Fuel Cells Rally

Zero to 60 mph in 10 seconds, a range of 300 miles, and 98 hp: Are these the performance statistics of the Toyota Camry? Or perhaps the Volkswagen New Beetle?

No, this car has composite tanks capable of storing 8 kg of hydrogen. The manufacturer officially lists the power in electrical kilowatts rather than the more mechanical horsepower. And there are radiators behind the taillights as well as the headlights. It's General Motors' Sequel, a fuel-cell concept car unveiled earlier this month at the North American International Auto Show in Detroit.

As with any concept shown off in the Motor City, the Sequel--a "sequel," GM says, to a century of internal combustion engines--is meant more to impress an audience than to advertise a car that will actually end up in showrooms. But it is also a breakthrough in fuel-cell design. Previous fuel-cell demonstration vehicles accelerated from zero to 60 mph in 12 to 16 seconds and had a maximum range of only 250 miles, GM says.

The company's goal is to build a fuel-cell system by 2010 that can compete with the internal combustion engine to power its versatile "skateboard" construction chassis. "The vision is real," said Larry Burns, GM's vice president of R&D and planning, upon the unveiling. "Not yet affordable, but doable."

Both performance and cost need to improve before fuel-cell vehicles can end up on the road. Responsibility for much of the improvement rests on the proton-exchange membrane, a barrier to hydrogen but not protons that is considered the "heart" of the fuel cell. The membrane allows a catalyst to strip electrons--which power the electric motor--from the hydrogen before the hydrogen reacts with oxygen to form water.

It is improvement of the membrane and membrane electrode assembly--comprising the membrane, anode, cathode, and catalyst layers--that will make fuel-cell vehicles affordable. Fluoropolymer producers such as DuPont, W. L. Gore & Associates, 3M, Solvay, Arkema, and Asahi Glass are applying their expertise to fuel-cell applications. Other firms such as PolyFuel, a spin-off from SRI International, and Pemeas, a former Celanese unit, are developing new nonfluoropolymer chemistries.

James D. Balcom, chief executive officer of PolyFuel, stresses the importance of the membrane to the overall effort. He ranks the membrane with on-board hydrogen storage and a hydrogen-fuel infrastructure as the chief obstacles to mass-marketed fuel-cell vehicles. "It is like controlling the central processing unit of a computer," he says. "The CPU determines the performance of the computer, and the membrane determines the performance of the fuel cell."

Complicating the task of improving membranes is the host of requirements they must meet, points out Andrew M. Weber, global business director of DuPont Fuel Cells. DuPont has been supplying Nafion membranes for fuel-cell applications since they were first used in the U.S. space program in the 1960s. "In fuel cells, the membrane and the polymers in the membrane are being asked to do quite a few things," he says. "They are being asked to provide mechanical stability. They have to act as a barrier from one standpoint and allow diffusion on the other."

Weber adds that the membrane also needs to be durable under the high-heat and austere chemical conditions in the fuel cell. "The trick is balancing properties and being able to deliver on all the attributes that are needed for an effective membrane in a fuel cell," he says. "That's the challenge for all the people trying to get into this field."

FUEL CELLS have a long way to go before they become commercial in automobiles. The Department of Energy has set a goal of fuel-cell demonstration vehicles by 2009 with a 300-mile range, a 2,000-hour lifetime, and a cost of $125 per kW of power. It also seeks laboratory fuel cells with a 5,000-hour lifetime and $45-per-kW cost by the same time. According to PolyFuel, current performance is a 1,000-hour lifetime and a $200-per-kW cost.

PolyFuel has projected that to meet DOE's goals, membranes would have to cost at most $150 per m², about half of what demonstration membranes cost today. The ideal cost, the company says, would be about $35. "There is about 10 m² of membrane in the fuel cell, so the cost of the membrane is not trivial," Balcom says. "Three thousand dollars' worth of membrane is more than engines cost today."

