- Email: [email protected]
Radiation induced recrystallisation and enhancement in photoluminescence from porous silicon
__ !iB
-H J&3 __ ELSEVIER
NINMI B
Beam Interactions with Materials 8 Atoms
Nuclear
Instruments
and Methods
in Physics
Research
B 132 (1997) 409417
Radiation induced recrystallisation and enhancement in photoluminescence from porous silicon Tejashree
M. Bhave ‘, S.V. Bhoraskar
‘.*, Prabhat
Singh b, V.N. Bhoraskar
a
d Department of Ph>~sics.University of Pune. Ganeshkhind, Pune 411 007. India ’ National Chemiul Received
19 February
Laboratory. Punr. India
1997; revised form received 21 May 1997
Abstract
Porous Silicon (PS) samples were irradiated with different ionizing radiations which included 1 and 6 MeV electrons, “‘Co gamma rays and 10 MeV silicon ions. Improvement in the efficiency of photoluminescence (PL) and its stability with time were invariably observed in all the irradiated PS samples. Improvement in the luminescent properties was best for samples irradiated with 10 MeV silicon ions. Partial restructuring of Si-0-Si and Si-H type species into Si-OH was confirmed from the infrared spectra of pre- and post- irradiated samples. Grazing angle X-ray diffraction (XRD) analysis revealed that preferential recrystallisation occurs in the irradiated region. The virgin PS sample exhibited only the (I 1 1) peak in the XRD pattern; whereas the irradiated PS sample showed a (3 1 1) peak along with the (1 1 1) peak. The average size of the microcrystallites was calculated from the diffraction peak broadening, using Scherrer’s formula. Depth profile studies, corresponding to the average sizes of the microcrystallites confirmed the existence of (3 1 1) planes, and revealed that the degree of recrystallisation is maximum at the end of the trajectories of silicon ions. 0 1997 Elsevier Science B.V.
1. Introduction Since the discovery of visible photoluminescence (PL) from porous silicon (PS) in 1990 by Canham [l], efforts are being made to improve the PL efficiency [24] as well as to understand the mechanism behind the PL. Several models as regards to the origin of the PL from PS have been suggested. In one of the models [5], the origin of
*Corresponding author. Tel.: 91 0212 352678: fax: 91 0212 350087; e-mail: [email protected]. 0168-583X/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. 1’11SO168-583X(97)00389-3
the PL having energy above the band gap energy of silicon is attributed to the effect of the quantum confinement of the carriers in the nanodimensional structures of PS. A few workers have suggested that the chemical species, such as siloxene or silicon back bonded species [6] are responsible for the observed PL. It has also been suggested that Si : H groups are essential for emission of PL [7]. Some have emphasized the importance of luminescence from SiO., layers covering the nanoscale silicon [8] while others have concluded that PL in PS is of molecular nature and not related to quantum confinement [9].
Results of all the earlier studies lead to the conclusion that the process of PL is very sensitive to the surface structure of PS. The high surface reactivity with respect to the ambients is also found to be responsible for the degradation of PL over a period of time. In addition to the origin of PL. its long term instability in the ambients remains an unanswered question. The present research activity is. therefore. focused to improve the efficiency and the stability of PL; by tailoring appropriately the surface conditions of PS. From the point of view of the technological applications it is necessary to enhance the photoemission as well as to hinder the slow aging phenomenon responsible for the degradation of the PL intensity with time. Attempt to modify the surface properties for this purpose has therefore become an important research problem. Some workers have reported the improvement in the luminescence efficiency by hydrogenation and oxidation [7,10]. Fu et al. [2] have observed an appreciable increase in the PL intensity of PS by gamma irradiation. whereas. Grunning et al. [l l] have observed 70 times increase in the PL intensity by remote oxygen containing hydrogen plasma treatment. In an attempt to improve the luminescence efficiency of porous layer. we have carried out several irradiation studies. Noteworthy of these are 10 MeV silicon ions. 10 MeV oxygen ions, 1 MeV electrons, gamma rays from ‘“Co and keV ions from plasma source. In our earlier publication, we have reported the remarkable improvement in the PL efficiency, as well as, its stability with time for 10 MeV silicon ion irradiated PS [4]. An increase in the intensity of PL by an order of magnitude has been reported which is also associated with a blue shift. In addition, very good stability of PL, over a period of one year. has been achieved. In the present paper these improvements in the PL properties have been discussed in relation to the induced structural modifications in the surface region after ion irradiation. Grazing angle X-ray diffraction (XRD) studies have revealed the preferential recrystallisation resulting into a slightly different morphology in the ion irradiated samples.
