Battery Scientists Seek Smaller, Lighter Cells

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI  January 1998 Column

R&D HORIZONS

Lithium chemistries have altered the industry, but safety is still a concern.

The recent boom in personal electronics—particularly laptops and cell phones—has fueled a general resurgence in battery research. Manufacturers are naturally looking for smaller, lighter, more powerful cells, and battery scientists have been answering the call. Although most technologies have targeted nonmedical applications, medical device manufacturers are sure to benefit from the latest advances, too.

A thin-film rechargeable lithium battery developed by researchers at Oak Ridge National Laboratories (Oak Ridge, TN).

Much of the current work has focused on two distinct areas: lithium chemistries and fuel cells. Lithium has the smallest atomic mass and the highest electrochemical potential of any metal and naturally lends itself to the production of lighter cells with greater specific energy. Primary (nonrechargeable) lithium batteries have been in use for some time, but secondary (rechargeable) batteries have received the greatest attention lately, thanks to the advent of lithium-ion cells. Fuel cells, generally designed for larger applications, do not need recharging—just a constant supply of fuel, usually in the form of hydrogen.

LITHIUM-ION BATTERIES

Steve Visco, a researcher for Lawrence Berkeley National Labs and Poly Plus Batteries (Berkeley, CA), recounts the evolution of lithium-ion batteries—or rather, of the electrolyte, which progressed from liquid to polymer to gel. "Liquid electrolyte is the most conventional," Visco says. "You still have some sort of a metallic lithium film for the negative electrode (anode) and then some kind of positive electrode (cathode)," such as manganese oxide or titanium disulfide, together with an organic electrolyte. Traditional batteries typically use an aqueous electrolyte, but lithium reacts with water, and so some other compound must be used. Early versions for cell phones, it turned out, did not adequately address the highly reactive nature of lithium and tended to explode after repeated recharging. "Almost everybody understood the potential safety problems with metallic lithium," Visco says, but no one admitted to problems in the lab until after the accidents. The problem was that the manufacturers tested fresh batteries, not batteries that had repeatedly cycled.

As Visco explains, a lithium battery is constructed with a thin metal foil. Discharging actually moves the lithium from one side of the battery to another. "When you recharge it," he says, by running an electric current in the opposite direction, "you hope it goes back to its original state. But it could deposit in a granular high-surface-area form," in what looks like a spongy growth. After repeated cycling, he says, "you get a very active metal with a high surface area, and if it gets sufficiently warm, you could get thermal runaway." Explosive venting might then result. The safety problem only manifested itself after 20 or 30 cycles, and even then, perhaps only 10 in a million batteries actually failed—but that's still too much, particularly for medical applications.

The next generation used a polymer material rather than a liquid to shuttle the ions across the battery. Unfortunately, the most logical dry polymers (from an electrochemical perspective) don't work well at room temperature. Of course, that might not be a problem for medical manufacturers. Visco and his colleagues have been experimenting with different cathode materials. "We've found variations that work at body temperature quite well," he says. "We've run these things at around 30ºC, and they have good performance. They're fully solid state, they've got no liquid, they cycle great and actually have an energy density exceeding that of the lithium-iodine pacemaker batteries."

Improving on the solid polymer electrolyte, the next generation of lithium-ion cells used a polymer gel. The electrolyte in these batteries, formed by adding a little organic liquid to a polymeric material, behaves like a solid even though it is not a true solid. Early attempts using a lithium metal anode and a vanadium-oxide cathode showed promise but ultimately failed to meet performance requirements. Visco's group had more success using a sulfur-based cathode with an electrolyte gel made, for example, from polyethylene oxide. The sulfur electrolyte interacts with the lithium to prevent the spongy deposit.

But Japanese researchers went even further, creating a lithium-ion battery using a carbon cathode. "When you recharge the battery," Visco explains, "the lithium just goes into this carbon sponge," practically eliminating the safety hazard. Production capacity has since grown exponentially, to the point that now most laptops use this type of lithium-ion battery. But there's still room for improvement. "Lithium technologies tend to transmit high energy per volume and high energy per weight," Visco explains. "One area they don't excel in yet is rate—so you're not going to see lithium-ion cells in power tools." Moreover, most lithium-ion batteries require integral microcircuits that literally watch the voltage of each cell to prevent overcharging (4.2 V is generally the maximum), which would cause them to detonate or vent explosively. "That makes them expensive," Visco says.

