Ž .Surface and Coatings Technology 131 2000 93]97

Growth characteristics of carbon nanotubes by plasmaenhanced hot filament chemical vapor deposition

Jae-hee Hana,U, Byung-Sik Moonb, Won Suk Yanga, Ji-Beom Yooa,Chong-Yun Parkb

aDepartment of Materials Engineering, Sungkyunkwan Uni ersity, 300, Chunchun-dong, Jangan-gu, Suwon, South Korea 440-746bDepartment of Physics, Sungkyunkwan Uni ersity, 300, Chunchun-dong, Jangan-gu, Suwon, South Korea 440-746

Abstract

We successfully obtained vertically aligned multiwalled carbon nanotubes on different nickel-coated substrates by plasmaenhanced hot filament chemical vapor deposition at low temperatures below 6508C. Acetylene and ammonia gas were used as thecarbon source and a catalyst. The surface roughness of the nickel layer increased as NH etching time increased. The diameters3of the nanotubes decreased and the density of nanotubes increased as NH etching time increased. The diameter of the3nanotubes was 30]70 nm. A nickel cap was observed on the top of the grown nanotube and a very thin amorphous-carbon-likelayer was found on the nickel cap. The morphology and microstructure of carbon nanotubes were measured using scanningelectron microscopy and transmission electron microscopy. Q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Carbon nanotube; Plasma enhanced hot filament vapor phase deposition; Nickel layer; FED

1. Introduction

Since the first discovery of carbon nanotubes in 1991w x1 , there has been increasing interest in the carbonnanotube because of its unique electrical, chemical andmechanical properties and many potential applications.Therefore, numerous papers have been reported on thegrowth characteristics and the properties of carbon

w xnanotubes, such as growth mechanisms 2]5 , structure,w xalignment 6]9 , and electron emission properties

w x10]13 . Carbon nanotubes are an especially promisingcandidate for cold cathode field emitters because oftheir unique electrical properties, their high aspectratios and small radii of curvature at their tips. Variousstudies on electron field emission have been reportedincluding electron emission from a single carbon nan-

w x w xotube 14,15 and a carbon nanotube matrix 10 . These

U Corresponding author.

results show that carbon nanotube films are extremelysuitable as field emitters for field emitting devices. Forapplications such as flat panel displays and vacuummicroelectronics, large area films of nanotubes ornanofibers producing uniform field emissions are re-quired. Particularly, alignment of the carbon nanotubesis important to both fundamental studies and applica-tions.

Recently, aligned carbon nanotubes have been grownabove 7008C on mesoporous silica embedded with ironnanoparticles by thermal decomposition of acetylene

w xgas in nitrogen gas 6 . The growth below 6608C ofwell-aligned carbon nanotubes on nickel-coated glassover areas of up to several square centimeters has beenreported using plasma enhanced hot filament chemical

Ž .vapor deposition PEHFCVD with a gas mixture ofw xammonia and acetylene 16 .

In this study, we report the growth of verticallyaligned, multiwalled carbon nanotubes on differentnickel-coated substrates over large areas at low tem-

0257-8972r00r$ - see front matter Q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S 0 2 5 7 - 8 9 7 2 0 0 0 0 7 6 6 - 0

( )J. Han et al. r Surface and Coatings Technology 131 2000 93]9794

Ž .peratures -6508C using PEHFCVD. We investigatethe effect of growth parameters on the growth charac-teristics of carbon nanotubes.

2. Experiments

The experiments were carried out in a PEHFCVDsystem, which is schematically shown in Fig. 1. Acety-

Ž .lene C H gas was used as the carbon source and2 2Ž .ammonia NH gas was used as a catalyst and dilution3

gas. The flow rate of acetylene was 80 sccm and that ofammonia was 160 sccm. A dc plasma was employed inthis work to deposit vertically aligned carbon nan-otubes. A thin nickel layer was deposited on glass with

Ž . Ž .and without indium tin oxide ITO and silicon 100 bydc magnetron sputtering. The nickel film was depositedunder the condition that the base pressure of a cham-ber was below 7=10y5 Torr, the operating pressurewas 5=10y3 Torr, the operating current density was6.58=10y3 Arcm2, the flow rate of argon gas was 10sccm, and the substrate temperature was room temper-

˚ature. The deposition rate of nickel film was 200 Armin.Prior to carbon nanotube growth, the substrate wascleaned in TCE, acetone, methanol for 10 min, andrinsed in de-ionized water to remove organic contami-nant. The substrate was transferred to a vacuum cham-ber and pumped down below 2=10y5 Torr by a me-chanical and a diffusion pump. After the chamberpressure reached 2=10y5 Torr, NH was introduced3into the chamber in order to maintain a working pres-sure of 1]5 Torr. After the working pressure had beenstabilized, the power to the tungsten filament coil andto the dc power supply were turned on to generate heatand plasma. The bias voltage for plasma generation

Žwas kept at 480]600 V. The pre-treatment surface.etching was conducted by NH plasma for 1]5 min.3

Then C H was introduced into the chamber for the2 2growth of carbon nanotubes.

