Thin Solid Films 473 (2005) 315 – 320 www.elsevier.com/locate/tsf
The electromagnetic interference shielding effect of indium–zinc oxide/silver alloy multilayered thin films Won Mok Kima,*, Dae Young Kub, In-kyu Leeb, Yong Woon Seoc, Byung-ki Cheonga, Taek Sung Leea, In-ho Kima, Kyeong Seok Leea a
Materials Design Lab., Korea Institute of Science and Technology, Seoul 136-791, South Korea Department of Materials Engineering, Hankuk Aviation University, Seoul 421-791, South Korea c ITM, Incorporated, Anyang, Kyunggi-do 431-060, South Korea
Received 28 June 2004; received in revised form 17 August 2004; accepted 25 August 2004 Available online 27 September 2004
Abstract A study was made to examine the electromagnetic interference (EMI) shielding effect of multilayered thin films in which indium–zinc oxide (IZO) thin films and Ag or Ag alloy thin films were deposited alternately at room temperature using a RF magnetron sputtering. The optical, electrical and morphological properties of the constituent layers were analyzed using an ultraviolet-visible photospectrometer, a 4point probe and an atomic force microscopy (AFM), respectively. The EMI shielding effect of the multilayered thin films was also measured using a coaxial transmission line method. A detailed analysis showed that the control of the film morphologies, i.e., the surface roughnesses of the constituent metal layers was essential to an accurate estimate of the electrical and optical properties of multilayered coatings. It was shown that properly designed IZO/Ag alloy multilayered thin films could yield a visible transmission of more than 70%, a sheet resistance of less than 1 V/sq., together with an EMI shielding effect larger than 45 dB in the range from 30 to 1000 MHz. D 2004 Elsevier B.V. All rights reserved. PACS: 73.50; 78.66; 78.20 Keywords: EMI shield; Multilayer; Resistivity; Transmission; Electrical properties and measurements
1. Introduction Recently, multilayered transparent conducting films, consisting of thin films of Ag or Ag alloys sandwiched between transparent conducting oxide films such as indium– tin oxide, tin oxide and zinc oxide have been extensively studied for applications such as transparent conducting electrodes in flat-panel displays, low-emissivity coatings in architectural glasses and electromagnetic interference (EMI) shielding of plasma display panel (PDP) [1–9]. Many of the previous works on the multilayered transparent conducting films were focused on the electrical and optical properties [1–5], and a few studies on the chemical durability of such coatings [6–9]. There appears to have been no report,
* Corresponding author. Tel.: +82 29585384; fax: +82 29585391. E-mail address: [email protected]
(W.M. Kim). 0040-6090/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2004.08.083
however, regarding the EMI shielding efficiency (SE) characteristics of the multilayered coatings in correlation with their electrical and optical properties. This is because the SE of a PDP filter has been measured according to an industrial procedure, which requires the use of a PDP module as a base device to which the PDP filter tested must be attached. Because PDP modules have different electromagnetic field characteristics depending upon the manufacturer’s circuit design, per se, use of a combined device would not make a proper way of comparing among different PDP filters made of different materials. Therefore, in this study, the SE characteristics of the multilayered coatings designed for PDP filter application were examined by use of the coaxial transmission line method described in American Society for Testing and Materials (ASTM)  together with their electrical and optical properties in order to get an insight into the relationship between SE and electrical properties of multilayered coatings.
