In situ removal of oxide (and hydrocarbon) surface layers

[Probeman]

This should probably be posted in a topic entitled "crazy ideas", but wouldn't it be great if we had a way to remove hydrocarbon contamination and even better, remove surface oxidation "in situ" in our instruments?  Maybe even while making an x-ray measurement?

Normally people think of mechanical sputtering using an Argon plasma to remove such surface layers, but because the resulting Ar poisoning of ion pumps, I was wondering if anyone has tried to use a hydrogen plasma to remove hydrocarbon and/or oxide layers and so I did a quick search and found these articles:

https://avs.scitation.org/doi/abs/10.1116/1.571307

http://iopscience.iop.org/article/10.1143/JJAP.37.L536

The first from 1981 and the 2nd more recent.

It seems this might be doable but I have no idea of the technological considerations.  Has anyone any comments on whether this sort of "in situ" hydrogen plasma etching might be possible?   Does anyone have any idea of the etch rates if we're talking about a few nano-meters?

Ideally we'd only need to etch an area of a mm or even less, so if the plasma could be focused to a small spot maybe the etch rates could be reasonable?

[Probeman]

This should probably be posted in a topic entitled "crazy ideas", but wouldn't it be great if we had a way to remove hydrocarbon contamination and even better, remove surface oxidation "in situ" in our instruments?  Maybe even while making an x-ray measurement?

Normally people think of mechanical sputtering using an Argon plasma to remove such surface layers, but because the resulting Ar poisoning of ion pumps, I was wondering if anyone has tried to use a hydrogen plasma to remove hydrocarbon and/or oxide layers and so I did a quick search and found these articles:

https://avs.scitation.org/doi/abs/10.1116/1.571307

http://iopscience.iop.org/article/10.1143/JJAP.37.L536

The first from 1981 and the 2nd more recent.

It seems this might be doable but I have no idea of the technological considerations.  Has anyone any comments on whether this sort of "in situ" hydrogen plasma etching might be possible?   Does anyone have any idea of the etch rates if we're talking about a few nano-meters?

Ideally we'd only need to etch an area of a mm or even less, so if the plasma could be focused to a small spot maybe the etch rates could be reasonable?

I'm bumping this post because I'm still interested in finding someone with a hydrogen (or other) plasma cleaning hardware to remove the native oxidation layer from EPMA samples in situ, for example for analysis of trace bulk oxygen in alloys...

I found this company that apparently provides plasma cleaning technology for XPS and other instruments using the airlock:

http://www.piescientific.com/

But I worry about cleaning large areas of epoxy mounted samples.

[Probeman]

I recently contacted an engineer at Pie Scientific, as they make plasma etchers and other equipment. We spoke and he said I could post our conversation:

http://www.piescientific.com/

I wanted to ask what could be done to perform an "in situ" cleaning of a small (~1 mm) sample area using, say, a hydrogen plasma.

Hi Xinam,
We recently spoke about possibilities for in situ cleaning of EPMA samples.  I would prefer a very localized spot cleaning rather than a general sample cleaning in an airlock, but you indicated that "spot" cleaning might be difficult inside the instrument.

What about samples that are mounted in epoxy?  Isn't there redistribution of debris from the epoxy onto other areas of the sample?

Here is the link where I ask about this question:

http://probesoftware.com/smf/index.php?topic=1073.msg7175#msg7175

I will try to send you a photo of the airlock later.  Thank-you.

john

Hi John,
If the plasma source is installed on the main imaging/analysis chamber, the hydrogen or oxygen radicals will exit through a 3mm or 8mm aperture from the plasma source. The flow of the radicals is like a nozzle or jet flow. But it won't go straight like a laser beam. It may spread like a water/air jet nozzle. Even if the exit nozzle is very close to the sample, the hydrogen or oxygen radicals will still spread out all over the chamber. Atomic oxygen or hydrogen will not just stay at the sample location. It will still go everywhere in your chamber. Even if we can make a plasma source that can generate plasma close to the sample stage, the performance or effect won't be much different from our current EM-KLEEN design.  Let's say if the polymer membrane in your chamber indeed breaks down due to long plasma cleaning, how much does it cost to replace it. Can you tolerate this risk? If the membrane has an aluminum coating, it can protect the membrane from being etched by the atomic hydrogen or atomic oxygen.

I believe it will work just fine to clean the sample in the load lock chamber. Here I attached an application data from our customer in UCSD. Professor Kummel in UCSD bought two SEMI-KLEEN sapphire plasma source from us in 2016. He placed another order for the third SEMI-KLEEN plasma source just last week.

In the application data, the user only cleaned InGaAs sample with hydrogen plasma for 2 seconds in the load lock. Then the native oxide and hydrocarbon are totally removed according to XPS analysis. Of course, their samples are semiconductor samples, which are usually very clean.

I think EM-KLEEN will work just fine no matter you install it on the load-lock chamber or on the main chamber. We provide two-month full refund return warranty. If it doesn't work for your application, you can return it for a full refund within two months.