BATTERY ECONOMICS

Expensive or not, most battery researchers and suppliers agree that lithium-ion cells are the hottest thing to hit the market in the past several years. "I would say the best chemistry so far is lithium-ion," declares Frank McLarnon, staff scientist at Lawrence Berkeley National Labs. "It didn't exist five years ago and has become the most actively researched rechargeable battery worldwide. I wouldn't say it's a revolution, but it is a major shift in terms of forward-looking battery technologies."

Rechargeable lithium-ion batteries from Battery Engineering, Inc. (Canton, MA).

McLarnon describes his work as somewhere between fundamental and applied research. In particular, he's been trying to identify the factors that limit the life and the performance of an electrode through spectroscopy and cell testing and characterization. "We try different chemistries and manufacturing methods, and we'll try something that's classically done and find why technology works the way it does."

McLarnon points out that lithium-ion cells are still far too expensive for many applications, although again, that might not be such an issue for medical OEMs. "If you had a medical application that required a rechargeable battery," he says, "longevity and reliability become paramount goals." A car manufacturer might want to spend $100 for 1 kWh, while a medical manufacturer might be willing to spend $100 for 1 Wh. "So it's orders of magnitude different."

Visco agrees, noting that "if you put a battery into a cell phone, you can bury some of the cost of the battery in there somewhere." The same is certainly not true for a cheap flashlight. As a result, says Visco, "the more-high-tech batteries can really command a higher margin."

Battery Engineering, Inc. (Canton, MA), is among those suppliers seeking to command these attractive margins, actively promoting rechargeable lithium-ion batteries for high-end equipment. The company has developed and patented its own technology, which could very well fill some niche applications in the medical device market. "OEMs are looking for replacements for NiMH," says Sal Piazza, sales and marketing manager, citing concerns over weight, price, safety, and capacity. "A lithium-ion polymer battery uses a solid electrolyte—there's no liquid, so it's inherently safer." By way of example, the company's main line of lithium-ion polymer rechargeable batteries provide 125 Wh/kg energy density, which is four times better than NiCd at about one-fifth the weight. Manufactured from a lithium-ion and cobalt oxide cell along with a proprietary polymer electrolyte, the batteries are rated at 3.7 V nominal with 5%-per-month self-discharge and can be assembled in series and parallel. Moreover, unlike NiCd, lead-acid, and NiMH chemistries, they are environmentally safe.

Of course, market competition is increasing and vicious. "One of the problems with lithium-ion technology," remarks Visco, "is that it's pretty easy to get a license to it. The materials are not proprietary." As a result, margins are getting squeezed, and the big battery companies that are not already involved are leery about entering the market. Even now, he says, suppliers are looking toward the next generation to surpass basic lithium-ion technology. The polysulfide technology Visco is developing for Poly Plus Batteries, for example, is "without a doubt the lightest-weight system you can conceive of," he claims, characterized by "a very high energy per weight and per volume—as much as a factor of two over lithium-ion."

THIN-FILM CELLS

Of course, in a few years even that may seem unremarkable—particularly in light of the recent work by researcher John Bates at the Oak Ridge National Laboratory in Tennessee. Bates and his associates have developed an entirely new type of rechargeable solid-state battery, made by successively sputtering thin films of electrodes and electrolytes onto a substrate together with a lithium anode deposited by thermal evaporation. The substrate can be, for example, the packaging for an integrated circuit or the packaging for the device itself. The batteries can also be made to conform to contoured substrates. The manufacturing process is similar to that used to make high-tech windows and readily lends itself to mass production.

According to Bates, there are no practical limits concerning size. Total battery thickness probably wouldn't exceed 10—15 µm, and area can be on the order of a few millimeters or fraction of a millimeter to hundreds of square centimeters. The one thing to keep in mind is that the batteries are not self-supporting but need to be put on a substrate." Energy density is "extraordinary," Bates says, "providing you can use the substrate that's already there. We're looking at 500—600 Wh/L and over 10,000 W/L—not counting the substrate." Integrating the batteries directly into a circuit would be most efficient, but possibly they can also be built upon very thin substrates such as 1-mil stainless-steel foils to reduce the relative volume and mass of the substrate. They can be built in parallel and connected in series to generate more power.

