WO1999022229A1

WO1999022229A1 - Volume efficient acid based galvanic oxygen sensor

Info

Publication number: WO1999022229A1
Application number: PCT/US1998/022458
Authority: WO
Inventors: John H. Gordon; Strahinja Zecevic
Original assignee: Ceramatec, Inc.
Priority date: 1997-10-24
Filing date: 1998-10-23
Publication date: 1999-05-06

Abstract

An efficiently-sized galvanic cell type oxygen sensor (10) has a long life, even in the presence of carbon dioxide. The oxygen sensor includes a housing (11), an oxygen permeable film (19), a cathode (21), an oxygen reduction catalyst, an anode (29), and an electrolyte (25). The electrolyte used comprises concentrated organic acid which has been buffered with alkali organic acid salt. Oxygen permeates the oxygen permeable film and is consumed at the cathode. The output signal from the sensor is substantially linear with respect to oxygen concentration. The sensor may also be made relatively insensitive to changes in orientation with the placement of an electrolyte permeable film between the cathode and bulk electrolyte.

Description

VOLUME EFFICIENT ACID BASED GALVANIC OXYGEN SENSOR

Technical Field. The invention relates in general to gaseous measurement devices. More specifically, the invention relates to a device for the measurement of gaseous oxygen, where oxygen permeates to, and is consumed at the cathode of a galvanic cell. The current passing through the galvanic cell is proportional to the concentration of oxygen in the gas exposed to the device.

Background Art. Several types of oxygen sensors are .known, including: galvanic-type sensors (often referred to as "fuel cell-type" sensors), polarograph-type sensors, paramagnetic sensors, and solid zirconia electrolyte-type sensors.

Galvanic-type sensors are popular for their low cost, relative simplicity, and ability to operate at room temperature. Galvanic-type sensors include capillary-type and membrane-type sensors. Membrane-type sensors are often preferred over capillary-type sensors because a linear relationship exists between the output signal of the sensor, and the oxygen concentration of the tested gas.

Galvanic-type sensors involve the reduction and consumption of oxygen at a cathode, concurrent with the oxidation of an anode. A typical galvanic sensor is disclosed in U.S. Patent 3,767,552 (Oct. 23, 1973) to Lauer. An electronic circuit connecting the lead anode and cathode passes an electrical current which is directly proportional to the rate of the oxygen consumed at the cathode. In the case of the membrane-type sensor, an oxygen permeable polymer film is placed between the cathode and the environment. This film limits the amount of oxygen reaching the cathode. If the cathode consumes a substantial majority of the oxygen reaching it, a concentration gradient is established across the film. The rate of oxygen reaching the cathode is then diffusion-controlled and is essentially directly proportional to the oxygen concentration in the environment. At constant temperature and pressure, the cathode reaction rate and electrical current are directly proportional to the oxygen concentration. The majority of galvanic-type oxygen sensors use an alkaline electrolyte having a pH greater than 7. However, these sensors have the disadvantage of being subject to the formation of insoluble carbonates and bicarbonates on the surface of the lead anode when carbon dioxide is present in the gas being tested. These carbonates and bicarbonates "passivate" the anode and eventually cause the sensor to fail. Because carbon dioxide is present in the air, and numerous process gasses such as combustion, fermentation, and respiration, sensors with such electrolytes frequently have an unacceptably short life, or have erratic performance from the onset of passivation.

Fujita et al. (U.S. Patent 4,495,051 (Jan. 22, 1985)) teaches the use of buffered organic acid aqueous electrolytes with a pH between 4 and 7 to prevent the formation of carbonate and bicarbonate on the anode. The organic acids disclosed by Fujita et al. include: acetic acid, propionic acid, and n-butyric acid. Organic acid salts disclosed by Fujita et al. include: formic acid salt, acetic acid salt, propionic acid salt, n-butyric acid salt, maleic acid salt, and glutamic acid salt. In these sensors, protons react with oxygen at the cathode to form water and hydrogen peroxide, while the anode dissolves, thus lowering the pH of the electrolyte until it neutralizes. Fujita et al. also teaches that some lead compound dissolved in the electrolyte is important to prevent hydrogen evolution at the cathode. Some problems which might arise from hydrogen evolution include: an increase in electrical current which was not a result of oxygen reduction, a disruption in ionic conduction caused by bubbles forming in the electrolyte within the sensor, and premature loss of the anode through corrosion from hydrogen evolution at the anode.

