Integrierte Lithiumbatterien für extrem miniaturisierte Sensoren Integrated lithium micro batteries for highly miniaturized sensors

Transcription

1 Integrierte Lithiumbatterien für extrem miniaturisierte Sensoren Integrated lithium micro batteries for highly miniaturized sensors Dr. Robert Hahn 1, Katrin Höppner 2, Marc Ferch, 2 Krystan Marquardt 1, Prof. Klaus-Dieter Lang 1 1 Fraunhofer IZM, Berlin, Germany, robert.hahn@izm.fraunhofer.de 2 Technische Univeristät Berlin, Germany Kurzfassung Zur Energieversorgung extrem miniaturisierter Sensoren werden Energiespeicher benötigt, die sehr klein und gut integrierbar sind. Lithium Ionenbatterien sind aufgrund ihrer hohen Energiedichte und Impulsbelastbarkeit sowie der geringen Selbstentladung dafür besonders gut geeignet. Insbesondere werden sie als Zwischenspeicher für Energy- Harvesting-Systeme eingesetzt. Zunächst wird ein Überblick zum Stand der Technik von Mikrobatterien sowie zur Forschung an integrierbaren Mikrobatterien für verschiedene Aufbaukonzepte gegeben. Anschließend werden die Arbeiten zur Entwicklung von Lithium-Mikrobatterien mit Interdigitalstruktur am Fraunhofer IZM beschrieben und die Resultate hinsichtlich Energiedichte und Stromtragfähigkeit im Vergleich zum internationalen Stand dargestellt. Abstract Highly miniaturized energy autarkic sensors require an energy storage which is small and can be integrated into the system. Lithium ion batteries are a favourite option since they show a superior energy density and pulse performance while self-discharge is very low. In particular they are used as buffer storage in energy harvesting devices. Starting with a description of the state of the art of secondary micro batteries an overview is given of the research directions towards integrated lithium micro batteries classified by the construction principles. The research on lithium micro batteries with coplanar electrode structure performed at Fraunhofer IZM is described in the main section of this paper. Capacity, discharge curves and rate retention as function of current is shown and compared with the international state of the art. 1 Introduction Effective intermediate energy storage is required for all energy harvesting concepts, due to the varying availability of ambient energy and varying energy requirements of the device. In most cases this intermediate storage is done at best with the help of a secondary micro battery or with a capacitor or super capacitor. Secondary batteries have a much higher energy density compared to capacitors and sufficient power pulse capability. Lithium batteries have the highest energy density compared to other chemistries since lithium is the lightest of the metals and also has the highest reduction capability, expressed as the lowest standard reduction potential of all the metals in the electrochemical series, at V. Therefore only lithium secondary batteries will be considered here. The development of miniaturized power sources with high volumetric energy density is necessary for small electronic applications such as sensor nodes, active smart labels or MEMS. The energy density of batteries decreases by reducing the size of the batteries. This is caused by the increasing amount of passive packaging material in proportion to active material in small batteries. There is a significant reduction of energy density at a thickness below 1 mm because the volume of the packaging becomes dominant. Therefore direct integration of the battery on wafer- or package level is crucial if miniaturization is extremely important. Another relevant parameter of electrochemical storage systems is power density. Most secondary batteries have sufficiently small impedance and are capable of high current drain. But in miniaturized systems the secondary battery will be designed as small as possible. As a result a typical load and power pulse (for example for data transmission) may lead to much higher power density of small secondary batteries in comparison to larger systems. While miniaturization is the main driving force for integrated micro batteries, self-discharge is another reason for using small batteries in energy harvesting applications even if enough space is available for a larger state of the art battery. For example: a 100 mah lithium polymer battery is used while a much smaller battery would be sufficient to fulfill the storage requirements. A typical value of 30 % self-discharge per year translates into 3µA continuous self-discharge current which has to be compensated by a larger energy harvester. An impressive example of miniaturization and full systems integration was shown by G. Chen et al. [1]. It was demonstrated that a micro battery of only 1µAh capacity is sufficient to power a fully autonomous cubic-millimeter energy-autonomous wireless intraocular pressure monitor. One should keep in mind, that the power demand of autarkic electronic systems will be further reduced by a factor of 10 to 100 during the next decade. That means much

