ISSN 10637826, Semiconductors, 2014, Vol. 48, No. 9, pp. 1242–1247. © Pleiades Publishing, Ltd., 2014. Original Russian Text © A.P. Dostanko, O.A. Ageev, D.A. Golosov, S.M. Zavadski, E.G. Zamburg, D.E. Vakulov, Z.E. Vakulov, 2014, published in Fizika i Tekhnika Poluprovodnikov, 2014, Vol. 48, No. 9, pp. 1274–1279.
FABRICATION, TREATMENT, AND TESTING OF MATERIALS AND STRUCTURES
Electrical and Optical Properties of ZincOxide Films Deposited by the IonBeam Sputtering of an Oxide Target A. P. Dostankoa, O. A. Ageevb, D. A. Golosova^, S. M. Zavadskia, E. G. Zamburgb, D. E. Vakulovb, and Z. E. Vakulovb a
Belarusian State University of Informatics and Radioelectronics, Minsk, 220013 Belarus ^email:
[email protected] b Southern Federal University, Taganrog, 347928 Russia Submitted November 25, 2013; accepted for publication December 3, 2013
Abstract—The influence of the parameters of the deposition process on the stoichiometric composition and electrical and optical properties of ZnO films deposited by the ionbeam sputtering of a ZnO target is studied. It is established that, upon sputtering of a ZnO target with stoichiometric composition, there is a deficit of oxygen in the films deposited. Even for the case of target sputtering in a pure O2 atmosphere, the stoichiom etry index of the films is no higher than 0.98. A decrease in the oxygen content in the films is accompanied by a sharp decrease in the resistivity to 35–40 Ω m, narrowing of the optical band gap, and a shift of the optical transmittance edge from 389 to 404 nm. All of the variations in the optical and electrical properties of the ZnO films can be attributed to variations in the concentration and mobility of free charge carriers in the films. DOI: 10.1134/S1063782614090073
1. INTRODUCTION In recent years, renewed interest in thinfilm zinc oxide (ZnO) layers has been noted. The increased interest in this field is caused by the unique combina tion of optical and electrical properties of zinc oxide. Zincoxide films that possess a high degree of chemi cal inertness and resistance to atmospheric effects are used as conductive coatings, transparent in the infra red (IR) and visible spectral regions, for electrical con tacts and buffer layers of thinfilm solar cells and information display units [1, 2]. The doping of ZnO films makes it possible to attain a resistivity up to 2 × 10–4 Ω cm [3, 4]. Zincoxide films possess good piezo electric and electroluminescence characteristics and can be used as functional layers in devices, whose operation is based on surface acoustic waves [5, 6], sources and detectors of ultraviolet (UV) and IR radi ation, optical shutters, and nonlinear optical elements [7–9]. The ability of heated ZnO films to adsorb cer tain gases and, in this case, to change conductivity makes possible the use of these films as active elements of gas sensors [10]. The widespread industrial use of devices based on ZnO layers is limited by the complexity of the produc tion of ZnO films with specified functional character istics. Zinc oxide belongs to the group of directgap semiconductors, and the properties of the deposited films depend to a large extent on the concentration of free charge carriers controlled by oxygen vacancies [2]. The currently most widely applied method of the formation of thinfilm ZnO layers is reactive magne
tron sputtering [11, 12]. This technique provides a means for varying the conditions of film deposition within wide limits and, thus, for controlling the elec trical and structural properties of the deposited coat ing. Among the disadvantages of this technique, the complexity of the production of stoichiometric layers without heating the substrates must be mentioned. The magnetron sputtering procedures developed thus far provide the production of ZnO layers at tempera tures higher than 200°C, which restricts the field of possible applications of the layers [13]. For example, in the case of deposition onto polymeric substrates, the temperature during deposition of the films must be no higher than 80–120°C (the typical temperatures are 70–80°C). In addition, it is found that, in the case of the magnetron deposition of ZnO films, the distri bution of the optical and electrical characteristics of the resulting ZnO layers is nonuniform because of bombardment of the growing film by highenergy neg ative oxygen ions and atoms [14]. At the same time, inadequate attention is given to the use of the ionbeam sputtering of oxide targets. This technique provides deposition rates comparable to those in the abovementioned technique [15] and the formation of stoichiometric layers without heating of the substrate. In addition, the technique allows comparatively easy scaling [16]. Thus, the purpose of this study is to explore the electrical and optical characteristics of ZnO films deposited by the ionbeam sputtering of a ZnO oxide target without heating of the substrate.
