Optical properties of gold nanoparticles on heavily-doped Si substrate synthesized with an electrochemical process

Article abstract of DOI:10.1149/1.3583646

Gold nanoparticles on a heavily-doped Si substrate were synthesized through direct electroreduction of bulk AuCl₄^- ions. The optical properties of the nanoparticles were determined from ellipsometric analysis. The Lorentz-Drude model was used to represent the effective dielectric function of the gold nanoparticles. A prominent peak at ～ 1.7 eV is identified in the imaginary part ₂, which is attributed to the surface plasmon resonance. There are also two minor peaks located at ～ 3.1 and ～ 4.2 eV, respectively, which are related to the interband transitions. The interband transitions are greatly suppressed, which could be due to the modified lattice structures of the nanoparticles.

Full text of DOI:10.1149/1.3583646

K152
Journal of The Electrochemical Society, 158 (6) K152-K155 (2011)
0013-4651/2011/158(6)/K152/4/$28.00 The Electrochemical Society
C
V
Optical Properties of Gold Nanoparticles on Heavily-Doped Si
Substrate Synthesized with an Electrochemical Process
S. Zhu,a T. P. Chen,a,z Z. Liu,a Y. C. Liu,b Y. Liu,c and S. F. Yud
aSchool of Electrical and Electronics Engineering, Nanyang Technological University, Singapore 639798
bSingapore Institute of Manufacturing Technology, Singapore 638075
cState Key Laboratory of Electronic Thin Films and Integrated Devices, University of Electronic Science and
Technology of China, Chengdu 610054, People’s Republic of China
dDepartment of Applied Physics, The Hong Kong Polytechnic University, Hong Kong
ꢀ
Gold nanoparticles on a heavily-doped Si substrate were synthesized through direct electroreduction of bulk AuCl4 ions. The
optical properties of the nanoparticles were determined from ellipsometric analysis. The Lorentz-Drude model was used to represent
the effective dielectric function of the gold nanoparticles. A prominent peak at ꢁ 1.7 eV is identified in the imaginary part e2,
which is attributed to the surface plasmon resonance. There are also two minor peaks located at ꢁ 3.1 and ꢁ 4.2 eV, respectively,
which are related to the interband transitions. The interband transitions are greatly suppressed, which could be due to the modified
lattice structures of the nanoparticles.
C
V
2011 The Electrochemical Society. [DOI: 10.1149/1.3583646] All rights reserved.
Manuscript submitted February 24, 2011; revised manuscript received March 30, 2011. Published April 22, 2011.
Recently, metallic nanoparticles have attracted great interest in
the exploitation of the surface plasmon resonance (SPR).1–3 The ex-
citation of SPR is not only dependent on the extrinsic effects such as
size and shape of the nanoparticles, the surrounding dielectric me-
dium, and the interaction between neighboring nanoparticles, it is
also related to the dielectric function of the metal nanoparticle
itself.1–4 Among variety of techniques employed for the deposition
of metallic nanoparticles, solution-based methods have been used
for decades to synthesize metallic nanoparticles (gold, silver).5,6
This can be done by using a reducing agent together with a metal
precursor to form free-standing metallic nanoparticles.7 Subse-
quently, the prepared nanoparticles are transferred from the solution
to the desired substrate. As a result, a high throughput of metallic
nanoparticles with a narrow distribution of diameter can be obtained.
However, there are two critical issues restrict the further develop-
ment of solution-based methods: (1) poor adhesion to the substrate
and (2) difficult control of particle coverage and aggregation.8
Besides the solution-based method, electrodeposition is an alterna-
tive technique to synthesize metallic nanoparticles.8–10 This is an
effective technique to deposit metallic nanoparticles on conductive
substrate.8 The particle coverage can be controlled by the applied
voltage, while the mean particle diameter can be adjusted by deposi-
tion time, concentration and temperature of the solution.11,12 Owing
to its low resistivity, a heavily-doped Si substrate can be used as an
electrode in the electrodeposition. The obtained nanoparticles are
directly adhered to the Si substrate with good control of coverage
and aggregation. Metallic nanoparticles on Si substrate may have
applications ranging from sensing13 to optoelectronics/photonics14
and surface enhanced spectroscopy.15 For example, in the applica-
tion of surface-enhanced Raman spectroscopy (SERS), the silicon
substrate exhibits a strong and easily recognized photon band at
ꢁ 520 cmꢀ1, offering a useful internal reference or standard. In the
case of electrodeposition of gold nanoparticles on Si substrate, the
gold nanoparticles are firmly adhered to the silicon surface, thus
the nanoparticle arrays have very good long-term stability. In addi-
tion, unlike glass, there is no interfering fluorescence background
from the silicon substrate.
transition are greatly reduced owing to the changes in the lattice
structure.
