Multilayer thin film barrier for protection of flex-electronics
03/01/2008
EXECUTIVE OVERVIEW
Continuous evolution marks the electronics industry’s progress toward even thinner, lighter and more flexible products. Although tremendous progress has been demonstrated, one fundamental obstacle that remains is the sensitivity of these electronics to the moisture and oxygen in the environment. So the search for an inexpensive, yet robust method to protect these devices continues. The current conventional encapsulation solutions, e.g., using a metal can or glass can with desiccant, although somewhat effective, add weight, thickness and cost. Therefore, a thin film barrier is viewed as the only viable solution to protect these devices from damage caused by moisture and oxygen while maintaining a thin and flexible form factor and light weight.
 |
Thin film encapsulation is gaining considerable traction in various markets for a variety of applications, such as organic light-emitting diode (OLED) displays, thin film batteries, and thin film photovoltaics.
Theoretically, a single layer of some ceramic materials, if made perfectly, would have a barrier quality as good as a sheet of glass. But this has never been demonstrated to be practical for commercial use because imperfections occur during deposition. Among the most commonly seen imperfections are defects caused by surface topography and particles. The temperature limitations of organic and flexible electronics also often limit the quality of barrier films that can be deposited. It is often the case that the device to be protected is damaged by the high temperature required to deposit a good barrier. Some deposition technologies known for depositing good quality film, such as atomic layer deposition (ALD), are often too slow and too expensive for commercial production purposes.
Example: Barix
Vitex’s Barix multilayer thin film barrier is comprised of alternating layers of organic and inorganic films applied in vacuum. This coating is typically only a few microns thick and can be applied directly on the device as encapsulation or on plastic film to be used as a substrate or laminating encapsulant.
The unique feature of this multilayer coating is how this organic layer is deposited. A liquid monomer precursor is flash evaporated to vapor, which then flows into a vacuum chamber. This monomer vapor then condenses and becomes liquid on the substrate surface. Because the monomer is deposited in liquid form, it completely planarizes the surface topography and smoothes the surface. It is then polymerized by UV and becomes a solid whose top surface is atomically smooth, creating an ideal surface on which to deposit a barrier film.
An inorganic film, only a few tens of nanometers thick, is deposited on top of the polymer layer. Because the polymer surface is so smooth, the inorganic film is grown with very few defects and is therefore an almost perfect moisture barrier. These two processes are then repeated to achieve the desired level of barrier performance. Figure 1 shows what this barrier structure looks like and how it can effectively planarize the surface topography and smooth surface defects.
Both organic and inorganic layers are deposited at low temperatures, usually less than 90°C. Both materials are highly transparent and their refractive indices are carefully matched to make the barrier film suitable for applications such as OLED displays and photovoltaic modules to allow the maximum amount of light to go in or out through the barrier film.
 Figure 2. Calcium test yield vs. number of organic/inorganic pairs after high temperature and high humidity test. |
Figure 2 is a yield study performed by Vitex using the so-called calcium test method. Each test group consists of 54 calcium test samples. The yield criteria were set based on stringent tests that thin film barriers need to pass for OLED displays. Although the yield is not at an acceptable level, it was shown that it is possible to protect the device with only one organic-inorganic pair. It was observed that the yield loss was mainly caused by the early failures related to particles so large that one layer of the organic film with thickness used in this study was not adequate to cover. As the number of pairs of organic and inorganic layers increases, the yield also improves. With a three-pair process, we were able to achieve almost 100% yield. This study is particularly important because it shows the pathway of increasing robustness and reducing production costs. Because different applications may have different yield criteria and may require different levels of protection, the multilayer approach provides the flexibility to adjust the barrier accordingly.
Enabling ultra-thin and flexible OLED displays
OLED displays have the advantages of self-emitting light, fast response time, low power consumption and wide viewing angle and are considered as the most promising next generation display to replace LCD. In TIME magazine’s November 12, 2007 issue, LG. Philips’ flexible OLED display was selected as one of the best innovations of 2007 in the computer category. This 4-in. QVGA (240 × RGB × 360) display was a joint collaborative effort of LG. Philips, Universal Display Corp. (UDC), and Vitex Systems.