2 more applications can be powered with the tiny storage batteries described here. Energy autarkic systems and their energy harvesting devices can vary widely in terms of load profile, size, and other parameters. Therefore not one single electrical buffer storage will be used but a variety of systems have to be adapted to individual applications. This is another advantage of the battery integration concepts described below: size (form factor), capacity and power performance can be optimized up to a certain degree for each application only by changing design parameters. Some of the concepts offer the additional advantage of choosing different electrode and electrolyte materials from a proven set of compatible materials to customise parameters like cell voltage and cycle number for a given application. In summary, there are some general requirements for secondary batteries as electrical buffers in energy autarkic sensor systems: The battery must be rechargeable and should have a high energy density. The voltage of the battery should match to the electrical circuit. The buffer storage should be capable of high pulse discharge rate. This is the case in most systems with transceiver module. Small geometrical dimensions and flexible form factors may be necessary. Minor self-discharge is of crucial importance to utilize nearly all of the harvested energy. A long life of 10 years and more is of significant importance since the idea of energy harvesting is to have a system which works maintenance free over decades. A life time of 10 years or more is still an issue for most of the rechargeable batteries. The energy storage device must be electrically connected to the system. Encapsulation of the micro batteries is another issue. A near hermetic encapsulation is required for long life time. On the other hand the encapsulation volume should be only a small fraction of the battery. For special applications the encapsulation must be biocompatible and pressure resistant. Safety is a concern of all lithium batteries. On the other hand the safety is less critical in case of very small batteries. 2 Micro battery state of the art 2.1 Solid state thin film batteries Solid state batteries are chemical cells that have both, solid electrodes and solid electrolytes. Even though the lithium ion conductivity of the solid electrolyte is less than for liquid electrolytes, it is sufficient because just 1 µm-thick films are adequate for a pinhole-free barrier over most thin film electrodes. Further, the electronic resistivity of the mostly used LiPON is relatively high [2]. In comparison to lithium ion batteries, were graphite is used as negative electrode to avoid dendrite growth during charge, thin film batteries use a metallic lithium since dendrites are prevented by the solid state electrolyte. LiCoO 2 is the state of the art material of the positive electrode. Thin film batteries are easy to miniaturize by vacuum deposition and lithographic patterning. There is no problem with electrolyte leakage. Liquid electrolytes are often highly corrosive. Long term stable packaging of small thin film batteries although still an issue, is easier than the packaging of integrated micro batteries with liquid electrolyte. Solid state batteries have very long shelf lives, and show usually only little changes in performance with temperature. These are very important characteristics for electrical buffers of energy harvesting systems. On the other hand, the deposition of the electrolyte layer is a particularly highly challenging and time consuming process that leads to high production costs. Electrical cycling of the solid-state battery leads to cyclic changes of crystallite volume and material tension. Therefore, the total capacity is rather low since layer thickness cannot be increased above a few micro meters. Capacities around ca. 100 µah/cm² are usually obtained with this technology. Today several companies are offering solid state thin film batteries in the 50 µwh 20 mwh range. The discharge voltage is between 4.2 and 3.0 Volts. Several thousand full cycles and a relatively wide operational temperature window (-40 C 150 C) have been reported. Ageing models were developed for thin film solid state batteries which allow life time estimation as function of temperature, current and cycle depth [3]. 2.2 Lithium polymer batteries Lithium-based rechargeable batteries are used in very high volumes in applications, from mobile phones up to electrical vehicles. Lithium-ion polymer batteries use liquid lithium-ion electrochemistry in a porous matrix with polymer binder that eliminates the free electrolyte from the cell. They have a flexible, foil-type (multi-layer laminate) case, which is called pouch package. Meanwhile many different types of materials have been developed which feature lower cost, higher cycle life, safety and increased specific capacity. Nevertheless for small batteries LiCoO 2 is still the material of choice since it has the highest volumetric capacity density. The mean discharge voltage is approx. 3.7 V. Lifetime and self-discharge are both strongly dependent on temperature. The maximum storage life is achieved at low temperatures and ca. 40 % state of charge. After a full charge, a Li-ion battery will typically lose about 5 % capacity in the first 24 hours, then approximately 2-3 % per month at 20 C (recoverable capacity loss). A multitude of degradation mechanisms are responsible for losses [4]. They are mainly a function of the number of full

