ELECTROCHEMICALLY FUNCTIONAL MEMBRANES

Abstract

A system includes an electrochemically functional membrane, and a support structure constructed and arranged so as to support the membrane while leaving within the membrane a chemically active area having an area utilization of at least about 50%. In some embodiments, the support structure may include a plurality of grids that are sized and shaped so that the contact area between the grids and the membrane is reduced to less than about 40%. In some embodiments, the support structure may include aerogels, for example PVA-reinforced CNT aerogels having a conductivity that is increased by pyrolysis. The system may be a gas separation system; a gas production system; a gas purification system; or an energy generation system such as an SOFC.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is based upon, and claims the benefit of priority under 35 U.S.C. §119(e), from U.S. Provisional Patent Application Ser. No. 61/378,900 (the “'900 provisional application”), filed Aug. 31, 2010, entitled “Large-Area Nanostructured Membranes For Solid Oxide Fuel Cells”; and from U.S. Provisional Patent Application Ser. No. 61/420,319 (the “'319 provisional application”), filed Dec. 6, 2010, entitled “Electrochemically Functional Membranes.” The contents of the '900 provisional application and the '319 provisional application are incorporated herein by reference in their entireties as though fully set forth.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

This invention was made with government support under grant NSF PHY-0646094 awarded by the National Science Foundation. The government has certain rights in the invention.

BACKGROUND

Electrochemically functional thin membranes are useful in a variety of electrochemical applications, examples of which include gas separation; gas purification; gas production; and renewable energy generation.

For these electrochemically functional membranes, the thinner a membrane, the lower its ionic resistance, and the lower the energy needed to heat it up. The thicker the electrochemically functional membrane, the higher the required operational temperature. Conventional technologies involving the above applications typically use membranes that have a thickness on the order of micrometers. High operational temperatures are required, typically on the order of 800 degrees Celsius or higher.

As one example, solid oxide fuel cells (SOFCs) are attractive direct electrochemical energy conversion devices used in renewable energy generation applications. Unfortunately, traditional SOFCs must be operated at very high temperatures, typically 800-1000° C., to be effective. Such high operation temperatures have been limiting the application of SOFCs to stationary applications. Also, high operation temperatures add significant constraints on the choice of materials and packaging methods. At such high temperatures, the choice of materials for the electrodes and structural components of the fuel cell (for example membrane support, gas handling and sealants) is severely constrained, particularly in the reactive oxygen- and hydrogen-rich environments of a fuel cell. Even very stable materials will corrode under these conditions and/or fatigue and fail under the stress of thermal cycling. Accordingly, methods and systems for operating SOFCs below 600° C. or so must be developed in order to be able to use this technology in a variety of applications.

The use of nanoscale thin film electrolytes can be an attractive option to lower the SOFC operation temperature, since reducing the electrolyte thickness results in a significant decrease of ohmic resistance. Such thin electrolyte membranes are extremely fragile on their own, however, and rupture easily as a result of slight pressure, vibration or thermal stress.

There have been studies on solid oxide fuel cells utilizing nanoscale thin film electrolytes. For example, power density up to 100-400 mW/cm² has been reported in the temperature range below 600° C. These results, however, were performed on very small free standing membranes, therefore total power output was also very small. This size limitation comes from the mechanical stress of the thin film cathode-electrolyte-anode assembly. Area utilization was extremely limited due i.a. to the constraints of KOH¹ etching. This has been one of the technical difficulties encountered during efforts to increase total power output of thin films. ¹Potassium Hydroxide

Therefore, there is a need for methods and systems for increasing the yield and electrochemical performance for the above-described electrochemically functional membranes, in particular for nanostructured membranes.

BRIEF DESCRIPTION OF THE DRAWING FIGURES

The drawings disclose illustrative embodiments. They do not set forth all embodiments. Other embodiments may be used in addition or instead.

FIGS. 1A-1D illustrate the lithographic processes for forming grid-supported membranes that have plated nickel grid supports, in one or more embodiments of the present disclosure.

FIGS. 2A-2D provide optical images of a nickel-grid supported membrane device.

FIG. 3 illustrates the electrical performance of the nickel grid-supported device shown in FIGS. 2A-2D.

FIG. 4A illustrates the dimensions of a Pt (platinum) grid supported structure with an active area of 17.8 mm², in one or more embodiments of the present disclosure.

FIG. 4B compares the structure of FIG. 4A to a typical Si (silicon) supported fuel cell structure.

FIG. 5 is a plot of the power density and power output as a function of current density for the Pt grid supported structure shown in FIGS. 4A and 4B.

FIGS. 6A-6D illustrate processes for fabricating a current collector and strengthening grid on the underside of a thin film SOFC, in one or more embodiments of the present disclosure.