LG. Philips first built the amorphous silicon thin film transistors (TFTs) on stainless steel foil; UDC then deposited the OLED with its FOLED and PHOLED technologies; and then the display was encapsulated by Vitex with its Barix encapsulation. The resulting display has total thickness of 0.15mm and is flexible.
 Figure 3. a) Samsung SDI’s 4-in. bendable AMOLED display; b) schematic cross- section view of the display structure (not to scale, courtesy of Samsung SDI). |
Additionally, Samsung SDI demonstrated a bendable 4-in. WQVGA (480 × RGB × 272) AM OLED display with low temperature polysilicon (LTPS) TFTs by using a glass substrate. The glass is etched back to 50µm thick after the completion of OLED construction and thin film encapsulation. This etching step is used to reduce the thickness and achieve the desired bendability of 10cm radius or less. The display, including encapsulation, was <55µm thick, while the total display thickness became 0.25mm after laminating a 200µm thick polarizer to the front. Figure 3 is a picture of this Samsung SDI bendable display and a cross-section view of this display’s construction.
Different substrate materials, such as stainless steel and glass, were used to build the ultra-thin and flexible displays. However, because of the relatively high temperatures involved in depositing TFTs, the multilayer thin film encapsulation was the common choice and was proven essential in order to achieve the thickness, weight, flexibility, and durability goals.
Protecting thin film batteries
As displays and electronics become thinner and more flexible, so does the requirement for power supplies to have a similar form factor. Therefore, thin film and flexible batteries have become indispensable. A lithium phosphorus oxynitride (LiPON) thin film battery is considered a prime candidate for flexible electronics because it has the advantages of ultra thinness, high energy density, fast charging, and high temperature stability.
However, the use of lithium as the anode material makes the battery sensitive to moisture so it needs to be protected by a hermetic seal. Just like glass encapsulation used in the OLED displays, a mechanical encapsulation using metal or glass lid greatly diminishes the inherent merits of a thin film device. Thin film encapsulation is therefore necessary in order to fully exploit a thin film battery’s advantages.
However, using a thin film barrier to encapsulate LiPON batteries is more challenging than encapsulating OLED displays because the device surface topography is typically greater and the surface is rougher. The thin film encapsulation structure needs to be adjusted in order to be effective. Figure 4a is a lithium test sample with thin film encapsulation after 24 hrs in the damp heat test at 85°C/85% relative humidity (RH). The sample was encapsulated with the typical barrier structure used to protect OLED devices that has been developed over the past several years.
The barrier was not able to adequately protect the lithium because much of the lithium (metallic color portion) reacted with moisture and disappeared in a short period of time, although the identical barrier structure was proven to be effective for over 1000 hrs in the same test conditions for OLED devices. After adjusting the barrier structure to overcome the rough surface and extreme topography, Fig. 4b shows a similar lithium test sample after 200 hrs in the same damp heat test condition. Significant improvement was made simply by adjusting the barrier structure. The same coating structure was then used on real LiPON batteries for a customer (results are not shown).
Encapsulating thin film photovoltaics
Thin film photovoltaics, such as copper indium gallium selenide (CIGS), have the potential of low manufacturing cost and a flexible form factor. These types of thin film solar cells are gaining considerable attention that is exacerbated by the shortage of silicon material. Many companies are attempting to make CIGS solar cells on stainless steel foil in a roll-to-roll format in order to make the cell flexible while maintaining low manufacturing cost.
Although the solar cells are made on flexible metal foil, they are usually protected by rigid glass cover at the end of the process in order to meet the lifetime and moisture resistance requirements because some of the materials used in the cells are sensitive to moisture. Such a solution, although effective, not only diminishes the flexibility, but also adds weight and cost to the module.
Some companies try protecting the cells with plastic film instead of rigid glass to maintain the flexibility, but can only maintain short lifetimes because of the poor moisture barrier property of the plastic film. Figure 5a shows a flexible CIGS solar cell protected with Vitex’s Flexible Glass (polyethylene naphthalate (PEN) coated with the Barix barrier) on top. The cell was tested in a damp heat test at 85°C and 85%RH for over 1000 hrs.
