Thin Film Deposition is a vacuum technology for applying coatings of pure materials to the surface of various objects. The coatings, also called films, are usually in the thickness range of angstroms to microns and can be a single material, or can be multiple materials in a layered structure. This paper discusses the basic principles of thickness and rate control by use of quartz crystal monitoring.
One major class of deposition techniques is evaporation, which involves heating a solid material inside a high vacuum chamber, taking it to a temperature which produces some vapor pressure. Inside the vacuum, even a relatively low vapor pressure is sufficient to raise a vapor cloud inside the chamber. This evaporated material condenses on surfaces in the chamber as a coating or “film”. This method, including the general type of chamber designs commonly used for it, is an excellent candidate for successful control of rate and thickness through the use of quartz crystals.
The key concept behind this type of measurement and control is that an oscillator crystal can be suitably mounted inside the vacuum chamber to receive deposition in real time and be affected by it in a measurable way. Specifically the oscillation frequency will drop as the crystal’s mass is increased by the material being deposited on it. To complete the measurement system, an electronic instrument continuously reads the frequency and performs appropriate mathematical functions to convert that frequency data to thickness data, both instantaneous rate and cumulated thickness.
Such sensors and instruments are readily commercially available, including in an integrated package that not only reads and displays the rate and thickness data, but also provides outputs for other deposition system elements. It will have an analog drive signal to drive the source power supply in a closed loop technique based on the rate data and thus is able to maintain a preset rate during deposition. And it will have other outputs to interface with functions such as a source shutter triggered to close when the preset final thickness is achieved.
We will not go into the technical details of the instrument itself, or the algorithms that convert frequency to thickness. But it is useful to remember that the actual physical measurement is frequency and the rest is mathematical interpretation based on certain assumptions, one of which is that all frequency changes are due solely to film thickness on the crystal face, which is not necessarily true. Such things as transient temperature changes of the crystal can also cause frequency changes and be misinterpreted as film thickness. Such transients are generally of finite duration, though, and have only a small overall effect on the measured deposition in most cases. You typically have a graph display of rate vs time while depositing, and if you see a surprising transient it could be caused by this temperature issue rather than actual deposition, such as a blast of heat hitting the crystal when the source shutter first opens.
There can be numerous physical configurations of evaporation process chambers, but one of the most common will have the evaporant material source/heat source at the bottom of the chamber, with its vapor stream to rise above it and deposit from the underside onto a rotating substrate holder/tooling (often called a ‘dome’ for its typical shape). There is normally a movable physical shutter above the source between it and the substrates. When open, the substrates as well as the crystal monitor are subjected to the deposition. When the shutter is closed, at least the substrates must be shielded from the deposition, and usually the crystal is, too. The crystal, with its electrical cable, cooling water lines, and mechanical elements are securely mounted in a suitable location with the face of the crystal aimed at the source.
Quartz crystal rate controllers typically allow you to program in a desired heat profile (in terms of source power rather than actual temperature as such) to pre-condition the material prior to opening the source shutter and beginning actual deposition. This profile is experimentally determined by the user for each source/material, and in the controller is normally formatted to have two separate sequential ramps of power to some ‘soak’ level and be held there for some time.
The first ramp up is usually slower and stops at some level just below vaporization and holds it to achieve an acceptable degree of equilibrium before making the second ramp up to vaporization. The latter should ideally be the actual power level for the desired deposition rate, and in most cases, since you will now be consuming material and coating the bottom of the shutter, this second ramp and soak is made fairly brief. But those details are user choices. The controller lets you decide how you want it.
Once the final soak is complete, the controller will open the shutter and transition to closed loop power control to hold the programmed deposition rate and continue until final programmed film thickness is achieved. At that time the shutter closes and the controller executes a user programmed cool down profile similar to the heat up one. The closed loop control is usually a PID controller with user control of the parameters including the option of not using all three – depending on the characteristics of the power source, P alone might work, for example. As mentioned earlier, the instrument usually offers a choice of data displays, the most popular being rate vs time.
There can be various other options in such controllers, such as establishing closed loop control at some deposition rate in angstroms per second, and then, remaining under such control, ramping the rate up (or down) to a new rate and continuing from there, still ultimately going to the final programmed thickness with potentially several different controlled rate segments along the way.
A crystal has details to attend to, such as calibrating its thickness readings. It can, of course, only “see” the deposition that lands directly on itself, which is not what a user wants to know or to control. The user wants to know what is on the substrates, so physical measurements need to be made to compare to the crystal reading and a calibration factor (normally called “tooling” factor) needs to be calculated, programmed into the instrument, and then verified.
Another important point is that crystals do not last forever. A new crystal will have an initial operating frequency (most commonly 6.0 MHz), which will drop as material is deposited on the crystal. The oscillator has a lower limit of frequency at which it can perform, which is definitely one way for a crystal to fail. But it is not at all uncommon for a given crystal in actual use to never make it that far and to fail earlier from noise issues or other reasons.
To resolve angstroms of deposition, the oscillation has to be very good, and very stable and it does not take too much for the instrument to reject one as “failed”. So make sure to order a supply of these vital consumables, and remember that the substrates inside being processed are undoubtedly much more valuable than a crystal so don’t try to push for maximum life on a cheap crystal and risk scrapping a load of valuable product.
There are too many other aspects and options to cover them all in this essential basics article, but these quartz crystal rate monitors/controllers are used on virtually all evaporators, either filament/boat type or e-beam type. A major reason is the variability of deposition rate vs source drive power in most such systems – with these controllers you can better hold a rate and a thickness in spite of any such variation.
While it is possible to use this same crystal monitoring and control technology in a sputtering system, it is actually rarely done for several reasons, one of which is that deposition rate vs cathode drive power is usually adequately stable for process control in sputtering. Also, typical sputter chambers are less amenable to positioning crystals, and plasma can interfere with them.