- Email: [email protected]
Mechano-chemical degradation of flexible electrodes for optoelectronic device applications
Thin Solid Films 549 (2013) 251–257
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Mechano-chemical degradation of flexible electrodes for optoelectronic device applications T.S. Bejitual, N.J. Morris, S.D. Cronin, D.R. Cairns, K.A. Sierros ⁎ Department of Mechanical & Aerospace Engineering, West Virginia University, Morgantown, 26506 WV, USA
a r t i c l e
i n f o
Available online 20 September 2013 Keywords: ITO CNT PET Flexible optoelectronics Fatigue Corrosion
a b s t r a c t The electrical, optical, and structural integrity of flexible transparent electrodes is of paramount importance in the design and fabrication of optoelectronic devices such as organic light emitting diodes, liquid crystal displays, touch panels, solar cells, and solid-state lighting. The electrodes may corrode due to acid-containing pressure sensitive adhesives present in the device stacks. In addition, structural failure may occur due to external applied loading. The combined action and further accumulation of both repeated mechanical loading and corrosion can aggravate the loss of functionality of the electrodes. In this study we investigate, using the design of experimental methods, the effects of corrosion, applied mechanical strain, film thickness, and number of bending cycles on the electrical and structural integrity of indium tin oxide (ITO) and carbon nanotube (CNT) films both coated on polyethylene terephthalate (PET) substrates. In situ electrical resistance measurements suggest that fatigue-corrosion is found to be the most critical failure mode for the ITO-based coatings. For example, the change in ITO electrical resistance increase under fatiguecorrosion (1% strain, 150,000 cycles) is 5.8 times higher than that of fatigue mode alone. On the other hand, a minimum change in electrical resistance of the CNT-based electrodes is found when applying the same conditions. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In recent years, there is a considerable interest in the development of flexible optoelectronic devices since they offer significant advantages such as large area, portability, conformability, robustness, and low cost [1]. Indium tin oxide (ITO) films on polyethylene terephthalate (PET) substrates are routinely used as the transparent conducting flexible electrode in optoelectronic devices such as touch screens, displays, solar cells, and solid state lighting. This is because they exhibit an interesting combination of relatively low electrical resistivity (~10−4 Ω cm) and high transparency (N90%) to visible light [2–5]. However, the stresses in the inorganic/organic (ITO/PET) composite resulting from thermal or mechanical loadings may lead to either cohesive failure of the inorganic layer or adhesive failure at the interface between the layer and the substrate [6]. Despite such significant thermal and mechanical mismatches, the ITO/PET system remains the most commonly used flexible electrode [7]. Many researchers have studied the mechanics of ITO on flexible substrates in the past [8–11]. For example, using a controlled buckling test, Chen et al. [12] investigated the fracture of ITO coated PET with the conductive layer placed both under tension and compression. They found that when ITO is under tension a channeling crack forms at a strain of 1.1%. On the other hand, when it is placed under compression, the film ⁎ Corresponding author. Tel.: +1 304 293 3420; fax: +1 304 293 6689. E-mail address: [email protected] (K.A. Sierros). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.09.044
delaminated, buckled and cracked in a tunneling motion at a strain equal to 1.7%. Failure initiation of conductive inorganic coatings is expressed in terms of crack onset strain (COS). An increase of 10% in electrical resistance is chosen arbitrarily to denote COS [13]. The COS indicates initial destruction of conductive paths through the film due to crack initiation. Alzoubi et al. [14] investigated the high cycle bending fatigue of ITO coated PET in order to establish a baseline for comprehensive reliability studies of ITO thin films on a flexible substrate. The percent change in electrical resistance was measured at specific numbers of cycles. It was found that the bending diameter and the number of bending cycles have a pronounced influence on the conductivity of the ITO layer. In particular, at 200 cycles, the percent change in electrical resistance for samples bent around 5.0 and 8.9 mm mandrel diameters was 860% and 640%, respectively. Furthermore, at 500 cycles, the percent change in electrical resistances for samples bent around the same diameters was 6250% and 3360%, respectively. Koniger and Munstedt [15] built a sophisticated device to investigate the electrical behavior of conductive layers on flexible substrates under oscillatory bending both in tension and under compression. For sputtered ITO coatings on PET substrates, a dramatic increase of the electrical resistance was observed for a bending radius smaller than 14 mm (strain ~ 0.6%) due to cracks spanning the whole sample width. The higher the oscillatory bending amplitude, the more pronounced the increase of ITO electrical resistance. Direct current (DC) magnetron sputtering is a typical method of depositing transparent conducting oxides on flexible substrates at low
252
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
temperatures. It offers large area uniformity, high deposition rate and adequate reproducibility [16]. However, the low temperature used often leads to an amorphous film microstructure. Amorphous ITO films are more susceptible to chemical degradation than their crystalline counterparts [17]. Bejitual et al. [18] studied the electrochemical stability of amorphous ITO films deposited on PET substrate in contact with pressure sensitive, acrylic acid containing, adhesives which are employed in the fabrication of flexible optoelectronic devices. They investigated the corrosion of ITO films using different concentrations of acid solution and immersion times. Electrochemical measurements indicated corrosion activities at concentrations as low as 0.05 M of acrylic acid. In addition, degradation of ITO films increased with increasing immersion time to corrosive media. There are additional studies that report on either mechanical or chemical degradation of ITO films [19–21]. However, there is little research to date addressing the combined effect of both repeated mechanical loading and corrosion on the ITO film's structural integrity. Sierros et al. [22] investigated the stress corrosion cracking of such systems. Their study showed that the combination of static bending stress and corrosion, using acrylic acid, can cause the conductive layer to crack at stresses less than a quarter of those needed for failure with no corrosion. Recently, studies investigating carbon nanotube (CNT) films deposited on polymer substrate using solution-based techniques have been reported. CNT coated polymer systems exhibit relatively low sheet resistance values (~200 Ω/sq) and optical transmittance above 80% [23–26]. Such films are being developed with the aim of replacing brittle transparent conductive oxides due to their abundance, relatively low cost of fabrication and superior mechanical flexibility [27,28]. Sierros et al. [29] studied the mechanical integrity of CNT and ITO films under monotonic and cyclic uniaxial tension with in situ electrical resistance monitoring. They observed that the critical strain for 25% electrical increase is 10 times higher for CNT films than that observed for their ITO counterparts. In addition, Hecht et al. [30] conducted single-point stylus-pen actuation tests on CNT coated PET films for resistive touch screens. They reported that no failure is observed up to 3 million actuations. Furthermore, Trottier et al. [31] studied the cyclic bending and monotonic tensile deformation of CNT films on PET substrates. It was reported that the maximum changes in electrical resistance under cyclic bending (0.7% strain and 2500 cycles) and monotonic tensile deformation (0.5% strain) were less than 0.5% and 5%, respectively. Statistical design of experiments (DOE) has been employed in many engineering and scientific studies to identify and quantify the effect of factors and their interactions on a response variable [32]. When there are several factors in an experiment, a factorial design helps to conduct the experiments and analyze the results in a statistically meaningful way [33]. As the number of factors in full factorial design increases, the number of runs required increases rapidly. In such a case, one can assume that certain high-order interactions are negligible and a fractional factorial design involving fewer than the complete set of runs can be used to obtain information on the main effects and low-order interactions [34]. Alzoubi et al. [35] used DOE methods to study the effects of temperature, humidity, bending radius, film thickness, and frequency on the high cycle bending fatigue of thin copper films on polymer substrates. In this work, we report on the effects of corrosion, applied strain, film thickness, and number of cycles (time) on the electrical and mechanical integrity of ITO and CNT films that are deposited on PET substrates. 2. Experimental 2.1. Materials ITO films (CPFilms Inc., USA) deposited at room temperature on PET substrates (180 μm thick) by DC magnetron sputtering (sputtering
