Chemical Physics Letters 435 (2007) 109–113 www.elsevier.com/locate/cplett
Nonlinear optical response of Mo6S4.5I4.5 nanowires James J. Doyle a, Valeria Nicolosi a, Sea´n M. O’Flaherty b, Damjan Vengust c, Anna Drury a, Dragan Mihailovic c,d, Jonathan N. Coleman a,*, Werner J. Blau a
a
Materials Ireland Polymer Research Centre, School of Physics, Trinity College Dublin, Dublin 2, Ireland b Optical Metrology Innovations Ltd., 2200 Cork Airport Business Park, Ireland c Jozef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia d Mo6, Teslova 30, 1000 Ljubljana, Slovenia Nano-Materials, Slovenia Received 31 August 2006 Available online 20 December 2006
Abstract Experimental measurements of Mo6S4.5I4.5 nanowires investigating nonlinear optical extinction (NLE) of nanosecond laser pulses are reported. The experiments were performed using the open aperture Z-scan technique at 532 nm and 1064 nm. Concentration dependent studies were performed for Mo6S4.5I4.5 nanowires dispersed in isopropanol at 0.1, 0.05, 0.025 and 0.0125 g/l, showing direct correlation between the NLE response and nanowire bundle diameter. Further to this, intensity dependent scattering experiments are reported for Mo6S4.5I4.5 nanowires dispersions at wavelengths of 532 nm and 1064 nm. Mechanistic implications of the optical extinction are discussed. The NLE was compared with that for a stabilized multi-walled carbon nanotube (MWNT)-polymer composite dispersion. 2006 Elsevier B.V. All rights reserved.
1. Introduction Numerous organic materials, including porphyrins [1], C60 [2–6], MWNT’s [7–10], single-walled nanotubes (SWNTs) [11–13] and phthalocyanines [3,14–20] have been studied to probe their nonlinear extinction (NLE) of high intensity light. However, the facile manipulation and ability to chemically tailor the material properties to suit industrial requirements has led to a huge and growing interest in inorganic nanotubes, nanoparticles and nanowires. The straight-forward synthesis [21], easy dispersability in common organic solvents [22,23] and uniformity, in terms of both diameter and electronic behaviour, makes Mo6S9 xIx nanowires one of the most interesting monodimensional materials available at present. In this Letter, we report the first NLE results for Mo6S4.5I4.5 nanowires, probed at 532 and 1064 nm using the Z-scan technique [24]. We also investigate the scattering contribution to the optical extinction from Mo6S4.5I4.5 nanowires dissolved in *
Corresponding author. E-mail address:
[email protected] (J.N. Coleman).
0009-2614/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2006.12.053
isopropanol (IPA) compared to that of multi-walled carbon nanotubes (MWNT) dispersed in toluene and an organic polymer (PmPV). Through simultaneous nonlinear optical extinction and intensity dependent scattering experiments we report a direct correlation between the observed optical extinction and the scattered intensity. 2. Experiments Mo6S4.5I4.5 nanowires were fabricated by direct synthesis from elemental material that had been mixed in the desired stoichiometries. This resulted in a powder composed of bundles of individual nanowires, each nanowire having a diameter of 0.94 nm [21]. This powder was then washed in acetone several times to eliminate any excess iodine remaining from the synthetic step. Once the iodine had been removed, the Mo6S4.5I4.5 powder was purified as described previously [22,23]. Dispersions were then prepared by mixing the purified nanowire material in isopropanol at different concentrations (in the range 0.1– 0.0125 g/l). These mixtures were sonicated for 2 min using a high power sonic tip (120 W, 60 kHz) followed by gentle
110
J.J. Doyle et al. / Chemical Physics Letters 435 (2007) 109–113
agitation using a low power ultrasonic bath for 2 h afterwards to ensure a uniform dispersion [25]. For comparison purposes, a dispersion of PmPV–MWNT composite was prepared as previously reported [26]. The open-aperture Z-scan technique was used to measure the total transmittance through the samples. All NLE experiments described in this study were performed with 6-ns pulses from a Q-switched Nd:YAG laser. The beam was spatially filtered to remove higher-order modes and tightly focused to a 20–25 lm spot for all experiments. The laser was operated at both first and second harmonics (1064 nm and 532 nm, respectively), with a pulse repetition rate of 10 Hz. All Mo6S4.5I4.5 nanowires dispersion samples were tested in 1 cm quartz cells. For all intensity dependent scattering experiments, a focusing lens set-up was arranged at 45 to the direct incident beam. Transmission electron microscopy (TEM) measurements were made with a Hitachi H-7000 and ‘holey’ carbon grids. In preparation for atomic force microscopy (AFM) studies, using a Nanoscope III, a small volume from each dispersion was deposited on highly ordered pyrolytic graphite (HOPG) under ambient conditions. 3. Results and discussion Information on nanowire bundle size was determined by both AFM and TEM analysis on all concentrations. Representative TEM images for Mo6S4.5I4.5 nanowires at concentrations of 0.1, 0.05, 0.025 and 0.0125 g/l are shown in
Fig. 1a–d, respectively. In addition, for each of the four nanowire concentrations prepared, AFM measurements were carried out. From the AFM images, the diameters of a large number of nanowire bundles were calculated. The average bundle diameter tended to decrease with concentration as shown in Fig. 2. This debundling effect is in agreement with previous observations reported for both single wall carbon nanotubes [27,28] and Mo6S4.5I4.5 nanowires [29]. Bundle lengths do not appear to change with the various concentrations investigated (within error) and an average length of (3.8 ± 1.7) lm was estimated from the TEM data. Linear absorption coefficients, a0, as defined by the Lambert Beer law, ln(I/I0) = a0Cl were measured at 532 nm and 1064 nm for each sample. In this equation I/I0 defines the ratio of transmitted to incident laser light, C is the samples mass concentration and l is the sample thickness. The a0 value measured for Mo6S4.5I4.5 nanowires in IPA (532 nm) at 0.1 g/l was 1.1 l g 1 cm 1, 1.8 l g 1 cm 1 for 0.05 g/l and 1.6 l g 1 cm 1 for 0.025 g/l. The linear absorption coefficient for the 0.0125 g/l nanowire dispersion at 532 nm was undetectable using our setup as the linear transmission at such a low concentration was close to 100%. At 1064 nm the values calculated for a0 were 1.3 l g 1 cm 1, for the 0.1 g/l nanowire dispersion and 2.4 l g 1 cm 1 for the 0.05 g/l sample. The two lower concentrations (0.025 g/l and 0.0125 g/l) did not exhibit a NLE response at 1064 nm and consequentially the optical data from these samples have been omitted in the interest of brevity.
