Patent Application
20140205927
The present invention pertains to a fuel cell having the features described in the preamble of patent claim 1 and to a process for manufacturing such a fuel cell.
The logging of measured values at poorly accessible or mobile sites will be increasingly assumed by autonomous microsystems. Such microsystems comprise, for example, sensors, actuators, a signal processing means and a power supply means, which are connected to one another. To output processed measured values, such microsystems preferably also have a transmitting unit or another interface for data output. Developments made so far in such fields show great advances in the development of the sensor system and the actuator system in terms of miniaturizability and in terms of a reduction of the power consumption. However, sufficient advances have not been achieved in terms of the power supply.
Especially integrated circuits according to the CMOS technology (CMOS: Complementary Metal Oxide Semiconductor) are used in the development of intelligent microsystems. p- and n-Channel MOSFETs (Metal Oxide Semiconductor Field Effect Transistors) are used here at the same time to manufacture integrated circuits. If the CMOS technology is used, all materials and processes used must be CMOS-compatible, and this is especially true when a CMOS process is to be carried out monolithically on a silicon chip, Essential aspects for an autonomous microsystem are that the storage media used will have a high energy density; the components used will be compatible with the environment, and the power supply will be able to be miniaturized and will be cost effective.
Batteries, storage batteries and micro fuel cells are known for power supply. Fuel cell systems have already become established as macroscopic energy systems. However, it will become technologically increasingly more complicated with increasing degree of miniaturization to manufacture fittings and guides with sufficiently good tolerances in the case of mechanical components such as valves and pressure regulators, which comprise movable parts. Attempts made so far at miniaturizing fuel cells are therefore limited above all to the fuel cell proper. The literature contains, in general, several approaches towards manufacturing PEM fuel cells (PEM: Proton Exchange Membrane) according to the silicon technology. Separate hydrogen storage units are made available, especially used, for power supply. In addition, direct methanol fuel cells are known, which are designed as passive systems, the oxidant methanol being supplied from a storage tank via capillary forces to the anode of the fuel cell. The CO2 gas bubbles formed during the reaction of the methanol are likewise removed by means of capillary forces.
The object of the present invention is to propose an alternative fuel cell or a process for manufacturing same, wherein miniaturization, including a hydrogen storage unit, will be made possible. In particular, a monolithic reaction will be made possible in conjunction with the CMOS technology.
This object is accomplished by the fuel cell having the features according to patent claim 1 and by a process for manufacturing a fuel cell, which has the features according to patent claim 11. In addition, a monolithic arrangement according to the CMOS technology with such an integrated fuel cell is advantageous. Advantageous embodiments are the subject of the dependent claims.
Accordingly, a fuel cell with a storage unit for storing hydrogen, with a proton conductive layer, which covers a surface of the storage unit; and with a cathode on a side of the proton conductive layer, which side is located opposite the storage unit, is preferred, wherein the storage unit is directly coupled with the anode, the storage unit itself forms an anode, and/or the storage unit is incorporated in or on a substrate of a semiconductor.
A fuel cell, in which the storage unit is connected to the substrate at least via a stress compensation layer, is preferred. The stress compensation layer is advantageously made of a ductile material, especially tin, gold, silver, lead, cadmium or indium. Tin is especially preferred because of its environmental compatibility.
A fuel cell, in which the stress compensation layer is formed in all areas of a direct vicinity of the substrate and of the storage unit for the hydrogen, is preferred. In other words, the stress compensation layer is formed around the storage unit towards all sides towards which the substrate is arranged adjacent to the storage unit directly or via additional inserted layers. In case of a trough structure, in which the storage unit is accommodated, the stress compensation layer correspondingly extends over the wall areas within the trough between the storage unit and the substrate.
A fuel cell in which a diffusion barrier is formed between the storage unit and the substrate such that a reduced quantity of or preferably no hydrogen can escape from the storage unit towards the substrate is preferred. The diffusion barrier may be formed for this especially from silicon nitride and/or silica and/or an oxynitride layer.
A fuel cell, in which the diffusion barrier is formed in all areas of a direct vicinity of the substrate and of the storage unit for receiving the hydrogen and optionally other layers, is preferred. Correspondingly, the storage unit is preferably additionally also surrounded by a diffusion barrier, which inhibits or preferably prevents the passage of hydrogen from the storage unit to the substrate.
