How Does a Quartz Crystal Microbalance Measure Film Thickness?

[sunshine111]

I want to know the basic principle of working for quartz crystal microbalance, how it is used to measure the thickness of the film?And why it can work at room temp (with water cooling) and at higher temp say 65deg ??

 

[Andy Resnick]

You mean something like this?

http://www.iop.org/EJ/article/0950-7671/43/4/417/siv43i4p263.pdf?request-id=f268710b-9d7b-4f2d-b449-30522933915d

 

[sunshine111]

Thanks for the reply, but I want to know about the Quartz crystal microbalance, which is used in Ultra high vacuum for monitoring the thickness of the film. I don't understand how the QCM measures the thickness of the film. I want to know its function in a simple way.
but thanks for the reply.

 

[Andy Resnick]

The vacuum microbalance operates on a very simple principle- one side is heavier than the other, so the balance tilts. The tilt is measured, and related back to the weight of the object. I suspect for the specific application you mention, the trick is how to mount the sample to allow flim growth while operating the balance.

Do you have some sort of reference that mentions this technique?

 

[sunshine111]

Quartz crystal microbalance is the vacuum chamber, where some material from some source is evaporated. The QCM vibrates with some frequency(I guess) and then the thickness of the film deposited on the Quartz crystal is measured with the instrument. What I don't understand is how the change in the Quartz crystal is freq is converted to thickness.
Thanks.

 

[vk6kro]

Piezo electric materials generate a voltage across them when they are deformed by a force, or they bend when a voltage is applied to them. Crystal microphones used to use this first principle, but they used Rochelle salt which was, unfortunately, hygroscopic.

So, you could measure this voltage and, having previously calibrated the balance, you would know what weight was being applied.

The other thing that quartz crystals can do is oscillate when placed in a suitable holder in an oscillator circuit.
Within limits, you can apply substances to the crystal and its resonant frequency will drop to a lower frequency.
I have done this with graphite from a pencil.

The balance would be calibrated with known weights and by measuring the oscillation frequency for each weight. This would be converted to a conversion formula and the weight would be displayed instead of the frequency.

Putting too much material on the crystal eventually loads it too much to oscillate.

 

[sunshine111]

can you kindly explain me a little bit more about it in detail?
Also, why is that, it operates at water cooled temp. and at 65 deg?
how does this change anything related to thickness measurement?

 

[vk6kro]

You mean... how does it work at low temperatures and all the way up to 65 deg C ?

I have no experience with these balances but a lot of experience with Quartz crystals.

They do vary in frequency with temperature but this can be minimised by the angle the crystal is cut from the quartz.
In any case, in percentage terms, the shift with temperature is quite small.

Also, in this application they would be measuring a difference in frequency, rather than an absolute frequency, so this could minimize errors.

 

[sunshine111]

No, I mean it is operated at two temperatures for calibration of film thickness, one is water cooled temp. and the other is 65 deg?
What is the effect of frequency vs time for quartz? is it stable for these two temp or what?
and as far as i understood, you mean to say...that with change in frequency of quartz, the thickness is measured. How does it know that NOW it is that much thick? How does it sense it that NOW it is (say) 1 Monolayer or so?

 

[MATLABdude]

Unfortunately, most of the responses in this thread have addressed everything EXCEPT what the OP was asking for. He means one of the thin film thickness monitoring instruments commonly used for in-situ semiconductor deposition monitoring.

The Wikipedia has a little something:
http://en.wikipedia.org/wiki/Quartz_crystal_microbalance

Most (including the one that some of my colleagues in the lab use) work on resonant frequency shifting as more of whatever you're depositing accrues on the sensor (thus making it heavier, and thus lowering the sensor's resonant frequency).

Not always accurate as you can't always have it in exactly the same conditions as whatever substrate you're depositing the film on (sometimes, you can use a correction or tooling factor), and sometimes you need to use more complicated models (or another sensor) as your film grows thicker and heavier.

 

[vk6kro]

He is also asking about the temperature coefficient of quartz crystals.

He hasn't said what the machine is doing so that leaves room to guess.

You could measure the frequency at two temperatures at the extreme range of the device's range and then assume a linear frequency shift in between those readings.
Then you know the unloaded frequency and can work out the shift in frequency from the measured frequency when a load is put on.
It would be good if he told us this stuff, but even making a bad guess brings a reply and explains a little more of the situation.

Matlab, do you feel the coating is put on the crystal that is oscillating or is it put on something else which then adds weight to the crystal and affects its oscillation? Seems like the second one would be less destructive.

 

[MATLABdude]

He is also asking about the temperature coefficient of quartz crystals.

He hasn't said what the machine is doing so that leaves room to guess.

You could measure the frequency at two temperatures at the extreme range of the device's range and then assume a linear frequency shift in between those readings.
Then you know the unloaded frequency and can work out the shift in frequency from the measured frequency when a load is put on.
It would be good if he told us this stuff, but even making a bad guess brings a reply and explains a little more of the situation.

Matlab, do you feel the coating is put on the crystal that is oscillating or is it put on something else which then adds weight to the crystal and affects its oscillation? Seems like the second one would be less destructive.

The OP does mention that it's in a vacuum, and used to measure film thickness, so I would assume it's something along those lines.

Regarding the second point about temperature coefficient: the unit which we have (Infinicon's TM-400) is supplied with crystals designed to have 0 temperature coefficient between 40 and 80 C:
http://www.inficon.com/download/en/...MemDirF/Maxtek.html&__utmv=-&__utmk=179076838

The measurement head is also water-cooled, allowing you to keep the sensor relatively cool under moderate deposition conditions. I suppose the best way to compensate for temperature would be to incorporate a temperature sensor on the measurement head, and just hash a table of values.

 

[leemeeyun]

may i know where i can get quartz crystal microbalance (QCM)?
quartz crystal microbalance (QCM)
fundamental resonance frequency, f0=10MHz
Crystal diameter = 13.7mm
Electrode diameter = 5.1mm

how much is it (cost) ? for each piece?
i need it for my final year project.
thank for your help.

 

[vk6kro]

Here is something similar, but it looks expensive:

http://webferret.search.com/click?w...thinksrs.com/products/QCM200.htm,,askjeeves,1

They don't give a price, so you would need to ask.

 

[leemeeyun]

is this one i need. Do you guys know where i can get it ? Any suggestion ?
http://en.wikipedia.org/wiki/File:Krystalové-výbrusy.jpg

 

[MATLABdude]

Welcome to PhysicsForums!

That looks like a custom job, and I'd doubt that it meets the specifications that you laid out in your first post.

For monitoring their thin film deposition, my labmates get their crystals from Infinicon, while the lab down the hall gets theirs from Lesker:
http://www.inficonthinfilmdeposition.com/en/quartzcrystals.html
http://www.lesker.com/newweb/Process_Instruments/processequipment_Crystals_6MHz.cfm

If I recall correctly, the crystals themselves are comparatively inexpensive (under $100 CAD$), but the instrument to utilize it cost in the low to mid few thousands. In any case, the crystals sold by Infinicon and Lesker are 5 and 6 MHz crystals instead of 10.

Nevertheless, some quick Googling (for '10 mhz quartz microbalance') brings up the following manufacturer:
http://www.icmfg.com/quartzmicrobalance.html