In addition to cost, PolyFuel says commercial membranes will have to provide sustained operation at 95 to 100 °C; membranes today hit the 80 °C mark. The firm also says membranes should be able to operate at a relative humidity of 50% or even 25% at 80 °C. Now, they operate at 80%.

Michael M. Lynn, commercialization manager of fuel cells at 3M, says improving the properties of fuel-cell membranes will help drive down the cost of the overall system as auxiliary equipment needed to make up for membrane shortcomings is eliminated. "If you can operate in dry conditions, you don't need humidifiers," he says. "If you can operate at higher temperatures, you don't need all the cooling systems and radiators."

The component suppliers and auto companies that are developing fuel-cell systems have made significant technological advances recently, Lynn says. "Our improvements from two years ago are two orders of magnitude better," he says.

DuPont's Weber agrees, noting that developers of components, like his company, are getting valuable feedback from the car companies. "You are getting demonstration vehicles out there more and more," he says. "Every day, you get a clearer and clearer picture of what's needed, not only from a technology standpoint, but actually what's needed and driving the commercial development."

And the membrane and membrane electrode assembly suppliers are keeping up with the pace. In October, PolyFuel announced a prototype hydrocarbon-based membrane that it says costs less than perfluorinated membranes such as Nafion while being stable at 95 °C and 35% relative humidity. The company also says the membrane is 16 times as stiff as Nafion, has only a quarter of the hydrogen permeability, and puts out 10 to 15% more power.

Balcom offers few specifics on what chemistry PolyFuel uses, other than to disclose that the membrane has sulfonate conductive sites similar to those in Nafion. These conductive blocks, as he calls them, are held together by hydrocarbon structural blocks.

The membrane has generated a lot of interest from fuel-cell system developers, according to Balcom. "It is our expectation that we will reach a close relationship with one of the automotive companies in the near future to help develop the next-generation automotive membrane," he says.

Pemeas, which was spun off from Celanese last April, has been working on its platform of polybenzimidazole-based (PBI) membranes for about a decade. It has a membrane electrode assembly with a temperature performance reaching 200 °C. The company is eyeing applications such as stationary power generation, portable electronics, and transportation.

Horst-Tore Land, Pemeas' CEO, says his company has also been improving other properties. "We were able to significantly improve the performance, robustness, and the durability over the past few years," he says. "Ongoing tests, with more than 11,000 hours of run time and a very small degradation, indicate the competitiveness of our membrane electrode assembly."

Balcom predicts that nonfluorinated membranes will prove to have advantages over perfluorinated membranes in the long run. Because the technology for hydrocarbon membranes is newer than that for perfluorinated membranes, he argues that hydrocarbons may have more room to improve. "There have literally been hundreds of millions of dollars spent on perfluorinated membranes, and still they haven't hit the commercial target," he says. "Hydrocarbon membranes are coming on strong, and I think they have the legs to go the distance."

Balcom points to PolyFuel's own work and earlier work Honda Motors did to validate the feasibility of hydrocarbon membranes in fuel-cell cars. Until Honda's endorsement, Balcom says, observers had little faith that hydrocarbon membranes could meet the requirements.

Moreover, Pemeas' Land says nonfluorinated membranes, particularly its PBI, may have a long-term cost advantage. "We see many more new opportunities to improve the production process and the material," he says. "In addition, the production of PBI itself is less costly than fluoropolymers."

However, perfluorinated membrane developers question whether hydrocarbon membranes will be able to meet chemical and high-temperature stability standards at the same time.

3M's Lynn concedes that hydrocarbon membranes have potential. "Today, the perfluorosulfonic acid membranes have achieved superior performance to the hydrocarbon membrane, but that doesn't mean that a hydrocarbon membrane couldn't be developed that can achieve it," he says.

He doesn't see this happening anytime soon, however. "In order to get the very high performance and very good durability, lifetime, and robustness that is required for an automotive membrane electrode assembly, the best candidate today is the perfluorosulfonic acid membrane," Lynn says.

