Photoluminescent properties of porous carbon films pyrolised on silicon

Article abstract of DOI:10.1002/1521-396X(200011)182:1<395::AID-PSSA395>3.0.CO;2-8

Porous carbon (PC) films with strong photoluminescence were prepared by the pyrolysis of semi-crystals of poly(vinyl alcohol) (PVOH) deposited either on c-Si or on porous silicon (PS) substrates. Analysis of the PC films by micro-Raman spectroscopy was un

Full text of DOI:10.1002/1521-396X(200011)182:1<395::AID-PSSA395>3.0.CO;2-8

V. Baranauskas et al.: PL Properties of Porous Carbon Films Pyrolised on Si
395
phys. stat. sol. ꢀa) 182, 395 ꢀ2000)
Subject classification: 61.16.Ch; 68.55.Jk; 78.55.Mb; S5
Photoluminescent Properties
of Porous Carbon Films Pyrolised on Silicon
V. Baranauskas1), M. C. Tosin, H. J. Ceragioli , J. G. Zhao,
A. C. Peterlevitz, and S. F. Durrant
Departamento de Semicondutores, Instrumentos e FotoÃnica, Faculdade de Engenharia
EleÂtrica e ComputacËaÄo, Universidade Estadual de Campinas, Av. Albert Einstein N. 400,
13083-970 Campinas ± SP, Brazil
ꢀReceived March 12, 2000)
Porous carbon ꢀPC) films with strong photoluminescence were prepared by the pyrolysis of
semi-crystals of polyꢀvinyl alcohol) ꢀPVOH) deposited either on c-Si or on porous silicon ꢀPS)
substrates. Analysis of the PC films by micro-Raman spectroscopy was undertaken simulta-
neously with micro-photoluminescence spectroscopy to enable correlation of the characteristics
of the luminescence spectra with the probable emission structures. The results suggest that the
strong luminescence observed in these PC films originates from hydrogenated structures, and
from the OH bonds present in the crystalline phases of the PVOH. Morphological data obtained
using scanning electron microscopy ꢀSEM) and atomic force microscopy ꢀAFM) are also dis-
cussed.
Introduction Nanoporous materials for the production of new types of microstructures
promise to have a large impact on industry as well as on fundamental research in mate-
rials science. Such materials are of interest for the production of gas detectors, molecu-
lar filters, photonic crystals, displays, waveguides, biologically compatible prostheses,
and so on [1, 2].
Porous silicon ꢀPS) is prominent among these nano-structured materials because of
its luminescence [3]. As carbon and silicon fall within the same group of the Periodic
Table of elements, structures of porous carbon may also have interesting optical proper-
ties, including luminescence. The technological difficulty to be overcome is the tailoring
of the porosity, the degree of hydrogenation, the composition of adsorbed radicals and
of the nano-crystallinity of such carbonaceous materials.
The work described here shows the possibility of obtaining porous carbon ꢀPC) films
with strong luminescence, starting from the pyrolysis of polyꢀvinyl alcohol) semi-crystals
ꢀPVOH) deposited onto c-Si or PS. The use of this polymeric compound allowed us to
obtain films adherent to the Si surface, whose structure, after pyrolysis, becomes close
to that of diamond-like carbon ꢀDLC). Micro-Raman and micro-photoluminescence
analysis were undertaken simultaneously. Therefore it was possible to correlate the
structures observed by micro-Raman analysis with the luminescence of the samples.
The morphologies identified by scanning electron microscopy ꢀSEM) and atomic force
microscopy ꢀAFM) are also presented.
1) Corresponding author; Tel.: ꢀ55) 19 3 7883851; Fax: ꢀ55) 19 3 2891395;
e-mail: vitor@dsif.fee.unicamp.br
*
26