1242
ELECTRICAL AND OPTICAL PROPERTIES OF ZINCOXIDE FILMS Neutralizer power supply 0−25 A Ion source power supply 1 0−5.0 kV 300 mA
S N
To pump
IS
ZnO target Ion source power supply 2 0−5.0 kV 300 mA
SN
Substrate MFC
Ar
O2
Fig. 1. Schematic representation of the experimental sys tem for the deposition of zincoxide layers by ionbeam sputtering. The abbreviations IS and MFS refer to the ion source and automated mass flow controller, respectively.
2. EXPERIMENTAL The experimental system for the deposition of ZnO films by the ionbeam sputtering of an oxide target is shown in Fig. 1. The system is constructed on the basis of a VU2 vacuum unit. The chamber of the system is equipped with a doublebeam ion source based on Hallcurrent accelerator DBIS001. The accelerator was used for preliminary ion cleaning of the substrate surfaces and for sputtering of the target material. To neutralize the ion beams and to compensate the sur face charge produced upon the sputtering of insulator targets, a filament compensator was used. To deposit ZnO thin films by ionbeam sputtering, we used a target 79.6 mm in diameter and 3.2 mm in thickness. The target was fabricated by static pressing of the pureforanalysis ZnO powder. As the sub strates, we used Si (100) wafers and BK7 optical glass wafers with a thickness of 2 mm. The substrates were mounted onto a rotating drumtype substrate holder, which allowed us to bring the substrates into the depo sition area by turns. Before depositing the layers, we pumped the chamber of the vacuum unit to a residual pressure of 10–3 Pa and conducted preliminary ion cleaning of the substrates. For this purpose, Ar working gas was sup plied to the assisting unit of the ion source to an oper ating pressure of 2.0 × 10–2 Pa. In all of the experi ments, the cleaning time, ion energy, and discharge current were constant and corresponded to 5 min (in the mode of rotation of the substrate holder), 500 eV, and 70 mA, respectively. SEMICONDUCTORS
Vol. 48
No. 9
2014
1243
Then we cleaned the target surface of contamina tions and adsorbed gases. To do this, we removed the substrates from the deposition area. The target was sputtered with Ar+ ions with an energy of 1300 eV; the ions were generated by the sputtering unit of the dublebeam ion source DBIS001. In all of the exper iments, the discharge current was constant and corre sponded to 150 mA (the target current was 110 mA). The target cleaning time was 10 min. After cleaning the target, we deposited the ZnO layers. The substrates were brought into the deposition area. The working gases were supplied to the gasdis tribution system of the ionsource sputtering unit. The oxygen content in the Ar/O2 gas mixture was varied from zero to 100%, with the total gas rate 25 sccm (at a pressure in the chamber of about 0.02 Pa). The rate of the Ar/O2 working gas supply to the sputtering ion source was controlled by RRG1 mass flow control lers. The use of automated mass flow controllers allowed us to maintain an exact ratio between the partial pres sures of the working gases during the entire process. Upon the deposition of ZnO layers, we carried out two series of experiments for different sputtering condi tions: (i) the discharge voltage Ud = 4.5 kV, the dis charge current Id = 130 mA, the current of the sole noid Ic = 8.0 A, the target–substrate distance 27 cm, and the deposition time 30 min; (ii) the discharge volt age Ud = 5 kV, the discharge current Id = 200 mA, the current of the solenoid Ic = 8.0 A, the target–substrate distance 27 cm, and the deposition time 20 min. The films were deposited to a thickness of 200–400 nm depending on the oxygen percentage in the working gas mixture. To form the crystal structure, we annealed some of the deposited films using an Isoprin setup for IR heating. Annealing was conducted in air and an O2 atmosphere at a temperature of 300–800°C for 1 h. The temperature was elevated and lowered with a rate of 20 °C min–1. The resistance of the films was estimated for the test structures. With this purpose, we deposited the films onto heavily doped singlecrystal silicon sub strates. Then, by ionbeam sputtering through a mask, we deposited the upper Ni electrode onto the ZnO layer. The size of the capacitors produced in this man ner was 0.8 × 0.8 mm. The conductivity of the films was determined by measuring the resistance of the test structure at the frequency 1 kHz. This was done with an E720 immittance meter. The resistivity of the lay ers was calculated by the formula RS ρ = , (1) d where R and S are, correspondingly, the resistance of the test structure and its area (S = 6.4 × 10–7 m2) and d is the film thickness. The distribution of the thickness over the layer area was determined using a POI08 optical interferometric profilometer. The optical thickness and the refractive index of the films n were analyzed by optical ellipsometry, with an LEF3M
1244
DOSTANKO et al. Stoichiometric index 1.0
Deposition rate, nm/s 0.26
a 0.22
0.9
0.18
b b
0.8 0.14 0.10 0.06
0.7
a
0
20
40
60
80
100 [O2], %
0.6
0
20
40
60
80
100 [O2], %
Fig. 2. Dependences of the deposition rate of the zinc oxide films on the oxygen content in the Ar/O2 gas mixture under different conditions of deposition: (a) Ud = 4.5 kV, Id = 130 mA; (b) Ud = 5.0 kV, Id = 200 mA.