power ≈ 1 kW) were used in this work. During sputtering, the processing gas was argon with a partial pressure of approximately 0.5 Pa whereas oxygen was used as a reactive gas with a partial pressure of approximately 2 MPa. The deposition rate of the ITO film was approximately 5 nm/min. The deposited films, as measured by a stylus profilometer (Veeco Dektak 150), were 70 and 200 nm thick with sheet resistances of 100 and 70 Ω/sq, respectively. The ITO target was composed of 90 wt.% In2O3 and 10 wt.% SnO2. Sheet resistances were measured using a four-point probe system (Electronic Design to Market, USA). Also, CNT films (Unidym, USA) on PET substrates (180 μm thick) were used for comparison purposes. The carbon nanotubes were grown via chemical vapor deposition and subsequently purified via air oxidation followed by acid washing to remove residual metal catalyst and amorphous carbon. The resulting CNTs consisted of largely singlewalled and double-walled tubes. The CNTs were dispersed in a 1% aqueous solution of sodium dodecyl sulfate surfactant at 0.01 wt.% and they were then spray-coated onto the PET sheet which was heated to 100 °C. The films were then rinsed in de-ionized water to remove the residual surfactant. The resulting dry CNT film was approximately 10–20 nm thick as measured by atomic force microscopy. A polymethyl methacrylate (PMMA) layer was coated with a Mayer rod from a 0.5 wt.% solution of PMMA in methyl ethyl ketone over the dry CNT film to enhance the film adhesion to the PET and provide enhanced chemical/ environmental resistance. The dry thickness of the binder layer is ~50 nm, thin enough such that it does not electrically insulate the surface of the CNT film. The polymer coating is not continuous and in some regions the CNT layer is exposed allowing a conducting path. The sheet resistance of the CNT film was approximately 350 Ω/sq with a thickness of 15 nm. 2.2. Electromechanical testing During corrosion testing, non-strained flat U-shaped samples were immersed in a 0.05 M acrylic acid solution, and the change in electrical resistance was monitored in situ. Bending, fatigue, bendingcorrosion, and fatigue-corrosion characterizations were conducted by means of a custom-built apparatus. A schematic of the apparatus can be found in the previous work [36]. During the test, one end of the specimen is fixed to a mandrel made from chlorinated polyvinyl chloride (c-PVC) using a high-density polyethylene (HDPE) plate whereas the other end is clamped between a pair of HDPE plates which are connected to a low tension spring (k = 68.4 N/m). A strip of copper foil is used for electrical contact. A relatively large c-PVC mandrel (R ≫ r) supports the specimen in a horizontal position and prevents it from sagging. A stepper motor (MDrive M-17, IMS) is used to rotate the mandrel at a frequency of 2 cycles/s. During both static bending-corrosion and fatigue-corrosion experiments the HDPE container is filled with a 0.05 M acrylic acid solution. The change in electrical resistance is monitored in-situ using an Agilent 34970A data acquisition/switch unit. Specimens were cut into strips of 10 mm in width and 184–200 mm in length in accordance with the bending diameter used. Mandrels with diameters equal to 29.1, 22.6, and 17.8 mm were used. The approximate bending strains using such mandrels correspond to 0.6, 0.8, and 1.0%, respectively. 2.3. Microscopy Specimen surfaces were observed, after testing, using a Leica optical microscope equipped with a frame grabber (Guppy, Allied Visions Technology). Also, the resulting surface morphology of the film was examined using a JEOL JSM-7600F scanning electron microscope (SEM) equipped with a field emission gun. Before SEM imaging, specimens were sputter-coated with an approximately 5 nm thick Au layer using a Cressington 108 sputter coater.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
253
Table 1 Factors and levels for DOE analysis of fatigue and fatigue-corrosion experiments. Factors
Acid concentration (M) Strain (%) Thickness (nm) Number of cycles
Levels 1
2
3
4
0 0.6 70 0
0.05 0.8 200 50,040
– 1 – 100,020
– – – 150,000
2.4. Design of experiments Full factorial DOE was used to further investigate the effect of various designs, processes, and environmental factors on the functionality of ITO films. Four study factors were considered while change in electrical resistance was selected as a response. The study factors were: acrylic acid concentration, applied strain, film thickness, and number of cycles (NOC) for fatigue and combined fatigue-corrosion experiments. In addition, for bending and bending-corrosion, time was considered as a study factor due to the static nature of these experiments. The levels of each factor chosen are presented in Table 1.
Fig. 2. Critical number of cycles for fatigue and fatigue corrosion of 200 nm thick ITOcoated PET films.
3. Results and discussion
In order to investigate the effect of acid concentration, applied strain, film thickness, and number of cycles on the mechano-chemical degradation of the electrodes under static and dynamic loading, in situ electrical measurements were conducted. Electrical resistance changes of the conducting layer during testing can be associated with loss of coating functionality [37]. A typical response of ITO electrical resistance versus number of cycles and/or time is shown in Fig. 1. During the initial stages of fatigue-only, two regimes of electrical resistance increase are observed, at first an increase of electrical resistance due to changes in dimensions of the compliant substrate, until an equilibrium size is attained, and then a gradual linear increase that may be due to cracking of ITO. Gorkhali et al. [38] reported that the initial increase of electrical resistance for an applied strain of 1.5% is 50–100 cycles. In our case the initial increase, for a 0.5% lower applied strain, is observed to lie around 800–1000 cycles. In all cases, the change in electrical resistance for specimens subjected to fatigue-corrosion is observed to be the highest. On the other hand, the lowest change is observed in the case of corrosion-only experimental conditions. Furthermore, the critical
number of cycles (Fig. 2) for the initial loss of functionality, under fatigue and combined fatigue-corrosion of ITO films, is determined from an increase of 10% in electrical resistance (Fig. 1). In general, specimens under fatigue-corrosion degradation are observed to exhibit a lower critical number of cycles for both 70 and 200 nm thick ITO films. This denotes the significance of the combined effect of fatigue and corrosion, which can be used as an accelerated degradation experimental protocol that can aid towards identifying the interrelation of the critical factors that influence the long-term reliability of the device component. The effect of applied strain on the electrical degradation of ITO films at different degradation conditions was investigated. A common observation is that, for each condition and film thickness, the higher the applied strain the larger the positive change in electrical resistance. For example, we observe that at 150,000 cycles the change in ITO (200 nm) electrical resistance during fatigue-corrosion is 300% for 0.6% strain and 1107% for 1.0% applied strain. This denotes more than a threefold increase. Such phenomena can be related to the increased film crack formation as the applied strain increases. Also, the effect of film thickness on the electrical degradation of ITO films was investigated at different conditions. Higher change in electrical resistance values are observed for the thicker films, suggesting that
Fig. 1. Normalized electrical resistance versus experimental time/number of cycles for ITOcoated PET at 1.0% applied strain and 200 nm film thickness.
