Characterization of bias-controlled carbon nanotubes

(1)

Characterization of bias-controlled carbon nanotubes

C.L. Tsai*, C.F. Chen

Department of Materials Science and Engineering, National Chiao Tung University, 1001 Ta Hsueh Road, Hsinchu 30050, Taiwan, ROC Received 26 August 2002; received in revised form 17 June 2003; accepted 25 June 2003

Abstract

In this study, we focus on the immediately improving quality of growing carbon nanotubes without any pre- or post-treatment. The applied biases during the reaction can directly control the diameter and the quality of carbon nanotubes.This simple step skips additional treatments and is easily used in many deposition systems.The diameter of carbon nanotubes noticeably varies from 45 nm without any amorphous carbon(under q80 V) to 120 nm (under y120 V).Raman spectrums indicate that I yID G

ratio decreases with increasing positive bias.This implies applying positive bias could enhance the graphitization of carbon nanotubes.However, positive and negative bias effects slightly vary the field emission enhancement.In addition, carbon nanotubes grown under positive bias possess better field emission characterization.This results from the following reasons: (I) smaller diameter;(II) pure surface; (III) more graphitized structure; and (IV) higher field enhancement b.

Keywords: Bias-effect; Carbon nanotubes; Raman spectrum; Scanning electron microscopy

1. Introduction

Carbon nanotubes _{(CNTs), since their first discovery} in 1991 w1x, have been considered for many different applications.Their small dimension, strength and the remarkable physical properties of these materials make them the most promising emitters for field emission devices.Due to the size-effect and structure diversity of nanomaterials, the physical properties strongly depend on their atomic-size structure, size and chemistry w2x. Different diameter, length, chirality of CNTs gives rise to diverse physical and mechanical properties.Before growing CNTs, the pre-treatment of catalyst to synthe-size higher density and smaller diameter of CNTs is needed.These methods include varying the morphology; thickness; distribution and size of as-deposited catalyst

w3–5x to reduce the grain size of catalyst.Besides, the

post-treatment to purify the CNTs is usually used to remove the amorphous carbonaceous impurities on the surface of CNTs w6x.

In this study, immediately improving the quality of CNTs during deposition is the main topic.Compared to

*Corresponding author.Tel.: q886-935889532; fax: q886-9-43020734.

E-mail address: [email protected](C.L. Tsai).

the aforementioned treatment; applying bias voltages during growth in situ reforms the growing CNTs directly. Furthermore, both positive and negative bias effects on the characterization of growing CNTs have been com-pletely discussed.

2. Experiment

The 150 nm Palladium (Pd) catalysts was deposited

on Si by using electron beam evaporation.The reactive gas mixture was CH4yH with a flow rate of 10y402 sccm.The applied microwave power and pressure during the growth of CNTs were 400 W and 15 Torr, respec-tively.An optical pyrometer was used to monitor the substrate temperature, that was maintained at approxi-mately 700 8C.Biasing system was conducted with upper and lower electrodes made of Mo with a distance of 1 cm.Samples were placed on the Mo holder, which is attached to the lower grounded electrode.Various biases were applied on the sample, ramped from y160 V to q160 V with the interval of 40 V.The growth time lasted for 10 min.Table 1 clearly lists the growth conditions.

(2)

Table 1

Deposition conditions of CNTs grown under various biases

Sample aApplied bias(V) b_Current_(mA)

A y160 16;20 B y120 12;13 C y80 8;10 D y40 3;5 E 0 0 F q₄₀ _2;4 G q₈₀ _6;7 H q₁₂₀ _11;13 I q₁₆₀ _15;18

Flow rate of reactive gases: H2yCH s40y10, Microwave power:4 400 W, Substrate temperature: ;700 8C, Growth time: 10 min, Pres-sure: 15 Torr.

means the bias applied to the lower electrode as compared with a

the upper electrode.Thus, conditions A;D present the samples grown under negative biases.

means the current measured from upper to lower lower electrodes. b

Fig.1.SEM photographs of CNTs grown under(a) y120 V; (b) y80 V; and (c) 0 V.