FIGS. 7A-7C illustrate processes for fabricating a current collector and strengthening grid by electrodeposition on a thin seed layer.

FIGS. 8A-8E illustrate processes used to form aerogel-supported membranes, in one or more embodiments of the present disclosure.

FIG. 9 is a schematic flow chart illustrating a method of creating PVA (polyvinyl alcohol)-reinforced CNT (carbon nanotube) aerogels, in one or more embodiments of the present disclosure.

DESCRIPTION

In the present disclosure, methods and systems are disclosed relating to the fabrication and use of nanostructured membranes with large active areas are described. Potential applications include: renewable power generation applications, such as SOFCs, for example; gas separation technologies; gas purification; and gas production, such as oxygen production, for example.

It is to be understood that the inventive subject matter in the present disclosure is not limited to the particular embodiments described below, as such may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting, since the scope of the inventive subject matter will be limited only by the appended claims. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the inventive subject matter belongs.

Although any methods and materials similar or equivalent to those described herein can also be used in the practice or testing of the inventive subject matter disclosed below, a limited number of the exemplary methods and materials are described.

In overview, a system may include an electrochemically functional membrane, and a support structure constructed and arranged so as to support the membrane while leaving within the membrane a chemically active area having an area utilization of at least about 50%. In some embodiments, the support structure may include a plurality of grids that are sized and shaped so that the contact area between the grids and the membrane is reduced to less than about 40%. In some embodiments, the support structure may include one or more aerogels.

In some embodiments of the present disclosure, a grid-based support structure is used to increase the active area of electrochemically functional membranes. Support provided by plated nickel grids was found to provide over 60% active membrane area. Devices that can use such grids to support large membranes, for example on the order of about 5 mm², and ultra-thin cathodes include without limitation: gas separation devices; gas production devices; gas purification devices; and energy generation devices. Examples of ultra-thin cathodes include La_0.6Sr_0.4Co_0.8Fe_0.2O_3-δ (LSCF) cathodes.

As described below, plated nickel grid supports were fabricated that are able to provide over 60% active membrane area. In one embodiment, a nickel grid was first deposited on the cathode side of the membrane. This grid could stabilize a 180-nm-thick LSCF/YSZ (yttrium-stabilized zirconium)/porous Pt membrane over a 5-mm square area. In particular, a grid of plated nickel on the cathode with 5-10 μm line width and 25-50 μm pitch could be used to support a roughly 200-nm-thick LSCF/yttrium-stabilized zirconium/platinum membrane, while covering less than 20% of the membrane area. Measurements of their electrochemical performance up to 500° C. showed that this geometry yielded a maximum performance of 1 mWcm⁻² and 202 mV open-circuit voltage at 500° C. Larger power output would be possible with optimized device processing.

FIGS. 1A-1D illustrate lithographic processes used to form the above-described plated nickel grid supported membranes. In the illustrated embodiment, a silicon wafer 170 was used that was coated on both sides with 200 nm of LPCVD (low pressure chemical vapor deposition) silicon nitride 160. 80 nm of 8 mol % YSZ 150 and 20 nm of LSCF 140 were sputtered onto the top surface. Both films were deposited at 1 nm min⁻¹ from oxide targets, in 5 mT of argon with the substrate held at 500° C.

An LOR (lift-off resist) 120 was then spun onto the sample, followed by photo resist 110, baking the layers at 170° C. and 115° C., respectively. In the illustrated embodiments, LOR-3A lift-offresist (Microchip Corp.) and S1813 photo resist (Rohm and Haas Co.) were used. Other embodiments may use different types of LOR and photo resist.

The resist was exposed with the grid pattern, and the S1813 was developed in CD-30 (Rohm and Haas Co.) and baked at 140° C. The LOR 120 was then etched in CD-26 (Rohm and Haas Co.), just long enough to reach the LSCF surface but no longer, to avoid significant undercutting. A physical mask was then placed over the chip, covering the edges, while a thin (˜100 nm) nickel seed layer 130 was sputtered onto the surface. The resulting product is shown in FIG. 1A.

The photo resist was then dissolved in acetone, lifting off the unwanted nickel seed material but leaving the LOR layer 120 intact. Next the chip was immersed in electroless nickel plating solution (Nickel, Transience Corp.) at 88° C. to plate 5-10 μm of nickel 180, as shown in FIG. 1B.

The LOR layer 120 was then dissolved in Remover PC (Microchip Corp.), then the back side nitride was patterned via reactive ion etch (RIE) using a mixture of CF₄ and O₂ through a physical mask, followed by etching through the wafer in 30 wt % KOH at 96° C., as shown in FIG. 1C. During this etch the chip was sealed into a stainless steel housing with o-rings to ensure that KOH only touches the bottom of the chip.

