 10⁻⁵ M, respectively, d is also smaller than λ (d < λ), but considering that the POSS group is not connected to the fluorescent PDI molecule through chemical bonding, the values of d given previously are only of statistically average meaning, which means that the distance is not for two specific scatterers and the near zone scattering model is not suitable for the solution.

Fig. 5 Schematic illustration of the distance between scattering group POSS in DPP solutions.

Enhanced near zone scattering of aggregates

Fig. 5 shows a schematic illustration of near zone scattering enhancement resulting from DPP molecules in which RL is produced from aggregates, and a single molecule can only produce fluorescence, although the single molecule could also form near zone scattering. At a low concentration, there is no aggregation of DPP formed according to the absorption spectra shown in Fig. 2c, and no RL produced as shown in Table 1, although the quantum efficiency of DPP molecules is higher than that of DPP aggregates in terms of general realization for fluorescence.⁵¹ Near zone scattering enhancement of aggregates is so strong that it not only compensates the quenching caused by aggregation, but also decreases l_s enough to boost RL in the very weak regime.

A further explanation of enhanced near zone scattering by aggregation can be obtained by comparing two geometrical model¹⁸ for a single molecule and a dimer aggregate as shown in Fig. 6. (θ_T is the sum of angles of emission light which could scatter back from one POSS group to the other), which is defined as a ratio of near zone scattering to all of the emission light from PDI when pumped by 532 nm laser, is compared with each other for a single molecule and a dimer of DPP in CS₂ solution. In Fig. 6a, two blue circles represent the POSS groups with a radius DG = 2.7 Å whereas PDI is simplified as a pointolite located at O (0,0) with OD = 8.5 Å calculated from a simulation of molecular structure. Most of the emission light from PDI will escape and only a very small part could be scattered back from one side of a POSS group to the other side and the calculated R_NS is only 2.0% (), which means the near zone scattering is very weak in a single molecule, thus it is not easy to boost RL in a dilute solution. However, when aggregation occurs, a dimer model is constructed (see Fig. 6b), and the calculated R_NS increases to 16.1% (), meaning that there is a much stronger near zone scattering in these aggregates, which partly explains the boosting of RL only when aggregation occurs. The model presented in this research is constructed in terms of geometrical optics and a near zone scattering model¹⁸ and is justified by the experimental observations. It was easily found that the extent of the enhanced scattering by aggregation was much smaller than that of the difference between the scattering mean free path l_s and the size of the samples L, although the scattering is really enhanced according to the experimental observation. The unsolved question is that the RL here was boosted by a single aggregate or lots of the aggregates, if the latter, how much the l_s between aggregates as well as the sample size L which produces outside glass reflection play roles on boosting RL. Detailed theoretical analysis of the enhanced near zone scattering based on nano-scale aggregates and corresponding experiments are still in progress.

Fig. 6 Schematic illustration of the ratio of near zone scattering (R_NS) in (a) a single molecule and (b) a dimer of DPP in CS₂ solution.

Conclusions

In conclusion, coherent RL is produced by a DPP solution in a concentration range of 10⁻⁵ M to 10⁻³ M under pumping of a 532 nm pulse laser. Experimental investigation shows that the RL is produced from aggregates formed by DPP in the solution, which results in enhanced near zone scattering. There is no RL observed in DPP solution with a lower concentration or in the reference sample with the solution of DCP as gain media and POSS as scatterer at the same concentration range. A physical model presented in this paper shows that the possible reason for boosting RL is enhanced near zone scattering induced by aggregation of DPP in the solution. The phenomenon not only extends NP-based RL to the extremely weak scattering system, but also opens the way for designing new nano-scale active scatterers and exploring their near zone scattering with “d ≪ λ”.

Acknowledgements

This work was supported by National Natural Science Foundation of China (NSFC) (51273186, 21574120, 51673178, and 11404087), the Basic Research Fund for the Central Universities (WK2060200012) and the Science and Technological Fund of Anhui Province for Outstanding Youth (1608085J01).