3. Results and discussion

3.1. Scanning electron microscope(SEM)

Fig.1 and Fig.2 present the scanning electron microscope (SEM) pictures of CNTs grown under

vari-ous biases.The photograph on the right of each figure is the enlarged image.It clearly displays that the morphology of CNTs changes with various applied biases.According to our SEM results, the relationship between diameter and different bias is plotted in Fig.3. Each data is the average value of five samples.Obvi-ously, the diameter of CNTs increases with increasing negative bias.Fig.1 a indicates that CNTs will grow to a sub-micrometer diameter under a higher bias _(more negatives than y120 V_{), revealing that negative biases} drastically enhance the growth of CNTs on Pd films.

However, Fig.2 shows the reverse trend toward increasing positive biases.The diameter of CNTs can

(3)

Fig.2.SEM photographs of CNTs grown under(a) q40 V; (b) q80 V; and (c) q120 V.

Fig.3.Diameter of CNTs as a function of applied biases.

be reduced from 65 nm (under 0 V) to approximately

45 nm (under q80 V).Although a positive bias effect

decreases the diameter of CNTs, the change of diameter slightly varies under a higher positive bias (more

posi-tives than q120 V).Except for the various biases

subjected on the samples, it is worth emphasizing that all samples grown under the same conditions, such as thickness of Pd, microwave power and growing time, and so on.Thus, the change of diameter of CNTs result only from different bias during growth.

Unlike the crooked CNTs grown under negative bias, straighter CNTs are easily found in a positive bias condition.Besides, the surface of CNTs grown under positive bias is pure without amorphous carbon (a-C).

It is believed that more ions bombard the sample in the plasma under negative bias.Meanwhile, the accelerated carbon species will be accumulated on the sample.This causes the higher growth rate of CNTs with larger diameter in negative bias condition than in positive one.

Many reports have presented the method to enhance the nucleation density of diamond by applying negative bias

w7–9x.However, under positive bias condition, more

(4)

Fig.4.TEM images of CNTs grown under(a) positive and (b) negative biases.

Thus, it makes the diameter decrease and the surface pure without a-C.

3.2. Transmission electron microscope(TEM)

Fig.4a and b exhibit the transmission electron micro-scope(TEM) of CNTs grown under negative and

posi-tive biases.TEM images give two same shapes (I)

fishbone-like tube and (II) tube filled with catalysts.

Although the morphology of CNTs differs from various biases, there is no difference on the nanostructure. Besides, the CNTs grown under 0 V also have two same shapes, indicating that bias effect does not vary the nanostructure of the CNTs.The dark part shown by an arrow is identified as Pd by energy dispersive spectrom-eters_(EDX).

3.3. Raman spectra

Fig.5a exhibits the Raman spectra of CNTs grown under various biases.All of them have two sharp peaks located at approximately 1345 cmy1 and 1580 cmy1,

respectively.The peaks imply that CNTs are character-istic of graphite.The first-order Raman spectrum of CNTs shows strong sharp peaks at 1581 cmy1_{(G line),} which is the high-frequency E g first-order mode and2 1350 cmy1 _{(roughly corresponding to the D-line} asso-ciated with disorder-allowed zone-edge modes of graph-ite_{).The 1350 cm}y1 band is normally explained by relaxation of the wave vector selection rule, due to the effect of the finite size of the crystal in the material

w10,11x.Normally, the intensity of the 1350 cmy1 peak increases(i) with an increase in the amount of

unorgan-ized carbon in the samples and (ii) with a decrease in

the graphite crystal size.In Fig.5b, it is found that the I yID G ratio decreases with increasing positive biases. Positive bias process makes CNTs possess lower I yID G ratio, implying the more graphitized structures, thereby improving their properties.This phenomenon is corre-sponding to the previous SEM results.

