WOW. I cleaned my kitchen cabinets today. Well...partially. Actually I got bored half way through, made popcorn and watched youtube. Anyway, at this level of sophistication (let's call it "upper Ben Krasnow division") I'm expecting full DIY ellipsometer for the next video.
Nice video Sam. If you want to measure film thickness variations in high (x,y) resolution you might want to consider doing a variant of "white light interferometry". This method is also suitable for measuring the absolute topology of your etched patterns.
Nice setup. You can get more accurate peak position by normalizing the spectrum by dividing the sample spectrum by the bare silicon spectrum. To get an accurate absolute reflectance spectrum, you can then multiply by the expected reflectance of bare Si (calculated from the complex refractive index). This let's you measure thin films that do not even have a peak if you find the best fit calculated spectrum that matches the measured. I used to design thin film measuring equipment if you have questions.
thanks for the notes, I am already doing normalization on the spectra (helped see peaks on the lower end of the spectra tremendously) but if I talk about these things in the video it will drag on even longer... its hard to keep things short as is.
Is everyone else jealous of this guy's set up. Keep up what your doing no one else really has as informative videos as you. Making a chip is a project I only wish I had enough time and brain power to make. May I ask what you went to school for?
Nice! I just wanted to add that this technique will only work as long as the index of refraction does not change very much in the wavelength range that you are measuring. So any optically transparent materials such as photoresist or SiO2 should be fine to measure, but if you deposit a semiconductor thin film then this method won't give you the correct results. Also, if you collimate the white light coming into the objective you should be able to get a small light spot on your sample and perform thickness scanning measurements just by moving the stage.
How does it feel to KNOW so much stuff? Omg?! It's like turning into a wizard :O Hats off so much! I wish I could know at least a tiny bit of all of this you've learned!
5:20 "destructive interference with the blue one, constructive with the red one" i think in this case for blue wave the optical thickness is about 2 * lambda, for red one it's about ~lambda, so both of them seem to interfere constructively.
can you see light dispersion this way? if some wavelengths have different refractive index than others, peaks on the spectrogram can get unevenly spaced.
i am trying to measure the oxide layer thickness of anodized titanium, can you help? I have reflectance spetra but Im not getting the right values with the formula
thin film interference does not even necessarily need such a complicated setup, you can just eyeball the color of the film (with non-infinite error margin) in white light and compare it to pre-rendered palette. it works for estimating oxide level thickness on heat treated steel for example.
the colors our eye perceives are periodic with increasing thickness, i.e. color alone cannot tell you absolute thickness because the same color maps to many possible thickness values. good for a spot check though, yes
@@SamZeloof it would be the case if the light was monochromatic. but since there are multiple wavelengths interfering, every film thickness results in its own distinct color in white light. besides, do can you use the same method to estimate color of a metallic thin film? metals do have their own weird refractive index, but I wonder if it actually works the same way as it does for dielectrics.
did you figure this out? I am trying to measure oxide layer thickness of anodized titanium and this formula that sam used in the video is out of the range of nm @@Alexander_Sannikov
@@pwdrr4628 I recommend you to google titanium oxide layer color chart. as Sam mentioned, it's somewhat periodic, so you might not be always able to easily tell which exact period your color falls into, but if you know what thickness to expect, it's much more reliable.
Man, enjoyed it! :-) As always! I see you're aiming for higher density here. :-) I hope one day you'll make device as popular as 3d printers but for making custom ICs. Can help you achieve that with any kind of soft development if you'll need :-)
Those are pretty much "the standard" probe station. They are quite common here in Silicon Valley. Or were, as they're not as useful for the single-digit-nanometer chips from the world-class fabs today.
Hi Sam, you seem to have a lot of experience in coatings. I was wondering if you have any thoughts on the best way to evenly distribute a phosphor coating? I use ZnS powder to detect an electron beam in my vacuum chamber, but I can never get the powder to stick very well
You should check out the video by applied sciences about self mixing of laser diodes. Slightly different technique but then you can get a before and after measurement to see how thick the coating is. Anyway love the videos and hope you start posting more.