396
V. Baranauskas et al.
Experimental Details The films were produced by the immersion of the substrate for 24
hours in the vertical position in a solution of polyꢀvinyl alcohol), previously diluted in hot
ꢀ333 K) deionized water to a concentration of 25 g/l. Then, the samples were carefully
removed from the solution and placed onto a hot-plate at 333 K, to dry for 10 min. Pyro-
lysis was undertaken in a thermal oven under helium at a pressure and total flow rate of
60 Torr and 100 sccm ꢀstandard cubic centimeters per minute), respectively. The oven
temperature was increased from room temperature to the final pyrolysis temperature in 2
h, and reduced at the same absolute rate, giving a total processing time of 4 h. These slow
heating and cooling rates were used to minimize the thermal stresses of the coatings and
the substrate. The maximum pyrolysis temperature was between 513 and 843 K.
Substrates of polished n-type c-Si of h100i crystalline orientation and a resistivity of
0.6 Wcm were used. For the formation of PS, anodization was carried out only in the
center of the substrate, using an electrolytic cell with boundaries ꢀdiameter ꢀ1 cm)
established by O-rings, as described in the literature [4]. This procedure allows the
simultaneous deposition of the PVOH films onto the PS ꢀcentral area of the substrate)
or onto the polished c-Si ꢀremainder of the substrate area). Etching was accomplished
using an electrolyte composed of a mixture of HF ꢀ48wt%), ethanol ꢀ98wt%) and
water in the proportions ꢀ1:2:2), with a current density of 10 mAcm±2 for 2 h. The PS
formed under these conditions has deep pores ꢀ>100 mm) and high porosity [5].
Surface morphology was examined by scanning electron microscopy and atomic force
microscopy using a JSM-5900 LV SEM and a Nanoscope II AFM. Both micro-Raman
and micro-photoluminescence spectroscopy were carried out at ambient temperature
using a Renishaw microprobe system, equipped with a microscope whose focal spot
‡
diameter was about 2 mm and employing the output from an Ar laser ꢀ6 mW power)
for excitation at a wavelength of 514.5 nm.
Results and Discussion Figure 1 shows a typical SEM image of the cross-section of a
deliberately fractured PC film deposited onto c-Si. The pyrolysis temperature of this
sample was 513 K. We observe that the film presents a packing structure of vertical
walls. Such walls may be related to the crystalline planes of the PVOH polymer. A
large porosity is observed at the PC/c-Si interface. In addition, the top-surface of the
PC film is planar, with a thickness of around 17 mm.
Figure 2a presents a typical top-view SEM image of a PC sample deposited on PS
ꢀpyrolised at 743 K). The dark features that generally have a cross-like shape are the
PS macro-pores, covered by the PC film. Some spherical particles ꢀwhite color) dis-
persed on the surface of the PC film are ob-
served. Figure 2b shows details of the verti-
cal structure of this sample obtained by a
deliberate fracture. We can observe that the
PC film adheres perfectly to the PS surface,
forming a film of approximately constant
Fig. 1. Typical SEM cross-sectional image of the
PC/c-Si interface. This sample was pyrolised at 513 K

Photoluminescent Properties of Porous Carbon Films Pyrolised on Silicon
397
Fig. 2. Typical SEM images of a PC sample depos-
ited on PS and pyrolised at 743 K: a) Top-view
image of the PC surface and b) cross-sectional
image showing the details of the PC/PS interface
thickness. Apparently there is penetration
of the two films at the interface, with filling
of the superficial pores. The resolution of
the SEM image does not allow us to obtain
more precise details because of electronic
charging effects in the film. We observe
that the thickness of the PC film on the PS
is about 1.5 mm and therefore much smaller
than the thickness of the PC films on c-Si.
The uniform thickness of the films is due
to the inherent characteristics of the PVOH
crystallization process. When the Si wafer is
removed from the diluted PVOH, the mole-
cular forces of adhesion between the poly-
mer and the Si are strong enough to main-
tain a uniform film on the surface of the
wafer. On exposure of this film to air, the surface stress at the PVOH±air interface
induces a local increase in the polymeric bonds and the most superficial layer ꢀskin)
suffers immediate crystallization [6]. The surface of the film acquires vitreous properties
while its interior remains rubbery, being adjusted perfectly to the Si surface until its
complete crystallization. After pyrolysis, the films continue to maintain a uniform
thickness, indicating that the adhesion forces between the polymer and Si are strong
enough to prevent a rupture, even after the thermal dilations and contractions induced
by heating and cooling. The smaller thickness observed for the PC films deposited on PS
ꢀcompared to the thickness of the films deposited onto c-Si) may be due to the small
adhesion force between the polymer and the PS or to the loss of part of the polymeric
material inside the PS pores.
Figure 3 shows typical micro-Raman spectra of the PC samples prepared at 333 and
843 K, respectively. The Raman spectra of the samples prepared at temperatures of 513
and 623 K could not be recorded because of their strong luminescence. The spectrum
of the dry films at the temperature of 333 K ꢀFig. 3a) presents a small peak at 1434
cm±1, whose attribution is unknown. Following pyrolysis, two more broad peaks appear,
an intense one around 1601 cm±1 and a weaker one around 1364 cm±1, with a shoulder
at ꢀ1434 cm±1 ꢀFig. 3b). The width of the peak at 1601 cm±1 is about 70 cm±1 and is
practically the same in the samples pyrolised at 743 and 843 K. In the spectrum of the
sample pyrolised at 843 K, a small peak at 1963 cm±1, of unknown origin, is also ob-
served. The peaks around 1364 and 1601 cm±1 are consistent with the Raman spectra of
glassy carbon and diamond-like carbon films which present two broad peaks around
1345±1355 cm±1 ꢀD-peak) and 1574±1591 cm±1 ꢀG-peak) [7±9]. The presence of the
shoulder at 1434 cm±1 may be related to the O±H bonds present in the samples, and

