NANO LETTERS
Local Spectroscopy of Individual Hexaphenyl Nanofibers
2002 Vol. 2, No. 12 1379-1382
Adam Cohen Simonsen and Horst-Gu1 nter Rubahn* Fysisk Institut, Syddansk UniVersitet, CampusVej 55, DK-5230 Odense M, Denmark Received September 23, 2002; Revised Manuscript Received October 24, 2002
ABSTRACT Long, isolated nanofibers made of light-emitting p-hexaphenyl molecules are grown in the focus of an argon ion laser and are investigated via local luminescence spectroscopy. The luminescence is excited in 20 µm wide sections of the nanofibers. As one approaches the end of the nanofiber, the line width narrows by a factor of 2 and a fwhm of less than 10 nm is observed. The spectral narrowing is correlated to a decreasing cross-profile along the nanofibers, which is measured by atomic force microscopy.
1. Introduction. It is well-known that the optical properties of organic aggregates and ultrathin films are strongly affected by the intrinsic degree of structural order. Systematic studies have been performed, e.g., on oligo-thienyl films1 or alphasexithienyl single crystals2 and films.3 Fluorescence and photoluminescence have turned out to be sensitive tools for studying subtle changes in the optical properties, since these techniques are sensitive to localized electronic states that may be induced via crystal imperfections or local disorder. A recent study of the influence of hydrostatic pressure on the photoluminescence of a methylated poly-para-phenylene film revealed that increasing pressure leads to a broadening of the transition lines due to an increase in intrachain coupling between two conjugated backbones.4 A powder of parahexaphenyl, on the other hand, did not show pressuredependent broadening, but line widths of the order of 24 nm (fwhm). Compared to ideal single crystals, the spectral lines of ultrathin films are usually far more congested by broadening due to structural defects, even for films that are grown on single crystalline substrates. In the case of nanoaggregates such as needles or nanofibers made of light-emitting organic material, one might expect to reduce the effect of disorder if the size of the aggregates becomes small enough. Previously, spectra of backscattered light from individual micrometer-long5 or submicrometer-long6 metal rods have been reported. Line widths (fwhm) of the order of 50 nm5 and 29 nm6 have been measured. In the case of metallic nanorods, the collective electronic excitation gives rise to surface plasmons which decay with characteristic time constants of the order of femtoseconds or tens of femtoseconds. For unordered polycrystalline films of para-hexaphenyl, the electronic excitation lasts for hundreds of picoto nanoseconds.7 A similar excitation lifetime might be * Corresponding author. E-mail:
[email protected] 10.1021/nl025807z CCC: $22.00 Published on Web 11/08/2002
© 2002 American Chemical Society
expected for organic nanofibers, which then should provide very sharp spectral lines of the order of small fractions of a nanometer. However, lattice vibrations give rise to additional broadening effects. These broadening effects should be greatly diminuished by cooling the samples. Recent photoluminescence studies on para-hexaphenyl films as a function of film temperature show indeed a decrease of line width, albeit only very weak. For the (0-1) band the line width (fwhm) decreases from 42 nm at 270 K to 33 nm at 30 K.8 Hence the main contribution to the observed large line width must stem from other inhomogeneous broadening mechanisms such as structural defects, emission from aggregates, and X-traps,9 etc. In the present experiment we investigate aggregates that are strongly confined in width and height. Due to the decreasing influence of intermolecular interactions and defects on the photoluminescence spectrum, one would expect a narrowing of the inhomogeneous part of the spectral lines. This narrowing should continue until one reaches the limit where the residual inhomogeneous broadening is given only by the contribution from nonequivalent lattice points at which the molecules are placed. We use para-hexaphenyl as the organic material for our studies. para-Hexaphenyl is an oligomer of the conjugated polymer, poly-para-phenylene. It consists of a linear chain of six phenyl rings and has a band gap of 3.2 eV. The vibronic S1 f S0 exciton spectrum in the solid state is dominated by intense Raman active modes that are due to C-C stretching vibrations of all carbon atoms of the molecule. From theory, these modes are predicted to possess energies around 0.2 eV,10 which agrees with experimental Raman spectra that provide 0.197 eV.11 Due to a strong coupling of the Raman mode to the singlet electronic transition and a reduction of the length of the central interring bond in the excited state by 0.004 nm12, the (0-0) mode is
Figure 1. (a) Fluorescence microscopy of two well-separated nanofibers, demonstrating the extremely good parallel alignment of the nanofibers. (b) Atomic force microscopy of the end of the fiber on the right-hand side of (a). The dots are organic material from which the nanofiber is grown. (c) Blow up of the end region. Height scale 70 nm.
