Direction-specific fluorescence of an engineered organic crystal and the appearance of a new face caused by mechanically induced shaping

Shotaro Hayashi * and Toshio Koizumi
Department of Applied Chemistry, National Defense Academy, 1-10-20 Hashirimizu, Yokosuka, 239-8686, Japan. E-mail: shayashi@nda.ac.jp

Received 27th June 2019 , Accepted 23rd August 2019

First published on 23rd August 2019


Here we have studied the directional nature of the fluorescence emission of a centimetre-scale organic crystal. Sky blue-colored fluorescence (λPL = 481 nm) of the (010) face and the green-colored fluorescence (λPL = 502 nm) of the (001) face were observed, because the (010) and (001) faces of the crystal have different directional orientations of the molecules. Interestingly, the mechanically induced shaping of the centimetre-scaled single crystal caused a new green-colored fluorescent (001) face to appear in the cross section.


Attractive molecular assembles of one-dimensional (1D) molecules such as supramolecular polymers,1–3 and of two-dimensional (2D) molecules such as supramolecular thin films,4–6 have been developed.7 Alternatively, organic bulk crystals are found as supramolecular three-dimensional (3D) assemblies. There are multiple “faces” of these organic molecular crystals that grow three-dimensionally. Since the molecules are oriented in a single crystal, it appears that the organic crystal should have direction- and surface-specific properties. In particular, in a single crystal that can be visualized (on the macroscopic scale), research should be designed that takes advantage of the properties of each of these “faces”. However, there have been very few studies focusing on the direction-specific properties of the crystal faces.8–16 For example, Sada and his co-workers have succeeded in selectively adsorbing molecules chemically on the surface of a micro-scale crystal.9–11 In addition, as a result of the optical waveguide phenomenon in fluorescent microcrystals (micro-scale crystalline thin films), the light-emission is observed only from the edge.12–16 It is thought that such results in larger bulk crystals may also be observed. However, there is no definite report on the direction-specific properties of large (centimetre-scale) bulk crystals. This is probably due to crystal defects, including stacking faults, grain boundaries and impurities. Therefore, the creation of 3D specific properties that depend largely on the anisotropy and structural integrity of the crystals is extremely interesting.

In recent years, large crystalline thin films with structural integrity have been reported.4,5 Self-assembly into large (centimetre-scale) crystals increases the risk of defects. It is challenging and interesting with respect to face-specific properties in large (centimetre-scale) crystals that can be visualized with the naked eye. To crystallize precise organic molecules, we focused on fluorescent π-conjugated molecules that we have developed.17–20 Herein we present the centimetre-scale single crystal of a π-conjugated molecule, showing blue-colored fluorescence on one face and green-colored fluorescence on the other face: direction-specific double fluorescence. In this work, we have successfully fabricated a large-scale single crystal with structural integrity via the slow crystal growth of 1,4-bis[2-(4-methylthienyl)]-2,3,5,6-tetrafluorobenzene for the appearance of direction-specific double fluorescence. Moreover, we have performed spectroscopic detection of the fluorescence by specially dissolved μ-photoluminescence analysis and mechanically induced shaping of the crystal for the appearance of the green-colored fluorescence face.

The centimetre-scale organic single crystals with face-specific properties were fabricated by crystallization of 1,4-bis[2-(4-methylthienyl)]-2,3,5,6-tetrafluorobenzene, 1 (Fig. 1a),15,16 under optimal conditions. The compound (50 mg) was dissolved in chloroform (2 mL) in a test tube, and then 20 mL of methanol was carefully loaded onto the chloroform solution, to give two organic phases. Slow evaporation of the solvents below 20 °C was carried out for 2 days, and the solution was then decanted and washed with cold methanol (<20 °C). The centimetre-scale crystal 1 (thickness: 79 μm; width: 1185 μm; length: 19 mm) was collected using tweezers (Fig. 1b–g). Macroscopic morphologies at each face of the crystal were determined by optical microscopy. One face (010) was very smooth (Fig. 1d and e), but the other face (001) looked like the lamellar morphology was observed in the direction of the a-axis (Fig. 1f and g). X-ray diffraction (XRD) analysis of the crystal at each face was performed (Fig. 1h and i). The patterns derived from each lamella structures were observed when the (010) and (001) faces were set to be parallel (c-axis is perpendicular to the substrate) and perpendicular (b-axis is perpendicular to the substrate), respectively (Fig. 1h and i). According to the Bragg equation, the length was calculated as 10.7 Å, which corresponds to one lamella layer in the c direction (Fig. 1g). Thus, the face shown in Fig. 1d is the (010) face. The other patterns were calculated to have lengths of 8.05 and 6.66 Å, which are the lamella layers in the b and x-axis directions (Fig. 1e). Thus, the macroscopic lamella face corresponds to the (001) face (Fig. 1f).


