- Email: [email protected]
Synthesis, Crystal Structural and Electrical Conductivity Properties of Fe-Doped Zinc Oxide Powders at High Temperatures
J. Mater. Sci. Technol., 2012, 28(3), 268–274.
Synthesis, Crystal Structural and Electrical Conductivity Properties of Fe-Doped Zinc Oxide Powders at High Temperatures Hakan C ¸ olak1,2)† and Orhan T¨ urkoˇglu1) 1) Department of Chemistry, Science of Faculty, University of Erciyes, Kayseri, 38039, Turkey 2) Deparment of Chemistry, Science of Faculty, University of C ¸ ankırı Karatekin, Ballıca Campus, C ¸ ankırı, 18200, Turkey [Manuscript received December 31, 2010, in revised form March 5, 2011]
The synthesis, crystal structure and electrical conductivity properties of Fe-doped ZnO powders (in the range of 0.25–15 mol%) were reported in this paper. I-phase samples, which were indexed as single phase with a hexagonal (wurtzite) structure in the Fe-doped ZnO binary system, were determined by X-ray diffraction (XRD). The solubility limit of Fe in the ZnO lattice is 3 mol% at 950 ◦ C. The above mixed phase was observed. And the impurity phase was determined as the cubic-ZnFe2 O4 phase when compared with standard XRD data using the PDF program. This study focused on single I-phase ZnO samples which were synthesized at 950 ◦ C because the limit of the solubility range is the widest at this temperature. The lattice parameters a and c of the I-phase decreased with Fe-doping concentration. The morphology of the I-phase samples was analyzed with a scanning electron microscope. The grain size of the I-phase samples increased with heat treatment and doping concentration. The electrical conductivity of the pure ZnO and single I-phase samples was investigated using the four-probe dc method at 100–950 ◦ C in air atmosphere. The electrical conductivity values of pure ZnO, 0.25 and 3 mol% Fe-doped ZnO samples at 100 ◦ C were 2×10−6 , 1.7×10−3 and 6.3×10−4 S·cm−1 , and at 950 ◦ C they were 3.4, 8.5 and 4 S·cm−1 , respectively. KEY WORDS: II-VI semiconductors; Zinc oxide and doped zinc oxide; Four point probe method
1. Introduction Zinc oxide (ZnO) powders are very important materials due to the many interesting properties inherent in this material, such as dielectric, piezoelectric, pyroelectric, semiconducting, acousto-optic, optical, electro-optical, nonlinear optical, photoelectrochemical and electrical properties[1] . ZnO is one of the most versatile and technologically interesting semiconducting materials because of its typical properties such as resistivity control over the range of 10−3 –10−5 Ω·cm, transparency in the visible range, high electrochemical stability, direct band gap, absence of toxicity and abundance in nature[2] . ZnO normally occurs in the hexagonal wurtzite crystal structure with a=0.32488 nm and c=0.52066 nm in the standard data (JCPDS, 36-1451). Electron doping in † Corresponding author. Assist. Prof., Ph.D.; Tel.: +90 506 9868893; Fax: +90 376 2181031; E-mail address: hakan [email protected] (H. C ¸ olak).
nominally undoped ZnO has been attributed to Zn interstitials and oxygen vacancies[3] . High purity ZnO crystals exhibit strong n-type conductivity with the electrons moving in the conduction band as charge carriers. This characteristic has traditionally been attributed to native defects such as oxygen vacancies and zinc interstitials[4,5] . ZnO has both good electronic and optical properties because of a stoichiometric deviation due to the existence of intrinsic defects such as O vacancies and Zn interstitials. However, these properties of pure ZnO are unstable due to the adsorption of atmospheric oxygen, which decreases the conductivity, and they cannot meet the increasing needs for present day applications. To stabilize them against such changes and enhance the properties of the ZnO, doping is necessary and this purpose was achieved by adding some dopants[6,7] . Moreover, doping leads to an increase in the conductivity of ZnO. ZnO doping was achieved by replacing Zn2+ atoms with the atoms of the dopant elements. The efficiency of the dopant
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
element depends on its electronegativity and the difference between its ionic radius and the ionic radius of zinc[8] . Extensive studies have been carried out to modify the properties of ZnO for different applications. Doping with transitional metal elements leads to many interesting properties of ZnO[9] . In addition, thermal annealing is a widely used method to improve crystal quality and to study structural defects in materials. For semiconductors, thermal annealing is also used to activate dopants. During the annealing process, dislocations and other structural defects will move in the material and adsorption/decomposition will occur on the surface, thus the structure and the stoichiometric ratio of the material will change, which influences the electrical and optical properties of ZnO. The structural, electrical and optical properties are vital for semiconductor devices. It is necessary to investigate how these properties are affected by thermal annealing[10] . Although there are a few reports on the doping elements in the ZnO varistor, the effects of some dopants, such as Fe, on the electrical conductivity of the ZnO varistor still remain unclear. This is because doping effects have been studied in quite different systems under different experimental conditions[11] . In this work, metallic iron-doped ZnO powders were synthesized at high temperatures by using the solid state reaction method. Then, their structural properties were studied and the dc electrical conductivity of the ZnO solid solution samples was measured at high temperatures. The solid state reaction was chosen because of its reproducibility, easy control and sufficent products for measurement[12] . 2. Experimental Synthesized iron (Fe) doped ZnO powders at high temperatures were measured by using the standard solid state reaction method. At first, solid mixtures of Fe-ZnO (0.25≤ x ≤15 mol%) were prepared by mixing and homogenizing with stoichiometric amounts of pure ZnO and metallic Fe in an agate mortar. As starting materials, commercially pure ZnO powders (99.9% Fluka) and metallic iron powders (99.5% Alfa Aesar) were used. The powder mixtures were first calcined at 600 and 650 ◦ C for 24 h. After grinding and homogenization, the pre-annealed mixtures were heat treated at 700, 750, 800, 850, 900 and 950 ◦ C, respectively. All these heat treatments were made in air atmosphere for 48 h, in alumina crucibles and without any compaction procedure. At the end of each heat treatment procedure, the annealed powders were slowly cooled in the furnace by switching it off (uncontrolled). Before and after heat treatment, each of the samples was ground in an agate mortar to homogenize powder size. The samples were then analyzed with X-ray powder diffraction (XRD). The XRD data of the samples were recorded on a computerinterfaced Bruker AXS D8 advanced diffractometer operating in Bragg–Brentano geometry (CuKα radi-