relatively weak as compared to the higher modes. Especially in the powder phase, reabsorption reduces the intensity of the (0-0) mode additionally. The effect of film order on the relative intensities of the vibronic modes has been studied thoroughly for a similar molecule, namely alpha-sexithiophene.13 Luminescence spectra of para-hexaphenyl thin films usually show vibronic excitations up to (0-4) or even higher.14 Since the quantum yield of the luminescence in the solid state is of the order of 0.315 and since the emitted light is strongly anisotropic,16 para-hexaphenyl is a very interesting material for micro- or nanoscaled device applications. If one adjusts the growth conditions properly, then needle-shaped aggregates (“nanofibers”) can be generated on alkali halide17 or mica surfaces.18,14 On alkali halides as substrates, the orientation of the nanofibers is given by cleavage edges on the substrate. In contrast, on mica a dipole-assisted epitaxial growth process has been observed.20,19 The latter growth mode depends critically on surface temperature and can thus be controlled to a large extend by laser surface heating.21 In the present paper we take strong advantage of that nanofiber growth control. 2. Experimental Results and Discussion. The growth of nanofibers made of p-6P molecules in the focal region of an argon ion laser has been described previously.21 Briefly, hexaphenylene molecules were evaporated at temperatures above 600 K from a Knudsen cell onto a cleaved and outgassed mica surface with deposition rates of less than 0.02 Å/s at a surface temperature of 356 K. In the course of film growth, an argon ion laser with a power of 0.5 W has been focused onto the substrate. The nominal mass thickness of the film used for the present experiment was 5 nm, as determined by a water-cooled quartz microbalance. Muscovite mica is a sheet silicate, which upon cleavage develops surface dipoles. The surface dipole field has a 1380
strength of the order of 107 V/cm,22 and the p-6P molecules grow parallel to the direction of this field since the interaction energy between the surface dipoles and the polarizable molecules is larger than the thermal energy of the molecules.19 Once the orientation of individual molecules is defined by their alignment parallel to the surface dipoles, needle-shaped aggregates are formed if the substrate temperature is sufficiently high. The aggregates grow nearly perpendicular to the direction of the individual molecules axes20 and they are mutually parallel oriented. The main effect of the laser irradiation under these conditions is to decrease the nanofiber density and to allow the growth of well separated nanofibers.21 The quality of this growth manipulation is demonstrated by a fluorescence micrograph of two parallel needles in Figure 1a. To obtain the micrograph, the substrate was placed inside a fluorescence microscope and irradiated with a mercury lamp, providing a spectrum with a central wavelength of 365 nm. The emission of the nanofibers has been observed behind a 420 nm long pass filter with a highresolution digital camera. The specific appearance of the environment around the nanofibers allows us to locate and to characterize the same aggregates by atomic force microscopy (AFM) as were observed optically. AFM was done with a PicoSPM (Molecular Imaging) operated in tapping mode. A larger area fluorescence micrography of the surroundings of the nanofibers of Figure 1a is shown in Figure 2a. The same characteristic environment is then observed via a microscope objective mounted below the AFM tip (Figures 2b and 2c). The nanofiber in Figure 1b, imaged by the AFM, is the nanofiber on the right-hand side of Figure 1a. Once the optically investigated nanofiber is identified, the AFM enables one to investigate in detail its morphological structure. A blow up of the fiber-end region (800 × 400 Nano Lett., Vol. 2, No. 12, 2002
Figure 3. Luminescence spectra of the end of the nanofiber of Figure 1 (circles) and of a homogeneous p-6P film24 (solid black line). The gray curve is a Gaussian fit to the fiber spectrum.
Figure 2. (a) Fluorescence microscopy of the surroundings of the two nanofibers from Figure 1. (b) Darkfield microscopy image, taken with a CCD camera below the atomic force microscope (AFM). The AFM tip used for the measurements is also visible. (c) Setup for identifying individual nanofibers.