The electrochemical synthesis of gold nanoparticles was carried
out by using a two-electrode cell.8 A 2 ꢂ 2 cm2 platinum mesh was
used as the anode. A piece of heavily-doped p-type silicon wafer
(size: ꢁ 2 ꢂ 2 cm2; resistivity: 0.001–0.006 X cm), whose native ox-
ide was removed by diluted HF solution, was used as the cathode.
The distance between the two electrodes was fixed at 5 cm. All
chemicals used in this experiment were of analytic grade and were
used without further purification. The electrolytic solutions used in
the synthesis of theꢀg4old nanoparticles consisted of 0.1 mol dmꢀ3
KNO3 and 5.0 ꢂ 10 mol dmꢀ3 HAuCl4. A constant current was
applied (via a Keithley 2400 source meter) to the electrolysis at
room temperature. Scanning electron microscope (SEM) was used
to characterize the structural properties of the gold nanoparticles
formed on the silicon substrate. Energy-dispersive X-ray spectros-
copy (EDX) analysis data of the gold nanoparticle film was recorded
by an Oxford Link Isis installed at the SEM (JEOL, JSM-5600LV)
system. The SE measurements were performed using a variable
angle spectroscopic ellipsometer (J. A. Woollam, Inc). The ellipso-
metric angles (W and D) were measured in the wavelength range of
300–1100 nm with a step of 5 nm at three different incident angles
(i.e., 65, 70, and 75ꢃ).
Figure 1(a) shows the representative plane-view SEM image of
the gold nanoparticles. It is observed that there are many large par-
ticles (tens of nanometers in size) randomly distributed on the sub-
strate surface. In addition, there are also many tiny nanoparticles
between the large particles. It is believed that the large particles
were formed as a result of the aggregation of the tiny nanoparticles.
The surface morphology, size and size distribution of these gold
nanoparticles on the Si substrate are strongly dependent on the dis-
tance between the anode and cathode electrodes, as well as the dura-
tion of deposition. For example, the nanoparticles could be more
monodisperse for a short duration of deposition.8 The particle size
distribution is shown in Fig. 1(b). A Gaussian profile was used to fit
to the size distribution of the nanoparticles. It is found that the mean
particle size and standard deviation are of 61.3 and 21.0 nm respec-
tively. The EDX analysis shown in Fig. 2 confirms that the nanopar-
ticles are indeed gold nanoparticles.
In this study, gold nanoparticles on a heavily-doped Si substrate
were synthesized using electrochemical method. The effective
dielectric function of the nanoparticles, which is represented by
three Lorentz oscillators and a Drude term, was determined from the
spectroscopic ellipsometric (SE) analysis. The result shows that
the SPR band is red-shifted due to a large average diameter of the
nanoparticles, while the amplitudes of the two bands of interband
The effective dielectric function of gold nanoparticles can be rep-
resented by the Drude term and Lorentz oscillators, which are attrib-
uted to the free and bound electron transitions, respectively. The
corresponding energy-dependent dielectric function is described as16,17
Ep2
n
X
Ai
eðEÞ ¼ e1 þ
ꢀ
[1]
E
D
E2i ꢀ E2 ꢀ iCiE E2 þ iꢀh s
z E-mail: echentp@ntu.edu.sg
i¼1
Downloaded on 2013-12-03 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 158 (6) K152-K155 (2011)
K153
eV and sD is electron relaxation time with unit of fs. The number of
Lorentz oscillators (n) is determined to be three (i.e., n ¼ 3), as two
oscillators model the interband transitions and the other oscillator
models the surface plasmon resonance. On the other hand, the gold
nanoparticle layer is considered as a composite layer consisting of
the gold nanoparticles and voids, whose effective dielectric function
can be calculated using the Maxwell-Garnett effective medium
approximation (EMA).16,17 As shown in Fig. 1(a), the surface coverage
of the gold nanoparticles is less than 30%, thus, the volume fraction
of the voids should be larger than the volume of the gold nanopar-
ticles. In fact, the volume fractions of gold nanoparticles and voids
yielded from the SE spectral fitting described below are ꢁ 20 and
ꢁ 80%, respectively. Therefore, the gold nanoparticles should be
treated as an inclusion in the EMA.
The spectral fitting to the experimental ellipsometric angles
(W and D) was performed by adjusting the parameters in Eq. (1) to-
gether with the volume fraction of the gold nanoparticles and the
composite layer thickness to minimize the mean-square error
between the measured and calculated ellipsometric angles.18 Figure 3
shows the spectral fittings to the experimental data at different inci-
dent angles. All spectral features can be fitted excellently. The
Lorentz model parameters obtained from the fittings are listed in
Table I.