269
ation, graphite monochromator, 40 kV and 40 mA) over a 10 deg.≤2θ≤90 deg. angular range. The divergence and receiving slits of 1 and 0.1mm, respectively, were located on the diffractometer. The diffraction patterns were scanned in 0.002 deg. (2θ) steps and the diffracted beams were counted with a NaI(Tl) scintillation detector. The Diffrac Plus and Win-Index programs were used to obtain information about the crystal structures of the samples. Then, for a comparison with standard data, the PDF program (Maint Powder Diffraction Database Manager Software) was used. In all cases, the XRD patterns of the samples, which were obtained as single phase, could be indexed on the basis of a hexagonal cell in the Fe–ZnO system. The single phase was named as I-phase. The other samples which were not indexed with a hexagonal crystal structure were named as II-phase. Samples which were out of the solubility range were heterogeneous solid mixtures (I+II). The morphologies of both annealed (950 ◦ C) and non-annealed undoped ZnO samples and I-phase samples (950 ◦ C) were observed by scanning electron microscopy (SEM, LEO 440) operated at 20 kV. Before SEM measurement, the I-phase samples in the Fe–ZnO system were pressed into pellets of 0.1 cm in thickness (t) and 1.3 cm in diameter (d) under ∼5 ton pressure (∼369 MPa) and were calcined at 950 ◦ C for 24 h in air. After heat treatment, the XRD measurements showed that no phase change in the hexagonal samples was observed. The average grain sizes of the I-phase samples were calculated by the image-pro plus 5.0 program from the SEM micrographs. The electrical conductivity (σdc ) measurements of the I-phase samples (950 ◦ C) were made using the standard four-probe dc method. The circular samples with a hexagonal structure, which were mentioned above as the pelleted samples, were used for the electrical conductivity measurements. To reduce contact resistance, fine platinum wires were directly attached to the surface of the samples. The ohmic character of the wire contacts was checked prior to each measurement. The contacts were positioned symmetrically with respect to the center of the circular pellet and the contact separations (s) were 0.2 cm. The conductivity measurements were made between 100 and 950 ◦ C for I-phase samples synthesized at 950 ◦ C. The increase in furnace temperature was initial in 100 ◦ C (in the range 100–400 ◦ C) steps, then in 50 ◦ C (450–950 ◦ C) steps in air. During the measurements, the sample temperature was determined by a thermocouple located 5 mm away from the sample. This thermocouple had a cold (0 ◦ C) junction. All experimental data were recorded by a Keithley 2400 sourcemeter and a Keithley 2700 electrometer which were controlled by a computer. 3. Results and Discussion 3.1 Structural properties Because data on the solubility of Fe ions in ZnO
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
270
Table 1 Observed phases in the binary system of Fe-doped ZnO (in the range of 0.25–15 mol%) x (Fe addition)/mol% 0.25 0.50 0.75 1 2 3 4 5 6 7 8 9 10 11 12 13 14 I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I—Fe-doped ZnO (hexagonal structure), II—cubic-ZnFe2 O4 , I+II—heterogeneous solid mixtures
Temp./◦ C 700 750 800 850 900 950 Notes:
Fig. 1 XRD patterns of Fe-doped ZnO: (a) undoped ZnO (950 ◦ C), (b) 0.25 mol% Fe-doped ZnO, (c) 3 mol% Fe-doped ZnO, (d) 6 mol% Fe-doped ZnO, (e) 15 mol% Fe-doped ZnO (950 ◦ C) (∗: the peaks of II-phase, ZnFe2 O4 )
are not known, annealing experiments at varying temperatures were performed[13] . Fig. 1 shows the XRD patterns of pure ZnO (950 ◦ C, 48 h), 0.25, 3, 6 and 15 mol% Fe-doped ZnO samples (950 ◦ C, 48 h). In this figure, all the XRD patterns show typical peak patterns, which can be indexed as the ZnO wurtzite structure. The peak appears at 2θ=34.44 deg., which corresponds to the (002) directions of the ZnO hexagonal wurtzite structure[14] . For 6 and 15 mol% Fedoped ZnO samples, XRD patterns show extra peaks, which indicate ZnFe2 O4 , (PDF No. 01–1108)[15–18] . The XRD patterns of other single I-phase samples, which were synthesized at other temperatures, were quite similar to the patterns of the I-phase samples given in Fig. 1. At 950 ◦ C, the XRD measurements revealed that doping with more than 3 mol% caused the coexistence of the I and II phase. No indication of iron metal impurities was observed in the samples. In the Fe–ZnO binary system, the observed single phases and heterogeneous solid mixtures depending on the reaction temperature and amounts of Fe doping are presented in Table 1. The widest range solubility in the ZnO lattice of Fe ions is at 950 ◦ C, and at this temperature the solubility limit is 3 mol%. At lower temperatures the solubility range is lower than
15 I+II I+II I+II I+II I+II I+II
950 ◦ C. The low solubility of Fe may be satisfyingly explained by the heterovalence of the Fe ions in the ZnO host lattice[13] . The mobility of Fe ions which diffuse in ZnO tetrahedral sites increases with temperature. Heat treatment temperature was found to have an effect on the solubility of Fe ions in ZnO[19] . In addition, both the ionic radii and the valence state are important factors in determining the solubility of the dopants. A smaller deviation of these factors from those of Zn2+ would be favorable for the dopants to have a higher solubility. This study focused on single I-phase ZnO samples synthesized at 950 ◦ C due to the limit of the solubility range, which is the widest at this temperature. The other I-phase samples synthesized at temperatures lower than 950 ◦ C, and the multiphase samples (I+II) were excluded from this study. For the wurtzite ZnO structure, the lattice constants determined at room temperature by various experimental measurements and theoretical calculations were in good agreement. The lattice constants of the samples mostly ranged from 0.32483 to 0.32501 nm for the a parameter and from 0.52023 to 0.52050 nm for the c parameter. The deviation from that of the ideal wurtzite crystal is probably due to lattice stability and ionicity[20] . Table 2 shows the dependence of the Table 2 Unit cell parameters of I-phase ZnO samples at 950 ◦ C x (Fe addition)/mol% 0.25 0.50 0.75 1 2 3
Unit cell parameters/nm a c 0.32501 0.52050 0.32499 0.52047 0.32493 0.52033 0.32488 0.52026 0.32488 0.52025 0.32483 0.52033
lattice parameters a and c, respectively, on the addition ratio of metallic Fe. The lattice parameters of the Fe-doped ZnO system are smaller than those of pure ZnO, both a and c parameters decrease with the substitution of Fe ions with Zn ions. Fe ions are substituted with Zn2+ ions which are in the tetrahedral sites of the wurtzite structure of ZnO. It is considered that valence state of Fe in ZnO is both +2 and +3, namely Fe in the ZnO matrix exists in a mixed valence state. It has been reported that local magnetic probes like electron paramagnetic resonance and
