DIA, 2010: estudo do AAWSAP avalia ultracapacitores para alimentar propulsão elétrica e armas de energia dirigida e sugere detectar suas descargas

UNCLASSIFIED//F8R [email protected] tJ!I!! eHt I 01 November 2010 !COD: 20 July 2010 DIA-08-1011 -005 Defense Intelligence Reference Document Defense Futures Ultracapacitors as Energy and Power Storage Devices for Commercial and Military Applications UNCLASSIFIED/f FQA &FFI&Iat.k ll6E 8Ptl¥

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I 1000 ....... ,...s C') ~ 100 Con~entmnal.c - battene s ~ 1 hour 1 second 10C' "cii C: _,I I_________10 hours (I) "C 1 I>, ...Cl I 0.03 second (I) 0.1 -- '---------C: w 0.01 .__ ______________ __, 10 100 1000 10000 Power density (W/kg) Figure 3. Ragone Plot. The graph illustrates the areas where ultracaps and batteries dominate in the power density versus energy density relationship. HISTORY The concept of storing electrical energy in the electric double layer that is formed at the interface between an electrolyte and a solid has been known since the late 1800s. In 1966, Standard Oil Company of Ohio (SOHIO) invented the device in the format now commonly used, which forms the basis for hundreds of patents and t housands of journal articles. Figure 4 shows a drawi ng from the electrochemical double-layer device invented by SOHIO. NEC introduced the SuperCapacitor™ in 1978 under license from SOHIO. CO UC. 0'2 Ole. SA . Alt(S04 3 50 10 lO PEQ.MEABLI:' Mct.162At-.lc- Figure 4. Electrolytic Capacitor as Designed and Patented by SOHIO. From this point on, ultracaps have rapidly evolved through several generat ions of designs. In the 1980s, Matsushita Electric Company developed a method of manufactur ing ultracapacitors with improved electrodes. Initia lly, they were used as backup power devices for volati le clock chips and complementa ry meta l-oxide­ semiconductor (CMOS) compute r memories . As the technology became more understood, more and more applications were developed for ultracaps. Many other applications have emerged over the past 30 years, including wire less communication, power quality, and improved energy efficiency through regenerat ive energy capture processes, as found in hybrid electric vehicles .10 UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 5

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I Initial ultracaps used aqueous electrolytes made with Ab(SO4)3 (aluminum sulfate), H2SO4 (sulfuric acid), or KOH (potassium hydroxide) . While these electrolytes have excellent ionic conductivity, the operating voltage is limited to 1.2 volts using carbon electrodes. Organic electrolytes have lower ionic conductivity, but the higher dielectric constant increases the nominal cell voltage up to as high as 3 volts. Carbon has been used as a high-surface-area electrode material since the inception of the electrochemical capacitor. It is still the material of choice for many ultracapacitors; one of the primary reasons is the low cost of carbon materials. However, there are many types of carbon that are available for use as an electrode material. In addition to carbonaceous electrodes, metal oxides and conductive polymers are find ing increasing use in ultracap design. Advancements in the understanding of the electric double-layer and ultracapacitor behavior have led to better materials utilization and, consequently, improved devices. Early electrochemical capacitors were rated at a few volts and had capacitance values measured from less than one farad up to several farads. Today cells range in size from small devices with exceptional pulse-power performance in the millifarad range up to devices rated at several kilofarads. There are even some specialized ultracapacitor cells now in production that have ratings of more than 100 kF. The technology is experiencing increasingly broader use, replacing batteries in some cases and in others complementing their performance. Ultracap technology has grown into an industry with sales of several hundred million dollars per year that is poised for rapid growth in the near term due to expansion of power quality needs and the emerging energy management/conservation applications. 11 Advancements in ultracapacitors have led to numerous devices from an array of manufacturers. Table 2 compares the various products on the market, showing voltage, capacity, power density, and additional energy-storage characteristics. Ultracaps have moved away from aqueous electrolytes and are typically organic electrolytes due to the increased voltage performance. Electrodes vary from carbon/carbon systems to hybrid systems using metal oxides or conductive polymers paired with a carbon electrode. Packaging and sizes of ultracapacitors covers a large range as these are now used from cellular communications and small electronics to power delivery and management for seaport cranes. Ultracapacitor technology development is focusing on delivering better energy density, and this is being approached by improved carbon electrodes, better electrolytes, and alternative electrodes that provide pseudocapacitive behavior, including battery-like electrodes. 12 UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 6

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I Table 2: Manufacturers of Ultracapacitors 13 Device V rated C (Fl R (mOhml RC (sec} Wh/kg (1) W/kg (95%) (2) W/kg Match. Imped. Wgt. (kal Vol. lit. Maxwell* 2.7 2 885 .375 1.08 4.2 994 8 836 .55 .414 Maxwell 2.7 605 .90 .55 2.35 1 139 9 597 .20 .211 Apowercap** 2.7 55 4 .22 5.5 5,695 50,625 .009 --- Aoowercap** 2.7 450 1.4 .58 5.89 2,574 24 595 .057 .045 Ness 2.7 1 800 .55 1.00 3.6 975 8 674 .38 .277 Ness 2.7 3 640 .30 1.10 4.2 928 8 010 .65 .514 Ness (cyl.) 2.7 3 160 .4 1.26 4.4 982 8 728 .522 .38 Asahi Glass (pc) 2.7 1,375 2.5 3.4 4.9 390 3,471 .210 (est.) .151 Panasonic (oc) 2.5 1 200 1.0 1.2 2.3 514 4 596 .34 .245 EPCOS 2.7 3 400 .45 1.5 4.3 760 6 750 .60 .48 LS Cable 2.8 3 200 .25 .80 3.7 1 400 12 400 .63 .47 BatScao 2.7 2 680 .20 .54 4.2 2 050 18 225 .50 .572 Power Sys. (AC, IPC) ** 2.7 1,350 1.5 2.0 4.9 650 5,785 .21 .151 Power Sys. (GC, pc)** 3.3 3.3 1,800 1 500 3.0 1.7 5.4 2.5 8.0 6.0 486 776 4,320 6 903 .21 .23 .15 .15 Fuji Heavy Industry- hybrid l(AC/GC) ** 3.8 1,800 1.5 2.6 9.2 1,025 10,375 .232 .143 JSR Micro (AC/GC)** 3.8 1,000 2 000 4 1.9 4 3.8 11.2 12.1 900 1 038 7,987 9 223 .113 .206 .073 .132 (1) Energy density at 400 W/kg constant power, Vrated - 1/2 Vrated (2) Power based on P=9/16 * (1-EF) * V2/R, EF=efficiency of discharge * Except where noted, all the devices use acetonitrile as the electrolyte (pc = carbonate) AC = Activated Carbon; GC = Graphitic Carbon ** Laminated oouches; all other devices are oackaaed in metal containers propylene UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 7

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Chapter 2: Materials Technology The standard electrochemica l capacitor is made of two electrodes, an electrolyte, and a separator, packaged in either a metal container or a laminated pouch. Ultracapacitors are often packaged in a fashion similar to a typical battery configuration. The separator is an ultrath in material that allows ion transport but is electrically insulating. There are several polymer separators on the market that have the desired porosity needed to create fast ion transport. The electrodes are traditionally the same material in an ultracapacitor-th is is in stark contrast to a battery. Electrodes are typically carbon; this is advantageous in making a cost-effective energy/power-storage device . Activated carbon has a very high surface area, vital to storing energy in the electrica l double layer between the electrode and the electrolyte. This th in double layer is respons ible for the impress ive high-power-storage capabilities of the ultracapacitor. The electrolyte's resistivity and dielectr ic constant play a vital role in behavior of the double layer; consequently, the choice of electrolyte impacts performance. The earliest electrochemical capacitors were introduced 30+ years ago; they were symmetric designs (two identical electrodes) in aqueous electrolyte . While these electrolytes have excellent ionic conductivity, the operating cell voltage was limited to ~1.2 V/cell and these had a nominal cell rating of ~0 .9 volts . In the second generation of electrochemical capacitors, the use of organic electrolyte led to an increase of the rated cell voltage from about 0.9 V/cell to 2.3-2.7 V/cell. Today, ultracapacitors using an organic electrolyte are the most popular. 14 ELECTROLYTES Initial ultracaps used aqueous electrolytes, saturated with Al2(SO4)3, or 30% concentrations of H2SO4 or KOH. While these electrolytes have excellent ionic conductivity, the operati ng voltage is limited to 1.2 volts using carbon electrodes. Improvements to ultracapacito r performance were made by trans itioning from aqueous electrolytes to an organic medium. Organic electrolytes have lower ionic conductivity, but the higher breakdown potential increases the nominal cell voltage. An increase in cell voltage up to as high as 3 volts has been realized by use of an organ ic electrolyte. These electrolytes are typically an ammonium salt dissolved in an organic solvent, such as propylene carbonate or aceton itrile. Table 3 highlights the comparison between the various electro lytes used in ultracapacitors. Ionic liquids are beginn ing to become more commo n in ultracapac itors , as they allow for an increase in cell voltage to as high as 4 volts. These ionic liquids have a higher r esistivity , especially at lower tempera tures. The tradeo ff in resistivity for volt age may be beneficial for some high-power applicati ons. Tab le 3: Properties of Various Electrolytes Used in Ultracapacito rs15 Electrolyte KOH Sulfuric acid Propylene carbona te Acetonitrile Ionic liquid Density (am/cm3) 1.29 1.2 1.2 0.78 1.3- 1.5 Resistivity (Ohm-cm) 1.9 1.3 5 52 18 125 (25° C) 28 (100°C) Cell Voltaae 1.0 1.0 2.5-3.0 2.5-3 .0 4.0 3.25 UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 8

