Electron field emission from fluorinated amorphous
carbon nanoparticles on porous alumina
S.H. Lai
a, K.P. Huang
b, Y.M. Pan
a, Y.L. Chen
a,
L.H. Chan
a, P. Lin
b, H.C. Shih
a,*a
Department of Materials Science and Engineering, National Tsing Hua University, 101, Sec. 2, Kuang Fu Road, Hsinchu, Taiwan 300, ROC
b
Department of Materials Science and Engineering, National Chiao Tung University, 1001 Ta Hsueh Road, Hsinchu, Taiwan 300, ROC Received 27 August 2003; in final form 16 October 2003
Published online:
Abstract
The fluorinated amorphous carbon nanoparticles (a-C:F NPs) films have been prepared on a porous alumina at a lower temperature of 100 °C by an ECR-CVD system with a mixture of CF4, C2H2 and Ar as precursors. The morphology of a-C:F NPs films was verified by FESEM. The microstructure and chemical bonding nature of the a-C:F NPs films were investigated by Raman spectroscopy and XPS, respectively. The fluorinated samples need a lower field (2.25 V/lm for 10.6% F a-C:F NPs films) to emit the electron than unfluorinated ones. The excellent field-emission properties of a-C:F NPs films are due to the effect of the shape, fluorine doping and dangling bonds.
Ó 2003 Elsevier B.V. All rights reserved.
1. Introduction
Carbon-based materials are one of the most well-known materials to human life and they are available with a variety of properties. Various car-bon-based materials such as chemical vapor de-posited (CVD) diamond [1], diamond-like carbon (DLC) films [2] and carbon nanotubes (CNTs) [3,4] have proven to be very interesting electron-emitting cathodes because of their excellent electron
field-emission properties. The CNTs seem to be the best field emitters among all carbon-based materials owing to their high aspect ratios, high chemical stability, and high mechanic strength. It is believed that the lower work function (compared with Si, Mo, and W materials) and field-enhancement effect are the main reasons for the low turn-on voltage and high current density of the CNTs. Compared to CNTs, carbon nanoparticles have their advantages, such as the ball shape and lower synthesis temperatures, for field-emission materials [5]. A porous alumina (PA) has recently attracted increasing attention as a key material for the fab-rication of nanodevices [6–8]. In this Letter, fluo-rinated amorphous carbon nanoparticles (a-C:F
www.elsevier.com/locate/cplett
*
Corresponding author. Fax: +886-35-710290. E-mail address:[email protected](H.C. Shih).
0009-2614/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2003.10.106
NPs) films have been prepared on a PA template at a lower temperature of 100 °C by an electron cyclotron resonance chemical vapor deposition (ECR-CVD) system and the field-emission prop-erties of the film also have been investigated.
2. Experimental
A commercially available membrane filter, 13 mm o.d. and 60 lm thick (Whatman Ltd., Anodisc 13), which consists of an array of tubular channels with a diameter of about 100 nm as a template. The ECR-CVD system used in this work has been described elsewhere [9,10]. The experi-mental conditions for the deposition of a-C:F NPs are as follows: a microwave power of 800 W, a total gas pressure of 5 103 Torr, CF
4 flow rate
of 5–20 sccm, C2H2 flow rate of 10–20 sccm, and
Ar flow rate of 10–20 sccm for 1 min under a r.f. bias of 100–200 V. The morphology of the sample was observed by field-emission scanning electron microscopy (FESEM; JEOL JSM-6300F). The chemical bonding of the a-C:F NPs films was in-vestigated by X-ray photoelectron spectroscopy (XPS Perkin–Elmer Model PHI1600 system) using a single Mg Ka (1253.6 eV) X-ray source operat-ing at 250 W. Energy calibration was made by reference to the Au 4f7=2 peak at 83.8 eV. The
channel width was 1.0 eV for survey scan spectra and 0.2 eV for core-level spectra. The film micro-structure was also analyzed by Raman spectros-copy (Ranishaw Ramanscope Model 2001), whose excitation was achieved by means of the 514.4 nm (2.41 eV) line of an argon ion (Arþ) laser
(Coherent Innovr Model 90). The electron field-emission characteristics of a-C:F NPs films were measured by the conventional diode method in the vacuum chamber with 105 Torr pressure. The
current density and the electric field (J –E) char-acteristics were measured using an electrometer (Keithley 237). The field-emission measurements were carried out using a parallel plate structure with the films being the cathode and an indium-tin-oxide (ITO) coated glass as the anode. The gap between the cathode and the anode was 200 lm by a glass fiber spacer. The results are reproducible over periods of weeks and the emission properties
of a-C:F NPs films do not degrade with the ex-posure to atmospheric conditions.