Finally, the silicon nitride on the back side and underneath the membrane was removed with another RIE step, and a 0-60 nm blanket coat of YSZ 150 is sputtered onto the back side (as previously but at room temperature) followed by a porous Pt anode 190 (sputtered at room temperature in 75 mT of argon). The completed device 100 is shown schematically in FIG. 1D.

The above-described patterning procedure for the nickel grid accomplishes several goals. First, the nickel plating solution can slowly etch LSCF. Thus, the LOR protects the LSCF over what will become active membrane areas, and undercutting of this layer (which would lead to a gap between LOR and nickel seed coverage) is minimized. Second, with no nickel seed near the edge of the chip, LSCF is removed there, minimizing the possibility of electrical shorts from Pt that may deposit up the side of the chip. Third, the thickness and therefore strength of the nickel grid may be increased without usurping additional active membrane area. Because plating is roughly isotropic, the grid lines grow outward as well as upward during plating. However, due to the LOR layer, the LSCF is still exposed to gas underneath this overhang.

In some embodiments of the present disclosure, a relatively thin (−300 nm) LOR layer is used, which may inhibit diffusion since the overhang may be as much as 10 μm wide. In other embodiments, a thicker LOR layer may be used.

FIGS. 2A-2D show optical images of a grid-supported membrane device, fabricated using the procedures described above in conjunction with FIGS. 1A-1D. In particular, FIG. 2A and FIG. 2B illustrate completed devices, whereas FIG. 2C and FIG. 2D illustrate the devices after KOH etching but prior to the final RIE step. In FIG. 2A, the entire chip is shown from above, with the active area visible as the lighter square in the center. The edge exclusion, i.e. the region where no nickel seed layer was deposited and thus the LSCF was partially etched, is seen as the dark border.

In FIG. 2B, the same chip is shown from below, coated with porous Pt on the unpolished, undetached bottom edge of the wafer, down the sloped sidewalls of the KOH etch (dark), and on the active area in the center. Zooming in, FIG. 2C shows the result of plating roughly 5 nm of nickel onto a 25-μm-pitch, nominal 5-μm-line width grid pattern. The nickel is relatively flat (bright area in the image) over approximately 5 μm but then slopes down (dark area in the image) for a total width of 13 μm. In addition, occasional spontaneous plating nucleation sites can be seen, covering part of the active area with unwanted nickel.

Considering only the image shown in FIG. 2C, it would appear that this grid has covered over 75% of the membrane area. However, the bottom view, namely FIG. 2D, shows another advantage of the LOR layer: the actual nickel grid lines in contact with LSCF are only about 3 μm wide and cover only 22% of the active area. The rest of the nickel is lifted off the LSCF surface, which alters the color of reflected light due to interference effects. The spontaneous nucleation sites can also be seen, generally lifted slightly higher off the surface, but close to flat on the bottom, as evidenced by the interference patterns.

The amount of active area covered by the grid lines may be further decreased by increasing the grid pitch at constant line width. With currently used process parameters, however, a 25 μm grid is more reliable than a 50 μm grid, although chips have not emerged from processing completely defect-free.

FIG. 3 illustrates the electrical performance of a grid-supported device having a 5-mm square active area and a 50-μm nickel grid. The observed open-circuit voltages (V_OC) were low enough (maximum 202 mV at 500° C.) that the I-V curve at any given time was essentially a straight line connecting the open-circuit and short-circuit conditions. This reflects the fact that with a low V_OC, the current produced by the cell is substantially less than the internal leakage current, thus the activation over potential is essentially constant due to its logarithmic dependence on current. Also, in a situation like this reactants are significantly oversupplied to the cell, such that concentration losses do not arise, and only linear contributions remain in the current-voltage relationship. Each I-V curve can then be described by two parameters; namely V_OC and area-specific resistance (ASR), but others can be calculated: short-circuit current Isc=V_OC/R_M peaks at 19 mA cm⁻² at 500° C., and maximum power P_max=Isc V_OC/4 reaches 1.0 mWcm⁻².

As seen from FIG. 3, V_OC shows an overall trend toward higher values at higher temperatures, which is to be expected since V_OC is determined by the ratio of ionic to leakage resistance, as described above, and the ionic conduction is thermally activated. However, no single activation energy may be extracted. Various slopes are seen, corresponding to activation energies around 0.43 eV in heating up to 400° C., 0.22 eV cooling from 500° C. to 400° C., and 0.8 eV cooling below 400° C.