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I ELECTRODES Carbon has been the optimal electrode mater ial for ultracapacitors since their commercial introduction over 30 years ago. Pseudocapacitor systems use a comb ination of a carbon electrode and a battery-like electrode, such as a conductive polymer or a metal oxide. The hig h surface area of carbon makes it an extremely attractive option for ultracap electrodes . Carbon chemistry is quite well known, and high surface area and low cost ensures carbo n's cont inued use for the near future. While carbon has been used in electrochemica l systems for many years, there are many nuances that make its properties and performance vary significantly amongst the many types of carbon. The use of a carbon electrode makes ult racaps affordab le for use in many systems. Activated carbo n is readily available, can be made from many source materia ls, and is inexpens ive. Materials science activities for electrodes are focused on improving the carbon electrodes and i ntroduci ng alterna tive materia ls to create an asymmetric electrode. Types of carbon can include activated carbon, carbon cloth, aerogels, porous carbon, carbon nanotubes, and graphene. The properties of carbon can change dramatically based upon processing, having a major im pact to the porosity and active surface area. Activated carbon has an extremely high surface area, is inexpensive, and is produced at a global scale for use in a number of applications. For these reasons, activated carbon is the traditional carbon of choice for ultracapacitor applications. Recent advances in carbon materials development have led to a number of options for ultracapacitor electrodes. Table 4 demonstrates a number of electrode materials in use today and t he corresponding performance of these ultracaps. Table 4: Properties of Various Materials Used in Electrochemical Capacitor Electrode Materials 16 Material Density (g/cm 3 ) Electrolyte F/g F/cm 3 Activated Carbon 0.7 KOH 160 112 Oroanic 100 70 Carbon Cloth 0.35 KOH 200 70 Organic 100 35 Aerogel Carbon 0.6 KOH Orqanic 75 125 84 Porous Carbon from SiC 0.7 KOH 175 122 Oroanic 100 70 Porous Carbon from TiC 0.5 KOH 220 110 Organic 120 60 Anhydrous RuO2 2.7 Sulfuric Acid 150 405 Hvdrous RuO2 2.0 Sulfuric Acid 650 1 300 Doped Conductive Polvmer 0.7 Orqanic 450 315 UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 9

UNCLASSIFIED1fFOR 8ffl@IAL tt.!! e"t I Activated carbons are synthesized by a heat treatme nt of carbon-r ich organ ic precursors in a controlled atmosphere. This carbonization process can be performed from natura l sources, such as fru it shells, wood, pitch, or coke. Activated carbons can also be produced from synthetic precursors, such as select polymers. The heat treatment process uses a controlled partial oxidation of the precursor combined with a high-temperature processing. The high-temperature processing can be done in an inert atmosp here, an oxidizing environment, or with a chemical modificat ion. The activated carbons contain a distributed porous network throughout, as shown in Figure 5 .17 These materials can have an extremely high surface area, upward of 3,000 m2/g, but process conditions to create activated carbons have litt le control over pore size distribution, result ing in incomplete utilizat ion of the surface area. Carbon cloth electrodes use processing conditions similar to those used to create activated carbons . The advantage with a fabric composed of activated carbon is these materials are directly used as active electrodes, requiring no binder or additional processing. However, these tend to be expensive to produce, which has thus far limited their use to specialized applications. rnacroporcs >50 run ~ -1-- mi r pre <2nm mesopore 2-50nm Figure 5. Activated Carbon Porosity. These can be produced with a distribut ion of pore sizes ranging from less than 2 nm to greater than SO nm. Continued improvements to the carbon electrodes are expected to provide additional performance enhancement to ultracaps. Current research is leading toward carbon materials with higher specific capacitance (F/g). Improved materials, such as carbon nanotubes and tailored porous carbons show promise as the next generation of carbon materials. Graphene is another advanced carbon material that shows promise to providing increased capacitance. One of the key design issues revolves around improved understanding of the relationship between the carbon pore size and the electrolyte ion. The nanostructured materials allow fine tuning of the porous structure to increase capacitance. Carbon nanotubes (CNTs) are commonly produced as powders where they can be cast as a distributed network to form an electrode, and they can be used as an additive to increase conductivity. Additionally, CNTs can be grown as a "forest," where the CNTs are grow n perpendicu lar to a substrate. These forests can be used as is or can be modified or coated to change specific structural or electronic properties. CNTs are grown by a number of methods, but most methods used currently are a derivative of a chemical vapor deposition (CVD) process. The specific propert ies of the nanotubes are extremely tailorable. CNT length, diameter, the number of tubes, and the electronic properties can all be tailored by growth conditions. The catalysts used to grow CNTs can affect tube properties, as well as CNT tempe rature and precursor variables. Carbon nanotubes are finding util ity in both battery and ultracapacitor applications, where they are used for conductivity enhancement and energy storage. Their capacitive behavior suggests they may have a fundamentally different storage mechanism than traditional carbon. Storage potential includes CNT outer walls, inner diameter, and interwall spaces UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 10

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I of some CNT types. Carbon nanotubes have been a rich area of study for many applications, which has translated to a broad understanding of synthesis techniques to produce CNTs with tailorable properties. Graphene can be thought of as a sheet of graphite, either a single layer or a few layers thick. The two-dimensional nature of graphene leads to extremely high electron transport across the surface, interesting magnetic properties, high strength, and a large accessible surface area.18 Graphene is generally made by exfoliation/separation of graphite or grown by CVD processes. Derivations from both of these methods can produce single-layer graphene or few-layer graphene. These methods include, but are not limited to, CVD methods, epitaxia l growth, solvothermal synthesis, micromechanica l exfoliation, and colloidal synthesis. 19 Graphene has generated an enormous amount of attention due to its potential for transforming electronics. A number of approaches to graphene synthesis are being used that consider purity, scalability, and cost. Carbide-derived carbons are a unique class of porous carbons with extremely fine control over pore size. 20 These porous carbons are derived by chlorination of metal carbides at very high temperatures. At temperatures of 500-1000°C, metals and metallo ids are removed as chlorides, leaving behind a finely tunable nanoporous carbon. These porous carbons have a pore-size distribution that is tunable within 0.05 nm. The extremely accurate tunability has allowed these systems to serve as a model for the relationsh ip of pore size and capacitance, thereby allowing for design of materials to be tailored for maximum capacitance within a specific system. The local maximum in capacitance is dependent on the ion salvation sphere, which is a product of the electrolyte ion, solvent, and interaction with the pores of the electrode. These findings will lead to new electrode materia ls and designs. Activated carbon and other porous carbon derivatives are commonly used for symmetric ultracapacitors. Metal oxides such as ruthenium oxide and conductive polymers replace one of the electrodes to create an asymmetric system. Asymmetric electrodes greatly increase the storage capacity of ultracapacitors. These are a different class of capacitor, as they store their charge both in the electrica l double layer and using a surface redox (faradaic) reaction (in the same manner as a battery). While these undergo electron transfer reactions, they behave in a capacitive fashion. Often termed pseudocapacitor, this class of materials includes a variety of metal oxides and conductive polymers. The capacitance of an asymmetric electrode is twice that of a symmetric design. The electron transfer electrode essentially has a fixed potential, whereas the potential of the carbon electrode changes with state of charge. Additional storage capacity arises from the higher working voltage, due to the different rest potentials of the different electrodes. Asymmetric electrodes are becoming increasingly common due to the increased energy­ storage capability. As energy is proportional to the voltage squared, the higher operating voltage has profound effects on the capacity. Ruthenium oxide (RuO2) has been shown to give capacitance of as high as 1300 F/cm3• By comparison, carbonaceous materials in an aqueous electrolyte have a capacitance of 110-125 F/cm 3. There are many additional metal oxides being investigated for insertion into the ultracapacitor system. Many of these have their roots in lithium ion battery cathode materials. As ultracaps gain wider acceptance, market size is increasing rapidly and is expected to grow at an increasing pace. Investigations are being conducted into alternative metal oxide systems to ensure that cost, supply, and performance can be met on a global scale. Manganese oxide (MnO2) and nickel metal oxides/hydroxides are UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 11