Fig. 3. Normalized electrical resistance versus experimental time/number of cycles for CNT-deposited PET at 1.0% applied strain.
3.1. In situ electrical resistance measurements
254
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
Table 2 Analysis of variance (ANOVA) for fatigue and fatigue-corrosion. Source
DF
Conc. 1 Strain 2 Thickness 1 NOC 3 Conc. ∗ strain 2 Conc. ∗ thickness 1 Conc. ∗ NOC 3 Strain ∗ thickness 2 Strain ∗ NOC 6 Thickness ∗ NOC 3 Conc. ∗ strain ∗ thickness 2 Conc. ∗ strain ∗ NOC 6 Conc. ∗ thickness ∗ NOC 3 Strain ∗ thickness ∗ NOC 6 Error 6 Total 47 S = 30.81, R-sq = 99.88%, R-sq (adj) = 99.06%.
Seq SS
Adj SS
Adj MS
F
P
1470543 730137 58334 1056107 396287 16322 565061 9949 261251 19637 6410 145552 5585 8380 5697 4755252
1470543 730137 58334 1056107 396287 16322 565061 9949 261251 19637 6410 145552 5585 8380 5697
1470543 365068 58334 352036 198144 16322 188354 4975 43542 6546 3205 24259 1862 1397 950
1548.74 384.48 61.44 370.76 208.68 17.19 198.37 5.24 45.86 6.89 3.38 25.55 1.96 1.47
0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.05 0.00 0.02 0.10 0.00 0.22 0.33
they are influenced, more than their thinner counterparts, by the pure mechanical or the mechano-chemical imposed degradation. It is believed that the higher concentration of pre-existing flaws and defects in the thicker layer may play a role for such a response. On the other hand, it has been reported in the past that thinner layers of ITO can better withstand the applied strain than thicker layers [19]. However, the increase in resistance is not observed to be as large as the one observed during the application of different strains. This indicates the higher significance of applied strain as a degradation factor. Furthermore, CNT films are tested under the same conditions. In contrast to the ITO film behavior, CNT-based electrodes exhibit a different trend as shown in Fig. 3. The change in electrical resistance is observed to be negative, indicating enhanced electrical conductivity of the CNT films. In the presence of a combined mechano-corrosive environment this negative change becomes more pronounced. It is believed that the observed behavior is due to the effect of the acrylic acid solution. The role of acid treatment on CNT networks has been reported in the past. The acid further treats and enhances the metallicity of the CNTs as suggested by both Geng et al. and Nirmalraj et al. [39,40]. In addition, Hecht et al. [41] reported that immersion of CNT films on polycarbonate substrates in 0.1 M acrylic acid solution leads to a negative change in electrical resistance (−2.5%). In our case, the combined effect of fatigue and acidic corrosion leads to increased binder cracking and, therefore, to a larger CNT area exposure. This in turn leads to an
increased number of acid-treated CNT bundles and, therefore, higher film conductivity. However, the change in electrical resistance is not as significant as in the ITO film case. It is therefore suggested that CNT films can better withstand externally applied strains than their ITO counterparts under the combined actions of fatigue and corrosion. 3.2. Design of experiments A further investigation of the effects of corrosion, applied strain, film thickness, and number of cycles (NOC) on the electrical degradation of ITO films was conducted using full factorial design of experiments. Table 2 shows the analysis of variance (ANOVA) and presents the error sum of squares and P-value of the main effects and their twoway and three-way interactions under fatigue and fatigue-corrosion. All design factors and their two-way interactions are found to be significant since their P-values at the 95% confidence level is less than 0.05. However, three-way interactions involving thickness are observed to be insignificant. From the main effect graphs (Fig. 4), the most significant effects are acid concentration, strain, and NOC. High initial rate of change of ITO electrical resistance is observed for both strain and NOC. The significant contribution of corrosion on the change in electrical resistance increase is shown by the two-way interaction plots (Fig. 5). Also, ANOVA was performed for bending and bending-corrosion experiments. All design factors and their two-way interactions are
Fig. 4. Main effects plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
255
Fig. 5. Interaction plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
observed to be significant. However, only one three-way interaction involving thickness is observed to be insignificant. In addition, comparable trends of main effects and two-way interactions are observed for bending and bending-corrosion as in fatigue and fatigue-corrosion. However, the maximum change in electrical resistance under bendingcorrosion is 59% whereas under fatigue-corrosion it is observed to be 1107%. This further confirms the long-term significance of fatiguecorrosion in the design of optoelectronic devices. The reliability of the ANOVA results is confirmed by the normal probability plots of the randomly distributed residuals around the mean as shown in Fig. 6. 3.3. Microscopy Fig. 7 shows a typical ITO film cracking progression under fatiguecorrosion degradation conditions. The synergistic effect of fatigue and
corrosion leads to the worst electromechanical performance as compared to all different deformation-degradation modes employed in this study. When the number of cycles reaches a critical value, straight line cracks perpendicular to the straining direction are observed as shown in Fig. 7a. As the number of cycles increases, more cracks are observed to form. However, the crack morphology is observed to be different in the case of fatigue and fatigue-corrosion experiments. In particular, secondary cracks in the transverse direction are observed to form at a higher NOC for specimens under fatigue-corrosion as shown in Figs. 7e and f. For instance, above 100,000 cycles, striations due to fatigue have formed and they become more pronounced as the cycle number increases. The striations, which are non-crystallographic ridges at the formed crack sides, may be the consequence of the incremental crack growth, blunting, and re-sharpening of the crack tip, which are caused by the viscoeleastic deformation of the substrate and the accelerating effect of the corroding environment. Fatigue striations due to an
Fig. 6. Normal probability plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
256
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
Fig. 7. Optical microscopy images of cracked 200 nm thick ITO films under fatigue-corrosion. The corresponding number of cycle values is indicated on the images. Arrows indicate the tensile direction. Applied strain is 1.0%.
increment of crack growth and simultaneous blunting and resharpening of the crack tip by plastic deformation are observed [42]. Furthermore, coating delamination is observed (Fig. 7f) for films subjected to fatiguecorrosion at 150,000 cycles. In addition, it is worthwhile to mention that the ranking of damage may change depending on the magnitude of the applied stress. Furthermore, SEM observations were performed for both ITO and CNT-based conducting surfaces subjected to fatigue-corrosion experiments. Severe coating delamination is observed for the ITO surface after 150,000 cycles of testing. On the other hand under the same conditions, very limited binder delamination is observed with the underlying CNTs being mostly intact (Fig. 8). This is in agreement with the electrical resistance changes observed for both materials.
exposed area, which in turn leads to increased electrode conductivity when in contact with the acid medium. We also show that the change in electrical resistance for ITO specimens under fatigue-corrosion is significantly higher as compared to corrosion, bending, bending-corrosion, and fatigue. The severe cohesive and adhesive ITO delamination, from the base PET substrate, observed under fatigue-corrosion is associated with the dramatic electrical resistance increase. This underlines the importance of the combined action of corrosion and fatigue. DOE analysis suggests that the effects of all designed factors are statistically significant. Change in electrical resistance is found to be higher at higher strains, higher thickness, and higher number of cycles when a corrosive environment is present.
