Gasless and gas bubble-free electrodes

Abstract

Gas bubble-free electrodes are necessary for stable long-term operation of millimeter-scale electrokinetic (EK) pumps when currents exceed 10-50 μA. An accompanying Technical Advance describes EK pumps that draw 1-3 mA. We have developed gasless and gas bubble-free electrodes that can run millimeter-scale (and smaller) EK pumps continuously at high current densities. Two types of gasless electrodes based on porous carbon and ruthenium/tantalum-on-titanium oxides have been developed that are supercapacitors which store ions from a fluid electrolyte. The gas bubble-free electrodes isolate gas generated by water electrolysis of the pump fluid from the fluid channels by means of an electrically-conductive polymer. Nafion® tubing is a cationic-selective polymer that is used to pass currents and water for electrolysis at titanium and platinum surfaces. The gas bubble-free electrodes are easy to fabricate and can operate well even with typical, low-conductivity electrolytes. The gas bubble-free cathode seals to 1500 psi for high-pressure microhydraulic actuation

Description

STATEMENT OF GOVERNMENT INTEREST

This invention was made with Government support under government contract no. DE-AC04-94AL85000 awarded by the U.S. Department of Energy to Sandia Corporation. The Government has certain rights in the invention, including a paid-up license and the right, in limited circumstances, to require the owner of any patent issuing in this invention to license others on reasonable terms.

TECHNICAL FIELD

The present invention relates to an embodiment comprising a miniature microfluidic transducer, and particularly to an actuator driven by an electrokinetic pump, wherein the hydraulic pressure is used to drive a piston or bellows.

BACKGROUND

Miniature pumps and valves have been a topic of increasing interest in recent years within the field of chemical analysis, especially in those applications where a variety of functions including pumping, mixing, metering, and species separation are necessary. In particular, there has been interest in integrating miniature pumps and valves with silicon and glass chip-based analysis systems designed to detect and identify trace amounts of chemical or biological material.

To meet these needs efforts have been made to develop and refine micro-scale pumps that rely on the well-known electroosmotic effect, so-called electrokinetic (“EK”) pumps, and related control and valving mechanisms for these devices. The phenomenon of electroosmosis, in which the application of an electric field to an electrolyte in contact with a dielectric surface produces a net force on a fluid and thus a net flow of fluid, has been known since the nineteenth century. The physics and mathematics defining electroosmosis and the associated phenomenon of streaming potential have been extensively explored in “Introduction to Electrochemistry,” by Glasstone, (1942) pp. 521-529 and by Rastogi, (J. Sci. and Industrial Res., v.28, (1969) p. 284). In like manner, electrophoresis, the movement of charged particles through a stationary medium under the influence of an electric field, has been extensively studied and employed in the separation and purification arts.

The use of electroosmotic flow for fluid transport in packed-bed capillary chromatography was first documented by Pretorius, et. al. (J. Chromatography, v.99, (1974) pp. 23-30). Although the possibility of using this phenomenon for fluid transport has long been recognized, its application to perform useful mechanical work has been addressed only indirectly. The present embodiment describes gasless and gas bubble-free electrodes for actuators based on EK pumps.

EK pumps are typically composed of a nanoporous packing or monolith (pore diameters from 10 to 500 nm) and a pair of high-voltage electrodes. For example, silica acquires a negative surface charge composed of deprotonated silanol groups (SiOH9 SiO⁻+H⁺) when an electrolyte with pH>4 is introduced. As illustrated in FIG. 1, a thin electric double-layer (10 nm for water with 1 mM NaCl) is known to develop adjacent to the walls of such devices. Application of an electric field exerts a body force on ions residing within the double layer and results in ion migration in the direction of the electric field gradient which induces viscous “drag” in the bulk fluid. Adding a flow restriction downstream of the porous EK pump monolith will result in an opposing pressure gradient. Hydraulic work, therefore, may be obtained after the fluid exits the porous EK pump monolith. The pressure-driven flow may be used for various applications, such as flow work against a capillary restriction, driving a piston, expanding a bellows, or fluid compression.