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I leading alternative candidates for asymmetric ultracaps. Using a pseudocapacitive design will most likely not be able to deliver the cycle life of a symmetric ultracap, but it may be possible to achieve tens of thousands of cycles out of a battery-like electrode. Better capacitance and higher operating voltage of the asymmetric electrode design allow ultracapacitors to begin approaching batteries in terms of energy density. Additional ultracapacitor improvements will come from materials and cell design. Cell design factors include optimizat ion and assembly techniques. Thinner electrodes and current collectors will be one of the primary paths for better performance. Electrolytes that provide cell voltages of at least 3 volts will be standard. Improvements such as these will likely lead to improvements in commercially available ultracaps from 4 Wh/kg to 5-6 Wh/kg. 21 UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 12

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Chapter 3: Applications ELECTRONICS AND TELECOMMUNICATIONS The most straightforward application of an ultracapacitor is to stabilize de voltages, which is where they have found widespread use. Areas where a dip in voltage may occur and negatively impact performance is another natural fit for ultracaps. Initially, they were used as backup power for clock chips and CMOS memory; now, they find utility in power smoothing, flash photography, and digital cameras. Flash for mobile phone cameras is an ever-increasing market. Nearly one billion phones equipped with cameras were sold in 2009. Ultracaps enable flash photography integrated with mobile communication. Today's market is driving toward electronics devices with multiple functions to eliminate the need for multiple specialized electronics that each performs one task. Integrating components into one electronic device requires the use of ultracapacitors to help stabilize the short power spikes from the individual components. Ultracaps also serve to reduce thermal loading from power dissipation. Ultracaps have found extensive application in wireless communication, where they are used in wireless data cards for GSM, GPRS, and WiMAX data transmission. The peak current required during data transmission can exceed what is available under USB or PC-card standards, and the excess current is delivered with an ultracapacitor. Automated meter reading is another application where ultracaps have become an integral component, as the power required to send and receive the meter informa tion is handled by an ultracapacitor. The mobile telecomm industry is currently using these in a number of ways; it is likely that additional uses for ultracaps will spring up along with needs to balance power requirements for increasingly complex and multifunctional devices contained within smaller areas. INDUSTRIAL Industrial applications are becoming much more ubiquitous as ultracaps are being used as a platform to deliver a large amount of power quickly. These are becoming increasingly important in cranes, forklifts, elevators, and power tools. As an example, power is stored by an ultracapacitor in hybrid forklifts every time a loading fork descends. The stored power is then delivered to the forklift when heavy lifting is required. Figure 6 shows the power requirements for a hybrid forklift. The peak power pulses are areas where using an ultracap to deliver or to receive the power greatly improves the operational efficiency. The blue shaded area represents the power distribution handled by a fuel cell, and the tan regio ns above and below the blue area represents the output from and the input to the ultracapacitor, respectively. 22 In cranes and elevators, a similar application is used. Peak power is supplied for lift operations by the ultracap. During descent of the elevator or lowering of the loaded crane, generated power can be used to recharge the ultracapacitor. Power tools are becoming increasingly disconnected from an outlet. Many of these cordless power tools require significant amounts of peak power, which many batteries cannot adequately supply. The ultracapacitor used in conjunction with a high energy battery supplies the pulse power required during heavy use. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 13

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Ult truck opercrtl ng profile 35 ~ ::.::.-=--=--=--=--=---------- ~~===--=--~ f ~ JJI - - -'~ ~~~ J [~~ J1L J! l 0 Fuel-<.ell output 5 0 t--""ffltTft''Mt---itlr--"'fflltM----,-m-rimr,,,_"'""rrm--tt-,-i - 5 - 100 l 2 Scuru : G IHydrogen lime (hr) Figure 6. Ultracapacitor Use in a Forklift. Ultracaps provide energy storage upon descent of a load and deliver peak power to lift heavy loads in fuel cell powered forklifts. 23 Grid storage management is gaining more visibility as energy conservation and efficiency are becoming increasingly important. Many utility power sources generate a relatively continuous supply of electr icity. The generated power is matched to expected use statistics. Daytime use and nighttime rest periods can cause power shortages or generate unused power if not managed efficiently. This is especially true during the hot summer months or heat waves, where air conditioning loads can tax a system. Ultracapacitors can store excess power generated in the night, which can then be delivered when insuffic ient power is available to match the daytime needs. Distributed power sources such as wind and solar require a higher level of operational efficiency with a fluctuating power supply. Wind farms experience power fluctuations due to wind shear or lack thereof. These spikes in power generation may be wasted on a system ill­ equipped to handle the burst of power delivered. During periods where wind may be lacking, ultracapacitors can be used to help create a more stabilized power supply. TRANSPORTATION Earliest uses of ultracapacitors were in motor startup for tanks and submarines. Early adoption by the military led to cost reductions, which was a critical enabler for incorporation into diesel trucks and railroad locomotives . Ultracapacitors are find ing use in transportation in engine startup, braking systems, acceleration, and waste-energy harvest ing. Ultracaps have been placed in rail cars to generate power for acceleration and recapture it during braking. In some cases, they are placed alongside the tracks, and t he power management is hosted by the rail stations. The ability to absorb and discharge energy rapidly makes ultracapacitors far better than batteries for regenerative braking schemes. Most of these applications have been in public transportation. Regenerative braking can recuperate as much as 38% of the propulsion energy when used in an ultracap/battery hybrid configuration. Electric hybrid public buses were some of the first adopters of ult racapacitors for automotive transportation. These buses also use ult racaps for acceleration, which enables a much faster acceleration. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 14

UNCLASSIFIED//F8R 8Ff!C!Jlct tt:!!! tJHL I Ultracapacitors as Energy and Power Storage Devices for Commercial and Military Applications The Defense Intelligence Referen ce Document provides nonsubstantive but authoritative reference information related to intelli ence to ics or methodolo ies. Prepared by: Technology Warning Division (DW0-4 ) Defense Warning Office Directorate for Analysis Defense Intelligence Agency Authors : AAP Person 84, AAP Person 85 (U) COPYRIGHT WARNING: Further dissemination of the photographs in this publication is not authorized. This product is one of a series of advanced technology reports produced in FY 2010 under the Defense Intelligence Agency, Defense Warning Office's Advanced Aerospace Weapons System Applications (AAWSA) Program. Comments or questions perta ining to this document should be addressed to !MP Person 1 IAAWSA Program Manager, Defense Inte lligence Agency, ATTN: JUIAF - DI/DWO -3, Bldg 6000, Washington D.C. 20340-5100 UNCLASSIFIED,<fiOA OFFI&l..tJk WSE e••tv

UNCLASSIFIED1/FOR 8ffl@IAL tt.!! 8flt I AEROSPACE Ultracapacitors are suitable for applications requ iring large bursts of power for relatively short peri ods of time, with the ability to be recharged rapidly. Ultracapacitors are find ing ubiqu itous use in electronics and transportation, and there are a number of applications in aerospace ranging from backup power to high pulse-power delivery. The electronics systems in milita ry vehicles requi re failsafe operation; ultracaps can be designed to provide a continuous power source in the event of a power interruption. Fire-control systems are another example where continuous power is requ ired, including higher power capability. Airbags require a backup power source to ensure they are deployed as requ ired with or without a powered system. Small portab le communications systems have ultracaps for backup power and to reta in memo ry. Memory retention is critical in many systems; for example, the bridge power used during a transfer from ground to onboard power in aircraft systems can be supplied by an ultracap. Ultracaps are excellent for stabilizing the power that is either delivered or received. For example, communication systems may requ ire additional power during transmission, which can be offset by using an ultracap to supplement the system. Bus voltages may need peak current stabilization. The high-power capabilities of ultracaps make them advantageous for severa l systems . The fast discharge is ideal for delivering the significantly higher power required for starting cold engines. Active suspension systems in vehicles require bursts of power that are delivered in response to the terra in; an ultracap can store the necessary power, deliver as necessary, and recharge quickly. GPS-guided missiles and projectiles may use ultracapacitors to provide the power required for commun ication and navigation. High-power discharge for naval systems has been identified as an area where ultracaps are especially useful, often used in hybrid systems in tandem with high-energy batteries. Extremely high power delivery makes ultracapacitors a potential veh icle for delivering the power required for directed-energy weapons. As pseudocapacitive systems reach higher energy densities, along with improvements to the materials and cell composition, ultracapacitors will have capabilities of delivering large amounts of power in a very short time, conta ined within an increasingly smaller footprint. Signatures and Vulnerabilities As mentioned above, ult racapacitors could be impo rtant in the development of ult rawideband energy weapons. Previously, this market has been dominated by bulk acoustic semiconductor switch (BASS) devices, but by stringing several capacitors together in parallel, a more formidable discharge can be created. Because ultracapacitors have very high discharge rates, they will emit a distinctive broadband RF signature when discharged. A modest wideband receiver may be able to detect this discharge; its proxim ity will depend upon the construction of the target device . Furthermore, the architectu re of such a device may yield a unique signatu re that, with the proper equipment, can be identified and certa inly warrants further study. When employed in a particular electronic device, this signature may be altered by its immed iate envi ronment, and this may further give an indication of both the type of use and device employed. When searching for ultracaps in the field that are not i n use, the large electrostat ic potentials may provide a clue. UNCLASSIFIEO/fFOA OFFl€il.t.k YSE O,.Llf 15

