Process control and optimization is critical for low-cost, high-throughput, thin-film applications, such as barrier layers used in the food packaging industry. One useful technique for determining thickness and optical constants of the barrier layers during and after deposition is spectroscopic ellipsometry, which is a nondestructive and noncontact optical technique. Spectroscopic ellipsometry measures the change in the polarization state of light as it is obliquely reflected off of a thin-film sample, as shown in Figure 1. This change in polarization is represented, at each wavelength, by two parameters: ψ, which is an amplitude ratio, and Δ, which is a phase difference. These two parameters are related to the Fresnel reflection coefficients (rp and rs) for the sample, according to ρ = tan ψ eiΔ = rp/rs, where rp and rs provide information about the optical constants of the sample.
Figure 1 – Schematic of a spectroscopic ellipsometry measurement in which linearly polarized incident light (right) changes to elliptically polarized light (left) after oblique reflection off of the sample. Modeling the change in polarization provides thin-film parameters such as thickness and optical constants.
In the case of a phase-modulated ellipsometer, which is the instrument used to collect the raw data shown below, the measured parameters are Is and Ic, which are related to ψ and Δ according to: Is = sin 2ψ sin Δ and Ic = sin 2ψ cos Δ. In order to analyze ψ and Δ, or Is and Ic data, a model, representative of the film structure, must be created. Regression analysis can then be applied to model the sample at hand in order to determine its properties, such as thickness and optical constants. Further analysis of the optical constants can provide additional information about the sample such as its composition, crystallinity, band gap, roughness, etc., making spectroscopic ellipsometry ideal for many applications. Spectroscopic ellipsometry can be performed either in situ or ex situ, depending on what one would like to know about the sample or material under study.
Figure 2 – a) Ex situ raw spectroscopic ellipsometry data (Is and Ic) for a bare PET substrate. b) Optical constants of a bare PET substrate, derived from the spectroscopic ellipsometry data shown in (a).
Generally, plastic and paper products used for food packaging are permeable to gases (e.g., CO2 and O2) and vapors (e.g., water and odor), all of which can cause the package contents to quickly spoil. To protect the contents and enhance shelf-life and quality, the plastic and paper products are usually coated with a thin, transparent barrier layer, which reduces the permeability to outside gases and vapors. This barrier layer must be flexible, conformal, of high quality, and thin enough to allow the consumer to see the contents inside the package, but thick enough to adequately protect the contents of the package. Also, the overall deposition process should be fast and as economical as possible, which requires strict process control.
Spectroscopic ellipsometry is ideal for monitoring the thickness and quality of the barrier layer, in real time, as it is deposited onto a moving web of plastic or paper. This real-time measurement allows one to vary the deposition parameters in order to obtain the ideal barrier film thickness and quality. Spectroscopic ellipsometry can also be performed ex situ, or after deposition, as will be shown below.
Figure 3 – a) One-layer model used to determine the thickness and optical constants of a barrier layer deposited onto a PET substrate. b) Ex situ raw spectroscopic ellipsometry data (Is and Ic) and modeled fit for a barrier layer deposited onto a PET substrate. c) Optical constants of a barrier layer deposited onto a PET substrate, obtained from the Lorentz dispersion function used to model the barrier layer.
For all data shown, the substrate of choice is polyethylene terephthalate (PET), which is a transparent plastic commonly used for food packaging applications. Before characterization of the barrier layer can be achieved, the optical constants of the underlying substrate must first be measured so that they can be used in the optical model of the sample containing the barrier layer. The raw ex situ spectroscopic ellipsometry data for the PET substrate are shown in Figure 2a. In the region of 1.5–4.0 eV, the PET substrate is transparent, and interference fringes, due to backside reflections, are visible.
At higher energies, however, the PET substrate becomes completely absorbing and the backside reflections do not need to be taken into account. For this reason, the region from 4.0 to 6.5 eV was used for characterization of the bare PET sample, as well as for characterization of the PET sample coated with the barrier layer. The optical constants of the bare PET, which were directly extracted from the raw data, via inversion, are shown in Figure 2b.
The PET sample coated with the barrier layer was accurately characterized and modeled using a single-layer model, as shown in Figure 3a. The optical constants used for the PET substrate are those obtained from Figure 2b. The barrier layer was modeled using a classical Lorentz oscillator in order to obtain thickness of the layer, as well as its optical properties. The raw data and the model generated fit to the raw data, over the spectral range from 4.0 to 6.5 eV, are shown in Figure 3b. Also shown, in Figure 3c, are the optical constants for the barrier layer, obtained after the fit. The optical constants show normal dispersion and confirm that the barrier layer is transparent over the entire spectral range.
Real-time, in situ measurement using a spectroscopic ellipsometer
Figure 4 – Barrier layer thickness, as a function of time, for a roll-to roll deposition process; the original film thickness was 570 Å, and then the process parameters were varied to obtain a thickness of 270 Å.
All of the previous measurements were performed ex situ, but spectroscopic ellipsometry is also well suited for real-time, in situ measurements. Such measurements can provide fast, reliable, and nondestructive characterization of thickness and quality of the barrier layer films for quality control purposes. In this case, the spectroscopic ellipsometer is mounted onto a roll-to-roll coater to measure the thickness of the barrier layer in real time. The main challenge of such measurements is acquiring accurate data on a quick-moving, vibrating surface. For these applications, the UVISEL system (HORIBA Scientific, Edison, NJ) allows for ultrafast measurements (a full spectrum in as little as 60 msec), allowing data to be collected every few centimeters along the web length.
While the ellipsometer constantly acquires and analyzes data, the overall process control parameters can be varied as needed to achieve the ideal barrier layer. An example of data that can be provided by an in situ measurement is shown in Figure 4. Here, for the first 2 min of data collection, the thickness obtained was around 570 Å. After 2 min, the deposition parameters were changed and the new thickness obtained was around 270 Å. Figure 5 shows a photograph of the roll-to-roll in-line spectroscopic ellipsometer used for these measurements.
Figure 5 – Photograph of roll-to-roll in-line ellipsometer used for studying barrier layers on PET substrates.
Other thin-film applications of spectroscopic ellipsometry
Spectroscopic ellipsometry is extremely useful for many thin-film applications since it can provide thickness and optical constants in a nondestructive manner. Ideally suited for in situ measurements, spectroscopic ellipsometry can also monitor thickness and optical constants during deposition for real-time process control. Also used to obtain information such as band gaps, crystallinity, and composition, spectroscopic ellipsometry can be applied to many fields besides food packaging—basically, any applications that utilize thin-film technology and can benefit from the use of spectroscopic ellipsometry for characterization and monitoring. In addition to food packaging, common applications in which spectroscopic ellipsometry is helpful include measuring thin films for solar, biosensing, microelectronics, optical coatings, and optoelectronics applications.
Michelle N. Sestak, Ph.D., is Ellipsometry Applications Scientist, HORIBA Scientific, 3880 Park Ave., Edison, NJ 08820, U.S.A.; tel.: 732-494-8660, ext. 8220; e-mail: [email protected]; www.horiba.com/ellipsometry