The grown carbon nanotubes were characterized byŽ .scanning electron microscopy SEM to investigate the

effect of growth parameters on their morphology. Tran-Ž .smission electron microscopy TEM was used for mi-

crostructure analysis of nanotubes.

3. Results and discussion

In order to study the initial growth of carbon nano-tubes, the growth of carbon nanotubes were carried outfor 2]5 min. Prior to the growth of carbon nano-tubes, surface etching was performed using NH for 43min. The pressure of the chamber was 3.2 Torr. Theflow rate of NH was 160 sccm, the filament current3was 16 A, and the plasma power was 522 V and 0.11 A.After nickel etching with NH , C H was introduced3 2 2

ŽFig. 1. Schematic diagram of PEHFCVD plasma enhanced hot.filament chemical vapor deposition .

for carbon nanotubes growth. The flow rate of C H2 2was 80 sccm. The plasma power during growth was 563V and 0.18 A. As shown in Fig. 2, uniform carbonparticles were formed on the substrate and a size ofcarbon particles increased as the growth time in-creased. It seems that the nuclei do not agglomeratewith each other. It implies that the density and dis-tribution of carbon nanotubes are determined at theinitial stage of growth.

The effect of NH etching of nickel layer on the3growth characteristic of carbon nanotubes was investi-gated. Fig. 3 shows a surface morphology of nickellayer etched by NH for different time. The etching3was performed under the condition that the workingpressure was 1.4 Torr, the flow rate of NH was 1603sccm, the filament current was 9 A, the plasma bias

Fig. 2. Initial growth of carbon nanotubes with different growth.

( )J. Han et al. r Surface and Coatings Technology 131 2000 93]97 95

Fig. 3. Effect of etching time with NH on the surface of the Ni3layer.

voltage and current were 460 V and 0.01 A, respec-Ž .tively. A surface scratch Fig. 3 in 1 min, upper left is

due to the peeling operation with tweezers. As shownin Fig. 3, the surface roughness of the nickel layerincreased as etching time increased.

We studied the effect of surface etching on thegrowth characteristic of carbon nanotubes. The growthtime was 14 min. The rest of the growth conditionswere the same as conditions previously used. As shownin Fig. 4, well-aligned carbon nanotubes were obtained.The diameters of the nanotubes decreased and the

density of nanotubes increased as NH etching time3increased. This result may be due to the fact that thenucleation of nanotubes on the nickel surface in-creased as NH etching time increased.3

Carbon nanotubes were grown on nickel-coated sub-Ž .strates such as silicon 100 - and ITO-coated glass. The

effect of substrates on the growth characteristic ofnanotubes was investigated and was shown in Fig. 5.The growth conditions of carbon nanotubes were sameas the condition mentioned above. As shown in Fig. 5,the diameter of the grown nanotubes on nickel-coatedglass was nearly same as that of nanotube on thesilicon. In case of nickel-coated glass with ITO, adhe-sion between the ITO layer and the nickel layer causeda problem. The ITO layer was peeled off. However, thecarbon nanotubes were also found in the separatednickel layer that remained on the substrate. This indi-cates that the growth of carbon nanotubes can beobserved on any substrate if a nickel layer exists andthe growth conditions are satisfied.

Ž .Fig. 6 presents TEM 300 kV, H-9000NA cross-sec-tional views showing the interior and wall structures ofthe carbon nanotubes grown on nickel-coated glasssubstrate. The outer diameter of carbon nanotube isalmost 70 nm, and the inner diameter of tube is ap-proximately 7.5 nm. It shows that the nanotube is amultiwalled tube that has a central hollow. The fringesin a multiwall are individual cylindrical graphitic layers.In the lower-right image of Fig. 6, the dashed line isnearly 7.5 nm long and this has 22 graphitic walls. Thisis clearly consistent with the fact that the graphite

Fig. 4. Variation of diameter and density of carbon nanotubes with etching time.

