Upload
m-khardani
View
222
Download
4
Embed Size (px)
Citation preview
phys. stat. sol. (c) 4, No. 6, 1986–1990 (2007) / DOI 10.1002/pssc.200674420
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Bruggeman effective medium approach for modelling optical
properties of porous silicon: comparison with experiment
M. Khardani, M. Bouaïcha*, and B. Bessaïs
Laboratoire de Photovoltaïque et des Semi-conducteurs, Centre des Recherches et des Technologies de
l’Energie, B. P. 95. 2050 Hammam-Lif, Tunisia
Received 17 March 2006, revised 15 September 2006, accepted 15 November 2006
Published online 9 May 2007
PACS 78.20.Bh, 78.20.Ci, 78.66.Db, 78.67.Hc, 78.67.Lt
While numerous works deal with the properties and applications of porous silicon (PS), some of the re-
lated topics are not complete or could be investigated from different aspects. The main objective of this
paper is to provide novel information associated with the optical properties of nano- and meso-PS by
studying the variation of the effective refractive index (neff) and the relative dielectric constant (εr,eff) as a
function of porosity. For this purpose various PS samples were prepared by electrochemical etching of p
and p+-type silicon wafers in order to form silicon supported nano-PS and free-standing meso-PS layers,
respectively. The experimental effective optical parameters (neff , εr,eff ) of the meso-PS films, were deter-
mined from the transmission spectra and the Bragg law. While, in the case of nano-PS layers, we applied
the Goodman method, and deduced the values of neff and εr,eff from the Bruggeman’s effective medium
approximation (EMA). In the EMA calculation, the PS structure was considered as being a physical com-
bination of three distinct phases formed by silicon, silicon dioxide and voids with a convenient volume
fraction. A good agreement between theory and experiment was found in the case of silicon-supported
nano-PS for all porosities. However, for free-standing meso-PS, the theory does not well fit the experi-
mental results for porosities lower than 50% and higher than 70%.
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 Introduction
The description of optical wave propagation through porous media provides an intriguing and distin-
guished physical problem. Porous silicon (PS) is a material that has awakened interest for a wide variety
of applications including optoelectronics, chemical sensing [1], optical components [2], and as a sub-
strate for carbon nanotubes growth [3]. The use of PS thin films in the above mentioned applications
requires the determination of their refractive index and dielectric constant. According to the IUPAC
guidelines, PS- based media have been classified depending on pore size. Nanoporous and mesoporous
silicon prepared from p and p+-type silicon, consists of randomly interconnected quantum dots (size
dimension ≤ 4 nm), and an arrangement of quantum wires (4 - 50 nm) respectively [4]. It is common
practice to use the Bruggeman effective medium theory [5] to estimate the average ‘’effective’’ optical
and/or electrical parameters [6, 7]. The first part of this work deals with a description of the technique
and the choice of the experimental conditions that allow transmission spectra measurements and estima-
tion of the refractive index and the dielectric constant of nano- and free-standing meso-PS using the
Goodman’s method [8] and the Bragg law, respectively. The experimental values of the refractive index
and the dielectric constant of nano and meso-PS were compared to those calculated from Bruggeman's
effective medium approximation.
* Corresponding author: e-mail: [email protected], Phone: + 216 71 430 160, Fax: + 216 71 430 934
phys. stat. sol. (c) 4, No. 6 (2007) 1987
www.pss-c.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2 Theory
In the present work, we assume that the PS films are an arrangement of interconnected quantum dots
(QDs) in nanoporous p-type silicon and quantum wires (QWs) in mesoporous p+-type silicon. When the
wavelength of the incident light is much larger (Infrared Radiation) than the typical nanocrystallites
(QDs and QWs) size (≈ 4 nm in the case of nanoporous silicon and 10 nm in the case of mesoporous
silicon), it is common practice to use effective medium theory of Bruggeman to estimate average ‘’effec-
tive’’ optical constants. Mesoporous and nanoporous silicon are therefore considered as a physical com-
bination of three distinct phases formed by silicon, silicon dioxide and voids. The effective refractive
index of the studied material ( effn ) is related to the refractive index of each phase (nSi, nOx and nv) as well
as to their volume fractions (νSi, νOx and νV). The Bruggeman equation for the three-component system
looks like:
0( 1) ( 1) ( 1)
Si eff Ox eff V eff
Si Ox V
Si eff Ox eff V eff
n n n n n n
n d n n d n n d nν ν ν
- - -
+ + =
+ - + - + -
(1)
Where d is the system dimensionality [5, 9] (i.e., d = 3 for the nano-PS and d = 2 for the meso-PS). The
volume fraction of each phase satisfies the following equation:
1Si Ox V
ν ν ν+ + = (2)
The effective dielectric constant (εr,eff) is then the contribution of all the material components. It may be
obtained by solving Eq. (3) deduced from the EMA theory:
, , , ,
, , , ,
0( 1) ( 1) ( 1)
r Si r eff Ox r eff V r eff
Si Ox V
r Si r eff Ox r eff V r effd d d
ε ε ε ε ε ε
ν ν ν
ε ε ε ε ε ε
- - -
+ + =
+ - + - + -
(3)
The relative silicon dielectric constant is written as ( )2
rnε = where
in n j n = + , n and
4in
λα
π=
are respectively the real and imaginary parts of the refractive index, and2
1j = - . Applications in IR
are extremely comfortable because in n≺≺ at 800λ nm and then absorption can be neglected.