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I Stimulated Discharge Depending on the application and the implementation and construction of ultracaps in sensitive or destructive applications, vulnerabilities likely exist and warrant study in areas of research such as circuit vulnerability analysis, reversible manufacturing and construction techniques used in fabrication, stimulation of nondestructive and controlled discharge evaluations, and laboratorY: stimulated discharge experiments. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 16

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Chapter 4: Recent Developments There have been several recent advances in ultracapacitors focused on increasing energy density by using different active materials 24 and designs, 25 activated carbons, 26 polymers, 27 and metal oxides. 28 Conducting polymers have good performance, 29 but lack microfabrication protocols. Oxide-based thin-fi lm ultracaps have shown high performance, approaching the theoretical limit for capacitance ( ~1,000 F/g for MnO2). 30 However, the poor electrical conductivity and high impedances associated with surface intercalation redox reactions of these oxides have limited practical film thicknesses to a few microns. Carbon nanotubes have been added to the films to increase the electrical conductivity, but the complexity of manufacturing limits practical applications of such composite electrodes. In th is section we will discuss advances in application of CNT and advanced carbons to ultracapacitor materials, the application of thin-film manufacturing processes, and a revolutionary new super capacitor using a giant magneto capacitive effect. CNTS AND ADVANCED CARBONS Carbon electrodes constitute both electrodes in a symmetric ultracapacitor and one of the electrodes in an asymmetric, or pseudocapacitive, design. Improvements to the carbon electrode rely upon increasing the specific capacitance (in Farads per gram). These improvements come by tailoring the surface area and porosity to achieve the best balance that maximizes the interaction with the electrolyte. There is a linear relat ionship to the surface area and the capacitance up to a point where capacitance plateaus with activated carbons. By controlling the porosity and surface area, it is possible to increase the capacitance beyond this plateau. Carbon nanotubes could provide performance increases with aligned CNT forests of tailored sizes. The characteristics of an ultracapacitor are highly dependent on the nanostructure of the carbon used for the thin-film electrodes. Advanced carbons will provide better control over the pore size and distribut ion, leading to an expected 50- to 100-percent improvement over the carbons in use today . Carbon nanotubes can be produced with a wide variety of properties. Depending on synthesis parameters, nanotubes can be single walled or multiwalled, with varying numbers of tubes. CNT diameters can be tailored from a few nanometers to tens of nanometers, with lengths up to hundreds of microns. CNTs can be grown in random orientations or as aligned forests . CNTs have a fully accessible surface area and very high electrical conductivity. Methods for incorporating CNTs into electrodes for ultracaps include using CNTs as an additive for conductivity enhancement, creating dense mats of randomly oriented tubes, and creating electrodes from vertically aligned forests of tubes . Initial results of CNT-enable d ultracaps tended to show much lower capacitance than expected, which has been attributed to the hydrophobic nature of the CNT walls. Surface functionalization is a common approach to mitigating the hydrop hobicity issues and thus enabling higher capacitance. Another benefit of the functionalization is the ability to introduce and control pseudocapacitance. Most efforts in CNT ultracaps are directed toward vertica lly aligned forests. It is possible to controllably grow a dense, aligned forest that is perpendicular to the current collector. The size and density of the tubes and the number of walls can be controlled with catalyst design and reaction parameters. Manipulation of the CNT forest leads to increased capacitance by fine -tuning the distance between tubes. Additionally, fine- UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 17

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I tuning the inner diameter may provide further enhancements. Figure 7 shows a vertically aligned CNT forest produced at Lockheed Martin's Advanced Technology Center (LMATC). Lockheed Martin has worked with a number of catalysts and surfaces to grow tailored CNT forests for a variety of applications . Select locations throughout Lockheed Martin have been developing CNT-based technologies, utilizing vertically aligned CNTs, CNTs dispersed onto surfaces, and CNTs dispersed into other media. LMATC possesses expertise in CNT growth processes and characterization and has capabilities for both materials development and ultracap testing. Lockheed Martin's NEARLab facility produces CNT-coated glass fibers, which may provide a cost-effective power storage solution that has structural elements built in. Ultracaps assembled from CNT forests appear extremely promising for use in microelectron ics. Figure 7. CNT Forest. CNT forests grown in Lockheed Martin's laboratories can be tailored for specific sizes, lengths, and densities. Graphene is a relatively new discovery amongst carbonaceous materials. There are a number of types of graphene that can be characterized by the number of layers of graphene, the functionalization, or the oxidation status. Most graphene for ultracapacitor applications is going to be few-layer graphene and large-area flakes. The fewer the layers, the higher the active surface area will be. Single layer is ideal, but manufacturing considerations make single layer difficu lt, even at the laboratory scale. Functionalization will be directed toward improv ing capacitance or modifications to enable battery-like performance. Theoretical values of graphene indicate it could become an important material for the next generation of ultracapacitors. Graphene may be used as the sole electrode material, or it could be used as a conductive additive that also provides capacitance. The surface area is calculated to be as high as 2,600 m2/g, the thermal conductivity is 5,000 W/m·K, and the charge carrier mobility is 200,000 cm 2/V•s. High surface area values and great conductivity are ideal properties for creating an electrode with very high capacitance and extremely favorable rate capabilities. Reported capacitances range from 135 to 205 F/g in aqueous electrolytes. 31• 32 Figure 8 shows a transmission electron microscope image of graphene flakes used for ultracapacitor electrodes. These measurements come from few-layer graphene, rat her than single layer, which suggest UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 18

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I there is room for improvement on these values. Optimization of graphene synthesis to produce finer control is an area receiving increased attention in research and development activities in unive rsities, national labs, and industria l research facilities. Lockheed Martin is actively engaged in research investigating graphene and applications where graphene could be a critical differentiator. There are a number of companies that supply graphene made from various methods, as well as emerging startups geared toward producing graphene-based ultracapacitors or materials designed for ultracaps. Figure 8. Transmission Electron Microscope Image of Graphene. This could be an excellent ultracapac itor electrode materia l with its high surface area and excellent conductivity. 33 Manipulation of the porosity of high surface area carbons leads to the largest difference in specific capacitance . Understanding the relationship between the electrolyte ion size and the carbon pore size is critical to improving performance. There are a number of strategies being investigated for fine control over the pore-size distribution to increase the specific capacitance. The most common methods used currently are template methods and carbide-derived carbons. Template methods create a controlled mesoporous structure with a fairly narrow range. These structures have pores that range from 2 to 10 nanometers and are maximized to pore sizes roughly twice that of the solvated ions. The template process involves filling the pores of an inorganic template host with a carbon precursor (such as an alumina template). The template is removed after carbonization by acid treatment . The pore size is then dictated by the template as the pores are the remaining void space once occupied by the template . Simi lar methods have shown that smaller pores, including those less than two nanometers, may provide high specific capacitance. The realizati on t hat smaller pores contribute to charge storage in an electric double layer has led to the need to develop a better understanding of the charge storage mechanism. Carbide-derived carbons have a unique pore-size distribution that is tunable with sub­ angstrom accuracy. These have served as models to study the charge storage behavior and ion adsorption in pore sizes ranging from 0.6 nm to 1.1 nm. 34 The normalized capacitance decreases with decreasing pore size unt il a critical value is reached. Figure 9 shows t he relationship between average pore size and the normalized specific capacitance. Pore sizes smaller than one nanometer significantly contribute to the charge storage despite the fact that the solvated ion size is larger than the pore diameter. The capacitance increase is explained by a distorted ion shell model. The ion salvation shell is perturbed such that it is capable of a closer approach of the ion and the carbon surface. The discoveries at Drexel University that utilize the fine control to create tailored porous carbon structures can maximize specific capacitance for a given ult racapacitor system. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 19