2. Experimental Electrical contacts were established on one of the surfaces of 10 mm x 10 mm, p-type silicon with a resistivity of 0.4 Q cm. The PS was obtained by anodic etching using an electrolyte containing 48% HF and ethanol in the proportion of 1 : 1. A specpure graphite plate was used as the cathode. Each PS sample was prepared by keeping identical conditions of etching and passing anodic current of density of 40 mA/cm’ for a period of 20 min. As prepared samples of PS appeared reddish brown in colour. The thickness of the PS layer was estimated from the cross sectional measurements with a scanning electron microscope (Model 100 made by CSIO, India), and was found to be around 9-12 urn. The average crystallite size was determined with transmission electron microscopy. The micrograph in Fig. l(a) shows almost spherical microcrystallites of PS. The average particle size was estimated to be around 5 nm. The diffraction pattern as shown in Fig. l(b) has been analysed to reveal the crystal structure of the microcrystallites. Interplaner spacing (d) was found to correlate with that of silicon. The PL spectra were recorded at an excitation wavelength of 300 nm, using a Perkin Elmer LS-50 B fluorescence spectrometer at room temperature. A depth profile information of the crystalline structure of the PS sample was then investigated, in order to understand the structural modification. if any, caused by irradiation, by recording the grazing angle XRD spectra for angles ranging from 0.3” to 5”. Infrared spectra were recorded with a Perkin Elmer Spectrophotometer model 230. The samples were irradiated in vacuum (-lo-’ Torr) with 10 MeV silicon ions. Different fluences ranging from 10” to 5 x lOI ions/cm were used for the irradiation. Ions of this energy were obtained by passing 70 MeV multiply charged (7+) silicon ions through an aluminium foil of appropriate thickness. The penetration depth of 10 MeV silicon ions in the crystalline silicon is - 5 urn (using TRIM(92) programme). The ion irradiation facility of the Nuclear Science Centre, New Delhi, India, was used for this work. The beam was
T. M. Bhace et al. I Nucl. Instr. and Meth. in Phys. Rex B 132 (1997) 409417
Fig.
1. (a) Transmission
micrograph
of virgin PS recorded
defocussed to cover the entire area of the sample. The ion fluence was measured by a current integrator connected to the target holder on which the sample was mounted. Electron beams of 1 and 6 MeV energies were available from the race track microtron [12]. The beam was scattered by a tungsten foil to obtain ;I uniform electron intensity over a circular area of 50 mm diameter.
in bright
field mode; (b) electron
411
diffraction
pattern
of virgin PS
The gamma irradiation was carried out with a 6oCo source, with a dose rate of 0.35 mrad/h for 60 min.
3. Results and discussion A typical PL spectrum sample is shown in curve
of an unirradiated PS a of Fig. 2. This spec-
412
T.M. Bhure et ul. I Nucl. Insir. and Meth. in PI~Ix Rrs. B 132 (1997) 409417 Table 1 Data giving the area under the PL curves of PS irradiated different radiations Irradiated
Total area under the PL curve (a.u.)
Area under the PL curve using baseline correction 1a.u.)
Silicon ion irradiated PS
21 910.2
21 623.X
1 MeV electron irradiated PS
14 663.3
14 457.6
6 MeV electron irradiated PS
7944.9
7837.8
18 396.9
18 103.0
5173.1
4974.8
Gamma irradiated 40 mA unirradiated Waveleng(h
(nm)
sample type
with
PS PS
-
Fig. 2. PL spectra of PS irradiated with different radiations: (a) Unirradiated virgin PS; (b) 10 MeV Si ion irradiated PS with PS; (d) fluence of lo’? ions/cm’; (c) (‘“Co gamma irradiated I MeV electron irradiated PS; (e) 6 MeV electron irradiated PS.
trum has a major peak at 730 nm and a weak shoulder at around 800 nm. Fig. 2 also shows the PL spectra of other PS samples, irradiated by 10 MeV Si ions (curve b), 6”Co gamma rays (curve c), 1 MeV electrons (curve d) and 6 MeV electrons (curve e). An increase in the intensity of the major peak was clearly seen for all these irradiated samples. A relative comparison indicates that the highest PL intensity was obtained in the PS samples irradiated with 10 MeV Si ions. A comparison of the PL intensities of irradiated samples with respect to that of the virgin PS sample is shown in Table 1. Although the peak shape slightly differs in each of these spectra, the point worth noticing is that the overall PL response, in each of the irradiated samples, has increased. These results are given in Table 1 in terms of area under the curve for each PL spectrum in arbitrary units. Dose dependent PL studies were carried out for the ion irradiated PS samples, since they yielded the highest PL intensities. Fig. 3 shows the variation of PL peak intensity with the total fluence. The maximum PL intensity was obtained for a fluence of lOi ions/cm’. Further detailed structural investigations with the irradiated PS were carried out in details only for the 10 MeV silicon ion irradiation with the flu-
ence of lOI ions/cm* since this system alone is projected out as the most efficient candidate. In all the Si ion irradiated samples, apart from the enhancement in the PL response, an appreciable stability to the structure has been imparted by ion irradiation which is revealed by the plot of Fig. 4. It has been observed that the enhanced PL intensity of the ion irradiated PS remains constant over a period of one year. whereas, the PL intensity of an unirradiated sample decayed by approximately 40”/;1 over the same period. The grazing angle XRD measurements provided some information about the crystallinity of the ion irradiated PS layers. Fig. 5 compares the XRD patterns recorded at different grazing angles ranging between 0.3” and 5.0” for silicon ion irradiated samples. These angles of incidence correspond to the information depths ranging between 77 A and 12 urn. These depths have been calculated [ 131 using the relations:
(1) and t = ?IT /l
for c( > x,.
(2)
where E.is the wavelength of incident X-rays, a the angle of incidence, p the linear absorption coefficient. The critical angle tl, for the total reflection is given by
i? M. Bhaue et al. I Nucl. Instr. and Meth. iv P~I~IL’s. Rrs. B 132 ( 1947) 409417
413
4 a.
x d
60-
c %
60-
: g
LO-
01
I
I
’ 12.0
12.5 ~
Fig. 3. A plot of the PL intensity
01 0
Loglo
xc = 1.6 x 10~3p’%
of PL intensity
mn
14.0
Fluence)
for 10 MeV silicon ion irradiated
L
100 Time
Fig. 4. A plot of variation
13.5 (51
(days)
with time for curve: (a) unirradiated
(3)
The X-ray diffraction spectra of the irradiated PS were of different nature as compared to those of the unirradiated samples. The XRD spectra of the irradiated samples, invariably exhibited diffraction peaks corresponding to (3 1 1) planes appearing at 28 z 57”; whereas, those for the unirradiated PS, exclusively, showed a peak at 20 z W, which corresponds to the (I 1 1) plane. An
I
i
I
13.0
IL.5
-
PS as a function
of ion fluence.
