AUTHOR COMMENTARIES

In the world of chemistry, there are polymers and then there are living polymers. The former are long-chained molecules, generally with inactive chain ends, composed mostly of the same uniform elements known as monomers. The latter are polymers with one or more active chain ends that will continue to grow, adding monomer after monomer until the chemist decides it’s time to terminate the process.

Living polymerization allows chemists to tailor-make polymers to suit the needs of whatever application they have in mind, from biomedicine to robotics to smart materials that respond to the most subtle imaginable changes in the environment. "The possibilities are infinite," says chemist Krzysztof Matyjaszewski, "and it all depends on conceiving and designing the molecules in the proper way and then synthesizing them precisely. One needs to control every aspect of the molecular structure to target the desired properties."

Living radical polymerization became a viable technology in 1995 when Matyjaszewski published his seminal paper on what he called "atom transfer radical polyermization," or ATRP (J.S. Wang, K. Matyjaszewski, Journal of the American Chemical Society, 117[20]: 5614-5, 1995). That paper has since been cited nearly 1,700 times.

Matyjaszewski himself remains one of the most influential scientists in all of chemistry. In the current update to Thomson Scientific’s Essential Science Indicators^SM database, Matyjaszewski is the third-most-cited author in chemistry, with nearly 14,000 total citations to papers published since 1997. His 2001 Chemical Reviews paper on ATRP, for example, has been cited over 1,300 times (see table), and nearly 50 of his papers have each received more than 100 citations.

Matyjaszewski, 57, received his bachelor’s degree in 1972 from the Technical University of Moscow and his Ph.D. four years later from the Polish Academy of Sciences. Since 1985, he has been at Carnegie Mellon University in Pittsburgh, Pennsylvania, where he now holds the title of J.C. Warner University Professor of Natural Sciences. He is also director of the Center for Macromolecular Engineering, as well as director of the CRP Consortium, which focuses on interacting with industrial corporations to commercialize products based on controlled radical polymerization.

Have you always been interested in the problem of living polymerization, or was that a relatively new theme in your research in the mid-1990s?

My research has always focused on polymer synthesis, on making well-defined polymers. When I came to Carnegie Mellon in 1985, I started working on living ionic polymerization and applying this concept to the synthesis of well-defined organometallic and inorganic polymers. And then, in 1992 or 1993, I began thinking about the principle of extending living polymerization to radical polymerization.

What does radical polymerization do for you that living polymerization does not?

Radical polymerization, as with any chain-growth polymerization, consists of several elementary reactions. Chain growth is first initiated, and then the chain continues to grow until eventually it terminates by interaction with a different radical. So you have the "birth" of the macromolecule, which is its initiation, then its growth or propagation, and then eventually it "dies" by termination. It can also "divorce" or generate "offspring," by chain transfer, in which case the growth transfers to another chain.

The main advantage of radical polymerization over prior-art living polymerization is that one can polymerize a very broad range of monomers. Radical polymerization today accounts for 50% of all commercial polymers—perhaps 100 million tons worth each year.

In 1956, Michael Szwarc invented living anionic polymerization, which is a process that proceeds without termination and transfer. These reactions, however, require very stringent conditions—practically zero moisture, zero air, and no impurities. This process has been successfully applied to the synthesis of block copolymers that enable commercial production of thermoplastic elastomers, an advanced rubber that can be recycled and reprocessed many times.

In a way, this was the very beginning of nanotechnology. Living polymerization allows incorporation of incompatible and immiscible polymer segments together in the same macromolecule; these segments phase-separate spontaneously into regular domains on the scale of nanometers. Commercial products based on living anionic polymerization started appearing in the 1980s, but were based on polymerization of a very limited range of non-polar monomers—such as styrene, isoprene, or butadiene.

The challenge has always been to make this polymerization process more user-friendly, to be able to use other monomers, and to relax the process conditions, which is what radical polymerization allows us to do.

So what’s the catch? Why isn’t radical polymerization itself sufficient to allow for designing polymers from the ground up?

The problem is that chain growth in a radical polymerization is very rapid; a monomer unit is added to the growing chain every millisecond. After perhaps 1,000 monomer units are added, the chain terminates.