Bates had no real background in battery technology when he began his research. "We started from scratch," he recalls. "The key to the whole thing—the toughest thing—was finding the right electrolyte." The compound had to conduct Li⁺ ions well enough, yet remain stable while in contact with the lithium anode. His group worked for a number of years before hitting upon lithium phosphorus oxynitride, or LiPON as it is generally called. Performance is determined by the cathode material, battery area and thickness, and operating temperature. A crystalline LiCoO₂ cathode is generally preferred, providing single-cell operation at 4 V and continuous discharge currents up to several milliamperes per square centimeter. Amorphous LiMn₂O₂ cathodes might be more suitable for applications requiring manufacture at ambient temperature. The cells operate across a wide temperature range and offer long service life. All batteries have an internal resistance to ion transport, but thin films transport ions faster and easier than bulk materials do, resulting in a greater current density and cell efficiency. As for safety and reliability, Bates sees "no comparison" between his thin-film batteries and other types. "Lithium-ion is a pretty good battery but still very dangerous," he says. The solid-state battery, on the other hand, presents no hazard even if it gets crushed—a claim that few, if any, battery manufacturers can make.

Bates has understandably received a lot of interest from medical device companies. "Basically, any implantable device that needs a small compact power source" would be an appropriate use of the technology. The battery could be recharged inductively through the skin using technology already available. ECG electrodes provide another possible application. "Let's suppose you wanted to improve the signal-to-noise in an electrocardiogram," Bates says. "You'd put an amplifier on your patches and boost the signal right there. The battery is very small to begin with, and you can package everything in such a small size, so you can put everything right on the end of the cables."

While conceding that the technology would be hard to improve, Bates reports that his team is working on a new inorganic anode material that survives higher temperatures than normal lithium, which would further expand the applications for lithium-ion cells. Preliminary results are encouraging. "We've taken one of our lithium-ion batteries rapidly up to 300^ºC and took it out and cooled it rapidly and saw no change in performance," he says. "They're very rugged." One cell survived more than 7000 cycles at body temperature before Bates's team got tired of cycling it. "You reach a point," he says, "where you're not learning anything."

HARD DATA

Bates is currently looking for appropriate accelerated-aging methods to try to predict the battery's 10-year stability. "You can leave it on the shelf for a year and nothing happens, but we haven't been in this long enough to have any on the shelf for 10 years."

Device manufacturers and researchers alike bemoan the absence of long-term data. As Paul Skarstad, head of Medtronic's battery development division, succinctly says, "We won't use any technology—whether it's lithium-ion or nickel—metal hydride or any other battery chemistry—for human implantation unless we're very, very sure there is no safety hazard. That goes without saying." Nonetheless, he admits to some interest in lithium-ion technology. Lithium-ion cells have been on the market "long enough to begin to give us confidence," Skarstad says. "We're looking seriously at them. Have we made all the evaluations we need to that would give us confidence to make an implant tomorrow? No. That's not because we've found anything wrong—it's just that this work takes a long time."

Currently, all of Medtronic's implantable devices use primary cells. Medtronic is perhaps unique among device manufacturers in having its own battery development department, which has been in operation for more than 20 years. As for rechargeable implantables, Skarstad notes that the idea is indeed not new; a pacemaker with a rechargeable NiCd battery was introduced back in the 1970s. "From a technological perspective, that device was successful," he says. From a marketing perspective, however, it failed, principally because lithium batteries with 10-year lifetimes reached the market at about the same time, making the implantable rechargeable device obsolete almost immediately. "We can make small pacemakers with 10-year longevity," Skarstad says, "but there are other applications that require higher power and higher current, and for those it might be desirable to have a rechargeable battery."

FUEL CELLS

Skarstad's caution about lithium-ion chemistries involves "understanding their modes of failure," a sentiment echoed by David Watkins at Los Alamos National Labs. Watkins heads a group conducting research into proton exchange membrane (PEM) fuel cells, which, though targeted primarily at electric cars, could have far broader uses.

