ISSN 10637850, Technical Physics Letters, 2014, Vol. 40, No. 12, pp. 1065–1068. © Pleiades Publishing, Ltd., 2014. Original Russian Text © A.V. Arkhipov, P.G. Gabdullin, N.M. Gnuchev, A.Yu. Emel’yanov, S.I. Krel’, 2014, published in Pis’ma v Zhurnal Tekhnicheskoi Fiziki, 2014, Vol. 40, No. 23, pp. 58–66.
LowVoltage Field Emission from Carbon Films Produced by Magnetron Sputtering A. V. Arkhipov*, P. G. Gabdullin, N. M. Gnuchev, A. Yu. Emel’yanov, and S. I. Krel’ St. Petersburg State Polytechnic University, St. Petersburg, Russia *email:
[email protected] Received February 21, 2014
Abstract—Emission properties of carbon films deposited on silicon substrates by magnetron sputtering have been studied. The structure of the films was varied by changing the substrate temperature. It was found that the best emission properties are obtained for a coating constituted by graphitized islands with transverse dimensions of 30–40 nm and a thickness of 3–4 nm. This result is in good agreement with the data previously obtained for films formed by chemical vapor deposition. This suggests that it is the structure of a carbon coat ing that determines its emission properties. A model of the emission mechanism for films of the type under study is discussed. DOI: 10.1134/S1063785014120037
Many nanocarbon materials and structures possess a capacity for lowvoltage field emission even when containing no morphological elements with a large geometrical aspect ratio, such as tips or ribs [1–5]. Nanoisland carbon films formed on silicon substrates by the chemical vapor deposition (CVD) method can
serve as an example of structures of this kind [6]. In the present study, we examined similar coatings produced by the simpler and more readily available method of magnetron sputtering. The carbon coatings were formed on a MAG30 installation (manufactured by ProtonMIET Co.) via
400°C
400 nm
500°C
400 nm
600°C
1 µm
700°C
1 µm
Fig. 1. ATM images of the surface of carbon films prepared at different substrate temperatures.
1065
1066
ARKHIPOV et al.
I, µA 3
ln (I/U 2)
2.0
1.5
(a)
2
4
3
1 2
2
1.0 0 0.3
0.5
0.4 0.5 0.6 1/U, kV−1
0.7
1 0 0
1
2 U, kV
3
I, µA (b)
2.0 560°C 620°C
1.5
510°C
400°C 270°C
1.0
20°C 0.5
0 0.5
1.0
1.5 U, kV
2.0
2.5
Fig. 2. (a) Variation of the emission characteristics of a car bon coating sample in the course of its thermal activation at a temperature of 560°C: (1) at the beginning of heating at this temperature, (2) after 100 min of heating, and (3) after 200 min of heating. The inset shows the same data plotted in Fowler–Nordheim coordinates. (b) Emission characteristics of a sample in the most activated state mea sured at different temperatures.
sputtering of a pyrolyticgraphite target at a discharge current of 100 mA, zero substrate potential, and pro cess duration of about 80 s. ptype silicon substrates of KDB10 brand (pSi:B with a resistivity of 10 Ω cm) were used. The properties of the coatings that were deposited were optimized by varying the substrate temperature during their fabrication and by using the thermal activation procedure [6]. It is known (see, e.g., [7]) that films formed by magnetron sputtering at low temperatures with application of an electric potential to the substrate mainly contain carbon in the amorphous sp3 state. In deposition onto a heated sub strate, or upon subsequent heating of the coatings [8– 12[, or upon exposure to light or irradiation with ions
[11], nanocrystalline inclusions, including graphite like ones (sp2), are gradually formed in a film. The emission capacity of the films is determined by their electronic properties and structure, and, therefore, strongly depends on the state of carbon they contain [6, 12–14]. The goal of our study was to determine the conditions in which carbon films formed by magne tron sputtering can serve as effective distributed sources of cold field electron emission. We compared the characteristics of carbon coatings formed in the same operating regime of the deposition installation at the same process duration, but at differ ent substrate temperatures. The surface topography of the samples was examined with a NanoDST atomic force microscope (ATM, Pacific Nanotechnology) (Fig. 1). On its surface, the sample formed at a temperature of 400°C had, in addition to the natural profile of oxi dized silicon, carbon islands with a transverse size of 10–30 nm and a height of up to 1–2 nm. Their density was 20–50 μm–2. Similar islands on the sample pre pared at 500°C had a somewhat larger transverse size (up to 30–40 nm) and a substantially larger height (up to 3–4 nm). As the substrate temperature was raised to 600°C, the height of the islands increased to ~30 nm and approached their transverse dimensions. In addi tion, separate needlelike elements appeared with a thickness on the order of tens of nanometers and lengths of up to 1 μm lying on the coating surface. The surface of a sample fabricated at a substrate tempera ture of 700°C is covered by a nearly continuous layer of needles of this kind. It is noteworthy that the needle images obtained in the contact operating regime of the atomicforce microscope were stable, which points to their good mechanical attachment to the surface. This makes it possible to state with a high degree of confi dence that the arrangement of the needles on the sur face remained unchanged upon application of an elec tric field (they did not “rise” in the direction of force lines). Therefore, the largest geometrical aspect ratio of morphological surface elements (the ratio between the “vertical” and transverse dimensions) presumably did not exceed unity even for this sample and the emis sion properties of the deposited coatings could not be determined by multiple local enhancement of the applied electric field at tips or ribs, as in the case of nanotubes [15] or graphene structures [16, 17]. The layer formed in the given mode had a substantial con ductivity, which is indicated by its strong IR absorp tion in a wide frequency range (1000–3000 cm–1), which exceeded the absorption by samples deposited at lower temperatures by more than a factor of 30. The emission characteristics of carbon coating samples were measured with a previously described [6] experimental device, in which the emission current from an area on the order of 10 mm2 was recorded on applying a potential difference of up to 5 kV to a 0.5mmwide planar gap. The tests were carried out at a residual gas pressure on the order of 10–7 Torr.