Conversely, external pressure-driven flows in these systems will generate electric fields that may be used to perform electrical work.

Many different microfluidic transducers have been implemented by micromachining of silicon and glass substrates. Transducers with pneumatic, thermo-pneumatic, piezoelectric, thermal-electric, shape memory alloy, and a variety of other actuation mechanisms have been realized with this technology. However, only the thermo-pneumatic and shape memory alloy designs have been incorporated in commercially-available products. Unfortunately, transducers utilizing the aforementioned actuation mechanisms are only able to generate modest actuation pressures and are therefore of limited utility.

What is needed is a transducer that can be used for microfluidic systems that can exert larger actuation pressures over longer distances (i.e., more work per stroke) than can be presently developed by conventional (non-explosive) transducer and provides both rapid “on” and “off” actuation. Millimeter-scale electrokinetic pumps are capable of such actuation, but require gasless or gas bubble-free electrodes in order to operate for more than a few seconds, as is explained in subsequent sections.

SUMMARY

EK pumps are known to exhibit a linear pressure-flowrate operating envelope for a given electric field. This linearity is due to the linearity of superposing linear electroosmotic and pressure-driven flows (ignoring property changes due to viscous heating or electrolyte composition). Because hydraulic power is the product of pressure and flowrate, the most efficient operating point for a given electric field is half the maximum pressure and half the maximum flowrate. The maximum power output increases linearly with electric field up to the point where property changes occur. For example, viscous heating at high electric fields decreases the viscosity which, in turn, increases the current draw and the power output.

Our prior efforts have demonstrated electrokinetic pumps in glass capillaries (100 μm I.D./360 μm O.D., length 3-cm to 30-cm) that are capable of pressure gradients of 250-500 psi/mm and average fluid velocities of 2 mm/s. The present embodiment describes advances in EK pump fabrication for developing larger-diameter porous monoliths and their application to mechanisms for performing mechanical work. In particular, the pumps described herein have been fabricated with diameters of 2.9-mm, and lengths from 6-mm to 10-mm. Moreover, while these pumps produce pressure gradients that are similar to those of their smaller diameter counterparts they also produce much larger flowrates, e.g., 200 μL/min for the present embodiments vs. 5 μL/min for prior-art EK pumps.

The force and stroke (i.e., work per stroke) delivered by the EK actuators of the present embodiments exceed the output of solenoids, stepper motors, and DC motors of similar size, despite the low electric-to-hydraulic power conversion efficiency of EK pumps (1-6%). Piezoelectric actuators of similar size can deliver much larger forces (e.g., 200 lbf), but their displacements are very small (e.g., 50-μm). The pump and electrodes contain no moving parts and operate silently, which is beneficial for applications requiring actuation with low noise and vibration levels.

The objective of the present invention is to provide gasless and gas bubble-free electrodes for high-pressure electrokinetic pumping requiring up to 3 mA of current in a millimeter-scale package.

These and other objectives and advantages of the present invention may be clearly understood from the detailed description by referring to the following drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows a schematic of the charge double layer that develops due to an electrolyte and wall interaction.

FIG. 2A shows two gas bubble-free cathodes comprised of a titanium frit disk having with ionomer tubing passing through a central aperture and sealed to the top an bottom faces of each disk.

FIG. 2B shows a second embodiment of the gasless cathode comprising the binary oxides co-precipitated on tantalum foil.

FIG. 3A shows an exploded view of the actuator assembly showing the un-assembled titanium/Nafion® cathode and O-rings.

FIG. 3B shows a view of the assembled titanium/Nafion® cathode and O-rings and their spatial relationship with the encapsulated EK monolith.