3. Results and discussion
The top surface of the PA template, directly exposed to the ECR plasma, was coated with a-C:F NPs films after completion of the experi-ment. Fig. 1(a) is a typical SEM top-view micro-graph of the deposited films, which shows the nanostructures of the a-C:F NPs. The large par-ticles are deposited on the interpore of PA and then the small particles are formed among the large particles. Although the size of the a-C:F NPs is not uniform (<100 nm), the small particles are distributed uniformly all over the films. Fig. 1(b) gives the cross-sectional view of the mono-layer a-C:F NPs on the PA template. The smaller particles were also observed in the tubular channels. It should be noted that the a-C:F NPs are irregular balls and the particle surface is ragged. Besides, the a-C:F NPs film was synthesized under several ex-perimental parameters, such as microwave power, gas flow rate, and a r.f. bias. We tried to increase the F content, but instead of nanoparticles, other morphologies were produced, i.e., the highest F content of the a-C:F NPs is 10.6 at.%. The dense, flat and covered films were formed on the porous alumina in most experimental parameters. Other morphologies are the nanowires in and the nano-porous films on the nano-porous alumina, and they will be discussed in another Letter.
The growth mechanism of a-C:F NPs film was proposed as shown in Figs. 1(c) and (d). Initially, the plasma stream of the ion fluxes is guided to the PA template by a r.f. bias and the a-C:F NPs are deposited on the interpore and nanochannels of PA template due to the agglomeration of CFx
radicals. As the deposited time increasing, a-C:F NPs are growing bigger and smaller a-C:F NPs are also produced among the bigger ones. Besides the growth process of a-C:F NPs, the etching process also proceed by the plasma at the same time. Fi-nally, a-C:F NPs film is deposited on the PA template.
Fig. 2 shows the C 1s and F 1s spectra of the a-C:F NPs film with three different fluorine
con-tents. The surface compositions (F/C) were deter-mined by the area ratio of C 1s and F 1s XPS spectra, corrected by the atomic sensitivity factors (ASF; C: 0.296 and F: 1). In the C 1s spectra, we observed the C–F and C–F2 binding at 288.7 and
291.2 eV, respectively, for the film with 10.6 at.% F. The peak observed at 288.7 eV are ascribed to sp3-hybridized carbon atoms with covalent C–F
bonds, which are similar to those in the covalent graphite fluoride compounds, (CF)nand (C2F)n. In
the F 1s spectra, the peaks are at 687 and 689 eV for the film with 4.2 and 10.6 at.% F, respectively. The peak at 687 eV is assigned to semi-ionic fluorine and 689 eV is close to covalent fluorine. It should be noted that higher CF4 and lower C2H2
flow rate lead to a relative increase of the intensity of C–Fx bonding when the fluorine content
in-creases, and the C 1s and F 1s peaks shift to the higher binding energy side and their intensities become stronger.