 Figure 5. a) Flexible CIGS solar cell after 1000 hrs of damp heat test at 85°C/85%RH; b) open circuit voltage Voc after damp heat test. |
Due to the test and equipment limitations, it was not possible to measure the cell efficiency, so only open circuit voltage (Voc) was measured. Figure 5b shows the cell’s open circuit voltage over time after being stored in a damp heat test chamber. The Voc started to drop in <200 hrs in the damp heat test if the CIGS solar cell is not protected. The Voc further dropped to almost zero, which means the cell is almost dead, after 500 hrs in the test.
A flexible glass and a bare PEN were laminated on top of the CIGS solar cell by using commercially available ethyl vinyl acetate (EVA) film to compare the effectiveness of these films. It was observed on the sample laminated with bare PEN that Voc drop started later compared to the sample without any protection, but the cell still failed after 700 hrs. Therefore, the protection from bare PEN film is far from adequate because the IEC 61646 Terrestrial Photovoltaic Modules Design Qualification and Type Approval requires less than 5% efficiency drop after 1000 hrs in such test condition. In comparison, the sample protected with Flexible Glass held its Voc for over 1000 hrs without any change.
Barrier stability under UV and sunlight
Although the barrier has been proven effective under high temperature and high humidity test, more stringent tests are required because solar cells used for rooftops or solar farms need to be extremely reliable. The solar panels have to survive 30 years of harsh environments, such as UV exposure and extreme temperature.
UV damage to the organic material is an obvious concern so the organic material used in the Barix encapsulation was tested in various UV and sunlight irradiation to verify stability. A 0.5µm thick film of this organic material was deposited on 2-in. square glass substrates with the standard process.
Transmission spectra were measured by using a broadband spectrophotometry from n&k Technology. Samples A and B were then both exposed to UVB fluorescent lamps in an arrangement that complies with IEC 61646 for three weeks. The dosages were calculated to be equivalent to 15kWh/m2 in wavelengths between 280 and 385nm and 5kWh/m2 in wavelengths between 280 and 320nm. Sample A was then removed for transmission measurement while sample B was removed to be exposed to 100X solar irradiation (100kW/m2) by a sunlight spectrum simulator for two weeks for an equivalent dose of 33,600kWh/m2. The transmission spectrum was then taken on sample B.
 |
The table shows the average transmission from 350nm to 1000nm for all the samples before and after the tests. It can be seen that the transmission changes after such high dose UV and sunlight exposure are minimal. Barrier integrity under similar exposure will be studied in the near future.
Barrier stability at high temperature
A lamination process is often used to attach the barrier film to the device to be protected, such as a solar cell. This lamination process typically applies higher temperatures, such as 130°C to 150°C to melt EVA or thermal plastic in order to bond the films. Therefore, to demonstrate the applicability of our technology for these markets we have to make sure that our barrier film is not damaged by the high process temperatures. The calcium test is again used to verify the barrier performance in this study. We have verified that the barrier performance is maintained after being tested at 140°C for over 500 hrs.
Conclusion
The multilayer barrier developed has been proven effective in protecting various organic and flexible electronic devices including OLED, thin film batteries, and solar cells. Accelerated aging tests at high temperature and high humidity for thousands of hours have been performed and the positive results demonstrated. Film properties under severe UV exposure and high temperature were studied. Yield data demonstrate that the Barix process is a production-worthy process.
Acknowledgments
The author would like to thank H.K. Chung, PhD, and D.W. Han, PhD, at Samsung SDI for their continuing support for the development of thin film encapsulation. The author would also like to thank LG. Philips, UDC, and other unidentified collaboration partners for supplying the solar cells, thin film batteries, and test devices. Barix is a trademark of Vitex Systems Inc.
Chyi-Shan Suen received his MS in material science and engineering and MBA from U. of California, Berkeley and is the director of sales and marketing at Vitex Systems Inc., 2184 Bering Drive, San Jose, CA 95131 USA; ph 408/325-0366, e-mail [email protected].
Xi Chu received his PhD in material science and engineering from Northwest U. of California and is the director of process engineering at Vitex Systems Inc.