I
I
200
300
-
and (b) irradiated
PS samples with 10 MeV silicon ions.
illustration of (3 1 1) and (1 1 1) planes in silicon structure, as obtained from computer simulation, are shown in Fig. 6. The XRD patterns corresponding to the grazing angles of 0.3” and 1” exhibited both the (3 1 1) and ( 1 1 1) planes, whereas. those recorded at the grazing angles of 1.5”~2.5” showed prominent peaks relating to (3 I 1) planes alone. Peaks related to (1 1 I ) reappear at the grazing angles of 4.0” and 5.0” along with comparatively feeble peaks related to (3 1 1)
__ Fig. 5. Grazing
28
-
(Degrees) -
angle XRD patterns
Fig. 6. An illustration
of (3
of PS for angle (z): (a) unirradiated
PS and (b)-(i)
28 mgree4
-
10 MeV Si ion irradiated
PS
I I) and (I I I) planes in Si structure as obtained from computer simulation.
planes. Average microcrystalline sizes, calculated from the measured values of FWHM (using Scherrer’s formula), and the corresponding intensities for the (3 1 1) peaks are plotted as a function of the penetration depth in Figs. 7 and 8, respectively. It is evident that the crystallite size follows a trend similar to that of the intensity; the highest crystallite size and the highest intensity appear at a depth of 4.97 pm from the surface which coincides with the projected range of the 10 MeV silicon ions in Si which is equal to 4.91 pm (TRIM 92). Although peaks related to (3 1 1) planes were also
observed in electron and the gamma irradiated samples; such significant variations in crystallite size and XRD peak intensity were not observed in case of 6oCo gamma rays, 1 and 6 MeV irradiated PS samples. These results, therefore, lead to the conclusion that ion irradiation with 10 MeV silicon ions gives rise to a new microstructure which is different in morphology or stoichiometry than the virgin sample. The structural rearrangement, on account of the ion irradiation, is termed as recrystallisation
41s
T.M. Bhave et ul. I Nucl. Instr. and Meth. in Phvs. Res. B 132 (1997) 409617
~
Penetration
depth
(pm)
-
Fig. 7. A plot of microcrystallite size calculated from the FWHM of XRD peaks of (3 I 1) planes versus penetration depth of X-rays for silicon ion irradiated PS.
by the earlier workers [ 151.This occurs as an effect of the primary interaction, with the medium, in which the ion deposits energy into the substrate in the form of kinetic energy in electronic and nuclear excitations. The energy deposited per unit path length (dE/&) of the ions due to the electronic loss is maximum in the surface region and gradually decreases as ion penetrates in the medium. Whereas the energy dissipated per unit path length in the nuclear interaction is low at the surface region and reaches a maximum in the vicinity of the end of the projected range of the ions. Now the fact that the size of the recrystallized crystallites shows a maximum at a depth of -4.97 pm which is in the vicinity of the end of the projected range of 10 MeV Si ions, therefore leads to a conclusion that the major recrystallisation in PS is caused by nuclear interaction. Similar recrystallisation has not been observed in the crystalline silicon subjected to the similar irradiation. The experimentally determined values of the grain sizes which exhibit the (3 1 1) planes in XRD may
~
Penetration
depth
(,um)
-
Fig. 8. A plot of XRD peak intensity corresponding to (3 1 1) planes versus penetration depth of X-rays for silicon ion irradiated PS. A maximum intensity is observed at a penetration depth of 4.9 urn which is also the range of the ions.