A fuel cell, in which the storage Unit is completely surrounded, in a part of its circumferential area, by a proton conductive and non-hydrogen conducting layer, and in the other areas of its circumference by a diffusion barrier for hydrogen is preferred. Such an arrangement ensures that hydrogen cannot diffuse over time from the hydrogen storage unit.
Preferred is a fuel cell, in which the storage unit is in contact with one side of the proton conductive layer and in which a self-breathing air diffusion layer is formed as the cathode on the side that is opposite thereto, wherein the self-breathing air diffusion layer is sufficiently porous for the passage of air, especially for the passage of oxygen.
The storage unit, which is capable of receiving hydrogen, is preferably embedded in a cavity or pit within a substrate, preferably one consisting of silicon, wherein a stress compensation layer and/or a diffusion barrier for hydrogen is formed between a surface of the storage unit and a surface of the substrate, which said surfaces are adjacent to each other.
Moreover, a process for manufacturing a fuel cell is accordingly preferred, in which a cavity or pit is formed in a substrate, preferably one consisting of silicon; a stress compensation layer consisting of a ductile material, especially tin, and a diffusion barrier for inhibiting or blocking the passage of hydrogen, is formed in said cavity; the remaining cavity is filled with a storage unit, preferably one consisting of palladium; and the structure comprising the diffusion barrier, stress compensation layer and storage unit is preferably covered with a proton conductive and non-hydrogen conducting layer, especially a polymer electrolyte membrane, on a remaining free surface, and a cathode is formed on the opposite side of the proton conductive side, wherein the cathode allows air and/or oxygen to pass through.
Such a fuel cell can be manufactured in a surprisingly simple manner and offers a large number of advantages. Thus, the mass of the hydrogen being fed to the fuel cell can be controlled by varying the material properties of the surface of the hydrogen storage unit as well as by varying the contact surface between the hydrogen storage unit and the fuel cell. The hydrogen preferably reaches the MEA (MEA: membrane electrode unit) from the hydrogen storage unit directly by diffusion. The embodiment as a self-breathing system, i.e., the use of atmospheric oxygen from the immediate environment, makes it possible to build up the fuel cell system completely without active components such as guiding systems and valves.
Such a fuel cell, including the storage unit for the hydrogen, is especially well suited for miniaturization based on the very simple design. If the fuel cell is constructed from CMOS-compatible materials, the fuel cell including the storage unit for the hydrogen can be monolithically integrated at the chip level.
The advantage of such a fuel cell design over other electrical energy sources, for example, storage batteries, is that the capacity and the output can be set separately from one another. The capacity of the fuel cell of the new design is set by setting the layer thickness of the integrated hydrogen storage unit because of the fixed surface area and the fixed volume. The output is obtained from the contact. surface between the integrated hydrogen storage unit and the polymer electrolyte membrane.
An exemplary embodiment will be explained in more detail below on the basis of the drawings. In the drawings,
A pit, which is filled with various layers and materials, is formed in a substrate 1 made of silicon. A diffusion barrier 2, which will reduce or prevent the passage of hydrogen, is located directly adjacent to the substrate 1. On the other side of the diffusion barrier 2, a stress compensation layer 3 consisting especially of tin follows through the substrate 1. The stress compensation layer 3 is followed by a storage unit 4 as a hydrogen storage unit, which at the same time forms an anode. Palladium is preferably used as the material for the storage unit 4. The arrangement of the diffusion barrier 2 and the stress compensation layer 3 is preferably selected such that these are arranged in a trough-shaped manner in the pit and end at the same level with their upper, outside edges as a surface of the substrate 1, which is adjacent thereto on the outside. The storage unit 4 also has a surface, which preferably ends flat with the surface of the substrate 1, diffusion barrier 2 and stress compensation layer 3. As a result, the storage unit 4 is completely surrounded first by the stress compensation layer 3 and then by the diffusion bather 2 in the direction of substrate 1.
When the storage unit 4 is loaded with hydrogen or when the storage unit 4 is unloaded, same undergoes an expansion or contraction by up to about 12%. The stress compensation layer 3 consists of such a material and is of such a dimension that the expansion and contraction of the storage unit 4 is compensated such that separation of the arrangement from the substrate 1 and/or cracking in the substrate I are prevented from occurring. Preliminary experiments showed that, for example, a stress compensation layer 3 with a layer thickness of 70 μm is suitable at a layer thickness of 130 μm for the storage unit 4.
