We present a new technique for porous semiconductor formation which is based on the exposure of semiconductor surfaces to gas phase etchants. The technique offers the possibility of fabricating light-emitting devices by selectively exposing a silicon surface to HF vapor. Photoluminescence measurements reveal an efficient emission at around 750 nm. FTIR analysis confirm the existence of strong hydrogen incorporation and oxidation as evidenced from the local bonding environment of hydrogen and oxygen atoms.
©2000 Optical Society of America
Having superior electronic properties, silicon has dominated the semiconductor industry for a long time. However, due to its poor radiative recombination efficiency as an indirect bandgap material, efficient light-emitting devices have not been achieved in Si technology. Nevertheless, recent demonstrations[1,2] and observations of room temperature photoluminescence from porous Si, have evoked and stimulated interest on the growth and the investigation of the physical mechanisms responsible for the luminescence. This is mainly due to its potential for the realization of Si-based optoelectronic integrated circuits, logic and display systems. Therefore here has been an extensive research on the formation of porous Si since the observation of the visible light emission at room temperature. Anodic oxidation is a widely used technique for the formation of porous semiconductors[3–5]. Both the thickness and the micro-crystal sizes can be adjusted with this technique, thus enabling us to control the optical and electrical behavior of porous Si layers. However, isolating the metal contacts from the solution, has been the major drawback of this technique. Stain-etching technique (electroless) can be also used to grow porous layers in Si[7–9]. Major difference between the two techniques is the growth rate and the thickness limitation in the latter one. From the experiments carried out under different conditions, we found that it was not possible to grow more than 0.3 µm porous layer with stain-etching.
In this letter, we present a novel technique for porous semiconductor growth. It is an electroless etching technique but, thicker layers can be grown compared to stain-etching. It can produce new materials with interesting structural as well as optical properties. It can be considered as a new alternative for growing porous semiconductors from the point of view of addressing current research issues related to light emission and introducing versitility and effectiveness into the growth process. The technique enables us to grow porous semiconductors from the gas phase etchants instead of chemical solution. These results suggest that it can be effectively used to process selectively on semiconductor surfaces. This technique can be likely used to prepare doped porous semiconductors, as well. The samples are characterized by room temperature PL and FTIR measurements.
2. Sample preparation
The technique is based on the exposure of Si surface to gaseous etchants prepared from the (HF:HNO3:H2O 2:1:5) mixtures. For this application, a teflon cell is used in order to expose the Si surface to solution vapors as shown in Fig.1. The cell was kept at around 30° C and gaseous etchants over atmospheric pressure were released through the openings around the wafer holder. So, after being exposed to Si surface, the vapors are constantly exhausted from the cell. The layers formed under these conditions resulted in growth rates of 100nm per hour. So, around 10 hours of growth duration is required to obtain a 1 micron thick layer. The porous layers were formed on (100) oriented p-type Si wafers having resistivities between 5–10 Ohm-cm.
3. Physical properties
Figure 2 shows a typical photoluminescence (PL) emission from the porous Si layer so grown from the precursors of gaseous etchants. The PL was excited by an Ar+- ion laser at 514 nm at room temperature. The emission is located at 750 nm with a spectral bandwidth of about 320 meV. Note that both the PL peak position and its width are different from those prepared in the chemical etching solution. The samples obtained with the stain-etching technique from the same solution exhibit a PL peak at around 630 nm with FWHM of around 400 meV for a series of samples grown under different conditions. The 750 nm emission has also been reported from thermally oxidized porous Si samples with oxyhydrides formed at the surface of nanocrystallites. Another important feature of the PL emission is the presence of a doublet structure separated by 120 meV. These features can be attributed to the existence of structural unhomogeneities over the surface. These findings are supportive by SEM measurements which reveal two types of surface textures. Same measurements indicated a layer thickness to be between 1 to 2 micrometer, thus suggesting a growth rate of 100 nm per hour. Also, the photoluminescence emission was found to be very stable, even after a prolonged time of storage in atmosphere. It has been more than two years, but the first samples are still luminescing.
Infra-red vibrational spectrum provides information about the local bonding environment of hydrogen atoms in the porous layer (Fig.3). We observe a strong dominant single peak of Si-H stretching modes at 2120 cm-1. Note that the samples grown from the stain-etching solution exhibit a peak at 2200 cm-1. Also, observed are the other Si-Hx (x=1-3) related vibrations at 630 and 660 cm-1 in the wagging mode region. Si-O stretching mode is located at 1080 cm-1 as a strong symmetrical band. The relative intensities of both the H and O related modes suggest the presence of heavy hydrogenation and oxydation in our samples. The Si-Hx bending modes are found to be at 860, 910, 930 and 960 cm-1. Another important feature of the IR spectrum is the presence of a new peak at 1240 cm-1. This peak has never been observed in our stain-etched layers.
