Elsevier

Thin Solid Films

Volume 517, Issue 8, 27 February 2009, Pages 2590-2595
Thin Solid Films

Stress–corrosion cracking of indium tin oxide coated polyethylene terephthalate for flexible optoelectronic devices

https://doi.org/10.1016/j.tsf.2008.10.031Get rights and content

Abstract

Stress corrosion cracking of transparent conductive layers of indium tin oxide (ITO), sputtered on polyethylene terephthalate (PET) substrates, is an issue of paramount importance in flexible optoelectronic devices. These components, when used in flexible device stacks, can be in contact with acid containing pressure-sensitive adhesives or with conductive polymers doped in acids. Acids can corrode the brittle ITO layer, stress can cause cracking and delamination, and stress–corrosion cracking can cause more rapid failure than corrosion alone.

The combined effect of an externally-applied mechanical stress to bend the device and the corrosive environment provided by the acid is investigated in this work. We show that acrylic acid which is contained in many pressure-sensitive adhesives can cause corrosion of ITO coatings on PET. We also investigate and report on the combined effect of external mechanical stress and corrosion on ITO-coated PET composite films. Also, it is shown that the combination of stress and corrosion by acrylic acid can cause ITO cracking to occur at stresses less than a quarter of those needed for failure with no corrosion. In addition, the time to failure, under ~ 1% tensile strain can reduce the total time to failure by as much as a third.

Introduction

Flexible optoelectronic devices are currently receiving great interest because of the potential advantages of their large area, low cost, and exciting new form factors. To fully realize the potential commercial success of such flexible devices they must be able to withstand the mechanical and thermal strains of roll-to-roll manufacturing processes and endure repeated loading in a wide range of environmental conditions for several years service. While there are many components that are used in devices, such as displays, touch-screens and solar cells, they all share a common component, which is the transparent conducting electrode [1].

Most of the optoelectronic technologies today utilize indium tin oxide (ITO) coated polyethylene terephthalate (PET) substrates as the transparent flexible electrode because of the combination of low electrical resistivity (~ 7  10 4 Ω cm) and high optical transparency (> 90%) [2], [3], [4].

However, this hybrid material system represents some important challenges because a thin brittle inorganic layer is deposited on a compliant organic substrate and therefore a mismatch in mechanical and thermal properties exists. The mechanical integrity of the flexible anode component when the optoelectronic device is conformed to various shapes represents a major challenge to both developing manufacturing processes and designing systems that can be reliable in the field.

The brittle nature of the ITO thin layer makes it susceptible to cracking when deformed. In tension the onset of cracks between 2% and 2.5% uniaxial tensile strain correlates with a sudden increase in the ITO electrical resistance [5]. The minimum strain to cause cracks is an important design parameter and gives an upper limit for any short-term deformation. In addition, under increasing cyclic loading, very little resistance is recovered after the occurrence of initial cracking [6].

Furthermore, important thin-film mechanical deformation modes are bending and controlled buckling under uniaxial stress states. Bouten [7] investigated the failure of ITO on polycarbonate substrates using the two-point bend test, originally developed for optical fiber characterization, coupled with electrical resistance monitoring. It was determined that ITO layers, 100 nm thick, can fail at a critical failure strain at about 1.2%. The minimum strain at which cracking occurs is a function of both film thickness and quality. Coating defects and particulate matter can act as crack initiators thus reducing the critical strain.

Chen et al. [8], [9] studied the fracture properties of thin, brittle, ITO coatings, deposited on PET, under controlled buckling conditions. They used the controlled buckling test because the PET film is extremely flexible and the usual three or four-point bend tests were inappropriate. By buckling the PET film so that the ITO coating is on the convex or the concave sides, the conducting layer can be tested either in tension or compression. The electrical resistance of the coating can be monitored in situ. They found that the critical failure strain for the coating under tension was 1.1% and under compression was 1.7%. They concluded that the mechanism of the ITO coating failure under tension is by cohesive cracks whereas under compression, the coating delaminates and buckles before cracking.

Additionally, Sierros et al. [10] reported preliminary results for the crack onset strain for biaxial fragmentation of ITO coatings on polyethylene naphthalate flexible substrates using a specially-designed bulge-test apparatus coupled with electrical-resistance monitoring. They calculated the critical onset ITO biaxial strain, using the spherical cap model, to be 1.4%. They also observed, using contact-mode atomic force microscopy, that the complex biaxial stress state lead to a deformation of the coating islands and to a rougher surface. Biaxial branching cracks were also observed using scanning electron microscopy (SEM).

ITO thin layers are most often deposited on polymer substrates at room temperature by direct current (DC) magnetron sputtering. Sputtering is the most common method for large-scale deposition of transparent conducting oxides on substrates [11]. When ITO is deposited on an unheated glass or polymer substrate its resulting microstructure is amorphous [12]. Yeom et al. [13] reported the effect of oxygen stoichiometry on the amorphous structure and crystallization kinetics of ITO on glass, initially at room temperature and then annealed in air at 250 °C using reactive DC magnetron sputtering. They observed using bright-field TEM that initially the material is in the amorphous state and then, after 5 min annealing at 250 °C, becomes partially crystalline. For 10 min annealing at 250 °C the ITO film becomes more crystalline and after 30 min of annealing at 250 °C it is fully crystallized.