An apparent activation energy of 3 eV is seen during the rapid performance improvement around 450° C. in heating. This may be related to a change in morphology of one or more layers of the membrane. Although the LSCF and YSZ were deposited at 500° C. and thus are expected to already be crystalline, the improvement at 450° C. suggests that one or both of these layers required further structural transformation to achieve better electrochemical performance.

In some embodiments of the present disclosure, nanostructured metal grids are used to greatly increase the active fuel cell area of micro solid oxide fuel cells (μ-SOFCs). μ-SOFCs with nanometer scale electrolytes have attracted significant attention in recent years in the context of mobile energy. Reducing the electrolyte thickness results in a decrease of ohmic resistance, thus the use of nanoscale thin film electrolytes makes it possible to lower the operation temperature of SOFCs to less than about 600° C.

For renewable power generation applications, one challenging aspect is scaling of the power generating membrane active area. In the case of micro solid oxide fuel cells (μ-SOFCs), high power density has been achieved in free standing membranes with small areas (typically about 0.01 mm²), due to large stresses originated from thin film membrane electrode assembly and limited area utilization (mostly <5%) resulting from a constraint from KOH etching of Si substrate support. For 550 μm thick Si substrates, the area utilization was only 1.3%. There is a need for significantly scaling up the active membrane areas in μ-SOFCs.

In one embodiment of the present disclosure, nanostructured metal grids were used to form large area μ-SOFCs with an active area larger than 10 cm² with an area utilization over 85%. First, 50-150 nm 8 mol % yttrium doped zirconium (electrolyte) and LSCF (cathode) were deposited on a silicon nitride coated Si wafer at about 550° C. by RF (radiofrequency) sputtering. Metal grids (Pt and/or Ag (silver)) of about 1 μm thick were created by the combination of photolithography and metal deposition by DC-sputtering on the surface of cathode. These metal grids provided mechanical support for nanoscale thin film structure and cathode current collectors.

Grid shapes, sizes and widths influence the stability of the fuel cells. A number of different shapes of the grids were tested, including square, circle, triangle, and hexagonal shapes. In one embodiment, the height of grids was about 1-2 μm thick, and the width of grids was 5-20 μm. In another embodiment, the sidelength/diameter of grids in the range from 50 μm to 500 μm was tested. As a result, circular and hexagonal shapes were found to be stable.

Following grid formation on the surface of the cathode, the back-side pattern was formed via reactive ion etching and via wet etching of Si substrate in 30 wt % KOH at 96° C. for 3 hours. Finally, a porous metal layer, for example Pt/Ni, was sputtered onto the backside to form an anode layer.

The above-described processes successfully formed large area μ-SOFC devices with a high power output, namely grid-supported μ-SOFCs with an active area larger than 10 cm² with an area utilization over 85%. The power density of grid-supported fuel cells was tested with a cell with ˜14 mm² and the open circuit voltage 0.4 V with a power output 4 mW was achieved at 570° C. with porous Pt anode.

FIGS. 4A and 4B illustrate membrane assemblies that are supported with metallic Pt grids, with an active area of about 17.8 mm². The dimensions of a Pt grid supported structure 405 are illustrated in FIG. 4A, which shows that the Pt grid supported structure 405 has an active area of 17.8 mm². This active area was achieved by utilizing sub-100 nm electrolyte.

In the illustrated embodiments, Pt films were deposited by DC-sputtering at room temperature. The comparison between the structure 405 illustrated in FIG. 4A to a typical Si supported fuel cell structure 410 is shown in FIG. 4B. For the SOFC 405 illustrated in FIG. 4A, area utilization was improved from 1.3% to 63%, where the 63% results from the following formula: KOH etch utilization 71.2%× grid structure utilization 89%.

While the grids can be of any suitable shape, suitable stable shapes include hexagons and circles, as described above. The film was supported by the grid and stable even after annealed at 450-550° C. for 2-3 hours. This approach is scalable to a wafer level as well. The performance of these Pt grid supported fuel cells were tested with 5 mm×5 mm square samples.

FIG. 5 illustrates the power density and power output as a function of current density, for the membrane assemblies illustrated in FIGS. 4A and 4B. As shown in FIG. 5, a power density of 28 mW/cm² was achieved. The total power output of 4 mW was achieved from 5 mm×5 mm, which is about two orders of magnitude larger than the power density reported from small 100 μm×100 μm structures.

The embodiments described above may constitute major milestones for commercializing micro-SOFC technology for various applications having space and weight limitations.

In some embodiments of the present disclosure, conducting supports were used as strengthening grid and current collector for SOFCs. In particular, a current collector and strengthening grid were made on the underside of a thin-film, wafer-supported device, without any patterning or lithography from the back side. In these embodiments, no patterning or lithography needed to occur after membrane release, when the device is fragile and prone to failure. Another advantage of this approach was that the rear electrode does not need to contact the wafer support at the edge of the active area, because the grid naturally performed this function.