While Fujita et al. offers solutions which extend sensor life in the presence of carbon dioxide, Fujita et al. does not offer a solution with the advantage of having a low specific volume. For example, the sensor with the highest concentration of electrolyte taught by Fujita et al. has 5 M (molar) acetic acid, 4 M potassium acetate and 0.1 M lead acetate. Because this type of sensor does not require considerable amounts of excess lead, the sensor's volume requirement is primarily driven by the electrolyte volume requirement. However, if a sensor designer, following the teachings of Fujita et al. : a) requires a particular current signal at a given concentration of oxygen using an acid concentration of 5 M, b) desires to prevent carbonate and bicarbonate formation by only operating the sensor within a range where the electrolyte is a pH of 7 or less, and c) chooses to initially have 20% more lead than could theoretically be reacted with the existing protons, the minimum volume requirements for only the electrolyte and the lead anode of the sensor would be 8 cm³/ Ah current required. This volume does not include the volume required by the housing, the cathode, or the electrical connections.

Gambert et al. (U.S. Patent 4,894,138 (Jan. 16, 1990)), adopts the teachings of Fujita et al. , and asserts the life of acid based galvanic oxygen sensors can be extended even more through the use of organic acids containing polyfunctional groups capable of yielding several protons from each acid molecule rather than only one. While this concept initially sounds attractive, there is, in practice, little or no advantage to using polyfunctional electrolytes. For example, by using citric acid as an electrolyte with 3 carboxylic acid groups per molecule instead of acetic acid with only 1 carboxylic acid group per molecule, the life of the sensor should theoretically be extended by a factor of 3 times. Such a life extension would be true except for the fact that the citric acid molecule is more than three times as large as the acetic acid molecule. Thus, in terms of proton density, citric acid offers no actual advantage. Citric acid is also less preferable than acetic acid with similar proton densities because a citric acid solution is (relatively) very viscous and subject to polarization unless the solution is dilute. In addition, it is believed that some polyfunctional organic acids cannot be used because they cannot be effectively buffered to avoid hydrogen evolution. If one or more of the protons has a low dissociation constant, for example, a pK value lower than 3.7, considerable amounts of lead ions or buffering salt must be added to the solution to prevent hydrogen evolution. The addition of these non-proton releasing constituents dilutes the solution and defeats the goal of increasing proton density. In addition, polyfunctional organic acids with high pK valued protons offer little advantage over non-polyfunctional acids because one or more of the protons is unavailable until the pH rises to an undesirable level above 7. Both Fujita et al. and Gambert et al. stress that electrolyte conductivity is an important parameter in the selection of electrolyte for galvanic-type sensors. However, as will be clear from the disclosure of this invention, this parameter is not that important given the low current densities typical of sensors of this type. Disclosure of Invention

The invention involves acid-based galvanic oxygen sensors having increased life and decreased size over prior art sensors. These advantages are accomplished by increasing the proton density of the acid electrolyte. This invention is of significant importance to the art because it is advantageous to have the sensor survive as long as possible, but often constraints exist which limit the size of an oxygen sensor. Maximizing sensor life, given a particular size constraint, is also advantageous from the viewpoint that very little incremental cost is incurred since the cost of electrolyte and lead are relatively small compared to the costs of the housing and other sensor components. Additionally, the cost to replace the sensor is several times higher than the cost to produce the sensor.

Because hydrogen evolution can cause an improperly functioning sensor, it is important to understand the conditions under which hydrogen evolution occurs to maximize proton density without creating an improperly functioning sensor. Hydrogen evolution occurs when the electrode potential is equal to, or more negative than, the reversible potential for the reduction of 2H⁺ to H₂. In the case of an acid- based oxygen sensor using a lead anode, the most negative potential of the cathode is equal to the potential of lead anode dissolution. Consequently, hydrogen evolution occurs when the reversible potential for the reduction of 2H⁺ to H₂ at the cathode is less than or equal to the reversible potential for the reduction of Pb²⁺ to Pb at the anode.

Under many conditions, a potential difference exists between the cathode and the anode solely because of the electrical current passing through resistance in the circuit connecting the cathode and anode. This potential difference is lowest when the oxygen concentration is very low. A low potential difference increases the likelihood of hydrogen evolution. The primary variables which affect the thermodynamics of the hydrogen evolution process are temperature, lead ion concentration, pressure and proton concentration. From a thermodynamic standpoint, the relationship which relates these variables to the pH level where hydrogen evolution will occur in the absence of oxygen is the following: pH>42.3/T+2.21-0.2\l\n{[Pb ]P} Equatⁱon 1:

where pH is the negative log of the disassociated proton concentration T is the absolute temperature of the system in K

[Pb ⁺] is the molarity concentration of lead ions P is the pressure in atmospheres