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
271
Fig. 2 SEM images of I-phase samples for pure ZnO (non-annealed) (a), pure ZnO (annealed) (b), 0.25 (c), 0.50 (d), 0.75 (e), 1 (f), 2 (g) and 3 mol% (h) Fe-doped ZnO samples at 950 ◦ C
M¨ ossbauer spectroscopy indicate the presence of Fe in ZnO in both valence states, Fe2+ and Fe3+ . And the presence of Fe3+ in ZnO is due to possible hole doping in the system by cation (Zn) vacancies[21] . It has also been reported in another study[22] that usually, if Fe is present in the substitutional site in a defect-free ZnO crystal, the valence state of Fe will be +2. In this report, also, the X-ray photoelectron spectroscopy (XPS) result confirmed the presence of uncoupled Fe3+ within the sample. In the Fe–ZnO binary system, Fe3+ ions are present due to the existence of cation vacancy. Cation vacancy near Fe can promote Fe2+ into Fe3+ and also mediate the Fe2+ –Fe2+ exchange interaction. Since the transi-
tion metal (TM) doping percentage is slightly on the higher side toward the cationic percolation threshold, Fe2+ –Fe3+ exchange, although being less in number in comparison to the Fe2+ –Fe2+ interaction, may also be possible[22] . However, the ionic radius (fourcoordinated) of the Fe2+ ion is 0.074 nm and that of the Fe3+ is 0.064 nm while the ionic radius of the Zn2+ is 0.074 nm[23] . Therefore, the decrease in the lattice parameters a and c can be explained by the ionic radii difference[24] . Fig. 2 shows the SEM micrographs of both annealed (950 ◦ C) and non-annealed undoped ZnO and I-phase ZnO samples (950 ◦ C). In the SEM micrographs of undoped ZnO samples (Fig. 2(a) and (b)),
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
272
there is a homogeneous grain distribution. From these images the average grain sizes were calculated and were found to be 0.63 and 1.03 μm, respectively[19] . The grain sizes of the annealed sample are obviously larger than those of the non-annealed samples, which could be caused by strain relaxation. The annealing process clearly produces a recovery of the crystal structure and an increase in the grain size[1,25] . As can be seen in the SEM micrographs, the grains of the I-phase samples are homogeneously distributed. The grain size of all synthesized Fe-doped ZnO samples are given in Table 3. From the SEM images, Table 3 Grain size values of I-phase ZnO samples at 950 ◦ C x (Fe addition)/mol% ZnO∗ ZnO 0.25 0.50 0.75 1 2 3 Note: ∗ —non-annealed
Grain size/μm 0.67 1.03 1.32 1.46 1.73 1.93 2.51 2.60
the grain size of the Fe-doped ZnO samples slightly increases with the addition of Fe because the substitution of Fe2+ and Fe3+ for Zn2+ increases the activity of ZnO by means of distortion of the ZnO lattice, which is beneficial to grain growth in Fe-doped ZnO samples[26] . The grains are regular in size and shape. Some pores can be noticed in the photographs. When increasing the dopant concentration to 3 mol%, the porosity increased[24] . In addition, the surface of the samples is smooth. 3.2 Electrical properties The total electrical conductivity is calculated using the following equation: σT =
I −1 G V
(1)
where G is the geometric resistivity correction factor[27] . Both the values and the variation of the electrical conductivity of the pelleted powder ZnO samples are in connection with their structure and its changes. On the other hand, thermal treatments of the respective samples modify their structural characteristics and consequently, their electrical properties. On this basis, the study of the temperature dependence of the electrical properties of ZnO samples, may offer useful information about the possible changes of the structural characteristics of ZnO samples[28] . The temperature dependence of electrical conductivity (σdc ) of the I-phase ZnO pelleted powder samples at 950 ◦ C was studied in the temperature range of 100–950 ◦ C. logσ T –103 /T graphs were drawn in order to show the conductivity change with respect to temperature. The graphs were compared with each other
and the conductivity characteristics of the samples were evaluated. The electrical conductivity plots vs temperature of undoped ZnO, 0.50, 0.75 and 2 mol% Fe-doped ZnO (950 ◦ C) samples are given in Fig. 3. The graphic curves of the other single I-phase samples were quite similar to the curves of the single I-phase samples given in this figure. As seen in Fig. 3, the electrical conductivity of I-phase samples increases with increasing temperature and shows the semiconducting behavior of pure and Fe-doped ZnO samples. It is well known that the electrical conductivity of ZnO samples is controlled by intrinsic defects (oxygen vacancy and interstitial zinc atoms) generated at high temperature[11] . Also, for ZnO samples, increasing electrical conductivity with temperature can be explained by the following: (2) ZnO ↔ Znxi + 1/2O2 (g) Znxi ↔ Zn•i + e−
(3)
Zn•i
(4)
↔
Zn•• i
−
+e
After ionization, carrier concentration is increased by two extra electrons[29] . The electrical conduction of the doped ZnO follows a mechanism in which the electron or hole hops from one localized site to the next. Whenever it is transferred to another site, the surrounding molecules respond to this perturbation with structural changes and the electron or hole is temporarily trapped in the potential well leading to atomic polarization. The electron resides this site until it is thermally activated to migrate to another site. Another aspect of this charge hopping mechanism is that the electron or hole tends to associate with local defects[30] . It is observed that conductivity slightly decreases with increasing Fe concentration (Fig. 4). It is well known that the electrical conductivity of ZnO samples at room temperature is due to intrinsic defects created by oxygen vacancies. These defects introduce donor states in the forbidden band slightly below the conduction band and hence result in the conducting behavior of ZnO. The electrical conductivity is controlled by the intrinsic defects generated during synthesis and by the presence of dopants. It is known that during the sintering process at high temperature, oxygen vacancies will be produced which are responsible for electrical conductivity. From the observed result of decrease in conductivity of ZnO on Fe doping, it seems that the Fe-doping affects the defect chemistry of the ZnO. Therefore, It is believed that Fe in ZnO acts as a deep donor and decreases the concentration of intrinsic donors. This reduction in intrinsic donor concentration increases with increasing Fe content, which in turn decreases the electrical conductivity[31] . Many researchers have studied the effects of 3d transition metal impurities on the electrical conductivity of the ZnO varistor. They investigate that 3d transition metals could enhance the excess oxygen concentration in the grain boundary region and a
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