While exemplary embodiments are described in which the device is a solid oxide fuel cell, the methods and systems described below are applicable to thin membranes used in a wide variety of electrochemical applications, including without limitation gas separation, gas purification, gas production, and energy production technologies

Although methods and systems are described that use a silicon wafer insulated by silicon nitride, other materials, including without limitation an electrically conducting wafer with an etch-stop coating, may be used as well.

FIGS. 6A-6D illustrate processes for fabricating a current collector and strengthening grid on the underside of a thin film SOFC. First, a silicon wafer 610 is provided that is insulated by silicon nitride 620 on the top and bottom. As shown in FIG. 6A, the top nitride is patterned with the support grid pattern. However, this pattern ends outside the membrane active area so that current collected in the grid can flow down into the support wafer.

On top of this patterned nitride, a thick layer of a material 630 that is to be used as the current collector grid is deposited. The grid material 630 may be silicon, amorphous or polycrystalline, or a metal like nickel. The grid material 630 has a thickness sufficient to provide physical support as well as current collection, namely a thickness on the order of several microns.

The electrolyte 640 and top electrode 645 layers are deposited on the layer of material 630. Another thin insulation layer may be deposited in between, optionally. At an appropriate time, the back side nitride is patterned to expose the area to be etched. The wafer has the appearance illustrated in FIG. 6A. In addition, a current collector grid aligned with the back side grid may also be deposited on the top surface.

As illustrated in FIG. 6B, the wafer 610 is then etched through from the back side. The grid material 630 is etched as well. In some embodiments, the grid material etches isotropically under the nitride pattern. One way to improve the timing accuracy is to etch only partway through the top nitride layer in the first step. The wafer etch could then be allowed to go to completion, then the nitride would be etched again, enough that the partially etched areas are exposed but the undetached areas remain, to mask a final, shorter and slower etch of the grid material. Other embodiments may use other methods to improve etch timing accuracy.

Finally, the remaining nitride is removed from the back side of the wafer, as seen in FIG. 6C. If another nitride layer was deposited between the grid material and the electrolyte, this is removed here as well, then the bottom electrode material 650 is deposited. The bottom electrode does not necessarily directly contact the support wafer, but it does contact the grid lines and these contact the support wafer.

FIG. 6D shows details of the top nitride pattern. The grid lines extend beyond the edge of the active area to be exposed by etching (shown as a dashed rectangle), but end shortly outside the area, in order to connect current to the wafer support.

A related technique is illustrated in FIGS. 7A-7C. In the illustrated embodiments, the current collector grid is built up by electrodeposition on a thin seed layer 730. This seed layer 730 is deposited and patterned on a silicon wafer 710, followed by the electrolyte layer 740 and the top electrode layer 745. A nitride layer 720 may be deposited either before or after the seed layer 730. If it is deposited before the seed layer, the edges of the nitride is exposed so that the seed layer will electrically contact the support wafer.

In the step illustrated in FIG. 7A, the etch mask is patterned on the back side. A top side grid may be deposited as well. This may be useful, because the bottom grid will not be very thick when the membrane is released, so it may not be able to support the membrane.

The wafer 710 is then etched through and remaining nitride is removed, as illustrated in FIG. 7B. This process also removes any extra top nitride layer from the active area of the membrane, whether it was deposited before or after the seed layer.

In the step illustrated in FIG. 7C, the extra material 730 is electrodeposited on the back side, followed by deposition of a bottom electrode material 750. Again, the thickness of the extra material 730 may be on the order of several microns.

The techniques described above may be used in a wide variety of applications in which thin membranes that are electrochemically functional may be used. In addition to energy generation technologies, such as the SOFC example described above, these applications include without limitation: gas separation systems, gas production systems, and gas purification systems. Examples of gas separation technologies may include splitting water to produce hydrogen, or separating oxygen from air.

The structure, function, fabrication and use of gas separation, gas production, and gas purification systems are described for example in a monograph entitled: “Industrial Gas Handbook: Gas Separation and Purification,” by Frank G. Kerry, CRC Press Feb. 22, 2007. The contents of this reference is hereby incorporated by reference in its entirety.

In some embodiments of the present disclosure, aerogel-supported membranes have been constructed to increase the active membrane area. One option for supporting submicron membranes is to essentially reverse the conventional electrode-supported geometry by depositing the membrane on a smooth substrate, covering it with a porous material, and then removing the substrate.

