A1.1 A History of the Battery

Just a short time ago, the world of convenience, comfort, communications, and immense power, available and controlled by the individual, was nonexistent. That was hardly more than 100 years ago. The self-propelled automobile, aircraft, electronics, and all the other vitally important and generally accepted amenities of life, such as air-conditioning and nuclear energy as well as nuclear instrumentation for medical purposes, weren’t even imagined as probable engineering accomplishments.

The storage of energy at that time was hardly an important concern since primary sources of electrical energy were scarce. Before the 1870s, no practical, rotating machinery generators used coal or any other hydrocarbon for fuel. Hence, there was no need for a secondary system, such as a secondary battery, to store electrical energy. Charging such a battery would present some rather difficult problems. Plante invented the lead-acid cell, or battery, in about 1860. It was at that time, and still is, an excellent electrochemical cell for storing energy for prolonged periods of time at fairly inexpensive costs. Unfortunately, at that time there was little use for such a device. The principal interest was in primary sources, if not electromagnetic systems, then certainly for primary batteries to power telegraph lines and other special purpose systems in the middle nineteenth century. In fact, if it was not for the development of the internal combustion engine and, to some lesser extent, the electric car, then Plante’s invention might have lain dormant for an even longer time. The advent of the electric starter (first put into a Cadillac by Charles Kettering in about 1912 for test purposes) was the main force that put the lead-acid battery into high volume production. It was, and still is, the only relatively inexpensive battery that had the power density to start a large internal combustion engine for a long enough time to get ignition. Anyone who tries to develop an alternative storage system to the lead battery would soon develop a respect and appreciation to the lead system.

Much has been written on this subject in the form of papers, magazine articles, and books. One of the best descriptions of the early forms of electrochemical systems is in Volume I of George Heise and N. Corey Cahoon’s book The Primary Battery. The mid and late nineteenth century was indeed exciting and eventful with regard to many aspects of engineering and applied electricity (pre-electronic era of vacuum tubes and semi-conductors). Perhaps the most familiar form of the primary battery is the LeClanche cell invented in about 1860. The early forms were wet cells with zinc as the reducing agent and manganese dioxide as the oxidizer (initially called depolarizer). The modern versions (alkaline cells) are dry cells and have remarkably high energy densities and long charge retention times, making them suitable for many applications ranging from the familiar flash lights to modern day electronic gadgetry.

With the advent of greater technological demands, especially in warfare, and the increase in power needs of societies, increased efforts were made to provide power sources for the evolving machinery that required immense electrical power for short periods of time, such as the naval torpedo, in rapid development. Powerful and expensive batteries such as the silver/zinc cell and the zinc/dichromate cells were developed for these purposes. They were able to deliver large amounts of power for many seconds to minutes in order to propel, for example, a torpedo on its way to an enemy ship. These batteries not only required careful handling and trained personnel for proper and safe use, but they were also very costly. In special applications for military or space vehicle use, cost is at best a secondary consideration since the importance of the mission, and the costs of the rest of the associated equipment, are so great. In these instances, reliability is just as important as energy and power density. The lead-acid battery simply cannot perform for these applications, so other systems had to be developed. Unfortunately, few of these batteries are valuable in the commercial or consumer market. Despite the fact that requirements in the consumer market place were increasing, little was new or in way of significant improvements to the existing batteries.

Only recently have such new developments as the lithium cell moved forward in the industry. It’s major application lies with small portable electronic products that require small amounts of stored energy. The large system applications, such as the all-electric car, still await a practical battery.

A1.2 The Electric Car and the Power Source Search

Most of our early development efforts were directed towards finding new methods or perfecting existing methods of transforming energy from one form to another, mostly for transducer, communication, and instrumentation purposes. The detection of missile paths and the conversion of sonic to electrical signals were predominant. As events progressed and nuclear energy became tamed sufficiently for commercial and military use, our work became more directed towards control methods in reactor systems. As time moved on, it became more obvious to many that the generation and supply of primary sources of power for industrial and commercial use were not the lagging technologies. Instead, means of storing energy for later use or for portability purposes was the greater problem, and we will be increasingly confronting it. At that time, in the 1960s, the only commercially viable storage device for directly storing energy in electric form was still the lead-acid battery. It certainly seemed incongruous that such an ancient technology would predominate in our daily lives while all the other advances were taking place at a ferocious pace.