W.M. Kim et al. / Thin Solid Films 473 (2005) 315–320
2. Experimental details Multilayered coatings containing Ag or Ag alloy sandwiched between indium–zinc oxide (IZO) layers were fabricated using a radio frequency (RF) magnetron sputtering. The sputter chamber was equipped with a load–lock system, and the base pressure of the sputter chamber prior to deposition was made to reach below 105 Pa. The IZO layers were deposited with an oxide mixture target containing 12 wt.% of ZnO. Ag and Ag alloy layers were deposited respectively from a 99.99% pure Ag target and from an Ag– Pd–Cu (APC) alloy target containing 1 wt.% Pd and 0.5 wt.% Cu. All the targets were 50.8 mm in diameter, and a target to substrate distance was kept at 70 mm. The electrical, optical and morphological properties of monolithic IZO films were examined by varying oxygen content in the sputter gas. In addition, monolithic Ag and APC films of different thickness were also prepared to investigate the morphological, electrical and optical properties. The RF sputtering power for IZO deposition was 50 W, and that for Ag and APC was 20 W. During deposition of Ag and APC layers, a high-purity Ar gas was introduced into the chamber at the flow rate of 20 sccm, and the operating pressure was kept at 0.4 Pa by controlling a throttle valve. During deposition of IZO layers, the total gas flow rate and the operating gas pressure were kept the same as above, except for substitution of O2 gas for Ar gas which was up to 3 sccm. Five and seven alternate layers of IZO and Ag and IZO and APC together with three alternate layers of IZO and Ag were prepared (see Table 1) for testing of EMI shielding efficiency (SE) and optical transmission. The deposition parameters of metal and IZO layers were the same as above, except that the oxygen flow rate in sputter gas during deposition of IZO was fixed at 1 sccm. All the depositions Table 1 List of calculated and measured sheet resistance and rms surface roughness of multilayered samples Sample ID
Substrate/IZO(35)/ Ag(15)/IZO(70)/ Ag(15)/IZO(35) Substrate/IZO(35)/ Ag(15)/IZO(70)/ Ag(15)/IZO(70)/ Ag(15)/IZO(35) Substrate/IZO(35)/ APC(15)/IZO(70)/ APC(15)/IZO(35) Substrate/IZO(35)/ APC(15)/IZO(70)/ APC(15)/IZO(70)/ APC(15)/IZO(35) Substrate/IZO(35)/ Ag(50)/IZO(35)
Sheet resistance (V/sq.) Calculated
rms Roughness (nm)
were carried out at room temperature. No posttreatment was performed unless specified otherwise. Si(100) p-type wafers were used as substrates of samples for atomic force microscopy (AFM), and slide glasses for optical and electrical measurements. For measurements of EMI SE, use was made of 3-mm-thick acryl plates machined to have the required dimensions for the reference and the load specimens according to ASTM D4935-89 . Film thickness was measured from the films deposited on strip-masked substrates using a profilometer (model Alphastep 200, Tencor). The morphologies, i.e., surface roughnesses and the structures of the films were examined with an AFM (model CP, Park Scientific) and an X-ray diffractometer (model XRD 3000PTS, Siefert), respectively. The optical properties of the films were measured using a spectrophotometer (model Lambda 35, Perkin-Elmer) for the wavelength in the range from 190 to 1100 nm, and the resistivities of the films were measured using a 4-point probe (model MCP-T600, Mitsubishi Chem.).
3. Results and discussions Prior to fabrication of multilayered structures, the optical, electrical and morphological properties of the component layers of IZO, Ag and APC were investigated. In Fig. 1(a) and (b), the root-mean-squared (rms) surface roughnesses and the specific resistivities (q) of monolithic IZO films are shown as a function of vol.% O2 in sputter gas. The thicknesses of IZO films were in the range from 110 to 137 nm, decreasing with increasing oxygen content in the sputter gas. All the IZO films showed very smooth surface morphologies of rms surface roughness values below 0.4 nm except the film deposited without O2 added to the sputter gas. The IZO films deposited with a sputter gas having 0.5 vol.% O2 were found to have the minimum specific resistivity of 404 AV cm. XRD analysis showed that all the IZO films were amorphous, which is in good agreement with the observation by Aoshima et al. . A representative XRD plot obtained for IZO film deposited with 0.5 vol.% O2 is shown as an inset in Fig. 1(a). The smooth surfaces of the IZO films are thought to stem from the amorphous nature of the films. In Fig. 1(c), the calculated figure of merits (FOM=r/a, where r and a are the electrical conductivity and the average absorption, respectively.) for IZO films are also shown. The average absorption coefficient, a, was obtained by taking an average over the wavelength range from 400 to 700 nm. The absorption at a given wavelength was calculated using the following relationship: a¼
1 ð1 R Þ ln d T
where d, R and T represent a film thickness, optical reflectivity and transmissivity, respectively. The calculated FOM gave the highest value of 0.99 for the films deposited
W.M. Kim et al. / Thin Solid Films 473 (2005) 315–320
Fig. 1. Properties of IZO films deposited with varying O2 content in the sputter gas. (a) rms surface roughness ( ), (b) specific resistivity (E), and (c) the figure of merit (n). The insertion in panel (a) represents the XRD plot obtained for IZO film deposited in 0.5% O2 atmosphere.