273
Fig. 3 Electrical conductivity plots for undoped ZnO (a), 0.25 (b), 0.75 (c) and 2 mol% (d) Fe-doped ZnO (950 ◦ C)
is expected between the nearest donors at low temperatures. The total electrical conductivity can be expressed as follows; σdc = σ0 exp(−Ea /kT )
(5)
where σdc is the electrical conductivity at any temperature, Ea is the activation energy, which corresponds to the energy difference between the donor level and the conduction level, σ0 is the pre-exponential factor, k is the Boltzmann constant, and T is the absolute temperature[8,11,30,36] . The activation energy can be calculated from the conductivity equation shown below, Fig. 4 Electrical conductivity variation for Fe-doped ZnO I-phase ZnO samples (950 ◦ C)
potential barrier is formed preferentially. Therefore, the electrical conductivity of Fe-doped ZnO samples is apparently lower than that of undoped ZnO samples, and the grain boundary is more resistive than the grain[32–34] . Also, it was observed from the SEM micrographs that Fe-doping produces a large number of pores which may be the main reason for the significant decrease in the conductivity of the considered samples[35] . In general, for a semiconducting material, dc conductivity increases exponentially with temperature indicating that conductivity is a thermally activated process. For doped semiconductors with low concentration of donors, the hopping transport of carriers
Ea 1 + ln σ0 (6) k T The slope of the linear part of the Arrhenius curve of the lnσT –1/T graph is equal to –Ea /k. For the temperature range the I-phase ZnO samples have Arrhenius-type electrical conductivity. From these curves, drawn for all I-phase ZnO samples, the activation energies Ea were calculated (Fig. 5). The increase in the activation energy of Fe-doped ZnO can be explained as follows: when ZnO is substituted with Fe ions which are in the 3+ ion valence state, they replace the Zn2+ ions and occupy zinc interstitial sites. The substituted Fe atoms cannot be easily ionized like zinc atoms because of their higher ionization potential (2 eV). Hence, the donor concentration is lowered by the addition of Fe which results in the decrease of electrical conductivity. Thus, a ln σT = −
274
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
Fig. 5 Effect of doping concentration on activation energy for I-phase ZnO samples (950 ◦ C)
higher value of activation energy (Ea ) is found for Fedoped ZnO samples and it increases with increasing iron concentration (x)[31] . 4. Conclusion Fe-doped ZnO powders were synthesized by using the solid state reaction method with metallic Fe powder and commercially pure ZnO powder. The XRD analysis results indicated that all powder samples in the Fe–ZnO binary system (0.25≤x≤3 mol%, at 950 ◦ C) had a wurtzite structure, namely the solubility limit of Fe in the ZnO lattice at this temperature was 3 mol%. The lattice constants of the I-phase ZnO samples decreased with Fe doping concentration. The grain size of the I-phase ZnO samples increased with both heat treatment and the amount of Fe doping. For pure and Fe-doped ZnO samples (950 ◦ C), electrical conductivity increased with heat treatment and decreased with increasing doping concentration. In addition, the activation energy of I-phase ZnO samples increased with increasing dopant concentration. Acknowledgement This work was financially supported by the Research Foundation of Erciyes University (Kayseri, Turkey). REFERENCES [1 ] S.Y. Chu, T.M. Yan and S.L. Chen: Ceram. Int., 2000, 26(7), 733. [2 ] V.R. Shinde, T.P. Gujar, C.D. Lokhande, R.S. Mane and S.H. Han: Mater. Chem. Phys., 2006, 96(2), 326. [3 ] S.J. Pearton, D.P. Norton, K. Ip and Y.W. Heo: Prog. Mater. Sci., 2005, 50(3), 293. [4 ] Z. Zhou, K. Kato, T. Komaki, M. Yoshino, H. Yukawa, M. Morinaga and K. Morita: J. Eur. Ceram. Soc., 2004, 24, 139. [5 ] B.J. Copa: Electrical, Chemical and Structural Characterization of Au-Schottky Contacts on Remote Plasma-Treated n-Type {0001}, Ph.D. Thesis, North Carolina State University, 2003. [6 ] D.W. Zeng, C.S. Xie, B.L. Zhu, B.L. Song and H. Wang: Mater. Sci. Eng. B, 2003, 104(1–2), 68. [7 ] R. Maity, S. Kundoo and K.K. Chattopadhyay: Sol. Energy Mater. Sol. Cells, 2005, 86(2), 217.
[8 ] P. Nunes, E. Fortunato, P. Vilarinho and R. Martins: Vacuum, 2002, 64(3–4), 281. [9 ] S.V. Bhat and F.L. Deepak: Solid State Commun., 2005, 135(6), 345. [10] L. Wang, Y. Pu, W. Fang, J. Dai, C. Zheng, C. Mo and C. Xiong: Thin Solid Films, 2005, 491(1–2), 323. [11] A. Sawalha, M. Abu-Abdeen and A. Sedky: Physica B, 2009, 404(8–11), 1316. [12] X. Mao, W. Zhong and Y. Du: J. Magn. Magn. Mater., 2008, 320(6), 1102. [13] O.K. Scherger, H. Schmid, N. Vanderschaeghe, F. Wolf and W. Mader: J. Am. Ceram. Soc., 2007, 90(12), 3984. [14] A.J. Chen, X.M. Wu, Z.D. Sha, L.J. Zhuge and Y.D. Meng: J. Phys. D: Appl. Phys., 2006, 39, 4762. [15] M. Zhang, L.M. Cao, F.F. Xu, Y. Bando and W.K. Wang: Thin Solid Films, 2002, 406(1–2), 40. [16] G.Y. Ahn, S.I. Park, S.J. Kim, B.W. Lee and C.S. Kim: IEEE T. Magn., 2005, 41, 2730. [17] R. Wang and A.W. Sleight: J. Solid State Chem., 1996, 125, 224. [18] R. Kumar, A.P. Singh, P. Thakur, K.H. Chae, W.K. Choi, B. Angadi, S.D. Kaushik and S. Patnaik: J. Phys. D: Appl. Phys., 2008, 41(15), 155002. [19] H. Colak and O. Turkoglu: J. Mater. Sci. Technol., 2011, 27(10), 944. ¨ Ozg¨ ¨ ur, Y.I. Alivov, C. Liu, A. Teke, M.A. [20] U. Reshchikov, S. Doˇ gan, V. Avrutin, S.J. Cho and H. Morko¸c: J. Appl. Phys., 2005, 98, 041301. [21] D. Karmakar, S.K. Mandal, R.M. Kadam, P.L. Paulose, A.K. Rajarajan, T.K. Nath, A.K. Das, I. Dasgupta and G.P. Das: Phys. Rev. B, 2007, 75(14), 144404. [22] H. Liu, J. Yang, Y. Zhang, Y. Wang and M. Wei: Mater. Chem. Phys., 2008, 112(3), 1021. [23] R.C. Weast: Handbook of Chemistry and Physics, 56th edn, CRC Press, 1975–1976. [24] Y. Zhang, L. Wu, H. Li, J. Xu, L. Han, B. Wang, Z. Tuo and E. Xie: J. Alloy. Compd., 2009, 473(1–2), 319. [25] X.X. Wei, C. Song, K.W. Geng, F. Zeng, B. He and F. Pan: J. Phys. Condens. Matter., 2006, 18, 7471. [26] K. Park and K.Y. Ko: J. Alloy. Compd., 2007, 430(1– 2), 200. [27] S. Yilmaz, O. Turkoglu and Belenli: Mater. Chem. Phys., 2008, 112, 472. [28] M. Girtan, G.G. Rusu, S.D. Seignon and M. Rusu: Appl. Surf. Sci., 2008, 254(13), 4179. [29] J. Han, A.M.R. Senos and P.Q. Mantas: J. Eur. Ceram. Soc., 2002, 22, 49. [30] M.S. Hossain, R. Islam and K.A. Khan: Chalcogenide Lett., 2008, 5, 1. [31] M.L. Dinesha, H.S. Jayanna, S. Ashoka and G.T. Chandrappa: J. Alloy. Compd., 2009, 485(1–2), 538. [32] N. Ohashi, Y. Terada, T. Ohgaki, S. Tanaka, T. Tsurumi, O. Fukunage, H. Haneda and J. Tanaka: Jpn. J. Appl. Phys., 1999, 38, 5028. [33] F. Oba, I. Tanaka and H. Adachi: Jpn. J. Appl. Phys., 1999, 38, 3569. [34] P.Q. Mantas and J.L. Baptista: J. Eur. Ceram. Soc., 1995, 15, 605. [35] J. Han, P.Q. Mantas and A.M.R. Senos: J. Eur. Ceram. Soc., 2001, 21(10–11), 1883. [36] Y. Natsume and H. Sakata: Thin Solid Films, 2000, 372(1–2), 30.