With the goal in mind of finding an electrochemical couple that would permit high energy storage density and long life, many of us embarked upon a long journey, only to realize how difficult it is to surpass the venerable lead-acid battery on all fronts. The two largest applications are automotive power and bulk storage for emergency and standby use. The lead battery really does not offer a solution due to its high cost and short cycle life. Especially in the case of the auto market, the ED is much too small.

To power a typical family car at 50+ mph, energy requirements range minimally between 250 and 400 watt-hours per mile of travel. If it is required that the vehicle be capable of traveling a minimum of 300 miles on one charge, we see that a total of 75,000 to 120,000 Wh of energy must be available from the battery. An upper limit for a passenger vehicle would seem to be in the vicinity of 1,000 lbs. That means that the battery must possess an energy density of 75 to 120 watt-hours per pound – a very high figure for any short-term foreseeable technology. The lead battery has an ED between 10 and 16 Wh/lb depending upon how it is discharged – a far cry from the required numbers shown above.

An interesting bit of information is the little known fact that the electric automobile began its short journey back in the nineteenth century along with many other competitive projects ranging from steam powered cars to compressed air propulsion systems. Ferdnand Porsche of VW first began his career with very advanced electric car designs. In fact he had four-wheel drive powered by separate motors in each wheel and regenerative braking in the 1890s. The photo shown in Figure A1.1 is an electric car of 1900.

The performance figures of these ancient vehicles are comparable to those presently attained by modern electric cars, ranging anywhere between 25 and 40 miles on fairly level ground at moderate city speeds. It appears that, despite all the updated electronic wizardry, there has been little, real progress in developing an electric all-purpose car. This is undoubtedly because of the lack of significant improvements in electric propulsion power, despite the sleek, lightweight body designs.

To add to the dilemma of the lead battery, its cost is too great, and its life is too short. Depending on the depth and rates of discharge, the lead battery will give useful service for between 500 and 1,000 cycles. These numbers are debatable, but the point is that its life and cost are outside the useful range. For example, a series of six, traction, lead batteries for a golf cart cost about $400+ for perhaps a total of 3,000 to 5,000 watt-hours of useful storage. For interesting reading on the subject of transportation, and particularly the history of the electric car, M. Schiffer’s book entitled “Taking Charge” is highly recommended, see bibliography reference 38.

A1.3 The Initial Survey

When attempting to find an alternative electrochemical process that might offer higher ED and longer life, the tendency is to search the known possible couples. If one wishes to produce a long-life system, then simplicity as well as reversibility are suggested. One approach to longevity of operation is to identify reactions that will not only proceed to completion but will also have reagents that will react directly if mixed together physically. In other words, the reagents in an electrochemical couple, if left alone, will eventually interact, leaving the state of the cell in its original condition before charging. This suggests that metal oxidizer couples such as the metal-halogens, e.g., metal reducers and halogen oxidizers, will not only react electrochemically in the appropriate circumstances but will also react by direct contact.

Of the five halogens, fluorine, chlorine, iodine, and bromine, the most attractive is bromine because of many factors. It is relatively inexpensive and plentiful, it is a liquid at room temperature with a tolerable vapor pressure, and it is electrochemically active with most metals. Bromine also has the ability to be stored easily because of its high solubility of bromide salts (and most of its salts) in water solutions, and active, porous materials readily adsorb it. Chlorine has many of these features, but it is not readily adsorbed, does no form complexes as easily, has a high vapor pressure, and hydrolyzes quickly in aqueous environments, thus creating HCl. Iodine (derived mostly from sea weed) would be a good competitor were it not for its high cost and high molecular weight. Fluorine, on the other hand, is totally unmanageable in aqueous solutions, and it is so reactive that few materials can withstand its very strong oxidizing nature. That leaves little choice but bromine for a workable halogen. Examination again of Figure 5.3 in Chapter 5 will show the relative ED for the various metal-halides. In an attempt early on to keep the cell simple and most compatible with ambient conditions (air and moisture environment, freezing to boiling points of water), only those metals that can exist in the free state in our environment were selected for further study. Despite the obvious attractiveness of lithium at over 1,200 watt-hours per lb, dry salt (lithium bromide), and other alkali metals, they were set aside for immediate study until methods of handling them should be developed in the future. Such possible methods include non-aqueous solvents (organics), high temperature, or fused salt operations.

