LETTERS

Reproducible on–off switching ofsolid-state luminescence by controllingmolecular packing throughheat-mode interconversionTOSHIKI MUTAI*, HIROYUKI SATOU AND KOJI ARAKI*Institute of Industrial Science, University of Tokyo, 4-6-1, Komaba, Meguro-ku, Tokyo 153-8505, Japan

*e-mail: mutai@iis.u-tokyo.ac.jp; araki@iis.u-tokyo.ac.jp

Published online: 21 August 2005; doi:10.1038/nmat1454

Organic luminescent solids are attracting increasinginterest in various fields of application1–3. Modification oralteration of the chemical structures of their component

molecules is the most common approach for tuning theirluminescence properties. However, for dynamic tuning orswitching of solid-state luminescence with high efficiencyand reproducibility successful examples are limited2,4 aschemical reactions in the solid state frequently encounterinsufficient conversion, one-way reactions or loss of theirluminescence properties. One promising approach is to controlthe luminescence properties by altering the mode of solid-statemolecular packing without chemical reactions. Here, we showthat 2, 2′:6′, 2′′-terpyridine5, practically non-luminescent inthe form of amorphous solid or needle crystal6, shows strongblue luminescence upon formation of a plate crystal. Efficientand reproducible on–off switching of solid-state luminescenceis demonstrated by heat-mode interconversion between theplate and needle crystals. Because alteration of the mode ofmolecular packing does not require chemical reactions, thepresent findings would open the way for the development ofnovel organic luminescent solids that can be switched on andoff by external thermal stimuli.

Tuning and switching of the solid-state luminescence oforganic compounds are attractive targets for both fundamentalresearch and practical applications. To use the mode of molecularpacking in the solid state for this purpose can be a promisingapproach. Although there have been some reports on the crystalpolymorph of organic fluorescent compounds7–9, the effects ontheir fluorescent properties were generally limited. Among them,several non-emissive organic molecules and metal complexes,such as siloles10,11, stilbenes12, α-pyrone13 and [Pt(bpy)2]Cl2

(bpy = 2,2′-bipyridine)14, have been found to show inducedluminescence in the solid state. However, efficient and reliableswitching of solid-state luminescence by the mode of molecularpacking has not yet been established.

Table 1 Absorption labs and luminescence lem of tpy in solution and in solids.

labs (nm) lem (nm) Φ τ (ns)

Solution∗ 279 335 0.003‖Amorphous 305±2.5† 348±0.7‡ <0.01¶

Needle 305±3.1† 353±0.4§ <0.01¶ 0.8Plate 308±3.1† 365±0.4§ 0.2¶ 4.5∗ In cyclohexane.

† Determined from diffractive reflection spectrum with standard error (n= 4).

‡ Standard error (n= 4).

§ Standard error (n= 13).

‖ Determined using 2-aminopyridine as the standard (ethanol; excitation at 285 nm; Φ = 0.37).

¶ Determined using anthracene as the standard (excitation at 300 nm, Φ = 0.5).

2,2′:6′,2′′-terpyridine (tpy) is generally non-fluorescent, andonly a limited number of its derivatives15–17 are fluorescent.Indeed, tpy after purification showed practically no fluorescencein various organic solvents. An amorphous tpy solid was obtainedby freeze-drying a cyclohexane solution, which was confirmed tobe amorphous from the absence of any specific patterns in powderX-ray diffraction (XRD). Two types of tpy crystals were preparedby a controlled decrease of the temperature of a hot hexanesolution. A faster temperature decrease gave the needle crystal(melting point 86–88 ◦C), whereas a slow decrease yielded the platecrystal (melting point 91–93 ◦C). Although the amorphous solidand the needle crystal of tpy were practically non-luminescent,the plate crystal displayed a strong luminescence upon irradiationwith ultraviolet light. It is often the case that a trace amount ofemissive impurity works as a luminescent centre because of efficientenergy transfer in the solid state18. In our case, however, repeatedinterconversions between the luminescent plate crystal and thenon-luminescent needle crystal by controlled recrystallization fromhexane ruled out possible contamination of the emissive impurity