As reported by Gaarenstroom et al. [14] a major limitation in the use of ITO coatings as the anode in photoelectrochemical devices is the short lifetime due to corrosion. They compared corrosion rates for ITO film electrodes grown using different sputter deposition conditions and they observed that the most corrosion-resistant films exhibited well-formed crystals and showed the most crystallinity. As noted by Raes and Smeets [15] when ITO anodes are exposed to voltage differences in the presence of moisture and contamination, they are also sensitive to corrosion.

Because of the wide application of polycrystalline ITO layers on glass as anodes in optoelectronic devices, there have been a number of studies of ITO corrosion. Most of the work in the literature focuses on halogenated acids, and ITO in contact with polyethylenedioxithiophene doped with polystyrene sulfonate in organic light emitting diodes (OLED) devices [18], [19]. Scholten and van den Meerakker [16] reported the chemical etching behaviour of DC magnetron sputtered ITO films in a large number of acids. They found that the etch rate of crystalline ITO in acids other than the halogenated acids is extremely low. Also, Meerakker et al. [17] observed that ITO films deposited at room temperature are etched by acidic solutions in a matter of seconds whereas crystalline ITO films, deposited at higher temperatures, dissolve in concentrated acids within minutes. It is therefore evident that the ITO etch rate is sensitive to film crystallinity with amorphous ITO being etchable with dilute acids including acetic acid [17]. The etching of ITO electrodes, used in OLED's, by residual acid in a solid conducting polymer layer has also been reported [18], [19]. For flexible optoelectronic devices, acrylate-based inks and adhesives can be widely used and residual acrylic acid from these polymer layers can cause premature failure of amorphous ITO films on flexible substrates.

In addition to being in contact with acrylic acid, the ITO anode may also be under the influence of an externally applied mechanical stress, which is needed in order to bend the flexible device. Little research has been done to date on this issue. The combination of mechanical stress and chemical corrosion can lead to stress–corrosion cracking. It is important in any flexible optoelectronic device design to understand the degradation of properties due to all three of these phenomena.

In this study we investigate and report on the stress–corrosion cracking of ITO coated PET substrates which can be used in numerous optoelectronic device technologies.

Section snippets

Sample preparation

ITO was deposited on PET substrates (DuPont ST504) by DC magnetron sputtering at room temperature, and the samples had a sheet resistance of 100 Ω/sq. The sputtering power was approximately 1 kW. Argon was used as the processing gas, along with a small quantity of oxygen. The oxygen partial pressure was approximately 2 mPa. Film deposition rate was between 4 and 5 nm/min. The chemical composition of the ITO target was In2O3:SnO2 = 90:10.

The thickness of the ITO coating (100 Ω/sq. sheet

Tensile testing of films

The critical-onset strain at which the ITO-coated PET starts to crack was determined by tensile testing of the composite film. A sudden increase in resistance was observed at around 2.25% tensile strain for the 200 Ω/sq sample and 1.75% for the 100 Ω/sq coated sample (see Fig. 1). The difference in the two crack-onset values is due to the difference in film thickness of the two samples. These critical strains are consistent with high quality sputter deposited films without macroscale defects [5]

Conclusions

In conclusion, TOF–SIMS data confirm that the ITO film interacts with an acrylate-acid containing PSA. Also, it is shown that the combination of stress and corrosion by acrylic acid can cause ITO cracking to occur at stresses less than a quarter of those needed for failure with no corrosion (Fig. 8). In addition, the time to failure, under ~ 1% tensile strain can reduce the total time to failure by as much as a third. It is also worth noting that corrosion is formed around rough ITO areas.

Acknowledgements

DRC and KAS gratefully acknowledge the support of WVEPSCoR Research Challenge Grant under award no. EPS08-01. We thankfully acknowledge Emily Cairns for editing the SEM images.

References (19)

  • Z. Chen et al.

    Eng. Fract. Mech.

    (2002)
  • R. Kukla

    Surf. Coat. Technol.

    (1997)
  • T. Minami et al.

    Thin Solid Films

    (1995)
  • H.-Y. Yeom et al.

    Thin Solid Films

    (2002)
  • J.E.A.M. van den Meerakker et al.

    Thin Solid Films

    (1995)
  • H. Aziz et al.

    Synth. Met.

    (1996)
  • W. Bijnens et al.

    Synth. Met.

    (1998)
  • D.R. Cairns
  • D.C. Paine et al.
There are more references available in the full text version of this article.

Cited by (137)

  • Constructing antibacterial surfaces with alkali treatment on polyethylene terephthalate nanofibers

    2023, Colloids and Surfaces A: Physicochemical and Engineering Aspects
  • Stretchable and recoverable acrylate-based pressure sensitive adhesives with high adhesion performance, optical clarity, and metal corrosion resistance

    2021, Chemical Engineering Journal
    Citation Excerpt :

    The decline in %T with increasing strain can be attributed to the change in the polymer structure (from a random coil to linear chain), which possibly increases the crystallinity or the fluctuation of the filler phase structure due to the reorientation or alignment of the filler [32,33]. Although the AA-incorporated PSAs are well known for their excellent adhesion performance, some reports have described that the addition of carboxylic acid to soft adhesives can lead to the corrosion of metal substrates [8,34]. Therefore, we evaluated the compatibility of 1AA with a corrosion-sensitive layer (ITO) by applying 1AA and 1MA (composed of 10 mol% of methyl acrylate) on ITO-coated glass substrates, and subjected to the specified hygrothermal conditions for four weeks; 1MA was used as a reference for comparing the acid effects.

View all citing articles on Scopus
View full text