Fig. 3. Dependences of the stoichiometric index of the zincoxide films on the oxygen content in the Ar/O2 gas mixture under different conditions of deposition: (a) Ud = 4.5 kV, Id = 130 mA; (b) Ud = 5.0 kV, Id = 200 mA.
ellipsometer. The transmittance spectra in the range 350–900 nm were obtained using a Proscan MS121 spectrophotometer. The elemental composition of the films was analyzed by Xray fluorescence analysis through the use of an Oxford ED2000 spectrometer and an AVALON 8000 (Princeton GammaTech) energydispersive spectrometer.
transmittance measurements, the films were deposited onto BK7 glass substrates. It is established that the average transmittance in the visible spectral region is at a level of 0.8–0.85. The positions of the transmittance peaks depend on the optical thickness of the deposited layers. For comparison, Figure 4 shows the transmit tance spectrum of the substrate (curve d). Optically transparent layers are formed even upon sputtering in an Ar atmosphere. It is established that the optical transmittance edge at a level of 0.5 corresponds to the wavelength ~395 nm. As the O2 content in the Ar/O2 gas mixture is increased, the optical transmittance spectra shift to shorter wavelengths (Fig. 5).
3. RESULTS AND DISCUSSION We studied the effect of the parameters of ionbeam sputtering and the composition of working gases on the growth rate and stoichiometry of the ZnO films. The dependences of the deposition rate of the ZnO films on the O2 content in the Ar/O2 gas mixture were obtained (Fig. 2). The deposition rate of the films steadily increased, as the oxygen content was reduced. This is caused by an increase in the average mass of the bombarding ions and by a partial reduction in zinc on the target surface when bombarded with Ar ions. Under variations in the discharge parameters of the ion source, the deposition rate varies proportionally to the discharge power. It should be noted that, even upon sputtering in a pure oxygen atmosphere, the deposition rate of zinc oxide during ionbeam sputter ing is comparable to or higher than the typical deposi tion rates during the processes of reactive magnetron sputtering [3, 17]. Analysis of the elemental composition of the deposited films shows that all of the films exhibit a def icit of oxygen (Fig. 3). Even if the target is sputtered in a pure O2 atmosphere, the stoichiometry index of the film does not exceed 0.98. As the sputtering rate is increased, the deficit of oxygen increases. We obtained the optical transmittance spectra of the zincoxide films deposited in the As/O2 gas mix tures containing different O2 fractions (Fig. 4). For the
Transmittance, % 100 d 80 c
a
b
60 40 20 0 300
400
500
600
700 800 900 Wavelength λ, nm
Fig. 4. Transmittance spectra of the zincoxide films deposited by ionbeam sputtering in Ar/O2 gas mixtures containing different oxygen fractions: (a) 0, (b) 50, and (c) 100% O2. Curve d refers to the transmittance spectrum of the initial substrate. SEMICONDUCTORS
Vol. 48
No. 9
2014
ELECTRICAL AND OPTICAL PROPERTIES OF ZINCOXIDE FILMS Optical absorption edge, nm 408
(αE)2, 1010 cm−2 eV2 6
404
5
1245
e
c
b a
4
400
d
3 396 2 392 388
1 0
20
40
60
80
100 [O2], %
0 3.0
3.2
3.1
3.3
3.4
3.5
3.6 E, eV
Fig. 5. Dependence of the wavelength corresponding to the optical absorption edge (at a transmittance of 50%) of zincoxide films on the oxygen content in the Ar/O2 gas mixture.