nature materials VOL 4 SEPTEMBER 2005 www.nature.com/naturematerials 685

Untitled-15 1 8/11/05, 1:12:49 PM

Nature Publishing Group© 2005

LETTERS

b

c

ac

b

a

b

b

a

Figure 1 X-ray crystallographic pictures of tpy crystals. The black parts in eachmolecule indicate nitrogen atoms. a, The plate crystal. b, The needle crystal alongthe a axis (left) and the c axis (right). Crystal data, plate crystal: C15H11N3,FW= 233.27, crystal size 0.48×0.28×0.25 mm3, monoclinic, P21/c,a= 11.7040(3) A, b= 15.5590(3) A, c= 13.5920(3) A, β = 109.105(2)◦,V= 2338.81(9) A3, Z= 8, calculated density = 1.325, l (Cu Kα) = 1.54184 A,R1 = 0.0690, Rw = 0.2048 (all data) and GOF= 1.087. Needle crystal5: crystal size0.23×0.11×0.09 mm3, orthorhombic, P212121, a= 3.8450(2) A,b= 16.5250(8) A, c= 17.7030(9) A, V= 1,124.82(10) A3, Z= 4, calculateddensity = 1.377, l (Cu Kα) = 1.54184 A, R1 = 0.0869, Rw = 0.2138 (all data) andGOF = 1.116. For definition of the crystallographic parameters, see ref. 22.

in the luminescent plate crystal. The non-luminescent needlecrystal became luminescent upon transformation to the platecrystal and returned to the non-luminescent needle crystal againby subsequent recrystallization, confirming that the tpy moleculein the plate crystal was a luminescent species.

The absorption spectra of tpy in the two crystal states appearedin the same region with a similar shape, suggesting a similarphoto-excitation process in both crystals. As shown in Table 1,the luminescence quantum yield (Φ) of the organic solution, theamorphous solid or the needle crystal are all smaller than 0.01 andpractically non-luminescent. However, that of the plate crystal ishigh enough to be 0.2, which is more than an order of magnitudehigher compared with that of the needle crystal or the amorphoussolid. As far as we know, such a distinct luminescence behaviourthat is caused by the difference in the mode of solid-state molecularpacking of organic compounds has not been reported before. Theemission maximum of the plate crystal was observed to be 12 nmlonger compared with that of the needle crystal. The luminescencedecay times (τ) of the plate and the needle crystals were 4.5 and0.8 ns, respectively, indicating higher stability of the photo-excitedstate in the plate crystal.

(111

) (222

)

Inte

nsity

(020

) (002

)

(121

)

(230

)

(011

)

(002

)(0

20) (012

)(0

21)

(022

) (031

)

(004

)

(013

)

2θ (deg)θ

(012

)

(102

)

(240

)

(102

)(1

11)

(112

)

–

–

10 3020

a

b

c

d

Figure 2 Powder XRD patterns of tpy solids. a,b, Simulated powder XRD patternsof the plate crystal (a) and the needle crystal (b) calculated from theircrystallographic data. c,d, Measured powder XRD patterns of the solid after heatingthe needle crystal at 88 ◦C for 10 min and then cooling to room temperature (c) andsubsequently heating to 120 ◦C then cooling to room temperature (d).

Figure 1 shows the structures of the plate (P21/c) and theneedle5 (P212121) crystals. At the single molecular level, tpy wasin the s-trans conformation, but was slightly deviated from theco-planar state in both crystals. The most notable conformationaldifference was the dihedral angles of the pyridine rings ineach crystal. In the plate crystal, the pyridine rings at bothends were twisted anticlockwise to the central pyridine ring(molecule A: 3.59◦, 5.30◦; molecule B: 7.28◦, 1.05◦), whereas theywere twisted anticlockwise and clockwise to the central pyridinering (5.96◦, −4.73◦) in the needle crystal (see SupplementaryInformation). The observed conformational difference must beclosely related to the molecular packing in each crystal. As tothe molecular arrangement, columnar stacks of the molecules(inter-plane distance = 3.37 A) were present in the needle crystal,but no such molecule stacks were found in the plate crystal and,instead, dimer-like units (shortest inter-plane distance = 3.47 A)were observed.

