PHTN1432: Vacuum Systems and Thin Film Technology
Lab - High Performance Antireflective Coatings (2015W)



Compute the deposition thicknesses and TIMES and bring these precalculations to the lab on a plastic sheet (overhead plastic sheets are ideal). Without these precalculations it is impossible to proceed in the lab.


The principles of single-layer (quarter-wave) coatings were covered in class. To recap, many practical antireflective coatings are manufactured using a single quarter-wave film of magnesium fluoride (n=1.38). For a design wavelength of 550nm, a 996 Angstrom layer is required. With an incident wave of 550nm, a reflection from the front surface (air-to-film) interface occurs with an intensity predicted by the Fresnel equations at 2.55% (0.1597 squared). The transmitted portion of the wave proceeds through the film, reflects from the film-to-glass (n=1.5) interface, and proceeds to partially cancel the first reflection since this wave is shofted by one-half wavelength (i.e. it is 180 degrees out of phase with the original reflected wave). This second wave has an intensity of 0.17% (0.04167 squared). Taking the difference in the intensity coefficients then squaring yields a reflection of 1.39% ... much better than the 4% per surface for uncoated glass!

Oddly, at some wavelengths (275nm), the reflection is a maximum. In this case the 275nm wave encounters, effectively, a half-wave thick coating and so encounters a full-wave shift. In this case the first reflection (air-to-film) and second reflection (film-to-glass) are in-phase and add to form a reflection of 4.05%. At 183.3nm, the wave inside the film encounters a phase shift of 1.5 waves (equivalent to a half-wave shift) and so another minima of reflection occurs.

Simple coatings such as this are preferred for some applications such as curved surfaces since the reflectivity of the coating will never exceed that of an uncoated substrate. The coating employed in this lab (a Quarter-Quarter coating) can result in high reflectivities at some wavelengths - the opposite of the intended effect - however such coatings can also yield much lower reflectivities at a single wavelength.

Quarter-Quarter coating performance

When two quarter-wave thick layers (design wavelength = 500nm) of ZnO and SiO are employed, the performance of the coating is predicted as per above (via 'FilmStar' by FTG software, a commercial thin-film design package). This is the simplest of a class of coatings called "V" coatings, so-called because of the shape of the curve.


The PVD-75 is configured with SiO as target #1 and ZnO as target #3. Refer to the SOP page for details on the operation of the system. Coatings will be controlled by means of deposition time at a known power/pressure (this is a sputtering system with great consistency in results). Precalculate (before entering the lab) the thickness of each layer as well as the deposition time required.


Quantitative analysis requires the use of a spectrophotometer. Using the Lambda-3B spectrophotometer in the spectroscopy lab, determine the transmission curve for the coating. Be sure to use a glass blank as a reference. Convert the %T to %R, enter into a spreadsheet, and compare the results to the theoretical model.

Since you are plotting the spectral data on the same graph as the theoretical data, you might do well to acquire data manually at 10nm intervals - simply use the Lambda-3B by starting at 800nm and entering new wavelengths at decreasing intervals of 10nm while recording the transmission.

Quarter-Quarter Prediction Data from the FilmStar design. Use this data to compare with the actual performance of the coating. Plot BOTH the experimental and theoretical performance on the SAME GRAPH.

V Coating performance

The performance of an actual prototype V-coating developed in our lab is seen here. Not believing that the high reflectivity around 340nm could not somehow be attributed to the characteristics of the glass substrate, the red curve illustrates a sweep of a blank glass substrate - the blue curve does indeed illustrate the actual spectral performance of the coating! When using the USB DAQ on the Lambda-3B, be sure to eliminate areas where the filter switches (of course, since you are using the data for the spreadsheet, it is easy to be precise).

Lab Report

Prepare a condensed lab report as follows:

Copyright (C) Niagara College, Canada, 2006-2007
This page is part of the PHTN1432 Course Page