Thin Solid Films 451 – 452 (2004) 74–80

Growth vs. nucleation of conducting polymers thin films obtained by

plasma-enhanced chemical vapor deposition
L. Martina, J. Esteveb, S. Borros
´ a,*
a
´
Materials Science Lab, Grup d’Enginyeria Molecular, Institut Quımic `
de Sarria-Universitat Ramon Llull, Via Augusta 390, 08017 Barcelona,
Spain
b
´
Facultat de Fısica, ´
Departament de Fısica `
Aplicada i Optica, Universitat de Barcelona, Avda Diagonal, Barcelona, Spain

Abstract

When dealing with conducting polymers it is important to control the growth to obtain the desired uniformity. With plasma-
enhanced chemical vapor deposition, highly uniform and pin-hole free thin films can be achieved by promoting the growth
processes in detriment of the nucleation processes. This can be done by different means: decreasing the power of the radio-
frequencies that generate the plasma so that less monomers are ionized or by using pulsed plasma CVD. Pyrrole and thiophene
are the monomers that have been used. The range of RF power used is from 4 to 10 W of continuous RF and pulsed RF (tons6
ms, toffs6 ms). Samples have been taken at different times over a period of an hour and have been examined by AFM. The
results show that the morphological characteristics of thin films obtained on the lower power are better than those obtained on
the higher power. It has been demonstrated that the layer growth of both conducting polymers can be controlled using pulsed
plasma.
䊚 2003 Elsevier B.V. All rights reserved.

Keywords: Polymers; Atomic force microscopy; Nucleation

1. Introduction order to have more efficient charge transport properties.

With plasma-enhanced chemical vapor deposition
Conducting polymers have been an increasingly inter- (PECVD), highly uniform and pin-hole free thin films
esting topic for many research groups because of its can be obtained w3,4x. Another approach used in electro-
suitable use in microelectronics and lately as photovol- chemical deposition of conducting polymers is to nanos-
taic coatings. Upon doping, conjugated polymers exhibit tructurate with self-assembled monolayers the substrate
an electron–hole conduction mechanism (similar to that so that each molecule in the monolayer can serve as a
of conventional semiconductors). To produce electrical nucleation site. Films done using this technique have a
photocurrent, the electron–hole pairs must be separated. denser packing morphology, good adhesion to the sub-
This photovoltaic effect has been observed and studied strate and better conductivity w5–7x.
recently w1,2x. It is important to control the growth process to obtain
The best photovoltaic polymer devices produced until the desired film structure and uniformity. With PECVD
now are less efficient than their silicon counterparts. To a great uniformity can be achieved by promoting the
improve the charge transport properties, the structural growth processes in detriment of the nucleation pro-
order must be improved. The chemical and electrochem- cesses. This can be done by different means. If the
ical syntheses of these polymers are well established, as power of the radio-frequencies that generate the plasma
well as the relationship between structure and conduc- is decreased fewer molecules are ionized and the depo-
tivity. The main challenge of the deposition techniques sition is slower, enabling a controlled layer mechanism
is to obtain highly uniform and pin-hole free films in growth w8x and a better structure retention w9,10x. On
*Corresponding author. Tel.: q34-93-589-0413; fax: q34-93-205-
the other hand, the use of pulsed PECVD enables
6266. controlled layer mechanism growth during the off times
E-mail address: [email protected] (S. Borros).
´ w11x and has been reported to be a convenient method

0040-6090/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.tsf.2003.11.149
 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80 75

Fig. 1. AFM images, 5=5 mm2, of (a) PyrC (pyrrole polymerized under continuous plasma), (b) PyrP (pyrrole polymerized under pulsed
plasma), (c) ThioC (thiophene polymerized under continuous plasma) and (d) ThioP (thiophene polymerized under pulsed plasma). Plasma
conditions described in Table 1.

to reduce opening of aromatic rings w12x and heterocy- performed. Growth and nucleation of the resultant films
clic compounds w13,14x. have been studied with the help of AFM images.
For this reason we have studied the growth and
nucleation of conducting polymers thin films obtained 2. Experimental section
by plasma CVD because the understanding of the first
stages of layer formation allow us to control the final Plasma-polymerized films were prepared using pyr-
morphology of the film. PECVD of pyrrole and thio- role (Sigma, as supplied) and thiophene (Sigma, as
phene with RF ranging from 4 to 10 W has been supplied) by radio-frequency plasma polymerization.

