Production of the thin films
All thin films were produced by PVD with in a BAK 550 box coater with a planetary system. Evaporation of the materials was done by an electron beam gun. The samples were manufactured in fall 2008 in Lint, Belgium (see Figure 24). The thin films were deposited on polished Si wafers and 1 mm thick borofloat glass from Schott. For the stress measurements, 0.1 mm thick borofloat glass was used as substrates. In each batch, all three substrates were coated simultaneously. The layers were made relatively thick, 2000-5000 Å, in order to make the etching observations easier and to be able to find the refractive index and thickness TP.
All 10 thin films were grown at a substrate temperature of 325 °C. The evaporation properties differed slightly among them, but the best known evaporation parameters were used shown Table A1. The thin films S1 and S2 were grown with an 80 Å thick aluminium interface layer as an endpoint layer for easier etch rate determination.
Five additional complete magenta filters were also fabricated, M200-M325 processed at substrate temperatures between 200–325 °C. These filters were fabricated at relatively high O2 pressure. Another magenta filter called N250 was also added in the study. This was processed at lower oxygen pressure at 250 °C. from a standard production batch in the late 2009. The N250 filter was used in the later stage of the study to evaluate etching performance.
Physical characterization of the thin films
The 10 different single layer thin films in the first series produced at the late 2008 were characterized by ellipsometry and by transmission spectral curve fitting, respectively. The refractive index n, physical thickness TP, thermal behaviour and stress were thus evaluated.
The ellipsometric measurements were done on a SE-400 Sentech instruments equipped with a HeNe-laser which generates light at a wavelength of 632 nm. For each measurement, three different incident angles were selected for the laser thereafter computer software calculated the most probable reflective index and thickness. It was very important to keep the surface as flat as possible and the alignment of the substrate was critical. Measurements were done only on the thin films deposited on Si-wafers. Coating S1 and S2 could not be measured because both contained aluminium layers underneath.
By the measurements from a spectrophotometer in the range of 200-900 nm, it was possible to make curve fitting for each coating in order to match a simulated curve from the TFCalc software. The curve fitting models were optimized around 500 nm and the surrounding wavelengths were not fitted equally well due to the dispersions of the thin films. The information was still enough to find the thickness TP. This method resulted in more reliable values compared to the ellipsometric measurements. It was also possible to extract dispersion data but these calculations were not done.
The physical thickness was measured by a Tencor Alphastep 200 stylus profilometer. In order to measure the TP, some minor areas of the Si wafers was covered by pieces of thin aluminium foil during deposition. Thereafter the aluminium pieces were removed, creating sharp edges on the thin films. The stylus could then travel over these edges to find the TP.
Because this instrument was not well-calibrated, this part of the measurements was not carried out for all the thin films. Only three measurements were done on Y1 and one measurement on Z2.
Thin film stress
Borofloat glasses with a thickness of 0.11 mm and 48 mm in diameter were coated with the selected compounds from Table 1. By observing the resulting curvature, it was possible to measure the internal stress of the films. A very simple setup was used where the test pieces were placed in front of a pinhole with a cross mark. By observations through the pinhole at different distances to the substrate, it was easy to find the radius of curvature of the substrate.
The surface stress caused a curvature of the substrate and created a focal point. This focal point could easily be found by the pinhole. By turning the substrate, so a concave surface always was directed towards the pinhole, both compressive and tensile stress could then be measured.
In order to measure the thermal shift, two spectra were taken from each coating. The first spectrum was taken at room temperature. After that the sample was heated to 300-320 °C, the second spectrum was then taken 10 minutes after this condition has reached. The spectral shift caused by the heating was recorded. This gave a rough indication about the thermal stability of the filter. In the case of a multilayer filter, the shift was measured at a point in the spectrum where the transmission was at 50 % at the right cut-off. In the case of a single layer filter, the shift was measured from a local minima and a local maxima point.
2 Theoretical background
2.1 OPTICAL FILTERS
2.2 THIN FILM PROCESSES
2.4 WET ETCHING CHEMISTRY
2.5 ETCHING ANISOTROPY
2.6 POURBAIX DIAGRAMS, PH AND REDOX ENVIRONMENT
2.8 THERMAL SHIFT
2.9 THIN FILM STRESS
2.10 OPTICAL CHARACTERISATION OF THE THIN FILMS
3.1 PRODUCTION OF THE THIN FILMS
3.2 PHYSICAL CHARACTERIZATION OF THE THIN FILMS
3.3 THE IN-SITU ETCHING INSTRUMENT
3.4 ETCHING TESTS
3.5 ETCH MATCHING OF Y2O3/SIO2
3.6 ETCHING RATE OF PHOTORESIST
4 Results and discussion
4.1 CHARACTERISATION OF THE THIN FILMS
4.2 ETCH TESTS
GET THE COMPLETE PROJECT
WET ETCHING OF OPTICAL THIN FILMS