Plasma CVD Carbon Nanotubes

1
Plasma CVD Carbon Nanotubes

• Instructor Yonhua Tzeng
• An-Jen Cheng
• April 19 2004

2
Questions

• What is the major effect for growing carbon
  nanotubes by HFPECVD?
• Why can carbon nanotubes be bent by ion
  bombardment?

3
Introduction

• Carbon nanotubes which have a hexagonal structure
  are synthesized in plasmas containing ionized
  carbon atoms, and carbon nanotubes can be divided
  into single- and multi-wall .
• Carbon nanotubes can be synthesized by MW-PECVD,
  RF-PECVD, HF-PECVD, pyrolysis, laser evaporation
  and so on.
• Carbon nanotubes exhibit semiconducting or
  metallic properties depending on their diameter
  and helicity of the arrangement of graphite rings
  in the wall.
• Carbon nanotubes have preeminent electric,
  thermal, and mechanical properties such as high
  aspect ratio, high stress, high resistance to
  chemical and physics attack.

4
Outline

• The effects and parameters of magnetron-type
  radio-frequency plasma for carbon nanotubes
  growth.
• Growing carbon nanotubes by hot filament plasma
  enhanced chemical vapor deposition, and control
  carbon nanotubes shape by ion bombardment.
• Effects of coating a ultrathin polymer films on
  carbon nanotubes by plasma treatment.

5
RF magnetron-type apparatus

• 1. A powered Ni RF electrode is installed in the
  center of a grounded cylindrical chamber.
• PECVD for nanotubes growth is performed under the
  conditions of 0.5 Torr and RF power of 1000W.
• A magnetic field (0Bz340G) is externally
  applied parallel to the powered cylindrical RF
  electrode using solenoid coil in order to achieve
  lower plasma sheath voltage and higher plasma
  density.
• The low-pass filter (LPF) is used to control the
  DC bias voltage component (V rf) and DC current
  density (J rf) toward the RF electrode.
• Gas sources Methane and hydrogen (91).

Ref G-H, Jeong, N. Satake,
T. Kato, T. Hirata, R. Hatakeyama, K. Tohji, Jpn.
J.Appl. Phys., 42 (2003), ppL1340-L1342
6
SEM images show the features of
nucleation and successive nanotubes growth during
PECVD with time evolution. (a) As
polished Ni RF electrode surface (b) after
sputtering for 15 min ( c) after 1 min
growth using the mixture of CH4 and H2 (d)
after 3 min growth (e) after 7 min (f) after
15min growth


Ref G-H, Jeong, N. Satake, T. Kato, T.
Hirata, R. Hatakeyama, K. Tohji, Jpn. J. Appl.
Phys., 42 (2003), ppL1340-L1342
7
Effects of magnetic field

• 1. SEM images showing the effects of magnetic
  field externally introduced to the vacuum
  chamber. (a) Bz 0G, (b) Bz 170G,(c) Bz 340G
• Image (a) shows that when Bz 0 and Vdc-890V,
  most of the creations consist of amorphous carbon
  and graphite material.
• The density of nanotubes grown in Bz 170G and
  Vdc-380V is higher than the result in the case
  of Bz 340G and Vdc-180V.
• It is found that plasma confinement and self-bias
  control by magnetic field introduction have
  critical effects on the carbon nanotubes growth

REF T.
Hirata, N. Satake, G.-H. Jeong, T. Kato, R.
Hatakeyama, K. Motomiya, K.Tohji, Appl. Phys.
Lett., 82, 1119, 2003
8
Effect of current density

• SEM images showing the MWCNTs produced on the RF
  electrode under the condition of (a) Jdc0mA/cm2
  , Vdc-180V (b) Jdc1.5mA/cm2 , Vdc-235V (c)
  Jdc4.0mA/cm2 , Vdc-570V,and (d) dependence of
  nanotubes density on Jdc.
• The DC bias voltage of the RF electrode (Vdc) is
  externally changed for typical magnetic fields (
  170, 240, and 340G ) . An uniform, dense, and
  straight MWCNTs grow in the externally bias case
  of Jdc1.5mA/cm2 , Vdc-235V as shown in Fig.
  (b).
• Carbon nanotubes are observed to grow along the
  local electric field due to the potential drop in
  the plasma sheath formed by Vdc.
• DC bias voltage which directly determine the ion
  bombarding energy in the plasma sheath.
• The typical parameters measured by Langmuir
  probe, such as Te5eV, ne51010cm-3, and fp10V,
  the sheath thicknesses were found to be 3-6.7mm.