image file: c9ce01002e-f1.tif
Fig. 1 (a) Chemical structure of 1. (b) Structure of 1 in crystal form. The crystal structure of 1 featured S⋯F (2.719 Å) and F⋯H (2.218 Å) intramolecular contacts that were significantly shorter than the sums of their van der Waals radii [(dSF = rS + rF = 3.27 Å); (dFH = rF + rH = 2.67 Å)] (top view). These contacts resulted in highly planar molecules with a maximum torsion angle of 1.27° between the tetrafluorophenylene and thiophene units. The molecules formed a slip-stacked structure. The stacking distance was 3.720 Å (side view). The centre-to-centre separation between the thiophene–tetrafluorobenzene–thiophene planes equalled 2.347 Å. The intermolecular distance between end methyl groups was 4.396 Å. (c) Crystal size. (d) Microscope image of the crystal at the (010) face. (e) Crystal morphology of the crystal at the (010) face. (f) Microscope image of the crystal at the (001) face. (g) Crystal morphology of the crystal at the (001) face. (h) 1D XRD patterns of the crystal in the direction of the c-axis. (i) 1D XRD patterns of the crystal in the direction of the b-axis.

XRD analysis of the crystal indicated that the crystal is pure and has long-range (centimetre-scale) structure integrity. Thus, it was assumed that the number of aligned molecules for the a-axis can be calculated using the value of the intermolecular distance between the terminal methyl groups of the π-stacked molecules (4.396 Å). 2.27 × 107 of the molecules were aligned per 1 cm length. The crystal shown here is that ca. 4.31 × 107 of the molecules aligned in the a-axis. On the other hand, the lamellar distances for the b-axis and c-axis were 7.076 and 9.627 Å, respectively (Fig. S1). The crystal shown here is that ca. 1.67 × 106 and ca. 8.20 × 104 of the molecules aligned in the b-axis and in the c-axis.

The optical properties of 1 are shown, before the fluorescence properties at each face of the crystal. Fig. S1 displays the absorption and fluorescence spectra of 1 in tetrahydrofuran (THF) and of the crystal. The absorption spectrum in THF showed a main peak (λmax) at 329 nm (Fig. S1a). On the other hand, the absorption spectrum of the crystal 1, which can be measured by setting the light paths perpendicular to the (010) face (Fig. S1b), showed broad and red-shifted absorption peaks occurring at 391 and 430 nm, compared with those of 1 in THF (Fig. S1a). The crystal, having 79 μm thickness, showed a pale yellow-color. The slip-stacking alignment21 of 1 along the a-axis delocalizes the π-electron, and this leads to a lowering of the HOMO–LUMO band gap. The absorption band of 1 in THF was similar to the excitation peak that occurs at 347 nm.15 The features of the absorption spectrum of the crystal were slightly different to those in the excitation spectrum. However, the characteristic peaks (395 nm and 427 nm) of the excitation spectrum corresponded to those of the absorption spectrum. The absorption and excitation peaks at 427 nm of the crystal are probably due to the specific stacking mode of 1 in the crystal. The fluorescence spectrum of the crystal showed a red-shift when going from the dilute THF solution to the crystal (Fig. S1c). The spectrum in THF can be derived from π–π* fluorescence (405 nm). However, the red-shifted fluorescence band in the crystal (501 nm) is probably due to the slip-stacking structure (J-aggregate) of 1. It can be considered that the highly concentrated solutions of 1 cause the decreasing of the band gap, or the fluorescence spectra of the solutions to be red-shifted. The spectrum of a 1.0 × 10−2 M THF solution of 1 was identical to the spectrum of a 1.0 × 10−6 M THF solution of 1. Therefore, the excimer formation of 1, [11]*, did not occur in a highly concentrated solution (1.0 × 10−2 M). The absorption and fluorescence properties of 1 originate in the crystal. From these experiments, the slip-stacking of 1 in the crystal is important for the evolution of the specific fluorescence properties. The excitation energy transfer across the slip-stacked molecular wires results in sensitized fluorescence.