Desirable properties of porous support material include: high diffusivity and a pore size substantially smaller than the maximum allowable freestanding membrane width. Also, it is desirable that the porous support material adhere to the membrane surface without wetting and blocking too many active sites, that it possess sufficient mechanical strength to counteract film stresses generated during either deposition or heating to the operating temperature, and that it be stable at high temperatures under possibly changing oxidative environments. Further desirable properties of porous support material include good electrical conductivity, so that it can function as a current collector, and good catalytic activity.

In some embodiments of the present disclosure, aerogels are used as support material In particular, silica acrogel reinforced with carbon fibers, formed in situ between the membrane surface and a silicon cover piece, is used to support electrochemical membranes in one or more embodiments.

FIGS. 8A-8E illustrate the processes used to form aerogel-supported membranes, in some embodiments of the present disclosure. The assemblies are formed from two chips 850, a top chip and a hot torn piece (bottom chip). Both start as 10-mm square silicon chips 850, coated on both sides with 200 nm of low pressure chemical vapor deposited (LPCVD) silicon nitride 840. The top chip, when bonded to the bottom chip, can serve as a flow channel for air and a container for the aerogel.

As shown in FIG. 8A, a large square of nitride 840 is removed from one side of the top chip 850 and two long slits from the other side, followed by etching in KOH from both sides until the etch holes meet mid-wafer.

FIG. 8B illustrates the next step, in which the bottom chip 850 is coated with YSZ 830 and LSCF 820 and a hole opened in the back side nitride 840 as described above. An adhesion layer of Cr and a bonding layer of Au is electron-beam evaporated onto the top of the bottom chip 850, masked to deposit only around the edge where the two chips will be bonded.

The remaining silicon nitride is removed from the top chip with phosphoric acid (85% in water, heated to 180° C.) and the same Cr/Au bonding layer is evaporated to its bottom surface, after which the chips 850 are bonded together in an EV Group Inc. EV501 wafer bonder, at 350° C. and 1000N for 30 min at 10⁻⁵ torr. The resulting structure is shown in FIG. 8C.

The top chip serves as a container for the aerogel. This leaves an open space inset into one side of the top chip in which the aerogel is located, accessible by two slits on the other side for gas inlet and outlet.

During the next step, illustrated in FIG. 8D, the space between the top and bottom chips is filled with a porous medium. In the illustrated embodiment, the porous medium is silica aerogel reinforced with carbon fibers In one or more embodiments, carbon fiber reinforced silica aerogels can be synthesized in a two-step acid-base catalyzed process.

To form the gel/fiber composite, in some embodiments milled carbon fiber (Toho Tenax Co.) can be manually packed into the two-chip fuel cell assembly with a thin rectangular foil until the space is filled. In one embodiment, commercially available prepolymerized tetraethyl orthosilicate (Silbond Corp., H5), ethanol, water, and NH₄OH (30% in H₂O) were mixed in a 1:1, 7:1, 5:0, 007 volume ratio, then directly pipetted into the fuel cell assembly containing carbon fiber. The mixture was left in ethanol-saturated air to gel for 1 h, and then submerged in a bath of water, ethanol, and 30% NH₃/H₂O in a 1:1:0.003 ratio for 18-24 h, to age the gel. Ethanol replacement, then CO₂ replacement followed which minimized capillary forces during drying.

In one embodiment, supercritical drying may be performed in a custom-built autoclave, constructed from a 2″ ID steel elbow as a pressure chamber, fed by a siphon-type carbon dioxide cylinder and submersed in a water bath connected to a recirculating thermoelectric temperature controller.

After the gel is dried, the structure can be completed using the same processing steps as in the grid support structures described above, as seen in FIG. 5E: KOH etching, nitride removal, optional deposition of YSZ 830, and deposition of porous Pt anode 870. One difference is that RIE, despite coming from the opposite side of the chip, could cause significant damage to the nanofibers of the aerogel, so in the embodiment described above, the back side nitride is removed by wet etching in phosphoric acid, instead of RIE, while the top and sides of the device are still protected inside the stainless steel housing.

As described above, one approach toward realizing large area fuel cell junctions may include depositing a membrane on a smooth substrate, covering it with a high-porosity material formed in situ, then removing the substrate. A composite of silica aerogel and carbon fiber may be used as the support. As shown above, this material can be created in flow channels etched into the underside of a silicon chip bonded to the top of the membrane.

Other types of aerogels, for example carbon aerogels synthesized with resorcinol and formaldehyde and used with fiber support as electrodes in super capacitors may be suitable as well due to their electronic conductivity.

Further, carbon nanotube aerogels can also work well. In these gels, polyvinyl acetate (PVA) can be used to strengthen the gels. While it decreases electrical conductivity, at SOFC operating conditions the PVA may pyrolyze and enhance its conductivity. The gels can be formed as a composite with carbon fiber, since a key criterion is low shrinkage of the gel after gelation, and studies have shown that forming the gel as a composite with larger fibers accomplishes this.