4. Conclusions
Acknowledgment
In conclusion, we investigated the effects of acrylic acid environment, strain, film thickness, and NOC on the electrical and structural integrity of ITO and CNT films both deposited on PET substrates. CNTbased electrodes are observed to outperform their ITO counterparts. In the CNT case we observe that fatigue leads to an increased CNT bundle
TBS, SDC, and DRC acknowledge the support of the National Aeronautics and Space Administration, Experimental Program to Stimulate Competitive Research (NASA WV EPSCoR), grant number NCC5-570. KAS and DRC acknowledge the support from DOE MARCEE under the award number DE-FC26-04NT42136.
Fig. 8. SEM image showing delamination of top coat on the surface CNT coated PET at 1.0% applied strain and 150,000 cycles under fatigue-corrosion. Arrows indicate the tensile direction.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
References [1] G.P. Crawford, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 1. [2] C. Coutal, A. Azema, J. Roustan, Thin Solid Films 288 (1996) 248. [3] D.C. Paine, H.Y. Yeom, B. Yaglioglu, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 79. [4] H. Kim, J.S. Horwitz, G.P. Kushto, Z.H. Kafafi, D.B. Chrisey, Appl. Phys. Lett. 79 (2001) 284. [5] J. Lee, D.N. Liu, S. Wu, Introduction to Flat Panel Displays, John Wiley & Sons, 2008. 1. [6] Y. Leterrier, A. Mottet, N. Bouquet, D. Gillieron, P. Dumont, A. Pinyol, L. Lalande, J.H. Waller, J.E. Manson, Thin Solid Films 519 (2010) 1729. [7] J. Lewis, Mater. Today 9 (2008) 30. [8] Z. Chen, B. Cotterell, W. Wang, E. Guenther, S. Chua, Thin Solid Films 394 (2001) 202. [9] Y. Leterrier, L. Medico, F. Demarco, J.A.E. Manson, U. Betz, M.F. Escola, M.K. Olsson, F. Atamny, Thin Solid Films 460 (2004) 156. [10] D.R. Cairns, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 163. [11] Y.F. Lan, W.C. Peng, Y.H. Lo, J.L. He, Org. Electron. 11 (2010) 670. [12] Z. Chen, B. Cotterell, W. Wang, Eng. Fract. Mech. 69 (2002) 597. [13] P.C.P. Bouten, P.J. Slikkerveer, Y. Leterrier, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 99. [14] K. Alzoubi, M.M. Hamasha, S. Lu, B. Sammakia, J. Disp. Technol. 7 (2011) 593. [15] T. Koniger, H. Munstedt, Meas. Sci. Technol. 19 (2008) 1. [16] S.I. Kim, T.D. Jung, P.K. Song, Thin Solid Films 519 (2010) 3085. [17] T. Minami, H. Sonohara, T. Kakumu, S. Takata, Thin Solid Films 270 (1995) 37. [18] T.S. Bejitual, K. Ramji, A.J. Kessman, K.A. Sierros, D.R. Cairns, Mater. Chem. Phys. 132 (2011) 395. [19] D.R. Cairns, R.P. Witte II, D.K. Sparacin, S.M. Sachsman, D.C. Paine, G.P. Crawford, R.R. Newton, Appl. Phys. Lett. 76 (2000) 1425. [20] A.A. Abdallah, P.C.P. Bouten, J.M.J. den Toonder, G. de With, Surf. Coat. Technol. 205 (2011) 3103.
257
[21] P.M.S. Monk, C.M. Man, J. Mater. Sci. Mater. Electron. 10 (1999) 101. [22] K.A. Sierros, N.J. Morris, K. Ramji, K.A. Sierros, D.R. Cairns, Thin Solid Films 517 (2009) 2590. [23] L. Hu, D.S. Hecht, G. Gruner, Chem. Rev. 110 (2010) 5790. [24] J. Li, L. Hu, L. Wang, Y. Zhou, G. Gruner, T.J. Marks, Nano Lett. 6 (2006) 2472. [25] Y.I. Song, C.M. Yang, D.Y. Kim, H. Kanoh, K. Kaneko, J. Colloid Interface Sci. 318 (2008) 365. [26] J.M. Park, Z.J. Wang, D.J. Kwon, G.Y. Gu, K.L. DeVries, Solid State Electron. 79 (2013) 147. [27] G. Gruner, J. Mater. Chem. 16 (2006) 3533. [28] C.M. Aguirre, S. Auvray, S. Pigeon, R. Izquierdo, P. Desjardins, R. Martel, Appl. Phys. Lett. 88 (2006) 183104. [29] K.A. Sierros, D.S. Hecht, D.A. Banerjee, N.J. Morris, L. Hu, G.C. Irvin, R.S. Lee, D.R. Cairns, Thin Solid Films 518 (2010) 6977. [30] D.S. Hecht, D. Thomas, L. Hu, C. Ladous, T. Lam, Y. Park, G. Irvin, P. Drzaic, J. Soc. Inf. Disp. 17 (11) (2009) 941. [31] C.M. Trottier, P. Glatkowski, P. Wallis, J. Luo, J. Soc. Inf. Disp. 13 (9) (2005) 759. [32] A. Dasgupta, M.G. Pecht, B. Mathieu, Finite Elem. Anal. Des. 30 (1998) 125. [33] M.M. Hamasha, K. Alzoubi, J.C. Switzer III, S. Lu, M.D. Poliks, C.R. Westgate, IEEE Trans. Compon. Packag. Technol. 2 (2012) 2007. [34] D.C. Montgomery, Introduction to Statistical Quality Control, John Wiley & Sons, 2009. 549. [35] K. Alzoubi, S. Lu, B. Sammakia, M. Poliks, J. Disp. Technol. 7 (2011) 348. [36] T.S. Bejitual, D. Compton, K.A. Sierros, D.R. Cairns, S.N. Kukureka, Thin Solid Films 528 (2013) 229. [37] K.A. Sierros, N.J. Morris, S.N. Kukureka, D.R. Cairns, Wear 267 (2009) 625. [38] S.P. Gorkhali, D.R. Cairns, G.P. Crawford, J. Soc. Inf. Disp. 12 (1) (2004) 45. [39] H.Z. Geng, K.K. Kim, K.P. So, Y.S. Lee, Y. Chang, Y.H. Lee, J. Am. Chem. Soc. 129 (2007) 7758. [40] P.N. Nirmalraj, P.E. Lyons, S. De, J.N. Coleman, J.J. Boland, Nano Lett. 9 (2009) 3890. [41] D.S. Hecht, K.A. Sierros, R.S. Lee, C. Ladous, C. Niu, D.A. Banerjee, D.R. Cairns, J. Soc. Inf. Disp. 19 (2) (2011) 157. [42] J.C. Anderson, K.D. Leaver, R.D. Rawlings, J.M. Alexander, Material Science, Chapmand & Hall, 1990. 181.
