Friday, September 30, 2011

The Research behind "The Perfect Polymer"

Following up on the previous post, I did get a chance to read the paper and all I can say is "wow!". I'm going to review the paper first, and then will give some more grief to the PR blurb.

In a nutshell, the paper shows that the researchers looked at low density polyethylene (LDPE) and were able to predict flow behavior - linear, nonlinear and transient (!) - based on reaction conditions. That's plenty impressive. From the reaction conditions, they were able to predict the output material (molecular weight, molecular weight distribution, branching...) and then from that prediction, they were further able to predict the rheology. Again, that is plenty impressive. If you work with LDPE, this should be able to greatly aid in the development of new materials.

There are limitations to this work of course. First, it is limited to LDPE, although LDPE is nothing to laugh at. Given its extensive branching, modeling it is far more challenging than modelling it's linear cousin, high density polyethylene (HDPE), so I would expect similar results for HDPE sometime soon. Second, it is only predictions of the melt properties, not the properties of the final product after the polymer has cooled. Melt flow properties are plenty important, as the LDPE will have to be melted and forced to flow at some point in order to make useful products from it, but ultimately, people buy products based on the properties of the solid LDPE, not the molten LDPE.

Neither of these limitations that I just mentioned are criticisms of the work. It's great work. Papers that find universal truths for all of polymer science are very rare indeed, and nobody expects that.

But let me return to the original PR blurb. I am now even more flabbergasted by it. How they could prepare such a pile of trash is entirely beyond me. The authors never use the phase "the perfect polymer" (they do use the work "perfect" twice, but only in connection with polymer models: "Figure S1 shows an example of a time-dependent correlation map of relaxation time and priority (for LDPE 1). It also indicates the two extreme structures of perfect combs and perfect Cayley trees that constitute bounds for such maps." Emphasis added). Even within the realm of LDPE, they don't claim to have found the perfect LDPE, only that they now feel confident in making the noted predictions about it and short cutting the amount of experimentation. The reason we have so many grades of LDPE (Matweb lists 1176 grades) is that LDPE is used in so many different ways. A perfect LDPE? It doesn't exist. And the fact that the work doesn't make any predictions about the solid state properties means that experimentation is still needed.

I guess that means that, just like in the case of "The Ultimate Adhesive", we haven't reached the end of the road yet in polymer research. Our jobs are secure for a least another week (until next week's issue comes out). "Back to the lab, people!"

ResearchBlogging.orgRead, D., Auhl, D., Das, C., den Doelder, J., Kapnistos, M., Vittorias, I., & McLeish, T. (2011). Linking Models of Polymerization and Dynamics to Predict Branched Polymer Structure and Flow Science, 333 (6051), 1871-1874 DOI: 10.1126/science.1207060

3 comments:

Anonymous said...

Another copy and paste blog post? Yours,
Acheter vimax en France.2011AVEF

John said...

Cut and paste? From where? Hint: take a key line and google it. See it anywhere else?

Anonymous said...

Hello! My name is Frank. I'm engineer.