For this experiment the thickness of thin water-based coating on PET was measured using an MProbe NIR system (900 nm -1700 nm wavelength range). A typical method of measuring coating thickness of PET is by the FFT approach, using the MProbeVis (400-1000 nm) or MProbeVisHR (700-1100 nm).
Limitations in the Measurement of Water-Based Coatings on PET Film
Although this approach gives quick, easy and accurate results, it is difficult to use for the samples here because:
- The PET coating is relatively thin (400-600 nm)
- The refractive index of the coating varies from sample to sample because solids are dissolved in water in different concentration.
- There is a surface layer ~ 300-400 nm thick on the PET due to plasma treatment
This makes it necessary to use curve fitting to determine the thickness of this coating i.e. to fit the model to measured data. In the visible light range, both PET and coating are transparent and the effect of the coating can clearly be seen in the measured data (Fig. 1).
However, both PET and coating will still absorb some visible light and the level of PET absorption is dependent on surface conditions i.e. it changes from sample to sample. This makes it difficult to make consistent measurement.
Figure 1. Measured reflectance spectra of the samples in the Visible wavelength range (400-1000 nm)
Figure 2. Measured reflectance spectra of the samples in the NIR wavelength range (900-1700 nm)
Figure 3. Fit of the model to uncoated PET measurement. Thickness of the PET, plasma treatment layer and effects of the backside reflection are determined. Optical dispersion of the plasma treatment layer is also adjusted.
Measuring the Coating Thickness
For measurements taken in the NIR range (900-1700 nm), it can be seen that there is just a trace amount of absorption and effect of the surface treatment is much weaker (Fig. 2)
The first step to build the model is to fit it to the uncoated PET sample. Backside reflection, the presence of the plasma treatment layer, etc. need to be incorporated.
Once the base model has been constructed, it is time to measure the coating thickness. Due to variations in solute concentration in the water solution the coating material refractive index is changing. A Bruggement effective medium approximation can represent this – basically a physical mixture of dispersions of a “material” and water.
If one of the calculated parameters is set as a volume fraction of the water, the refractive index can then be adjusted in a controlled fashion in alignment with the actual physical sample. But first the dispersion of the base material must be calculated. One of the samples has a coating with the highest concentration of the “solid material” in the coating material. The dispersion of this material can be determined and defined as a reference material.
Figure 4. Fit of the model to data. The thickness of the coating (501 nm) and plasma treatment layer (300 nm) is determined. Coating material dispersion is represented using Cauchy approximation (see Fig. 5 for determined dispersion)
Figure 5. Dispersion of the base coating material is determined from the measurement.
Figure 6. Fit of the model to measured data (sample#2). Thickness of the coating: 648.7 nm, plasma treatment layer: 384 nm, water volume fraction in coating material: 29%
Figure 7. Coating material dispersion represented by EMA. Volume fraction of water 29% (relative to base) determined from the measurement
Figure 8. Fit of the model to measured data (sample#4). Thickness of the coating: 486.3 nm, plasma treatment layer: 359 nm, water volume fraction in coating material: 2%
Finally, the dispersion of the base material can be used in the production of an EMA material (base material+water) and we have a model ready for analysis of all the samples.
Measurement of the water-based coating on the PET film was simplified through the use of the MProbe NIR system. Although there were a relatively high number of variable parameters, it was possible to build a simple model that calculated accurate and consistent measurements.
This information has been sourced, reviewed and adapted from materials provided by SemiconSoft.
For more information on this source, please visit SemiconSoft.