Above this arrangement is located an anode contact 5, which extends from a top-side edge area of the storage unit 4 via the top-side edge sections of the stress compensation layer 3 and the diffusion barrier 2 to the surface of the substrate 1 for electrically contacting the anode formed by the storage unit 4 in order to make possible the electrical connection of the anode. In addition, the surface of the entire arrangement comprising the storage unit 4 and the top-side or outside edge sections of the stress compensation layer 3 surrounding same and of the diffusion barrier 2 is covered with a membrane 6, which is designed as a polymer electrolyte membrane or proton conductive layer. Membrane 6 is preferably covered completely by a cathode 7 acting as a second electrical connection contact on the side of the membrane 6 that is located opposite this arrangement and is hence the outer side. Cathode 7 preferably extends, on at least one circumferential edge section, laterally from the membrane 6, up to the substrate 1 and extends over a certain section in parallel over the surface of the substrate 1 in order to form a connection contact point. An electrical user 8 can thus be connected to the cathode 7 and to the anode contact 5 in order to be supplied with electric power.
In case of the especially preferred arrangement, the storage unit 4 is s hydrogen storage unit consisting of palladium Pd, which is partially filled with hydrogen. When an electric load, such as the user 8, is connected, PdHx is correspondingly reacted into Pd+Hx in the storage unit 4. Transition of hydrogen to protons and electrons takes place in membrane 6. Water, H2O, is formed by the reaction ½O2+2H++2e− by the reaction with atmospheric oxygen, which is fed in on the outside via the cathode 7.
In such an arrangement, the storage unit 4, based on palladium, is applied directly to the silicon-based substrate 1 and is rigidly connected to same. The hydrogen storage unit can thus be directly integrated on a chip. To ensure long-term stability of hydrogen storage in the palladium of the storage unit 4, all the surfaces that have no contact with the fuel cell, i.e., the membrane 6, are shielded with the diffusion barrier 2 over the environment in order to suppress or prevent the diffusion of hydrogen.
To prevent hydrogen from escaping in the area of the top side, the membrane 6 covers the entire surface of both the storage unit 4 and of other adjoining components or layers in the surface area up to the diffusion barrier 2. The membrane 6 is preferably gas-tight now for hydrogen and is coupled with the storage unit 4 over the full surface area.
Such a diffusion barrier 2, which will be used as a hydrogen diffusion barrier layer, can be preferably deposited by means of silicon nitrite layers or oxynitride layers by means of CVD (Chemical Vapor Deposition). Good adhesion to the substrate 1 consisting of silicon was achieved in preliminary experiments by a combination of polysilicon and palladium with subsequent formation of palladium silicide.
The stress compensation layer 3 is additionally placed between the diffusion barrier 2 and the storage unit 4 in the embodiment being shown compensate or at least sufficiently reduce stresses that develop because of an enlargement of the volume of the storage unit 4 consisting of palladium during loading with hydrogen. CMOS-compatible solutions are hereby made possible in manufacture. The mechanical stresses developing at the interface to the substrate 1 consisting of silicon during the loading of the integrated storage unit 4 with hydrogen and during the unloading of hydrogen from said storage unit 4 are compensated by a tin layer preferably deposited by electroplating, which forms the stress compensation layer 3.
As an alternative, the stress compensation layer 3 and the diffusion barrier 2 may, however, also be formed and arranged in the reverse order between the substrate 1 and the storage unit 4.
In such an arrangement, the hydrogen being stored in the palladium of the storage unit 4 diffuses in the atomic form to the boundary surface between the coupled membrane 6 and the storage unit 4 after the connection of a load such as the user 8. Based on the catalytic action of palladium, the hydrogen dissociates into a proton and an decimal. The protons migrate through the polymer electrolyte membrane, whereas the electrons reach the cathode 7 of the fuel cell via the user 8 to be operated. The protons react at the cathode 7 with the electrons and the atmospheric oxygen from the environment to form water.