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I 15 Ill II 't ~-Y °! 10 (Cll,Cfl1),N• 0.68rvnclame. I Is 0 0 L--+--1----= ======= :'.....J .______,... 0 3 s Aveqge pore Size {Ml) Figure 9. Relationship Between Average Pore Size and Normalized Specific Capacitance. As the average pore size decreases below 1 nm, the specific capacitance increases due to distorted electrolyte Ion salvat ion. THIN FILMS CNT films have been used in 2D thin-film ultracapacitors having respectable gravimetric and volumetric capacitance. Activated carbon powders have a better performance in terms of energy per unit area, 35 but they need to be processed into films before use, and th is lim its their choice for microdevice fabrication, which allows mass production of capacitors on semiconductor wafers. Carbide-derived carbo n (CDC) is a class of carbon materials produced by selectively etching metals from meta l carbides using chlorine at elevated temperatures in a process similar to current dry-etching techniques used in MEMS and microchip fabrication. CDC has been shown to have excellent performance as the active material in traditionally processed ultracaps 36 since it can have its microstructure precisely tuned by tailoring the synthesis conditions for a particular electrolyte. 37 For microfabricated supercapacitors, CDC is attractive for several reasons. The precursor carbides are conductive and can be deposited in uniform t hin and th ick films by well-known chemical and physical vapor deposition (CVD and PVD) techniques. 38 I n addition, the chlor ination process can be performed at tempe ratures at least as low as 200°C, 39 and the resulting coatings are well-adhered with an atomically perfect interface, 40 which minimizes device impedance. This technology can be used to produce the microfabricated ultracaps on the same chip as the integrated circuits, which they are powering (shown in Figure 10).41 Chlorine-containing plasma etching of materials in semiconductor manufacturing is a well-established technique and is similar to the chlorination procedure in CDC manufacturing. Continuous porous carbon films cannot be produced by conventional CVD, PVD, or other techniques, and the high-temperature activation needed to produce the microstructures necessary for ultracapacitor performance in CVD carbons would destroy the devices they were intended to power. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 20

UNCLASSIFIED1/FOR 8ffl@IAL tt.!! 8flt I Sintered TIC plate TIC plate - CDC film D Teflon plates \ - Electrolyte + eparato r C-CDC film 2. C II a embly Figure 10. CDC Synthesis and Electrochemical Test Cell Preparation Schematic. 42 Electrochemical measurements of CDC films were carried out in a three-electrode configuration with a large, overcapacitive activated carbon counter-electrode in both lM TEABF4 and lM H2SO4, as well as two-electrode cells with symmetric bulk CDC film electrodes. Volumetric capacitance was calculated for each of the different film thick nesses (shown in Figure 11). For both TEABF4 and H2SO4, the volumetric capacitance decreases with increasing coating t hickness. This is especia lly pronounced in the organic electro lyte, where there is a huge increase in volumetr ic capacitance as the film thickness decreases from 200 to ~2 mm. A 180 6j' E 160 .!:? ~ 140 Q) 0 C: 120l!! -~ 100 a. ca 0 80 0 ·c: '1ii 60 E :::, g 40 • 20 B 180 r;, E 160 ~ 140 ¢1 0 C: 120 19 'ij 100 ~ <11 0 80 0 ~ 60 •E :::, g 40 20 0 50 100 150 200 250 0 20 40 60 80 100 120 140 Coating thickness (µm) Coaling thickness (µm) Figure 11. Volumetric Capacitance of the Films in (A) TEABF4 and ( B) H2S0 4.43 MAGNETIC CAPACITORS Giant magneto capacitance (GMC) is a quantum mechanical effect observed in t hin-fi lm structures composed of alternating ferromagnetic and nonmagnet ic layers. 44• 45 Such materials have been reported to have increased permittivity by 109 .46 This material behavio r n orma lly occurs at tempe ratures well below 273K and has only recently been reported in the literature above this temperature.47 However, since 2007, Northern Lights Semiconductor Corp. {NLSC) has been developing a magnetic capacitor {MCap) that utilizes a material which exhibits the GMC above 300K. Structured like the basic flat-plate capacitor, an MCap nanocell featu res a proprietary dielectric layer sandwiched between two magnetized layers. Each magnetic material layer is made of multip le nanomete r-thick thi n-film layers. Millions of MCap nanocells are created and linked using a semiconductor thin-fi lm process, producing the MCap cell on a wafer {Figure 12). UNCLASSIFIEO/fFOA OFFl€il.t.k YSE O,.Llf 21

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I r------- / MCap nanocell / /1.6µm o.55µmJ• •I / / Magnetic materials "'i / / Dielectric layer -- / MCap cell Magnetic materials ,----,•...____ / Made up of millions (all proprietary compositions) / / MCap nanocells ~- - - - - - - - ../ Figure 12. MCap Structure. MCap nanocell and MCap cell are fabricated using traditional semiconductor fabrication processes . MCap cells are then packaged into larger MCap modules. The capacitance of the device is given by the equation , C = so k A/d, where A is the area of the plate, d is the separation between plates, so the permittivity of free space, and k is t he dielectric constant (or relative permittivity) of the material. The MCap increases capacitance by increasing the dielectric constant k through the GMC effect (fGMc) given by t he relationship, k' = k*fGMc- GMC acts like a charge trap that brings electrons closer, thus increasing electron densities at the plates. Based on quantum theory, GMC brings about a capacitance which, to date, has been measured to be 109 times larger than that observed in electrostatic capacitors. As shown earlier in equation (2), energy is proportional to the capacitance and the voltage squared. Moreover, capacitor leakage and self-discharge are essentially eliminated as electrons are "trapped" in the magnetic field. Due to the above effects, the available energy for a large range of storage devices, like smart cards and other products, can be increased substantially by simply packaging them in series-parallel connected modules to meet the energy and voltage requirements. Coupled with Lockheed Martin's extensive nanomaterials and device physics experience, it is likely that NLSC will be able to use its MRAM development expertise to accomplish this. Exper imental Results Over the last year, NLSC has had MCap structures and nanocells tested by independent parties with results shown in Figure 13. These unbiased tests demonstrate the validity of the high-capac itance claims. Figure 14 quantifies the progress in improved performance throughout the development of the device. The number of devices that demonstrate the GMC behavior has increased substant ially in 2009, which is reportedly due to improved materials development rather than an increased production load. The performances of the nanocells as well as the variability are plotted, which shows that potent ial improvement in the materials could lead to an increased GMC factor. There is promising evidence that improvements may allow a higher level of performance, which could match lit hium ion batter ies at ambient temperature and perhaps exceed them at low and high temperatures. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 22

------------------------ UNCLASSIFIED1/FOR 8ffl@IAL tt.!! 8flt I 4284A-pad 01, Cp0.20F -+- AClVrms,0\/dc 180000 +---il-+------ ---1 ....,. ACtVnns, 1Vdc 160000 +--"'-+--------1 --+--AC.tVrms,3Vdc 140000 G:' 120000 2. 100000 Q. U 80000 60000 40000 20000 0 AC.tVnns,5Vdc Frequency vs Glpacitance 5.0 4.5 1.0 0.5 0.0 a CSM mV !Vn,/ tn 4M - No.l._41/\..W - No.1_41.A_IV - - ~. 1_411\._lV - No.1_41A_SV \, -- No.1_41A....10V - No.1_411,Jllll - No,1 41A41N -\ "- ~~ \. ~ I 1L I . 0.1 10 100 1000 10000 0.01 0.1 10 100 Frequency [kHz) Frequency [k I Figure 13. Preliminary MCap Test Results. These graphs demonstrate NLSC capabilities: MCap microstructure tested at differe nt VDC (by Taiyo Yuden) (left) and sing le nanocell tests (right). 350 (/) a> .... 300 C) .Q o> fal250 C Q) o...O ~~200 ~CJ -0 g,150.._ C Q) ·- .0 s: §_g100 z (/) 50 - - ----- - ---- ----- -- ----- --------- ► Maximum GMC attained Replace Li- _______ __ _____________ __ _ • ion battery Replace HV Capacitor / Ultracapacito Few Mcap nanocells with MC due to processing & equipment issues t 1 E+12 ~ >.c 1 E+ 11 g·o - 0.. 1 E+10 g c .c C'Cl 1 E+09 ~ :a ..... Q) 1E+08 0 ~ -0 1 E+07 a> c:5 C'Cl 1 E+06 ~ ~ Q)·- 1 E+05 E .s 1 E+04 eel~ I I 1 E+03 ~ ~ 1E+O~~ 1E+01 1 E+00 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2007 2008 2009 Figure 14. Summary Chart of MCap Results to Date. Right Axis: High-Low value lines in red, Median value in black dot. Left Axis : number of nanocells exhibiting GMC phenomenon in blue bar . Figure 15 shows an estimate that has been made by NLSC of the potential specific energies and projected manufacturing costs of the MCap for compar ison to current energy storage devices, UNCLASSIFIEO/fFOA OFFl€il.t.k YSE O,.Llf 23

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I 10,000 00 $10 Li-ionprice 0.1 0.01 Pb-Acid price .001 Fuel cell U-ion Mcap cap Attainable ojected Energy Density ominal Cost ($/Wh) 1, 000.00 100.00 10.00 1.00 0.10 0.01 (Whig) 0.00 I I .. • • ... I I=117 ... ... odq, 10 rte r 1e1 ' ~ ' , I 1 II • .j,•. ➔• I rn m e. #, p I I I Jrtial1 I Ii I p i II ~ 2~ :r: 2= II ~ ~:j; .. ~ I ~ n $ $ $ $0 M Pr N 1E+06 1E+07 1E+08 1E+09 1E+10 1E+11 fGMC Figure 15. Specific Energy and Estimated Costs vs. foMc The mass production target of MCaps on a wafer of fo Mc of ~10 9 would enable the achievement of exceeding t he specific energy of lithium ion batteries at a significant ly reduced cost . MCaps at t hese prod uction targets would disrupt both the lithi um ion battery and the ultracapacitor market . UNCLASSIFIEO/fF OA OFFI &I:.l.b YS& ONLY 24