The molecular orbital calculations for tpy at the singlemolecular level with geometries in the optimized (co-planars-trans conformation) and crystal states were performed by time-dependent density functional theory19 using the B3LYP/6-31G∗

basis set. The results suggested that the difference in twistingangles of the pyridine rings altered the contributions of the n–π∗

(forbidden) and π–π∗ (allowed) transitions to the lowest energyabsorption band (see Supplementary Information), suggesting thatthe conformational fixation due to the crystal packing could be apossible reason for the notable luminescence induction. However,the results are still at the preliminary stage and they have to beexamined further by assessing the effect of the intermolecularinteraction in order to understand the solid-state luminescence.

Furthermore, we could efficiently alter the mode of thesolid-state molecular assembly of tpy by a controlled heating–coolingprocess of the solid sample, opening the way for a simple and

686 nature materials VOL 4 SEPTEMBER 2005 www.nature.com/naturematerials

Untitled-15 2 8/11/05, 1:12:50 PM

Nature Publishing Group© 2005

LETTERS

Rela

tive

inte

nsity

Number of cycles (times)

10

1

1 2 30

Figure 3 A plot of relative intensity of the solid-state luminescence of tpyagainst the number of heat–cooling cycles starting from the needle crystal. Theopen and filled circles indicate the tpy solid after the heating and coolingprocesses, respectively.

effective heat-mode switching of the solid-state luminescence by adry process. In the first heating curve (heating rate = 10 ◦C min−1)of differential scanning calorimetry, the plate crystal showed a clearsingle endothermic peak (melting enthalpy �H = 17.5 kJ mol−1;melting entropy �S = 48.0 J mol−1 K−1) at its melting point(91.1 ◦C), but the needle crystal showed vague (83.7 ◦C, peak top)and sharp (91.1 ◦C) endothermic peaks. The latter sharp peak wasidentical to that of the plate crystal in terms of temperature andthe values of �H and �S, indicating that the first vague peak wasascribed to the transition of the needle crystal to the plate crystal.This was further confirmed by XRD analysis. The powder XRDpattern of the solid (Fig. 2c) after heating the needle crystal at 88 ◦Cfor 10 min was different from that of the needle crystal (Fig. 2b) butwas almost identical to that of the plate crystal (Fig. 2a), confirmingthe vague peak to be the needle-to-plate crystal phase transition.When the tpy solid either in the crystal or the amorphous statewas heated above its melting point to 120 ◦C and cooled to roomtemperature (cooling rate 10 ◦C min−1), the resultant solid showedpowder XRD peaks (Fig. 2d) corresponding to those of the needlecrystal. Relatively weak diffraction signals in this case suggestedthat the solid was in a mixed state of the amorphous solid and theneedle crystal.

On the basis of these observations, we tested the reversibleheat-mode switching of the solid-state luminescence of tpy. Whenthe needle crystal was heated at 89.5 ◦C for 10 min and cooledto room temperature (cooling rate 5 ◦C min−1), the solid showedthe strong blue luminescence of the plate crystal. The solid wasthen reheated to 100 ◦C and cooled down to room temperature(cooling rate 5 ◦C min−1). The resultant solid showed only aweak luminescence. The repeated thermal cycles demonstratedthat luminescence switching was reproducible without any sign ofdegradation or chemical reaction of the tpy molecule; the on/offratio of the luminescence intensities was sufficiently large to benearly one order of magnitude, confirming that effective heat-modeluminescence switching was achieved (Fig. 3). It is to be notedthat both the luminescent and non-luminescent states were stableenough during our experimental period (nearly a year) when theywere kept at room temperature without heating.