Fig. 2 displays the single crystal of 1 (a/b/c = 19[thin space (1/6-em)]200/1185/79 μm) at each face, under UV irradiation. The (010) face of the crystal showed sky blue-colored fluorescence (Fig. 2a). On the other hand, green-colored fluorescence was observed at the (001) face (Fig. 2b). This was observed regardless of the direction of UV irradiation. A large single crystal of 1 showed different fluorescence behaviours at each face of the crystal. It is conceivable that the fluorescence spectrum of the crystal (Fig. S1d) is probably a mixture of the fluorescence bands at each face. Thus, we were challenged to determine the fluorescence band of the crystal at the local position. The measurements at each face of the crystal were performed using a fiber probe system that connected an excitation light source and a detector spectrometer (Fig. S2). The excitation and detection were performed coaxially. The spectrum of the (010) face showed an emission (400–600 nm) peak occurring at 481 nm (Fig. 2c), corresponding to the sky blue-colored fluorescence. On the other hand, the bands of the (001) face showed a red-shifted band (420–670 nm) and the peak occurred at 502 nm (Fig. 2c). The fluorescence band correlates to the green-colored fluorescence. The fluorescence spectrum of the crystal measured under the normal method exhibited a broad band (400–670 nm), and the peak occurred at 501 nm with a shoulder peak between 450–490 nm, and this corresponds to the sum of the fluorescence spectra of both the (010) face and the (001) face, respectively (Fig. 2c, black line). The slip-stacked structure of 1 offers the different direction of the molecule in both faces. Therefore, this interesting fluorescence property suggests that the different waveguide occurred according to the orientation of molecule 1.


image file: c9ce01002e-f2.tif
Fig. 2 (a) Fluorescence image of the (010) face of the crystal under UV (365 nm) irradiation. (b) Fluorescence image of the (001) face in the crystal under UV irradiation. (c) Fluorescence spectra of the crystal. The black line was determined from the regular measurement of the crystal. The red and blue lines were determined from the method illustrated in Fig. S2. Red line: (010) face. Blue line: (001) face.

The crystal can be shaped into single crystal fibers. Wherever it is cut, a new face is seen in the cross section. A piece of the crystal 1 can be cut along its length using a surgical knife to produce long and fine fiber shapes.19 If the crystals were smoothly shaped, the obtained crystals would show the green-colored fluorescence (001) face in the cross section. Fig. 3 displays the fluorescence images of the crystals at each face before and after the shaping process. The crystal at the (010) and (001) faces before the shaping process showed sky blue-colored and green-colored fluorescence, respectively (Fig. 3a and b). A green-colored fluorescence line was observed when crystal 1 was cut along its length using a surgical knife (Fig. 3c). The obtained crystal showed the green-colored fluorescence (001) face in the cross section (Fig. 3d and e).


image file: c9ce01002e-f3.tif
Fig. 3 (a) Fluorescence image of the crystal at the (010) face. (b) Fluorescence image of the crystal at the (001) face. (c) Fluorescence image of the crystal cut along its length using a surgical knife. (d) Fluorescence image of the obtained crystal at the (010) face. (e) Fluorescence image of the crystal at the (001) face. (f) Fluorescence image of the obtained crystal fibers (sky blue: b and green: g). (g) 1D XRD patterns of the crystal fibers (* = (021) face). (h) Excitation spectra of a crystal (red line) and the crystal fibers (blue line), excited at 530 nm. (i) Normal fluorescence spectra of a crystal (red line) and the crystal fibers (blue line).