with a sputter gas having 0.5 vol.% O2. Because of the highest FOM and the lowest specific resistivity of these films, together with their reasonably good surface morphology, sputter gas with 0.5 vol.% O2 was employed for deposition of IZO films during fabrication of multilayered structures. Fig. 2(a) and (b) show the rms surface roughnesses and the specific resistivities of monolithic Ag and APC films of varying thicknesses. APC films appeared to have very smooth surface morphologies of rms surface roughness values in the range from 0.29 to 0.62 nm. On the other hand, the rms surface roughnesses of Ag films were more than four times as large as those of APC films, lying in the range from 1.8 to 2.3 nm. In Fig. 2(a), the 2-D AFM images of Ag and APC films with thickness of 15 nm were also shown for comparison. The specific resistivity of a 100-nm-thick Ag film was 2.38 AV cm which is about 1.5 times larger than that of a bulk Ag; 1.62 AV cm . The specific resistivity of a 100-nm-thick APC film was 3.3 AV cm, which is about 40% larger than that of Ag film due to the effect of alloying. The specific resistivities of both Ag and APC films increased with decreasing film thickness. It should be observed that APC films maintained larger specific resistivity values than Ag films down to the film thickness of 10 nm but that the trend was reversed with further reduction in film thickness. The latter finding is believed to have much to
do with the particulate nature of thin Ag films. A similar trend was also found in the plot of average total absorption (=1RT, averaged over the wavelength range from 400 to 700 nm) of Ag and APC films versus film thickness. The average total absorptions of Ag films thicker than 15 nm were in the range of 2–2.9%. As the film thickness decreased, the average total absorption increased very rapidly, rising to 4% at 10 nm and reaching as high as 19.2% at 7 nm. On the contrary, the average total absorptions of APC films stayed at a relatively constant value of 5.4–6.3% down to the film thickness of 10 nm, and were about 9.2% for 7-nm-thick APC film. From these results, it should be clear that continuous films were formed for films thicker than 10 nm for both Ag and APC films. In Table 1, a list is shown of the structures of the fiveand the seven-layered samples containing Ag or APC layers examined in this study. The thickness of Ag and APC layers sandwiched between IZO layers was fixed at 15 nm. In each structure, the thickness of the first and the last IZO layers was 35 nm while that of the intermediate IZO layers being 70 nm. The metal layer thickness of 15 nm was chosen to obtain a sheet resistance sufficiently low enough to get a high SE as well as to have a relatively high transmission in the visible range from an assuredly continuous film. The IZO thicknesses in the multilayered structures were chosen to yield roughly appropriate transmissions in the visible range from simulations using the measured optical constants of IZO and 100-nm-thick Ag films. Theoretical transmission
Fig. 2. Dependence of the properties of monolithic Ag (n) and APC (E) films on film thickness. (a) rms surface roughness and (b) specific resistivity. The 2-D AFM images of Ag and APC films with thickness of 15 nm were inserted into panel (a) for comparison.
