[labnetwork] SnO2 via ALD

[Owain Clark]

Possibly also worth noting that we have started to see some ALD precursor suppliers recommend that sources are refrigerated when not connected to a tool. I do not know if this is based on observed data or common sense best practise. We have noticed with a couple of precursors that processes tend to shift over a period of years using the same source, usually manifesting as uniformity or deposition rate shifts. Tool variation could be a factor, but we are talking relatively simple chambers such as the Veeco S200 with not a lot to go wrong. Source temperature is perhaps the largest point of variation as jackets are manually applied to each source per change, although one suspects if users wait long enough for the source and jacket to equalise it should be not be that significant. User impatience for source heating could well be a factor.

I have more questions that answers at this point regarding long term ALD source stability, we will experiment with cold storage when removed from tool and keep tracking process data. I suspect if it is a factor that it will be compound dependent. If it does turn out to be a factor then questions about an ALD source's shelf life arise, given that decomposition must be accelerated when they are heated for processing. At this point does the clock start ticking? We have always lowered source oven/heating jacket temperature when not processing which in hindsight may have been a good idea.

BR, Owain

From: labnetwork <labnetwork-bounces at mtl.mit.edu> On Behalf Of Hathaway, Malcolm R
Sent: 06 February 2026 20:57
To: Martin, Michael <michael.martin at louisville.edu>; Network Fab <labnetwork at mtl.mit.edu>
Cc: Naveed, Abdul Basit <abdulbasit.naveed at louisville.edu>
Subject: Re: [labnetwork] SnO2 via ALD

CAUTION: This e-mail originated outside the University of Southampton.
Hi Michael,

So, tin oxide is interesting, in that the TDMASn is, in my estimation, very sensitive and not particularly shelf- or bottle-stable.  How old is your precursor?

I have had this tin precursor precipitate all the tin out while the ampule was still sealed (very light sensitive, as it turns out.  They put a "light sensitive" sticker on the bottle for me the next time...)

It also dies pretty quickly while on the machine, if the cylinder temp is too hot.  I don't have my notes in front of me, but it seems to prefer the lower end of the TDMAx range (50-60C, perhaps).

Do you see pressure pulses when you pulse the precursor?  If not, the precursor is dead (or the valve is blocked/stuck).  Even if you do, the precursor could be decomposed in the bottle.  There will still be liquid present, perhaps sending vapor to the chamber, but the Sn stays back in the bottle.

We have crafted glass precursor cylinders (for non-pyrophorics) specifically to be able to see the state of the precursor in the cylinder, which is pretty helpful in situations like this.

If you have a glove-box, you could do some things to see what's the status of the precursor.  We can walk through those off-line if you have an interest.

Heating of lines:  As I'm sure you know, you generally want the line heated to at least the temperature in the bottle, and sometime a bit higher, to avoid any chance of precursor condensation in the line before it gets to the chamber.  This can happen in a very short distance (a "cold spot") of unheated line.   If this is happening, it will manifest as a lack of pressure pulses in the chamber when the TDMASn valve is pulsed.  It may be that the parts of the line that are heated are close enough to the "unheated" segments that the unheated segments stay warm enough.  If the system is still "stock" I would imagine the line heating should be okay, but worth checking with a temperature gauge.

On the other hand, if some part of the plumbing is too hot, it could lead to thermal decomposition of the precursor before it gets to the chamber.   Our system with valve manifold heated to 150C did not show any issue like this.

It may be worth giving Beneq a call to see if they have specific suggestions.  I don't have one of their tools but they seem to know their way around ALD, so they probably have some suggestions.


Mac
Harvard CNS
617-259-0859



________________________________
From: Martin, Michael <michael.martin at louisville.edu<mailto:michael.martin at louisville.edu>>
Sent: Wednesday, February 4, 2026 4:01 PM
To: Network Fab <labnetwork at mtl.mit.edu<mailto:labnetwork at mtl.mit.edu>>
Cc: Hathaway, Malcolm R <hathaway at cns.fas.harvard.edu<mailto:hathaway at cns.fas.harvard.edu>>; Naveed, Abdul Basit <abdulbasit.naveed at louisville.edu<mailto:abdulbasit.naveed at louisville.edu>>
Subject: SnO2 via ALD

Hello everyone,
   We are having a heck of a time depositing SnO2 in our ALD onto bare (with native oxide) silicon at a substrate temperature of 100 to 200C.  This is our first time attempting this material in our Beneq TFS200.  We are using TDMASn + H2O as precursors. We have tried heating the TDMASn anywhere between 50 to 100C, using headspace loading and via the material's vapor pressure.  We have looked at TDMASn pulse times from 200ms to 1.4 seconds with dwell times of 1 to 3 seconds. Water vapor is pulsed at our usual 100ms and a purge of 3 seconds.
  So here's a couple of questions: 1. I have read anecdotally that the material will decompose above 60C. Do you have any experience with this? 2. While the line between our heated bubbler and the reactor is quite short, it is not heated. Do your tools depositing tin oxide require a heated line to the reactor?
   Any advice would be fantastic.

Regards,

   Michael



Engineering Senior Researcher

Micro/Nano Technology Center


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20260209/072bff8c/attachment.html>