TECHNICAL PHYSICS LETTERS
Vol. 40
No. 12
2014
LOWVOLTAGE FIELD EMISSION FROM CARBON FILMS
1.2 µA
7 µA
80 µA
1067
12 µA
140 µA
Fig. 3. Evolution of the emission images of a sample with increasing discharged current. The size of the region shown in the figure is 5 × 5 mm.
The carbon coating samples shown in Fig. 1 did not exhibit any emission capacity in the freshly prepared state even at the maximum field strength (10 V/μm). The coatings were then subjected to an activation pro cedure, which consisted in prolonged heating of a sample, with periodic monitoring of its emission prop erties [6]. The temperature was gradually raised from room temperature until an emission current appeared or up to 850°C. Among the abovedescribed kinds of coatings, the activation procedure was successful only for the sample fabricated at 500°C. In the rest of the cases, a measurable emission current was either not obtained at all even at the highest field strengths and temperatures, or was unstable and soon was instanta neously terminated, probably due to the disintegration of a small number of “accidental” emission centers or of even a single center of this kind. The thermal activation was effective for the sample deposited on the substrate at 500°C. A measurable emission current appeared already on its being heated to a temperature of 520°C, which only slightly exceeds the coating deposition temperature. Keeping the coat ing at a temperature of 560–580°C for 3 h led to a gradual improvement of its emission properties. The emission characteristics of the sample, measured at a temperature of 560°C in different stages of its activa tion, are shown in Fig. 2a. These dependences are approximately linear in the Fowler–Nordheim coor dinates. The threshold electric field strength at which a 1nA current was recorded in the best activated state was about 1.6 V/μm. After this state was reached, the emission properties of the coating were stabilized and remained unchanged even upon its heating to higher temperatures (>700°C). Lowering the temperature of an activated sample also did not lead to loss of its emission capacity; i.e., TECHNICAL PHYSICS LETTERS
Vol. 40
No. 12
the activation was irreversible. Figure 2b shows how the emission curves depend on temperature. The char acteristics measured at temperatures higher than 500°C nearly coincide. At lower temperatures, the emission current is somewhat smaller, but the thresh old field strengths are rather low for all the character istics presented in the figure (<2 V/μm). The existence of the temperature dependence may be due to the low degree of doping of the silicon substrates used in the study, with their resistance limiting the emission cur rent in the cold state and becoming insignificant at high temperatures. Visual inspection of the spatial distributions of the emission current by replacing the anode with a lumi nescent screen showed that the number of operative emission centers grows with increasing applied field (Fig. 3). This behavior is typical of carbon field emit ters [18] and indicates that there is a mechanism lim iting the current obtained from a single emission cen ter, which improves the stability of the centers against overheatingcaused disintegration. Our experiment showed that this property is also inherent in the films of the type under study. Thus, the fundamental aspects of the emission behavior exhibited by carbon island films produced by magnetron sputtering of a graphite target were found to be similar to those previously observed for CVD grown coatings [6]. In both cases, the most effective emission was observed for coatings constituted by iso lated graphitized (originally, or as a result of subse quent thermal treatment) islands having transverse dimensions of 30–40 nm at a thickness of 3–4 nm. This observation can be naturally explained under the assumption that facilitated emission from films of this kind is due to the emission into vacuum of hot elec trons receiving additional energy as a result of the tun 2014
1068
ARKHIPOV et al.