therefore be expected to arise from the energy deposited in nuclear excitation. Similar phenomenon has been observed by Wang et al. [14]. They have formulated empirical equations describing the rate of ion induced recrystallisation at a crystal/amorphous interface of Si. They infer this process to be arising from the formation of knock-ons in the collision cascades which result into the bond rearrangement by a series of displacements and recombinations to the original lattice points. Similar crystallisation has also been observed by Nakata and Kajiyama [ 151in their amorphous silicon films which were irradiated by 2.5 MeV heavy ions of 75As and s4Kr. Since in this process the interatomic bonds are broken by knock-ons, a series of knockons and recombination processes can induce bond rearrangement. The bonding configuration of the surface of the PS sample before and after irradiation was investigated by using the infrared spectroscopy. Table 2 shows the results of infrared spectra recorded for
416
T M. Bhuw et (11.I Nucl. Imtr. and MetA. it1 Pltys. Rrs. B 132
Table 2 Assignments of infrared absorption peaks obtained diated and IO MeV Si ion irradiated samples
for unirra-
Wave number (cm ‘)
Unirradiated PS Bonds present
Irradiated PS Bonds present
460 640 835 870 1040-l 160 2100 2250 3420-3580
Si-0-S; bend SiGH bend Si&H? waggmg Si&H2 scissors SiLO -Si stretch Si&H stretch _
SiC0 Si bend Si~-H bend Si Hz wagging Si H2 scissors Si&O-Si stretch SiCH stretch O-SiL H stretch 0-H stretch
O--H stretch
a typical. as synthesized PS and of a PS sample irradiated with 10 MeV ions. It can be seen that in addition to Si-0-Si and Si-H bonds which exist in the unirradiated PS. new absorption peaks have appeared in the irradiated PS sample at 2250 cm ’ . This peak corresponds to O-Si-H stretch mode. These results, therefore, indicate that though there is no significant change in the signal intensity corresponding to Si-0-Si bonds in the ion irradiated PS, there are additional 0-Si-H bonded species. The stability and improvement in the PL intensity over a long period, observed only in the Si ion irradiated PS samples. seems to be related to the O-Si-H types of species covering the surface. This layer might be providing a protective and passivating effect, as a result, the number of nonradiative recombination centres can be reduced [4]. These results thus indicate that the PL in PS is not only related to the nanocrystallites but also governed to a large extent by the molecular species. Anderson et al. [9] have found the importance of surface stoichiometry and Si-OH alloy in controlling the PL behaviour and have favoured the “molecular origin” as the basis of PL in PS. This point is worth noticing since there are reports where it was inferred that oxygen plays a key role in enhancing the intensity of PL from PS. Earlier workers argued that oxygen atoms could passivate the silicon dangling bonds, thereby reducing the probability of nonradiative recombination, as a result the intensity of PL increases. However in our present experiments no change in the intensity of Si-0-Si related peak in the IR
( 19971 409-417
spectra was observed. On the other hand appearance of 0-Si-H related peaks in the ion irradiated PS samples indicate that instead of only oxygen, the O-H species are perhaps better passivating species for improving the intensity and stability of PL. In addition to these observations the associated blue shift in the PL marks the reduction in the average crystallite size as also inferred from the line width of XRD peaks appearing from the recrystallized silicon crystallites. However the role of the new crystalline phase is not very clear and needs more detailed investigations. In summary, we have made an attempt to study some optical and structural behaviour of heavy ion irradiated PS. The PL intensity is enhanced by 20 times L+S-u-VISthe stability against the ambients is extensively improved. The structural studies have revealed that the irradiation with 10 MeV silicon ions gives rise to a new microstructure which is different in morphology or stoichiometry than the virgin sample; as a result of recrystallisation or chemistry. Infrared studies have revealed the bond rearrangement of bonds on the surface, leading to Si-OH kind of passivation. The high degree of electronic loss per unit of path length of IO MeV silicon ions (dE/dx) in silicon seems to be responsible for the observed rearrangement of the bonds at the surface, whereas, the energy deposited due to nuclear loss induces the recrystallisation. Although in the silicon ion irradiated PS the improvement in optical properties and the recrystallisation in a layer existing at a particular depth seem to have occurred simultaneously, it is not very clear whether there exist any correlation between these two effects. A detailed structural investigation is in progress for the samples irradiated with different high energy heavy ions.
Acknowledgements Thanks are due to Nuclear Science Center, New Delhi, India, for providing the irradiation facilities and to Mr. S.S. Hullavarad for help extended in irradiating the PS samples. One of us (TMB) wishes to acknowledge CSIR, New Delhi. India, for financial support.
TM.
Bhavr et al. I Nucl. Instr. and Meth. in Phys. Rex B 132 (1997) 409417
References [I] L.T. Canham. Appl. Phys. Lett. 57 (1990) 1046. ]2] J.S. Fu. J.C. Mao, E. Wu, Y.Q. Jia, B.R. Zhang. G.C. Qin. G.S. Wui. Y.H. Zhang. Appl. Phys. Lett. 63 (1993) 1830. 13) A.J. Kontkeiwicz, A.M. Kontkeiwicz, J. Siejka, S. Sen, G. Nowak. A.M. Hoff, P. Sakthivel, K. Ahmed, P. Mukherjre. S. Witanachchi. J. Lagowski. Appl. Phys. Lett. 65 (1994) 1436. [4] T.M. Bhave, S.V. Bhoraskar, S. Kulkami. V.N. Bhoraskar. J. Phys. D 29 (1996) 462. [5] L. Brus. Nature 353 (1991) 301. [6] M. Stutzmann, M.S. Brandt, M. Rosenbauer, T. Weber, H.D. Fuchs, Phys. Rev. B 47 (1993) 4806. [7] SM. Prokes. W.E. Carlos. V.M. Bermudez, Appl. Phys. Lett. 61 (1992) 1447.
[8] C.-H. Lin, S.-C. (1993) 902.
Lee. Y.-F.
417 Chen,
Appl.
Phys.
Lett. 63
[9] O.K. Anderson E. Veje. Phys. Rev. B 53 (1996) 15643. [lo] S. Shih, K.M. Jung, D.L. Kwong, M. Kowar. J.M. White, Appl. Phys. Lett. 62 (1993) 1780. [ll] Y. Gruning. S.C. Gujrathi, S. Poulin. Y. DiaWard. A. Yelon, J. Appl. Phys. 75 (1994) 8075. [12] V.B. Asgekar. R.K. Bhalla, B.S. Raye, M.R. Bhiday. V.N. Bhoraskar. Pramana 15 (1980) 479. [13] G. Lim. W. Parrish, C. Ortiz. M. Bellotto. M. Hart, J. Mater. Res. 2 (1987) 471. [14] Z.-L. Wang, N. Itoh. N. Matsunami, Q.T. Zhao. Nucl. Instr. and Meth. B 100 (1995) 493. [15] J. Nakata, K. Kajiyamd. Appl. Phys. Lett. 40 (1982) 686.