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I Contents Summary ................................................................................................................ iv Chapter 1: Concept Overview ................................................................................ 1 Batteries and Ultracaps ...................................................................................... 3 History ............................................................................................................... 5 Chapter 2: Materials Technology ........................................................................... 8 Electrolytes ........................................................................................................ 8 Electrodes ............................................................................................................................................................................ 9 Chapter 3: Applications ....................................................................................... 13 Electronics and Telecommunications ................................................................ 13 Industrial ......................................................................................................... 13 Transportation ................................................................................................. 14 Aerospace ........................................................................................................ 15 Chapter 4: Recent Developments ........................................................................ 17 CNTs and Advanced Carbons ............................................................................ 17 Thin Films ........................................................................................................ 20 Magnetic Capacitors ......................................................................................... 21 Chapter 5: Future Developments ......................................................................... 25 Chapter 6: Conclusions ........................................................................................ 27 Chapter 7: Endnotes ............................................................................................ 28 Figures Figure 1. Ultracaps in Various Configurations . ...................................................... 1 Figure 2. Operation of an Ultracapacitor ................................................................ 3 Figure 3. Ragone Plot ............................................................................................. 5 Figure 4. Electrolytic Capacitor as Designed and Patented by SOHIO ..................... 5 Figure 5. Activated Carbon Porosity ..................................................................... 10 Figure 6. Ultracapacitor Use in a Forklift ................................................................ 14 Figure 7. CNT Forest ................................................................................................ 18 Figure 8. Transmission Electron Microscope Image of Graphene ......................... 19 Figure 9. Relationship Between Average Pore Size and Normalized Specific Capacitance ............................................................................................................. 20 UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY ii

UNCLASSIFIED1/FOR 8ffl@IAL tt.!! 8flt I Chapter 5: Future Developments There have been significant advancements in ultracapacitor technology in the last few years. Increased understanding of the physical mechanism behind th is technology, combined wit h advancements in mate rials science, part icularly on the nanoscale, has led to a rapid increase in capability. Improvements to ultracaps in the next 10 years (2010-20) will focus on electrodes, better electrolytes, packaging, and alternative designs. Electrode improvements will most likely include migration away from activated carbon, for both electrodes. Asymmetric electrodes will eliminate carbon on one side, and new carbon materials will provide better performance at a competitive price . Electrolytes will likely move toward additional organic materials and ultracaps with high temperatu re performance requirements likely will use ionic liquids. In the near term, a transition to the asymmetric design is expected. Continued improvements to the electrodes and cell design will prov ide better capacitance. Many of the short-term improvements will likely be directed toward manufactur ing capacity . Ultracapacitors are becoming more common, and as improvements to performance and cost make them more accessible to multip le applications, this trend will continue. Over the next few years, it is expected that the market for ultracaps will increase dramat ically . Trends in materials for ultracaps will continue forward with improvements and utilization of activated carbons, thinner current collectors, and improved cell design and packaging. Research being done at universities and national labs will continue, but these will not become common materia ls for ultracapacitors in the next few years. However, as the demand for ultracapacitors increases, specialty materials will begin to see utility for some applicat ions, driving down manufacturing costs, which in turn will enable their use in more systems. Significant use will be made in integrating into power converters to reduce size, mass, and cost. Both th in-film and MCap devices will become prevalent. If MCaps achieve lithium ion specific energy their adoption will be rapid and revolutionary as has occurred with LiFePO4 (lithium-iron-phosphate) batteries for both transportation and extremely high pulsed-power systems such as lasers. Within the next 10 years, ult racaps will begin to see the incorporat ion of advanced carbon materials and electrodes designed with features on the nanometer scale. Carbon nanotubes, graphene, and porous carbons all have extreme ly high potent ial to unseat activated carbon as the electrode of choice. The replacement of carbon systems being used today (including activated carbon, aerogels, and carbon cloths) will occur as cost reductions take place in the manufactur ing of nanostructured carbonaceous mater ials. The tradeoff between surface area and pore size can be exploited at the nanoscale. Optimization of these parameters will likely yield fairly significant improvements to the capacitance and result in higher power and better energy density. The mid-range (2020 -30) developme nt of ultracapacitors wil l most likely be the full incorporatio n of advanced carbons and hybrid systems . It is hard to anticipate which technology has the most to offer, as there are unique benefits and hurdles for each. Hybrid systems will be common, utilizing a battery-like electrode combined with one of the advanced carbon electrodes. The pseudocapacitive electrode options will become diversified, with manufacturers using unique materials to differentiate their product. An alternative electrode to the ruthenium oxide (RuO2) used today will be used for both performance and economical reasons. Supply issues will drive electrode materia ls toward those with larger availability. UNCLASSIFIEO/fFOA OFFl€il.t.k YSE O,.Llf 25

UNCLASSIFIED1fFOR 8ffl@IAL tt.!! e"t I In the 2020-2030 timeframe, ultracapacitor cell designs will be different. Simple plate designs, coin cells, and laminated pouches will still be used for some applications. However, advanced designs and form-factors will be used to fit specialized applications. Ultracapacitors will be designed for incorporation into chip architectures. Several options will exist for micro- and nano-electronics. Electro lyte options will most likely be different from those used today. New electrolyte solvents and better salts will be discovered that produce higher voltage cells while maintaining good ionic conductivity. Ionic liquids and complex solvents that withstand higher voltages are likely candidates. Three dimensional architectures may begin to be introduced, which will greatly improve both energy density and power density. Manufacturing techniques will enable small- and medium-format three-dimensional electrodes, where the electrical double layer is stored perpendicular to the current collector. The manufacture and assembly of large­ format ultracapacitors with extremely high power capabilities will be enabled by three­ dimensional architectures of nano-engineered carbons and current collectors. Beyond 2030 it is diff icult to project the future developments of ultracapacitors since materials science is experiencing a nanomaterials revolution. It can be expected, however, that nearly all microelectronic circuitry will incorporate thin-film ultracaps or MCaps, and laptop computers will become the size of an iPhone with mostly voice­ activated functions. As depicted in Figure 16, advancements will occur on a much faster timescale than what has happened over the last 50 years. In the energy storage area, particularly for PHEV and EV batteries, the major impact for ultracapacitors may allow not only reduced mass and size but also order-of-magnitude reduction in costs. 010 2020 2030 2050 ■ •ULTRACAPS ► Figure 16. Evolution of Ultracapacitors UNCLASSIFIEO/fFOA QFFI€il.t.le YSE O,.LY 26

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I Chapter 6: Conclusions Ultracapacitors are becoming increasingly important in applications ranging from portable electronics to cargo transport cranes. Their introduction in electronic circuits for conversion and storage assistance will become the next major milestone. The virtually unlimited cycle life combined with high power capabilities has made them an integral component to many energy-management schemes. Early adoption of ultracaps by the military was an important differentiator for vehicle performance and a technology driver. Improvements to activated carbons and manufacturing techniques have brought costs down. The ultracapacitor is relatively young and is expanding rapidly as ultracaps are reaching a more widespread audience. Increased adoption has led to recognition of the benefits, and the attention has allowed researchers to develop a better understanding of the power-storage mechanisms. New materials and cell designs will produce increased capacitance and higher voltages, which will in turn give ultracapacitors better energy density. Advanced carbons show great promise to generate a carbon electrode with much higher capacitance than the activated carbons in use today. These include porous carbons, carbon nanotubes, and graphene. Capacitance increases by the carbon electrode alone could double the energy density. Additional advantages in conductivity and the ability to functionalize these materials will generate improvements in power density as well as energy density. Hybrid systems utilizing a battery-like electrode are beginning to gain acceptance. These pseudocapacitive systems maintain the high power density of a symmetric ultracap and also have substantially higher energy density. While these systems sacrifice some cycle-life capability, it appears as though they will survive tens of thousands of cycles. Ultracapacitors are a solution for generating or absorbing high pulse power. They also are excellent devices for backup power and stabilization of fluctuating power requirements. Increased power densities and energy densities will open up new applications for ultracaps, replacing fuel and batteries in some cases, supplementing them in others. Power management and generation for military systems will see increasing uses for ultracaps. Their unique capabilities will enable future designs. Their specific performance capabilities must be optimally utilized and we must have a keen understanding and awareness of the unique spectral signatures these devices generate when storing high power and in charge or discharge modes. Development of these advanced ultracaps must be performed, understood, and adopted by the U.S. government and those who are developing the systems to be used by the government. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 27