3 charge/discharge cycles, battery voltage and temperature. Battery ageing is accompanied by increase of cell impedance, power fading, and capacity decay. The smallest batteries available in pouch package are in the capacity range between 50 and 100 mah. 2.3 Coin type cells In a coin-type battery the active masses of and are filled into a metal cup and a metal cover, respectively, in the form of powder or pastes in order to provide the two electrical contacts to the outside. Both parts are separated by a separator and joined by means of a polymer sealing ring. All coin cells for memory back-up can be used only with low currents and are therefore not sufficient for most energy harvesting applications. Nevertheless, they show a low self-discharge rate and withstand over-discharge and over-charge. Various combinations of and materials are available to obtain mean discharge voltages between 1.3 and 2.8 V. The rechargeable coin cells with lithium titanium oxide as differ in terms of cycle life greatly from the cells with LiAl. While cells with LiAl can be cycled more than 1000 times at 10 % depth of discharge (DOD), they can be cycled only approx. 50 times at full discharge. In contrast the cells with titanium oxide can be cycled 500 times at 100 % depth of discharge which make them well suited for a lot of energy autarkic systems. The self-discharge is between 2 and 5 % per year. Energy and power of typical state of the art secondary micro batteries is shown in fig. 1. The power was calculated as product of maximum current and nominal voltage. The coin cells have the lowest power density while some thin film batteries are capable of a current delivery similar to lithium polymer cells. 3 Micro battery research 3.1 Stacked electrode configuration So far the significant majority of all battery designs relay on layer-by-layer construction. The electrodes are coated on metal foils, dried and calandered (pressed between rolls). Those electrodes are stacked with separators to complete the electrochemical cell. For micro batteries thin and thick film technology can be applied as well Integrated lithium polymer battery Our first attempt towards a silicon integrated micro battery was to pre-fabricate battery laminates which were cut to millimeter sized pieces and placed into silicon cavities. Near hermetic packaging was achieved with help of a glass lid [5]. Although functional prototypes were fabricated on a laboratory scale, the serial assembling process was slow. Another drawback of that concept was the unreliable electrical contact which was dependent on the contact pressure between battery substrate and the lid. separator top side contact glass lid seal passivated silicon substrate with cavity Figure 2 Integrated micro battery with stacked electrodes Thin film solid state batteries Figure 1 Overview of state of the art lithium secondary micro batteries. Comparison of maximum power and stored energy of coin, polymer and thin film solid state cells Recently a new type of coin cell has been introduced which can be described as in-between of lithium polymer and coin type. In this case millimeter small battery foil stripes of the lithium polymer type are fabricated as a multi turn winding and inserted into a coin cell housing [5]. A 50 mah cell is achieved in a 12 mm diameter, 5 mm thick housing with graphite and LiNi x Mn y Co z O 2. Thin film deposited LiCoO 2 or LiMn 2 O 4 s of traditional solid state thin film batteries require an annealing step (700/400 C) to get the required crystal structure. Therefore alternative positive electrodes including V 2 O 5, TiS 2 and TiO y S z were investigated although the potential window against lithium is reduced ( V) [6]. The chemical (resistance to humidity) and thermal stability of the LiPON electrolyte is enhanced by additions of boron [7]. Until now, the capacity of the thin film batteries was limited by the thickness of the LiCoO 2 layer. Battery performance deteriorates rapidly as the thickness exceeds 6 µm. It was recently shown, that thickness and thus capacity can be increased up to 400 µah/cm 2 if the LiPON electrolyte is replaced by a solid sulfide electrolyte (amorphous Li 2 S-P 2 S 5 ) which is fabricated by pulsed laser deposition [8].