Contents lists available at ScienceDirect
Thin Solid Films journal homepage: www.elsevier.com/locate/tsf
Mechano-chemical degradation of flexible electrodes for optoelectronic device applications T.S. Bejitual, N.J. Morris, S.D. Cronin, D.R. Cairns, K.A. Sierros ⁎ Department of Mechanical & Aerospace Engineering, West Virginia University, Morgantown, 26506 WV, USA
a r t i c l e
i n f o
Available online 20 September 2013 Keywords: ITO CNT PET Flexible optoelectronics Fatigue Corrosion
a b s t r a c t The electrical, optical, and structural integrity of flexible transparent electrodes is of paramount importance in the design and fabrication of optoelectronic devices such as organic light emitting diodes, liquid crystal displays, touch panels, solar cells, and solid-state lighting. The electrodes may corrode due to acid-containing pressure sensitive adhesives present in the device stacks. In addition, structural failure may occur due to external applied loading. The combined action and further accumulation of both repeated mechanical loading and corrosion can aggravate the loss of functionality of the electrodes. In this study we investigate, using the design of experimental methods, the effects of corrosion, applied mechanical strain, film thickness, and number of bending cycles on the electrical and structural integrity of indium tin oxide (ITO) and carbon nanotube (CNT) films both coated on polyethylene terephthalate (PET) substrates. In situ electrical resistance measurements suggest that fatigue-corrosion is found to be the most critical failure mode for the ITO-based coatings. For example, the change in ITO electrical resistance increase under fatiguecorrosion (1% strain, 150,000 cycles) is 5.8 times higher than that of fatigue mode alone. On the other hand, a minimum change in electrical resistance of the CNT-based electrodes is found when applying the same conditions. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In recent years, there is a considerable interest in the development of flexible optoelectronic devices since they offer significant advantages such as large area, portability, conformability, robustness, and low cost [1]. Indium tin oxide (ITO) films on polyethylene terephthalate (PET) substrates are routinely used as the transparent conducting flexible electrode in optoelectronic devices such as touch screens, displays, solar cells, and solid state lighting. This is because they exhibit an interesting combination of relatively low electrical resistivity (~10−4 Ω cm) and high transparency (N90%) to visible light [2–5]. However, the stresses in the inorganic/organic (ITO/PET) composite resulting from thermal or mechanical loadings may lead to either cohesive failure of the inorganic layer or adhesive failure at the interface between the layer and the substrate [6]. Despite such significant thermal and mechanical mismatches, the ITO/PET system remains the most commonly used flexible electrode [7]. Many researchers have studied the mechanics of ITO on flexible substrates in the past [8–11]. For example, using a controlled buckling test, Chen et al. [12] investigated the fracture of ITO coated PET with the conductive layer placed both under tension and compression. They found that when ITO is under tension a channeling crack forms at a strain of 1.1%. On the other hand, when it is placed under compression, the film ⁎ Corresponding author. Tel.: +1 304 293 3420; fax: +1 304 293 6689. E-mail address: [email protected] (K.A. Sierros). 0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.09.044
delaminated, buckled and cracked in a tunneling motion at a strain equal to 1.7%. Failure initiation of conductive inorganic coatings is expressed in terms of crack onset strain (COS). An increase of 10% in electrical resistance is chosen arbitrarily to denote COS [13]. The COS indicates initial destruction of conductive paths through the film due to crack initiation. Alzoubi et al. [14] investigated the high cycle bending fatigue of ITO coated PET in order to establish a baseline for comprehensive reliability studies of ITO thin films on a flexible substrate. The percent change in electrical resistance was measured at specific numbers of cycles. It was found that the bending diameter and the number of bending cycles have a pronounced influence on the conductivity of the ITO layer. In particular, at 200 cycles, the percent change in electrical resistance for samples bent around 5.0 and 8.9 mm mandrel diameters was 860% and 640%, respectively. Furthermore, at 500 cycles, the percent change in electrical resistances for samples bent around the same diameters was 6250% and 3360%, respectively. Koniger and Munstedt [15] built a sophisticated device to investigate the electrical behavior of conductive layers on flexible substrates under oscillatory bending both in tension and under compression. For sputtered ITO coatings on PET substrates, a dramatic increase of the electrical resistance was observed for a bending radius smaller than 14 mm (strain ~ 0.6%) due to cracks spanning the whole sample width. The higher the oscillatory bending amplitude, the more pronounced the increase of ITO electrical resistance. Direct current (DC) magnetron sputtering is a typical method of depositing transparent conducting oxides on flexible substrates at low
252
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
temperatures. It offers large area uniformity, high deposition rate and adequate reproducibility [16]. However, the low temperature used often leads to an amorphous film microstructure. Amorphous ITO films are more susceptible to chemical degradation than their crystalline counterparts [17]. Bejitual et al. [18] studied the electrochemical stability of amorphous ITO films deposited on PET substrate in contact with pressure sensitive, acrylic acid containing, adhesives which are employed in the fabrication of flexible optoelectronic devices. They investigated the corrosion of ITO films using different concentrations of acid solution and immersion times. Electrochemical measurements indicated corrosion activities at concentrations as low as 0.05 M of acrylic acid. In addition, degradation of ITO films increased with increasing immersion time to corrosive media. There are additional studies that report on either mechanical or chemical degradation of ITO films [19–21]. However, there is little research to date addressing the combined effect of both repeated mechanical loading and corrosion on the ITO film's structural integrity. Sierros et al. [22] investigated the stress corrosion cracking of such systems. Their study showed that the combination of static bending stress and corrosion, using acrylic acid, can cause the conductive layer to crack at stresses less than a quarter of those needed for failure with no corrosion. Recently, studies investigating carbon nanotube (CNT) films deposited on polymer substrate using solution-based techniques have been reported. CNT coated polymer systems exhibit relatively low sheet resistance values (~200 Ω/sq) and optical transmittance above 80% [23–26]. Such films are being developed with the aim of replacing brittle transparent conductive oxides due to their abundance, relatively low cost of fabrication and superior mechanical flexibility [27,28]. Sierros et al. [29] studied the mechanical integrity of CNT and ITO films under monotonic and cyclic uniaxial tension with in situ electrical resistance monitoring. They observed that the critical strain for 25% electrical increase is 10 times higher for CNT films than that observed for their ITO counterparts. In addition, Hecht et al. [30] conducted single-point stylus-pen actuation tests on CNT coated PET films for resistive touch screens. They reported that no failure is observed up to 3 million actuations. Furthermore, Trottier et al. [31] studied the cyclic bending and monotonic tensile deformation of CNT films on PET substrates. It was reported that the maximum changes in electrical resistance under cyclic bending (0.7% strain and 2500 cycles) and monotonic tensile deformation (0.5% strain) were less than 0.5% and 5%, respectively. Statistical design of experiments (DOE) has been employed in many engineering and scientific studies to identify and quantify the effect of factors and their interactions on a response variable [32]. When there are several factors in an experiment, a factorial design helps to conduct the experiments and analyze the results in a statistically meaningful way [33]. As the number of factors in full factorial design increases, the number of runs required increases rapidly. In such a case, one can assume that certain high-order interactions are negligible and a fractional factorial design involving fewer than the complete set of runs can be used to obtain information on the main effects and low-order interactions [34]. Alzoubi et al. [35] used DOE methods to study the effects of temperature, humidity, bending radius, film thickness, and frequency on the high cycle bending fatigue of thin copper films on polymer substrates. In this work, we report on the effects of corrosion, applied strain, film thickness, and number of cycles (time) on the electrical and mechanical integrity of ITO and CNT films that are deposited on PET substrates. 2. Experimental 2.1. Materials ITO films (CPFilms Inc., USA) deposited at room temperature on PET substrates (180 μm thick) by DC magnetron sputtering (sputtering
