Electron field emission from phase pure nanotube films grown in a methane/hydrogen plasma

Article abstract of DOI:10.1063/1.122395

Phase pure nanotube films were grown on silicon substrates by a microwave plasma under conditions which normally are used for the growth of chemical vapor deposited diamond films. However, instead of using any pretreatment leading to diamond nucleation we deposited metal clusters on the silicon substrate. The resulting films contain only nanotubes and also onion-like structures. However, no other carbon allotropes like graphite or amorphous clustered material could be found. The nanotubes adhere very well to the substrates and do not need any further purification step. Electron field emission was observed at fields above 1.5 V/μm and we observed an emission site density up to 10⁴/cm² at 3 V/μm. Alternatively, we have grown nanotube films by the hot filament technique, which allows to uniformly cover a two inch wafer.

Full text of DOI:10.1063/1.122395

Electron field emission from phase pure nanotube films grown in a methane/hydrogen
plasma
Olivier M. Küttel, Oliver Groening, Christoph Emmenegger, and Louis Schlapbach
Citation: Applied Physics Letters 73, 2113 (1998); doi: 10.1063/1.122395
View online: http://dx.doi.org/10.1063/1.122395
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/73/15?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Increased field-emission site density from regrown carbon nanotube films
J. Appl. Phys. 97, 104309 (2005); 10.1063/1.1897836
Highly efficient electron field emission from decorated multiwalled carbon nanotube films
Appl. Phys. Lett. 85, 4487 (2004); 10.1063/1.1819521
Electron field emission properties of gamma irradiated microcrystalline diamond and nanocrystalline carbon thin
films
J. Appl. Phys. 92, 3311 (2002); 10.1063/1.1499996
Origin of electric field enhancement in field emission from amorphous carbon thin films
Appl. Phys. Lett. 78, 2339 (2001); 10.1063/1.1366369
Effect of sp 2 -phase nanostructure on field emission from amorphous carbons
Appl. Phys. Lett. 76, 2627 (2000); 10.1063/1.126430
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
160.36.178.25 On: Sat, 20 Dec 2014 02:22:26