FIG. 4 shows a schematic of the gas bubble-free electrodes showing the relevant electrochemistry and species transport. Water is oxidized at the anode to form protons and oxygen and water is reduced at the cathode to form hydrogen and hydroxyl ions.

FIG. 5A shows a side view close-up image of the gas bubble-free anode bonded inside of the electrolyte reservoir.

FIG. 5B shows a top-down close-up view of the gas bubble-free anode coiled inside the electrolyte reservoir. The anode comprises a 10-cm length of 0.005″ platinum wire threaded inside a 0.008″ I.D./0.011″ O.D. Nafion tube.

FIG. 6 shows the constant current performance of a with ¼″ diameter ruthenium-tantalum oxide supercapacitor electrode in 0.5-M sulfuric acid.

FIG. 7 shows the constant current performance of a ¼″ diameter ruthenium-tantalum oxide supercapacitor electrode in 10-mM sulfuric acid.

FIG. 8A shows the one view of the assembled actuator pump and electrode interconnections.

FIG. 8B shows a close-up views of the assembled actuator pump and electrode interconnections.

FIG. 9 shows a cross-sectional cut away of the actuator pump/electrode housing assembly.

FIG. 10 shows the transient response of the actuator of the present embodiment using a silica monolith EK pump at an electric field strengths of 100 V/cm.

FIG. 11 shows the transient pressure response of the actuator of the present embodiment with a closed exit and a syringe plunger isolator present that does not permit the pressure to drop below 50 psi.

FIG. 12 shows the performance of the actuator of the present embodiment operating at 8 psi/V at electric field strengths of 150, 300, and 600 V/cm.

DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION

Electroosmotic flow is not an efficient method of converting electrical work to mechanical work because the mechanism is based on viscous coupling of ion motion to fluid motion in the nanometer-scale electric double layer, which results in high shear stress and corresponding viscous dissipation. EK pumps are therefore inefficient (our pumps have demonstrated efficiencies between about 1% and about 6%) and draw substantial current densities when large electric fields are applied (e.g. 100 mA/cm² for 1000 V/cm). Moreover, for capillary EK pumps with 0.1-mm O.D. porous monoliths, typical currents of 5-10 μA result in current densities at the electrode surfaces that are insufficient to nucleate bubbles (for 0.38-mm-diameter platinum wires), and the electrolysis gases simply dissolve into solution. However, increasing the pump cross-section to a diameter of 2.5-mm results in currents up to 3 mA, which is sufficient to generate visible bubble growth in a few seconds. These gas bubbles cause the current to fluctuate and decrease to a trivial magnitude (nonzero due to water films around the bubbles). Hence, gas bubble-free electrodes are necessary for stable long-term operation.

Fabrication of Bubble-Free Electrodes:

The electrodes achieve gas bubble-free operation by using Nafion® tubing to isolate electrolysis gas, generated at the surface of the metal electrode, from the pump fluid. Nafion® is a cation-selective polymer (i.e., a charge-selective salt bridge) that permits diffusive and electrophoretic transport of cations and water.

Isolation of electrolysis gas from the fluid stream is achieved with Nafion® tubing in two high-current, high-pressure, compact electrodes: a flow-through titanium/Nafion® cathode and a flexible platinum/Nafion® electrode. The cathode is shown in FIG. 2A (and a cross-sectional schematic of the electrodes is shown in FIG. 4). It consists of Nafion® tubing bonded inside a titanium frit with faces that have been sealed with epoxy. A titanium frit with 2-μm porosity is used as a cathode metal because it provides an exit path for electrolysis gases, and the frit provides a rigid mechanical support for squeezing the O-ring to achieve a high-pressure seal. Titanium is electrochemically inert as a cathode, i.e., it does not passivate or corrode.