Synthesis, Crystal Structural and Electrical Conductivity Properties of Fe-Doped Zinc Oxide Powders at High Temperatures Hakan C ¸ olak1,2)† and Orhan T¨ urkoˇglu1) 1) Department of Chemistry, Science of Faculty, University of Erciyes, Kayseri, 38039, Turkey 2) Deparment of Chemistry, Science of Faculty, University of C ¸ ankırı Karatekin, Ballıca Campus, C ¸ ankırı, 18200, Turkey [Manuscript received December 31, 2010, in revised form March 5, 2011]
The synthesis, crystal structure and electrical conductivity properties of Fe-doped ZnO powders (in the range of 0.25–15 mol%) were reported in this paper. I-phase samples, which were indexed as single phase with a hexagonal (wurtzite) structure in the Fe-doped ZnO binary system, were determined by X-ray diffraction (XRD). The solubility limit of Fe in the ZnO lattice is 3 mol% at 950 ◦ C. The above mixed phase was observed. And the impurity phase was determined as the cubic-ZnFe2 O4 phase when compared with standard XRD data using the PDF program. This study focused on single I-phase ZnO samples which were synthesized at 950 ◦ C because the limit of the solubility range is the widest at this temperature. The lattice parameters a and c of the I-phase decreased with Fe-doping concentration. The morphology of the I-phase samples was analyzed with a scanning electron microscope. The grain size of the I-phase samples increased with heat treatment and doping concentration. The electrical conductivity of the pure ZnO and single I-phase samples was investigated using the four-probe dc method at 100–950 ◦ C in air atmosphere. The electrical conductivity values of pure ZnO, 0.25 and 3 mol% Fe-doped ZnO samples at 100 ◦ C were 2×10−6 , 1.7×10−3 and 6.3×10−4 S·cm−1 , and at 950 ◦ C they were 3.4, 8.5 and 4 S·cm−1 , respectively. KEY WORDS: II-VI semiconductors; Zinc oxide and doped zinc oxide; Four point probe method
1. Introduction Zinc oxide (ZnO) powders are very important materials due to the many interesting properties inherent in this material, such as dielectric, piezoelectric, pyroelectric, semiconducting, acousto-optic, optical, electro-optical, nonlinear optical, photoelectrochemical and electrical properties[1] . ZnO is one of the most versatile and technologically interesting semiconducting materials because of its typical properties such as resistivity control over the range of 10−3 –10−5 Ω·cm, transparency in the visible range, high electrochemical stability, direct band gap, absence of toxicity and abundance in nature[2] . ZnO normally occurs in the hexagonal wurtzite crystal structure with a=0.32488 nm and c=0.52066 nm in the standard data (JCPDS, 36-1451). Electron doping in † Corresponding author. Assist. Prof., Ph.D.; Tel.: +90 506 9868893; Fax: +90 376 2181031; E-mail address: hakan [email protected] (H. C ¸ olak).
nominally undoped ZnO has been attributed to Zn interstitials and oxygen vacancies[3] . High purity ZnO crystals exhibit strong n-type conductivity with the electrons moving in the conduction band as charge carriers. This characteristic has traditionally been attributed to native defects such as oxygen vacancies and zinc interstitials[4,5] . ZnO has both good electronic and optical properties because of a stoichiometric deviation due to the existence of intrinsic defects such as O vacancies and Zn interstitials. However, these properties of pure ZnO are unstable due to the adsorption of atmospheric oxygen, which decreases the conductivity, and they cannot meet the increasing needs for present day applications. To stabilize them against such changes and enhance the properties of the ZnO, doping is necessary and this purpose was achieved by adding some dopants[6,7] . Moreover, doping leads to an increase in the conductivity of ZnO. ZnO doping was achieved by replacing Zn2+ atoms with the atoms of the dopant elements. The efficiency of the dopant
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
element depends on its electronegativity and the difference between its ionic radius and the ionic radius of zinc[8] . Extensive studies have been carried out to modify the properties of ZnO for different applications. Doping with transitional metal elements leads to many interesting properties of ZnO[9] . In addition, thermal annealing is a widely used method to improve crystal quality and to study structural defects in materials. For semiconductors, thermal annealing is also used to activate dopants. During the annealing process, dislocations and other structural defects will move in the material and adsorption/decomposition will occur on the surface, thus the structure and the stoichiometric ratio of the material will change, which influences the electrical and optical properties of ZnO. The structural, electrical and optical properties are vital for semiconductor devices. It is necessary to investigate how these properties are affected by thermal annealing[10] . Although there are a few reports on the doping elements in the ZnO varistor, the effects of some dopants, such as Fe, on the electrical conductivity of the ZnO varistor still remain unclear. This is because doping effects have been studied in quite different systems under different experimental conditions[11] . In this work, metallic iron-doped ZnO powders were synthesized at high temperatures by using the solid state reaction method. Then, their structural properties were studied and the dc electrical conductivity of the ZnO solid solution samples was measured at high temperatures. The solid state reaction was chosen because of its reproducibility, easy control and sufficent products for measurement[12] . 2. Experimental Synthesized iron (Fe) doped ZnO powders at high temperatures were measured by using the standard solid state reaction method. At first, solid mixtures of Fe-ZnO (0.25≤ x ≤15 mol%) were prepared by mixing and homogenizing with stoichiometric amounts of pure ZnO and metallic Fe in an agate mortar. As starting materials, commercially pure ZnO powders (99.9% Fluka) and metallic iron powders (99.5% Alfa Aesar) were used. The powder mixtures were first calcined at 600 and 650 ◦ C for 24 h. After grinding and homogenization, the pre-annealed mixtures were heat treated at 700, 750, 800, 850, 900 and 950 ◦ C, respectively. All these heat treatments were made in air atmosphere for 48 h, in alumina crucibles and without any compaction procedure. At the end of each heat treatment procedure, the annealed powders were slowly cooled in the furnace by switching it off (uncontrolled). Before and after heat treatment, each of the samples was ground in an agate mortar to homogenize powder size. The samples were then analyzed with X-ray powder diffraction (XRD). The XRD data of the samples were recorded on a computerinterfaced Bruker AXS D8 advanced diffractometer operating in Bragg–Brentano geometry (CuKα radi-
269