Hence, we obtain: 2
rnε ª .
Calculation of the effective refractive index for porosities ranging from 30% to 80% has been performed
using the data summarized in Table 1, where only Ox
ν was taken as the fitting parameter.
Table 1 Refractive indexes of the various PS composed phases used to calculate the
effective refractive index of the PS films.
Phase Volume fraction Refractive index
Void νV: porosity 1.00
Oxide νOx 1.43
Silicon νSi 3.44
Equation (1) and Eq. (3) are solved by considering values of nSi from [10] and nOx from [11].
1988 M. Khardani et al.: Bruggeman effective medium approach for modelling optical properties of PS
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-c.com
3 Sample preparation
Nano- and meso-PS layers were obtained by electrochemical etching of (100) oriented boron-doped p+-
type (resistivity of about 0.02 Ωcm) and p-type (resistivity of about 1.00 Ωcm) monocrystalline –Si,
respectively. The meso and nanoporous silicon layers were grown in a HF (40%): absolute ethanol (1:1)
electrolyte at different current density values in order to have layers having different thickness and po-
rosities. Preparation of free-standing mesoporous silicon was realized by a subsequent electrochemical
etching at the end of the anodization process. In order to prevent breaking of the nanostructures forming
the PS layers, the latter were kept under ethanol.
4 Optical characterizations
4.1 Mesoporous silicon
For the calculation of the experimental values of the refractive index and the dielectric constant of free-
standing meso-PS films for different porosities, plane-parallel films were investigated by Fourrier Trans-
form Infrared (FT-IR) spectroscopy using a Nicolet spectrometer in transmission mode. The measure-
ments were performed in the 500 cm–1–5000 cm–1 spectral range. The meso-PS refractive index values,
nPS, were deduced from the spectral position of the interferential fringes detected in the 2000 cm–1–5000
cm–1 spectral range, using the simple relation:
2PSn e p λ= (4)
where e denotes the thickness of the porous film and λ the wavelength of the pth fringe. Equation (4) can
be used for the determination of the thickness if the refractive index is known and vice versa. Knowing
that the thickness of the meso-PS films is determined by the gravimetric method, it is easy to determine
the refractive index.
4.2 Nanoporous silicon
Using the Goodman method, the refractive index nPS of the silicon-supported nano-PS may be obtained
independently of the thickness, by calculating the value of the maximum to minimum amplitude ratio
Tmax/Tmin. Figure 1 depicts the variation of the refractive index of silicon-supported nano-PS and meso-
PS layers for porosities ranging from 30% to 80%.
30 40 50 60 70 80
1,2
1,4
1,6
1,8
2,0
2,2
2,4
2,6
2,8
3,0
Refr
acti
ve i
ndex
Porosity (%)
Fig. 1 Experimental refractive index of the PS medium vs. porosity:
() mesoporous silicon and () nanoporous silicon.
phys. stat. sol. (c) 4, No. 6 (2007) 1989
www.pss-c.com © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
It appears obvious that for both nano- and meso-PS layers, nPS decreases as the porosity increases due to
the fact that the volume fraction of voids rises as the porosity increases. At specific porosities the nPS of
the nano-PS is lower than that of the meso-PS film. Such experimental result can be attributed to the
nanocrystallite size dimension of the PS structure. In fact, the doping level can affect the morphological
aspects of the PS structure which in turn affects wavepropagation through nanocrystallites arrangements.
The dielectric constant of silicon-supported nano- and meso-PS is obtained by applying the appropriate
approximation ( 2
rn=ε ).