Fabrication of the titanium/Nafion® cathode starts with drilling a hole in the frit with a diameter that provides a close fit to the desired Nafion® tube outer diameter. A small length of Nafion® tubing (2-cm to 3-cm) is pushed through the drilled hole, and epoxy is painted around the interface. The epoxy is cured for 20 minutes at 90° C. The excess tubing length is trimmed, and the faces are ground and/or sanded smooth. The finished cathode of FIG. 2A, is sealed between two O-rings, as shown in the ‘exploded’ parts assembly of FIGS. 3A and 3B. An electrical connection is made by wires pressed against the electrode edge via holes in the union.

The second electrode is shown in FIGS. 5A and 5B deployed inside a modified syringe barrel. This electrode consists of a flexible anode comprising of a small diameter (0.005″ O.D.) platinum wire inside a small diameter (0.008″ I.D., 0.011″ O.D.) Nafion® tube. The platinum wire is threaded through the Nafion® tubing to the desired length (more than 10 cm is difficult for a 0.005″ wire due to friction). The tubing is then epoxy bonded to the wire at one end. The other end is threaded through and bonded to a via hole in a housing (here, a syringe barrel). The wire is then soldered to a structural electrical terminal. A high-pressure version of the flexible platinum/Nafion® electrode is composed of braided platinum wire. The resulting crevice volumes provide an exit path for the electrolysis gases. Testing of a three-wire braid inside a Nafion® tube with a 0.003″ wall thickness (0.021″ I.D., 0.027″ O.D.) has shown leak-free operation at 3000 psi.

The Nafion® tube has a thin wall (0.001″) that results in rapid water diffusion to the platinum, despite the opposing electrophoretic transport of hydrated protons. Platinum is the preferred metal because it is electrochemically inert for both anodic and cathodic operation, i.e., it does not passivate or corrode. Small diameter (0.005″) platinum wire is ductile and relatively affordable. Testing has shown that this anode configuration can supply more than 2 mA per cm of length (23 mA/cm² and 31 mA/cm² based on the tube O.D. and I.D., respectively) in 5 mM TRIS-HCl at pH 8.5. However, during high-current testing (2 mA/cm) a small number of bubbles appeared during startup and shutdown. Bubbles did not appear to be evolving during constant current operation over a period of minutes. Testing with EK pumps has shown the electrodes to be bubble free.

FIG. 4 is a schematic of the electrode cross-sections, mass fluxes, and electrode reactions during steady-state operation with pure water. Electron transfer occurs at the metal-Nafion® interface and involves water and proton reduction at the cathode and water oxidation at the anode. In steady-state, water is transported to the cathode by diffusion and electrophoretic transport of hydrated protons. Water is transported to the platinum anode by diffusion alone, which requires a large surface area and/or a thin membrane. For common EK pump electrolytes such as 5 mM TRIS-HCl at pH 8.5, the schematic in FIG. 4 does not change substantially. The reactants and products are the same because the concentration of TRIS⁺ is negligible relative to water (1.3 mM TRIS ⁺ vs. 55 M water), and electrophoretic transport of TRIS⁺ through Nafion® is slower than proton transport. A pH gradient from high pH at the cathode to low pH at the anode will develop because electrophoresis of TRIS⁺ and Cl⁻ will carry most of the current due to their large concentration relative to protons and hydroxyls. A reduction in pH adversely affects silica EK pumps by reducing the magnitude of the zeta potential (fewer charge sites per area) and, hence, their efficiency. In practice, this problem is avoided by flowing sufficient buffer to avoid pH changes. Alternately, some of the high pH fluid exiting the pump may be returned to the entrance reservoir.

Examples of Gasless Electrodes Used in EK Pump Actuators:

A class of gasless electrodes has been developed for reciprocating EK actuation.