A1.4 Review of a Research Path for a Long-Life, High ED Battery

The aqueous zinc/bromine system, at about 200 watt-hours per pound for dry salt and delivering 1.8 volts open circuit, offers most of the attractive features we would like. Zinc is an inexpensive and readily available metal, and its bromide salt is extremely soluble in water with a reasonably high electrical conductivity. Many configurations of this system were designed and extensively tested by many researchers, including the first author. The problems associated with this system are (1) the hydrolysis of bromine, (2) zinc dendrite growth when plating with subsequent shorting and/or fall off of metal from the negative electrode, (3) the formation of hydrogen gas as zinc is attacked by the HBr formed by hydrolysis, and (4) the loss of bromine in storage by hydrolysis and molecular diffusion.

Among the many attractions of the Zn/Br cell are its remarkably well-behaved performance, its high efficiency, its quite flat discharge characteristics, and its ability to return to its original discharged state of just ZnBr₂ solution when left for prolonged periods of time or when discharged completely.

The above listed problems are formidable and become more evident as development efforts progress. The storage of bromine has been accomplished rather poorly by mechanical means such as compartment or cuplike structures on the positive electrode and then later by merely storing the bromine in solution as the complex ion, Br^–, but diffusion is so severe in a static electrolyte cell that self-discharge occurs at a prohibitively high rate. This method does work if circulating electrolyte systems are employed and the positive electrolyte is largely stored in a reservoir external to the cell. Membranes of some sort must obviously be employed as well.

Some other techniques have also been developed by the first author, such as complexing the bromine with an agent that physically immobilizes the reagent without affecting its chemical activity. There are numerous such agents, such as the alkyl-ammonium halides, that will accomplish the task. The agents we selected for extensive study were various weights of polyethyleneglycol. Depending on the molecular weight of the complexer, the association with bromine structures ranges from a liquid to a solid. The diffusion rate of these types of complexes is much lower than free bromine, but it is still too fast for good charge retention. The method that has proven more successful is the adsorption onto activated carbon as part of the electrode. Between 20 and 30 Wh/lb was realized in large multi-cell modules. This electrochemical couple has been well investigated, and many fair sized systems have been constructed and operated.

At about this time there was an increasing interest in full flow electrolyte configurations. There are many reasons to seriously explore these types of battery systems because of the advantages offered by resorting to such increased complexities. Some of these advantages include the following:

• The separation of energy density from power density
• The possibility of “refueling” a battery rather than the slow electrical recharging process
• The increased charge retention times because of the separation of reagents during idle periods
• The greater control of operation in general, especially thermal control
• The increased design latitude for different applications, in principle
• The charge imbalance problems can be better handled than with static cells.

However, there are a few negative factors that must also be considered, such as the necessity for better leak proof designs to withstand the increased pressures internal to the battery, the electric current losses due to electrolyte interconnectivity through a common reservoir, the pump power losses, and the increased cost and complexity for such systems.

Examples of the Zn/Br battery systems are shown in Figures A1.2 through A1.5. The first is an installation that was made in about 1978 for the Duke Power Company in Charlotte, NC (Figure A1.2). Its capacity was about 150 KWh, and it has a discharge/charge cycle period of 16 hours. It delivered about 20 kW for almost 8 hours and was operated as an early test setup for about two years.

Small high drain units were also designed and constructed with very closely spaced electrodes. The unit shown in Figure A1.3 had a maximum discharge period of five minutes and delivered 5 to 15 KW depending on the length of time. The system experienced a fair degree of deterioration in capacity over the time it was tested. Postmortem inspections revealed a serious attack by bromine on almost all non-fluorocarbon materials as well as on many of the plastic components such as membranes. Using materials that are more resistant to oxidation by bromine could solve all these problems.

Figure A1.6 shows a battery fabricated for the Delco Division of General Motors in late the 1970s as one of a series of multiple cell modules to be tested for possible use in electric vehicles. Interest in electric vehicle projects diminished greatly, and the industry began looking long term at the possibilities of lithium-ion batteries that were being developed in Japan at the time.

