Metamaterials | History

John Pendry and the Wire Medium

A material with a negative ? is often called a plasmonic material, since the coupled light-electron excitations that dance along a metal surface are called surface plasmons. In 1997, even before the field of plasmonics exploded, there were probably thousands of papers on plasmonic materials. Surface plasmons were believed to be the key mechanism in lots of exciting but (at the time) poorly understood and controversial optical phenomena, such as surface enhanced Raman scattering (SERS). Theories about the role of plasmons could be found everywhere, but it was incredibly difficult to decisively and quantitatively connect the theory to measured data. A plasmon is an optical, nanoscale thing - you can't hold a plasmon in your hand and look at it. You have to infer a lot of information, based on lots of different microscopy and light scattering techniques.

The optical work I was doing as a postdoc was vastly different from the microwave scattering work I'd done as a graduate student. By comparison, microwaves were easy! You could model just about any kind of structure, and whatever you modeled you could measure almost exactly. Very little guesswork involved. You could also make samples and do measurements really, really fast. Part of the reason is that the microwaves we were using were electromagnetic waves, with wavelengths of many inches, unlike light waves which are just a few hundred nanometers in size. If you want to make something that reflects or scatters microwaves, it's big! - usually about the size of your hand or larger. You can really get an intuitive feel for how microwaves interact with structured materials, and can try lots of experiments quickly. By the time I had finished my graduate work, I could design a photonic crystal, simulate its properties, fabricate the structure and make the measurements all in one day. That was awesome. We definitely could not do that with optical plasmons.

I began to obsess about whether there could be a way to create a microwave analog to the metal nanoparticle.

The big problem with creating a microwave "plasmonic" material was that there were no known materials that had a negative ε at microwave frequencies. Negative ε was considered an optical phenomenon that occurred only in conductors at near-visible and ultraviolet wavelengths. At microwave frequencies metals are—well, just metals: They exclude electromagnetic fields. If the fields can't even get into the metal, they can't interact with the metal in any interesting way. Metals at microwave frequencies don’t support surface plasmons, and definitely cannot be considered "plasmonic."

But, maybe there was a possibility. While browsing through the thousands of papers on plasmons, I ran across a paper by John Pendry and colleagues, published in Physical Review Letters in 1996, in which they suggested an artificial material—one composed of wires - could behave exactly like a plasmonic material, but at any frequency. Including microwave frequencies. It was exactly what I was looking for!


Pendry's structure required really, really thin wires. Much thinner than any typical commercially available wire. If you could get those wires, you'd have to be really careful in how you arranged them and held them together - it would almost be like weaving a material out of thread. It wasn't anything we could build, at least not without a lot of effort. Could there be another approach?

As simple a structure as Pendry's wire structure was, the theory was fairly complicated. I took Pendry’s paper around to several of our theoretical colleagues at UCSD, and none of them could understand it, at least without having much more time to spend on it. Moreover, Pendry's theory was becoming wildly controversial, with lots of other scientists and theorists objecting to both Pendry’s approach and the results. Plasmons at microwave frequencies? Not a chance, according to Pendry's critics.

So, without a way to make the thin wire structure; with no one around who understood the paper; and with the paper mired in controversy, I really couldn’t justify delving much further into the subject.

A Trip to France!

I had always wanted to visit France. It was a lifelong goal. I’d been to a lot of different countries for various conferences, but never had been invited to one in France.

In 1998, however, an email showed up, inviting me to a conference called PIERS - Progress in Electromagnetic Research Symposium. I hadn't heard of the conference before, and I didn't know quite what it was about, but it was in Nantes, France, and I saw my opportunity. I talked it over with Shelly, and he agreed it was a good thing to do, so I was set. I just needed a topic. I quickly put together some ideas based on the work I was doing with Olivier Martin, and bought my tickets.

And here is where the randomness of life really comes into play. That conference turned out to be truly fortuitous and pivotal. In the session that I was in, it turned out there were lots of people talking about negative ε and even Pendry's wire medium. A couple of groups were actually doing detailed numerical simulations, and had succeeded in verifying Pendry's prediction. There were no experiments, but at least there was growing evidence that the theory was right. Still, it required really, really thin wires.

It also turned out that Eli Yablonovitch, who along with Sajeev John was one of the founders of the field of photonic crystals - was attending the conference and that session. Both Shelly and I had known Eli for many years, and so when I saw him we started comparing notes on the session. At the time, he was very interested in wire structures as well, and was also interested in the possibility of microwave plasmons. So, we had common interests.

"You know," Eli told me, "I've organized a meeting on photonic crystals in a few months in Laguna Beach this year. Why don't you come and talk about microwave plasmons?"

Laguna Beach. It sounded great. I was in. But I didn’t know anything about microwave plasmons, other than that I was hoping someone would propose a structure that we could make.