power ≈ 1 kW) were used in this work. During sputtering, the processing gas was argon with a partial pressure of approximately 0.5 Pa whereas oxygen was used as a reactive gas with a partial pressure of approximately 2 MPa. The deposition rate of the ITO film was approximately 5 nm/min. The deposited films, as measured by a stylus profilometer (Veeco Dektak 150), were 70 and 200 nm thick with sheet resistances of 100 and 70 Ω/sq, respectively. The ITO target was composed of 90 wt.% In2O3 and 10 wt.% SnO2. Sheet resistances were measured using a four-point probe system (Electronic Design to Market, USA). Also, CNT films (Unidym, USA) on PET substrates (180 μm thick) were used for comparison purposes. The carbon nanotubes were grown via chemical vapor deposition and subsequently purified via air oxidation followed by acid washing to remove residual metal catalyst and amorphous carbon. The resulting CNTs consisted of largely singlewalled and double-walled tubes. The CNTs were dispersed in a 1% aqueous solution of sodium dodecyl sulfate surfactant at 0.01 wt.% and they were then spray-coated onto the PET sheet which was heated to 100 °C. The films were then rinsed in de-ionized water to remove the residual surfactant. The resulting dry CNT film was approximately 10–20 nm thick as measured by atomic force microscopy. A polymethyl methacrylate (PMMA) layer was coated with a Mayer rod from a 0.5 wt.% solution of PMMA in methyl ethyl ketone over the dry CNT film to enhance the film adhesion to the PET and provide enhanced chemical/ environmental resistance. The dry thickness of the binder layer is ~50 nm, thin enough such that it does not electrically insulate the surface of the CNT film. The polymer coating is not continuous and in some regions the CNT layer is exposed allowing a conducting path. The sheet resistance of the CNT film was approximately 350 Ω/sq with a thickness of 15 nm. 2.2. Electromechanical testing During corrosion testing, non-strained flat U-shaped samples were immersed in a 0.05 M acrylic acid solution, and the change in electrical resistance was monitored in situ. Bending, fatigue, bendingcorrosion, and fatigue-corrosion characterizations were conducted by means of a custom-built apparatus. A schematic of the apparatus can be found in the previous work [36]. During the test, one end of the specimen is fixed to a mandrel made from chlorinated polyvinyl chloride (c-PVC) using a high-density polyethylene (HDPE) plate whereas the other end is clamped between a pair of HDPE plates which are connected to a low tension spring (k = 68.4 N/m). A strip of copper foil is used for electrical contact. A relatively large c-PVC mandrel (R ≫ r) supports the specimen in a horizontal position and prevents it from sagging. A stepper motor (MDrive M-17, IMS) is used to rotate the mandrel at a frequency of 2 cycles/s. During both static bending-corrosion and fatigue-corrosion experiments the HDPE container is filled with a 0.05 M acrylic acid solution. The change in electrical resistance is monitored in-situ using an Agilent 34970A data acquisition/switch unit. Specimens were cut into strips of 10 mm in width and 184–200 mm in length in accordance with the bending diameter used. Mandrels with diameters equal to 29.1, 22.6, and 17.8 mm were used. The approximate bending strains using such mandrels correspond to 0.6, 0.8, and 1.0%, respectively. 2.3. Microscopy Specimen surfaces were observed, after testing, using a Leica optical microscope equipped with a frame grabber (Guppy, Allied Visions Technology). Also, the resulting surface morphology of the film was examined using a JEOL JSM-7600F scanning electron microscope (SEM) equipped with a field emission gun. Before SEM imaging, specimens were sputter-coated with an approximately 5 nm thick Au layer using a Cressington 108 sputter coater.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
253
Table 1 Factors and levels for DOE analysis of fatigue and fatigue-corrosion experiments. Factors
Acid concentration (M) Strain (%) Thickness (nm) Number of cycles
Levels 1
2
3
4
0 0.6 70 0
0.05 0.8 200 50,040
– 1 – 100,020
– – – 150,000
2.4. Design of experiments Full factorial DOE was used to further investigate the effect of various designs, processes, and environmental factors on the functionality of ITO films. Four study factors were considered while change in electrical resistance was selected as a response. The study factors were: acrylic acid concentration, applied strain, film thickness, and number of cycles (NOC) for fatigue and combined fatigue-corrosion experiments. In addition, for bending and bending-corrosion, time was considered as a study factor due to the static nature of these experiments. The levels of each factor chosen are presented in Table 1.
Fig. 2. Critical number of cycles for fatigue and fatigue corrosion of 200 nm thick ITOcoated PET films.
3. Results and discussion
In order to investigate the effect of acid concentration, applied strain, film thickness, and number of cycles on the mechano-chemical degradation of the electrodes under static and dynamic loading, in situ electrical measurements were conducted. Electrical resistance changes of the conducting layer during testing can be associated with loss of coating functionality [37]. A typical response of ITO electrical resistance versus number of cycles and/or time is shown in Fig. 1. During the initial stages of fatigue-only, two regimes of electrical resistance increase are observed, at first an increase of electrical resistance due to changes in dimensions of the compliant substrate, until an equilibrium size is attained, and then a gradual linear increase that may be due to cracking of ITO. Gorkhali et al. [38] reported that the initial increase of electrical resistance for an applied strain of 1.5% is 50–100 cycles. In our case the initial increase, for a 0.5% lower applied strain, is observed to lie around 800–1000 cycles. In all cases, the change in electrical resistance for specimens subjected to fatigue-corrosion is observed to be the highest. On the other hand, the lowest change is observed in the case of corrosion-only experimental conditions. Furthermore, the critical
number of cycles (Fig. 2) for the initial loss of functionality, under fatigue and combined fatigue-corrosion of ITO films, is determined from an increase of 10% in electrical resistance (Fig. 1). In general, specimens under fatigue-corrosion degradation are observed to exhibit a lower critical number of cycles for both 70 and 200 nm thick ITO films. This denotes the significance of the combined effect of fatigue and corrosion, which can be used as an accelerated degradation experimental protocol that can aid towards identifying the interrelation of the critical factors that influence the long-term reliability of the device component. The effect of applied strain on the electrical degradation of ITO films at different degradation conditions was investigated. A common observation is that, for each condition and film thickness, the higher the applied strain the larger the positive change in electrical resistance. For example, we observe that at 150,000 cycles the change in ITO (200 nm) electrical resistance during fatigue-corrosion is 300% for 0.6% strain and 1107% for 1.0% applied strain. This denotes more than a threefold increase. Such phenomena can be related to the increased film crack formation as the applied strain increases. Also, the effect of film thickness on the electrical degradation of ITO films was investigated at different conditions. Higher change in electrical resistance values are observed for the thicker films, suggesting that
Fig. 1. Normalized electrical resistance versus experimental time/number of cycles for ITOcoated PET at 1.0% applied strain and 200 nm film thickness.