In some embodiments of the present disclosure, an approach is used that simultaneously attains the strength of PVA-reinforced gels and the uniform conductivity of pure CNT aerogels. Pyrolysis of PVA-CNT aerogels is used to produce aerogels for supporting nanostructured electrochemical membranes. These aerogels have been shown to simultaneously attain the strength of PVA-reinforced gels and the uniform conductivity of pure CNT aerogels. Such PVA-CNT aerogels can have numerous applications, including but not limited to electrodes for batteries and super capacitors and supports for thin-film fuel cells.

FIG. 9 illustrates a method 900 of creating PVA-reinforced CNT aerogels, in one or more embodiments of the present disclosure. In these embodiments, CNT aerogels are created through gelling by suspension of the CNTs in water.

In step 910, gel is formed by suspending CNTs in water using sanitation and a surfactant. This is followed by PVA infiltration in step 920, during which the gel is covered with a water/PVA solution. Then, in step 930, water is replaced with alcohol and then with CO₂. In step 940, CO₂ is supevcritically removed. Finally, in step 950, the gel is pyrolyzed by heating it to between about 400 and 1000° C. in an inert atmosphere.

In some embodiments, pyrolysis is used with resorcinol-formaldehyde (RF) aerogels in order to remove atoms other than carbon, and the remaining carbon is either amorphous (at lower temperatures) or forms small graphitic clusters (at higher temperatures).

During pyrolysis, PVA loses mass between 300 and 500° C., and at this point it is composed almost entirely of carbon atoms. This process by itself can substantially increase the conductivity of the reinforced CNT aerogel, because the amorphous carbon framework, though still not highly conductive itself; is smaller than the PVA and thus the CNTs will be closer together and thus more likely to form a conductive link. RF aerogels, however, need heating to around 800° C. to become highly conductive, because at this temperature most of the carbon converts to graphite-like bonding configurations (i.e. sp² rather than sp³). In some embodiments, the gls may have to be heated as high as about 800° C. to reap the full conductivity benefits.

The above-described gels may be used to support free standing membranes for micro-fuel cells, as well as to provide a support structure for a variety of electrochemical devices utilizing free standing membranes or membranes on porous substrates. Examples include membranes that are mixed conductors used for hydrogen production or catalytic reduction or oxidation at high temperatures.

In sum, methods and systems have been disclosed for fabricating nanostructured, electrochemically functional membranes for a wide variety of applications including without limitation gas separation, gas production, gas purification, and energy generation devices. These methods and systems allow the operational temperature in these applications to be substantially reduced, and dramatically enhance the performance of these devices.

By scaling up active membrane areas and allowing for electrochemically functional membranes to be fabricated that are nanostructured but have very large active electrochemical areas, these methods and systems significantly enhance the yield and electrochemical performance of gas separation, gas production, gas purification, and energy generation devices.

The integrated devices and structures described above are useful in a wide variety of applications, and in particular will advance low-temperature devices for portable applications.

The components, steps, features, objects, benefits and advantages that have been discussed are merely illustrative. None of them, nor the discussions relating to them, are intended to limit the scope of protection in any way. Numerous other embodiments are also contemplated, including embodiments that have fewer, additional, and/or different components, steps, features, objects, benefits and advantages. The components and steps may also be arranged and ordered differently.

Nothing that has been stated or illustrated is intended to cause a dedication of any component, step, feature, object, benefit, advantage, or equivalent to the public. While the specification describes particular embodiments of the present disclosure, those of ordinary skill can devise variations of the present disclosure without departing from the inventive concepts disclosed in the disclosure.

While certain embodiments have been described of systems and methods relating to nanostructured membranes for use in gas production, gas purification, gas separation, and energy generation devices, it is to be understood that the concepts implicit in these embodiments may be used in other embodiments as well. In the present disclosure, reference to an element in the singular is not intended to mean “one and only one” unless specifically so stated, but rather “one or more.” All structural and functional equivalents to the elements of the various embodiments described throughout this disclosure, known or later come to be known to those of ordinary skill in the art, are expressly incorporated herein by reference.

Claims

1. A system comprising:

an electrochemically functional membrane; and

a support structure constructed and arranged so as to support the membrane while leaving within the membrane a chemically active area having an area utilization of at least about 50%;

wherein the support structure comprises at least one of:

a plurality of grids that are sized and shaped so as to reduce the contact area between the grids and the membrane to less than about 40%; and

one or more aerogels.

2. The system of claim 1, wherein the system is one of:

a gas separation system; a gas production system; a gas purification system; and

an energy generation system.

