Control of Silicon Nanocrystallite Luminescence Behavior Through Surface Treatments

  • PDF / 410,162 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 0 Downloads / 210 Views

DOWNLOAD

REPORT


hydrogen passivated and etched using dips in 48 W/o Hydrofluoric Acid (HF). Films were oxidized using 20 W/o Nitric Acid (HNO3) dips or by allowing them to age in normal room air. Methanol dips were performed using HPLC grade methanol. Iodine dips were performed with a solution of -7 x 10-4 M 12 in methanol. After chemical treatment, the samples were characterized using photoluminescence emission spectroscopy (PL). The excitation source was a pulsed 3xNd:YAG laser (QX,= 355 nm) and the emission was collected using a Si photodiode array detector cooled to -20 'C. Spectra were analyzed using a computerized multichannel analyzer, and corrected for background emissions and detector response. The X-ray Photoelectron Spectroscopy (XPS) was performed using a Perkin Elmer Model 5100 XPS system. During XPS measurements, the binding energy was referenced to the adventitious carbon peak at 284.6 eV present on all samples. The base pressure of the system was less than 5 x 10-9 torr. The incident source was a 300 W, Mg Ke (E=1253.6 eV) un-monochromatized x-ray beam. RESULTS AND DISCUSSION A series of experiments were performed in which films were processed using chemical treatments to study the role that surface and size play in the luminescence behavior. One set of samples were etched in HF for 60 seconds to remove the oxide layer that grows on the nanocrystallites over time as they are exposed to air. A ten minute HNO3 dip was then used to grow a new oxide layer on the sample, and then the etch/oxide regrowth cycle was repeated. Each new oxide regrowth step reduces the size of the crystalline Si particle core and further enhance the confinement of excited carriers. Photoluminescence spectra of the films during these cycles are shown in Figure 1. These spectra show that as the etch/oxidation cycles shrink the particle sizes, the peak PL emission shifts to higher energies. This indicates a strong, consistent correlation between Energy (eV) particle size and emission wavelength. 1.6 1.8 2.0 2.4 2.2 A d I Similarly, dipping films in a mixture of HF, Aged HNO3, and H 2 0 will also simultaneously oxidize and etch the Si nanocrystallites. By extending the length .2------------------of this etch time it should be possible to reduce ., Aged+HF particle size and change the luminescence wavelength. As shown in Figure 2, the peak PL < _wavelength again blueshifts as the etch time SA--d+HF+HNOincreases and reduces the particle size. Once again, the changing size of the nanostructure correlates with the changing emission energy, supporting Cwell __- -...---models of luminescence based on quantum Aged+HF+HNO3+HF

confinement.

Figure 3 shows the PL behavior of series of films dipped in HF for different lengths of time. The 500 550 600 650 700 750 800 Wavelength (nm)

Figure 1 PL spectra of samples undergoing HF etch/HNO 3 oxidation cycles. As the cycles reduce the particle size, there is a blueshift of the peak emission,

HF serves to etch away the native oxide layer on the particles and leads to a hydride passivation of the underlying Si