Flexographic printing of ultra-thin semiconductor polymer layers

In order to achieve the desired sub 100 nm dry film, a low polymer concentration of 8 mg ml⁻¹ was dispersed in Xylene. The solution was placed in a glass bottle on a roller until fully dissolved (typically overnight to ensure complete mixture of polymer). The solution was then stored in an air conditioned laboratory out of direct sunlight.

The solution surface tension was then characterised by employing the pendant drop approach [14–16] determined through using a Fibrodat DAT1100 (Fibro System AB, Sweden). 1 ml of the solution was weighed using a scale to determine the density and 8 droplets of 8 μl volume were measured.

Prior to treatment of the glass substrate and printing plate, the effect of printing parameters on deposition quality was investigated. The glass tiles were cleaned with IPA (isopropanol) in an ultrasonic bath for 3 min at 30 °C, rinsed with deionised water and then wiped dry using a lint free cloth.

The effect of flexographic printing parameters was investigated using an IGT F1 tester (figure 3).The device has a digital display through which the printing variables (printing speed, printing force, anilox force) can be varied. An anilox with a capacity of 12 c.c. m⁻² was loaded into the system and the glass was placed onto a rigid substrate carrier which allowed it to pass under the printing plate when it rotated. Once the substrate was loaded, 1 ml of solution was added directly to the anilox where the surface of the anilox meets the doctor blade. To perform a print, the anilox was engaged with the plate and rotated. The plate was then engaged with the substrate and rotated further to produce a print. The glass was then transferred to a hot plate at 100 °C for 30 s to allow the solvent to fully evaporate. The settings were varied to determine optimum printing conditions. The anilox to printing plate force was kept constant at 150 N as preliminary testing showed sufficient transfer of solution between the anilox and printing plate. The printing plate to substrate force was varied from 20 N to 80 N in 20 N intervals, the printing force was varied from 20 N to 80 N in 20 N intervals and the printing speed was varied from 0.2 m s⁻¹ to 1.0 m s⁻¹ in 0.2 m s⁻¹ intervals.

Zoom In Zoom Out Reset image size

The topography of the printed samples was assessed using a white light interferometer (Veeco NT9300 wide area white light interferometer). The WLI was set to scan in PSI (phase shift interference mode) due to the sub 160 nm features present on the samples. Images were captured at the centre of the samples rendering a 3D topography plot for a projected area of 1.25 mm  ×  0.94 mm (a resolution of 640  ×  480 pixels with sampling at 1.96 μm intervals). This showed the amount of dried polymer film remaining on the glass and the print quality. 2D profiles were captured for film continuity analysis and to compare varying print quality for each experiment. Film thickness was determined by making small cuts into the film. This allowed step height measurements to be made.

The glass substrates were cleaned using IPA and exposed to UV/Ozone using Novascan^™ PSD-UV system for 30, 90, 120, 240, 360, 480 and 600 s. Surface energy was determined through the Liftshitz-van Oss theory [17–19] where the contact angles of water, diiodomethane and ethylene glycol were inputted. The theory used the Liftshitz-van der Waals component (${{\gamma}^{\text{LW}}}$) and the acid–base component (${{\gamma}^{\text{AB}}}$) to formulate the following equation:

$$\begin{eqnarray}&&{{\left({\gamma}_{\text{s}}^{\text{LW}}{\gamma}_{\text{l}}^{\text{LW}}\right)}^{0.5}}+{{\left({\gamma}_{\text{s}}^{+}{\gamma}_{\text{l}}^{-}\right)}^{0.5}}+{{\left({\gamma}_{\text{s}}^{-}{\gamma}_{\text{l}}^{+}\right)}^{0.5}}=0.5\left(1+\cos \theta \right)\end{eqnarray}$$

The properties of the 3 liquids and reference properties of O-Xylene are shown in table 1 [20–22]. This is a diverse triplet with diiodomethane being purely dispersive and water and ethylene glycol being both polar and dispersive, but with differing negative and positive polar components. Using these values, together with the measured contact angles, the free surface energy of both the glass samples and the substrate were estimated. Measurements were taken for uncleaned, solvent cleaned and UV/Ozone treated glass. Dedicated equipment for measuring contact angle (Fibrodat) could not be used due to the inability of the equipment to measure low wetting angles. Therefore to perform the measurements, drops of liquid were applied via a micropipette and photographed with a camera. Contact angle was measured using image analysis software (Image J 1.46r, U.S. National Institutes of Health).