Many versions and applications were explored for the zinc/bromine battery, both in static as well as full flow electrolyte designs. Figure A1.4 is a photo of a golf cart provided by the Cushman Company to TRL in 1970 as a test bed for some early batteries. A set of six Zn/Br batteries powered the cart almost 40% further and with even more acceleration power than the equivalent lead-acid batteries. However, the cost of manufacturing these units in the limited quantities was far too great to be practical. The low lead cost of lead batteries is primarily due to the automotive industry where the demands are in the tens of millions of units merely for new car production and many times larger for the replacement market. And, since the lead battery performs so well as the SLI system (starting, lighting, and ignition), there is little need for a new replacement technology. It becomes sort of the “chicken or the egg” situation; the costs remain high until a large market is established, and a large market requires low cost to get started.

All of these systems were full flow electrolyte developments (Figure A1.5), but the bromine was stored within the porous positive electrode. The purpose of the Circulation of electrolyte was to maintain a uniform temperature throughout the cells and prevent internal shorting by zinc dendrite. As interesting as these systems were, they were beset by many technical problems associated with the ones identified above.

Also, there is serious and understandable reluctance on the part of consumers as well as the industry to place products containing free bromine. That leaves the market potentials for such systems for special industrial and perhaps military uses, where proper training and precautions would be available.

In the 1960s there was an awakening of full flow batteries in the form of the NASA, Lewis-sponsored program to develop an iron/chromium redox cell. The chemistry is, as in other redox systems, quite simple. The reaction is

(A1.1)

where the ferric ions are the oxidizing agents.

This system is resplendent with problems as well. The effective operation of such a cell depends on a functioning anion exchange membrane. These types of membranes are not nearly as efficient as cations, and, consequently, the chloride ion usually transfers the electric charge in the cell. Eventually, the iron is transported over to the chromium die and vice versa with the result that it ceases to function. Costs and operational problems of this type have made the Fe/Cr cell impractical.

Because of the hazard and chemical attack problems encountered with bromine, we searched for alternative systems that posed less danger but still had the attractive properties of small failure modes. In 1971, some early engineering work was started at TRL, Inc. with what is known as the iron-redox cell. The attractive features of this cell include (1) low cost of materials, (2) safe chemicals, (3) a completely reversible operation, (4) a reasonable energy density of ~ 80 watt-hours per pound of dry salts, and (5) an acceptable potential of 1.2 volts open circuit.

Some of the problems encountered with the systems are (1) the low conductivity of the electrolyte, (2) the necessity for low pH to maintain salts in solution, (3) the poor quality of iron plating, and an inexorable rise in electrolyte pH due to hydrogen evolution at low pH.

Again, we embarked on an exploratory and development journey to see what practical problems existed and how we might produce practical devices. From about 1971 intermittently to 1993, extensive work was performed to develop and fabricate carbon-polymer bonded and compression molded electrodes with high conductivity and low porosity. Static electrolyte cells yielded poor results. Diffusion of ferrous and ferric chloride across almost any type of diffusion barrier is too great to provide cells with reasonable capacity and charge retention. Cation exchange membranes proved to be too costly and had much too high electrical resistance to enable cells to deliver useable power. If the pH of cells were kept very low, ion exchange membranes would function quite well, but the iron plating is consumed by the high acidity, and hydrogen gas evolution becomes a serious problem along with high self-discharge. If pH is permitted to raise the salts of iron, then chloride and oxygen begin precipitating in great quantities, thus rendering the cell inoperative.

Nevertheless, we proceeded to develop many ways of handling the problem, including ancillary automatic pH control units, which functioned quite well but needed some attention and consumed some of the charge stored in the cell. Numerous single cells and multiple cell arrays were built, and many of them were installed in various apparatus for test and demonstration purposes. Figure A1.6 is an early eight cell module employing a method of encapsulation of electrodes, membranes, and support structure, which essentially eliminates any possibility of electrolyte leaks within, from cell to cell, as well as to the outside. Prior to this form of construction, modules or arrays of cells were fabricated as plate-and-frame construction typical of most fuel cells, electrodialysis systems, and batteries, and they were fraught with leakage problems that rendered them impractical if not unsafe for general use.