398
V. Baranauskas et al.
Fig. 3. Typical micro-Raman spectra
of the top-surface of the PC sam-
ples deposited on c-Si as a function
of the pyrolising temperature: a)
333 K, b) 843 K
which form part of the crystal-
line structure of PVOH ꢀFig.
3a). Raman spectroscopy was
also undertaken in the range
400±1300 cm±1, but for all the
samples deposited on c-Si, no
peaks were observed. For the
samples deposited on PS, since
the PC thickness was less, it
was possible to identify a peak
at 520.5 cm±1, corresponding to
the longitudinal optical ꢀLO)
Raman phonon of the c-Si [10].
Figure 4 compares the lumi-
nescence of the samples pre-
pared at several temperatures
on c-Si, as indicated. The sharp
peaks that appear at the left of
most of the spectra correspond
to the D and G active Raman
modes of the carbon structure.
The sample dried at 333 K ꢀcurve a) is almost non-luminescent, except for the small peak
at ꢀ605 nm. The highest luminescence intensity was observed for the samples pyrolised
at 513 K ꢀcurve b). The spectrum of these samples is approximately gaussian, centered
at ꢀ630 nm, with a shoulder around 608nm. With an increase in the pyrolysis tempera-
ture there is a decrease in the luminescence intensity, but retention of the gaussian
profile centered at ꢀ630 nm, as indicated in the spectrum of the sample pyrolised at
623 K ꢀspectrum c). For further increases in the pyrolysis temperature the luminescence
continues to drop, as we can see in spectra d and e, taken from the samples prepared at
743 and 843 K, respectively. In these spectra we can observe the presence of the peak at
ꢀ605 nm and the appearance of another two peaks at ꢀ615 nm ꢀsmall) and at ꢀ658nm
ꢀbroad).
The origin of the luminescence of these films at ꢀ605 nm may be attributed to the
PVOH structure since this peak is present in almost all the samples, from the sample
dried at 333 K to the sample pyrolised at 843 K. The intense gaussian luminescence
profile observed for the samples pyrolised at 513 and 623 K may be attributed to the
hydrogenated-carbon structures present in the films after pyrolysis. At higher tempera-
tures ꢀ743 and 843 K), the rate of release of hydrogen during the pyrolysis may be
larger and may therefore be related to the decrease in the luminescence.

Photoluminescent Properties of Porous Carbon Films Pyrolised on Silicon
399
Fig. 4. Typical micro-photoluminescence spec-
tra of the PC films deposited on c-Si as
a
function of the pyrolising temperature:
ꢀa) 333 K, ꢀb) 513 K, ꢀc) 623 K, ꢀd) 743 K
and ꢀe) 843 K
Figure 5 allows comparison of the lu-
minescence of the samples deposited on
c-Si and on PS. The deposition proce-
dures were absolutely the same since
the surfaces of the c-Si and of the PS
were part of the same wafer, being simul-
taneously processed. Spectra a and b re-
present the luminescence of PC samples
deposited onto PS ꢀspectrum a) and onto
c-Si ꢀspectrum b), and pyrolised at the
temperature of 513 K. We can observe
that the luminescence of the sample de-
posited onto PS has approximately triple the intensity of the sample deposited on c-Si,
and also a small blue-shift ꢀꢀ30 nm) in the peak of the spectrum. These differences are
also observed in the samples prepared at the temperature of 623 K, where spectrum c
corresponds to the PC deposited onto PS and spectrum d to the PC deposited onto c-Si.
The higher intensity observed for the samples deposited onto PS cannot be attributed
to the intrinsic luminescence of the PS, which ꢀfor the PS used in this study) has a
comparative intensity ten times less than the PC samples. Another intriguing fact is that
the apparent thickness of the PC deposited onto PS is smaller than that of the PC
film deposited onto c-Si, and therefore its luminescence intensity should be smaller,
which is not observed. These results suggest
that the PS structure is working either as a
catalyst that modifies the polymer structure
or that the polymer is diffusing into the PS
pores, increasing the effective thickness of
the carbon film much above the thickness
evaluated through the depth measurement
of the apparent polymer/PS interface.
Fig. 5. Photoluminescence spectra of PC films de-
posited ꢀa), ꢀc) on c-Si or ꢀb), ꢀd) on PS at differ-
ent temperatures. Curves a and b correspond to
the pyrolysis temperature of 513 K, and curves c
and d to the pyrolysis temperature of 623 K