In our experiments, we observe that both the emission and the vibrational properties of the layers grown from the gas phase are entirely different from those obtained by chemical solution. The PL peak energy is red-shifted, indicating probably the existence of larger crystallite sizes or different chemical groupings if one considers a chemical confinement. Also, the samples show a much more efficient luminescence with respect to the ones prepared by stain etching technique, as can be seen by the naked eye. This is suggestive of the possibility of growing thicker layers using this technique. From these experiments, it is difficult to distinguish between the quantum and chemical confinement effects. However, there is a correlation between the PL emission and IR vibrations. For samples prepared by chemical solution(stain-etching), the PL peak is located at around 630 nm and the Si-H stretching modes with a backbonded O atom are dominated by a single band at 2200 cm-1. Whereas, gas phase grown layers exhibit a PL peak at 750 nm and the Si-H stretching modes are located at 2120 cm-1. Moreover, we observe a new band at 1240 cm-1. These observations are probably indicative of some microstructural changes in the local bonding environments of both H and O atoms in the new layers. The band at 1240 cm-1 can be assigned to unintentional impurities or Si=O stretching bonds in oxyhydrides bound to the surface of the Si nanocrystals. Recent calculations suggested the possibility of having a silanone based oxyhydride fluorophor (-Si(O)-OSiH3) surface coating of the Si crystallites. Nevertheless, the annealing experiments have shown that this band is not very stable. However, additional surface analysis would be necessary to confirm these hypothesis.
In conclusion, we report the possibility of fabricating light emitting porous silicon by exposing a silicon surface to gaseous etchants. The layers exhibit an efficient and a red shifted emission compared to those prepared by stain etching technique. Also, they exhibit different structural features as evidenced from the local bonding environment of H and O atoms. The photoluminescence was found to be stable over a long period of time(two years) without any degradation. The work suggests that this technique can be used to process selectively on semiconductor surfaces, thus making its integration into microelectronics possible..
This work was supported by NATO-Scientific Affairs Division under the contract No: 950830
References and Links
1. L. T. Canham, “Silicon quantum wire array fabrication by electrochemical and chemical dissolution of wafers,” Appl. Phys. Lett. 57, 1046 (1990) [CrossRef]
2. V. Lehmann and U. Gosele, “Porous silicon formation: A quantum wire effect,” Appl. Phys. Lett. 58, 856 (1991) [CrossRef]
3. L.T. canham, W.Y. Leong, T.I. Cox, and L. Taylor, “Efficient visible electroluminescence from highly porous silicon under cathodic bias,” Appl. Phys. Letts. 21, 2563(1992) [CrossRef]
4. H.D. Fuchs, M. Stutzmann, M.S. Brandt, M. Rosenbauer, J. Weber, A. Breitschwerdt, P. Deak, and M. Cardona, “Porous silicon and siloxene: vibrational and structural properties,” Phys. Rev. B48, 8172(1993)
5. A.G. Cullis, L.T. Canham, and P.D.J. Calcott, J. Appl. Phys.82, 909(1997) [CrossRef]
6. Z. Gaburro, H. You, and D. Babie, “Effect of Resistivity and current density on photoluminescence in porous silicon produced at low HF concentration,” J. Appl. Phys. 84, 6345 (1998) [CrossRef]
7. J. Sarathy, S. Shih, K. H. Jung, C. Tsai, K. -H. Li, D. L. Kwong, and J. C. Campbell, “Demonstration of photoluminescence in nonanodized silicon,” Appl. Phys. Lett. 60, 1533 (1992) [CrossRef]
8. A. Ksendzov, R. W. Fathauer, T. George, W. T. Pike, and R. P. Vasquez, “Visible photoluminescence of porous Si1-xGex obtained by stain etching,” Appl. Phys. Lett. 63, 200 (1993) [CrossRef]
9. S. Kalem and Rosenbauer, “Optical and structural investigation of stain-etched silicon,” Appl. Phys. Lett. 67, 2551 (1995) [CrossRef]
10. Y. Kanemitsu, T. Futagi, T. Matsumoto, and H. Mimura, “Origin of the blue and red photoluminescence from oxidized porous silicon,” Phys. Rev. B49, 14732 (1994)
11. J. L. Gole and D. A. Dixon, “Evidence for oxide formation from the single and multiphoton excitation of a porous silicon surface or silicon nanoparticles,” J. Appl. Phys. 83, 5985 (1998). [CrossRef]