3. The system of claim 2, further comprising a substrate, and an electrode layer deposited on each side of the membrane.

4. The system of claim 1, wherein the grids comprise plated metallic grids, and wherein the energy generation system comprises an SOFC.

5. The system of claim 4, wherein the grids are constructed by:

depositing the membrane and a cathode layer on a substrate;

etching metallic grids by DC-sputtering on the surface of the cathode layer; and

forming a back-side pattern by reactive ion etching and wet etching of the substrate.

6. The system of claim 5, wherein the grids comprise nickel and have a line width of about 5-10 μm and a pitch of 25-50 μm; and wherein the membrane is a LSCF/yttrium-stabilized zirconium/platinum membrane having a thickness of about 200 μm.

7. The system of claim 4, wherein shapes of the grids comprise one of: a hexagon; and a circle.

8. The system of claim 4, wherein the grids comprise Pt or Ag, and have a thickness of about 1 micrometers.

9. The system of claim 8, wherein the membrane is a 8 mol % yttrium doped zirconium membrane having a thickness of about 50 to about 150 nm, and wherein the area utilization of the membrane is more than 85%.

10. The system of claim 1, wherein the aerogels comprise a silica aerogel reinforced with carbon fibers.

11. The system of claim 10, wherein the silica aerogel is created in flow channels etched into the underside of a silicon chip bonded to the top of the membrane.

12. The system of claim 1, wherein the aerogels comprise a PVA (polyvinyl alcohol)-reinforced CNT (carbon nanotube) aerogel having a conductivity that is increased by pyrolysis.

13. The system of claim 12, wherein the PVA-reinforced CNT aerogel is fabricated by:

suspending CNTs in water to form a gel;

performing PVA infiltration by covering the gel with a water/PVA solution;

replacing the water with alcohol and then with CO2;

super critically removing the CO2; and

pyrolyzing the gel by heating it to a temperature between about 400 and 1000° C.

14. A system comprising:

a substrate;

an electrochemically functional membrane deposited on the substrate;

a cathode layer deposited on the membrane; and

a plurality of supporting grids etched on the surface of the cathode layer according to a grid pattern;

wherein the grids are shaped and sized so as to support the membrane while reducing the contact area between the grids and the membrane to less than about 15%, thereby leaving within the membrane a chemically active area having an area utilization of at least about 60%.

15. The system of claim 14, wherein the grid pattern ends outside the active area of the membrane so that current collected in the grid can flow down to the substrate.

16. The system of claim 14, wherein the substrate is a Si substrate coated with an insulating layer; and wherein the insulating layer is patterned with the grid pattern.

17. The system of claim 16, wherein the plurality of grids comprise a current collector grid etched on the patterned insulating layer.

18. The system of claim 14, wherein the membrane is a nanostructured electrolyte membrane comprising yttrium doped zirconia; and wherein the cathode layer comprises LSCF.

19. The system of claim 14, wherein the substrate has a back-side pattern etched thereon by reactive ion etching and wet etching.

20. The system of claim 14, wherein the grids comprise one of: a metal, and a semiconductor material; and wherein the shapes of the grids comprise one of: a hexagon; and a circle.

21. A method of fabricating grids for supporting an electrochemically functional membrane in a device, the method comprising:

depositing a layer of the electrochemically functional membrane on a silicon wafer that is coated both sides with silicon nitride;

sputtering a layer of metal onto the surface of the wafer, and patterning the layer of matter into a grid pattern using photolithography;

patterning the nitride from the back side, and etching the wafer through; and

removing the silicon nitride from the back side of the wafer,

wherein the grid pattern shapes a plurality of grids so as to decrease the contact area between each grid and the membrane to less than about 40% of the total area of the membrane.

22. The method of claim 21, wherein the grid comprises a current collector grid for the device.

23. A method of increasing the active area of an electrochemically functional membrane, the method comprising:

depositing the membrane on a substrate;

covering the membrane with a high porosity material; and

removing the substrate.

24. The method of claim 23, wherein the high porosity material comprises silica aerogel reinforced with carbon fibers.

25. The method of claim 24, wherein the aerogel is created in flow channels etched into the underside of a silicon chip bonded to the top of the membrane.

26. A method of fabricating a PVA (polyvinyl alcohol) reinforced CNT (carbon nanotube) aerogel for supporting an electrochemically functional membrane in a system, the method comprising:

suspending CNTs in water to form a gel;

performing PVA infiltration by covering the gel with a water/PVA solution;

replacing the water with alcohol and then with CO2;

super critically removing the CO2; and

pyrolyzing the gel by heating it to a temperature between about 400 and 1000° C.