W.M. Kim et al. / Thin Solid Films 473 (2005) 315–320
spectra were calculated using a standard matrix transfer method. A three-layered sample with a 50-nm-thick Ag layer, Ag3, was also fabricated in order to see the effect of a low sheet resistance on SE characteristics as compared with other samples while neglecting the optical transmission. The calculated and the measured sheet resistances, together with the rms surface roughnesses of the multilayer samples, are also listed in Table 1. Because the multilayers can be considered as a parallel circuit of five and seven resistors for five- and seven-layered structures, respectively, the total sheet resistance of a multilayer, R s, can be calculated using the relationship, X di 1 ¼ Rs qi
where d i and q i are the thickness and specific resistivity of i-th layer in the multilayers. In the calculation, use was made of the measured specific resistivity values of 4.08 and 5.27 AV cm, respectively, for Ag and APC films of 15-nm thick. For IZO films, the specific resistivity value of 404 AV cm determined for a 130-nm-thick IZO film deposited with a sputter gas having 0.5 vol.% O2 was used without taking the thickness effect into account. This is reasonable considering that the IZO thicknesses of 35 and 70 nm used in the multilayers are relatively thick enough and that the major contribution to electrical property comes from the metal layer because of its conductivity almost 100 times higher than that of IZO layer. It would be quite natural that the calculated sheet resistance is increasing in the order of Ag3, Ag7, APC7, Ag5 and APC5. The measured sheet resistance was in the same order as the calculated one. However, as can be seen from Table 1, the measured sheet resistances were larger than the calculated ones, especially for the samples Ag5 and Ag7; larger by almost 26% or more. By contrast, the differences between the measured and the calculated sheet resistances for the samples APC5 and APC7 were much smaller. The measured sheet resistances of five-layered samples were close to those of seven-layered samples, although it is apparent that the samples containing Ag as a metal layer gave slightly smaller sheet resistance than those containing APC. The multilayered samples containing Ag displayed approximately three times larger overall rms roughness than those containing APC. The smaller roughness values of Ag5 and Ag7 than those of single Ag films might be due to a smoothening effect caused by the formation of relatively thick and smooth IZO layers. Shown in Fig. 3 are the plots of the EMI SEs in the far field measured as a function of frequency by use of the coaxial transmission line method described in ASTM designation D4395-89. Despite some scatterings in the data with respect to frequency, it is apparent that the SEs of the seven-layered samples are similar to each other, and those of five-layered samples are also similar to each other. The SE was the highest for Ag3 sample with the smallest sheet
Fig. 3. The EMI shielding efficiencies of the multilayered coatings.
resistance among all the samples. In Fig. 4, the frequencyaveraged shielding efficiencies shown in the Fig. 3 were plotted against the sheet resistance. The line shown in the plot is drawn according to the relation, which was developed to describe SE in the far field for an electrically thin film using good conductor approximation , SEðdBÞ ¼ 20log10 ð1 þ Zo =2Rs Þ
where Z 0 is the impedance of free space (377 V). The frequency-averaged SE measured experimentally could be well represented by the above relation for the limited range of sheet resistance examined in this study. Fig. 5 shows a plot of transmission spectra obtained for the multilayered samples. The transmission was the highest for Ag5 sample, and the lowest for APC7 sample. With increasing number of layers, the spectral range for transmission became narrower. The average transmissions in the wavelength range from 450 to 650 nm were 68%, 59%, 61% and 50% for Ag5, Ag7, APC5 and APC7, respectively. As mentioned previously, the layer structures examined in this study were designed using optical constants of 100-nmthick Ag films, without taking account of the thickness and materials effect. It has been reported that the optical constants of thin films vary with thickness of films . Especially, for metal films, which follow Volmer–Webertype island growth mode in the early stage of thin film growth, the discrepancy in optical constants between bulk and thin films is pronounced . In addition, the optical constants of the APC films would be different from those of Ag films of the same thickness due to alloying effect. In Fig. 5, the calculated transmissions for Ag5 and APC5 samples using the optical constants obtained respectively for 15-nmthick Ag and APC films are compared to the experimentally measured transmission. For comparison, the calculated
W.M. Kim et al. / Thin Solid Films 473 (2005) 315–320
Fig. 4. Relationship between the frequency-averaged shielding efficiency and the sheet resistance of the multilayered coatings.
transmission using optical constants of 100-nm-thick Ag film is also included in the plot, which grossly overestimated the measured values. The calculated transmission for APC5 sample led to a very good fit to the measured transmission. On the other hand, for the Ag5 sample, some discrepancy between the calculated and measured transmission could still be seen. This discrepancy observed in the Ag-containing sample might be due to the errors originating from the rough film surface, of which the roughness is comparable to the dimension of the film thickness. Besides high chemical instability of pure Ag, the large surface roughness values make Ag films unsuitable for the use in multilayered transparent conducting films. Using the optical constants obtained for 15-nm-thick APC film, simulation for optically optimized structures was conducted to yield average transmissions of 77% and 74%, respectively, for five- and seven-layered structures containing 15-nm-thick APC layers in the wavelength range from 450 to 650 nm. Such improvement in optical transmission is believed to be achieved experimentally as well considering
Fig. 6. Postannealing effect of the multilayered coatings on the transmission and the sheet resistance. The symbols at the start of arrow represent the properties of the as-deposited samples, and the symbols at the end of arrow represent the properties after annealing.
