Hello dear, if you have only PLA in the reactor, it is not possible, because the PLA suffer degradation when it is exposed to high temperature. Therefore the MW will be reduced. I think that possibly your reactor had some compound that produce this increasing on MW.

It is possible but less likely. How was you batch mixer cleaned? I would attempt to repeat it to see if it is something that is reproducible in the first place. But certainly interesting. 

Hello dear, if you have only PLA in the reactor, it is not possible, because the PLA suffer degradation when it is exposed to high temperature. Therefore the MW will be reduced. I think that possibly your reactor had some compound that produce this increasing on MW.

One of these compounds can be the chain extenders,

and also the reaction catalyst can provide linkage the polymer chains leading to higher MW PLA

Dear Ihsan,

Is there any way to distinguish between chain extenders and remaining catalyst in increasing the Mw? Do they manifest themselves in Mw distribution after processing?

I attached the GPC results

If your PLA have alcohols as end-functional groups, then the compounds that raise molec.wt. may be for example diacids or anhydrides.  Alternatively, you might have polyols with functionality greater than 2 (for example pentaerithrytol), which might be thought of as "crosslinkers" and thus raise molec.wt.  

@C Esnaashari:  By the way, I cannot see your GPC results.  Can you post again.

Dear Kaoru,

I am working with a pla with unknown trade name. But its properties (crystallinity, thermal stability, melt strength, ...) are excellent. Torque vs time graph shows a small and very broad peak which confirms GPC results. I want to find more about  the chain extender in this PLA and its amount.

attached is the distribution curve (a photo taken from the printed results)

Thank you.

Interestingly, the Garlotta reference (attached as link below) observed what you are seeing as well.  I'm not completely sure why this happens.

One thought that came to mind is, what solvent did you dissolve the original PLA pellet in?  For semicrystalline PLA, the only common solvents that adequately solubilize PLA crystals were chloroform and dichloromethane, when I was working with the polymer.  If it is amorphous PLA, then you could use a lot of different kinds of solvent without a problem.

The reason I ask is, if you used for example THF (a common GPC solvent), then semicrystalline PLA would not completely solubilize, in which case there is the possibility that the unsolubilized PLA crystals were of higher molec.wt. than what is seen in the graph.  Melt-processing and quenching down to room temperature would turn the semicrystalline PLA into amorphous PLA and make it more soluble, so that it becomes conceivable that this difference alone made the molec.wt. to appear to increase.

Maybe this isn't the problem, but I thought I'd ask all the same.

http://naldc.nal.usda.gov/download/4048/PDF

Dear Kaoru,

Thank you so much for your answer. Yes I've already seen this paper. They ascribed it to an increase in hydrodynamic volume and consequently the intrinsic viscosity, caused by chain entanglements during processing. But they measured  Mw by dilute solution viscometry and the increase in Mw determined by SEC was statistically insignificant. 

I used THF as solvent and pla was quenched to RT after processing. The unprocessed pla granules were transparent indicating being amorphous. By the way, I had no difficulty solving the pla in THF (of course I had to heat it for a few hours). 

But how do you explain the increase in torque? I processed two other grades of PLA (a low viscosity grade and a high viscosity one from natureworks) and neither one showed such a torque increase at that processing condition.

I attached the torque vs time graphs for comparison.

Without all the information, I would be speculating, but here are my thoughts:

* I'm not sure if MW really is increasing; and if it is increasing, I'm not sure what's causing it.

* Was dry THF used, and was the PLA dry as well, when you dissolved in THF?  Since you mention the THF / PLA mixture was heated, I was concerned that residual water can cleave enough chains to cause initial MW to go down.

* On the torque - I can only guess, I'm assuming that your PLA resins are not crystallized to the maximum extent, before putting in the melt processor (which is not unusual)?  If so, and the PLA resins could crystallize some more, then what you are observing from the torque measurement may be coming from the PLA further crystallizing as it heats up in the instrument.  Temperature takes some time to come up to target, before the polymer is melted.  Also, the shear that I would expect to be present in such a process should further enhance strain induced nucleation and crystallization, as has been observed often in PLA films and fibers (and of course, also in PET films and fibers which have been probably much more studied in the past due to its very high commercial volume).

--So I would suspect that crystallization of your PLA resin is giving the initial rise in torque.  If so, then the drop in torque would be due to the PLA crystals subsequently melting as the resin temperature reaches your target temperature (the "melt" temperature).

Again, it's all speculation since I don't know the details of your experiments, but maybe this would help in trying to understand the phenomenon.

Dear Kaoru,

Here is some more information:

I heated the PLA/dry THF mixture only up to 50 C, and dried neither one of the PLAs.

The processing temperature was 230 C (far above the melting temperature of PLA). I am sure torque increase is not due to PLA further crystallization.

Mn  and Mw increased 49% and 126% respectively.

If your column can run GPC on chloroform, that might be most straightforward - as you can dissolve even semicrystalline PLA in chloroform at room temperature.  That said, since Garlotta is reporting an increase in molec.wt. (and M_v by viscometry is a fairly acceptable way to monitor molec.wt.), there's something in the melt processing that causes it.  I would suggest examining all the additives, cleaning agents (maybe another polymer is used to clean out the melt extruder, in Garlotta's case, for example), anything that contacts PLA, and analyzing PLA before and after each step very carefully.  It sounds like a problem with many parts, and I'm not sure it can be answered in this forum alone (although I'm open to surprises!)

Sorry if this was not sufficient help.  Best of luck!