-H J&3 __ ELSEVIER
NINMI B
Beam Interactions with Materials 8 Atoms
Nuclear
Instruments
and Methods
in Physics
Research
B 132 (1997) 409417
Radiation induced recrystallisation and enhancement in photoluminescence from porous silicon Tejashree
M. Bhave ‘, S.V. Bhoraskar
‘.*, Prabhat
Singh b, V.N. Bhoraskar
a
d Department of Ph>~sics.University of Pune. Ganeshkhind, Pune 411 007. India ’ National Chemiul Received
19 February
Laboratory. Punr. India
1997; revised form received 21 May 1997
Abstract
Porous Silicon (PS) samples were irradiated with different ionizing radiations which included 1 and 6 MeV electrons, “‘Co gamma rays and 10 MeV silicon ions. Improvement in the efficiency of photoluminescence (PL) and its stability with time were invariably observed in all the irradiated PS samples. Improvement in the luminescent properties was best for samples irradiated with 10 MeV silicon ions. Partial restructuring of Si-0-Si and Si-H type species into Si-OH was confirmed from the infrared spectra of pre- and post- irradiated samples. Grazing angle X-ray diffraction (XRD) analysis revealed that preferential recrystallisation occurs in the irradiated region. The virgin PS sample exhibited only the (I 1 1) peak in the XRD pattern; whereas the irradiated PS sample showed a (3 1 1) peak along with the (1 1 1) peak. The average size of the microcrystallites was calculated from the diffraction peak broadening, using Scherrer’s formula. Depth profile studies, corresponding to the average sizes of the microcrystallites confirmed the existence of (3 1 1) planes, and revealed that the degree of recrystallisation is maximum at the end of the trajectories of silicon ions. 0 1997 Elsevier Science B.V.
1. Introduction Since the discovery of visible photoluminescence (PL) from porous silicon (PS) in 1990 by Canham [l], efforts are being made to improve the PL efficiency [24] as well as to understand the mechanism behind the PL. Several models as regards to the origin of the PL from PS have been suggested. In one of the models [5], the origin of
*Corresponding author. Tel.: 91 0212 352678: fax: 91 0212 350087; e-mail: [email protected]. 0168-583X/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. 1’11SO168-583X(97)00389-3
the PL having energy above the band gap energy of silicon is attributed to the effect of the quantum confinement of the carriers in the nanodimensional structures of PS. A few workers have suggested that the chemical species, such as siloxene or silicon back bonded species [6] are responsible for the observed PL. It has also been suggested that Si : H groups are essential for emission of PL [7]. Some have emphasized the importance of luminescence from SiO., layers covering the nanoscale silicon [8] while others have concluded that PL in PS is of molecular nature and not related to quantum confinement [9].
Results of all the earlier studies lead to the conclusion that the process of PL is very sensitive to the surface structure of PS. The high surface reactivity with respect to the ambients is also found to be responsible for the degradation of PL over a period of time. In addition to the origin of PL. its long term instability in the ambients remains an unanswered question. The present research activity is. therefore. focused to improve the efficiency and the stability of PL; by tailoring appropriately the surface conditions of PS. From the point of view of the technological applications it is necessary to enhance the photoemission as well as to hinder the slow aging phenomenon responsible for the degradation of the PL intensity with time. Attempt to modify the surface properties for this purpose has therefore become an important research problem. Some workers have reported the improvement in the luminescence efficiency by hydrogenation and oxidation [7,10]. Fu et al. [2] have observed an appreciable increase in the PL intensity of PS by gamma irradiation. whereas. Grunning et al. [l l] have observed 70 times increase in the PL intensity by remote oxygen containing hydrogen plasma treatment. In an attempt to improve the luminescence efficiency of porous layer. we have carried out several irradiation studies. Noteworthy of these are 10 MeV silicon ions. 10 MeV oxygen ions, 1 MeV electrons, gamma rays from ‘“Co and keV ions from plasma source. In our earlier publication, we have reported the remarkable improvement in the PL efficiency, as well as, its stability with time for 10 MeV silicon ion irradiated PS [4]. An increase in the intensity of PL by an order of magnitude has been reported which is also associated with a blue shift. In addition, very good stability of PL, over a period of one year. has been achieved. In the present paper these improvements in the PL properties have been discussed in relation to the induced structural modifications in the surface region after ion irradiation. Grazing angle X-ray diffraction (XRD) studies have revealed the preferential recrystallisation resulting into a slightly different morphology in the ion irradiated samples.