the close fit between the calculated and the measured transmission as shown in Fig. 5. Furthermore, there is still more room for further improvement in optical and electrical properties of the multilayered transparent conducting films. Fig 6 shows the variations of the sheet resistances and the average transmissions of the multilayer samples before and after annealing at 250 8C for 30 min in Ar atmosphere. Annealing treatment yielded the reduction of sheet resistance combined with the increase of transmission for all the samples, which is in good agreement with the results observed by Aoshima et al. . The sheet resistances of annealed seven-layered samples were even lower than 1 V/ sq. From the above observations and analysis, it follows that the multilayered PDP filters may be fabricated to have transmission higher than 70%, sheet resistance less than 1 V/sq., yielding in turn, frequency-averaged SE characteristics larger than 45 dB.
Fig. 5. Measured transmission spectra (open symbols) of the multi-layered coatings. Comparison between the measured and the calculated transmission spectra of the five-layered samples are made. Dashed, dotted and solid lines are the transmission spectra calculated by using the optical constants for 100-nm-thick Ag film, 15-nm-thick Ag film, and 15-nm-thick APC film, respectively.
Multilayered coatings containing Ag and APC layers sandwiched between IZO layers were fabricated, and their SE characteristics combined with electrical and optical properties were examined. The behavior of the frequencyaveraged SE of the multilayer films investigated could be well represented by the relationship developed for SE in the far field for an electrically thin film using a good conductor approximation. It was shown that it would be possible to fabricate the multilayered PDP filters with visible transmission higher than 70%, sheet resistance less than 1 V/sq., which would yield a frequency-averaged SE characteristics larger than 45 dB. The analysis on the optical and the electrical properties showed that it was important to have smooth surface of metal layer in order to make an accurate
W.M. Kim et al. / Thin Solid Films 473 (2005) 315–320
estimate of the electrical and the optical properties of multilayer coatings.
References  M. Bender, W. Seelig, C. Daube, H. Frankenberger, B. Ocker, J. Stollenwerk, Thin Solid Films 326 (1998) 67.  A. Kfppel, W. Kriegseis, B.K. Meyer, A. Scharmann, C. Daube, J. Stollenwerk, J. Trube, Thin Solid Films 365 (2000) 139.  A. Kfppel, B. Meyer, J. Trube, Thin Solid Films 392 (2001) 311.  M. Farland, P. Karlsson, C. Charton, Thin Solid Films 392 (2001) 334.  K.H. Choi, Y.J. Kim, Y.S. Lee, H.J. Kim, Thin Solid Films 341 (1999) 152.  E. Ando, M. Miyazaki, Thin Solid Films 392 (2001) 289.
 Y. Aoshima, M. Miyazaki, K. Sato, Y. Akao, S. Takaki, K. Adachi, Jpn. J. Appl. Phys. 39 (2000) 4884.  T. Okamura, F. Yamazaki, R. Oi, M. Yoshikai, M. Koyama, U.S. Patent No. 6104530, 15 Aug. 2000.  S. Takaki, K. Sato, M. Miyazaki, Y. Kawamura, H. Nishimura, U.S. Patent No. 6221520, 24 Apr. 2001.  Standard test Method for Measuring the Electromagnetic Shielding Effectiveness of Planar Materials (D4935-89), ASTM, Philadelphia, 1994.  G.T. Duys, T. Farrel (Eds.), Electrical Resistivity Handbook, Peter Peregrinus, London, UK, 1992, p. 535.  N.F. Colaneri, L.W. Shacklette, IEEE Trans. Instrum. Meas. 41 (1992) 291.  X.S. Miao, T.C. Chong, Y.M. Whang, K.G. Lim, P.K. Tan, L.P. Shi, Jpn. J. Appl. Phys. 38 (1999) 1638.  S. Lee, J. Hong, S. Oh, Jpn. J. Appl. Phys. 36 (1997) 3662.