Fig. 3. Normalized electrical resistance versus experimental time/number of cycles for CNT-deposited PET at 1.0% applied strain.
3.1. In situ electrical resistance measurements
254
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
Table 2 Analysis of variance (ANOVA) for fatigue and fatigue-corrosion. Source
DF
Conc. 1 Strain 2 Thickness 1 NOC 3 Conc. ∗ strain 2 Conc. ∗ thickness 1 Conc. ∗ NOC 3 Strain ∗ thickness 2 Strain ∗ NOC 6 Thickness ∗ NOC 3 Conc. ∗ strain ∗ thickness 2 Conc. ∗ strain ∗ NOC 6 Conc. ∗ thickness ∗ NOC 3 Strain ∗ thickness ∗ NOC 6 Error 6 Total 47 S = 30.81, R-sq = 99.88%, R-sq (adj) = 99.06%.
Seq SS
Adj SS
Adj MS
F
P
1470543 730137 58334 1056107 396287 16322 565061 9949 261251 19637 6410 145552 5585 8380 5697 4755252
1470543 730137 58334 1056107 396287 16322 565061 9949 261251 19637 6410 145552 5585 8380 5697
1470543 365068 58334 352036 198144 16322 188354 4975 43542 6546 3205 24259 1862 1397 950
1548.74 384.48 61.44 370.76 208.68 17.19 198.37 5.24 45.86 6.89 3.38 25.55 1.96 1.47
0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.05 0.00 0.02 0.10 0.00 0.22 0.33
they are influenced, more than their thinner counterparts, by the pure mechanical or the mechano-chemical imposed degradation. It is believed that the higher concentration of pre-existing flaws and defects in the thicker layer may play a role for such a response. On the other hand, it has been reported in the past that thinner layers of ITO can better withstand the applied strain than thicker layers [19]. However, the increase in resistance is not observed to be as large as the one observed during the application of different strains. This indicates the higher significance of applied strain as a degradation factor. Furthermore, CNT films are tested under the same conditions. In contrast to the ITO film behavior, CNT-based electrodes exhibit a different trend as shown in Fig. 3. The change in electrical resistance is observed to be negative, indicating enhanced electrical conductivity of the CNT films. In the presence of a combined mechano-corrosive environment this negative change becomes more pronounced. It is believed that the observed behavior is due to the effect of the acrylic acid solution. The role of acid treatment on CNT networks has been reported in the past. The acid further treats and enhances the metallicity of the CNTs as suggested by both Geng et al. and Nirmalraj et al. [39,40]. In addition, Hecht et al. [41] reported that immersion of CNT films on polycarbonate substrates in 0.1 M acrylic acid solution leads to a negative change in electrical resistance (−2.5%). In our case, the combined effect of fatigue and acidic corrosion leads to increased binder cracking and, therefore, to a larger CNT area exposure. This in turn leads to an
increased number of acid-treated CNT bundles and, therefore, higher film conductivity. However, the change in electrical resistance is not as significant as in the ITO film case. It is therefore suggested that CNT films can better withstand externally applied strains than their ITO counterparts under the combined actions of fatigue and corrosion. 3.2. Design of experiments A further investigation of the effects of corrosion, applied strain, film thickness, and number of cycles (NOC) on the electrical degradation of ITO films was conducted using full factorial design of experiments. Table 2 shows the analysis of variance (ANOVA) and presents the error sum of squares and P-value of the main effects and their twoway and three-way interactions under fatigue and fatigue-corrosion. All design factors and their two-way interactions are found to be significant since their P-values at the 95% confidence level is less than 0.05. However, three-way interactions involving thickness are observed to be insignificant. From the main effect graphs (Fig. 4), the most significant effects are acid concentration, strain, and NOC. High initial rate of change of ITO electrical resistance is observed for both strain and NOC. The significant contribution of corrosion on the change in electrical resistance increase is shown by the two-way interaction plots (Fig. 5). Also, ANOVA was performed for bending and bending-corrosion experiments. All design factors and their two-way interactions are
Fig. 4. Main effects plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
255
Fig. 5. Interaction plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
observed to be significant. However, only one three-way interaction involving thickness is observed to be insignificant. In addition, comparable trends of main effects and two-way interactions are observed for bending and bending-corrosion as in fatigue and fatigue-corrosion. However, the maximum change in electrical resistance under bendingcorrosion is 59% whereas under fatigue-corrosion it is observed to be 1107%. This further confirms the long-term significance of fatiguecorrosion in the design of optoelectronic devices. The reliability of the ANOVA results is confirmed by the normal probability plots of the randomly distributed residuals around the mean as shown in Fig. 6. 3.3. Microscopy Fig. 7 shows a typical ITO film cracking progression under fatiguecorrosion degradation conditions. The synergistic effect of fatigue and
corrosion leads to the worst electromechanical performance as compared to all different deformation-degradation modes employed in this study. When the number of cycles reaches a critical value, straight line cracks perpendicular to the straining direction are observed as shown in Fig. 7a. As the number of cycles increases, more cracks are observed to form. However, the crack morphology is observed to be different in the case of fatigue and fatigue-corrosion experiments. In particular, secondary cracks in the transverse direction are observed to form at a higher NOC for specimens under fatigue-corrosion as shown in Figs. 7e and f. For instance, above 100,000 cycles, striations due to fatigue have formed and they become more pronounced as the cycle number increases. The striations, which are non-crystallographic ridges at the formed crack sides, may be the consequence of the incremental crack growth, blunting, and re-sharpening of the crack tip, which are caused by the viscoeleastic deformation of the substrate and the accelerating effect of the corroding environment. Fatigue striations due to an
Fig. 6. Normal probability plot for change in electrical resistance for ITO films under fatigue and fatigue-corrosion.
256
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
Fig. 7. Optical microscopy images of cracked 200 nm thick ITO films under fatigue-corrosion. The corresponding number of cycle values is indicated on the images. Arrows indicate the tensile direction. Applied strain is 1.0%.
increment of crack growth and simultaneous blunting and resharpening of the crack tip by plastic deformation are observed [42]. Furthermore, coating delamination is observed (Fig. 7f) for films subjected to fatiguecorrosion at 150,000 cycles. In addition, it is worthwhile to mention that the ranking of damage may change depending on the magnitude of the applied stress. Furthermore, SEM observations were performed for both ITO and CNT-based conducting surfaces subjected to fatigue-corrosion experiments. Severe coating delamination is observed for the ITO surface after 150,000 cycles of testing. On the other hand under the same conditions, very limited binder delamination is observed with the underlying CNTs being mostly intact (Fig. 8). This is in agreement with the electrical resistance changes observed for both materials.
exposed area, which in turn leads to increased electrode conductivity when in contact with the acid medium. We also show that the change in electrical resistance for ITO specimens under fatigue-corrosion is significantly higher as compared to corrosion, bending, bending-corrosion, and fatigue. The severe cohesive and adhesive ITO delamination, from the base PET substrate, observed under fatigue-corrosion is associated with the dramatic electrical resistance increase. This underlines the importance of the combined action of corrosion and fatigue. DOE analysis suggests that the effects of all designed factors are statistically significant. Change in electrical resistance is found to be higher at higher strains, higher thickness, and higher number of cycles when a corrosive environment is present.