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I Chapter 7: Endnotes L Tecate Group website 2 B. E Conway, Electrochemical Supercapacitors: Scientific, Fundamentals, and Technological Applications, Kluwer, New York (1999) . 3 Wikipedia Commons 4 J. R. Miller, P. Simon, Fundamenta ls of Electrochemical Capacitor Design and Operation; ECS Interface, 17, Spr. 2008, 31-33 5 EPCOS White Pa per on UltraCap Technology 6 P. Simon, A. Burke, Nanostr uctu red Carbons: Double-La yer Capacitance and More; ECS Interface, 17, Spr. 2008, 38-43 7 EPCOS White Pa per on UltraCap Technology 8 D. Tuite, Get the Lowdown On Ultracapacito rs; Electronic Design, Nov 2007. 9 J. R. Miller, A. F. Burke, Electrochemical Capacitors: Challenges and Opportu nities for Real-World Applicati ons; ECS Interface, 17, Spr. 2008, 53-57 10 J. R. Miller, A. F. Burke, Electrochemical Capacitors: Challenges and Opportu nities for Real-World Applications; ECS Interf ace, 17, Spr. 2008, 53-57 LI J. R. Miller, A. F. Burke, Electrochemical Capacitors: Challenges and Opportunities for Real-World Applicati ons; ECS Interface, 17, Spr. 2008, 53-57 12 K. Naoi, P. Simon, New Materials and New Configurations for Advanced Electrochemical Capacitors ; ECS Interface , 1 7, Spr. 2008, 34-37 13 Burke, Ultracapacitor Technologies and Application in Hybrid and Electric Vehicles; Internat ional Journal of Energy Research; July 2009 14 J. R. Miller, P. Simon, Fundamentals of Electrochemical Capacitor Design and Operation; ECS Interface, 17, Spr. 2008, 31-33 15 P. Simon, A. Burke, Nanostructured Carbons: Double-Laye r Capacitance and More; ECS Interface , 17, Spr. 2008, 38-43 16 EPCOS White Paper on UltraCap Technology 17 P. Simon, Y. Gogotsi, Charge Storage Mechanism in Nanoporous Carbons and Its Consequence for Electrical Double Layer Capacitors, Phil. Trans. R. Soc. A, 368, 3457 (2010) 18 C. N. R. Rao, A. K. Sood, R. Voggu, K. S. Subrahmanyam, Some Novel Attributes of Graphene, J. Phys. Chem. Lett., 1, 572 (2010) 19 S. Park, R. Ruoff, Chem ical Methods for the Production of Graphenes; Nature Nanotechnology; 4, 217 (2009) 20 J. Chmiola, G. Yushin, Y. Gogotsi, C. Portet, P. simon, P. L. Taberna, Anomalous I ncrease in Carbon Capacitance at Pore Sizes Less than 1 Nanometer; Science, 313, 1760 (2006) 21 P. Simon, A. Burke, Nanostructured Carbons: Double-Laye r Capacitance and More; ECS Interface, 17, Spr. 2008, 38-43 22 D. Tuite, Get the Lowdown On Ultracapacitors ; Electronic Design, Nov 2007. 23 D. Tuite, Get the Lowdown On Ultracapacitors; Electronic Design, Nov 2007. 24 A. S. Arico, P. Bruce, B. Scrosati, J.-M. Tarascon,W. van Schalkwijk, Nat. Mater. 4, 366 (2005) . 25 S. W. Lee, B.-S. Kim, S. Chen, Y. Shao-Horn, P. T. Hammond, J. Am. Chem. Soc. 131, 671 (2009) . 26 D. Pech et al., J. Power Sources 195, 1266 (2010). 27 w. Sun, X. Y. Chen, Microelectron . Eng. 86, 1307 (2009). 28 J. Wang, J. Polleux, J. Lim, B. Dunn, J. Phys. Chem. C 111, 14925 (2007) . 29 K. Naoi, P. Simon, New Materials and New Configurations for Advanced Electrochemical Capacitors; ECS Interface, 17, Spr. 2008, 34-37 30 H. J. Ahn, W. B. Kim, T. Y. Seong, Electrochem. Commun . 10, 1284 (2008). 31 C. N. R. Rao, A. K. Sood, R. Voggu, K. S. Subrahmanyam, Some Novel Attributes of Graphene, J. Phys. Chem. Lett. , 1, 572 (2010) 32 M. D. Stoller, S. Park, Y. Zhu, J. An, R. S. Ruoff, Graphene-Based Ultracapacitors, Nano Lett, 8, 3498 (2008) 33 M. D. Stoller, S. Park, Y. Zhu, J. An, R. S. Ruoff, Graphene-Based Ultracapacitors, Nano Lett, 8, 3498 (2008) 34 J. Chmiola, G. Yush in, Y. Gogotsi, C. Portet, P. simon, P. L. Taberna, Anomalous Increase in Carbon Capacitance at Pore Sizes Less than 1 Nanometer; Science, 313, 1760 (2006) 35 D. Pech et al., J. Power Sources 195, 1266 (2010). 36 J. Leis, M. Arulepp, A. Kuura, M. Latt, E. Lust, Carbon 44, 2122 (2006) . 37 A. Janes, E. Lust, J. Electrochem. Soc. 153, A113 (2006) . 38 E. N. Hoffman, G. Yushin, B. G. Wendler, M. W. Barsoum,Y. Gogotsi, Mater. Chem. Phys. 112, 587 (2008) . 39 R. K. Dash, G. Yush in, Y. Gogotsi, MicroporousMesoporous Mater. 86, 50 (2005) . 40 z. G. Cambaz, G. N. Yushin, Y. Gogotsi, K. L. Vyshnyakova, L. N. Pereselentseva, J. Am . Ceram . Soc. 89, 509 (2006) 41 J. Chmiola, C. Largeot, P. Taberna, P.Simon, Y.Gogotsi, Monolithic Carbide-Derived Carbon Films for Micro­ Supercapacitors, Science,328, 480 (2010) 42 A. Janes, E. Lust, J. Electrochem . Soc. 153, A113 (2006) . 28 UNC LASSI FIEO/fFOA OFFI&I:.l.b YS& ONLY

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I 43 J. Leis, M. Arulepp, A. Kuura, M. Latt, E. Lust, Carbon 44, 2122 (2006) . 44 T. Goto, T. Kimura, G. Lawes, et. al. Ferroelectricity and Giant Magnetocapacitance in Perovskite Rare-earth Manganites. Physical Review Letters, 92, 257201 (2004 ). 4 5 N. Hur, S. Park, P.A. Sharma, S. Guha, and S-W. Cheong, Colossal Magnetodielectric Effects in DyMn2O5, Physical Review, Lett. 93, 107207 (2004). 46 R. F. Mamin, T. Egami, z. Marton, and S. A. Migachev, Giant dielectric permittivity and Magnetocapacitance in La0.875Sr0.125MnO3 Single Crystals, Phys. Rev. B 75, 115129 (2007) . 47 J. Wan, Qi Lu, Bo Chen, Fengqi Song, J. Liu, J.g Dong, and G.Wang, Giant Room-Temperature Magnetocapacitance in Co2+ Doped SnO2 Dielectric Films, Applied Physics Letters, 95, 152901 (2009) UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 29

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I Figure 10. CDC Synthesis and Electrochemical Test Cell Preparation Schematic. 21 Figure 11. Volumetric Capacitance of the Films in (A) TEABF4 and (B) H2S04...... 21 Figure 12. MCap Structure ................................................................................... 22 Figure 13. Preliminary MCap Test Results ........................................................... 23 Figure 14. Summary Chart of MCap Results to Date ............................................ 23 Figure 15. Specific Energy and Estimated Costs vs. foMc ...................................... 24 Figure 16. Evolution of Ultracapacitors ............................................................... 26 Tables Table 1: Contrasting Properties of Batteries and Ultracapacitors ........................... 4 Table 2: Manufacturers of Ultracapacitors .............................................................. 7 Table 3: Properties of Various Electrolytes Used in Ultracapacitors ....................... 8 Table 4: Properties of Various Materials used in Electrochemical Capacitor Electrode Materials ................................................................................................. 9 UNCLASSIFIEll/FQA QFFl&IAk W&lii OHL¥ iii