APPLIED PHYSICS LETTERS
VOLUME 73, NUMBER 15
12 OCTOBER 1998
Electron field emission from phase pure nanotube films grown
in a methane/hydrogen plasma
a)
Olivier M. K u¨ ttel, Oliver Groening, Christoph Emmenegger, and Louis Schlapbach
University of Fribourg, Physics Department, P e´ rolles, 1700 Fribourg, Switzerland
͑
Received 5 May 1998; accepted for publication 10 August 1998͒
Phase pure nanotube films were grown on silicon substrates by a microwave plasma under
conditions which normally are used for the growth of chemical vapor deposited diamond films.
However, instead of using any pretreatment leading to diamond nucleation we deposited metal
clusters on the silicon substrate. The resulting films contain only nanotubes and also onion-like
structures. However, no other carbon allotropes like graphite or amorphous clustered material could
be found. The nanotubes adhere very well to the substrates and do not need any further purification
step. Electron field emission was observed at fields above 1.5 V/␮m and we observed an emission
4
2
site density up to 10 /cm at 3 V/␮m. Alternatively, we have grown nanotube films by the hot
filament technique, which allows to uniformly cover a two inch wafer. © 1998 American Institute
of Physics. ͓S0003-6951͑98͒00141-7͔
1
Since Iijima’s original work, carbon nanotubes have
been recognized as fascinating materials with promising ap-
plications in carbon chemistry and physics. At present, there
are several approaches to produce nanotubes. The arc dis-
98%͒ gas mixture, otherwise used for the growth of diamond
films.14 The growth parameters used are standard growth
conditions for the deposition of chemical vapor deposition
͑CVD͒ diamond films, except for the slightly increased sub-
strate temperature. However, instead of scratching the silicon
substrate with diamond powder or biasing it with a negative
dc voltage in order to get a high diamond nucleation density,
we used a different pretreatment. Metal was deposited either
by sputter coating a very thin film of Ni ͑300–400 Å͒ or by
spraying Fe͑NO ͒ dissolved in ethanol onto the silicon sub-
2
charge method and the carbon vapor method use the pyroli-
sis of hydrocarbons ͑e.g., benzene at approx. 1000 °C), by
laser vaporization and variant or combination of these meth-
ods. A comprehensive overview is given in Ref. 4.
Catalytic methods for making nanotubes have their ori-
3
5
gin in the corresponding work on carbon fibers. All methods
3
3
discussed in the previous paragraph have also been used to
make nanotubes by using a metallic catalyst. In the arc
strate. After introducing the sample into the growth chamber
Fe͑NO ͒ is chemically reduced by the high concentration of
3
3
6,7
method such catalysts as Co, Fe, or Ni, as well as Y and
atomic hydrogen to metallic Fe and forms little clusters due
to the increased mobility at 900 °C at the surface. A similar
effect is observed on the sputter deposited Ni cluster. Any
oxygen impurities are removed and the continuous thin layer
of Ni forms little islands. These Ni or Fe clusters act as
catalytic growth centers for nanotubes. A typical growth time
was 15 min. Afterwards the films were analyzed either by a
high resolution scanning electron microscope Zeiss DSM
8,9
Gd were used to grow nanotubes. Mixed catalysts such as
Fe/Ni, Co/Ni, and Co/Pt are reported to give an improved
nanotube yield.1
0,11
Each method has its strengths and weak-
ness. The carbon arc method remains the most useful be-
cause of its ease with which large amounts of nanotubes can
be produced.
A common drawback of all the mentioned methods are a
number of impurities incorporated in the nanotube material,
whose type and amount depend on the deposition technique.
The most common are of carbonic nature such as graphitic or
amorphous nanoparticles. During the last years a number of
purification processes have been developed: gas phase
9
͑
82 ͑HRSEM͒ or by a transmission electron microscope
TEM͒.
The best deposition conditions for nanotube were found
by placing the substrate in a remote position where the
plasma ball, which is typical for this kind of discharge is
located well above the sample. Otherwise, the high concen-
tration of atomic hydrogen etches the growing film. Alterna-
tively, the sample surface could be tilted by 180° in order to
protect the growing nanotube film from the reactive species
of the plasma. Figure 1 shows a HRSEM picture of a nano-
tube film. The picture at the top of Fig. 1 ͑a͒ clearly shows
how the nanotubes grow from the metallic clusters in all
directions like a pin holder. The image in Fig. 1͑b͒ shows
nanotubes arranged in a ‘‘spaghetti-like’’ structure. At higher
resolution the diameter of the individual nanotubes can be
determined to be in the range of 20–60 nm, while its length
can be as long as 100 ␮m. Nanotubes uniformly grow over
the whole substrate and adhere pretty well to it.
1
1
12
purification, liquid phase purification, and purification by
1
3
intercalation. Metallic impurities, used during the catalytic
growth, can be removed for the most part by heating the
sample up to the evaporation temperature of the impurities.8
In this letter we present an alternative method for the
growth of nanotubes. We describe a process by which nano-
tubes can be deposited on a silicon substrate as a thin film. It
is very phase pure ͑only nanotubes or onion-like structures͒
and adheres very well to the substrate.
The films were grown on silicon substrates via micro-
wave plasma chemical vapor deposition in a tubular deposi-
tion system ͑2.45 GHz͒ at a gas pressure of 40 mbar and a
substrate temperature of 900–1000 °C from a CH /H ͑2%/
4
2
At the moment it is rather difficult to speculate about the
growth process. It seems that nanotubes and diamond can
a͒Electronic mail: olivier.kuettel@unifr.ch
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/98/73(15)/2113/3/$15.00 2113 © 1998 American Institute of Physics
60.36.178.25 On: Sat, 20 Dec 2014 02:22:26
0
1