Table 1 2.1. Plasma reactor system

Deposition rate of different samples
The plasma polymerization apparatus consists of a
Sample Monomer Type of plasma Deposition rate stainless steel discharge vessel (diameter 30 cm; width
(nm miny1) 25 cm). The RF power supply (27 MHz) is capacitively
PyrC Pyrrole Continuous 11.5 coupled to the internal electrode. The vapor container
PyrP Pyrrole Pulsed 2.2 with the monomer is attached to the deposition chamber
ThioC Thiophene Continuous 21.4 through a vacuum cock and a needle valve. The reactor
ThioP Thiophene Pulsed 4.8
chamber was evacuated to a base pressure of 0.15 mbar
76 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80

Fig. 2. Grain size distribution.

and then the chamber was filled with monomer vapor 2.4. XPS
until the pressure reached the desired level. Ten minutes
later, when the pressure was stabilized, the power source The XPS analyses were performed using a Perkin
was turned on. The films were deposited onto cleaned Elmer PHI 5500 spectrometer. The operating conditions
and degreased glass slides. were as follows: the X-ray source wAl Ka (1486.6 eV)x
was operated at a power of 350 W. The spectrometer
2.2. Infrared spectroscopy was operated in the fixed analyzer transmission mode
at a pass energy of 23.50 eV and a take-off angle of
45 8. The base pressure in the sample chamber was
For IR studies, specimens were scraped from the glass
maintained below 5=10y9 Torr. Total spectral acquisi-
substrate and pelletized after mixing with KBr; these
tion times per sample were typically 10–15 min. Spec-
pellets were used to record IR spectra. For recording IR
tral analyses were carried out using MULTIPAK software
spectra a FTIR Nicolet Magna 560 was used.
for quantification and peak-fitting. The energy scale of
the spectra is corrected to the C–H (1s) at 285.0 eV.
2.3. AFM
3. Results
The surface morphology of the films has been studied
with a Nanoscope III (Digital Instruments, Santa Bar- 3.1. Deposition rate
bara, CA) equipment with an extended electronic mod-
ule. All images were collected on tapping mode and The deposition rate of the films has been calculated
treated with Nanoscope software. The images for the from thickness measurements done using the AFM
morphological studies had a scan size of 5=5 mm2. For technique onto scrapped films at different deposition
thickness measurements, images of a scan size of 10=10 times over a range between 30 s and 1 h (Fig. 1). For
mm2 were done onto scrapped films. all the samples, the power was maintained at 10 W (and
 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80 77

Fig. 3. AFM images, 5=5 mm2, of ThioC (a) 5 min of deposition, (b) 10 min of deposition and (c) 15 min of deposition.

15 kHz for the pulsed plasma) and the monomer The effect of the use of pulsed plasma can be seen
pressure at 0.4 mbar. The results are summarized in first in the reduction of the roughness. Roughness
Table 1. measurements were performed on several images taken
The fact that pyrrole is more labile than thiophene from the same sample in order to assure homogeneity.
w15x explains its lower deposition rate. Most of the RF Comparing films obtained at 1 h of deposition, it can
power is used to break its bonds thus reducing the be seen that PPy films reduce their roughness from 4.85
reaction pathways that lead to polymerization. As it was to 2.99 nm and PTh shows even a greater reduction
expected, the use of pulsed plasma greatly decreases the from 5.52 to 1.84 nm. Although this reduction is
deposition rate. important, there is a second effect that has a greater
impact on the visual appearance of the films. The
3.2. Morphology
reduction in grain size combined with a narrower distri-
Regarding the images obtained with AFM (Fig. 1) it bution of grain size (Fig. 2) leads to obtain films with
can be concluded that pulsed plasma-polymerized films greater uniformity. In continuous plasma films the size
have better morphological characteristics. distribution is less uniform because the grain growth vs.
78 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80

Fig. 4. AFM images, 5=5 mm2, of ThioP (a) 5 min of deposition, (b) 10 min of deposition and (c) 15 min of deposition.

nuclei formation ratio is smaller than that in the pulsed monomer molecules is so intense that nuclei formation
plasma films. This affirmation is supported by the fact overcomes the layer growth. First, a single nucleus is
that the number of grains increases when pulsed plasma formed but soon after that an important number of new
is applied. In PyrC the average number of grains is 156 nuclei is formed. These initial nuclei start growing but
in front of 160 for PyrP. And this trend is even stronger before they can cover the exposed surface new nuclei
with thiophene films: ThioC having 114 grains and are formed above the polymer layer that has not been
ThioP 176. covered completely. On the other hand, in pulsed plasma,
To understand the mechanism, samples were taken at nuclei formation is not so fast and when the first nuclei
different times and examined by AFM. It was observed are formed there is enough time for them to grow and
that, in all the images taken from the continuous plasma cover the surface before a new nucleus deposits over
samples, an important number of grains was present this polymer layer and starts the formation of the second
(Fig. 3). In contrast, only some of the pulsed plasma layer. The flat images are indeed taken at the moment
samples presented grains and the others were extremely where a layer has been formed and no nucleus has been
flat (Fig. 4). In continuous plasma the ionization of the deposited to continue the deposition.
 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80 79

Fig. 5. XPS survey scan of ThioC.