The strong blue luminescence of tpy in the plate crystal presentsa clear example of an induced luminescence based on the modeof molecular packing instead of chemical reactions in organicsolids. The effective and reproducible switching of this solid-state luminescence was realized by the heat-mode interconversionbetween the plate and needle crystals. Heat-mode phase-changetechnology has become the mainstream of rewritable opticalmedia20, where change of reflectivity is mostly used as output.Because using luminescence output has many advantages, such

as high signal-to-noise ratio, sensitive detection and the potentialfor two-dimensional imaging, novel types of organic luminescentmaterials that can be switched on and off reliably and efficientlyby external thermal stimuli are attractive candidates for rewritableoptical media, information storage and imaging, and other newphoto-electronic applications.

METHODSTpy solids were prepared from commercial tpy (Aldrich) after being purified by aluminium oxidecolumn chromatography eluted with hexane/tetrahydrofuran (4:1). The absorption spectra of theamorphous and crystal tpy were obtained by measurement of the diffractive reflection spectrumfollowed by Kubelka–Munk conversion. The value of Φ for each solid was determined by usinganthracene as the standard21 (Φ = 0.5). Differential scanning calorimetry was measured on a Pyris 1system (Perkin Elmer) with a temperature range of −5 to 120 ◦C at the rate of 10 ◦C min−1 .

HEAT-MODE SWITCHING OF SOLID-STATE LUMINESCENCEThe needle crystal (1 mg) was ground and placed on a glass plate 2 mm×10 mm in size. Heating andcooling of the sample glass plate were done with a hot stage (Mettler FP 82 HT) at therate of 5 ◦C min−1 .Off-to-On process: The sample glass plate was heated from 75 to 89.5 ◦C and maintained for 10 min.After cooling, the luminescence spectrum was measured at room temperature (excitation at 300 nm).On-to-Off process: The sample glass plate was heated from 85 to 100 ◦C and maintained for 1 min.The luminescence spectrum was measured after cooling.

Received 3 March 2005; accepted 23 June 2005; published 21 August 2005.

References1. Coe, S., Woo, W. K., Bawendi, M. & Bulovic, V. Electroluminescence from single monolayers of

nanocrystals in molecular organic devices. Nature 420, 800–803 (2002).2. Irie, M., Fukaminato, T., Sasaki, T., Tamai, N. & Kawai, T. Organic chemistry: A digital fluorescent

molecular photoswitch. Nature 420, 759–760 (2002).3. Taniguchi, Y. Organic solid-state laser. J. Photopolym. Sci. Technol. 15, 183–184 (2002).4. Papaefstathiou, G. S., Zhong, Z., Geng, L. & MacGillivray, L. R. Coordination-driven self-assembly

directs a single-crystal-to-single-crystal transformation that exhibits photocontrolled fluorescence.J. Am. Chem. Soc. 126, 9158–9159 (2004).

5. Albano, G., Balzani, V., Constable, E. C., Maestri, M. & Smith, D. R. Photoinduced processes in4′-(9-anthryl)-2,2′ :6′ ,2′′-terpyridine, its protonated forms and Zn(II), Ru(II) and Os(II) complexes.Inorg. Chim. Acta 277, 225–231 (1998).

6. Bessel, C. A., See, R. F., Jameson, D. L., Churchill, M. R. & Takeuchi, K. J. Structural considerations ofterdentate ligands – crystal structures of 2,2′ :6′ ,2′′-terpyridine and 2,6-bis(pyrazol-1-yl)pyridine.J. Chem. Soc. Dalton Trans. 3223–3228 (1992).

7. Davis, R., Rath, N. P. & Das, S. Thermally reversible fluorescent polymorphs ofalkoxy-cyano-substituted diphenylbutadienes: role of crystal packing in solid state fluorescence.Chem. Commun. 74–75 (2004).

8. Botta, C. et al. Optical properties and photoluminescence of tetrahexyl-sexithiophene allotropes.Synth. Met. 122, 395–399 (2001).