UNCLASSIFIED1fFOR 8ffl@IAL tt9! e"t I Ultracapacitors as Energy and Power Storage Devices for Commercial and Military Applications Summary Ultracapacitors (ultracaps) are energy storage devices capable of extremely rapid charge and discharge rates with the ability to be cycled hundreds of thousands of times. These unique capabilities make ultracaps attractive for a number of applications. The advantages over lithium ion batteries are somewhat mitigated by the fact that they have less than 10% of the specific energy (Wh/kg) and require a power converter to regulate their voltage. The high capacitance is achieved by the enormously high surface area of the carbon electrodes compared to planar electrolytic capacitors. Ultracapacitors are commonly used to provide voltage stabilization of power converters and to supplement peak-power loading of electrical systems. Their application has expanded from electronics and telecommunications to industrial load leveling and transportation. Due to their durability, their use in aerospace has been for munitions fusing and missile power load leveling. Ultimately, application will be pulsed power storage and delivery for electric propulsion and directed­ energy weapons. Recent developments in ultracapacitors have focused on different active materials and electrode designs. By using advanced carbons, polymers, and metal oxides, increased power density and energy density can be achieved. Oxide-based thin-film ultracaps have demonstrated high-performance energy and power density, approaching theoretical limits. Carbon nanotubes and advanced carbons have been used as additives to and solely as electrode materials. Nanostructured materials and thin-film manufacturing process improvements will generate breakthroughs in mass production. Magnetic capacitors are a recent development, manufactured using semiconductor processes on silicon wafers. They possess a different storage mechanism and will transform energy storage and pulsed-power applications. Ultracapacitors are reaching widespread adoption and they are now available to support electronics through transportation. New advancements in materials will enable high-energy density devices with exceptional power capabilities. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY iv

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Chapter 1: Concept Overview Ultracapacitors, also known as supercapacitors, hybrid capacitors, electrochemical capacitors, electrochemical double-layer capacitors, or ultracaps, are energy-storage platforms that offer energy storage capable of extremely rapid charge and discharge rates. Ultracapacitors also have the ability to be cycled hundreds of thousands of times. The phenomenal charge and discharge capabilities make them ideal for supporting volatile memory and comput ing applications, energy efficiency/capture processes, and high­ power applications. Ultracaps store their charge in the electrical double layer between the electrode and the electrolyte. Charge storage is a physical mechanism rather than a chemical phase change, so these are theoretically capable of cycling an infinite number of times. 2 One of the Figure 1. Ultracaps in Various Configurations.' disadvantages of ultracaps can be a high self-discharge rate. 3 The charge and discharge voltage output for an ultracap is a sloping linear curve, which allows for straightforward and accurate state-of-charge monitoring. 4 The real advantage of an ultracapacitor is the ability to deliver or accept bursts of power in a short time. Ultracaps were int roduced in 1966 and found init ial use 12 years later as backup power devices for volatile memory and clocks. Over the last 30 years, numerous advances have been made that have led to many uses of ultracaps, from transportation to portable electronics and more. Ultracaps are becoming more affordable as activated carbon electrodes and manufacturing improvements have driven costs down. Several companies now make ultracapacitors to fill a broad spectrum of applications. Figure 1 shows a number of ultracaps with various capacities to fill a range of functions. Transportation, microelectronics, and aerospace markets are some of the many areas where ultracaps have become enablers. Advanced materials and improved cell designs will lead to improvements in both power density and energy density. The improved performance is expected to make ultracaps important components for efficient power leveling and high-power receiving and delivery. The unique signature of ultracapacitors must be understood for aerospace and military applications. As an electrochemical capacitor, ultracapacitors store energy within the electric double layer formed at the interface between the electrode and electrolyte. In a conventional capacitor, the energy is stored by moving charge carriers from one plate to another, and the charge separation creates a potential. Voltage differentials in a conventional capacitor are dependent upon the dielectric material separating the plates. In the ultracap, the electrical double layer is the separation of charge in a vanish ingly thin gap between two plates. Figure 2 depicts the components and basic design of the ultracap, UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 1

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I showing the electrochemical double layer responsible for the characteristic performance. 5 The basic principles of operation have not changed with improved technologies, only the materials and cell design. The two electrode plates are traditionally the same material, which is usually a carbonaceous material, such as activated carbon. Each layer is capable of storing a low voltage, and multiple layers serve to increase the potential. Individua l cells are placed in series to create a higher voltage, in much the same manner as a battery. Recent advances in ultracapacitors have moved toward employing dissimilar electrodes, which create a higher potential; since the electrodes are now different, these are sometimes referred to as "pseudo" or hybrid capacitors . The pseudocapacitor uses a battery-like electrode to replace one of the carbon electrodes, yielding a high-energy-storage electrode and a highly capacitive electrode within the same system. The capacitance of an ultracapacitor can be determined by the Helmholtz equation (equation 1), which describes the relationship between the electrolyte and the electrode: C = EA/d (1) Here, £ is the dielectric constant of the electrolyte, A is the availab le surface area, and d represents the distance between the center of the double layer and the electrode surface. Increasing the double-layer capacitance in an ultracap is generally accomplished by either manipulating the electrode (carbon) surface area or the electrolyte. Energy density (equation 2) is the product of the capacitance and the square of the voltage: E = ½ cv2 (2) Strateg ies for increasing the stored energy target improvements to both the electrodes and the electrolyte. Changing from an aqueous electrolyte to an organic electrolyte with a higher dielectric constant will increase the voltage from approximately 1 volt to more than 2.5 volts. Increasing the surface area of the electrode is another approach to increasing storage capability. There is a tradeoff between porosity and surface area that must be considered when constructing an electrode with an extremely high surface area. An activated carbon approaching a measured surface area of 3,000 m2/g may have less than half that as accessible or useable area.6 Increasing the pore size sacrifices surface area but provides more accessible material. By providing more interfac ial area between the electrode and the electrolyte, a better electrochemical double layer can be produced, yielding a better ultracapacitor. UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 2

UNCLASSIFIED1/FOR 8ffl@IAL tt91! e"t I Activated Materi al selection and construction carbon electrodes Electrolyte Activated carbon electrodes 0000000000 Separator soaked in electrolyte 0000000000 ■ Electrodes ■ Separator ■ Electrolyte ■ Wound component ■ Case Figure 2. Operation of an Ultracapacitor. A simple ultr acapacitor utilizes two similar electrod es (typica lly carbon) with a separator and electrolyt e to create the electroc hemical double layer that governs operation . 7 BATTERIES AND ULTRACAPS For an ultracapacitor, energy is stored within the electrica l double layer, which is a physical storage mechanism . Contrast this with batter ies, where charge and discharge take place through a chemical (Farada ic) reaction . Since charge is stored physically, there is no direct degradation mechanism that limits cycle life. Side reactions do occur, which prevents infinite cycle life, but ultracaps can often be cycled tens of millions of times .8 Batteries store their energy through chemica l reaction; consequently, cycle life is much shorter and heavily dependent on the depth of discharge . In contrast, there are no limitations to the discharge depth with an ultracapac itor. Table 1 addresses some of the general characteristics of batteries and ultracapacitors, which can help to determine the preferred system for a specified application . The values are based upon 2008 data, so some numbers are outdated, but the relat ive values remain consistent. There are now hybrid devices that blur the boundaries further by introducing a battery-like electrode to replace one of the two identical electrodes in a traditional ultracapacitor. UNCLASSIFIEO/fFOA OFFI&I:.l.b YS& ONLY 3

UNCLASSIFIED11FOR. 8FFl@IAL tt.!! e"t I Table 1: Contrasting Properties of Batteries and Ultracapacitors 9 Characteristic State of the Art Lithium Ion Battery Ultra capacitor Charge time ~3-5 minutes "'1 second Discharge time ~3-5 minutes ~1 second Cycle life <5,000 @ 1C rate >500,000 Specific Energy (Wh/kg) 100-200 5 Specific power (kW/kg) 0.5 -1 5-10 Cycle efficiency (%) <50% to >90% <75 to >95% Cost/Wh $1-2/Wh $10-20/Wh Cost/kW $75-150/kW $25-50/kW The voltage curve for an ultracapacitor is proportional to the depth of discharge, whereas batteries deliver a relatively constant voltage over a long discharge period. The sloping voltage curve of an ultracap can be advantageous for state-of-charge determinations. However, this is also responsible for the decreased energy density available in an ultracapacitor. For applications requiring energy to be delivered over a longer time scale or delivered at a constant voltage, the relatively flat voltage curve found in most batteries would be the preferred option. In instances where power is to be delivered or received quickly, or where many cycles are required, an ultracap is ideal. To better understand the differences between batteries and ultracaps, it helps to think of batteries as storing watt-hours of energy and ultracaps as storing watts of power. A Ragone plot illustrates the distinctions by plotting power versus energy in a logarithmic scale. As can be seen in Figure 3, ultracaps make an excellent option where high power density is required. The times shown are rough estimates for a full charge or discharge, which are estimates to help understand the relationship between energy density and power density. Batteries have made significant improvements to power delivery recently, but batteries remain the high-energy-density solution . Ultracapacitors are often thought of as a stop-gap between conventional capacitors and batteries. Recent developments of hybrid capacitors have led to considerable progress toward higher energy density. However, there is a tradeoff between high energy density and high­ power devices, and the distinctions between the two require consideration of application requirements when choosing the energy storage platform for a system. The differences between ultracaps and batteries do not make the two mutually exclusive. There are applications where either a battery or an ultracap is the preferred energy storage platform. Quite often, these two can be used together in systems to perform complementary roles. UNCLASSIFIEO/;LFOA OFFI&il:.l.b YS& OrtLY 4