PHTN1432: Vacuum Systems and Thin Film Technology
Lab - Dielectric Mirror Coatings (2018W)


The prelab (with unique parameters for each individual lab group section) is REQUIRED to be completed by each student and submitted on the date specified (and at a time PRIOR to the scheduled start of the lab session). Students who fail to submit the prelab or submit it late (late prelab submissions are worth ZERO) will be penalized -20% on the lab. If all students in the entire group fail to submit a complete design with a predicted spectrum matching the required design, the entire group will not be admitted into the lab, will not be able to fabricate a device, will be marked as absent, and all members of that lab group will forfeit 50% of their marks in this lab immediately! NO EXCEPTIONS!

eBeam Deposition System


Each person in each lab group must submit a prelab consisting of a Filmstar design report for the mirror coating showing all layers including the GLASS SUBSTRATE and the PROTECTIVE TOP LAYER on the design. Without this design it is impossible to proceed in the lab. Multiple, independently developed designs will help detect errors in a particular design.

Embed your name in the TITLE of the graph and ensure your prelab submission is displayed in Physical thicknesses (for all layers).

Mirror wavelengths are chosen based on the fact that the reflectivity peak spans over a range of 0.9 to 1.1 times the target wavelength and so will cover a good portion of the gain spectrum of the target laser. Design the mirror based on your specific lab group as follows:

Lab Group Lab Date Center Wavelength (nm)
Wednesday a.m. Group A 2018/04/11 532
Wednesday p.m. Group A 2018/04/11 555
Thursday Group A 2018/04/12 593.5

FAILURE TO USE THE CORRECT WAVELENGTH IN YOUR FILMSTAR DESIGNS WILL RESULT IN AUTOMATIC FAILURE OF THE LAB - If the wavelength in your designs does not match that of your assigned lab group it will be considered plagiarism and a mark of zero will result for the entire lab

Filmstar models MUST BE UNIQUE TO EACH STUDENT. Plagiarized models (including the prelab) will result in a mark of ZERO for the entire lab and a trip to the Chair's office for academic dishonesty resulting in possible expulsion.

Failure to submit this prelab at least fifteen minutes prior to the lab will result in an immediate -20% penalty on the entire lab mark. NO LATE SUBMISSIONS OF THE PRELAB will be accepted.

Building the FilmStar Model

First, review the structure of a dielectric mirror and specifically the order of the layers from class notes. Now, leaving the basic units in OPTICAL thickness, begin by building the stack from quarter-wave layers (i.e. "0.25x" where "x" represents the symbol for the material used such as MgF2 or ZnS). Begin with a massive glass layer (5mm of glass for this OC) and build the stack from there. The order of the layers is very, very important! Be sure to set the design wavelength as mentioned in the lectures.

In the course of your prelab research you might have found two possible values for the index of refraction of ZnS: choose one for this model however during analysis of results try both values to see if one "fits" better than the other.

Each layer should be, nominally, of thickness λ/4n where λ is the wavelength of the design and n is the refractive index of the individual material used. Other combinations will work however ther may not be optimal and, regardless, will be much thicker so that the deposition may take longer than anticipated and may well consume more material than is available in a crucible!

When done, convert all layers to physical thickness for fabrication in the lab (remembering that these thicknesses are in nanometres and our XTM/2 monitor reads in Angstroms).

NOTE: This lab will be completed on Sparky, a high vacuum eBeam evaporator system in which layer thickness will be monitored using an Inficon XTM/2. Density, Z-Ratio, and Tooling Factor are pre-programmed into the unit for each material used: the target selection on the unit matches the film # on the monitor.

Target Configuration (Current as of 2018/04)
#2: Magnesium Fluoride (ρ=3.150, z=0.637, Tf=220%)
#3: Zinc Sulphide (ρ=3.980, z=0.775, Tf=220%)
#6: Aluminum (ρ=2.700, z=1.080, Tf=220%)

The tooling factor was verified correct (within 7%) for substrates on the lower rung of the holder only.


The coating is to be applied using SPARKY, a six-target cryopumped eBeam system. When you enter the lab, the CTI cryopump will already be running (since it requires at least several hours to precool). Verify that the cryopump is ready (via the attached monitor) before proceeding.