The electrodes are supercapacitors that store charge by intercalation and adsorption to the surface. One version of these electrodes is a co-precipitated binary oxide of ruthenium and tantalum on titanium shown in FIG. 2B. The second version is a porous carbon electrode such as a carbon xerogel or a paste suspension supercapacitor. The ruthenium-based electrodes derive a portion of their capacitance from pseudo-capacitance in which the ruthenium changes oxide states (III to V). The recipes for these electrodes comes from supercapacitor patent literature. Their application to EK flow is a particular challenge. A typical electrolyte in supercapacitors is a strong acid (12 M sulfuric acid) that provides low electrical resistance and, in the case of ruthenium supercapacitors, a large concentration of protons for rapid intercalation and adsorption. EK pumps work most efficiently when operating with low conductivity, low molarity buffers such as 5 mM TRIS with 1.3 mM HEPES at pH 8, both of which are large ions that move slowly. As seen in a comparison between FIGS. 6 and 7, the performance of the supercapacitor electrodes drop sharply when EK pump electrolyte (the low pH electrolyte) is used instead of the higher pH 0.5 M sulfuric acid. The pH increase from 0 to 8 drops the proton concentration by a factor of 10⁸ from 1 M to 10 nM, and the conductivity drops from 1 S/cm to 100 μS/cm. To improve the performance, lithium hydroxide was used to titrate TAPS (N-Tris(hydroxymethyl)methyl-3-aminopropanesulfonic acid, pKa=8) to pH 8.5. Lithium is a small cation with a large diffusion coefficient, but the amount that can be added is limited by the molarity of the buffer against which it is titrated. In this case, a molarity of 2.5 mM Li was the final concentration. This increase in cation strength was sufficient to increase the storage rate and capacity of the ruthenium supercapacitors. For a fixed actuation time of 10 seconds, the constant current response increases from 0.5 to 2 mA for a ¼″ O.D. electrode plate. The improvement comes at the expense of the conductivity increasing from 100 to 150 μS/cm. The net improvement in actuation current is a factor of 2. The same concepts may be used for porous carbon supercapacitors. Initial testing in the lithium-enhanced buffer shows that the currents are smaller but the capacity is larger. The ruthenium electrodes are physically rigid, and the porous carbon electrodes are easily cracked. The ruthenium electrodes are thus the preferred method for reciprocating EK actuation.

Bubble-free electrodes have been developed that permit water hydrolysis but which isolate the gases that develop at the surfaces of the electrodes from the electrolyte. Compact, high current electrodes have been developed using a perfluorosulfonic acid co-polymer tubing produced by Perma Pure LLC (Toms River, N.J.) under the DuPont trade name Nafion®, a cation-conductive co-polymer, i.e., a charge-selective salt bridge, that permits diffusive and electrophoretic transport of cations and water.

The first of these electrodes is the high-pressure flow-through cathode shown in FIG. 2A and schematically in cross-section in FIG. 4. The cathode is comprised of a titanium frit disc with 2 μm porosity and a sleeve of Nafion® tubing (1.3-mm I.D., 1.5-mm O.D.) bonded to the inside diameter of a central through hole cut into porous disc. In addition, the two faces of the disc and the outside edges of the Nafion tubing adjoining those surfaces are sealed with an epoxy sealant. A metal frit is used because it provides an exit path for electrolysis gases and a rigid mechanical support for squeezing an O-ring to achieve a high-pressure seal. Moreover, titanium is also chemically stable as a cathode in EK pump electrolytes.

The second electrode is again the flexible anode bonded inside a modified syringe barrel shown in FIGS. 5A and 5B. The syringe acts as a reservoir for an electrolyte solution with which to operate the actuator pump.

The Nafion® tube has a thin wall (25-40 μM) that results in rapid water diffusion to the platinum, despite the opposing electrophoretic transport of hydrated protons.

Platinum is the preferred anode metal because it does not passivate (oxidize) while receiving electrons and is chemically inert in strong acids and bases. Testing has shown that this anode configuration can supply more than 2 mA/cm (23 mA/cm² and 31 mA/cm² based on the tube O.D. and I.D., respectively) in 5 mM Tris(hydroxymethyl) aminomethane-hydrochloride (TRIS-HCl) at pH 8.5.