2114
Appl. Phys. Lett., Vol. 73, No. 15, 12 October 1998
K u¨ ttel et al.
FIG. 2. TEM images of the nanotubes. Open as well as closed nanotubes
can be observed ͑a͒. A closer look reveals that also onion-like structures are
formed at the film surface ͑b͒.
FIG. 1. HRSEM pictures of the nanotube film. Nanotubes grow from the
surface of the metallic island in pin-holder manner ͑a͒ ͑image size 9ϫ9
2
␮
m ). The spaghetti-like structure is clearly shown in ͑b͒ ͑image size 30
ϫ30 ␮m ).
2
tiwall nanotubes. However, in many cases onion-like struc-
tures were found side by side with nanotubes ͓Fig. 2͑b͔͒.
Little is known about the role of the metal clusters. In
grow side by side depending on the nature of the nucleation
center. Metallic clusters give rise to the growth of nanotube
films and diamonds or diamond precursors at the surface lead
to CVD diamond growth. These findings were already re-
ported by other groups.15 It seems that the only stable carbon
structure growing at typical diamond CVD conditions en-
countered in a microwave reactor are diamond and nano-
tubes. As for the growth of diamond, the high atomic hydro-
gen concentration is certainly a key parameter. It induces the
plasma chemistry, saturates the dangling bonds during the
growth process and etches nondiamond/nanotube phases.
This is a rather important advantage of this growth process.
The films consists of pure nanotube without any carbon
nanophase impurities. Hence, these films are well suited for
many applications and do not need any subsequent purifica-
tion step. TEM investigations at these films revealed that
open and closed nanotubes are formed ͓Fig. 2͑a͔͒. This might
be due to the high concentration of atomic hydrogen in the
gas phase, where deposition and etching of the nanotube film
many models the metal cluster is considered to be a seed,
setting the outer radius of the filament.1
6,17
However, some
models assume that carbon diffuses through the metal seed
and condenses at the bottom hence, pushing up the seed.18 In
contrast to results presented in Ref. 15 where metal particles
were found at the top of the growing nanotubes, we could
find metal cluster as seeds only. It seems that in our growth
process the size of the metal cluster can be substantially
larger as the diameter of the nanotube growing on its surface.
This is clearly illustrated by looking at Fig. 1͑a͒ where metal
islands in the ␮m range can be observed.
These nanotube films are well suited for electron field
emission applications. First of all they can be grown on large
surfaces, are phase pure, without large clusters of carbon,
and adhere well to the substrate. The electron emission is
governed by the tangle of nanotubes where statistically some
nanotubes protrude out of the film. Hence, it is not necessary
to align the nanotube after the deposition process as it was
19
done by other groups. The field emission measurements
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
are in a delicate balance. So far, we could only observe mul-
were performed by taking current–voltage (I–V) curves at
160.36.178.25 On: Sat, 20 Dec 2014 02:22:26

