Have there been any big nanotech patent conflicts yet?

No, but there are certainly a lot of deals being worked on. And companies are flexing their IP [intellectual property] muscles and saying, "We have all these patents and you [competing companies] are trying to do something in our space, so let's make a deal." That's always going to happen in the beginning stages of a new technology.

If they don't make a deal and products start to come out, at that point you're going to see litigation—once there are damages based on products being sold. That's when there's a basis for litigation. Right now [the value of patents], from a litigation perspective, is speculative, because in most cases there are no products on which to base a damages calculation. The best you can do is get an injunction to try to stop research from happening. And that may not be all that useful. Litigation is so expensive, and if you don't really know if a product is going to come out or not, you might be wasting precious dollars on the lawyers that could be spent on research.

Nanosys Inc. in Palo Alto, Calif., which is planning to go public, has a large cache of patents but no products. Is this situation going to be typical of nanotechnology companies?

If you look at the spectrum of nanotech business models, Nanosys is on the patent-heavy side of the spectrum. That's in contrast to those companies that are already putting out nanoparticles, such as those used in sunblock and for polishing semiconductor wafers, who may rely on trade secret protection for their manufacturing processes. There are several other companies that are working on accumulating key groups of patents, but I wouldn't say that acquiring patents dominates their business plans. It's just one strategy to create an asset that will help those companies in the event that their product development plans don't go as well as they'd hoped. IP is an asset that can exist independently of product development. Certainly you can generate significant revenue from patents—you've probably heard of Rambus. [Rambus, Inc., Los Altos, Calif., earned more than US $23 million in 2003 from its random-access memory architecture patents.] But since there are not so many nanotech products on the market right now, there isn't much of a royalty base in place to generate revenues.

How are people valuing nanotech intellectual property?

You can take the example of Nanogram Devices Corp. It was able to generate a valuation of $45 million, as I understand it, based primarily on the [rights to exploit] IP. [Investors] had initially invested $9 million; so they got a pretty healthy return. [The valuation came when Wilson Greatbatch Technologies Inc., in Clarence, N.Y., bought the company in March.]

There's another example. A company called Nanogen, Inc., in San Diego, saw its stock jump 50 percent in one day, based on a press release [on 3 December 2003] about one process patent that was just issued.

[These examples] show that the market reacts quickly and not necessarily rationally; so I think what you really need to do in terms of valuing nanotech patents is to dig in and look at the scope of the patent. Get the file history out of the patent office—that is, the history of the objections that were made to the patents and how the attorneys overcame them. You look at the whole patent, not just the issued document, and you figure out what weak spots are there and how broad it really is when you consider what representations were made to get the claims allowed. Also ask, Does the patent cover critical areas, things that other companies will have to go through to get to a certain endpoint?

Let's say you're talking about nanowires [nanometer-scale rods usually made from a semiconductor]. There's a lot of hope that nanowires will be used in transistors and all kinds of electrical devices in the future, once [engineers] figure out how to orient them the right way, connect them, and build up more complex architecture. Well, say that you have a patent on nanowires; let's figure out how much it's worth. First, does it cover all nanowires or just some nanowires? What are the limitations? Does it cover nanowires of a particular length and width, and how likely is it that these particular nanowires are going to end up becoming the standard of choice for a given electrical component? Those are hard questions to answer, but that's the process you have to go through.

How good a job is the patent office doing at handling nanotech?

The patent office always has a lag period when a new technology comes along. The first thing you have to have is a classification system. The first thing that happens when you file a patent is that the patent office classifies it. If it's a fishing lure, it goes into a certain class. If it's a group IV quantum dot [a nanometer-scale crystal with size-related electronic and optical properties], then it should go into another class.

