3906
J. Phys. Chem. B 1998, 102, 3906-3911
Trap State Dynamics in MoS2 Nanoclusters R. Doolen, R. Laitinen, F. Parsapour, and D. F. Kelley* Department of Chemistry, Colorado State UniVersity, Fort Collins, Colorado 80523-1872 ReceiVed: December 16, 1997; In Final Form: February 19, 1998
The trap-to-trap relaxation and recombination dynamics of photogenerated electron/hole pairs in MoS2 nanoclusters have been studied. Static and time-resolved emission experiments have been performed on 3.0 and 4.5 nm diameter nanoclusters in ternary inverse micelles, acetonitrile, and octane at room temperature and at 20 K. The results indicate that, following synthesis in ternary inverse micelles, the nanoclusters have both shallow and deep traps. The deep traps are retained upon extraction into acetonitrile and passivated upon charge neutralization and reextraction into octane. The emission kinetics show that trap-to-trap relaxation is fast ( 2.0. We suggest that sulfide ions dangling on the nanocluster edges may be protonated or deprotonated, controlling the overall charge. These extractions have no detectable effect on the absorption spectrum but dramatically alter the surface chemistry and trap dynamics, as discussed below. Nanoclusters having an absorption maximum at 470 nm may also be synthesized from a MoCl4 (10-3 M) solution in ternary micelle. In this case, approximately 4 mol equiv of 22% aqueous (NH4)2S is injected into the MoCl4 solution. Previous characterization studies have shown that these nanoclusters have diameters of about 4.5 nm. These nanoclusters remain in the octane phase upon extraction with acetonitrile, indicating that they are not initially charged. This is probably due to protonation, caused by the presence of the water. Time-resolved emission results were obtained by timecorrelated single photon counting, using an apparatus that has been previously described.17 In this apparatus, an approximately 1 mm2 spot of the sample is irradiated with 7 ps, 318 nm
ultraviolet laser pulses at 4 MHz, having an average power of ca. 1 µW. This is a very low fluence, and saturation effects are completely avoided. The apparatus has a temporal response of about 80 ps. Wavelength selection was obtained by using a 1/ m monochromator with a 150 groove/mm grating. Static 4 emission spectra were obtained using a home-built instrument that has been described earlier.18 All spectra reported here are uncorrected for instrument response. However, the instrument response is quite flat in the 400-600 nm region. Lowtemperature results were obtained using a closed cycle helium displex (Air Products, 202E). The low-temperature cell consisted of two quartz windows separated by a Teflon-coated O-ring and secured to a brass plate at the coldfinger tip. Thermal contact was ensured using indium gaskets, and the temperature was monitored with a thermocouple mounted at the tip of the coldfinger. Results and Discussion Trap State Relaxation Dynamics. Room-temperature static emission spectra of small (approximately 3.0 nm diameter) nanoclusters in acetonitrile exhibit an intensity maximum at about 420 nm, as shown in Figure 2. Similar room-temperature spectra are obtained from the small nanoclusters in ternary micelles and following reextraction into octane. The emission kinetics in the ternary micelle or in acetonitrile show two distinct components: a fast, instrument response limited component and a slower (several nanoseconds) component (see Figure 3). The relative intensities of the fast and slow components are somewhat wavelength dependent. The fraction of fast component increases at further red observation wavelengths. These observations indicate that the fast component is due to deeper traps which undergo rapid nonradiative recombination. Extraction from wet acetonitrile into octane dramatically alters the kinetics. Specifically, charge neutralization and extraction into octane greatly reduces or eliminates the fast component, indicating elimination of the deep traps. The room-temperature static spectra are essentially identical in all three (ternary micelle, acetonitrile, and octane) environments. This is because in all cases the slower (shallow trap) component has a much larger integrated intensity and therefore dominates the static spectrum. Although the lifetimes of both the fast and slow components increase at low temperatures, the slow component continues to dominate at 20 K, which is why the static spectrum in acetonitrile changes very little with temperature.
3908 J. Phys. Chem. B, Vol. 102, No. 20, 1998
Figure 3. Emission kinetics of small nanoclusters in room-temperature acetonitrile. The observation wavelengths were 420 nm (dotted curve) and 520 nm (solid curve).
Figure 4. Emission kinetics of small nanoclusters in ternary micelles at 20 K. The observation wavelengths were 420 nm (dotted curve) and 520 nm (solid curve).
No emission rise times are observed in any of the roomtemperature kinetics at any wavelengths. Thus, in all roomtemperature samples the lowest energy available traps appear to be populated rapidly (
