In this research, we used photoluminescence (PL) and photoluminescence excitation (PLE) to visualize the electronic band structure in porous silicon (PS). From the combined results of the PLE measurements at various PL emission energies and the PL measurements under excitation at various PLE absorption energies, we infer that three di\ufb00erent electronic band structures, originating from di\ufb00erent luminescent origins, give rise to the PL spectrum. Through either thermal activation or di\ufb00usive transfer, excited carriers are moved to each of the electronic band structures.
Since the discovery of visible photoluminescence (PL) at room temperature, researchers have studied porous silicon (PS) with respect to its potential in new elec- tronic devices . Speci\ufb01cally, reports have focused on manufacturing methods for devices such as vapor sen- sors, light-emitting devices, waveguides, ballistic electron surface-emitting displays, microcavities,etc. [2\u201310].
The mechanism underlying PS PL emission remains poorly understood. The quantum con\ufb01nement model as- sumes that PL occurs as a result of the recombination of localized excitons in silicon nanocrystals [11\u201313]. The chemical model assumes that PL depends on some chem- ical entity such as an Si-H bond, a polysilane chain, or a siloxene group [14,15]. Furthermore, a quantum con\ufb01ne- ment/luminescence centers model has recently been pro- posed. This model comprises a combination of the quan- tum con\ufb01nement model and the chemical model [16,17]. According to this model, the PL originates from lumi- nescence centers surrounding silicon nanocrystals with a radiative recombination of the carrier being photocreated inside the silicon nanocrystals. In yet another model, PL is posited to occur through an interface region between the silicon nanocrystals and the silicon compound layer; it occurs as a result of the recombination of the carrier being transferred from the silicon nanocrystals [18, 19]. Because several di\ufb00erent models have been proposed to explain the mechanism underlying PL, PS may, at this point, may be considered to have a few di\ufb00erent PL ori-
To clearly explain the mechanism underlying PL in PS, we are \ufb01rst required to understand its electronic band structure. Measuring PLE spectra is a useful method for examining the correlation between the light absorp- tion band and the PL emission band, or, more precisely, the relationship between and the transfer process of the excited carriers in the electronic band structure.
In this paper, we investigated PLE measurements at various PL emission energies. We also measured PL un- der excitation at various PLE absorption energies. From these results, we analyzed the electronic band structure of PS.
We produced the PS used by anodization in a 10 % hy- dro\ufb02uoric acid-ethanol solution of a p-type, boron doped, (100)-oriented silicon wafer with a resistivity in a range between 0.06\u223c 0.12 \u2126\u00b7cm. The contact area between the HF-ethanol solution and the wafer was 0.49 cm2. We conducted the anodization at room temperature in the dark by using a potentiostat/galvanostat (Hokuto Denko, HA 301), and we used an arbitrary function gen- erator (Hokuto Denko, HB 105) to induce variations in the anodization current density and time. For the PL and the PLE measurements, we used a spectro\ufb02uorome- ter (SLM-AMINCO, 8100) with an attached 300 W Xe- lamp for an excitation source.
Figure 1 shows a PL spectrum of PS, measured at room temperature. We used a 40 mW He-Cd laser to measure the PL shown in Fig. 1(a) under an excitation energy of 2.81 eV, and we observed a maximum peak in the range of 1.55 to 1.65 eV. In general, the visible light emission from PS manifests as a broadband, as shown in Fig. 1(a). We used a spectro\ufb02uorometer to measure the PL shown in Fig. 1(b) under an excitation energy of 2.70 eV.
In contrast to Fig. 1(a), Fig. 1(b) shows four nar- rowband emissions: 2.00 eV, 1.89 eV, 1.79 eV, and 1.55 eV. This phenomenon clearly demonstrates that these narrow band emissions are PL components of the broad- band emission. When we measured the PL spectrum with a spectro\ufb02uorometer, we were able to observe each component due to the reduced spectral bandwidth. The spectro\ufb02uorometer made this possible because, unlike a high-powered laser, its weak excitation power signif- icantly limits the formation of carriers. We took PLE measurements for the four PL peaks; from these mea- surements, we deduced the electronic band structure of the PS.
Figure 2(a) shows the PLE spectra monitored at the 1.89 eV PL peak. As the \ufb01gure indicates, we observed three PLE peaks: 3.57 eV, 3.10 eV, and 2.71 eV. The PLE peaks appeared in Fig. 2(a) because the excited carriers from the absorption band of these three PLE peaks were recombined at the 1.89 eV emission band. In order to ensure that excited carriers fromthe absorption
Fig. 1. (a) Room temperature PL spectrum, measured at an excitation energies of 2.81 eV, exerted by 40 mW He-Cd laser. (b) Room temperature PL spectrum measured under an excitation energy of 2.70 eV using a spectro\ufb02uorometer.
band of the PLE peaks were recombined at the 1.89 eV emission band, we measured the PL characteristics un- der the excitation energy of the 3.10 eV PLE peak, as shown in Fig. 2(b). In contrast to Fig. 1(b), only two peaks appeared, 1.92 eV and 2.00 eV, in the PL spec- trum. We assume that the 1.92 eV and the 1.89 eV PL peaks shown in Fig. 1 originate from the same emis- sion band, though their energy values di\ufb00er slightly. If the peaks do originate fromthe same emission band, the 1.92 (1.89) eV and 2.00 eV PL peaks must be formed by a radiative recombination of excited carriers from an absorption band at 3.10 eV.