So the life of a propagating chain—the time during which you can do some chemistry—is only about one second. That’s a very short time, and it makes radical polymerization very difficult to control. The question was, how could one extend the life of the growing chains from one second to one day, to provide enough time to do some chemistry and functionalize the chain ends?

How did your 1995 JACS paper solve the problem?

The idea I had in the early 1990s was to allow chains to add maybe one or two monomer units and then send them to sleep. They would grow for one or two milleconds and would then become inactive or dormant. Then, after a few seconds or perhaps a minute, they would wake up again, add a couple of polymer units, and again go back to sleep.

Thus, a one-second lifetime of a propagating chain is divided into a thousand pieces of one millisecond each, and, between these one-millisecond bursts of activity, a one-minute period of dormancy is inserted. This way, a one-second chain lifetime is now expanded to approximately 1,000 minutes, or one day. Therefore, we now have a day to perform a number of complex chemical transformations.

One can add a second monomer and controllably incorporate it; one can use multifunctional initiators to prepare very sophisticated structures, such as star polymers, brushes, or even dendritic or hyper-branch polymers; and one can make block copolymers, graft copolymers, or gradient copolymers. This is the basic advantage of controlled or living radical polymerization.

In the paper published originally in JACS in 1995, we applied this concept to copper-mediated atom transfer radical polymerization, or ATRP. We used very inexpensive alkyl halides as initiators and copper complexes with simple commercially available ligands as the catalytic system.

ATRP adapts a known organic-chemistry reaction, atom transfer radical addition, to polymer synthesis. Other transition metals can be applied as catalysts for ATRP, most notably ruthenium, used by Sawamoto. Other controlled/living radical polymerization systems developed in the past 20 years, based on nitroxides or dithioesters, were introduced by Rizzardo, Georges, Hawker, Tordo, and other groups.

Did you realize how big this was going to be??

Yes, because this level of control over radical polymerization was the Holy Grail in synthetic polymer chemistry. There are thousands of possible monomers, and so many possible structures, and there is a real need to control how they are incorporated into functional materials.

How has your research and the field itself evolved in the dozen years since your breakthrough?

Well, early on it was very easy for us. Whatever we did in ATRP, we were always first. We used new ligands, new initiators, applied new conditions. We made many new structures—various block copolymers, stars, hyperbranched polymers--and whatever we touched, it was something new.

Now we’re reaching a level of saturation in the field; there are many groups throughout the world working on ATRP. Approximately 1,000 papers on ATRP are published every year. At the same time, a deeper understanding of the process allows new possibilities for precise macromolecular engineering. The interest now is not only in understanding how these systems work, but also in making new copolymers with properties targeting specific applications.

Can you can give us an idea of the range of these applications?

Well, we can start with biomedical applications—say, for drug delivery, or signaling, or the engineering of tissue or bone. The polymers might carry a particular drug to a bio-target and release the agent with a controlled speed. A very interesting area is bio-conjugation, which covalently links natural products and synthetic polymers. There are many publications on applying ATRP to synthesis of "smart" materials that respond to changes in temperature, pH, salt concentration, or even to light. They can be designed to expand or collapse, and can be used as artificial muscles in soft robotics.

Some materials made by ATRP can be very tough but also super-soft, a material a thousand times softer than rubber. There are many new opportunities in optoelectronic applications, in advanced coatings, in sealants—but also in health-and-beauty products.

What’s been the biggest challenge in turning these applications into reality?

We can now make many new polymers that didn’t exist before, but we need to understand and correctly determine their properties. One of the biggest challenges, therefore, is the precise correlation between the molecular structure and the final material’s properties. These properties depend not only on molecular structure but also on processing.

Therefore, we progressively focus more on the entire macromolecular engineering process that includes not only the rational design of polymers with controlled architecture, but also the precise synthesis of these structures—controlling the processing in such a way that we can achieve the final, targeted properties of these advanced materials. In addition, for commercial production, the appropriate balance between cost and performance is needed, and the recent advances in reducing the concentration of catalyst in an ATRP will contribute to lower costs.

Until now, we focused on making all polymers exceptionally uniform. However, another challenge is to actually control the disorder—controlled heterogeneity, or controlled chaos, if you wish. Copolymers with irregular (but still controlled) branching, segment size, and monomer sequences (like in gradient copolymers) can provide materials with entirely new morphologies and new properties.