FIG. 4 shows a schematic diagram of the relevant electrochemistry and transport associated steady-state operation of the gas bubble-free electrodes with pure water. Electron transfer occurs at the metal-Nafion® interface and involves water and proton reduction at the cathode and water oxidation at the anode. In steady-state, water is transported to the cathode by diffusion and electrophoretic transport of hydrated protons. Water is transported to the platinum anode by diffusion alone, which requires a large surface area and/or a thin membrane. For common EK pump electrolytes such as 5-mM TRIS-HCl at pH 8.5, the schematic in FIG. 4 does not change substantially. The relative amounts of reactants and products of these reactions remain unchanged because the concentration of TRIS⁺ is negligible relative to water (e.g., 1.3-mM TRIS⁺ vs. 55 M water), and electrophoretic transport of TRIS⁺ through Nafion® is slower than proton transport.

FIGS. 8A and 8B show two views of the assembled high-pressure microhydraulic actuator of the present embodiment, while FIG. 9 shows a schematic cut away of EK pump/electrode assembly 10. The actuator comprises EK pump 1, itself comprised of porous media 2 mounted in housing 3, rubber O-rings 6 for connecting and containing pump 1 and electrodes 7 and 8, so as to provide respective inlet and outlet ends 4 and 5. The assembled prototype actuator, shown in FIGS. 8A and 8B, has a source of an electrolyte (not shown), contained within liquid-tight housing 20 and liquid-tight cylinder and piston assembly 30 comprising a 100-μL syringe with a 1.4-mm diameter plunger attached to the outlet side of the EK pump. Syringes with a larger plunger area may be used to achieve proportionally larger forces but smaller velocities.

As an example, FIG. 10 shows the speed with which the pressure may be increased. The transient time is determined by the volume displaced during static system loading over the same pressure range. The actuator must displace that volume of liquid before steady state is achieved. In this example, the total volume displaced can be estimated by assuming a linear pressure rise to 1500 psi and an unrestricted EK pump flowrate of 100 μL/min for the given electric field strength (based on other experiments at partial load conditions and assuming linearity). The resulting displacement is 0.8 μL, i.e., 0.8 mm³, which seems reasonable given the syringe plunger isolator and the possibility that the isolator had a small bubble.

FIG. 11 shows the transient pressure response to step changes in electric field for an actuator comprising a silica monolith EK pump with a closed exit. The pump fluid is 5-mM TRIS with 1.25-mM 4-(2-hydroxyethyl)-I-piperazineethanesulfonic acid (HEPES) at pH 8. The apparent thermodynamic efficient, η_max , is 2.5 for the first three voltage steps and decreases to 2.3 for higher field strengths because there is a small leak at high pressure. The transients for each voltage step take 5 to 7 seconds to reach steady-state due to system compressibility.

FIG. 12 shows the performance for an actuator capable of producing 8 psi/Volt (55 kPa/V). The partial load conditions were observed using masses on a syringe plunger to maintain constant pressure during 15 to 60 seconds of piston displacement to measure the flowrate. Pressures above 900 psi (6.2 MPa) were not measured due to seal limitations. A pressure of 900 psi (6.2 MPa) at the actuator face corresponds to 6.3 lbf (28 N) distributed along the monolith. The observed η_max decreases slightly above 500 psi (3.4 MPa) due presumably to small leaks, past the O-ring on the outlet side of the EK actuator. However, no visible leaks were seen during the experiments. A least-squares fits of the data for constant field strength show a non-linear rise in the y-intercept, i.e., higher η_max, due to pump heating that lowers the viscosity and allows more current for a given field strength.