( )J. Han et al. r Surface and Coatings Technology 131 2000 93]9796

Ž . Ž . Ž .Fig. 5. Effects of substrate on the growth of carbon nanotube: upper left glass; upper right Si; and lower glass coated with ITO.

w xspacing is Cr2s0.34 nm 5 . The graphitic walls run-ning across the diameter of the nanotube looks likehaving some cut-off-like defects. The lighter contrastpart in the lower right of Fig. 6 indicates the inside ofthe nanotube that has the hollow structure. A nickelcap was observed at the end of each carbon nanotubesas shown in the upper left image of Fig. 6. A reason tocall it nickel is that lattice fringes at the tip of thecarbon nanotube are confirmed to be those of metal by

Žusing the other very high resolution TEM 1250 kV,. Ž .JEM-ARM1250 fringes image is not shown here and

that the carbon nanotubes have been known to be

grown with the catalytic transition metals, such as Ni,w xCo, Fe, etc., in other researches 2]5,16 . The shape of

the nickel cap is like a long ellipsoid with a sharp end.The nanotube has a narrow appearance at the end ofthe nanotube. As shown in the upper-right image ofFig. 6, the diameter of the nickel cap is approximately17 nm and 5 nm thickness of an amorphous-carbon-likelayer was deposited on the nickel cap. Because itsnon-graphitic lattice image unlike the images of multi-wall structure shown in the lower left or right image ofFig. 6, we could suggest that it is an amorphous-carbon-like layer.

Fig. 6. TEM images of carbon nanotubes grown on Ni-coated glass layer

( )J. Han et al. r Surface and Coatings Technology 131 2000 93]97 97

4. Conclusion

Vertically aligned carbon nanotubes on nickel-coatedglass substrates were obtained at low temperaturesbelow 6508C by plasma enhanced hot filament chemicalvapor deposition using acetylene gas as the carbonsource and ammonia gas as the dilution gas and cata-lyst. The surface roughness of the nickel layer in-creased as NH etching time increased. The diameters3of the nanotubes decreased and the density of nan-otubes increased as NH etching time increased. The3growth characteristics of carbon nanotubes are depen-dent on the under layer, such as glass, ITO and silicon.A nickel cap was observed at the top of the grownnanotube. A very thin amorphous-carbon-like layer wasfound on the nickel cap from transmission electronmicroscopy. Highly oriented carbon nanotubes ob-tained in this work can be applied to the field emission

Ž .display FED .

References

w x Ž . Ž .1 S. Iijima, Nature London 354 1991 56.

w x2 T.W. Ebbesen, Carbon Nanotubes: Preparation and Properties,Chemical Rubber, Boca Raton, 1997.

w x Ž .3 J.C. Charlier, A. De Vita, X. Blase, R. Car, Science 275 1997´646.

w x4 M. Yudasaka, R. Kikuchi, Y. Ohki, E. Ota, S. Yoshimura,Ž .Appl. Phys. Lett. 70 1817 1997.

w x5 S. Amelinckx, X.B. Zhang, D. Bernaerts, X.F. Zhang, V. Ivanov,Ž .J.B. Nagy, Science 265 1994 635.

w x6 W.Z. Li, S.S. Xie, L.X. Qian, B.H. Chang, B.S. Zou, W.Y.Ž .Zhou, R.A. Zhao, G. Wang, Science 274 1996 1701.

w x7 P.M. Ajayan, O. Stephan, C. Colliex, D. Trauth, Science 265Ž .1994 1212.

w x Ž .8 W.A. de Heer et al., Science 268 1995 845.w x Ž . Ž .9 M. Terrones et al., Nature London 388 1997 52.

w x Ž .10 P.G. Collins, A. Zettl, Phys. Rev. B 55 1997 9391.w x Ž .11 H. Schmid, H.W. Fink, Appl. Phys. Lett. 70 1997 2679.w x12 A.G. Rinzler, J.H. Hafner, P. Nikolaev, L. Lou, S.G. Kim, D.

Tomanek, P. Nordlander, D. Colbert, R.E. Smalley, ScienceŽ .269 1995 1550.

w x Ž .13 W.A. de Heer, A. Chatelain, D. Ugarte, Science 270 19951179.

w x14 J.-M. Bonard, J.-P. Salvetat, T. Stockli, W.A. de Heer, L. Forro,¨ ´Ž .A. Chatelain, Appl. Phys. Lett. 73 1998 918.ˆ

w x15 Y. Saito, K. Hamaguchi, T. Nishino, K. Hata, K. Tohji, A.Ž .Kasuya, Y. Nishina, Jpn. J. Appl. Phys. Part 2 36 1997 L1340.

w x16 Z.F. Ren, Z.P. Huang, J.W. Xu, J.H. Wang, P. Bush, M.P.Ž .Siegal, P.N. Provencio, Science 282 1998 1105.