In addition to the chemical analysis performed by XPS, Raman spectroscopy was employed to
survey the C–C structural modifications. The Ra-man spectra present two partially overlapping bands characteristic of all a-C films. They usually referred as the D (disorder) and the G (graphite) bands, since they recall more or less close to the Raman spectrum of polycrystalline graphite [11] as well as other forms of non-crystalline graphitic materials [12]. Two Raman spectra obtained from the different fluorine content of the deposited films are shown in Fig. 3. The lower spectrum is typical of a diamond-like film, while the upper one, ob-tained from a film with 10.6 at.% F, presents a strong luminescence background, characteristic of a polymer-like material. The first band, centered at approximately 1350 cm1, corresponds to the
so-called D-band and is associated with disorder-allowed zone edge modes of graphite that become Raman active due to the lack of long-range order [11]. The second one, which peaked at approxi-mately 1580 cm1, is known as the G-band and is
attributed to E2g-symmetry optical modes
occur-ring at the Brillouin-zone center of crystalline
Fig. 1. FESEM image of the a-C:F NPs films, showing: (a) surface morphology and (b) the cross-sectional view. Schematic of the growth mechanism of a-C:F NPs films, (c) initial stage and (d) final stage.
graphite [11]. No other band was observed in our a-C:F NPs films.
The J –E curves and its Fowler–Nordheim (F–N) plots for the a-C:F NPs films are shown in Fig. 4. Here, we define the turn-on field (Eto) as the
field needed to produce an emission current density of 10 lA/cm2. The turn-on fields of the a-C:F NPs
films are 3.45 V/lm (0% F), 2.75 V/lm (4.2% F) and 2.25 V/lm (10.6% F). The turn-on field of the fluorinated samples is lower than that of the un-fluorinated one. For the conventional flat-panel
displays, a current density of 1 mA/cm2is usually
required. The threshold fields (Eth) producing a
current density of 1 mA/cm2 are 7 V/lm (0% F),
5.05 V/lm (4.2% F) and 4.55 V/lm (10.6% F) for the a-C:F NPs films. From the results, the unfluo-rinated samples need a higher field to emit the electron than fluorinated ones. It should be noted that the emission current density is of the order of 0.5 mA/cm2 without large fluctuations for a test
period of 4 h. Then the samples were taken out to observe the changes of the morphology. The mor-phology of a-C:F NPs films have no apparent
Fig. 2. C (1s) and F (1s) core-level spectra of a-C:F NPs films with different fluorine contents.
Fig. 3. Raman spectra of a-C:F NPs films.
Fig. 4. The J –E curves and its Fowler–Nordheim (F–N) plots of the a-C:F NPs films.
changes and are not destroyed after the electron field-emission measurements and the field-emission properties of the films are excellent compared with a-C [2,13] and CNTs [14,15].
The reason why a lower field is required to emit electrons from a-C:F NPs films is explained as follows: (1) The shape effect: the end of a nanotube is a cap consisting of pentagons and hexagons. The curve position of the nanotube consists of penta-gons, hexagons and heptagons. The pentagons and heptagons are considered defects as compared with the hexagons. It is known that the defects are the possible emission sites. It is observed from the SEM images that the a-C:F NPs are irregular spheres and the surface of which is ragged. The nanoparticle has more defects than normal, straight nanotube in a unit area, i.e., the nano-particle has more possible emission sites than the nanotube. Therefore, we attributed the low turn-on field of a-C:F NPs films to the shape effect [5]. (2) The donor effect: the nitrogenation of a-C:H has been investigated and the nitrogen acts an n-type dopant in a-C [16]. A model based on the a-C:H:N (nitrogen containing hydrogenated amorphous carbon) film acting as a Ôspace charge interlayerÕ with a lower electron affinity is pro-posed to explain the emission at lower electric fields than a-C:H [2]. The fluorinated carbon film has more electrons can be emitted than the nitro-genated one when the electric field is applied, be-cause the electronegativity of F atom is higher than N atom. (3) The dangling bonding effect: by theoretical calculation, the open end of CNTs with dangling bonds is predicted to be the most favorable structure for field emission [17]. Rob-ertson [18] proposed an emission mechanism for diamond-like carbon films based on the non-uniform hydrogen termination on the film surface. The higher CF4 concentration, the greater is the
probability for the formation of the fluorinated film due to the agglomeration of CFxradicals [19],
but when only CF4 reactive gas, there is no
for-mation of the fluorinated film. Fluorine atoms can react with hydrogen atoms that are from the hydrocabon radicals on the surface, forming vol-atile HF and leaving dangling bonds on the film surface [20]. When the fluorine concentration of the a-C:F NPs films increases, more dangling
bonds are created on the film surface, which can influence the electronic properties of a-C:F NPs films. The dangling bonds are more reactive and the foreign atoms or molecules may be adsorbed when the samples are exposed to the atmosphere. By an extremely strong local field during the emission process, the attached atoms or molecules may be desorbed, and the dangling bonds may be exposed to contribute greatly to electron emission.