nel injection from the substrate into carbon nanois lands. According to [19, 20], the relaxation of hot electrons in these islands is strongly suppressed at their “correct” size by the socalled “bottleneck effect” [21, 22], which disrupts the electron–phonon coupling. As a result, the lifetime of hot electrons (according to the experimental data of [23]!) may reach a value of 1 ns, which is sufficient for them to drift across an island to the vacuum barrier, with a high probability of their fur ther tunneling across this barrier into vacuum. In the case of a solid conducting coating, the operation of the described emission mechanism is impossible because the externally applied electric field hardly penetrates into the substrate and no conditions for generation of hot electrons are created. If the film has an island structure, the potential difference between the sub strate bulk and an island grows with its increasing size at a fixed applied field strength [24, 25]. With a pre served width of the tunnel gap and the operative mech anism of electric charge removal from the island (e.g., of the emission type), this must lead to an increase in the efficiency of hot electron emission into these islands. At the same time, the energy relaxation time of electrons injected into the island becomes shorter if the island size grows to above a certain limit. The simultaneous effect of these two tendencies must lead to the existence of an island size that is optimal as regards the emission properties. Presumably, islands with dimensions of about 30 nm at a thickness of 3– 4 nm satisfy these optimal conditions, which accounts both for the above experimental results and for the pre viously obtained data [6]. Acknowledgments. This study was in part financially supported by the Ministry of Education and Science of the Russian Federation, grant no. 11.G34.31.0041. REFERENCES 1. A. V. Karabutov, V. D. Frolov, V. I. Konov, et al., J. Vac. Sci. Technol. B 19, 965 (2001). 2. A. V. Okotrub, L. G. Bulusheva, A. V. Gusel’nikov, et al., Carbon 42, 1099 (2004). 3. Z. Shpilman, Sh. Michaelson, R. Kalish, and A. Hoff man, Diamond Relat. Mater. 15, 846 (2006). 4. K. Uppireddi, B. R. Weiner, and G. Morell, J. Vac. Sci. Technol. B 28, 1202 (2010).
5. K. Nose, R. Fujita, M. Kamiko, and Y. Mitsuda, J. Vac. Sci. Technol. B 30, 011204 (2012). 6. A. V. Arkhipov, P. G. Gabdullin, S. I. Krel, et al., Fuller. Nanotub. Carbon Nanostruct. 20 (4–7), 468 (2012). 7. C. A. Dimitriadis, N. A. Hastas, N. Vouroutzis, et al., J. Appl. Phys. 89, 7954 (2001). 8. B. K. Tay, D. Sheeja, S. P. Lau, et al., Surf. Coat. Tech nol. 130, 248 (2000). 9. M. Chhowalla, A. C. Ferrari, J. Robertson, and G. A. J. Amaratunga, Appl. Phys. Lett. 76, 1419 (2000). 10. X. Chen, J. P. Sullivan, T. A. Friedmann, and J. M. Gibson, Appl. Phys. Lett. 84, 2823 (2004). 11. H. Naramoto, X. Zhu, Y. Xu, et al., Phys. Solid State 44, 668 (2002). 12. P.C. Huang, W.C. Shih, H.C. Chen, and I.N. Lin, J. Appl. Phys. 109, 084309 (2011). 13. J. D. Carey and S. R. P. Silva, Phys. Rev. 70, 235417 (2004). 14. O. S. Panwar, M. A. Khan, B. S. Satyanarayana, et al., J. Vac. Sci. Technol. B 28, 411 (2010). 15. A. V. Eletskii, Phys. Usp. 53 (9), 863 (2010). 16. A. Malesevic, R. Kemps, A. Vanhulsel, et al., J. Appl. Phys. 104, 084 301 (2008). 17. J. Liu, B. Zeng, X. Wang, J. Zhu, and Y. Fan, Appl. Phys. Lett. 101, 153104 (2012). 18. A. N. Obraztsov, I. Yu. Pavlovskii, and A. P. Volkov, Tech. Phys. 46, 1437 (2001). 19. A. Pandey and P. GuyotSionnest, Science 322, 929 (2008). 20. W. A. Tisdale, K. J. Williams, B. A. Timp, et al., Science 328, 1543 (2010). 21. H. Benisty, Phys. Rev. 51, 13 281 (1995). 22. T. Inoshita and H. Sakaki, Physica B 227, 373 (1996). 23. K. Mukai and M. Sugawara, in SelfAssembled InGaAs/GaAs Quantum Dots, Ed. by M. Sugawara (Aca demic Press, San Diego, 1999), Semiconductors and Semimetals Series, Vol. 60, p. 209. 24. G. C. Kokkorakis and J. P. Xanthakis, Surf. Interface Anal. 39, 135 (2007). 25. A. V. Arkhipov, P. G. Gabdullin, and M. V. Mishin, Fuller. Nanotub. Carbon Nanostruct. 19, 86 (2011).
Translated by M. Tagirdzhanov
TECHNICAL PHYSICS LETTERS
Vol. 40
No. 12
2014