Another challenge is to continuously increase control over the polymerization and to exceed 99.99% selectivity. If you think about typical organic chemistry, when people get a 90% yield, they’re happy. But in polymer chemistry, 90% selectivity is a disaster. You can only add 10 monomer units before chain transfer or termination occurs. The selectivity is only 9 to 1. So we need to go from 90% selectivity to 99.9% or more.

Naively, one would think that 99.99% selectivity is impossible, more wishful science than reality. What makes you think otherwise?

This has already been demonstrated. Polymers with very high molecular weight, exceeding one million, have been prepared. This means that the same reaction occurred repeatedly, 10,000 or more times, without error. But doing that for one single polymer, for a homogenous polymer, is boring.

What’s exciting is to be able to switch to another monomer as desired, so that the end result is, say, a block copolymer, with two different segments in the same chain. These segments will phase-separate, depending on their composition and size, or length, of each segment. There might be five-nanometer separation between the two segments, or 100 nanometers between them.

Then, depending on the mole fraction of each monomer, the phase-separated block copolymers adopt various regular structures: spheres or cylinders, like wires, or lamellae—layered structures—but also more complex morphologies with many potential applications in nanotechnology.

You recently had a breakthrough in lowering the amount of catalyst needed. Can you tell us about that?

This was an idea based on the concurrent use of reducing agents. Even in living radical polymerization, a small amount of unavoidable termination occurs, and this process irreversibly consumes the catalyst. Thus, relatively large amounts of catalysts were added to overcome this problem.

The novel thought was to use a very small amount of catalysts, essentially a few parts per million, together with environmentally friendly reducing agents, which could be vitamin C, ascorbic acid, or a little sugar like glucose to regenerate the catalyst.That kind of environmentally friendly solution is what industry needs to reduce catalyst removal/recycle costs, and that’s what we reported just last year.

How difficult is it to do ATRP in the laboratory?

It’s so simple that undergraduate students can do it without special training. When I was a graduate student 30-odd years ago, doing living anionic or cationic polymerization was extremely difficult. We had to learn how to do glass blowing and solvent purification, and it would take a month of training just to do the initial experiment.

We can now teach undergraduate students how to make almost-perfect block copolymers on a bench in two hours.

What are your hopes for the future of radical polymerization?

It’s now 12 years since this discovery and our first patent on ATRP. Right now, this is still a process that’s only used commercially for highly specialized products. I hope it will be used in more applications and also for larger-volume applications.

I’d like to see people like physicists begin making these polymers themselves, because it’s so easy. The same goes for biologists, who can make new bio-conjugates: you take a natural product—say, a segment of DNA or a peptide—and then link it together with one of the polymers made by ATRP. Because you can design these polymers intelligently, and force them to do what you want and when you want, you can attach such a polymer to an enzyme and either shut it down or activate it on demand. You can steer its activity this way.

Modification of surfaces is another tremendously important area. By grafting well-defined polymers from a surface, you can change its behavior and materials with totally new properties. Imagine growing polymers from a surface by adding units one by one to create the desired responsive surface. This allows preparation of extremely dense polymer films that can change the lubrication properties of the surface, get a different compressibility, prevent corrosion, and so on.

You can create such a dense polymer film layer that nothing can get beneath it and detect the nature of the original surface. This will be very important for biomedical applications and preparation of biocompatible and protein-repellant materials.

Bridging between these sibling technologies—let’s say between surface science and engineering, including optoelectronicss, robotics , and bioengineering—is opening up exciting new areas of research at surprising speed. I recall my very first discussions with some engineers, and they couldn’t believe it was possible to create block copolymers from almost any two monomers. Now we can do it easily and mareally useful materials that never existed previously.
 

RELATED> Krzysztof Matyjaszewski answers a few questions about his fast moving front paper (from the archives of ESI-Topics.com) for January 2008, "Macromolecular engineering: From rational design through precise macromolecular synthesis and processing to targeted macroscopic material properties," (PROG POLYM SCI, 30 (8-9): 858-875 AUG-SEP 2005).

Note: links to external sites will open in a new browser window. Thomson Scientific and ScienceWatch.com does not endorse external sites.