Tuesday, November 11, 2008



Forrest has just reminded me that I should have blogged this ages ago.

Bath undergraduate Nick Grudgings did some experiments on making polylactic acid. Here's his setup:

The flask contains lactic acid, and is resting in a heater. The tube is feeding in dry nitrogen to eliminate moisture. There's a magnetic stirrer flea under the lactic acid that you can't see. We also put in a thermometer. The whole thing was in a fume cupboard, as you don't want to breathe lactic acid vapour...

We warmed it up gradually, with the stirrer turned on. When the lactic acid started to melt we added 1/600 by weight of tin octoate as a polymerisation catalyst. The melt became quite viscous at around 140oC as polymerization started. You can control the mechanical properties of the result by the amount of catalyst: the polymer chains form from each catalyst molecule, so more catalyst means shorter chains.

Here's the result:

At the bottom of the flask is a cool solid clear lump of polylactic acid, with - it turned out - a little residual lactic acid embedded in it.

After a while (weeks) that lactic acid started to absorb atmospheric water vapour and thereby to disrupt the polymer, from which we learned that we needed to cook it for longer (we used about half an hour) to complete polymerisation. Obviously the longer the polymer chains you go for, and thus the less catalyst you use, the longer the polymerisation will take.

Labels: ,

Excellent! Great first step! :-D

Now to get from starch to lactic acid.
So what happens if you break up some PLA, mix it with fresh lactic acid, and cook that? If that does produce more but longer-chain PLA, can we then partially decompose that back into shorter-chain PLA, and maintain vaguely stable material properties? It would be very cool to avoid needing a material supply other than lactic acid, even if it is in trace amounts.
jonored: Only the chains with catalyst will grow, but any decomposition process will break up all chains equally. Each chain cleavage leaves catalyst at the end of, at most, one daughter chain.

As material is added and taken away, the proportion of catalyst-terminated chains will gradually decline. I don't think "vaguely stable material properties" would be the result...especiall because properties are often quite dependent on polydispersity (i.e., chain lenght distribution).
There's an emulsion polymerization method which might be amenable to PLA manufacture. Pretty much just get the lactic acid into a solvent and then emulsify that with an immiscible liquid carrying a catalyst. The tiny droplet size in emulsions limits the chain length but supply of catalyst and temperature controls lower chain lengths. I realize that this isn't amenable to quick-and-dirty usage but it might be useful in other settings.
the guy:

The finished emulsion may also be less energy-intensive to extrude than a melt, if the time to dry each layer of emulsion is long enough, and it can be made thick enough to hold up as a reasonably-sized layer.
Er...I meant, "if the time needed to dry each layer is tolerably short."

sorry, Joel, my explanation was too brief for correct understanding. The emulsion stage is just a step in the formation of a controlled-length polymer. The emulsion itself is not (as far as I know) extrudable. It's just a set of carrier solvents for the polymerization to occur in. The polymer powder would still have to be filtered out, rinsed and melted down into a manageable form.
this might be helpful:
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]