Fig. 3a and f display fluorescence images of the single crystal before (thickness: 79 μm; width: 1185 μm; length: 19 mm) and after shaping (thickness: ∼50 μm; width: ∼50 μm; length: ∼19 mm). The fiber-shaped organic single crystal under UV irradiation showed two types of sky-blue and green-colored fluorescence. The sky blue-colored fluorescence was derived from the (010) face of the crystal fiber (Fig. 3f), and this is similar to the results in Fig. 2a and 3a. On the other hand, the green-colored fluorescence at the (001) face of the crystal fiber was also observed (Fig. 3f). This is because the direction of the crystal fiber faces towards the detector (camera). XRD analysis of the fibers on a glass plate showed the patterns of the (010), (001) and (001) faces (Fig. 3g). Blue-coloured fluorescence was observed when the (010) face of the fibers is facing upward (c-axis is perpendicular to the substrate). On the other hand, the (001) face of the fibers (b-axis is perpendicular to the substrate) showed green-colored fluorescence. The fluorescent properties of the crystal fibers remained after shaping. The mechanical processing (cutting) of crystal 1 into fine crystalline fibers increases the total area of the green-colored fluorescent face.

The excitation and fluorescence spectra of both the crystal and the fibers were measured using the normal setup (Fig. S3). The excitation and fluorescence bands of the fibers dominated at 445 and 500 nm, respectively (Fig. 3h and i). These bands are thought to be derived from the (001) face, while the bands around 390 and 460 nm are derived from the (010) face (Fig. 3h and i). A large band change was caused by the increase of the (001) face area. The absolute fluorescent quantum yield (Φ) of the crystalline materials was increased by the shaping. The Φ value of the crystal was 25%, but the value for the fibers was 33%. The fibers have a larger surface of the (010) face than that of the (001) face. This is because the fluorescence efficiency of the green-fluorescent (001) face is higher than that of the sky-blue fluorescent (010) face.

XRD analysis of the crystal fibers at each face was performed (Fig. 4). The fibers were fixed by gel on a glass plate, and they showed blue-colored fluorescence at the (010) face and green-colored fluorescence at the (001) face (Fig. 4a and b). Patterns derived from each lamella structure were also observed when the (010) and (001) faces were set to be parallel (c-axis is perpendicular to the substrate) and perpendicular (b-axis is perpendicular to the substrate), respectively (Fig. 4c and d). Moreover, we found the angle that shows the patterns of the (021) face (Fig. 4e). These patterns were attributed to each lamella spaces (Fig. 4f).


image file: c9ce01002e-f4.tif
Fig. 4 (a) Fluorescence image of the crystal fiber at the (010) face. (b) Fluorescence image of the crystal fiber at the (001) face. (c) 1D XRD patterns of the crystal fiber in the direction of the c-axis. (d) 1D XRD patterns of the crystal fiber in the direction of the b-axis. (e) 1D XRD patterns of the crystal fiber in the direction of the x-axis. (f) Crystal structure at the (100) face and a schematic illustration of the lamella spaces.

In summary, we have observed the facet-specific fluorescence behaviour of a centimetre-scale single crystal based on a π-conjugated molecule, 1,4-bis[2-(4-methylthienyl)]-2,3,5,6-tetrafluorobenzene. The crystal was on the centimetre-scale with long-range 3D structure integrity. The (010) face of the crystal showed sky blue-colored fluorescence, but the green-colored fluorescence was observed at the (001) face of the crystal. Interestingly, the mechanically induced shaping of a centimetre-scale single crystal also exhibited direction-specific fluorescence behaviours at each face. Other fluorescent fibril lamella crystals are candidates for this work.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

S. H. acknowledges KAKENHI (Grant-in-Aid for Scientific Research B: no. 18H02052, Grant-in-Aid for Scientific Research on Innovative Areas ‘π-figuration’: no. 17H05171 and ‘coordination asymmetry’: no. 19H04604) of the Japan Society for the Promotion of Science (JSPS).