3.3. Composition The low oxygen content on the thiophene samples

confirms the lower bond cleavage due to RF power.
The use of continuous plasma produces two negative This was demonstrated with the XPS spectrum (Fig. 5),
effects on the chemical structure of the films. The first where an oxygen content of 13.4% was revealed.
one is that the constant ion bombardment of the growing Regarding the IR spectra, ThioC samples consist of
film that results in some extent in the formation of broad with absorption peaks at 2930–2800 cmy1 (O–H
trapped free radicals in the film. Upon exposure to stretching), 1750–1650 cmy1 (C_ O stretching) and
atmosphere, these free radicals are oxidized and form 1030 cmy1 (C–O stretching). The same effect as with
C_ O and O–H moieties. The second one is that the pyrrole is observed: broad peaks when using continuous
constant irradiation produced by the glow discharge can plasma vs. narrow peaks and reduction of oxidized
lead to a massive bond cleavage thus breaking complete- forms peaks when using pulsed plasma.
ly the monomer structure.
Both effects are clearly present in the plasma polym- 4. Conclusions
erization of pyrrole. The structure of PPy and PTh films
was investigated by IR and XPS. The IR spectra of It has been demonstrated that the morphological
PyrC samples consist of broad peaks due to the inhomo- characteristics of thin films obtained with pulsed plasma
geneities in the structure. It should also be noted the CVD are much better than those obtained with contin-
presence of absorption peaks at 2930–2800 cmy1 uous plasma. They grow from fewer and smaller nuclei,
(assigned to O–H stretching), 1750–1650 cmy1 (C_ O giving time to the growth processes to cover the surface
stretching) and 1099 cmy1 (C–O–C asymmetrical before new nuclei start a new layer. Therefore, these
stretching). This is consistent with XPS results, which films have an extremely low roughness (as low as 0.3
reveal an oxygen content of 36%. Furthermore, the Cy nm). However, this kind of growth implies that the
N ratio (which has been widely used as a fragmentation overall growth rate is small. This approach would be
indicator) is 5.2, due to the loss of nitrogen atoms from very interesting to applications such as of photovoltaic
the monomer structure. The CyN ratio in the monomer coatings, which require very uniform films because it
is 4. enables better morphological characteristics.
The PyrP sample shows an IR spectrum with more
defined peaks, revealing a more uniform structure. This References
is a direct consequence of the mild character of pulsed
plasma, which is responsible for the ring retention. w1x E. Conwell, Trends in Polymer Science, 5 (1997) 218–222.
Another effect is the reduction in the formation in w2x C.O. Too, G.G. Wallace, A.K. Burrell, G.E. Collis, D.L. Officer,
E.W. Boge, S.G. Brodie, E.J. Evans, Synth. Metals 123 (2001)
trapped free radicals. As a consequence, there is a 53–60.
reduction of the peaks of the oxidized forms (C_ O, w3x F.F. Shi, Surf. Coat. Technol. 82 (1996) 1–15.
O–H and C–O–C). XPS results reveal an oxygen content w4x H. Yasuda, Plasma Polymerization, Academic Press, Orlando,
of 8% and a CyN ratio of 4.6. 1985.
80 L. Martin et al. / Thin Solid Films 451 – 452 (2004) 74–80

w5x C.G. Wu, S.C. Chiang, C.H. Wu, Langmuir 18 (2002) w11x N.M. Mackie, D.G. Castner, E.R. Fisher, Langmuir 14 (1998)
7473–7481. 1227–1235.
w6x J.F. Kang, J.D. Perry, P. Tian, S.M. Kilbey, Langmuir 18 (2002) w12x M.A. Leich, N.M. Mackie, K.L. Williams, E.R. Fisher, Mac-
10196–10201. romolecules 31 (1998) 7618–7626.
w7x E. Smela, Langmuir 14 (1998) 2996–3002. w13x L.M.H. Groenewoud, G.H.M. Engbers, J.G.A. Terlingen, H.
w8x D. Sato, T. Suwa, M.A. Kakimoto, Y. Imai, J. Photopolym.
Wormeester, J. Feijen, Langmuir 16 (2000) 6278–6286.
Sci. Technol. 10 (1997) 149–150.
w9x X. Gong, L. Dai, A.W.H. Mau, H.J. Griesser, J. Polym. Sci., w14x L.M. Han, R.B. Timmons, D. Bogdal, J. Pielichowski, Chem.
Part A: Polym. Chem. 36 (1998) 633–643. Mater. 10 (1998) 1422–1429.
w10x M.E. Ryan, A.M. Hynes, S.H. Wheale, J.P.S. Badyal, C. w15x S. Borros, M.P. Diago, J. Esteve, N. Agullo, Mater. Res. Soc.
Hardacre, R.M. Ormerod, Chem. Mater. 8 (1996) 916–921. Symp. Proc. 725 (2002) 225–230.