3.4. I–V characterization

The field emission tests are performed on a diode structure, in which CNTs are separated from the anode,

(5)

Fig.6.The I–V curve and an insert of F-N plots of CNTs grown under(a) q120 V, (b) 0 V and (c) y120 V, respectively.

Fig.5.(a) Raman spectra of CNTs grown under various biases and (b) the I yI ratio as a function of applied biases.D G

indium-tin-oxide glass, using 500 mm glass as spacers. The emission current_{(I) is then measured as a function} of anode-to-cathode voltage in a vacuum of 1=10y6 Torr.The Fowler–Nordheim theory w12x is the most commonly used model for the emission of cold electrons from a metal under a strong applied field.The total current as a function of the local field at the emitter surface F is approximately given by IA(F y2 f)

exp_(Bf3y2yF), with Bs6.83=10 , and7 f the work function in electron volt. F is ususlly taken as FsbEs

bVyd, where V is the applied voltages, d the distance between cathode and anode, b the field enhancement,

and EsVyd the macroscopic field.When Ln (IyV ) is2

plotted vs.1yV(F–N plot), one should obtain a straight

line with a slope that depend on b, f and d.Fig.6a and b displays the I–V curve and F–N plot of CNTs grown under y120, 0 and q120 V.We further estimate the field enhancement b from the constant F–N slope in the low current regime with ds500 mm and fs₅

eV as for graphite and C60 w13x. b is obtained for approximately 2590, 2690 and 2770 corresponding to

y_{120 V, 0 V and q120 V, respectively.It seems that}

positive and negative bias effects slightly vary the field

emission enhancement.The difference may be resulted from the various diameters, geometry and the graphitized structure of CNTs.According to I–V result, it clearly exhibits that CNTs grown under positive bias possess the best field emission property, that is, a turn-on field of 1.38 Vymm and 655 mAycm under 2.2 Vymm than2 those grown under zero and negative bias.It is attributed to the following reasons such as: (I) smaller diameter; (II) pure surface, (III) more graphitized structure of

CNTs; and(IV) higher field enhancement b. Acknowledgments

The authors would like to thank the National Science Council of the Republic of China for financially sup-porting this research under Contract No.NSC 91-2216-E-009-030.

References

w1x S.Iijima, Nature(London) 354 (1991) 56.

w2x Z.L. Wang (Ed.), Characterization of Nanophase Materials,

(6)

w3x Y.C. Choi, Y.M. Shin, Y.H. Lee, B.S. Lee, G.-S. Park, W.B.

Choi, et al., Appl. Phys. Lett 76(2000) 2367.

w4x M.Yudasaka, R.Kikuchi, T.Matsui, Y.Ohki, S.Yoshimura,

E.Ota, Appl.Phys.Lett.2477(1995) 67.

w5x C.Bower, O.Zhou, W.Zhu, D.J.Werder, S.Jin, Appl.Phys.

Lett.2767(2000) 77.

w6x U.Kim, R.Pcionek, D.M.Aslam, D.Tomanek, Diamond Relat.Mater.1947(2001) 10.

w7x S.Yugo, T.Kanai, T.Kimura, T.Muto, Appl.Phys.Lett 1036 (1991) 58.

w8x J.T.Huang, W.Y.Yeh, J.Hwang, H.Chang, Thin Solid Films

35(1998) 315.

w9x R.Stockel, K.Janischowsky, S.Rohmfeld, J.Ristein, M.¨ Hundhausen, L.Ley, Diamond Relat.Mater.321(1996) 5. w10x F.Tuinstra, J.L.Koenig, J.Chem.Phys.1126(1970) 53. w11x R.J. Nemanich, S.A. Solin, Phys. Rev. B 392(1979) 20. w12x R.H. Fowler, L.W. Nordheim, Proc. Royal. Soc. London A 173

(1928) 119.