Begin by venting the chamber then opening it. Load one substrate per student into the holders. Ensure the required target crucibles are filled (this is important since many layers will be deposited). Ensure that the crystal (XTM/2) is operating properly. Now lower the chamber, rough-pump to below 100mTorr, and open the gate valve. When the pressure reaches the low 10-5 torr range (< 2*10-5, usually about 20 minutes) deposition may begin.

Turn the CV-8 eBeam supply ON, ensuring all interlocks are closed. Before applying power, ensure both the operating voltage and the emission current controls are set to zero. Now proceed to deposit all layers in the proper order.

The deposition process is as follows:

  • Select the proper target crucible (first layer next to the glass
  • Select the corresponding film number on the XTM/2 deposition monitor
  • Ensure the shutter is closed
  • Turn the high voltage ON, adjust for 8kV
  • Turn the emission current to zero, the filament ON, and increase the emission current SLOWLY until a faint violet glow is seen (the emission current will be very low at this point and will likely barely register on the meter)
  • Align the electron beam to sweep across the surface of the material in the crucible - ALIGNMENT OF THE ELECTRON BEAM IS CRITICAL .... if the beam hits the crucible either the crucible will be damaged or circuit breakers will blow when the emission current is increased. Adjust the width of the sweep to zero (it is preset) and the longitudinal sweep control such that the square swept area is centered-on and just inside the crucible. The beam will stay aligned even when a new crucible is selected. Note that alignment is easiest to perform on the MgF2 crucible since that material glows white when struck with energetic electrons.
  • eBeam Alignment - Crucible

  • The first time a crucible is used, ramp the power VERY slowly to allow it to outgas and prevent splattering
  • Keep the shutter closed and ramp-up the emission current to the point where a bright violet glow is seen
  • Open the shutter and increase the emission current to control the deposition rate as required
  • Use the shutter to stop the deposition
  • Set the emission current to zero and wait one minute for the melt to cool before switching crucibles
  • When switching layers decrease the emission current to zero but the high voltage (and filament) may be left on
  • Verify the beam position for each target (via the violet glow at low emission current) before ramping the power up for the next material - realignment should NOT be required (so deposition of multiple layers will go reasonably quickly)

eBeam Deposition System - Crucible The violet glow as seen through the window of the chamber. The filament is visible in the lower part of the photo and the crucible is seen in the upper part. The crucible is not visible directly, but through the periscope arrangement which ensures the window is always kept clear of the deposit. Of course, the deposited film is also deposited on the lower mirror and so the colour of the image may well change during deposition.

With MgF2 use a rate of 5 to 10 Angstroms/second and for ZnS a rate of 4 to 6 Angstroms/second. Since ZnS sublimes, be careful when controlling the rate of deposition - it may "run away" quickly. While exceedingly high rates are possible (especially since Sparky features a 13kW supply), splattering may occur resulting in poor surface quality of the coating and possible damage to the monitor crystal - be especially careful with a 'freshly' loaded crucible. The other concern is evolved gases (again, a problem with freshly loaded crucibles) which may cause a pressure increase sufficient to permit an electrical discharge (i.e. arcing) of the eBeam supply into the chamber - low rates, especially on the first few depositions, will prevent this! Remember, too, to degas each material the for the first use only by slowly heating it with the shutter closed (subsequent layers will not require this process but the shutter is still required to terminate the deposition).

Deposit ten alternating layers. Be prepared when you enter the lab as this is all you will have time for in the two-hour lab: each layer will take about 2 minutes to deposit plus a minute of cool-off between switching the hearth targets plus ramp-up time. Be sure to program the XTM/2 monitor for each film (simply select the new film number, all other parameters are preprogrammed including density, z-ratio, and tooling factor). Write-down the actual layer thickness deposited. More layers means higher reflectivity of the finished mirror. Be sure to also deposit the top protective layer.

When all layers are complete shut down the CV-8 power supply (remove the key for safety), close the gate valve, vent and open the chamber, and USE THE SAFETY GROUND TO DISCHARGE THE FILAMENT TERMINALS (Wherever a safety grounding probe is provided, USE IT!).

Check the XTM/2 monitor, as well, for the current tooling factor.


Quantitative analysis requires the use of a spectrophotometer. Using the Lambda-3B spectrophotometer in the spectroscopy lab, determine the reflection curve for the coating (all substrates should be the same - verify this by sampling at least three from different locations in the chamber) - the reflection curve may be determined using the reflection adapter.