Might be nice to plot the thickness in a time-graph. Then if the controls are sensitive, you can measure different depths by moving the position of the chip.
can you suggest a method to calculate the thickness(approx) using some equation and knowing the properties of the materia ...can you please tell the accuracy of such equation
There is no virus where I am, nor is there an awsome lab to tinker in! Not sure which I'd prefer if I had a choice. With the setup and skills you have developed so far could you fab a ROM and if so how many bits would it store?
Thanks! Great as always. Why the roach killer with the other chems on the bench? Also, might you be willing to share your Python code for the STS spectrometer? Also, Ossila has a very readable guide to spin coating, for the rest of us. Any special fitting on the STS to the trinoc port? Thanks!!
![เขามัทรี - เอ็กซ์ ศุภกฤต - จอนนี่มิวสิค [ Official MV ]](http://i.ytimg.com/vi/AS9ufKuQIVs/mqdefault.jpg)
Really nice setup! Your projects are always interesting.
Man, I'd love to see you and Sam do a collaboration on something.
do a collab! also Sam looks a little bit like mini-Ben. both are amazing, however.
"So, I'm in the home chip lab..." as if everyone has a home chip lab :) Love it!
ANOTHER HIT, THIS MAN JUST CANNOT MISS
HUH
WOW. I cleaned my kitchen cabinets today. Well...partially. Actually I got bored half way through, made popcorn and watched youtube.
Anyway, at this level of sophistication (let's call it "upper Ben Krasnow division") I'm expecting full DIY ellipsometer for the next video.
Nice video Sam. If you want to measure film thickness variations in high (x,y) resolution you might want to consider doing a variant of "white light interferometry". This method is also suitable for measuring the absolute topology of your etched patterns.
Nice setup. You can get more accurate peak position by normalizing the spectrum by dividing the sample spectrum by the bare silicon spectrum. To get an accurate absolute reflectance spectrum, you can then multiply by the expected reflectance of bare Si (calculated from the complex refractive index). This let's you measure thin films that do not even have a peak if you find the best fit calculated spectrum that matches the measured. I used to design thin film measuring equipment if you have questions.
thanks for the notes, I am already doing normalization on the spectra (helped see peaks on the lower end of the spectra tremendously) but if I talk about these things in the video it will drag on even longer... its hard to keep things short as is.
hey,
I am spin coating ZnO on glass. Could you help me on how to measure its thickness(could be around 50-200nm)?
Is everyone else jealous of this guy's set up. Keep up what your doing no one else really has as informative videos as you. Making a chip is a project I only wish I had enough time and brain power to make. May I ask what you went to school for?
4:04 well if that isn't a faithful representation of a Python, i don't know what is.
Nice! I just wanted to add that this technique will only work as long as the index of refraction does not change very much in the wavelength range that you are measuring. So any optically transparent materials such as photoresist or SiO2 should be fine to measure, but if you deposit a semiconductor thin film then this method won't give you the correct results.
Also, if you collimate the white light coming into the objective you should be able to get a small light spot on your sample and perform thickness scanning measurements just by moving the stage.
Agreed on all points - thanks for your note !
A good point! Thanks Brenesrob! Do you happen to know if there is any technique to measure semiconductor thin film? Thanks
How does it feel to KNOW so much stuff? Omg?! It's like turning into a wizard :O Hats off so much! I wish I could know at least a tiny bit of all of this you've learned!
You’ve gotta be a naturally born genius there’s no other way to describe this.
Making the most of school closure! Great stuff, man!
So rad!
5:20 "destructive interference with the blue one, constructive with the red one"
i think in this case for blue wave the optical thickness is about 2 * lambda, for red one it's about ~lambda, so both of them seem to interfere constructively.
can you see light dispersion this way? if some wavelengths have different refractive index than others, peaks on the spectrogram can get unevenly spaced.
Nice video- could I check what software you used for the STS spectrometer module?
Thanks
New favorite channel.
i am trying to measure the oxide layer thickness of anodized titanium, can you help? I have reflectance spetra but Im not getting the right values with the formula
Sorry, but where do i get $Delta_m$ from? Is it always 1 for maxima und 1/2 for minima, so the first point or can it be second or third as well?