2. Experimental Electrical contacts were established on one of the surfaces of 10 mm x 10 mm, p-type silicon with a resistivity of 0.4 Q cm. The PS was obtained by anodic etching using an electrolyte containing 48% HF and ethanol in the proportion of 1 : 1. A specpure graphite plate was used as the cathode. Each PS sample was prepared by keeping identical conditions of etching and passing anodic current of density of 40 mA/cm’ for a period of 20 min. As prepared samples of PS appeared reddish brown in colour. The thickness of the PS layer was estimated from the cross sectional measurements with a scanning electron microscope (Model 100 made by CSIO, India), and was found to be around 9-12 urn. The average crystallite size was determined with transmission electron microscopy. The micrograph in Fig. l(a) shows almost spherical microcrystallites of PS. The average particle size was estimated to be around 5 nm. The diffraction pattern as shown in Fig. l(b) has been analysed to reveal the crystal structure of the microcrystallites. Interplaner spacing (d) was found to correlate with that of silicon. The PL spectra were recorded at an excitation wavelength of 300 nm, using a Perkin Elmer LS-50 B fluorescence spectrometer at room temperature. A depth profile information of the crystalline structure of the PS sample was then investigated, in order to understand the structural modification. if any, caused by irradiation, by recording the grazing angle XRD spectra for angles ranging from 0.3” to 5”. Infrared spectra were recorded with a Perkin Elmer Spectrophotometer model 230. The samples were irradiated in vacuum (-lo-’ Torr) with 10 MeV silicon ions. Different fluences ranging from 10” to 5 x lOI ions/cm were used for the irradiation. Ions of this energy were obtained by passing 70 MeV multiply charged (7+) silicon ions through an aluminium foil of appropriate thickness. The penetration depth of 10 MeV silicon ions in the crystalline silicon is - 5 urn (using TRIM(92) programme). The ion irradiation facility of the Nuclear Science Centre, New Delhi, India, was used for this work. The beam was
T. M. Bhace et al. I Nucl. Instr. and Meth. in Phys. Rex B 132 (1997) 409417
Fig.
1. (a) Transmission
micrograph
of virgin PS recorded
defocussed to cover the entire area of the sample. The ion fluence was measured by a current integrator connected to the target holder on which the sample was mounted. Electron beams of 1 and 6 MeV energies were available from the race track microtron [12]. The beam was scattered by a tungsten foil to obtain ;I uniform electron intensity over a circular area of 50 mm diameter.
in bright
field mode; (b) electron
411
diffraction
pattern
of virgin PS
The gamma irradiation was carried out with a 6oCo source, with a dose rate of 0.35 mrad/h for 60 min.
3. Results and discussion A typical PL spectrum sample is shown in curve
of an unirradiated PS a of Fig. 2. This spec-
412
T.M. Bhure et ul. I Nucl. Insir. and Meth. in PI~Ix Rrs. B 132 (1997) 409417 Table 1 Data giving the area under the PL curves of PS irradiated different radiations Irradiated
Total area under the PL curve (a.u.)
Area under the PL curve using baseline correction 1a.u.)
Silicon ion irradiated PS
21 910.2
21 623.X
1 MeV electron irradiated PS
14 663.3
14 457.6
6 MeV electron irradiated PS
7944.9
7837.8
18 396.9
18 103.0
5173.1
4974.8
Gamma irradiated 40 mA unirradiated Waveleng(h
(nm)
sample type
with
PS PS
-
Fig. 2. PL spectra of PS irradiated with different radiations: (a) Unirradiated virgin PS; (b) 10 MeV Si ion irradiated PS with PS; (d) fluence of lo’? ions/cm’; (c) (‘“Co gamma irradiated I MeV electron irradiated PS; (e) 6 MeV electron irradiated PS.
trum has a major peak at 730 nm and a weak shoulder at around 800 nm. Fig. 2 also shows the PL spectra of other PS samples, irradiated by 10 MeV Si ions (curve b), 6”Co gamma rays (curve c), 1 MeV electrons (curve d) and 6 MeV electrons (curve e). An increase in the intensity of the major peak was clearly seen for all these irradiated samples. A relative comparison indicates that the highest PL intensity was obtained in the PS samples irradiated with 10 MeV Si ions. A comparison of the PL intensities of irradiated samples with respect to that of the virgin PS sample is shown in Table 1. Although the peak shape slightly differs in each of these spectra, the point worth noticing is that the overall PL response, in each of the irradiated samples, has increased. These results are given in Table 1 in terms of area under the curve for each PL spectrum in arbitrary units. Dose dependent PL studies were carried out for the ion irradiated PS samples, since they yielded the highest PL intensities. Fig. 3 shows the variation of PL peak intensity with the total fluence. The maximum PL intensity was obtained for a fluence of lOi ions/cm’. Further detailed structural investigations with the irradiated PS were carried out in details only for the 10 MeV silicon ion irradiation with the flu-
ence of lOI ions/cm* since this system alone is projected out as the most efficient candidate. In all the Si ion irradiated samples, apart from the enhancement in the PL response, an appreciable stability to the structure has been imparted by ion irradiation which is revealed by the plot of Fig. 4. It has been observed that the enhanced PL intensity of the ion irradiated PS remains constant over a period of one year. whereas, the PL intensity of an unirradiated sample decayed by approximately 40”/;1 over the same period. The grazing angle XRD measurements provided some information about the crystallinity of the ion irradiated PS layers. Fig. 5 compares the XRD patterns recorded at different grazing angles ranging between 0.3” and 5.0” for silicon ion irradiated samples. These angles of incidence correspond to the information depths ranging between 77 A and 12 urn. These depths have been calculated [ 131 using the relations:
(1) and t = ?IT /l
for c( > x,.
(2)
where E.is the wavelength of incident X-rays, a the angle of incidence, p the linear absorption coefficient. The critical angle tl, for the total reflection is given by
i? M. Bhaue et al. I Nucl. Instr. and Meth. iv P~I~IL’s. Rrs. B 132 ( 1947) 409417
413
4 a.
x d
60-
c %
60-
: g
LO-
01
I
I
’ 12.0
12.5 ~
Fig. 3. A plot of the PL intensity
01 0
Loglo
xc = 1.6 x 10~3p’%
of PL intensity
mn
14.0
Fluence)
for 10 MeV silicon ion irradiated
L
100 Time
Fig. 4. A plot of variation
13.5 (51
(days)
with time for curve: (a) unirradiated
(3)
The X-ray diffraction spectra of the irradiated PS were of different nature as compared to those of the unirradiated samples. The XRD spectra of the irradiated samples, invariably exhibited diffraction peaks corresponding to (3 1 1) planes appearing at 28 z 57”; whereas, those for the unirradiated PS, exclusively, showed a peak at 20 z W, which corresponds to the (I 1 1) plane. An
I
i
I
13.0
IL.5
-
PS as a function
of ion fluence.