nm2; Figure 1c) reveals a nanofiber height of 60 nm and a width of 100 nm at a position 50 nm apart from the end. The nanofiber surface has step edges that are tilted by 70 to 80 degrees with respect to the long axis. The value of this tilt is in agreement with the crystallographic structure that has been determined via X-ray diffraction and via low energy electron diffraction,20 albeit not for individual nanofibers. Recent investigations of nanofibers via selected area electron diffraction suggest that the nanofibers are made of crystallites with orientations parallel or perpendicular to the nanofiber long axis.23 The crystallites are separated by grain boundaries. Nano Lett., Vol. 2, No. 12, 2002
Figure 3 shows spectrally resolved luminescence from the nanofiber of Figure 1b (circles). The excitation radiation from the high-pressure mercury lamp was focused to a spot with a Gaussian radius of about 10 µm, which illuminated the nanofiber 20 µm apart from its tip. The resulting luminescence was spectrally resolved by a miniature spectrometer (Ocean Optics S2000), which was coupled to the fluorescence microscope via an optical fiber with 50 µm core diameter. The spectral resolution of the spectrometer is 2.5 nm. The solid gray line is a Gaussian fit to the measured line, providing a deconvoluted line width (fwhm) of 7.4 nm. For comparison, the spectrum of a homogeneous, polycrystalline hexaphenyl film on an indium-tin oxide/glass substrate is also shown as solid black line.24 The line widths of the three visible lines in that case are of the order of 20 nm. Next we translated the nanofiber laterally within the fluorescence microscope with piezo-driven translational stages. The measured luminescence line width after Gaussian fitting and deconvolution is shown as a function of spatial position on the nanofiber in Figure 4a. The zero-point is taken as that point where the signal-to-noise ratio becomes less than 5. At lower signal levels a determination of the line width becomes virtually impossible. Note, however, that due to the spatial extension of the excitation source, the whole curve could well be shifted 10 micrometers to the left. Obviously, a significant increase of line width is observed if one compares the measured line widths at distances 25 µm and 58 µm from the nanofiber end. Following the optical investigation, the same nanofiber has also been investigated with the AFM. From highresolution scans such as the one in Figure 1c we have determined the width of the nanofiber (see X-scan in Figure 1c) as a function of distance to the end (Figure 4b). Here, the zero point is taken as the point where the nanofiber height is just 5 nm above the horizontal level of the mica substrate. Whereas the height of the nanofiber rises to its final value of 60 nm within the first 100 nm apart from the end (see Y-scan in Figure 1c), its width reaches the value of 100 nm about 60 nm apart from the end. It then remains constant for the next 20 µm, increases steeply by a factor of 2, and again remains constant for distances from the end larger than 1381
We note finally that even for the present case of nanofibers on samples that are hold at room temperature the obtained line widths are comparable to those obtained for isolated quantum dots, where one finds typically line widths (fwhm) of a few nanometers in the visible spectral regime. This observation points to a huge potential applicability of the nanofibers as the optically active elements in future nanooptronics. Acknowledgment. We are grateful to the Danish Research Foundation SNF for financial support and to F. Balzer for preparation of the samples. A.C.S. thanks the Danish National Research Foundation for support via a grant to the MEMPHYS-Center for Biomembrane Physics. References
Figure 4. (a) Measured spectral half width of the (0-1) vibronic transition in a single nanofiber as a function of distance from the end of the nanofiber. (b) Measured distance dependence of the width of the nanofiber.