Appl. Phys. Lett., Vol. 73, No. 15, 12 October 1998
K u¨ ttel et al.
2115
FIG. 3. I–V measurements on a nanotube film with the inset being the
Fowler–Nordheim representation of the data. Emission starts at 2.5 V/␮m.
1
0Ϫ8 mbar with a Keithley 237 instrument. The detection
limit was 1 pA at maximum voltage (Ϯ1100 V͒. The emis-
sion current is collected by a highly polished steel sphere of
4
mm diameter, mounted on a linear piezo drive, with a step
size of 1 ␮m and 10 mm travel. The anode is grounded over
a 100 M⍀ ballast resistor. Alternatively, the emission could
be monitored with a phosphorus coated indium tin oxide
FIG. 4. Electron field emission as monitored by a phosphorous screen. ͑a͒ is
taken at 2.5 V/␮m, ͑b͒ at 3.1 V/␮m. The size of the image is 1.5ϫ3 cm .
2
͑
ITO͒ glass electrode in order to determine the emission site
The authors are indebted to R. Wessicken at ETH Z u¨ rich
for performing the TEM investigations. Part of this work was
supported by the Swiss National Science Foundation
density. In Fig. 3 we show I–V measurements with the inset
being the Fowler–Nordheim representation of the data. The
emission starts at a field of a 1.5 V/␮m. Using a work func-
tion of 5 eV one can deduce from the slope of the straight
line in the Fowler–Nordheim representation an aspect ratio
of approx. 800–1000 which is further confirmed by HRSEM.
The mechanism of field emission is clearly governed by the
field enhancement at the apex of the nanotubes. A more rig-
orous investigation of the work function of nanotubes, using
energy resolved field emission, gives a value of 5.3 eV.20 In
Fig. 4 the emission was monitored on a phosphorus screen.
At a field of 2.5 V/␮m the emission site density is approx.
͑NFP36͒ and the Swiss Priority Program of Materials ͑PPM͒.
1
2
S. Iijima, Nature ͑London͒ 354, 56 ͑1991͒.
S. Iijima, T. W. Ebbesen, and P. M. Ajayan, Nature ͑London͒ 358, 220
͑
1992͒.
3
4
M. Endo, K. Takeuchi, S. Igarashi, K. Kobori, M. Shiraishi, and H. W.
Kroto, J. Phys. Chem. Solids 54, 1841 ͑1993͒.
Carbon Nanotubes, edited by Th. W. Ebbesen ͑Chemical Rubber, Boca
Raton, 1997͒; Carbon Nanotubes, edited by M. Endo, S. Iijima, and M. S.
Dresselhaus ͑Pergamon, New York, 1996͒.
5
Graphite Fibers and Filaments, edited by M. S. Dresselhaus, G. Dressel-
haus, I. L. Spain, and H. A. Goldberg ͑Springer, New York, 1988͒.
S. Iijima and T. Ichihashi, Nature ͑London͒ 363, 603 ͑1993͒.
D. S. Bethune, C. H. Kiang, M. S. de Vires, G. Gorman, R. Savoy, J.
Vasquez, and R. Beyers, Nature ͑London͒ 363, 605 ͑1993͒.
J. M. Lambert, P. M. Ajayan, P. Bernier, J. M. Planeix, V. Brotons, B.
Coq, and J. Castaing, Chem. Phys. Lett. 226, 264 ͑1994͒.
S. Subramoney, R. S. Ruoff, D. C. Lorents, and R. Malhotra, Nature
͑London͒ 366, 637 ͑1993͒.
S. Seraphin and D. Zhou, Appl. Phys. Lett. 65, 1593 ͑1994͒.
T. W. Ebbesen, P. M. Ajayan, H. Hiuram, and K. Tanigaki, Nature ͑Lon-
don͒ 367, 519 ͑1994͒.
H. Hiura, T. W. Ebbesen, and K. Tanigaki, Adv. Mater. 7, 275 ͑1995͒.
F. Ikazaki, S. Ohsima, K. Uchida, Y. Kuriki, H. Hayakawa, M. Yumura,
K. Takahashi, and K. Tojima, Carbon 32, 1539 ͑1994͒.
3
2
1
0 /cm and increases to its maximum at roughly 3 V/␮m
6
7
when the whole screen is illuminated, making it even impos-
4
2
sible to determine an emission site density ͑about 10 /cm ).
In this letter we presented the deposition of nanotube
films onto silicon substrates by a microwave plasma from a
CH /H gas mixture under similar conditions than what is
used for the growth of CVD diamond films. The film quality
is excellent with regards to the purity of the film as well as
the adhesion to the substrate. These films are very well suited
as material for cold cathodes. They do not need any further
post treatment and by masking the deposition of the metal
catalyst we have succeeded in structuring the films. This is
very important for technical applications.
We also have used a hot filament reactor to deposit
nanotube films onto two inch silicon wafers. The advantage
using a hot filament reactor lies in the fact that atomic hy-
drogen is less abundant and hence, we could not observe an
etching of the nanotube films. Films on two inch silicon sub-
strates are black in appearance, adhere very well to the sub-
8
9
4
2
1
1
0
1
12
13
14
E. Maillard-Schaller, O. M. K u¨ ttel, P. Gr o¨ ning, O. Gr o¨ ning, R. Agostino,
and L. Schlapbach, Phys. Rev. B 55, 15 895 ͑1997͒.
15
R. Stockel, K. Janischowsky, M. Stammler, L. Ley, M. Albrecht, and H.
P. Strunk, Diamond Relat. Mater. 7, 147 ͑1998͒.
¨
16
S. Amelinckx, X. B. Zhang, D. Bernaerts, X. F. Zhang, V. Ivanov, and J.
B. Nagy, Science 265, 635 ͑1994͒.
G. G. Tibbetts, J. Cryst. Growth 66, 632 ͑1984͒.
R. E. Smalley, Mater. Sci. Eng., B 19, 1 ͑1993͒.
W. A. de Heer, A. Ch aˆ telain, and D. Ugarte, Science 270, 1179 ͑1995͒.
O. Gr o¨ ning, O. M. K u¨ ttel, Ch. Emmenegger, P. Gr o¨ ning, and L. Schlap-
17
1
1
2
8
9
0
bach ͑unpublished͒.
In collaboration with CSEM in Neuenburg, Switzerland, Detailed results
21
strate, and are very uniform. These films are very suitable
21
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
for applications as cold electron sources. 1
will be published elsewhere.
60.36.178.25 On: Sat, 20 Dec 2014 02:22:26