Notes and references

  1. J. Kang, D. Miyajima, T. Mori, Y. Inoue, Y. Itoh and T. Aida, Science, 2015, 347, 646–651 CrossRef CAS .
  2. T. Fukui, M. Takeuchi and K. Sugiyasu, Sci. Rep., 2017, 7, 2425 CrossRef .
  3. S. Ogi, K. Sugiyasu, S. Manna, S. Samitsu and M. Takeuchi, Nat. Chem., 2014, 6, 188–195 CrossRef CAS .
  4. N. Seiki, Y. Shoji, T. Kajitani, F. Ishiwari, A. Kosaka, T. Hikima, M. Takata, T. Someya and T. Fukushima, Science, 2015, 348, 1122–1126 CrossRef CAS .
  5. F. K.-C. Leung, F. Ishiwari, T. Kajitani, Y. Shoji, T. Hikima, M. Takata, A. Saeki, S. Seki, Y. M. A. Yamada and T. Fukushima, J. Am. Chem. Soc., 2016, 138, 11727–11733 CrossRef CAS .
  6. H. Shioya, Y. Shoji, N. Seiki, M. Nakano, T. Fukushima and Y. Iwasa, Appl. Phys. Express, 2015, 8, 121101 CrossRef .
  7. F. Ishiwari, Y. Shoji and T. Fukushima, Chem. Sci., 2018, 9, 2028–2041 RSC .
  8. T. Tachikawa, S. Yamashita and T. Majima, J. Am. Chem. Soc., 2011, 133, 7197–7204 CrossRef CAS .
  9. N. Tokunaga, Y. Fujiki, S. Shinkai and K. Sada, Chem. Commun., 2006, 3617–3618 RSC .
  10. Y. Fujiki, N. Tokunaga, S. Shinkai and K. Sada, Angew. Chem., Int. Ed., 2006, 45, 4764–4767 CrossRef CAS .
  11. Y. Fujiki, S. Shinkai and K. Sada, Cryst. Growth Des., 2009, 9, 2751–2755 CrossRef CAS .
  12. S. Hotta and T. Yamao, J. Mater. Chem., 2011, 21, 1295–1304 RSC .
  13. S. Hotta and M. Goto, Adv. Mater., 2002, 14, 498–501 CrossRef CAS .
  14. S. Hotta, M. Goto, R. Azumi, M. Inoue, M. Ichikawa and Y. Taniguchi, Chem. Mater., 2004, 16, 237–241 CrossRef CAS .
  15. T. Yamao, Y. Taniguchi, K. Yamamoto, T. Miki, S. Ota, S. Hotta, M. Goto and R. Azumi, Jpn. J. Appl. Phys., 2007, 46, 7478–7482 CrossRef CAS .
  16. Z. Yu, Y. Wu, Q. Liao, H. Zhang, S. Bai, H. Li, Z. Xu, C. Sun, X. Wang, J. Yao and H. Fu, J. Am. Chem. Soc., 2015, 137, 15105–15111 CrossRef CAS .
  17. S. Hayashi and T. Koizumi, Angew. Chem., Int. Ed., 2016, 55, 2701–2704 CrossRef CAS .
  18. S. Hayashi, A. Asano, N. Kamiya, Y. Yokomori, T. Maeda and T. Koizumi, Sci. Rep., 2017, 7, 9453 CrossRef .
  19. S. Hayashi and T. Koizumi, Chem. – Eur. J., 2018, 24, 8507–8512 CrossRef CAS .
  20. S. Hayashi, Polym. J., 2019, 51, 813–823 CrossRef CAS .
  21. D. J. Crouch, P. J. Skabara, M. Heeney, I. McCulloch, S. J. Coles and M. B. Hursthouse, Chem. Commun., 2005, 1465–1467 RSC .

Footnote

Electronic supplementary information (ESI) available: Procedures, images of the crystal, crystal data, absorption and fluorescence spectra, XRD results and elasticity data. See DOI: 10.1039/c9ce01002e

This journal is © The Royal Society of Chemistry 2019