40 µm. Most interestingly, the change in morphology apparently occurs at a similar position on the nanofiber as where we also observe a change in luminescence line width. The offset between the curves can be attributed to the uncertainty in defining the zero point for the optical measurements. 3. Conclusions and Outlook. We have investigated systematically the change in luminescence line width of an individual nanofiber as a function of spatial position down to the end of the fiber. Even for positions far away from the end we find relatively small line widths of the order of 16 nm (fwhm), which narrow down to less than 10 nm at the very end of the fiber. A comparison of change in morphology (viz., width of the nanofiber) and change in line width suggests that the inhomogeneously broadened line width is dominated by residual disorder due to the different orientations of the crystallites from which the nanofiber most probably is made.23 This means that neither a better spectral resolution nor a decrease of surface temperature would additionally decrease the observed line width. Of course, stimulated emission or lasing25 would decrease the line width. For methylated poly-para-phenylene films on quartz a decrease of the line width to 4.4 nm has been reported once the threshold for amplified stimulated emission of 14 mJ/cm2 has been overcome.26 In the present case, however, due to the weak excitation rate and since the width of the needle is below the threshold width for waveguiding (222 nm27), we can rather neglect line narrowing effects due to amplified spontaneous emission. 1382
(1) Gebauer, W.; Sokolowski, M.; Umbach, E. Chem. Phys. 1998, 227, 33. (2) Muccini, M.; Lunedi, E.; Taliani, C.; Beljonne, D.; Cornil, J.; Bredas, J. L. J. Chem. Phys. 1998, 109, 10513. (3) Marks, R. N.; Muccini, M.; Lunedi, E.; Michel, R. H.; Murgia, M.; Zamboni, R.; Taliani, C.; Horowitz, G.; Garnier, F.; Hopmeier, M.; Oestreich, M.; Mahrt, R. F. Chem. Phys. 1998, 227, 49. (4) Guha, S.; Graupner, W.; Yang, S.; Chandrasekhar, M.; Chandrasekhar, H. R.; Leising, G. Phys. Status Solidi B 1999, 211, 177. (5) Mock, J. J.; Oldenburg, S. J.; Smith, D. R.; Schultz, D. A.; Schultz, S. Nano Lett. 2002, 2, 465. (6) Soennichsen, C.; Franzl, T.; Wilk, T.; von Plessen, G.; Feldmann, J.; Wilson, O.; Mulvaney, P. Phys. ReV. Lett. 2002, 88, 077402. (7) Piaggi, A.; Lanzani, G.; Bongiovanni, G.; Loi, M. A.; Mura, A.; Graupner, W.; Meghdadi, F.; Leising, G. Optical Materials 1998, 9, 489. (8) Guha, S.; Rice, J. D.; Yau, Y. T.; Martin, C. M.; Chandrasekhar, M.; Chandrasekhar, H. R.; Guentner, R.; Scandiucci de Freitas, P.; Scherf, U. Phys. ReV. B, submitted. (9) Wolf, H. C. Z. Naturforsch. 1958, 139, 413. (10) Cuff, L.; Kertesz, M. Macromolecules 1994, 27, 762. (11) Louarn, G.; Athouel, L.; Froyer, G.; Buisson, J. P.; Lefrant, S. Synth. Met. 1993, 55-57, 4762. (12) Zojer, E.; Koch, N.; Puschnig, P.; Meghdadi, F.; Niko, A.; Resel, R.; Ambrosch-Draxl, C.; Knupfer, M.; Fink, J.; Bredas, J. L.; Leising, G. Phys. ReV. B 2000, 61, 16538. (13) Lane, P. A.; Liess, M.; Wei, X.; Partee, J.; Shinar, J.; Frank, A. J.; Vardeny, Z. V. Chem. Phys. 1998, 227, 57. (14) Andreev, A. Y.; Matt, G.; Sitter, H.; Brabec, C. J.; Badt, D.; Neugebauer, H.; Sariuciftci, N. S. Synth. Met. 2001, 116, 235. (15) Stampfl, J. et al. Synth. Met. 1995, 71, 2125. (16) Yanagi, H.; Okamoto, S. Appl. Phys. Lett. 1997, 71, 2563. (17) Mikami, T.; Yanagi, H. Appl. Phys. Lett. 1998, 73, 563. (18) Andreev, A. Y.; Matt, G.; Brabec, C. J.; Sitter, H.; Badt, D.; Seyringer, H.; Sariciftci, N. S. AdV. Mater. 2000, 12, 629. (19) Balzer, F.; Rubahn, H.-G.; Appl. Phys. Lett. 2001, 79, 3860. (20) Balzer, F.; Rubahn, H.-G. Surf. Sci. 2002, 507-510, 588. (21) Balzer, F.; Rubahn, H.-G. Nano Lett. 2002, 2, 747. (22) Mu¨ller, K.; Chang, C. C. Surf. Sci. 1969, 14, 39. (23) Plank, H.; Sariciftci, N. S.; Andreev, A.; Sitter, H.; Hlawacek, G.; Teichert, C.; Thierry, A.; Lotz, B.; Resel, R. Nanotechnology, submitted. (24) Meghdadi, F.; Tasch, S.; Winkler, B.; Fischer, W.; Stelzer, F.; Leising, G. Synth. Met. 1997, 85, 1441. (25) Bauer, C.; Schnabel, B.; Kley, E.-B.; Scherf, U.; Giessen, H.; Mahrt, R. AdV. Mater. 2002, 14, 673. (26) Schweitzer, B.; Wegmann, G.; Giessen, H.; Hertel, D.; Ba¨ssler, H.; Mahrt, R. F.; Scherf, U.; Mu¨llen, K. Appl. Phys. Lett. 1998, 72, 2933. (27) Balzer, F.; Bordo, V. G.; Simonsen, A. C.; Rubahn, H.-G. Phys. ReV. B, to be published.
NL025807Z Nano Lett., Vol. 2, No. 12, 2002