REF T. Hirata, N. Satake, G.-H. Jeong,
T. Kato, R. Hatakeyama, K. Motomiya, K.Tohji,
Appl. Phys. Lett., 82, 1119, 2003
9

• (a) Variation of the nanotubes length with growth
  time and (b) diameter distribution of the MWCNTs
  grown for 3 min and 15min (c) TEM image of
  individual MWCNT, and (d) Raman spectrum of the
  uniformly grown MWCNTs.
• We found that the nanotubes growth rate in this
  study is 0.5µm/min, and the nanotubes diameter
  distribution becomes narrower in other words,
  although MWCNTs grown at the early stage have
  various diameters but most of the MWCNTs after 15
  min growth have similar diameter of 100-120nm

Ref G-H, Jeong, N. Satake, T. Kato,
T. Hirata, R. Hatakeyama, K. Tohji, Jpn. J. Appl.
Phys., 42 (2003), ppL1340-L1342
10
HFPECVD Carbon nanotubes growth

• During deposition, the gas pressure of 10 Torr
  (50C2H2 and 50 NH3) was kept constant, and the
  temperature of the substrate was about 650C. The
  power of the DC plasma was 160W (400V, 0.4A) and
  the time of deposition was 3 min. The distance
  between filament and substrate was about 5-10mm.
• Acetylene(C2H2) and ammonia(NH3) gas were used
  as a carbon source and catalyst.
• Prior to carbon nanotubes growth, the substrate
  was cleaned in acetone and methanol for 10 min.

Fig. Schematic diagram of plasma enhanced hot
filament chemical vapor deposition reactor
REF J.-H Han, W.-S. Yang, J.-B. Yoo, C-Y Park,
J. Appl. Phys., 88, 7363, (2000) H. Lim, H. Jung,
S.-K. Joo, Microelectronics engineering 69 (2003)
81-88
11
Effect of plasma density for nanotubes growth

• The bias voltage for plasma decreased from 550 to
  429V , the substrate temperature change from 580
  to 530C with decrease in the bias voltage. As
  shown in Fig.(a), the vertically aligned carbon
  nanotubes were grown at the plasma power of 550V
  and 0.15A. But as shown in Figs. (b) and (c), the
  growth of carbon nanotubes was not observed at
  plasma power lower than 550V(0.15A)
• The filament current is expected to play
  important roles such as heating the substrate and
  electron generation. The fact that the growth of
  carbon nanotubes changes under the same filament
  current implies that plasma power plays a more
  important role than the filament current on
  carbon nanotubes growth.

Fig. Effect of plasma power on the growth
of carbon nanotubes at a constant filament
current of 14 A (a) 550V
(0.15A) (b) 504V (0.11A) (c) 450V (0.06A) (d)
430V (0.02A)


12
Effect of plasma density for nanotubes growth

• The plasma power and the filament current were
  simultaneously varied to maintain a constant
  substrate temperature (556C) in the HFPECVD
  system.
• As shown in Fig.1 (a)-(d), an increase in plasma
  power enabled the carbon nanotubes to be grown
  dramatically even though the filament current
  decreased.
• The plasma power may be the most crucial factor
  for determining the growth characteristics of
  carbon nanotubes.
• Without the filament current, the growth of
  vertically well-aligned carbon nanotubes was also
  observed and shown in Fig.2 (b)

Fig.1.Effect of
plasma power on the growth characteristics of
carbon nanotubes

at
a constant temperature of 556C (a) 500V (0.13A)
(plasma power),12.1A


(filament current) (b) 551V (0.15A), 11.1A (c)
600V(0.21A,(A,(d) 650V (0.25A,7A




Fig.2. SEM images of carbon nanotubes grown on Ni
coated glass without filament current (a) 650V
(0.23A) , 25min (b) 640V (0.23A), 40 min





REF
J.-H. Han, W.-S. Yang, and J.-B Yoo, C.-Y Park,
J. Appl. Phys., 88, 7363 (2000)
13
Ion mass doping system

Fig. Schematic of
the ion mass doping system

Fig. Two cases
of CNTs specimens (a) perpendicular,
(b) at 45 to the ion
shower

REF H. Lim, H. Jung,
S.-K. Joo, Microelectronics Engineering 69
(2003) 81-88
14
Effect of Ion Bombardment

• Ion bombardment was used to bend the carbon
  nanotubes structure, the shape of the CNTs did
  not change when the ion beam was perpendicular to
  the substrate however, CNTs were bent when the
  CNTs was tilted to 45 degree.
• The reasons for the CNTs to bend
• 1.The unique features of the process in
  IMDS such as DC bias and physical etching.
• A strong electric field during the ion
  shower may be the reasons that cause the CNTs to
  bend.

Fig. Carbon nanotubes after Ar ion bombardment
(a) with substrate

perpendicular to ion shower (b) at angle of 45C


REF H. Lim, H. Jung, S-K.
Joo, Microelectronics Engineering 69 (2003), 81-88
15
Cont.

• After placing the carbon nanotubes tilted for an
  hour at the same electric field without ion
  bombardment, there is no trace to find out carbon
  nanotubes have been bent in such a high electric
  field.
• The physical damage on the carbon nanotubes wall
  may be another reason.
• The carbon nanotubes were damage on only one side
  of the wall and bent toward the etched side,
  being etched means that carbon nanotubes loses
  some carbon atoms. It reported that carbon
  nanotubes yield to plastic deformation under
  tensile stress.