FLUOROPOLYMER membranes are hardly at the end of their rope. Producers say they have made great strides in recent years and are promising many more.

Mike Houser, leader of sales, marketing, and strategic alliances for fuel-cell technologies at W. L. Gore, says his firm's platform of membrane electrode assemblies based on composite reinforced polytetrafluoroethylene is durable and offers high power density. It also operates at 80 °C and has a 1,400-hour lifetime. Its output, Houser says, made it the membrane of choice for the fuel cell in the GM Sequel. It is also used in the fuel cells that GM installed at Dow Chemical's plant in Freeport, Texas.

Japan's Asahi Glass announced in September that it had developed a fluorine-based proton-conductive polymer composite membrane that is capable of 2,000 hours of continuous operation and can withstand temperatures as high as 120 °C.

Because 3M uses a different monomer than does DuPont, Lynn says, his company's membranes have a glass transition temperature of 125 °C, 25° higher than that of Nafion. The company is also developing systems that work with lower humidity.

Solvay is working on two different membrane chemistries. Through its 1990s purchase of the French company Morgane, Solvay acquired technology for radio-grafted membranes--made through radiation-induced grafting polymerization based on sulfonic acid groups and a fluoropolymer backbone. These membranes were originally used in chlor-alkali production. Guy Laurent, a program manager for fuel-cell research at Solvay, says the company has since found that they are suitable for low-temperature direct-methanol fuel cells. In such cells, methanol is used rather than hydrogen gas.

With its 2002 purchase of Ausimont, Solvay acquired capabilities in perfluorinated membranes. It claims that the monomer it uses permits an operating temperature of 110 °C. The company is looking at stationary and automotive applications with these membranes.

Arkema, the soon-to-be-independent chemical unit of Total, is in fuel cells through its fluoropolymer business. Last year, the company, along with membrane electrode assembly developer Johnson Matthey, fuel-cell system developer United Technologies, Georgia Institute of Technology, and the University of Hawaii, received a $5.8 million grant to support a DOE fuel-cell project.

DuPont's Nafion, which has been around for about 40 years, is the product that every other developer compares itself to and is gunning for. "Nafion has been the workhorse in this whole area for years and years, not by accident, but because it does balance all the properties," Weber says. "Everybody is going to be targeting the leader."

He maintains that DuPont is staying ahead by making steady improvements in its membrane platform. "There is a lot of engineering that can be done in Nafion."

Weber also contends that the economies of scale that will come when fuel cells are commercialized will lower much of their cost. "Right now, the volumes are in the demonstration and developmental level," he says. "I think a lot of the discussion around cost is actually misplaced at this point."

But commercialization will be a long time coming. Automakers don't expect fuel-cell cars to be competitive with internal combustion vehicles until the end of the decade. Mass marketing will take another five or 10 years.

Most observers expect other markets to come before automobiles, largely because of the massive hydrogen infrastructure that will be needed. One is direct-methanol fuel cells for consumer electronics, where technology for methanol storage cartridges is readily available. Another is stationary power, where the fuel-cell performance requirements aren't quite so stringent.

According to Laurent, Solvay is focusing first on methanol-based portable electronic applications. "We believe the market will be one of the first markets to grow up in the next two to three years," he says.

Land says most of Pemeas' cash still comes from venture capital, but he expects that its revenue in a couple of years will come from portable electronics and stationary fuel cells. Pemeas could use the money. Although the company raised $26 million in 2004, Celanese recently disclosed that it lost $7 million in the first nine months of 2004. Celanese is looking to sell its remaining stake in the company.

While nearer term markets beckon for some, Gore's Houser is counting on making enough progress in automotive markets to put vehicles on the road. "Fuel-cell vehicles have to be just as convenient, just as safe, and no more time-consuming than vehicles right now," he says. "We are enabling that trade-off where it makes sense from a cost/performance standpoint to switch."

See an animation of how fuel cell works