EM-KLEEN plasma source can also be used on XPS, FEI Quanta, and FEI Helois systems in your lab.

Please let me know if you need a quote.

Best regards,

Ximan Jiang

So he is suggesting an airlock cleaner similar to what JEOL has on some of their systems.  But I am concerned about plasma etching of epoxy mounted samples, which is why I was hoping for some sort of "focused" etching system for a small area directly under the beam. I then replied:

Hi Ximan,
Thanks for the info.  I understand what you are saying about the spreading of the H atoms from the nozzle.   Have you done any studies about making the stream more directional?

May I post your comments/pdfs below to the EPMA forum?

What do you think about a sample that is mounted in epoxy?  Will plasma cleaning just create more of a mess?

john

Hi John,
We didn't do any research to make it more directional. Yes, You can post my reply on the forum.

I don't have any XPS surface analysis data for samples mounted on XPS. Epoxy is not UHV compatible. Plasma cleaning won't make the sample on epoxy more of a mess. It will help. But I don't know how fast the sample will be recontaminated by epoxy.

Is electron probe sensitive enough to detect angstrom layer of native oxide like XPS? Auger electron spectroscopy (AES) might be more sensitive to the surface layer. X-ray emission volume in SEM/EPMA from high energy e-beam bombardment is usually below the surface layer.

EPMA should be more sensitive to chamber hydrocarbon vapor than the native oxide. Due to a high concentration of hydrocarbon vapor inside the main chamber, electron radiation will soon deposit a layer of carbon on the samples.

I think you should install the plasma source on the main chamber to remove the hydrocarbon vapors inside the chamber. I attached another application data here. The user used our plasma source to remove the hydrocarbon vapor inside their XPS chamber.  The cleaning results have been confirmed by RGA spectrum. But you have to figure out the risk of damaging the polymer membrane. 

Best regards,

Ximan Jiang

Just thought I would share this.  Does anyone have any observations on whether a plasma air lock cleaner would work on epoxy mounted samples?  Or just make a mess.  Any one know of a directional (focussed) plasma cleaner (hydrogen?) that could only clean the area under the beam?  Sort of like depth profiling in a SIMS...

See the pdf attachments Ximan sent me below.

[Beamsys]

I haven't heard about semi-focused plasma cleaning being commercially available for SEM or Microprobe, but this is really a straight-forward engineering project. I agree with previously posted discussion in that the most logical place to locate such cleaning is in LLK, while cleaning sample immediately after pump-down and prior to transfer to the main chamber. Market is probably too small for making blank-page design, but adapting ion gun from existing ion mill (Fischione, for example) would be straight-forward. For "proof of concept" just place ion gun above sample location in LLK, for more sophisticated approach make some crude X/Y in-LLK positioning. As cleaning is done in LLK Ar poisoning of ion pumps is, probably, of not so big concern. Also, there is no any reason why ion mill gun wouldn't fire with H2, providing that backpressure/gas flow/power parameters could be varied. I'd happy make such retrofit, providing there are adequate money behind the interest. Or try talking to Noel Smith of Oregon Scientific - I bet his team would be more then happy to custom-make H2 plasma column for your load lock....

[Jacob]

Normally people think of mechanical sputtering using an Argon plasma to remove such surface layers, but because the resulting Ar poisoning of ion pumps, I was wondering if anyone has tried to use a hydrogen plasma to remove hydrocarbon and/or oxide layers and so I did a quick search and found these articles:

The ion pump poisoning isn't as big a deal as you'd think.  Other ion pumped systems use argon sputter cleaning to great effect.  Scanning Auger microprobes will use an argon for cleaning and depth profiling, and those chambers are pumped entirely by ion pumps, severed from the foreline, when in operation.  A big enough ion pump of the right type and you're golden.  Even then, you could close the gun valve while digging your hole, and avoid the problem entirely.

You can see the same type of cleaning systems in XPS/ESCA systems.  So I say buy an Argon gun and let 'er rip.

[Probeman]

Normally people think of mechanical sputtering using an Argon plasma to remove such surface layers, but because the resulting Ar poisoning of ion pumps, I was wondering if anyone has tried to use a hydrogen plasma to remove hydrocarbon and/or oxide layers and so I did a quick search and found these articles:

The ion pump poisoning isn't as big a deal as you'd think.  Other ion pumped systems use argon sputter cleaning to great effect.  Scanning Auger microprobes will use an argon for cleaning and depth profiling, and those chambers are pumped entirely by ion pumps, severed from the foreline, when in operation.  A big enough ion pump of the right type and you're golden.  Even then, you could close the gun valve while digging your hole, and avoid the problem entirely.

You can see the same type of cleaning systems in XPS/ESCA systems.  So I say buy an Argon gun and let 'er rip.

Hi Jacob,
Thanks. 

I knew that other techniques utilize Ar sputtering with ion pumps, but I didn't know how dependable/useful such systems were in practice.

Now we might be inspired to implement a prototype Ar sputtering system in our SX100, that is if our engineer ever finds the time (we keep him pretty busy)! 
john