Printing tests were repeated using glass substrates subjected to varying durations of UV/Ozone exposure. Furthermore, the flexographic plate was exposed to UV/Ozone and the effect on print quality assessed. Both the glass slides and the printing plates were exposed for up to 10 min with measurements taken on separate samples for different exposure times within this range to determine the effect on print quality.

The contact angles of 3 liquids (water, diiodomethane and ethylene glycol) on glass substrate and the effect of UV/Ozone on these are shown in figure 7. Both water and ethylene glycol showed a significant decrease in contact angle with UV/Ozone treatment, with longer treatment times giving greater reductions in contact angle for up to 6 min with contact angles below 5° indicative of almost complete wetting. The mechanisms that contribute towards this are thought to be removal of hydrocarbon contaminants as well as introduction of oxygen groups to the surface [23–25]. The contact angle of diiodomethane did not decrease to the same extent as deionised water or ethylene glycol but some reduction was observed with contact angles below 10° measured. The solid free surface energy of the glass tiles after various UV/Ozone exposure times was estimated and is shown in figure 8. Even after the first treatment of 30 s of UV/Ozone exposure, the glass substrate showed an increase in surface free energy from 47.1 mN m⁻¹ to 51.7 mN m⁻¹. The surface energy continued to increase to around 53.6 mN m⁻¹ after 10 min of exposure. Through introduction of oxygen species and the removal of organic contaminants, an overall surface energy change of 6.5 mN m⁻¹ was observed.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The contact angles of the 3 liquids (deionised water, diiodomethane and ethylene glycol) were determined and are plotted in figure 9. The plate material displayed higher contact angles than the glass substrate. Deionised water showed a decrease in contact angle from 100° to 50°. Diiodomethane did not show a consistent decrease in contact angle however. Within the first 30 s of exposure, the contact angle increased from 34° to 46° before decreasing to a value of 24° after 10 min of exposure. Ethylene glycol showed a slight decrease in contact angle from 68° to 63° after 10 s of exposure before increasing to 68° after 2 min and then showing a significant drop to 30° after 10 min of exposure. The increasing contact angle resulted in a changing surface energy of the rubber printing plate (figure 10). The overall surface free energy of the plate decreased after the first 30 s of exposure from 42 mN m⁻¹ to 34 mN m⁻¹. The surface energy then increased to 49 mN m⁻¹ after the full 10 min of exposure. This reduction in surface energy was seen to be consistent over repeated measurement. Although the reason behind this phenomenon has yet to be fully understood, one explanation could point to oil-based chemicals such as waxes, present in some photopolymer plates, migrating to the surface of the plate through ultraviolet exposure [26]. This effect could then result in the decrease in surface energy which was observed in this experiment.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Surface topography plots for prints on UV/Ozone exposed glass are shown in figure 11 alongside corresponding prints made on untreated glass at a print force of 40 N and a varying print speed between 0.2 m s⁻¹ and 1.0 m s⁻¹. It was observed that generally more solution was deposited as a result of treatment of the glass. As the print speed increased to 0.4 m s⁻¹ and 0.6 m s⁻¹ more solution was deposited and spread across the glass surface due to the increased wetting nature. When printing at a speed of 1.0 m s⁻¹, the film formed had fewer local peaks due to contaminants but more holes, or gaps, were observed due to the increased wetting nature of the surface.

Zoom In Zoom Out Reset image size

To exploit the decrease in surface energy of the printing plate, another film was printed after exposing the substrate and printing plate to UV/Ozone (for 10 min and 30 s respectively) and at a printing force of 40N and printing speed of 1.0 m s⁻¹ (figure 12). The quality of the printed film increased considerably, with even coverage of polymer over the glass substrate. This improved film is thought to be a result of the greater surface energy differential between the glass substrate and rubber printing plate so as to make transfer to the glass more energetically favourable and thus induce greater solution transfer. The film thickness was determined through white light interferometry of cuts made into the film (as demonstrated in figure 13). A representative sample showed an average film thickness of 56 nm (with a standard deviation of 11 nm based on 9 measurements) which is ideal for the hole injection layer present within OLED.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size