If using the reflectance accessory ...
The UV lamp is not required, only the VIS. Before starting the unit ensure the alignment mirror is in place on the accessory. Once ready, carefully remove the alignment mirror and replace it with your dielectric mirror (coated side down). Now, run the spectrum from 800nm to 315nm (using the "800-315" reference file). When all runs are complete, replace the alignment mirror.

Transmission only ...
If the reflection adapter is not installed, simply run the transmission spectrum as per the previous lab and take the inverse of the transmission curve (in Excel) by subtracting 100% minus transmission for each point. Be sure to use a glass blank the same as used for the deposition as a reference.

Compare the observed results to the theoretical prediction from FilmStar (using the actual layer thicknesses deposited) and try to "fit" the FilmStar model to the observed spectrum by modifying layer thicknesses as well as changing the n value of the ZnS layer (as outlined below).

Lab Report

Prepare a condensed lab report as follows:


  1. Outline how a Dielectric high-reflector coating works. Include a diagram showing each reflection at the interface between materials and the relative phase of each reflection.
  2. The initial design of your coating - include a "cross-section" diagram of the layers (which shows the glass substrate and the relative location of all layers deposited). The design must include the thickness of each layer as well as an unedited FilmStar report (of the initial design).
  3. Calculations (done manually, and with all work shown explicitly) showing the predicted maximum reflectivity of the coating at the design wavelength. Use the Fresnel equations to compute the reflection coefficient of all major reflection components - at a minimum consider the first term in the series for each component (i.e. you may neglect small sub-reflections). For example, if the transmitted component from the first air-to-film interface was 95%, the reflection from the next film-to-film interface would be 0.95 times the reflection, and so on. How does this computed reflection figure compare to the prediction of the FilmStar design?

  5. Submit a corrected FilmStar report outlining the expected performance of the mirror that was actually fabricated in the lab (using thicknesses actually deposited as read from the XTM/2 - the tooling factor given in the lab is correct). The FilmStar report must be included verbatim and without modification - sharing of FilmStar models will be considered plagiarism and will result in a mark of zero for the entire lab.
  6. Submit a copy of the spectrophotometric reflection curve for the mirror (only one required since the chamber is designed for consistency). Identify λcenter. This might be a calculated reflection based off the measured transmission of the device (see the lab).
  7. Submit several FilmStar models which attempt to match the observed results from the lab as closely as possible. This will attempt to rationalize and explain discrepancies between the theoretical curve of the FilmStar model and the actual performance of the coating. Two parameters must be modified in an attempt to "fit" the model:

    #1: Try changing the n value of the ZnS layer. In your research will will have noted that ZnS has two values of n listed, depending on the source. Try to model the filter using BOTH values to see if one is closer than the other and include both outputs to allow comparison between the predicted output in each case. Include at least ONE FilmStar model using each n value which attempts to "fit" the model to the results achieved! Explain, in each case, which value was used and ultimately which produces the best fit to observed results.
    #2: Use the n-value above which matches the closest and change the thickness of each type of layer (i.e. scale all low-index or all high-index layers simultaneously using the interactor) until a similar curve to that which was found experimentally in the lab is achieved. When the model matches the observed results as closely as possible, produce a final FilmStar model and include it in the report. Once finally optimized, compute a new tooling factor for that material/layer.

    The FilmStar reports must be included verbatim and without modification. Sharing of FilmStar models will be considered plagiarism and will result in a mark of zero for the entire lab.
    For this question, at least three FilmStar outputs are required as well as an explanation of how each was made (i.e. parameters which were changed for each model).

  8. Many datasheets mention "Stoichiometric composition" with respect to evaporation (both eBeam and sputtering techniques). One of the most interesting examples of stoichiometric change during evaporation is in the use of titanium oxides as described here. Describe (in a paragraph) HOW starting with Ti3O5 can lead to perfect deposits of TiO2 - describe the reaction that occurs and leads to the desired composition of the film.

Marking Scheme: Total Marks = 22, Q1=2, Q2=2, Q3=5, Q4=2, Q5=2, Q6=7, Q7=2

Copyright (C) Niagara College, Canada, 2018
This page is part of the PHTN1432 Course Page