"With a fuel cell," Watkins explains, "you're supplying a continuous stream of fuel as the chemical reactant. In a normal battery, you're going through some chemical change in an electrode structure. In a fuel cell, you put hydrogen into one electrode and oxygen into the opposite electrode and, through a process that is directly analogous to everything else that a battery does, you create a voltage." The electrodes in PEM fuel cells use a polymer membrane coated with thin films of catalysts. The first electrode catalyzes the hydrogen to produce protons and electrons. The protons traverse the electrolyte and react with the oxygen on the other electrode, producing water. The cell requires some sort of mechanism to get both the hydrogen and oxygen into the system while letting the water out. The Los Ala-mos device takes oxygen out of the air and releases water as vapor. Hydrogen is contained and introduced through a metal hydride canister. The fuel cell does not require recharging but will produce energy as long as fuel is supplied.

Table I. Comparison of common batteries and a very small fuel cell. In designing a fuel cell for higher energy but not higher power, any additional weight comes only from the increased hydrogen storage. As a result, a fuel cell can have a higher specific energy and energy density than a comparable battery.

Watkins's group has a patent on a small fuel cell system that is potentially ready for commercialization, but he, like Skarstad, would like to see more confidence-building data. "We just don't know enough about the possible failure mechanisms because people haven't built enough of the cells. We've built maybe a dozen. My concern about anybody's technology is, have they really got a firm feeling for how reliable it is?" As a benchmark, Watkins points to the traditional lead-acid battery, which, he says, works "awfully well." Citing the achievements of project scientists Shimshon Gottesfeld and Mahlon Wilson, Watkins says his group can achieve the same total energy—and reliability—at a fraction of the weight (Table I). "We could make it maybe about one-fifth the weight in a small, portable device (10 W, 10 hours) and even better in a moderate-size low-power system. That's a big difference." A medical application that required the patient to carry a heavy battery could ultimately benefit, although it may take another four or five years to accumulate the data to support such applications. "If they have to carry five pounds, and we can make it half a pound, people are going to be happy," Watkins says. "But if it isn't reliable it doesn't matter, because they're not going to be both happy and dead."

Watkins concedes that many people might be somewhat wary about lugging hydrogen around but notes that any battery presents a potential hazard—as clearly demonstrated by the explosive failure of some early lithium-ion cells. "Any time you have a battery or anything compact that can produce energy, and the energy is in chemical form to be converted to electronic form, you end up with essentially some form of hazard. That's true of existing battery technologies; that's true of the hydrogen that you'd have in a fuel cell. It's a matter of how do you engineer a system for effective safety."

The Los Alamos team has run fuel cells in the lab for more than 2000 hours, though Watkins points out that actual battery life would depend on the size of the fuel-storage device and the power extracted from the cell. "What is important to understand is that 2000 hours of operation is a lower limit on how long such a system might operate," he explains. Anyone who has used a laptop, for example, has probably noticed that after several charging cycles, the battery begins losing its charge faster and faster and eventually dies completely. "In our case," says Watkins, "2000 hours would represent 200 'recharges' (i.e., replacing the hydrogen in the supply cylinder) without an appreciable decay in performance."

As an added benefit, PEM fuel cells are constructed from environmentally benign materials and present no problem of disposal. True, a precious-metal catalyst, principally platinum, is used on the electrodes, but Watkins's team has been very successful at reducing the amount of platinum needed without adversely affecting performance. "Depending on how you design it, you might have enough platinum where you'd want to recycle it," he says, adding that such a decision is based on economic—not environmental—concerns.

CONCLUSION

By all accounts, lithium-ion batteries represent a major milestone in battery science, but the industry is changing rapidly. As competition heats up, battery suppliers will have to find ways of increasing performance or reducing price. Visco expects his polysulfide batteries to top the energy density of a comparable lithium-ion cell by a factor of two. Bates hopes to make two thin-film batteries as cheap as 15—20¢ per square centimeter. Piazza's team is working on increasing both the operating and storage temperatures for their lithium-ion batteries. Ultimately, no one battery type will dominate the field; rather, specific applications will use the most suitable from a wide selection of cells. Unfortunately for device manufacturers, a specified battery might well become obsolete long before its service life ends.

Copyright ©1998 Medical Device & Diagnostic Industry