I
I
200
300
-
and (b) irradiated
PS samples with 10 MeV silicon ions.
illustration of (3 1 1) and (1 1 1) planes in silicon structure, as obtained from computer simulation, are shown in Fig. 6. The XRD patterns corresponding to the grazing angles of 0.3” and 1” exhibited both the (3 1 1) and ( 1 1 1) planes, whereas. those recorded at the grazing angles of 1.5”~2.5” showed prominent peaks relating to (3 I 1) planes alone. Peaks related to (1 1 I ) reappear at the grazing angles of 4.0” and 5.0” along with comparatively feeble peaks related to (3 1 1)
__ Fig. 5. Grazing
28
-
(Degrees) -
angle XRD patterns
Fig. 6. An illustration
of (3
of PS for angle (z): (a) unirradiated
PS and (b)-(i)
28 mgree4
-
10 MeV Si ion irradiated
PS
I I) and (I I I) planes in Si structure as obtained from computer simulation.
planes. Average microcrystalline sizes, calculated from the measured values of FWHM (using Scherrer’s formula), and the corresponding intensities for the (3 1 1) peaks are plotted as a function of the penetration depth in Figs. 7 and 8, respectively. It is evident that the crystallite size follows a trend similar to that of the intensity; the highest crystallite size and the highest intensity appear at a depth of 4.97 pm from the surface which coincides with the projected range of the 10 MeV silicon ions in Si which is equal to 4.91 pm (TRIM 92). Although peaks related to (3 1 1) planes were also
observed in electron and the gamma irradiated samples; such significant variations in crystallite size and XRD peak intensity were not observed in case of 6oCo gamma rays, 1 and 6 MeV irradiated PS samples. These results, therefore, lead to the conclusion that ion irradiation with 10 MeV silicon ions gives rise to a new microstructure which is different in morphology or stoichiometry than the virgin sample. The structural rearrangement, on account of the ion irradiation, is termed as recrystallisation
41s
T.M. Bhave et ul. I Nucl. Instr. and Meth. in Phvs. Res. B 132 (1997) 409617
~
Penetration
depth
(pm)
-
Fig. 7. A plot of microcrystallite size calculated from the FWHM of XRD peaks of (3 I 1) planes versus penetration depth of X-rays for silicon ion irradiated PS.
by the earlier workers [ 151.This occurs as an effect of the primary interaction, with the medium, in which the ion deposits energy into the substrate in the form of kinetic energy in electronic and nuclear excitations. The energy deposited per unit path length (dE/&) of the ions due to the electronic loss is maximum in the surface region and gradually decreases as ion penetrates in the medium. Whereas the energy dissipated per unit path length in the nuclear interaction is low at the surface region and reaches a maximum in the vicinity of the end of the projected range of the ions. Now the fact that the size of the recrystallized crystallites shows a maximum at a depth of -4.97 pm which is in the vicinity of the end of the projected range of 10 MeV Si ions, therefore leads to a conclusion that the major recrystallisation in PS is caused by nuclear interaction. Similar recrystallisation has not been observed in the crystalline silicon subjected to the similar irradiation. The experimentally determined values of the grain sizes which exhibit the (3 1 1) planes in XRD may
~
Penetration
depth
(,um)
-
Fig. 8. A plot of XRD peak intensity corresponding to (3 1 1) planes versus penetration depth of X-rays for silicon ion irradiated PS. A maximum intensity is observed at a penetration depth of 4.9 urn which is also the range of the ions.