Figure 4 shows the effective dielectric function e of the gold
nanoparticles as a function of photon energy. A prominent peak at
ꢁ 1.7 eV can be identified in the imaginary part e2, while two minor
peaks at ꢁ 3.1 and ꢁ 4.2 eV can be barely observed. All these peak
positions are consistent with the resonance energies of the Lorentz
oscillators yielded from the spectral fittings (see Table I). The real
part e1 of the effective dielectric function is positive over the entire
Figure 1. (a) A plane-view SEM image of the gold nanoparticles electro-
chemically synthesized on a heavily-doped Si substrate. (b) Size distribution
of the gold nanoparticles.
where e is a background constant or an offset, Ai is the amplitude
1
of the ith oscillator with the unit of (eV)2, Ei is the resonance energy
with the unit of eV, Ci is the damping factor of the oscillator with
the unit of eV, Ep is the unscreened plasma energy with the unit of
Figure 3. Spectral fittings to the experimental ellipsometric data at different
incident angles.
Figure 2. EDX analysis of the sample of gold nanoparticles on Si substrate.
Downloaded on 2013-12-03 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

K154
Journal of The Electrochemical Society, 158 (6) K152-K155 (2011)
was observed in the EDX analysis (see Fig. 3). On the other hand,
Table I. Values of the Lorentz oscillator parameters of Eq. 1
yielded from the SE fittings.
the Lorentz oscillators with the resonance energies of ꢁ 3.1 and
ꢁ 4.2 eV are attributed to the interband transitions. The oscillator of
3.1 eV corresponds to the transitions from the uppermost 5d-valence
electron bands to states in the s, p bands just above the Fermi level;
while the oscillator of 4.2 eV represents the transitions from the
lower-lying d bands to band 6 and from s, p states in band 6 to s-, p-
like states in band 7.24 The transitions would be affected by the
structures of the gold nanoparticles. As can be concluded from Fig. 4,
the amplitudes of the two bands of interband transitions are very
small. This should be related to the structures of the gold nanopar-
ticles, but the detail is not known yet.
Lorentz 1
Lorentz 2
Lorentz 3
Ai (eV2)
Ei (eV)
Ci (eV)
4.48
1.73
0.87
1.29
3.07
1.20
4.70
4.32
2.70
spectral range of the ellipsometric measurement, which is consistent
with the previous study of discrete metal islands.19,20
The contributions of the three Lorentz oscillators to the dielectric
function are also shown in Fig. 4. The first oscillator with the reso-
nance energy of ꢁ 1.7 eV is associated with the surface plasmon res-
onance (SPR) of the gold nanoparticles. As the size of the gold
nanoparticle becomes much smaller than the wavelength of incident
light, light can easily penetrate through the whole nanoparticle,
which will cause a redistribution of free electrons with respect to the
positively charge ions. The formed metallic lattices can be consid-
ered as resonators, and their oscillation amplitude will be magnified
if excited resonantly.21 In general, the SPR band is determined by
the size and shape of the nanoparticles, as well as the surrounding
dielectric medium. In our study, the SPR band is located at ꢁ 1.7 eV.
This energy is lower than the commonly proposed characteristic
band position (ꢁ 2.2 eV).22,23 The red-shifted SPR band observed
here can be explained by the large average diameter of the nanopar-
ticles.23 Image charge effect from Si substrate was also proposed for
a red-shifted SPR band.16 However, the image charge effect might
not play an important role here, because the Si substrate could have
been oxidized during the electroreduction process as a clear O signal
The refractive index n and the extinction coefficient k of the gold
nanoparticles can be obtained from the calculation using the dielec-
tric function shown in Fig. 4 with the following equations
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
e21 þ e22 þ e1
n ¼
2
[2]
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
e21 þ e22 ꢀ e1
k ¼
2
where e1 and e2 are the real and imaginary parts of the dielectric
function, respectively. The refractive index and extinction coeffi-
cient of the gold nanoparticles as a function of photon energy
obtained from the above calculation are shown in Fig. 5. As can be
observed in the figure, the refractive index and extinction coefficient
of the gold nanoparticles are very different from that of a gold
film.25 The difference is due to the large contribution of the surface
plasmon resonance of the gold nanoparticles and the significant
Figure 4. Effective dielectric function of the gold nanoparticles: (a) real
part e1 and (b) imaginary part e2. The contributions of the three Lorentz
oscillators are shown in the figure also.
Figure 5. (Color online) Refractive index n and extinction coefficient k of
the gold nanoparticles. The optical constants of a gold film (Ref. 25) are also
shown in the figure for comparison.