5 Comparison with the EMA calculations
Figure 2, shows experimental and theoretical values (deduced from the Bruggeman effective medium
approximation) of the refractive index as a function of porosity for nanoporous silicon layers. In this
case, we obtained a good agreement between experiment () and theory (). Figure 3 shows the experi-
mental and theoretical values (deduced from the Bruggeman effective medium approximation) of the
refractive index as a function of porosity for mesoporous silicon films.
30 40 50 60 70 80
1,2
1,4
1,6
1,8
2,0
2,2
2,4
Refr
acti
ve i
ndex
Porosity (%)
Fig. 2 Refractive index as a function of porosity for
nano-PS deduced from experiment () and the prediction
of the EMA theory for nanoporous silicon ()
30 40 50 60 70 80
1,2
1,4
1,6
1,8
2,0
2,2
2,4
2,6
2,8
3,0
Refr
acti
ve i
ndex
Porosity (%)
Fig. 3 Refractive index as a function of porosity for
meso-PS deduced from experiment () and the predic-
tion of the EMA theory ()
In the case of meso-PS, it was found (Fig. 3) that the calculated values are in good agreement with the
experimental ones in the 50% - 67% porosity range. However, beyond this range of porosities, the calcu-
lated value of the refractive index is lower than the experimental one. This mismatch may be attributed to
be due to the gravimetric method used for the estimation of the thickness of meso-PS films. In fact, the
porosity p is defined as the fraction of voids in the porous structure and satisfies the following equation:
1 2
1 3
m mp
m m
-
=
-
(5)
where, m1, m2 and m3 denote the weight of the original wafer, anodized wafer and stripped wafer (in 1N
NaOH solution), respectively. The small weight differences (m1 - m2) and (m1 - m3) induce some errors in
the p values, especially for high and low porosities. Moreover, the thicknesses of the meso-PS films,
were estimated using the same gravimetric method. Systematically, for high and low porosities, accumu-
lation of errors leads to the observed discrepancy between experimental results and theoretical predic-
tion. Figure 4 depicts the evolution of the experimental and theoretical dielectric constant versus poros-
ity, taking into account the above mentioned approximations.
1990 M. Khardani et al.: Bruggeman effective medium approach for modelling optical properties of PS
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-c.com
30 40 50 60 70 80
1
2
3
4
5
6
7
8
Die
lectr
ic c
on
sta
nt
Porosity (%)
Fig. 4 Experimental dielectric constant as a function of porosity for meso () and nano-PS () and theoretical
dielectric constant for meso () and nano-PS ()
6 Conclusions
In this work, PS was assumed to be an optically isotropic medium composed of three components: sili-
con, silicon dioxide and voids. The effective refractive indexes of nano- and meso-PS were evaluated.
Using the Bruggeman effective medium theory, we obtained good agreement between theoretical and
experimental results for nano-PS layers. However, for meso-PS, the Bruggeman effective medium theory
fits experimental results only for porosities ranging between 50% and 70%. The observed disagreement
for large and low porosities was attributed to the fact that for meso-PS, the experimental values of neff
was evaluated using the Bragg law, where the thickness was estimated by the gravimetric method. The
accuracy of the latter was found to induce additional errors especially at high and low porosity ranges.
References
[1] M. Archer, M. Christophersen, and P. M. Fauchet, Sens. Actuators B 106, 1 (2005).
[2] G. Lérondel, R. Romestain, J. C. Vial, and M. T. Thönissen, Appl. Phys. Lett. 71, 196 (1997).
[3] K. Kordás, A. E. Pap, J. Vähäkangas, A. Uusimäki, and S. Leppävuori, Appl. Surf. Sci. 252, 5 (2005).
[4] A. G. Cullis, L. T. Canham, and P. D. J. Calcott, Appl. Phys. 82, 3 (1997).
[5] D. A. G. Bruggeman, Ann. Phys. 24, 636 (1935). [6] M. Bouaïcha, M. Khardani, and B. Bessaïs, Mater. Sci. Eng. C 26, 2-3 (2006).
[7] M. Khardani, M. Bouaïcha , W. Dimassi, M. Zribi, S. Aouida, and B. Bessaïs, Thin Solid Films 495(1-2), 243
(2006).
[8] A. M. Goodman, Appl. Opt. 17, 2779 (1978).
[9] R. Landauer, AIP. Conf. Proc. 40, 2 (1978).
[10] M. Q. Brewster, Thermal Radiative Transfer and Properties (John. Wiley & Sons, Inc., New York, 1992).
[11] E. A. Taft, J. Phys. C: Solid State Phys. 17, 35 (1984).