9. Sanz, N., Baldeck, P. L., Nicoud, J. F., Fur, Y. L. & Ibanez, A. Polymorphism and luminescenceproperties of CMONS organic crystals: bulk crystals and nanocrystals confined in gel-glasses. SolidState Sci. 3, 867–875 (2001).

10. Luo, J. et al. Aggregation-induced emission of 1-methyl-1,2,3,4,5-pentaphenylsilole. Chem. Commun.1740–1741 (2001).

11. Chen, J. et al. Synthesis, light emission, nanoaggregation, and restricted intramolecular rotation of1,1-substituted 2,3,4,5-tetraphenylsiloles. Chem. Mater. 15, 1535–1546 (2003).

12. An, B.-K., Kwon, S.-K., Jung, S.-D. & Park, S. Y. Enhanced emission and its switching in fluorescentorganic nanoparticles. J. Am. Chem. Soc. 124, 14410–14415 (2002).

13. Hirano, K., Minakata, S., Komatsu, M. & Mizuguchi, J. Intense blue luminescence of 3,4,6-triphenyl-α-pyrone in the solid state and its electronic characterization. J. Phys. Chem. A 106,4868–4871 (2002).

14. Miskowski, V. M., Houlding, V. H., Che, C.-M. & Wang, Y. Electronic spectra and photophysics ofplatinum(II) complexes with α-diimine ligands. Mixed complexes with halide ligands. Inorg. Chem.32, 2518–2524 (1993).

15. Mutai, T., Arita, S. & Araki, K. Phenyl-substituted 2,2′ :6′ ,2′′-terpyridine as a new series of fluorescentcompounds - their photophysical properties and fluorescence tuning. J. Chem. Soc. Perkin Trans. 21045–1050 (2001).

16. Mutai, T., Cheon, J.-D., Tsuchiya, G. & Araki, K. 6-Amino- 2,2′ :6′ ,2′′-terpyridines as highlyfluorescent compounds - effect of the number of pyridine rings on fluorescence properties. J. Chem.Soc. Perkin Trans. 2 862–865 (2002).

17. Wong, W. L., Huang, K. H., Teng, P. F., Lee, C. S. & Kwong, H. L. A novel chiral terpyridinemacrocycle as a fluorescent sensor for enantioselective recognition of amino acid derivatives. Chem.Commun. 384–385 (2004).

18. Benz, K. W. & Wolf, H. C. Concentration dependence of the energy transfer in anthracene mixedcrystals. Z. Naturforsch. A 19, 177–181 (1964).

19. Frisch, M. J. et al. Gaussian 98 Revision A.9. (Gaussian, Inc., Pittsburgh, 1998).20. Kolobov, A. et al. Understanding the phase-change mechanism of rewritable optical media. Nature

Mater. 3, 703–708 (2004).21. Ates, S. & Yildiz, A. Determination of the absolute quantum efficiency of the luminescence of

crystalline anthracene and of meso-dimeso derivatives using photoacoustic spectroscopy.J. Chem. Soc. Faraday Trans. 1 79, 2853–2861 (1983).

22. Hall, S. R., Allen, F. H. & Brown, I. D. in International Tables for Crystallography Volume G:Definition and Exchange of Crystallographic Data (eds Hall, S. R. & McMahon, B.) 210–257 (Kluwer,Netherlands, 2005).

AcknowledgementsThe authors thank K. Ogawa and T. Fujiwara, University of Tokyo, for their advice on measuringdiffractive reflectance spectra. This work was supported by a Grant-in-Aid for Scientific Research onPriority Areas (417) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT)of Japan.Correspondence and requests for materials should be addressed to T.M. or K.A.Supplementary Information accompanies this paper on www.nature.com/naturematerials.

Competing financial interestsThe authors declare that they have no competing financial interests.

nature materials VOL 4 SEPTEMBER 2005 www.nature.com/naturematerials 687

Untitled-15 3 8/11/05, 1:12:51 PM

Nature Publishing Group© 2005