Fig. 6. Spectra of the absorption coefficient squared for zincoxide films deposited in Ar/O2 gas mixtures contain ing different oxygen fractions: (a) 0, (b) 30, (c) 51, (d) 74, and (e) 100%.
The band gap was determined from the fundamen tal absorption edge of zinc oxide [18]. In the case of direct interband transitions, the dependence of the absorption coefficient α on the photon energy E is described by the relation
varies within wider limits. Such behavior is apparently due to more pronounced variations in the stoichiome try index of the films at high sputtering rates. It should be noted that films with nearly stoichiometric compo sition possess smaller values of the refractive index. We determined the dependences of the resistivity of the zincoxide films on the O2 content in the Ar/O2 gas mixture (Fig. 9). As the O2 content in the Ar/O2 gas mixture is increased, the resistivity first decreases, reaches a minimum of 35–40 Ω m at an oxygen con tent of 10–30% and then sharply increases to 106 Ω m at an oxygen content of 40–50%. As the O2 content is increased further, the resistivity only slightly increases to 6 × 106 Ω m. Variations in the conductivity of the
α ( E ) = A E – Eg .
(2)
Here, A is a constant independent of frequency and Eg is the band gap. In the ideal case, if the results are plot ted on the scale α2(E), the experimental points are bound to fall on a straight line with the slope A2, and at α = 0, this line intersects the E axis at the point E = Eg. Figure 6 shows the spectra of the absorption coeffi cient of the zincoxide films deposited in the Ar/O2 gas mixture containing different oxygen fractions (Ud = 5.0 kV, Id = 200 mA). In the plots shown in Fig. 6, we can clearly distinguish a linear portion; approximation of this portion with relation (1) gives a photon energy corresponding to the optical band gap. It is found that the optical band gap of the zincoxide thin films varies from 3.215 to 3.265 eV, as the O2 con tent in the Ar/O2 gas mixture is increased (Fig. 7). The widening of the band gap with increasing O2 content during deposition can be attributed to an increase in the concentration of charge carriers and to the Burst ein–Moss shift. Such widening of the band gap was noticed in nonstoichiometric ZnO films previously [19]. Analysis of the dependence of the refractive index n on the oxygen content in the Ar/O2 gas mixture (Fig. 8) shows that, for the film produced at a lower sputtering rate, the refractive index n remains practi cally unchanged under variations in the relative O2 content in the workinggas mixture and varies in the region of 1.95–2.0. Upon highrate sputtering of a zincoxide target (Fig. 8, curve b), the refractive index SEMICONDUCTORS
Vol. 48
No. 9
2014
Eg, eV 3.27 3.26 3.25 3.24 3.23 3.22 3.21
0
20
40
60
80
100 [O2], %
Fig. 7. Dependence of the optical band gap on the oxygen content in the Ar/O2 gas mixture.
1246
DOSTANKO et al. Resistivity, Ω m 107
Refractive index n 2.1
106 a
2.0
105 104
1.9 b 1.8
103 102 101
1.7
0
20
40
60
80
100 [O2], %
Fig. 8. Dependence of the refractive index of the zinc oxide films on the oxygen content in the Ar/O2 working gas mixture at different deposition parameters: (a) Ud = 4.5 kV, Id = 130 mA; (b) Ud = 5.0 kV, Id = 200 mA.
zincoxide films are associated with variations in the concentration of oxygen vacancies in the films. If the zincoxide target is sputtered in an argon atmosphere with a low oxygen fraction, zinc oxide is partially reduced, and we observe a deficit of oxygen in the films. This deficit favors an increase in the conductiv ity of the deposited films. Postgrowth annealing of the films in air and in an O2 atmosphere yields a sharp increase in the resistivity of the films. Irrespective of the conditions of deposi tion, the resistivity of the films increases upon anneal ing and reaches 106 Ω m. 4. CONCLUSIONS The optical and electrical properties of zincoxide films deposited by the ionbeam sputtering of a target without substrate heating are studied. It is established that the deposition rate of zinc oxide upon ionbeam sputtering is comparable to or higher than the deposi tion rates typical of the magnetronsputtering tech nique. The stoichiometry index of the films depends on the O2 content in the Ar/O2 gas mixture. Upon sputtering of the zincoxide target of stoichiometric composition even in a pure oxygen atmosphere, a def icit of oxygen is observed in the films, and the stoichi ometry index of the deposited zinc oxide does not exceed 0.98. A decrease in the oxygen content in the films is accompanied by a sharp decrease in the resis tivity to 35–40 Ω m, a narrowing of the optical band gap, and a shift of the optical absorption edge from 389 to 404 nm. All variations in the optical and electrical properties of the ZnO films can be attributed to varia tions and the concentration and mobility of free charge carriers in the films.