Fig. 1. TEM images of a series of Mo6S4.5I4.5 nanowire in IPA. Concentrations vary from 0.1 g/l (a), 0.05 g/l (b), 0.025 g/l (c), 0.0125 g/l (d). The scale bar for each image represents 1 lm.
J.J. Doyle et al. / Chemical Physics Letters 435 (2007) 109–113
Table 1 Nonlinear optical coefficients for Mo6S4.5I4.5 nanowires at various concentrations, at 532 nm and 1064 nm, including focal intensity (I0), linear absorption coefficient (a0) and effective nonlinear absorption coefficient (beff)
35 Average Diameter
25
1.6
20
1.2
-9
βeff [ x 10 cm W ]
15 0.8 10
-1
Average Diameter [nm]
2.0
β eff
30
0.4
5 0 0.00
Wavelength (k)
Concentration (g/l)
I0 (GW cm 2)
a0 (cm 1)
beff (cm W 1)
532
0.1 0.05 0.025 0.1 0.05
1.0 1.0 0.9 1.6 1.6
0.11 0.09 0.04 0.13 0.12
(1.6 ± 0.6) · 10 (5.1 ± 0.4) · 10 (3.7 ± 0.3) · 10 (1.1 ± 0.3) · 10 (1.2 ± 0.2) · 10
1064
9 10 10 10 10
0.0 0.02
0.04
0.06
0.08
0.10
Concentration [g/l] Fig. 2. This plot is presented as the mean nanowire diameter (measured from at least 50 nanowire or nanowire bundles per sample), as a function of concentration (filled squares). Also plotted is the effective nonlinear absorption coefficient as a function of nanowire concentration (circles).
In order to probe the resulting NLE response at 532 and 1064 nm, for a variety of Mo6S4.5I4.5 nanowires concentrations, the open-aperture of the Z-scan technique was employed. Fig. 3 plots the normalised transmission as a function of laser energy density for the four Mo6S4.5I4.5 nanowire concentrations in solution, at 532 nm. All four curves displayed a characteristic NLE profile, where above a certain incident energy threshold, the degree of transmitted light detected is greatly reduced, due to the intensity dependent nonlinear response of the material. To quantify the magnitude of the optical dissipation we estimate the NLE coefficient, beff, from experimental fitting of the Z-scan spectra [24]. Table 1 presents the calculated values for a0 and beff at specific focal intensities for four concentrations at 532 nm and two concentrations at 1064 nm. The value for beff measured for Mo6S4.5I4.5 nanowires was (1.6 ± 0.3) · 10 9 cm W 1 for 0.1 g/l, (5.1 ± 0.4) · 10 10 cm W 1 for 0.05 g/l, (3.7. ± 0.3) · 10 10 cm W 1
1.0
Normalised Transmission
111
0.9
0.8
0.1g/l (532nm) 0.05g/l (532nm) 0.025g/l (532nm) 0.0125g/l (532nm)
0.7
0.6 0.1
1
10 -2
Pulse Energy Density [J cm ] Fig. 3. Plot of normalised transmission against laser pulse energy density for Mo6S4.5I4.5 nanowires at 0.1 g/l, 0.05 g/l, 0.025 g/l and 0.0125 g/l at 532 nm.
for 0.025 g/l and (3.7. ± 0.3) · 10 10 cm W 1 for 0.0125 g/l. At concentrations of 0.1 g/l and 0.05 g/l, studied at 1064 nm, beff values of (1.1 ± 0.3) · 10 10 and (1.2 ± 0.2) · 10 10 cm W 1 were measured for the nonlinear absorption coefficient respectively. At the lower concentrations, no NLE response was observed under 1064 nm irradiation. The beff values for the 532 nm irradiation are plotted versus concentration in Fig. 2. Further analysis revealed a Mo6S4.5I4.5 nanowire diameter dependent NLE of incident irradiation. In Fig. 3, normalised transmission is presented as a function of energy density (J cm 2) for the four nanowire dispersions at different concentrations. It can be noted from this plot that the NLE response of the Mo6S4.5I4.5 nanowires at 532 nm falls into one of two distinct regimes. The weaker NLE responses are from the two lower concentrations of 0.025 g/l and 0.0125 g/l. From AFM imaging studies, this corresponds to an average nanowires bundle diameter of