High-pressure microhydraulic actuation, therefore, has been demonstrated with gas bubble-free electrodes, an EK pump, and syringes with different plunger areas. Using the actuator shown in FIG. 8A, several 1-mm thick glass microscope slides (25-mm×76-mm) were fractured in 3-point bending with a 3.3-mm diameter piston driven at 530 psi (3.7 MPa), corresponding to a output force of 6.2 lbf (28 N). The current and voltage were 2 mA at 1500 V. Silica EK pumps have demonstrated flowrates and pressures of 200 μL/min at 400 psi (2.8 MPa) and 100 μL/min at 1000 psi (6.9 MPa) for driving loads of 2.1 and 5.3 lbf (9.3 and 24 N) at velocities of 1 mm/s and 0.5 mm/s, respectively, with a 2.2-mm diameter piston. Long-stroke actuation has been demonstrated by lifting 2 lbf (9 N) at 1 mm/s over 7-cm using a 100-μL syringe. Hydraulic power up to 17 mW has been demonstrated by an 8-psi/volt (6.9 kPa/V) pump delivering 164 μL/min at 900 psi (6.2 MPa). These forces and strokes exceed the output of solenoids, stepper motors, and DC motors of similar size. Piezoelectric actuators of similar size can deliver much larger forces (200 lbf), but their displacements are very small (50 μm). The pump and electrodes contain no moving parts and operate silently, which is beneficial for applications requiring actuation with low noise and vibration levels.

High-pressure microhydraulic actuation driven by millimeter-scale electrokinetic pumps with gas bubble-free electrodes has been demonstrated. High performance porous polymer and sintered silica monoliths have been developed that give 1% and 3% electric-to-hydraulic work conversion efficiencies, respectively. Flowrates up to 200 μL/min, pressures up to 1500 psi (8.3 MPa), and hydraulic powers up to 17 mW have been observed. Electrokinetic pressures of 3 psi/Volt (21 kPa/V) and 8 psi/Volt (6.9 kPa/V) have been demonstrated. Gas bubble-free electrodes have been developed that permit extended hermetic operation. EK actuators are capable of delivering more work per stroke than electromechanical actuators of similar size.

Claims

1. An electrode assembly, comprising:

an electronically conducting body, said conducting body having one or more exterior surfaces;

a conductor attached to and making electrical contact with said conducting body; and

a cation-selectable polymer jacket covering some portion of said one or more exterior surfaces, wherein an interface between said covered portion is open to an ambient exterior atmosphere

2. The electrode assembly of claim 1, wherein the conducting body comprises a metal foil or a metal wire.

3. The electrode assembly of claim 2, wherein the metal foil or a metal wire is titanium or tantalum.

4. The electrode assembly of claim 1, wherein the conducting body comprises a metal frit,

5. The electrode assembly of claim 4, wherein the metal frit comprises titanium or tantalum.

6. An electrode assembly, comprising:

an electronically conducting collector substrate;

an ionically conducting porous layer disposed on one or more surfaces of said collector substrate; and

an electronic conductor making electrical contact with said collector substrate.

7. The electrode assembly of claim 6, wherein said ionically conductive porous layer is an oxide layer.

8. The electrode assembly of claim 7, wherein the oxide layer is a co-precipitated binary oxide.

9. The electrode assembly of claim 8, wherein the co-precipitated binary oxide are oxides of ruthenium and tantalum.

10. The electrode assembly of claim 6, wherein the metal collector substrate is an expanded metal screen or a metal foil.

11. The electrode assembly of claim 10, wherein the expanded metal screen or metal foil is titanium.

12. The electrode assembly of claim 6, wherein said ionically conductive porous layer is a porous carbon layer.

13. The electrode assembly of claim 12, wherein the porous carbon layer is a carbon xerogel or a carbon paste suspension.

14. The electrode assembly of claim 13, wherein the collector substrate is an expanded aluminum screen.

15. A flow-through electrode assembly, comprising:

a metal body, comprising generally parallel first and second surfaces and a central opening therethrough, said first and second surfaces and said central opening covered by a cation-selective polymer jacket surrounding and isolating said surfaces and said central opening, wherein an interface between said surfaces and said central opening is open to an ambient exterior atmosphere; and

a conductor attached to said metal body.