New Type of Superlattice: An Epitaxial Semiconductor-Atomic Superlattice, SAS
- PDF / 1,771,474 Bytes
- 11 Pages / 415.8 x 637.2 pts Page_size
- 91 Downloads / 223 Views
351 Mat. Res. Soc. Symp. Proc. Vol. 592 © 2000 Materials Research Society
silicon, we presented the experimental results on a new type of silicon multilayers consisting of alternating layers of silicon and monolayers of adsorbed gases, Si/LAG multilayers, constructed by repeated interruptions of silicon deposition with adsorbed gases of oxygen and hydrogen [11,12]. It was found [13] that the particle size of deposited Si may be controlled by depositing at a relatively low substrate temperature, Ts, followed by annealing the as-deposited amorphous film. Hydrogen was used for further passivation of dangling bond defect, and oxygen was used for minimizing the grain boundaries. Figure 1 shows the photoluminescence spectra of two samples taken from Ref. 12. 2
9 PERIODS Si! IAG (d=10nml -45Z9"nn
_No
I
1.7eV
-
102 + Hz 1
V7
,,z
LU
/
z U
.... /
2.34 eV
RAMAN
-
" 23
/ol.""m
V1
PHOTON ENERGY YeVI Fig. 1 PL of two samples: nine periods of Si/IAG (0 2+H2 ), d 10 nm; and five periods of Si/lAG (02 only). Hydrogen passivation of dangling bonds appears to be important as far as strong photoluminescence is concerned. The silicon Raman line is included, to provide an estimate of the strength of the observed PL. Photoluminescence in silicon dioxide [14], and sub-oxides [15] has been observed. Both cases require an excitation with much larger photon energy. Our excitation is only 2.7 eV, far below what is needed to photo-excite the oxides. Recently photoluminescence at 1.9 eV has been observed in the so-called non-bridging oxygen hole center [16], however again, the excitation energy of the XeCl laser is 4 eV far above the 2.7 eV Ar laser line used in our experiments. We detected no significant luminescence before annealing. Therefore, we can discount the possibility of photoluminescence from a-Si:H formed by plasma decomposition of Sill4 [17], as a main source of our observed visible light in the Si/IAG structures. It was concluded [12], that the Si/IAG provides an alternative solution to porous silicon for the significant increase in the luminescent efficiency in the visible spectrum, the improvement in the structure over porous silicon is obvious. The Si/IAG superlattice represents the precursor to what our main endeavor discussed in this paper: the discovery that silicon may be grown epitaxially beyond a monolayer of adsorbed oxygen, the formation of an entirely epitaxial system, the SAS.
352
Without a suitable heterojunction barrier, silicon has not played a significant role in quantum devices. Silicon dioxide with a barrier height of 3.2 eV in the conduction band of silicon is amorphous, preventing the building of a quantum well structure on top of the a-Si0 2 barrier. Several years ago, it was proposed that the oxides of one or two monolayers might allow the continuation of epitaxy. [10] After several failed attempts to realize the SLB (Superlattice Barrier) with thin silicon layers separated by thin oxides, a new method involving the exposure of oxygen followed by epitaxial growth of silicon using the
Data Loading...