4 3.2 Interdigitated electrodes Alternating rows of s and s are referred to as interdigitated structures. The main advantage is the ease of fabrication since both electrodes can be depositied on the same substrate. Material deposition or electrode assembly on top of the electrolyte layer is not required. If electrode width and distance is below the thickness dimensions of the stacked layer-by-layer battery, power performance of the interdigitated battery can be much higher. On the other hand an innate non-uniform current density must be taken into consideration Electrodes based on open nanoporous metal structure Electrodes with extreme high internal surface area can be fabricated based on nanoporous structures. Batteries with interdigitated electrodes were demonstrated based on 3D nanoporous bicontinuous nickel scaffolds (fig. 3). Active electrode material (NiSn and LiMnO 2 ) of approx. 50 nm thickness was deposited on the Ni nano structure by pulsed electroplating [9]. An energy density of approx. 150 µwh/cm 2 was reported. The main advantage seems to be the high current and pulse capability approaching values of super capacitors. On the other hand, voltage and capacity is relatively low and the fabrication of the Ni scaffold with polystyrene spheres seems to be an expensive process. electrolyte 0.5 µm Figure 4 Layer by layer printed lithium ion battery [11] Lift-off based electrodes Battery fabrication could be much simpler in contrast to the technologies shown in section and if electrode slurries used for planar batteries could be adapted to interdigitated structures. Dispensing or printing of these electrode pastes however lead to only low aspect ratios and the electrodes can t be placed close enough to result in a low diffusion length for the lithium ions. Therefore a lift-off process was developed were the electrode slurries were deposited with help of a micro-injection system into photoresist channels which were removed afterwards (fig. 5) [12]. This resulted in ca. 20 µm wide 15 µm thick Li 4 Ti 5 O 12 and LiCoO 2 stripes. Functional cells were fabricated with help of a gel electrolyte. Due to the high aspect ratio the narrow electrode structure of these batteries show a high current capability. Maximum discharge current was approx. 5 ma/cm 2. A capacity and mean cell potential was 270 µah/cm 2 and 2.35 V respectively. substrate with interconnects 30 µm Figure 3 Micro battery with nano porous interdigitatedelectrodes [9] gel electrolyte Additive manufacturing Printing of functional inks of (Li 4 Ti 5 O 12 ) and (LiFePO 4 ) materials was reported as a process of additive manufacturing to create stacked, interdigitated electrodes [10]. By repeated printing of several layers one on top of the other, a very high aspect ratio was achieved. Anodes and s were place closer to each other than before (fig. 4). The material is sintered after printing, resulting in true solid state electrodes without polymer binder. Electrochemical behavior was reported to be similar to state of the art batteries with the same electrode materials. substrate with current collectors 40 µm Figure 5 Battery with micro casted paste electrodes [12] Cavity integrated electrodes The approaches shown in fig. 3 to 5 do not feature a separator structure. The small separation between and is directly filled with electrolyte. Thus, these structures are prone to short circuits which may result from particle generation during fabrication or dendrite growth which can occur at high currents or low temperature charge. Therefore we introduced a modified structure with electrodes immersed in cavities. The transport of lithium ions between and takes place in the electrolyte layer on top of the electrodes.