The problem with nanotech is that there is no classification system yet. The only class is for fullerenes, such as carbon nanotubes and buckyballs, because that's a particularly well-known area. So if you file a fullerene application, the patent office will recognize it as such. If you file a patent application on a molecular computing device that uses [mirror-image] molecules to switch back and forth, the patent office is not going to have a classification for it. It might go to the chemical group. It might go to the electrical group. It might get passed back and forth causing significant delays. And I know from some recent meetings with patent office officials that they're working to address that and trying to build that classification system for nanotech. In the meantime, nanotech companies are well advised to have their attorneys work proactively with the patent examiners to prevent unnecessary delays.

But it's a big challenge because nanotech cuts across so many industries. I sympathize with the Patent Office. They face a difficult path but a very important one, because until we have a good classification system, we can't recognize [patents] that are nanotech and those that perhaps are not. Nor can the Patent Office ensure that the most qualified examiner in a particular nanotech sector is assigned to a common group of nanotech applications. In nanotech, we're dealing with technology that's not described in prior patents. It's mostly described in journals. So if you don't search those areas, you don't come up with the best and closest prior developments, and you may make a mistake and grant the patent too broadly.

A patent that's granted too broadly has the presumption of validity and then it's difficult to correct that mistake without resorting to expensive litigation, which most companies cannot afford. So companies end up being forced to consider licenses and expensive design-arounds when, really, they shouldn't have [been put in that position], because the patent shouldn't have been granted in the first place.

What are the potential trouble spots in nanotechnology intellectual property?

We've seen a relatively high percentage of situations where clients have patents that overlap with competitors.

Of course, it's a lot easier to discover these situations now because we have publication of U.S. patents, which is a change in law that occurred in 1999. Almost all U.S. patents are published 18 months after their filing date, which makes it possible for us to see competitors' filings sooner than we used to. They used to be kept secret until they were issued, and that could take many years.

A problem area that we see is in what to do when you have an overlapping patent with a competitor. Say you're a patent lawyer and your client filed around the same time [as the competitor]. There's a procedure that's used within the patent office called an interference proceeding, to determine who should obtain the rights to the overlapping subject matter. The U.S. system is the only system in the world that still relies on "first to invent" rather than "first to file" [to settle disputes]. This means that even if you filed the patent first, say a month before your competitor, there is an option of provoking an interference proceeding to determine who actually invented it first, based on lab notes and other evidence.

These interference proceedings are a real problem, because they take years and years to decide, if they are not settled. In fact, we're seeing biotech patents that are just now being issued, after long-fought interference proceedings on patents that go back to the 1980s. These patents will emerge from an interference many years later, and during that time there's uncertainty about who really owns the technology.

It's not surprising that so much overlap is happening in nanotech, because in any sort of hot pocket of research, there are going to be groups competing to hit milestones and doing very fundamental things that result in broad claims that are necessarily going to have a higher chance of overlapping.

It is a challenge for researchers in nanotech to be aware of what competitors are doing and also to ensure the integrity of their invention records in case they are needed for an interference proceeding. Being aware of competitors' activities requires monitoring of the published patents. At our law office, we do a lot of monitoring, which enables us to pursue our clients' applications from an informed perspective.

What's the impact of an interference proceeding on a nanotech start-up?

There's a lot of strategy involved. When start-ups are trying to attract funding and they're faced with an interference, it creates a certain amount of unpredictability, which investors don't like. So it can be a hindrance to securing investors.

But interference can also be an opportunity. There are ways to carve out as much subject matter away from the interference as you can. So if you have nonoverlapping pockets that are described in your patent, you can spin those off as separate patents and hopefully get them issued quickly without interference, Then only the overlap area is left for the interference. When people see some patents issuing that you will control, it helps to secure investors.

You can also settle an interference proceeding just like you settle litigation. For example, you can use a cross-license agreement if the competitor is only interested in a narrow field of use. Say that your own field of use is molecular electronics, and your competitor is interested in diagnostics. You could settle this interference by giving each party the field of use it desires.

It doesn't always work that way. As I said, there are some very fundamental biotech patents that are just now issuing that have been stuck in long, long interference proceedings. It really shakes up the water when one of those comes out. People have to scramble to get a license.