Great work and awesome home lab.
Stay safe everybody.
Stay safe sir.
thin film interference does not even necessarily need such a complicated setup, you can just eyeball the color of the film (with non-infinite error margin) in white light and compare it to pre-rendered palette. it works for estimating oxide level thickness on heat treated steel for example.
the colors our eye perceives are periodic with increasing thickness, i.e. color alone cannot tell you absolute thickness because the same color maps to many possible thickness values. good for a spot check though, yes
@@SamZeloof it would be the case if the light was monochromatic. but since there are multiple wavelengths interfering, every film thickness results in its own distinct color in white light.
besides, do can you use the same method to estimate color of a metallic thin film? metals do have their own weird refractive index, but I wonder if it actually works the same way as it does for dielectrics.
did you figure this out? I am trying to measure oxide layer thickness of anodized titanium and this formula that sam used in the video is out of the range of nm @@Alexander_Sannikov
@@pwdrr4628 I recommend you to google titanium oxide layer color chart. as Sam mentioned, it's somewhat periodic, so you might not be always able to easily tell which exact period your color falls into, but if you know what thickness to expect, it's much more reliable.
Man, enjoyed it! :-) As always! I see you're aiming for higher density here. :-) I hope one day you'll make device as popular as 3d printers but for making custom ICs. Can help you achieve that with any kind of soft development if you'll need :-)
That is a sweet probe station for a home setup!
Those are pretty much "the standard" probe station. They are quite common here in Silicon Valley. Or were, as they're not as useful for the single-digit-nanometer chips from the world-class fabs today.
Can this method work for multi coating measurements ?
Hi Sam, you seem to have a lot of experience in coatings. I was wondering if you have any thoughts on the best way to evenly distribute a phosphor coating? I use ZnS powder to detect an electron beam in my vacuum chamber, but I can never get the powder to stick very well
You should check out the video by applied sciences about self mixing of laser diodes. Slightly different technique but then you can get a before and after measurement to see how thick the coating is. Anyway love the videos and hope you start posting more.
Did you check the accuracy of your measurement? I think your experiment needs validation (cross-checking).
Might be nice to plot the thickness in a time-graph. Then if the controls are sensitive, you can measure different depths by moving the position of the chip.
can you suggest a method to calculate the thickness(approx) using some equation and knowing the properties of the materia
...can you please tell the accuracy of such equation
Love your lab setup! Also great video :-)
Nice!
If we merge the capabilities of you, cody's lab, nilered, applied science, etc the stuff that could be done...
How do You know that light source contains all components of the spectrum?
its white
I wonder if it would be possible to produce finFETs with similar methods to this chip lab.
Wonder if you can measure metal film this way. Awesome video
Awesome video, Sam.
Amazing your all video.
What your job . MCU company
Sir, what are the factors affecting the thin film and how to protect the thin film?
There is no virus where I am, nor is there an awsome lab to tinker in! Not sure which I'd prefer if I had a choice. With the setup and skills you have developed so far could you fab a ROM and if so how many bits would it store?
Thanks! Great as always. Why the roach killer with the other chems on the bench? Also, might you be willing to share your Python code for the STS spectrometer? Also, Ossila has a very readable guide
to spin coating, for the rest of us. Any special fitting on the STS to the trinoc port? Thanks!!
The roach killer contains boric acid which is used as a dopant
Keep up the great work😁👍. Out of curiosity what's your major? assuming you're in college lol
Can you spare a small piece of crystalline silicon wafer for me.
Inspired!
Put steppers on the handwheels to map it!
How you get this tools🔥🔥🔥🔥🔥🔥 how much
3D printed maps of the surface would be rad
Which I had the money for these things, Ebay yea but these things are extremely expensive anyway. Cool stuff though.
I though it's about TFTM based on crystal
In a few years: So i build a Synchrotron...
So i built a cpu that can play crysis
Hi Sam
Hmmm if i would have that much money on hardware...
It`s chip porn!
eh bro u want a job?