400
V. Baranauskas et al.: PL Properties of Porous Carbon Films Pyrolised on Si
Conclusions Strongly luminescent PC films were obtained by the pyrolysis of PVOH
crystals. These films have a uniform thickness, are adherent to either the c-Si surface
or to the PS surface, and possess a polymodal distribution of pores and little spheri-
cal particles ꢀof a few nanometers) dispersed on their surfaces. Such PC coatings
apparently cover the PS surface very well, including the horizontal surface of the
macro-pores. Micro-Raman analyses of the samples prepared at several temperatures
revealed an active Raman mode at ꢀ1434 cm±1, probably related to the original
structure of the PVOH, and broad peaks at around 1364 and 1601 cm±1, which are
compatible with the structure of vitreous carbon or DLC. The pyrolysis of PVOH at
513 K strongly activates its luminescence, producing a gaussian spectrum centered at
ꢀ630 nm with a shoulder at ꢀ608nm. Higher pyrolysis temperatures decrease the
luminescence, and small peaks appear at around 605, 615 and 658nm. The peak at
ꢀ605 nm is present in almost all the spectra, even in that of the sample prepared at
333 K, which suggests its attribution to structures that are part of the PVOH crystal.
The pyrolysis of the films at high temperatures provokes larger loss ꢀeffusion) of
hydrogen and since they show a decrease in the luminescence, it is suggested that the
strong luminescence may be due to the activation of the hydrogenated structures of the
crystalline phases of the PVOH that may occur in the mid temperatures ꢀꢀ513±623 K).
The luminescence of the PC films deposited onto PS was much larger than the lumi-
nescence of the PC films deposited onto c-Si. Such an increase cannot be attributed
to only the PS luminescence, because the type of PS used has a low intrinsic lumines-
cence. A possible mechanism is the molecular diffusion of the polymer, or of its com-
ponents, inside of the PS structure.
Acknowledgements The authors thank V. J. Trava Airoldi, E. J. Corat and N. F. Leite
of INPE, Brazil, for useful discussions. The electron microscopy work was performed
with the JSM-5900 LV microscope of the LME/LNLS, Campinas. We gratefully
acknowledge the Brazilian agencies FAPESP, CAPES and CNPq for partial financial
support.
References
[1] L. T. Canham and D. Bellet, in: Proc. New Developments in Porous Silicon, Thin Solid Films
297, xi ꢀ1997).
[2] V. Parkhutik and L. T. Canham, in: Proc. Porous Semiconductors ± Science and Technology
in Porous Silicon, J. Porous Mater. 7, 9 ꢀ2000).
[3] L. T. Canham, Appl. Phys. Lett. 57, 1046 ꢀ1990).
[4] V. Baranauskas, G. P. Thim, and A. Peled, Appl. Surf. Sci. 86, 398ꢀ1995).
[5] V. Baranauskas, M. C. Tosin, A. C. Peterlevitz, H. J. Ceragioli, and S. F. Durrant, Mater.
Sci. Engng. B 69/70, 171 ꢀ2000).
[6] M. O. Ngui and S. K. Mallapragada, J. Polymer Sci. 36, 2771 ꢀ1998).
[7] R. J. Nemanich and S. A. Solin, Phys. Rev. B 20, 392 ꢀ1979).
[8] M. Yoshikawa, G. Katagiri, H. Ishida, A. Ishitani, M. Ono, and K. Matsumura, Appl. Phys.
Lett. 55, 2608ꢀ1989).
[9] D. S. Knight and W. B. White, J. Mater. Res. 4, 385 ꢀ1989).
[10] Z. Igbal and S. Veprek, J. Phys. C. 15, 46 ꢀ1982).