"Sure," I responded, "but I don't know anything about microwave plasmons."

Eli, being one of the giants in the field, could be very persuasive. He replied "that's ok, you've got a few months. Just get some ideas together and come and talk about them."

I was still a little worried. I doubted there would be much I could do in just a few months.

"Ok," I said, "but if I can't come up with anything, can I switch my topic to photonic crystal accelerator cavities?" Photonic crystal accelerator cavities were what I had studied as a graduate student. I had lots to say on that topic. It was a good failsafe topic.

"Of course," Eli assured me. And that was that.

I spent two weeks traveling around France, and it was spectacular! Thoughts of wire structures and plasmons left my mind. There was no rush and no impending deadline--I could always switch my topic.

When I returned to San Diego, there were many other priorities to worry about, and I only had a little bit of time to think about the microwave plasmon problem. After a few weeks, seeing what I thought was the writing on the wall, I called Eli and told him I was going to switch my topic to accelerator cavities as we had agreed.

Eli thought about it. "No, I still like microwave plasmons. Keep trying."

"Ok," I said, "but if I really can't come up with anything, I can still switch, right?"

"Yeah, sure.

So, I wasn't off the hook, at least not entirely. But I still had too many other things to do, still didn't understand Pendry's paper, and still couldn't get past the issue of super thin wires. So, another couple of weeks passed, and again I still had nothing. It was definitely time to bail out.

I called Eli again.

"Look, I really haven't been able to come up with anything on microwave plasmons, so I've got to switch back to accelerator cavities."

Eli didn't hesitate. "No, that's boring. I really want to hear about microwave plasmons. Just put together what you have and do a poster."

It was only three weeks before the conference! Now I was in a state of panic.

That night, I pulled out Pendry's paper, and finally started to think about it seriously. Somehow, under the duress of Eli's enthusiasm, I started to make progress. John had provided a very high-level, elegant theory for the effective medium response of a bunch of thin wires, and he had wrapped the results in the language that would come naturally to a condensed matter theorist. But maybe there was an alternative way to arrive at his formulas, which were in the end fairly uncomplicated. And maybe this alternative approach would provide the key to designing a structure easier to fabricate.

As soon as I started thinking about an alternative derivation, everything came together almost instantly. The wire could be viewed as an electrical circuit, and making the wires thin was an elegant way of increasing the self-inductance - a simple circuit parameter. The larger the inductance, the more the wires would behave like a plasmonic material. But thin wires weren’t the only way to increase inductance! You could, for example, take an ordinary wire and curl it around to form a series of loops, kind of like a solenoid or a slinky. That's the usual way, in fact, to create an inductance. An array of loop-wires would form a plasmonic medium in the same way that an array of thin wires would. After having this epiphany, I went back through Pendry's paper and discovered one brief line in which he, in fact, had mentioned the thin wires were a way of increasing self-inductance. So, I had only rediscovered the result that was in Pendry's paper the whole time. But Pendry had not stressed this particular aspect, probably feeling that the thin wire description was more physical.

Now, with less than three weeks before the Laguna Beach conference, I finally understood Pendry's paper and had an idea how to make a microwave plasmonic material. But there were many things to do to create a poster for the conference. I needed to confirm the idea with a specific structure, and I also wanted to do the experiment. I needed help!

The entire microwave plasmon research project was, at that point, just a hobby. It had nothing to do with the optical nanoparticle microscopy we were doing, except as maybe a fun little analogy, and certainly had nothing to do with the company Seashell we had started. There was not much of a way to justify using lab resources to officially investigate this topic.

But, one of the great features of university research is that a good idea can almost always find a way to be realized. The students and postdocs in our group were all friendly, and were happy and eager to work on something new, even if it meant more work for them and projects that had to be done outside of the normal workday. This would definitely have to be a nights and weekends activity. At the time, David Vier, a senior research scientist in the group, was running simulations for the photonic accelerator project, and I knew he would be able to help me with the loop wire simulations if he could spare the time. Syrus Nemat-Nasser, a graduate student in our group, could help both with simulations and measurements. And Willie Padilla, a first year graduate student who was just beginning his PhD at UCSD, was also available to help with constructing the sample and making measurements. Given that we had just a couple of weeks, I appealed to all of them to help make the plasmonic medium concept a reality.

Everyone was enthusiastic and agreed to help!

So, in less than two weeks, we had generated the theory for the loop-wire medium, performed the simulations, fabricated the material, and made the microwave equivalent of a plasmonic nanoparticle. It was kind of amazing. Surprisingly enough, everything worked exactly as it should. And just in time. Eli's friendly pressure had been just the right sort of catalyst!

Story Sections: 1 | 2 | 3 | 4