4. Conclusions
Acknowledgment
In conclusion, we investigated the effects of acrylic acid environment, strain, film thickness, and NOC on the electrical and structural integrity of ITO and CNT films both deposited on PET substrates. CNTbased electrodes are observed to outperform their ITO counterparts. In the CNT case we observe that fatigue leads to an increased CNT bundle
TBS, SDC, and DRC acknowledge the support of the National Aeronautics and Space Administration, Experimental Program to Stimulate Competitive Research (NASA WV EPSCoR), grant number NCC5-570. KAS and DRC acknowledge the support from DOE MARCEE under the award number DE-FC26-04NT42136.
Fig. 8. SEM image showing delamination of top coat on the surface CNT coated PET at 1.0% applied strain and 150,000 cycles under fatigue-corrosion. Arrows indicate the tensile direction.
T.S. Bejitual et al. / Thin Solid Films 549 (2013) 251–257
References [1] G.P. Crawford, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 1. [2] C. Coutal, A. Azema, J. Roustan, Thin Solid Films 288 (1996) 248. [3] D.C. Paine, H.Y. Yeom, B. Yaglioglu, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 79. [4] H. Kim, J.S. Horwitz, G.P. Kushto, Z.H. Kafafi, D.B. Chrisey, Appl. Phys. Lett. 79 (2001) 284. [5] J. Lee, D.N. Liu, S. Wu, Introduction to Flat Panel Displays, John Wiley & Sons, 2008. 1. [6] Y. Leterrier, A. Mottet, N. Bouquet, D. Gillieron, P. Dumont, A. Pinyol, L. Lalande, J.H. Waller, J.E. Manson, Thin Solid Films 519 (2010) 1729. [7] J. Lewis, Mater. Today 9 (2008) 30. [8] Z. Chen, B. Cotterell, W. Wang, E. Guenther, S. Chua, Thin Solid Films 394 (2001) 202. [9] Y. Leterrier, L. Medico, F. Demarco, J.A.E. Manson, U. Betz, M.F. Escola, M.K. Olsson, F. Atamny, Thin Solid Films 460 (2004) 156. [10] D.R. Cairns, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 163. [11] Y.F. Lan, W.C. Peng, Y.H. Lo, J.L. He, Org. Electron. 11 (2010) 670. [12] Z. Chen, B. Cotterell, W. Wang, Eng. Fract. Mech. 69 (2002) 597. [13] P.C.P. Bouten, P.J. Slikkerveer, Y. Leterrier, in: G.P. Crawford (Ed.), Flexible Flat Panel Displays, John Wiley & Sons, 2005, p. 99. [14] K. Alzoubi, M.M. Hamasha, S. Lu, B. Sammakia, J. Disp. Technol. 7 (2011) 593. [15] T. Koniger, H. Munstedt, Meas. Sci. Technol. 19 (2008) 1. [16] S.I. Kim, T.D. Jung, P.K. Song, Thin Solid Films 519 (2010) 3085. [17] T. Minami, H. Sonohara, T. Kakumu, S. Takata, Thin Solid Films 270 (1995) 37. [18] T.S. Bejitual, K. Ramji, A.J. Kessman, K.A. Sierros, D.R. Cairns, Mater. Chem. Phys. 132 (2011) 395. [19] D.R. Cairns, R.P. Witte II, D.K. Sparacin, S.M. Sachsman, D.C. Paine, G.P. Crawford, R.R. Newton, Appl. Phys. Lett. 76 (2000) 1425. [20] A.A. Abdallah, P.C.P. Bouten, J.M.J. den Toonder, G. de With, Surf. Coat. Technol. 205 (2011) 3103.
257
[21] P.M.S. Monk, C.M. Man, J. Mater. Sci. Mater. Electron. 10 (1999) 101. [22] K.A. Sierros, N.J. Morris, K. Ramji, K.A. Sierros, D.R. Cairns, Thin Solid Films 517 (2009) 2590. [23] L. Hu, D.S. Hecht, G. Gruner, Chem. Rev. 110 (2010) 5790. [24] J. Li, L. Hu, L. Wang, Y. Zhou, G. Gruner, T.J. Marks, Nano Lett. 6 (2006) 2472. [25] Y.I. Song, C.M. Yang, D.Y. Kim, H. Kanoh, K. Kaneko, J. Colloid Interface Sci. 318 (2008) 365. [26] J.M. Park, Z.J. Wang, D.J. Kwon, G.Y. Gu, K.L. DeVries, Solid State Electron. 79 (2013) 147. [27] G. Gruner, J. Mater. Chem. 16 (2006) 3533. [28] C.M. Aguirre, S. Auvray, S. Pigeon, R. Izquierdo, P. Desjardins, R. Martel, Appl. Phys. Lett. 88 (2006) 183104. [29] K.A. Sierros, D.S. Hecht, D.A. Banerjee, N.J. Morris, L. Hu, G.C. Irvin, R.S. Lee, D.R. Cairns, Thin Solid Films 518 (2010) 6977. [30] D.S. Hecht, D. Thomas, L. Hu, C. Ladous, T. Lam, Y. Park, G. Irvin, P. Drzaic, J. Soc. Inf. Disp. 17 (11) (2009) 941. [31] C.M. Trottier, P. Glatkowski, P. Wallis, J. Luo, J. Soc. Inf. Disp. 13 (9) (2005) 759. [32] A. Dasgupta, M.G. Pecht, B. Mathieu, Finite Elem. Anal. Des. 30 (1998) 125. [33] M.M. Hamasha, K. Alzoubi, J.C. Switzer III, S. Lu, M.D. Poliks, C.R. Westgate, IEEE Trans. Compon. Packag. Technol. 2 (2012) 2007. [34] D.C. Montgomery, Introduction to Statistical Quality Control, John Wiley & Sons, 2009. 549. [35] K. Alzoubi, S. Lu, B. Sammakia, M. Poliks, J. Disp. Technol. 7 (2011) 348. [36] T.S. Bejitual, D. Compton, K.A. Sierros, D.R. Cairns, S.N. Kukureka, Thin Solid Films 528 (2013) 229. [37] K.A. Sierros, N.J. Morris, S.N. Kukureka, D.R. Cairns, Wear 267 (2009) 625. [38] S.P. Gorkhali, D.R. Cairns, G.P. Crawford, J. Soc. Inf. Disp. 12 (1) (2004) 45. [39] H.Z. Geng, K.K. Kim, K.P. So, Y.S. Lee, Y. Chang, Y.H. Lee, J. Am. Chem. Soc. 129 (2007) 7758. [40] P.N. Nirmalraj, P.E. Lyons, S. De, J.N. Coleman, J.J. Boland, Nano Lett. 9 (2009) 3890. [41] D.S. Hecht, K.A. Sierros, R.S. Lee, C. Ladous, C. Niu, D.A. Banerjee, D.R. Cairns, J. Soc. Inf. Disp. 19 (2) (2011) 157. [42] J.C. Anderson, K.D. Leaver, R.D. Rawlings, J.M. Alexander, Material Science, Chapmand & Hall, 1990. 181.