Fig. No change of
carbon nanotubes after expose to

electric field of 5 kV/cm without ion
bombardment

Fig. TEM image
of asymmetrically etched carbon nanotubes


REF H. Lim, H. Jung, S-K.
Joo, Microelectronics Engineering 69 (2003), 81-88
16
Polymer film coating

• The carbon nanotubes were vigorously stirred at
  the bottom of the tube and thus the carbon
  nanotubes surface can be continuously rotated and
  exposed to the plasma for thin film deposition
  during the plasma polymerization process, and the
  magnetic bar was used to stir the powders.
• These ultrathin coating could act to activate,
  passivate or functionalize the particle to
  achieve both desirable bulk and surface
  properties.
• To be able to distinguish the deposited polymer
  thin film and the surface of carbon nanotubes, a
  small fraction of C6F14 is introduced to
  copolymerize with pyrrole monomer.

Fig. Schematic diagram of the plasma
reactor for thin polymer film coating of
nano-particles



REF P.He, J. Lian,
D. Shi, L. Wang, D. Mast. Wim J. van Ooij,
M.Schulz, Mat. Res. Soc. Symp. Proc. Vol.740
(2003), 13.19.1-13.19.7
17
HRTEM CNTs images

• The original Pyrograf III PR-24-PS PR-24-HT
  nanotubes shows the graphite structure with the
  interlayer spacing d0020.34nm, based on the
  bright-field TEM and HRTEM images the wall
  thickness of the nanotubes can be speculated to
  be 20-30nm for both kinds of nanotubes.
• In Fig. (b), we can see the nanotubes that were
  treated by plasma polymerization process, the
  thickness of the ultrathin film is approximately
  2-7nm.

Fig. HRTEM images of pyrograf III PR-24-PS
PR-24-HT nanotubes

(a) The fragments of the wall with
inclined planes (002) showing

lattice space on
the outer and inner surfaces of uncoated
nanotubes

with slight roughness(lt1nm) on the
surface (b) An ultrathin film of


pyrrole can be observed on both outer and inner
surfaces of coated nanotubes




REF P.He, J. Lian, D. Shi, L. Wang, D. Mast.
Wim J. van Ooij, M.Schulz, Mat. Res. Soc. Symp.
Proc. Vol.740 (2003), 13.19.1-13.19.7
18
Cont.

• Fig. (b) is the HRTEM image of coated SWCNT in a
  bundle, due to high energies these SWCNTs tend
  to cluster together in an aligned form.
• The polymer film is deposited on the outer
  surface of the bundle as show in Fig. (b)

Fig. HRTEM images of single wall carbon
nanotubes (a) an isolated

SWCNT coated
with pyrrole (b) a bundle of SWCNT coated with
pyrrole




REF P.He, J. Lian, D. Shi, L. Wang, D.
Mast. Wim J. van Ooij, M.Schulz, Mat. Res. Soc.
Symp. Proc. Vol.740 (2003), 13.19.1-13.19.7
19
TOFSIMS

• Time-of flight secondary mass spectroscopy
  (TOFSIMS) was used to investigate the surface
  films of carbon nanotubes.
• The spectrum of the untreated nanotubes show an
  appreciable intensity of carbon, hydrogen, and
  oxygen, which is a characteristic of untreated
  natural surface.
• The spectrum in Fig. (b) shows carbon-fluorine in
  the forms of C4 F7, C3 F7, C4 F6, and C5 F7,,
  indicating highly branched and cross-linked
  polymer structure in the deposited thin film.

Fig. (a)
SIMS data showing uncoated MWCNTs

Fig. (b) SIMS data showing
coated MWCNTs



REF P.He, J. Lian, D. Shi, L. Wang, D.
Mast. Wim J. van Ooij, M.Schulz, Mat. Res. Soc.
Symp.
Proc. Vol.740 (2003),
13.19.1-13.19.7
20
Answer

• Plasma power
• The physical damage on the carbon nanotubes.
  Carbon nanotubes were damage only on one side of
  the wall and bent toward the etched side

21
Reference

• REF G-H, Jeong, N. Satake, T. Kato, T. Hirata,
  R. Hatakeyama, K. Tohji, Jpn. J.Appl. Phys., 42
  (2003), ppL1340-L1342
• 2. REF T. Hirata, N. Satake, G.-H. Jeong, T.
  Kato, R. Hatakeyama, K. Motomiya, K.Tohji, Appl.
  Phys.Lett., 82, 1119, (2003)
• 3. REF J.-H Han, W.-S. Yang, J.-B. Yoo, C-Y
  Park, J. Appl. Phys., 88, 7363, (2000)
• 4. REF H. Lim, H. Jung, S.-K. Joo,
  Microelectronics engineering 69 (2003) 81-88
• 5. REF P.He, J. Lian, D. Shi, L. Wang, D.
  Mast. Wim J. van Ooij, M.Schulz, Mat. Res. Soc.
  Symp. Proc. Vol.740 (2003), 13.19.1-13.19.7
• 6. REF D. Shi, J. Lian, P. He, L.M. Wang, Wim
  J. van Ooij, M. Schulz, Y. Liu, David B. Mast,
  Appl. Phys. Lett. 81,5216,(2002)