ation, graphite monochromator, 40 kV and 40 mA) over a 10 deg.≤2θ≤90 deg. angular range. The divergence and receiving slits of 1 and 0.1mm, respectively, were located on the diffractometer. The diffraction patterns were scanned in 0.002 deg. (2θ) steps and the diffracted beams were counted with a NaI(Tl) scintillation detector. The Diffrac Plus and Win-Index programs were used to obtain information about the crystal structures of the samples. Then, for a comparison with standard data, the PDF program (Maint Powder Diffraction Database Manager Software) was used. In all cases, the XRD patterns of the samples, which were obtained as single phase, could be indexed on the basis of a hexagonal cell in the Fe–ZnO system. The single phase was named as I-phase. The other samples which were not indexed with a hexagonal crystal structure were named as II-phase. Samples which were out of the solubility range were heterogeneous solid mixtures (I+II). The morphologies of both annealed (950 ◦ C) and non-annealed undoped ZnO samples and I-phase samples (950 ◦ C) were observed by scanning electron microscopy (SEM, LEO 440) operated at 20 kV. Before SEM measurement, the I-phase samples in the Fe–ZnO system were pressed into pellets of 0.1 cm in thickness (t) and 1.3 cm in diameter (d) under ∼5 ton pressure (∼369 MPa) and were calcined at 950 ◦ C for 24 h in air. After heat treatment, the XRD measurements showed that no phase change in the hexagonal samples was observed. The average grain sizes of the I-phase samples were calculated by the image-pro plus 5.0 program from the SEM micrographs. The electrical conductivity (σdc ) measurements of the I-phase samples (950 ◦ C) were made using the standard four-probe dc method. The circular samples with a hexagonal structure, which were mentioned above as the pelleted samples, were used for the electrical conductivity measurements. To reduce contact resistance, fine platinum wires were directly attached to the surface of the samples. The ohmic character of the wire contacts was checked prior to each measurement. The contacts were positioned symmetrically with respect to the center of the circular pellet and the contact separations (s) were 0.2 cm. The conductivity measurements were made between 100 and 950 ◦ C for I-phase samples synthesized at 950 ◦ C. The increase in furnace temperature was initial in 100 ◦ C (in the range 100–400 ◦ C) steps, then in 50 ◦ C (450–950 ◦ C) steps in air. During the measurements, the sample temperature was determined by a thermocouple located 5 mm away from the sample. This thermocouple had a cold (0 ◦ C) junction. All experimental data were recorded by a Keithley 2400 sourcemeter and a Keithley 2700 electrometer which were controlled by a computer. 3. Results and Discussion 3.1 Structural properties Because data on the solubility of Fe ions in ZnO
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
270
Table 1 Observed phases in the binary system of Fe-doped ZnO (in the range of 0.25–15 mol%) x (Fe addition)/mol% 0.25 0.50 0.75 1 2 3 4 5 6 7 8 9 10 11 12 13 14 I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I I I I I I I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I+II I—Fe-doped ZnO (hexagonal structure), II—cubic-ZnFe2 O4 , I+II—heterogeneous solid mixtures
Temp./◦ C 700 750 800 850 900 950 Notes:
Fig. 1 XRD patterns of Fe-doped ZnO: (a) undoped ZnO (950 ◦ C), (b) 0.25 mol% Fe-doped ZnO, (c) 3 mol% Fe-doped ZnO, (d) 6 mol% Fe-doped ZnO, (e) 15 mol% Fe-doped ZnO (950 ◦ C) (∗: the peaks of II-phase, ZnFe2 O4 )
are not known, annealing experiments at varying temperatures were performed[13] . Fig. 1 shows the XRD patterns of pure ZnO (950 ◦ C, 48 h), 0.25, 3, 6 and 15 mol% Fe-doped ZnO samples (950 ◦ C, 48 h). In this figure, all the XRD patterns show typical peak patterns, which can be indexed as the ZnO wurtzite structure. The peak appears at 2θ=34.44 deg., which corresponds to the (002) directions of the ZnO hexagonal wurtzite structure[14] . For 6 and 15 mol% Fedoped ZnO samples, XRD patterns show extra peaks, which indicate ZnFe2 O4 , (PDF No. 01–1108)[15–18] . The XRD patterns of other single I-phase samples, which were synthesized at other temperatures, were quite similar to the patterns of the I-phase samples given in Fig. 1. At 950 ◦ C, the XRD measurements revealed that doping with more than 3 mol% caused the coexistence of the I and II phase. No indication of iron metal impurities was observed in the samples. In the Fe–ZnO binary system, the observed single phases and heterogeneous solid mixtures depending on the reaction temperature and amounts of Fe doping are presented in Table 1. The widest range solubility in the ZnO lattice of Fe ions is at 950 ◦ C, and at this temperature the solubility limit is 3 mol%. At lower temperatures the solubility range is lower than
15 I+II I+II I+II I+II I+II I+II
950 ◦ C. The low solubility of Fe may be satisfyingly explained by the heterovalence of the Fe ions in the ZnO host lattice[13] . The mobility of Fe ions which diffuse in ZnO tetrahedral sites increases with temperature. Heat treatment temperature was found to have an effect on the solubility of Fe ions in ZnO[19] . In addition, both the ionic radii and the valence state are important factors in determining the solubility of the dopants. A smaller deviation of these factors from those of Zn2+ would be favorable for the dopants to have a higher solubility. This study focused on single I-phase ZnO samples synthesized at 950 ◦ C due to the limit of the solubility range, which is the widest at this temperature. The other I-phase samples synthesized at temperatures lower than 950 ◦ C, and the multiphase samples (I+II) were excluded from this study. For the wurtzite ZnO structure, the lattice constants determined at room temperature by various experimental measurements and theoretical calculations were in good agreement. The lattice constants of the samples mostly ranged from 0.32483 to 0.32501 nm for the a parameter and from 0.52023 to 0.52050 nm for the c parameter. The deviation from that of the ideal wurtzite crystal is probably due to lattice stability and ionicity[20] . Table 2 shows the dependence of the Table 2 Unit cell parameters of I-phase ZnO samples at 950 ◦ C x (Fe addition)/mol% 0.25 0.50 0.75 1 2 3
Unit cell parameters/nm a c 0.32501 0.52050 0.32499 0.52047 0.32493 0.52033 0.32488 0.52026 0.32488 0.52025 0.32483 0.52033
lattice parameters a and c, respectively, on the addition ratio of metallic Fe. The lattice parameters of the Fe-doped ZnO system are smaller than those of pure ZnO, both a and c parameters decrease with the substitution of Fe ions with Zn ions. Fe ions are substituted with Zn2+ ions which are in the tetrahedral sites of the wurtzite structure of ZnO. It is considered that valence state of Fe in ZnO is both +2 and +3, namely Fe in the ZnO matrix exists in a mixed valence state. It has been reported that local magnetic probes like electron paramagnetic resonance and
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
271
Fig. 2 SEM images of I-phase samples for pure ZnO (non-annealed) (a), pure ZnO (annealed) (b), 0.25 (c), 0.50 (d), 0.75 (e), 1 (f), 2 (g) and 3 mol% (h) Fe-doped ZnO samples at 950 ◦ C
M¨ ossbauer spectroscopy indicate the presence of Fe in ZnO in both valence states, Fe2+ and Fe3+ . And the presence of Fe3+ in ZnO is due to possible hole doping in the system by cation (Zn) vacancies[21] . It has also been reported in another study[22] that usually, if Fe is present in the substitutional site in a defect-free ZnO crystal, the valence state of Fe will be +2. In this report, also, the X-ray photoelectron spectroscopy (XPS) result confirmed the presence of uncoupled Fe3+ within the sample. In the Fe–ZnO binary system, Fe3+ ions are present due to the existence of cation vacancy. Cation vacancy near Fe can promote Fe2+ into Fe3+ and also mediate the Fe2+ –Fe2+ exchange interaction. Since the transi-
tion metal (TM) doping percentage is slightly on the higher side toward the cationic percolation threshold, Fe2+ –Fe3+ exchange, although being less in number in comparison to the Fe2+ –Fe2+ interaction, may also be possible[22] . However, the ionic radius (fourcoordinated) of the Fe2+ ion is 0.074 nm and that of the Fe3+ is 0.064 nm while the ionic radius of the Zn2+ is 0.074 nm[23] . Therefore, the decrease in the lattice parameters a and c can be explained by the ionic radii difference[24] . Fig. 2 shows the SEM micrographs of both annealed (950 ◦ C) and non-annealed undoped ZnO and I-phase ZnO samples (950 ◦ C). In the SEM micrographs of undoped ZnO samples (Fig. 2(a) and (b)),