The principal energy storing reaction for the iron redox cell is shown in Figure 5.2 of Chapter 5 in this book. It is a unique situation where the iron ion is moved both upwards and downwards during charging and during the discharge mode. There are few elements that offer similar properties for practical exploitation, vanadium being one other. The cost of the iron system is markedly lower, but it does have its own collection of problems.

Among the early working hardware that was constructed to demonstrate the potential practicality of iron/redox, a small power source was constructed and put in place of the gasoline-fueled engine in a go-cart. The version shown in Figure A1.7 was assembled in 1973 and operated as a fun vehicle for three years. It was a dual full flow electrolyte system in which the ferric (ferric chloride) solution was stored in the larger tank, and the ferrous solution flowed past the plated iron on the negative electrode. A microporous plastic separator (Daramic, produced by the W.R. Grace Co. for the lead-acid battery) was employed to keep the two solutions from mixing too quickly and internally discharging the cells. A twelve-volt auto starter motor was used to propel the cart.

Later in about 1974, a program was initiated with the sponsorship of EPRI and the Mississippi County Community College (MC3) in Arkansas to develop and design a storage system for a solar collector field to provide uninterrupted power to parts of the college. The total energy capacity was about 20 KWH.

Figures A1.8 and A1.9 show portions of the system that was built and tested in Durham, North Carolina prior to installation.

The larger tank held the ferric chloride solution, and the smaller one held the ferrous solution. The electrical load was a series of electrical heaters immersed in a cooled tank of water. As discussed earlier, the iron system was plagued with a couple of serious problems regarding pH drifting upward and iron plating falloff from the electrode surfaces.

It is possible that resorting to non-aqueous solutions can solve these problems, but not nearly enough data has been obtained to justify expenditures of large sums until more real evidence is acquired.

In order to avoid all the problems associated with plating solids onto electrode surfaces, the sulfide/bromine couple was identified and pursued as a possible solution to large-scale, station ary electrical energy storage. TRL proceeded to investigate this possibility in 1989 and accumulated enough understanding and empirical data of the couple S⁼/Br₂ to fabricate and test numerous laboratory single cells. Later, in 1991 National Power, PLC in the UK developed enough interest to support research at TRL and within National Power. The experimental results continued to be promising, and over three years of intensive R & D was invested in the couple.

As always, there are sets of attributes and problems associated with any approach to solving practical energy storage. In this case, the main issues that presented the greatest concern are as follows:

• Electrodes with long life at high current densities, i.e., greater than 1 ampere per square inch of active area
• Cation membranes that will stand up in high concentrations of free bromine, other than NAFION, a DuPont material that functions reasonably well in such cells but is much too costly at low current densities
• The loss of bromine to the entire system when sulfides diffuse into the positive electrolyte with high bromine concentrations (In rich bromine solutions, sulfur and sulfides will be oxidized all the way to the sulfates. When that occurs there is no electrical means of restoring or returning the sulfur component to the negative side in a usable form. That then provides one mechanism that will inexorably bring about the malfunctioning of a cell, and it can be corrected only by treating the electrolyte externally and with chemical means).

Again, it is possible to overcome these shortcomings, but only with the investment of sizable funding and testing. Non-aqueous electrolytes may indeed hold an answer to some of the problems.

Finally, we arrived at the concentration cell approach. This mechanism is entirely different than an electrochemical couple and appears to offer an answer to virtually all of the seemingly insurmountable issues encountered repeatedly with other electrochemical means. The concentration cell, which has been described in thorough detail throughout this book, was explored by the first author and associates beginning in 2002, and it is still being pursued.

The following are the main reasons for the optimism:

1. There are no electrically conductive solids deposited on electrodes to cause shorting or loss problems.
2. pH stays very steady, in the case of the sulfide system, at relatively high values.
3. The reactions are completely reversible.
4. Reagents are inexpensive and quite safe in normal usage.
5. Voltages and energy densities are reasonably high.
6. Reactions are well behaved with very few side reactions.
7. It operates very well in simple configurations with no necessity for the complexities of circulating electrolytes or compensating networks.

The hope now is that others will perform additional work to determine the ultimate possibilities of such an approach to practical energy storage for industrial and commercial applications.