therefore be expected to arise from the energy deposited in nuclear excitation. Similar phenomenon has been observed by Wang et al. [14]. They have formulated empirical equations describing the rate of ion induced recrystallisation at a crystal/amorphous interface of Si. They infer this process to be arising from the formation of knock-ons in the collision cascades which result into the bond rearrangement by a series of displacements and recombinations to the original lattice points. Similar crystallisation has also been observed by Nakata and Kajiyama [ 151in their amorphous silicon films which were irradiated by 2.5 MeV heavy ions of 75As and s4Kr. Since in this process the interatomic bonds are broken by knock-ons, a series of knockons and recombination processes can induce bond rearrangement. The bonding configuration of the surface of the PS sample before and after irradiation was investigated by using the infrared spectroscopy. Table 2 shows the results of infrared spectra recorded for
416
T M. Bhuw et (11.I Nucl. Imtr. and MetA. it1 Pltys. Rrs. B 132
Table 2 Assignments of infrared absorption peaks obtained diated and IO MeV Si ion irradiated samples
for unirra-
Wave number (cm ‘)
Unirradiated PS Bonds present
Irradiated PS Bonds present
460 640 835 870 1040-l 160 2100 2250 3420-3580
Si-0-S; bend SiGH bend Si&H? waggmg Si&H2 scissors SiLO -Si stretch Si&H stretch _
SiC0 Si bend Si~-H bend Si Hz wagging Si H2 scissors Si&O-Si stretch SiCH stretch O-SiL H stretch 0-H stretch
O--H stretch
a typical. as synthesized PS and of a PS sample irradiated with 10 MeV ions. It can be seen that in addition to Si-0-Si and Si-H bonds which exist in the unirradiated PS. new absorption peaks have appeared in the irradiated PS sample at 2250 cm ’ . This peak corresponds to O-Si-H stretch mode. These results, therefore, indicate that though there is no significant change in the signal intensity corresponding to Si-0-Si bonds in the ion irradiated PS, there are additional 0-Si-H bonded species. The stability and improvement in the PL intensity over a long period, observed only in the Si ion irradiated PS samples. seems to be related to the O-Si-H types of species covering the surface. This layer might be providing a protective and passivating effect, as a result, the number of nonradiative recombination centres can be reduced [4]. These results thus indicate that the PL in PS is not only related to the nanocrystallites but also governed to a large extent by the molecular species. Anderson et al. [9] have found the importance of surface stoichiometry and Si-OH alloy in controlling the PL behaviour and have favoured the “molecular origin” as the basis of PL in PS. This point is worth noticing since there are reports where it was inferred that oxygen plays a key role in enhancing the intensity of PL from PS. Earlier workers argued that oxygen atoms could passivate the silicon dangling bonds, thereby reducing the probability of nonradiative recombination, as a result the intensity of PL increases. However in our present experiments no change in the intensity of Si-0-Si related peak in the IR
( 19971 409-417
spectra was observed. On the other hand appearance of 0-Si-H related peaks in the ion irradiated PS samples indicate that instead of only oxygen, the O-H species are perhaps better passivating species for improving the intensity and stability of PL. In addition to these observations the associated blue shift in the PL marks the reduction in the average crystallite size as also inferred from the line width of XRD peaks appearing from the recrystallized silicon crystallites. However the role of the new crystalline phase is not very clear and needs more detailed investigations. In summary, we have made an attempt to study some optical and structural behaviour of heavy ion irradiated PS. The PL intensity is enhanced by 20 times L+S-u-VISthe stability against the ambients is extensively improved. The structural studies have revealed that the irradiation with 10 MeV silicon ions gives rise to a new microstructure which is different in morphology or stoichiometry than the virgin sample; as a result of recrystallisation or chemistry. Infrared studies have revealed the bond rearrangement of bonds on the surface, leading to Si-OH kind of passivation. The high degree of electronic loss per unit of path length of IO MeV silicon ions (dE/dx) in silicon seems to be responsible for the observed rearrangement of the bonds at the surface, whereas, the energy deposited due to nuclear loss induces the recrystallisation. Although in the silicon ion irradiated PS the improvement in optical properties and the recrystallisation in a layer existing at a particular depth seem to have occurred simultaneously, it is not very clear whether there exist any correlation between these two effects. A detailed structural investigation is in progress for the samples irradiated with different high energy heavy ions.
Acknowledgements Thanks are due to Nuclear Science Center, New Delhi, India, for providing the irradiation facilities and to Mr. S.S. Hullavarad for help extended in irradiating the PS samples. One of us (TMB) wishes to acknowledge CSIR, New Delhi. India, for financial support.
TM.
Bhavr et al. I Nucl. Instr. and Meth. in Phys. Rex B 132 (1997) 409417
References [I] L.T. Canham. Appl. Phys. Lett. 57 (1990) 1046. ]2] J.S. Fu. J.C. Mao, E. Wu, Y.Q. Jia, B.R. Zhang. G.C. Qin. G.S. Wui. Y.H. Zhang. Appl. Phys. Lett. 63 (1993) 1830. 13) A.J. Kontkeiwicz, A.M. Kontkeiwicz, J. Siejka, S. Sen, G. Nowak. A.M. Hoff, P. Sakthivel, K. Ahmed, P. Mukherjre. S. Witanachchi. J. Lagowski. Appl. Phys. Lett. 65 (1994) 1436. [4] T.M. Bhave, S.V. Bhoraskar, S. Kulkami. V.N. Bhoraskar. J. Phys. D 29 (1996) 462. [5] L. Brus. Nature 353 (1991) 301. [6] M. Stutzmann, M.S. Brandt, M. Rosenbauer, T. Weber, H.D. Fuchs, Phys. Rev. B 47 (1993) 4806. [7] SM. Prokes. W.E. Carlos. V.M. Bermudez, Appl. Phys. Lett. 61 (1992) 1447.
[8] C.-H. Lin, S.-C. (1993) 902.
Lee. Y.-F.
417 Chen,
Appl.
Phys.
Lett. 63
[9] O.K. Anderson E. Veje. Phys. Rev. B 53 (1996) 15643. [lo] S. Shih, K.M. Jung, D.L. Kwong, M. Kowar. J.M. White, Appl. Phys. Lett. 62 (1993) 1780. [ll] Y. Gruning. S.C. Gujrathi, S. Poulin. Y. DiaWard. A. Yelon, J. Appl. Phys. 75 (1994) 8075. [12] V.B. Asgekar. R.K. Bhalla, B.S. Raye, M.R. Bhiday. V.N. Bhoraskar. Pramana 15 (1980) 479. [13] G. Lim. W. Parrish, C. Ortiz. M. Bellotto. M. Hart, J. Mater. Res. 2 (1987) 471. [14] Z.-L. Wang, N. Itoh. N. Matsunami, Q.T. Zhao. Nucl. Instr. and Meth. B 100 (1995) 493. [15] J. Nakata, K. Kajiyamd. Appl. Phys. Lett. 40 (1982) 686.