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
272
there is a homogeneous grain distribution. From these images the average grain sizes were calculated and were found to be 0.63 and 1.03 μm, respectively[19] . The grain sizes of the annealed sample are obviously larger than those of the non-annealed samples, which could be caused by strain relaxation. The annealing process clearly produces a recovery of the crystal structure and an increase in the grain size[1,25] . As can be seen in the SEM micrographs, the grains of the I-phase samples are homogeneously distributed. The grain size of all synthesized Fe-doped ZnO samples are given in Table 3. From the SEM images, Table 3 Grain size values of I-phase ZnO samples at 950 ◦ C x (Fe addition)/mol% ZnO∗ ZnO 0.25 0.50 0.75 1 2 3 Note: ∗ —non-annealed
Grain size/μm 0.67 1.03 1.32 1.46 1.73 1.93 2.51 2.60
the grain size of the Fe-doped ZnO samples slightly increases with the addition of Fe because the substitution of Fe2+ and Fe3+ for Zn2+ increases the activity of ZnO by means of distortion of the ZnO lattice, which is beneficial to grain growth in Fe-doped ZnO samples[26] . The grains are regular in size and shape. Some pores can be noticed in the photographs. When increasing the dopant concentration to 3 mol%, the porosity increased[24] . In addition, the surface of the samples is smooth. 3.2 Electrical properties The total electrical conductivity is calculated using the following equation: σT =
I −1 G V
(1)
where G is the geometric resistivity correction factor[27] . Both the values and the variation of the electrical conductivity of the pelleted powder ZnO samples are in connection with their structure and its changes. On the other hand, thermal treatments of the respective samples modify their structural characteristics and consequently, their electrical properties. On this basis, the study of the temperature dependence of the electrical properties of ZnO samples, may offer useful information about the possible changes of the structural characteristics of ZnO samples[28] . The temperature dependence of electrical conductivity (σdc ) of the I-phase ZnO pelleted powder samples at 950 ◦ C was studied in the temperature range of 100–950 ◦ C. logσ T –103 /T graphs were drawn in order to show the conductivity change with respect to temperature. The graphs were compared with each other
and the conductivity characteristics of the samples were evaluated. The electrical conductivity plots vs temperature of undoped ZnO, 0.50, 0.75 and 2 mol% Fe-doped ZnO (950 ◦ C) samples are given in Fig. 3. The graphic curves of the other single I-phase samples were quite similar to the curves of the single I-phase samples given in this figure. As seen in Fig. 3, the electrical conductivity of I-phase samples increases with increasing temperature and shows the semiconducting behavior of pure and Fe-doped ZnO samples. It is well known that the electrical conductivity of ZnO samples is controlled by intrinsic defects (oxygen vacancy and interstitial zinc atoms) generated at high temperature[11] . Also, for ZnO samples, increasing electrical conductivity with temperature can be explained by the following: (2) ZnO ↔ Znxi + 1/2O2 (g) Znxi ↔ Zn•i + e−
(3)
Zn•i
(4)
↔
Zn•• i
−
+e
After ionization, carrier concentration is increased by two extra electrons[29] . The electrical conduction of the doped ZnO follows a mechanism in which the electron or hole hops from one localized site to the next. Whenever it is transferred to another site, the surrounding molecules respond to this perturbation with structural changes and the electron or hole is temporarily trapped in the potential well leading to atomic polarization. The electron resides this site until it is thermally activated to migrate to another site. Another aspect of this charge hopping mechanism is that the electron or hole tends to associate with local defects[30] . It is observed that conductivity slightly decreases with increasing Fe concentration (Fig. 4). It is well known that the electrical conductivity of ZnO samples at room temperature is due to intrinsic defects created by oxygen vacancies. These defects introduce donor states in the forbidden band slightly below the conduction band and hence result in the conducting behavior of ZnO. The electrical conductivity is controlled by the intrinsic defects generated during synthesis and by the presence of dopants. It is known that during the sintering process at high temperature, oxygen vacancies will be produced which are responsible for electrical conductivity. From the observed result of decrease in conductivity of ZnO on Fe doping, it seems that the Fe-doping affects the defect chemistry of the ZnO. Therefore, It is believed that Fe in ZnO acts as a deep donor and decreases the concentration of intrinsic donors. This reduction in intrinsic donor concentration increases with increasing Fe content, which in turn decreases the electrical conductivity[31] . Many researchers have studied the effects of 3d transition metal impurities on the electrical conductivity of the ZnO varistor. They investigate that 3d transition metals could enhance the excess oxygen concentration in the grain boundary region and a
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
273
Fig. 3 Electrical conductivity plots for undoped ZnO (a), 0.25 (b), 0.75 (c) and 2 mol% (d) Fe-doped ZnO (950 ◦ C)
is expected between the nearest donors at low temperatures. The total electrical conductivity can be expressed as follows; σdc = σ0 exp(−Ea /kT )
(5)
where σdc is the electrical conductivity at any temperature, Ea is the activation energy, which corresponds to the energy difference between the donor level and the conduction level, σ0 is the pre-exponential factor, k is the Boltzmann constant, and T is the absolute temperature[8,11,30,36] . The activation energy can be calculated from the conductivity equation shown below, Fig. 4 Electrical conductivity variation for Fe-doped ZnO I-phase ZnO samples (950 ◦ C)
potential barrier is formed preferentially. Therefore, the electrical conductivity of Fe-doped ZnO samples is apparently lower than that of undoped ZnO samples, and the grain boundary is more resistive than the grain[32–34] . Also, it was observed from the SEM micrographs that Fe-doping produces a large number of pores which may be the main reason for the significant decrease in the conductivity of the considered samples[35] . In general, for a semiconducting material, dc conductivity increases exponentially with temperature indicating that conductivity is a thermally activated process. For doped semiconductors with low concentration of donors, the hopping transport of carriers
Ea 1 + ln σ0 (6) k T The slope of the linear part of the Arrhenius curve of the lnσT –1/T graph is equal to –Ea /k. For the temperature range the I-phase ZnO samples have Arrhenius-type electrical conductivity. From these curves, drawn for all I-phase ZnO samples, the activation energies Ea were calculated (Fig. 5). The increase in the activation energy of Fe-doped ZnO can be explained as follows: when ZnO is substituted with Fe ions which are in the 3+ ion valence state, they replace the Zn2+ ions and occupy zinc interstitial sites. The substituted Fe atoms cannot be easily ionized like zinc atoms because of their higher ionization potential (2 eV). Hence, the donor concentration is lowered by the addition of Fe which results in the decrease of electrical conductivity. Thus, a ln σT = −