5 cell voltage [V] cell voltage [V] cell voltage [V] gel electrolyte 4,2 4,0 3,8 3,6 0.4 ma 1.0 ma 2.0 ma 3.9 ma 5.9 ma 7.9 ma 9.8 ma Si substrate with current collectors Figure 6a micro battery with cavity electrodes The cavities were fabricated in silicon substrates by means of dry [13] or wet etching [14]. Composition and fabrication of the electrode pastes has been optimized for good electrical contact and process ability [15]. As shown in fig. 6a relatively wide electrodes were used to facilitate the cavity filling. Despite the large dimensions the current capability is quite good as will be shown in chapter 4 (table 1). The reason is that the lithium ions move during charge and discharge most of the way in the electrolyte and not in the electrode layer D Electrodes Three dimensional micro battery configurations have been introduced primarily to increase the electrode surface in a given volume and thus to increase the maximum current. The concepts and characteristics of 3D electrode configurations have been reviewed in [16]. 3D electrodes have not been practically realized due to numerous fabrication difficulties such as conformal electrode coating on high aspect ratio structures and assembly of 3D and 3D. 4 Results µm 200 µm 1.8 mm Micro batteries were fabricated with graphite (LiC 6 ) s, LiNi 1/3 Co 1/3 Mn 1/3 O 2 (NMC) and LiCoO 2 (LCO) s and 1M LiPF 6 in EC:DMC (1:1 wt%) electrolyte. The first prototypes used the coplanar design shown in fig. 6a with only one stripe of and each as shown in fig. 6b. graphite LCO separator 3,4 3,2 3,0 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6 discharge capacity [mah] Figure 7 Constant current discharge curves as function of current of planar micro battery (fig. 2). Anode: LiC 6, : NMC, size: 1.2 cm 2 4,2 4,0 3,8 3,6 3,4 3,2 3,0 0,00 0,02 0,04 0,06 0,08 0,10 0,12 discharge capacity [mah] Figure 8 Constant current discharge curves as function of current of co-planar electrode configuration (fig. 6). Anode: LiC 6, : LCO, electrode width: 900 µm, battery size: 0.16 cm 2 4,2 4,0 3,8 3,6 3,4 3,2 17 µa 34 µa 85 µa 170 µa 340 µa 850 µa 48 µa 120 µa 240 µa 480 µa 720 µa 3,0 0,00 0,05 0,10 0,15 0,20 0,25 0,30 discharge capacity [mah] Figure 9 Constant current discharge curves as function of current of co-planar electrode configuration (fig. 6). Anode: LiC 6, : NMC, electrode width: 1800 µm, battery size: 0.3 cm 2 Figure 6b Photographic and SEM image of fabricated coplanar lithium micro battery At first conventional batteries according to fig. 2 were fabricated to evaluate the electrochemical performance of the electrode paste recipes. For that purpose planar electrodes with exactly the same composition were fabricated

6 % of nominal discharge capacity by doctor blading or dispensing which were afterwards used for the coplanar. The resulting capacity and maximum current is 1.3 mah/cm 2 and 5-10 ma respectively. Discharge curves at various currents of planar and coplanar electrode batteries can be compared in figures 7-9. The results were highly reproducible. Batteries of the same design showed nearly identical discharge curves and capacities. No mayor difference between 200 and 400 µm wide separators occurred for the batteries with 900 and 1800 µm wide electrodes. Up to 100 charge discharge cycles have been performed with the coplanar batteries. The capacity degradation is relatively high, up to 0.5 % per cycle [14]. Since state of the art battery materials have been used, this result must be attributed to the micro battery package which still has to be improved. In all tests, there was no indication of short circuits or dendrite growth as expected due to the cavity design. The electrical characterization results are summarized in table 1 and fig. 10. It is quite obvious, that the capacity per area of the interdigitated design can only