Notes and references

1. M. Segev, Y. Silberberg and D. N. Christodoulides, Nat. Photonics, 2013, 7, 188.
2. M. Segev, Y. Silberberg and D. N. Christodoulides, Nat. Photonics, 2013, 7, 197.
3. V. S. Letokhov, JETP Lett., 1967, 5, 212.
4. L. Cerdán, A. Costela, I. García-Moreno, O. García and R. Sastre, Opt. Express, 2010, 18, 10247.
5. Z. Hu, Q. Zhang, B. Miao, Q. Fu, G. Zou, Y. Chen, Y. Luo, D. Zhang, P. Wang, H. Ming and Q. Zhang, Phys. Rev. Lett., 2012, 109, 253901.
6. F. Luan, B. Gu, A. S. L. Gomes, K.-T. Yong, S. Wen and P. N. Prasad, Nano Today, 2015, 10, 168.
7. H. Cao, Y. G. Zhao, S. T. Ho, E. W. Seelig, Q. H. Wang and R. P. H. Chang, Phys. Rev. Lett., 1999, 82, 2278.
8. M. A. Noginov, G. Zhu, I. Fowlkes and M. Bahoura, Laser Phys. Lett., 2004, 1, 291.
9. W. Guerin, F. Michaud and R. Kaiser, Phys. Rev. Lett., 2008, 101, 093002.
10. K. S. Lee, P. N. Prasad, G. Huyet and C. H. Tan, Opt. Mater. Express, 2012, 2, 578.
11. R. C. Polson, A. Chipouline and Z. V. Vardeny, Adv. Mater., 2001, 13, 760.
12. J. Zhang, L. Xu, H. Wang, F. Huang, X. Sun, H. Zhao and X. Chen, Appl. Phys. Lett., 2013, 102, 021109.
13. K. Firdaus, T. Nakamura and S. Adachi, Appl. Phys. Lett., 2012, 100, 171101.
14. N. M. Lawandy, R. M. Balachandran, A. S. L. Gomes and E. Sauvain, Nature, 1994, 368, 436.
15. S. Ferjani, V. Barna, A. De Luca, C. Versace and G. Strangi, Opt. Lett., 2008, 33, 557.
16. G. Strangi, S. Ferjani, V. Barna, A. De Luca, C. Versace, N. Scaramuzza and R. Bartolino, Opt. Express, 2006, 14, 7737.
17. R. C. Polson and Z. V. Vardeny, Appl. Phys. Lett., 2004, 85, 1289.
18. O. Merchiers, J. M. Saiz, F. González and F. Moreno, J. Opt. Soc. Am. A, 2005, 22, 497.
19. A. Ishimaru, Proc. IEEE, 1991, 79, 1359.
20. Y. Shkuratov, A. Ovcharenko, E. Zubko, O. Miloslavskaya, K. Muinonen, J. Piironen, R. Nelson, W. Smythe, V. Rosenbush and P. Helfenstein, Icarus, 2002, 159, 396–416.
21. M. J. McGill, Encyclopedia of Optical Engineering, 2003, vol. 22, pp. 1103–1113.
22. G. Wurm, H. Relke and J. Dorschner, Astrophys. J., 2003, 595, 891.
23. V. A. Markel', J. Mod. Opt., 1992, 39, 853.
24. X.-D. Zhuang, Y. Chen, B. Zhang, Y. Li, B. Yu and W. Qiao, New J. Chem., 2010, 34, 1120.
25. Y. D. Kim, J. P. Kim, O. S. Kwon and I. H. Cho, Dyes Pigm., 2009, 81, 45.
26. Y. Liu, K. R. Wang, D. S. Guo and B. P. Jiang, Adv. Funct. Mater., 2009, 19, 2230.
27. C. Li and H. Wonneberger, Adv. Mater., 2012, 24, 613.
28. Y. Avlasevich, C. Li and K. Müllen, J. Mater. Chem., 2010, 20, 3814.
29. X. Ren, B. Sun, C.-C. Tsai, Y. Tu, S. Leng, K. Li, Z. Kang, R. M. Van Horn, X. Li, M. Zhu, C. Wesdemiotis, W.-B. Zhang and S. Z. D. Cheng, J. Phys. Chem. B, 2010, 114, 4802.
30. Y. S. Ma, C. H. Wang, Y. J. Zhao, Y. Yu, C. X. Han, X. J. Qiu and Z. Q. Shi, Supramol. Chem., 2007, 19, 141.
31. C. Haines, M. Chen and K. P. Ghiggino, Sol. Energy Mater. Sol. Cells, 2012, 105, 287–292.
32. L. L. Ma, Q. B. Wang, G. F. Lu, R. P. Chen and X. Sun, Langmuir, 2010, 26(9), 6702–6707.
33. S. G. Liu, G. D. Sui, R. A. Cormier, R. M. Leblanc and B. A. Gregg, J. Phys. Chem. B, 2002, 106, 1307.
34. P. Yan, A. Chowdhury, M. W. Holman and D. M. Adams, J. Phys. Chem. B, 2005, 109, 724.
35. X. Zhang, Z. J. Chen and F. Wurthner, J. Am. Chem. Soc., 2007, 129, 4886.
36. M. Schneider, J. Hangen, D. Haarer and K. Mullen, Adv. Mater., 2000, 12, 351.
37. T. J. Tang, J. Q. Qu, K. Mullen and S. E. Webber, Langmuir, 2006, 22, 26.
38. B. A. Gregg, J. Phys. Chem. B, 2003, 107, 4688.
39. Y. K. Che, X. M. Yang, K. Balakrishnan, J. M. Zuo and L. Zang, Chem. Mater., 2009, 21, 2930.
40. H. Cao, Waves Random Media, 2003, 13, R1.
41. X. Zhao, Z. Wu, S. Ning, S. Liang, D. Wang and X. Hou, Opt. Express, 2011, 19, 16126.
42. D. S. Wiersma and A. Lagendijk, Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top., 1996, 54(4), 4256–4265.
43. X. H. Wu, A. Yamilov, H. Noh, H. Cao, E. W. Seelig and R. P. H. Chang, J. Opt. Soc. Am. B, 2004, 21, 159.
44. R. Chen, M. L. B. Utama, Z. Peng, B. Peng, Q. Xiong and H. Sun, Adv. Mater., 2011, 23, 1404.
45. S. Ghosh, X. Q. Li, V. Stepanenko and F. Würthner, Chem.–Eur. J., 2008, 14(36), 11343.
46. K. Balakrishnan, A. Datar, T. Naddo, J. Huang, R. Oitker, M. Yen, J. Zhao and L. Zang, J. Am. Chem. Soc., 2006, 128(22), 7390.
47. A. E. Clark, C. Y. Qin and A. D. Q. Li, J. Am. Chem. Soc., 2007, 129, 7586.
48. Y. Wang, H. Chen, H. Wu, X. Li and Y. Weng, J. Am. Chem. Soc., 2009, 131, 30–31.
49. F. Donati, A. Pucci, C. Cappelli, B. Mennucci and G. Ruggeri, J. Phys. Chem. B, 2008, 112, 3668–3679.
50. P. Bauer, H. Wietasch, S. M. Lindner and M. Thelakkat, Chem. Mater., 2007, 19, 88–94.
51. P. Anger, P. Bharadwaj and L. Novotny, Phys. Rev. Lett., 2006, 96, 113002.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra10511d

---