Downloaded on 2013-12-03 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 158 (6) K152-K155 (2011)
K155
8. D. A. Brevnov and C. Bungay, J. Phys. Chem. B, 109, 14529 (2005).
changes in the interband transitions associated with the structures
(e.g., size and shape) of the nanoparticles, as discussed above.
In conclusion, gold nanoparticles have been synthesized on a
heavily-doped Si substrate using electrochemical approach. The
morphology of the nanoparticles can be manipulated by adjusting
the distance between the anode and cathode of the electrochemical
cell as well as the process time. The effective dielectric function of
the gold nanoparticles can be modeled with three Lorentz oscillators
together with a Drude term. A prominent peak in the imaginary part
e2, which is attributed to the surface plasmon resonance of the gold
nanoparticles, is observed. Besides, two minor Lorentz bands due to
the interband transitions are also identified.
9. S. X. Huang, H. Y. Ma, X. K. Zhang, F. F. Yong, X. L. Feng, W. Pan, X. N. Wang,
Y. Wang, and S. H. Chen, J. Phys. Chem. B, 109, 19823 (2005).
10. X. Jiang, R. Ferrigno, M. Mrksich, and G. M. Whitesides, J. Am. Chem. Soc., 125,
2366 (2003).
11. A. Radisic, G. Oskam, and P. C. Searson, J. Electrochem. Soc., 151, C369
(2004).
12. R. M. Penner, J. Phys. Chem. B, 106, 3339 (2002).
13. J. Spadavecchia, P. Prete, N. Lovergine, L. Tapfer, and R. Rella, J. Phys. Chem. B,
109, 17347 (2005).
14. S. Pillai, K. R. Catchpole, T. Trupke, and M. A. Green, J. Appl. Phys., 101,
093105 (2007).
15. Z. H. Zhu, T. Zhu, and Z. F. Liu, Nanotechnology, 15, 357 (2004).
16. S. Zhu, T. P. Chen, Z. H. Cen, E. S. M. Goh, S. F. Yu, Y. C. Liu, and Y. Liu, Opt.
Express, 18, 21926 (2010).
17. S. Zhu, T. P. Chen, Z. H. Cen, Y. C. Liu, and Y. Liu, Electrochem. Solid-State
Lett., 13, K39 (2010).
Acknowledgments
18. L. Ding, T. P. Chen, Y. Liu, C. Y. Ng, and S. Fung, Phys. Rev. B, 72, 125419
(2005).
19. A. J. De Vries, E. S. Kooij, H. Wormeester, A. A. Mewe, and B. Poelsema, J.
Appl. Phys., 101, 053703 (2007).
20. T. W. H. Oates, D. R. McKenzie, and M. M. M. Bilek, Phys. Rev. B, 70, 195406
(2004).
This work has been financially supported by National Research
Foundation of Singapore (NRF-G-CRP 2007-01). Y. Liu acknowl-
edges NSFC under project No. 60806040, and the Grant under pro-
ject No. 2008ZC80.
21. V. M. Shalaev, Nanophotonics With Surface Plasmons, Elsevier, Amsterdam,
Boston (2007).
References
1. N. J. Halas, Nano Lett., 10, 3816 (2010).
22. E. S. Kooij and B. Poelsema, Phys. Chem. Chem. Phys. 8, 3349 (2006).
23. S. Zhu, T. P. Chen, Y. C. Liu, S. F. Yu, and Y. Liu, Electrochem. Solid-State Lett.,
13, K96 (2010).
24. M. Losurdo, M. M. Giangregorio, G. V. Bianco, A. A. Suvorova, C. Kong, S.
Rubanov, P. Capezzuto, J. Humlicek, and G. Bruno, Phys. Rev. B, 82, 155451
(2010).
2. E. Hutter and J. H. Fendler, Adv. Mater., 16, 1685 (2004).
3. W. A. Murray and W. L. Barnes, Adv. Mater., 19, 3771 (2007).
4. S. Link and M. A. El-Sayed, Annu. Rev. Phys. Chem., 54, 331 (2003).
5. E. S. Kooij, H. Wormeester, E. A. M. Brouwer, E. van Vroonhoven, A. van Silfh-
out, and B. Poelsema, Langmuir 18, 4401 (2002).
25. E. D. Palik, Handbook of Optical Constants of Solids, Academic, Orlando, FL
(1985).
6. D. V. Goia, J. Mater. Chem., 14, 451 (2004).
7. A. N. Shipway, E. Katz, and I. Willner, ChemPhysChem, 1, 18 (2000).
Downloaded on 2013-12-03 to IP 137.99.31.134 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use