be half of the stacked electrodes if the same electrode thickness is used. The capacity is in any case higher than that of solid state thin film batteries with improved thickness [8]. The current capability of the coplanar batteries with 900 and 1800 µm wide electrodes is only slightly lower in comparison to the planar battery. In case of interdigitated electrodes and smaller width (300 µm) the same rate capability should be achieved. If the electrode width is further reduced, much higher discharge rates can be achieved as was demonstrated in [12] (fig. 10). electrode configuration stacked coplanar/ interdigitated source * [8] * * [12] electrode width [µm] capacity [mah/cm 2 ] energy [mwh/cm 2 ] maximum discharge current** [ma/cm 2 ] Table 1 Summary of micro battery parameters * this work, ** continuous discharge current at 50% capacity retention stacked electrodes (fig.2) coplanar electrodes 900 µm coplanar electrodes 1800 µm interdigitated 20 µm [12] 0 0, C Rate Figure 10 Capacity retention at various rates (C-rate measures the discharge current in relation to the capacity) 5 Conclusions The technology of integrated lithium micro batteries matured significantly during last years. The combination of optimized electrode pastes with an interdigitated electrode design results in high capacity and rate capability at low production cost mah/cm 2 and 2-5 ma/cm 2 can be achieved. Future work will include the proof of long term stability of the battery package and the development of high throughput manufacturing. 6 References [1] G. Chen et al., A Cubic-Millimeter Energy-Autonomous Wireless Intraocular Pressure Monitor, ISSCC 2011 [2] J. B. Bates, N. J. Dudney, B. Neudecker, A. Ueda, C. D. Evans, (2000). Thin-film lithium and lithium-ion batteries, Solid State Ionics, vol. 135, 2000, pp [3] Igor Bimbaud, STMicroelectronics-Tours: Implementation of energy harvesting technologies JNRSE 2013, March 27-28, 2013 Toulouse, France [4] J. Vetter, P. Novák, M.R. Wagner, C. Veit, K.-C. Möller, J.O. Besenhard, M. Winter, M. Wohlfahrt-Mehrens, C. Vogler and A. Hammouche (2005), Ageing mechanisms in lithium-ion batteries, Journal of Power Sources 147, pp [5] K. Marquardt, R. Hahn, T. Luger, H. Reichl, Assembly of and hermetic Encapsulation of wafer level secondary batteries, MEMS2006, Istanbul, Jan., pp [6] B. Fleutot et al. Characterization of all-solid state Li/LiPONB/TiOS microbatteries, Journal of Power Sources 196 (2011), p [7] Yongsub Yoon et al., The mixed former effect in lithiumborophosphate oxynitride thin film electrolytes, Electrochemical acta 111 (2013) p.144 [8] Mitsuyasu Ogawa et al. High capacity thin film lithium batteries, Journal of Power Sources 205 (2012) p. 487 [9] Pikul J H, et al. High-power lithium ion microbatteries from interdigitated three-dimensional bicontinuous nanoporous electrodes, Nat. Commun., 4 (2013), 1732 [10] Ke Sun et al. 3D Printing of Interdigitated Li-Ion Microbattery Architectures, Advanced Materials Vol. 25 (2013) Issue 33, p [11] [12] Kazuomi Yoshima et al., Fabrication of micro lithium-ion battery using polymer wall, Journal of Poser Sources 208 (2012), p.404 [13] R. Hahn, et al., Silicon integrated micro batteries based on deep reactive ion etching and through silicon via technologies, ECTC2012, San Diego, CA, May 29-June 01, 2012, pp [14] K. Höppner et al., Design, fabrication and testing of silicon integrated micro batteries, PowerMEMS, December 3-6, 2013, London, UK, p. 411 [15] M. Ferch, K. Hoeppner, K. Marquardt, R. Hahn, Development and Investigation of electrode slurries for micro batteries,2 nd International Conference on Materials for Energy, EnMat May 12-16, 2013 [16] B. Dunn, J. W. Long et al., Rethinking Multifunction in Three Dimensions for Miniaturizing Electrical Energy Storage The Electrochemical Society Interface Fall 2008, pp.49-53