hello kwon,

Did you also measure the pH of the TDE before adding HCl (It´s  a bit
difficult because if no water is present, H3O+ are not built affecting the
potential on the glas elektrode; but you can add a small amount of pure
water to measure it)? 
The pH of the commercially available TDE varies from batch to batch ranging
from around 5.5 to 11. The lower the purity the lower the pH. Maybe there is
an acidic contamination in the lower purity batches. But this should not be
an issue since the concentration of impurities are very low (not detectable
by GCMS) and the pH can be adjusted to around 7 if you consider one
important fact: The liquid is quite viscous which hinders the establishment
of the equilibrium. You should measure the pH for at least one hour to
equilibrate the electrode. If you start to adjust the pH with a batch
exhibiting a pH of around 6 i can imagine that you end up at a pH of around
2 or three adding HCl. For each individual batch you have to preadjust the pH. 

The crystallisation problem. 

Maybe NaCl is precipitating since you added 1µl 1N HCl and 1µl 1N NaOH to
the TDE (unfortunately i don´t know the volume) and PBS 1x is also added to
match the refractive index. I would just replace the buffer by pure water
(The performance should not change dramatically) and search for crystals
inside the sample. I doubt that the TDE itself causes the problem. I think
it is a salt from the used buffer system.