100
0
20
40
60
80
100 [O2], %
Fig. 9. Dependence of the resistivity of the zincoxide films on the oxygen content in the Ar/O2 gas mixture.
ACKNOWLEDGMENTS The study was supported by the Russian Founda tion for Basic Research and the Federal targeted pro gram “Scientific and Scientific–Pedagogical Person nel of Innovative Russia” for 2009–2013, projects nos. 120890045Bel_a and 14.A18.21.0900, and the Belarussian Republican Foundation for Fundamental Research, project no. T12R191. REFERENCES 1. C. G. Granqvist, Solar Energy Mater. Solar Cells 91, 1529 (2007). 2. K. Ellmer, A. Klein, and B. Rech, Transparent Conduc tive Zincoxide: Basics and Applications in Thin Film Solar Cells (Springer, 2008). 3. M. K. Jayaraj, A. Antony, and M. Ramachandran, Bull. Mater. Sci. 25, 227 (2002). 4. M. Kon, P. K. Keun, Y. Shigesato, P. Frash, M. Akio, and K. Susuki, Jpn. J. Appl. Phys. 41, 6174 (2002). 5. A. F. Belyanin, N. V. Suetin, P. V. Pashchenko, M. A. Timofeev, D. V. Lopaev, V. G. Pirogov, S. I. Poly akov, N. I. Sushentsov, L. V. Pavlushkin, and V. A. Krivchenko, Sist. Sredstva Svyazi, Televid. Radioveshch., Nos. 1–2, 76 (2006). 6. D.S. Liu, C.Y. Wu, C.S. Sheu, F.C. Txai, and C. H. Li, Jpn. J. Appl. Phys. 45, 3531 (2006). 7. V. A. Krivchenko, D. V. Lopaev, P. V. Pashchenko, V. G. Pirogov, A. T. Rakhimov, N. V. Suetin, and A. S. Trifonov, Tech. Phys. 53, 1065 (2008). 8. A. N. Gruzintsev, V. T. Volkov, K. Bartkhou, and P. Benaul, Semiconductors 36, 701 (2002). 9. A. N. Gruzintsev, V. T. Volkov, and E. E. Yakimov, Semi conductors 37, 259 (2003). SEMICONDUCTORS
Vol. 48
No. 9
2014
ELECTRICAL AND OPTICAL PROPERTIES OF ZINCOXIDE FILMS 10. N. L. Hung, J. Korean Phys. Soc. 57, 1784 (2010). 11. W. L. Dang, Y. Q. Fu, J. K. Luo, A. J. Flewitt, and W. I. Milne, Superlatt. Microstruct. 42, 89 (2007). 12. V. A. Aleksandrov, A. G. Veselov, O. A. Kiryasova, and A. A. Serdobintsev, Tech. Phys. Lett. 38, 843 (2012). 13. S. Youssef, P. Combette, J. Podlecki, R. Al Asmar, and A. Foucaran, Cryst. Growth Des. 9, 1088 (2009). 14. T. Minami, T. Miyata, T. Yamamoto, and H. Toda, J. Vac. Sci. Technol. 18, 1584 (2000). 15. Zhu Chang, Mi Gaoyuan, A. P. Dostanko, D. A. Golosov, and S. M. Zavatskiy, J. Appl. Opt. 31, 855 (2010).
SEMICONDUCTORS
Vol. 48
No. 9
2014
1247
16. D. A. Golosov, S. M. Zavadski, I. V. Svadkovski, and S. N. Melnikov, Vakuum. Tekh. Tekhnol. 20, 227 (2010). 17. T. K. Subramanyam, B. Srinivasulu Naidu, and S. Uthanna, Cryst. Res. Technol. 35, 1193 (2000). 18. S. I. Sadovnikov, N. S. Kozhevnikova, and A. A. Rem pel, Semiconductors 44, 1349 (2010). 19. A. P. Roth, B. W. James, and D. F. Williams, Phys. Rev. B 25, 7836 (1982).
Translated by E. Smorgonskaya