We also measured the PL characteristics under an ex- citation energy of 3.57 eV, and observed the PL peaks at 1.92 and 2.00 eV. However, in contrast to Fig. 2(b), the 2.00 eV PL peak was more intense than the 1.92 eV PL peak. These results indicate that a large portion of the excited carriers fromthe 3.10 eV and the 3.57 eV ab- sorption bands recombine in the emission bands at 1.92 eV and 2.00 eV, respectively; furthermore, some of the excited carriers from the 3.10 eV and the 3.57 eV ab- sorption bands appear to recombine at the 2.00 eV and the 1.92 eV emission bands, respectively. Consequently, the electronic band structure (I) of luminescent origin, which gives rise to the 1.89 eV and the 2.00 eV PL peak seen in Fig. 1(b), is comprised of the 3.10 eV and the 3.57 eV absorption bands, and the 1.92 eV and the 2.00 eV emission bands. We will address the 2.71 eV PLE peak seen in Fig. 2(a) in the discussion of Fig. 3.
Figure 3(a) shows the PLE spectra monitored at the 1.79 eV PL peak of Fig. 1(b). We observed two peaks, 2.71 and 3.10 eV. In Fig. 3(a), the intensity of the 2.71 eV peak was stronger than that of the corresponding peak in Fig. 2(a), implying that many of the excited carriers from the 2.71 eV absorption band recombined at the 1.79 eV emission band. This phenomenon indi-
cates that another electronic band structure exists in PS, which di\ufb00ers from the electronic band structure I. This electronic band structure (II) is comprised of a 2.71 eV absorption band and a 1.79 eV emission band.
Another important feature of Fig. 3(a) is the 3.10 eV PLE peak, which is due to an absorption band of electronic band structure I. The carrier transfer between the electronic band structures explains this phenomenon. That is, the appearance of the 3.10 eV PLE peak in Fig. 3(a) indicates that excited carriers fromthe 3.10 eV absorption band in electronic band structure I recombine at the 1.79 eVemission band in electronic band structure II. We can explain the 2.71 eV PLE peak seen in Fig. 2(a) in the same way. That is to say, the appearance of the 2.71 eV PLE peak in Fig. 2(a) indicates that excited carriers from the 2.71 eV absorption band in electronic band structure II recombine at the 1.89 eVemission band in electronic band structure I.
Figure 3(b) shows the PLspectrum measured under an excitation energy of 2.71 eV. The spectrum has three PL peaks: 1.90 eV, 1.79 eV, and 1.55 eV. Of these peaks, as mentioned above, the 1.79 eV PL peak is the most intense, as a large proportion of the excited carriers re- combine at the 1.79 eV emission band. In addition, the 1.90 eV PL peak is due to the radiative recombination of excited carriers from the 2.71 eV absorption band in electronic band structure II at the 1.90 eVemission band in electronic band structure I. We will address the 1.55 eV PL peak seen in Fig. 3(b) in the discussion of Fig. 4.
Figure 4(a) shows the PLE spectrum monitored at the 1.55 eV PL peak of Fig. 1(b). The appearance of only one strong peak at 2.34 eV indicates that all the excited carriers fromthe 2.34 eV absorption band recombine at the 1.55 eV emission band. This result indicates the ex- istence in the PS of yet another electronic band structure (III), which di\ufb00ers from electronic band structures I and
II. Fig. 4(b) shows the PL spectrum measured under an excitation energy of 2.34 eV. The spectrum shows only one strong peak, occurring at 1.55 eV. This result clearly proves that electronic band structure III is comprised of a 2.34 eV absorption band and a 1.55 eV emission band. We can now explain the appearance of the 1.55 eV PL peak seen in Fig. 3(b). This peak is clearly the result of the recombination of excited carriers fromthe 2.71 eV absorption band in electronic band structure II at the 1.55 eV emission band in electronic band structure III.
In summary, Fig. 5 shows the electronic band struc- ture acquired from the PL and the PLE measurements. As this \ufb01gure shows, electronic band structure I is com- prised of 3.57 eV and 3.10 eV absorption bands and 2.00 eV and 1.92 eV emission bands. Electronic band struc- ture II is comprised of a 2.71 eV absorption band and a 1.79 eV emission band. Electronic band structure III is comprised of a 2.34 eV absorption band and a 1.55 eV emission band.
This action might not be possible to undo. Are you sure you want to continue?