The stress compensation layer 3 will have the lowest modulus of elasticity possible in order to reduce the mechanical stresses towards the substrate 1. To compensate the mechanical stresses of the palladium storage unit, a material that is as ductile and as reversibly deformable as possible is therefore selected for the stress compensation layer 3. In addition, the material of the stress compensation layer 3 will have a good adhesive strength both to the storage unit 4 consisting of palladium and to the substrate 1 consisting of silicon or the thin layers, which are applied to the substrate 1 and which are usually inserted. In addition, materials are preferably used that can be deposited according to methods employed in semiconductor technology or compatible methods or can be manufactured, in case of greater layer thicknesses, by means of thick-layer processes, for example, electroplating, screen printing or casting processes. In addition, the materials used to form the stress compensation layer 3 will preferably be environmentally compatible. These conditions are met by ductile materials such as especially gold, silver, lead, cadmium, indium or tin. Tin is especially preferred in terms of environmental compatibility and cost-effective manufacture. A fuel cell system which is formed essentially from silicon, palladium, a polymer electrolyte membrane with palladium current collector and tin is made available as a result. These selected materials advantageously are materials that do not harm the environment.
The diffusion barrier 2, i.e., the barrier for hydrogen, is formed at the walls of the pit 11 in a next process step. This may be carried out especially by preparing an oxide layer from SiO and the subsequent preparation of the nitride layer proper by CVD to form Si3N4.
A bonding agent layer, preferably one consisting of polysilicon and being preferably prepared by CVD processes, and an electroplating starting layer, preferably one consisting of palladium and being preferably applied by means of PVD processes (PVD: Physical Vapor Deposition), are subsequently preferably deposited. The Pd silicide formation proper takes place by tempering. Polysilicon is used according to the standard CMOS process as an adhesive layer for palladium. Palladium is formed at the boundary to the palladium silicide, because polysilicon is preset under it. The polysilicon thus forms a bonding agent layer 12. This is used primarily for contacting, and, above all, reinforcement with aluminum is possible as well.
In another process step shown the stress compensation layer 3 proper is formed in the pit 11 on the layer structure located therein. After masking the silicon surface of the substrate 1 by means of usual lithographic processes, the stress compensation layer 3 proper consisting of Tin Sn is applied by means of an electroplating process.
The storage unit 4 proper, consisting of palladium Pd, is formed by means of a usual Pd electroplating process in the remaining pit 11 after the preparation of the stress compensation layer 3. Planarization of the surface is finally performed by means of a polishing machine in order to obtain a uniform surface over the substrate 1, storage unit 4 and edges of the stress compensation layer 3, which reach the surface between these, and the diffusion barrier 2. The storage unit 4 is ultimately contacted by, for example, a corresponding masking of the surface by means of a lithographic process, the preparation of a gold layer to form the contacts by means of, e.g., PVD by vapor deposition and ultimately structuring of gold strip conductors from the gold layer 9.
After the palladium has been filled into the pit 11 as an anode or storage unit 4 and before the membrane 6 is applied, the storage unit 4 is filled with hydrogen. Hydrogen now diffuses into the palladium. The proton conductive membrane 6 is subsequently applied as a clover, which does not let hydrogen through, on the one hand, but does let protons through, on the other hand. The structuring may also be carried out by means of usual process steps from CMOS processes, for example, by the use of a RIE process (RIE: Reactive Ion Etching). The use of auxiliary layers, which may be used as a protective function for the silicon substrate, is advantageous. Adhesion of the polymer electrolyte membrane to the palladium storage unit is preferably achieved by the use of a bonding substance, which is added to the polymer dispersion. Atomic hydrogen is thus advantageously already present in the palladium of the storage unit 4 or of the anode before coverage with the membrane 6.
The polymer electrolyte membrane is subsequently formed as the membrane and the cathode 7 is formed on the surface of the substrate 1 or of the materials introduced into the pit 11. To design a semiconductor technological embodiment of a self-breathing air diffusion layer, this is preferably constructed on the membrane 6 from a current collector for electrical return to the substrate 1 and from a fine, catalytically active palladium lamellar structure. The palladium lamellar structure at the same time forms the self-breathing air diffusion layer and the cathode 7 of the fuel cell. Microstructured sputtering masks, which are prepared by means of deep etching process known as advanced silicon etching from, for example, silica wafers having a thickness of 300 μm, may be used to structure the palladium current collector and the palladium air diffusion electrode. The web width of the lamellar structure formed was 100 μm in preliminary experiments. Such a microstructured palladium catalyst is highly porous and hence permeable to air. A CVD silicon nitride layer was used as a suitable electrical insulation layer between the anode and cathode according to preliminary experiments.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2007 048 807.8 | Oct 2007 | DE | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 12249699 | Oct 2008 | US |
| Child | 14221955 | US |