274
H. C ¸ olak et al.: J. Mater. Sci. Technol., 2012, 28(3), 268–274.
Fig. 5 Effect of doping concentration on activation energy for I-phase ZnO samples (950 ◦ C)
higher value of activation energy (Ea ) is found for Fedoped ZnO samples and it increases with increasing iron concentration (x)[31] . 4. Conclusion Fe-doped ZnO powders were synthesized by using the solid state reaction method with metallic Fe powder and commercially pure ZnO powder. The XRD analysis results indicated that all powder samples in the Fe–ZnO binary system (0.25≤x≤3 mol%, at 950 ◦ C) had a wurtzite structure, namely the solubility limit of Fe in the ZnO lattice at this temperature was 3 mol%. The lattice constants of the I-phase ZnO samples decreased with Fe doping concentration. The grain size of the I-phase ZnO samples increased with both heat treatment and the amount of Fe doping. For pure and Fe-doped ZnO samples (950 ◦ C), electrical conductivity increased with heat treatment and decreased with increasing doping concentration. In addition, the activation energy of I-phase ZnO samples increased with increasing dopant concentration. Acknowledgement This work was financially supported by the Research Foundation of Erciyes University (Kayseri, Turkey). REFERENCES [1 ] S.Y. Chu, T.M. Yan and S.L. Chen: Ceram. Int., 2000, 26(7), 733. [2 ] V.R. Shinde, T.P. Gujar, C.D. Lokhande, R.S. Mane and S.H. Han: Mater. Chem. Phys., 2006, 96(2), 326. [3 ] S.J. Pearton, D.P. Norton, K. Ip and Y.W. Heo: Prog. Mater. Sci., 2005, 50(3), 293. [4 ] Z. Zhou, K. Kato, T. Komaki, M. Yoshino, H. Yukawa, M. Morinaga and K. Morita: J. Eur. Ceram. Soc., 2004, 24, 139. [5 ] B.J. Copa: Electrical, Chemical and Structural Characterization of Au-Schottky Contacts on Remote Plasma-Treated n-Type {0001}, Ph.D. Thesis, North Carolina State University, 2003. [6 ] D.W. Zeng, C.S. Xie, B.L. Zhu, B.L. Song and H. Wang: Mater. Sci. Eng. B, 2003, 104(1–2), 68. [7 ] R. Maity, S. Kundoo and K.K. Chattopadhyay: Sol. Energy Mater. Sol. Cells, 2005, 86(2), 217.
[8 ] P. Nunes, E. Fortunato, P. Vilarinho and R. Martins: Vacuum, 2002, 64(3–4), 281. [9 ] S.V. Bhat and F.L. Deepak: Solid State Commun., 2005, 135(6), 345. [10] L. Wang, Y. Pu, W. Fang, J. Dai, C. Zheng, C. Mo and C. Xiong: Thin Solid Films, 2005, 491(1–2), 323. [11] A. Sawalha, M. Abu-Abdeen and A. Sedky: Physica B, 2009, 404(8–11), 1316. [12] X. Mao, W. Zhong and Y. Du: J. Magn. Magn. Mater., 2008, 320(6), 1102. [13] O.K. Scherger, H. Schmid, N. Vanderschaeghe, F. Wolf and W. Mader: J. Am. Ceram. Soc., 2007, 90(12), 3984. [14] A.J. Chen, X.M. Wu, Z.D. Sha, L.J. Zhuge and Y.D. Meng: J. Phys. D: Appl. Phys., 2006, 39, 4762. [15] M. Zhang, L.M. Cao, F.F. Xu, Y. Bando and W.K. Wang: Thin Solid Films, 2002, 406(1–2), 40. [16] G.Y. Ahn, S.I. Park, S.J. Kim, B.W. Lee and C.S. Kim: IEEE T. Magn., 2005, 41, 2730. [17] R. Wang and A.W. Sleight: J. Solid State Chem., 1996, 125, 224. [18] R. Kumar, A.P. Singh, P. Thakur, K.H. Chae, W.K. Choi, B. Angadi, S.D. Kaushik and S. Patnaik: J. Phys. D: Appl. Phys., 2008, 41(15), 155002. [19] H. Colak and O. Turkoglu: J. Mater. Sci. Technol., 2011, 27(10), 944. ¨ Ozg¨ ¨ ur, Y.I. Alivov, C. Liu, A. Teke, M.A. [20] U. Reshchikov, S. Doˇ gan, V. Avrutin, S.J. Cho and H. Morko¸c: J. Appl. Phys., 2005, 98, 041301. [21] D. Karmakar, S.K. Mandal, R.M. Kadam, P.L. Paulose, A.K. Rajarajan, T.K. Nath, A.K. Das, I. Dasgupta and G.P. Das: Phys. Rev. B, 2007, 75(14), 144404. [22] H. Liu, J. Yang, Y. Zhang, Y. Wang and M. Wei: Mater. Chem. Phys., 2008, 112(3), 1021. [23] R.C. Weast: Handbook of Chemistry and Physics, 56th edn, CRC Press, 1975–1976. [24] Y. Zhang, L. Wu, H. Li, J. Xu, L. Han, B. Wang, Z. Tuo and E. Xie: J. Alloy. Compd., 2009, 473(1–2), 319. [25] X.X. Wei, C. Song, K.W. Geng, F. Zeng, B. He and F. Pan: J. Phys. Condens. Matter., 2006, 18, 7471. [26] K. Park and K.Y. Ko: J. Alloy. Compd., 2007, 430(1– 2), 200. [27] S. Yilmaz, O. Turkoglu and Belenli: Mater. Chem. Phys., 2008, 112, 472. [28] M. Girtan, G.G. Rusu, S.D. Seignon and M. Rusu: Appl. Surf. Sci., 2008, 254(13), 4179. [29] J. Han, A.M.R. Senos and P.Q. Mantas: J. Eur. Ceram. Soc., 2002, 22, 49. [30] M.S. Hossain, R. Islam and K.A. Khan: Chalcogenide Lett., 2008, 5, 1. [31] M.L. Dinesha, H.S. Jayanna, S. Ashoka and G.T. Chandrappa: J. Alloy. Compd., 2009, 485(1–2), 538. [32] N. Ohashi, Y. Terada, T. Ohgaki, S. Tanaka, T. Tsurumi, O. Fukunage, H. Haneda and J. Tanaka: Jpn. J. Appl. Phys., 1999, 38, 5028. [33] F. Oba, I. Tanaka and H. Adachi: Jpn. J. Appl. Phys., 1999, 38, 3569. [34] P.Q. Mantas and J.L. Baptista: J. Eur. Ceram. Soc., 1995, 15, 605. [35] J. Han, P.Q. Mantas and A.M.R. Senos: J. Eur. Ceram. Soc., 2001, 21(10–11), 1883. [36] Y. Natsume and H. Sakata: Thin Solid Films, 2000, 372(1–2), 30.