Discoloration – LiCl vs. CsCl vs. NaCl
A couple of weeks ago, I posted a discoloration array experiment that showed monotonic discoloration patterns here. This was a result that we’d been expecting for several months, but we never got it with NaCl and NaHCO3. Simplifying the solutions to NaCl and NaOH was the key. These results spurred some discussion on the mechanism of discoloration (salt ions, Debye length, ion concentrations) and Bill Burnhard suggested some experiments with LiCl and CsCl. It was hypothesized that lithium (atomic # 3) and cesium (atomic # 55) ions might affect the Debye length and entropy of the system since they are different sized cations.
I made arrays of these salts and NaOH. These are lower NaOH concentrations than the last experiment because I wanted to slow down the discoloration. I also included an array with NaCl as a control. These were w680 pieces that I cut off the wafer, without RTP and performed at room temperature.
Here is the time=0 pH data:
The pH jumps up with the addition of NaOH, and higher salt concentrations at a given NaOH concentration decrease the pH. The pH of the different salt solutions is approximately the same.
Here is the discoloration info:
The discoloration was monotonic again, with the highest salt and OH- concentrations discoloring the fastest. It seems like the discoloration rates for all 3 salts are approximately the same. You could possibly argue that cesium caused slightly faster discoloration, but it’s subtle (and could be due to wafer variability).
Unfortunately, I’m not sure how to interpret these results. Different size cations in the Stern layer of the pnc-Si surface evidently don’t affect the access of OH- ions to the surface. I plan on repeating this discoloration array but with a divalent salt.
By my eye, your results indicate that Li protects the Si surface better than either Na or Cs (the latter two being about the same). At 45 m and 2 h, the Li is outperforming the Na/Cs.
This seems to fit with Paul’s newest results. At 10 mM, he observed that permeability of DNA was higher for Li than Cs or Na. This suggests that the electric potential drops off faster for Li, with the consequence of making pores look bigger to DNA. Or coming back to Barret’s earlier thoughts on Debye length, Li appears to reduce the Debye length relative to Na or Cs.
I agree with Bill at those time points and even for the longer time points, when the LiCl concentrations are low. Curiously, at 72 hours, LiCl alone appeares to attack the Si when it is at high concentration. If LiCl actually attacks the Si, it would make these comparisons difficult. Do you think the 72 hour data is correct, or some type of anomally?
Chris, I think the LiCl data is correct, especially since I’ve seen chips discolor in salts alone after a couple days at elevated temperatures.
I see the protection offered by LiCl, but I hesitate to believe it entirely because of membrane variability (and because this is only n=1). If it gels with other explanations like DNA separations, it’s worth repeating.
It’s interesting that LiCl both protects and attacks. That said, this could be pandora’s box, and given the amount of time it must have taken to collect this data, it seems like quite a labor intensive project.
Is there some aspect of this that should be pursued by someone? I guess we should think about how we would use the potential results, and the likelihood of getting sufficient evidence that would stand up in a publication.