4. Conclusions
In this Letter, a-C:F NPs films with three dif-ferent fluorine contents were deposited on PA template by an ECR-CVD system. When the fluorine content of the a-C:F NPs films increases, the C 1s and F 1s peaks shift to the higher binding energy side and their intensities become stronger and the microstructure changes from diamond-like film to polymer-like one. From our results, a-C:F NPs films are excellent field emitters compared with a-C and CNTs. The excellent field-emission properties of a-C:F NPs films are due to the effect of the shape, fluorine doping and dangling bonds.
Acknowledgements
The authors acknowledge the financial support of this work by the National Science Council of the Republic of China under the contract of NSC 91-2120-E-007-006.
References
[1] M.W. Geis, J.C. Twichell, J. Macaulay, K. Okano, Appl. Phys. Lett. 67 (1995) 1328.
[2] G.A.J. Amaratunga, S.R.P. Silva, Appl. Phys. Lett. 68 (1996) 2529.
[3] A.G. Rinzler, J.H. Hafner, P. Nikolaev, L. Lou, S.G. Kim, D. Tomanek, P. Nordlander, D.T. Colbert, R.E. Smalley, Science 269 (1995) 1550.
[4] W.A. de Heer, A. Chatelain, D. Ugrate, Science 270 (1995) 1179.
[5] J. Yu, E.G. Wang, X.D. Bai, Appl. Phys. Lett. 78 (2001) 2226.
[6] C.A. Huber, T.E. Huber, M. Sadoqi, J.A. Lubin, S. Manalis, C.B. Prater, Science 263 (1994) 800.
[7] C.R. Martin, Science 266 (1994) 1961.
[8] Z.H. Yuan, H. Hung, H.Y. Dang, J.E. Cao, B.H. Hu, S.S. Fan, Appl. Phys. Lett. 78 (2001) 3127.
[9] S.H. Tsai, F.K. Chiang, T.G. Tsai, F.S. Shieu, H.C. Shih, Thin Solid Films 366 (2000) 11.
[10] X.W. Liu, S.H. Tsai, L.H. Lee, M.X. Yang, A.C.M. Yang, I.N. Lin, H.C. Shih, J. Vac. Sci. Technol. B 18 (2000) 1840. [11] F. Tuinstra, J.L. Koenig, J. Chem. Phys. 53 (1970) 1126. [12] D.S. Knight, W.B. White, J. Mater. Res. 4 (1989) 385. [13] S.R.P. Silva, G.A.J. Amaratunga, J.R. Barnes, Appl. Phys.
Lett. 71 (1997) 1477.
[14] P.G. Collins, A. Zettl, Appl. Phys. Lett. 69 (1996) 1969.
[15] X. Xu, G.R. Brandes, Appl. Phys. Lett. 74 (1999) 2549. [16] S.R.P. Silva, G.A.J. Amaratunga, Thin Solid Films 270
(1995) 194.
[17] S. Han, J. Ihm, Phys. Rev. B 61 (2000) 9986. [18] J. Robertson, Proc. IVMCÕ98 (1998) 162.
[19] W. Schwarzenbach, G. Cunge, J.P. Booth, J. Appl. Phys. 85 (1999) 7562.
[20] K. Teii, M. Hori, T. Goto, N. Ishii, J. Vac. Sci. Technol. A 18 (2000) 1.