It should be understood that the constants in the above relationship are dimensioned so as to render the relationship dimensionally correct, i.e. , balanced. From the relationship described by Equation 1 , it can be seen that the highest pH will result at low temperatures and low pressures. By way of example, assuming 273K (0°C) and 0.8 atm as design criteria for a commercial sensor, for given values of lead ion concentrations, the minimum possible pH of the electrolyte to prevent hydrogen evolution at the anode is:

[Pb²⁺ι pH minimum

0.001 M 3.97

0.01 M 3.47

0J M 2.97

1 M 2.47

For a 1 atmosphere and 50°C process, the sensor pH could be as low as 2.4 without evolving hydrogen if the lead ion concentration were 1 M; a very feasible condition. While the relationship for this minimum pH is based on calculations using thermodynamic relationships and data, it can be verified through experimentation. Thus, it is believed that the constraints of a pH between 4 and 7 set forth by Fujita et al. are overly conservative and are not conducive to the minimization of sensor volume.

The invention involves the use of organic acid electrolyte solutions which have little or no water, are buffered with alkali salts of organic acids, contain enough lead ions in solution to prevent hydrogen evolution, and have significantly lower specific volume requirements than previously thought possible. Further, this invention establishes that the condition of high electrolyte conductivity taught by both Fujita et al. and Gambert et al. is not a requirement for successful operation of galvanic oxygen sensors.

The maximization of proton concentration involves the minimization of all constituents which cannot release protons. Specifically, it is preferable to minimize the water used. Concentrated formic acid and concentrated acetic acid are the only two organic weak acids which exist in the liquid state without requiring any water as a solvent. Undiluted formic acid is approximately 25 M while undiluted acetic acid is approximately 20 M. We have found it possible to dissolve adequate amounts of alkali organic acid salt and lead organic acid salt in these non-aqueous, concentrated acids to form electrolytes which do not evolve hydrogen, even in the absence of oxygen.

Other features, advantages, and objects of the present invention will become apparent from a consideration of the following figures and further description.

Brief Description of Drawings

FIG. 1 shows a cross-sectional view of a galvanic cell-type oxygen sensor built according to the invention.

FIG. 2 is a graph showing the relationship between sensor output voltage (mV) and oxygen concentration (per cent) utilizing a sensor built according to a preferred embodiment of this invention.

FIG. 3 is a plot comparing the specific volume requirements of acid-based galvanic oxygen sensors (cc / Amp-hour) to electrolyte concentration (moles / liter) and the relative differences between the examples Fujita et al. and those of the present invention.

Best Mode for Carrying Out the Invention FIG. 1 shows a cross-sectional view of a preferred embodiment of a galvanic cell-type oxygen sensor according to the present invention. The sensor, generally 10, is comprised of a housing 11 divided into two parts: a main body 13 and a cap 15. The housing 11 is preferably made of a material which is compatible with concentrated acetic acid or formic acid, has low oxygen permeability, and is electrically insulative. Polyvinylidene fluoride and equivalent polymers are materials which meet these criteria.

A sensing zone 17 comprises a diffusion film 19 which is typically a fluoropolymer such as polytetrafluoroethylene ("PTFE"), or fluorinated ethylene propylene ("FEP"). Adjoining the diffusion film 19 is a cathode 21 which is porous (e.g., via pores 27) to the electrolyte 25. The cathode 21 is a conductor which comprises, in part or in total, an electrocatalyst suitable for the reduction of oxygen. The cathode is made of a metal effective for the electrolytic reduction of oxygen such as gold, although other metals such as ruthenium, iridium, platinum, and mixtures thereof may alternatively (or in conjunction with the gold) be used.

Adjoining the cathode 21 is a separator 23, an optional component, but one which is highly preferred. If used, the separator 23 should be permeable to protons and water but should also prevent mixing of electrolyte 25 with the relatively small amount of liquid which may be found in the pores 27 of the cathode 21 and between the diffusion film 19 and cathode 21. Suitable materials for the separator 23 include microporous hydrophilic films such as treated polyolefin films utilized for a similar purpose in the battery industry, or cellulose fiber papers. Proton-conductive solid electrolyte films such as perfluorinated ion exchange membranes with sulfonic acid exchange groups are also suitable for use as a separator. The separator 23 performs two basic functions: (1) It prevents internal shorting of the galvanic cell sensor if a portion of the anode 29 breaks off and would otherwise come in contact with the cathode 21, and (2) it prevents the mixing of electrolyte 25 with the cathode 21, thereby substantially stabilizing the sensor signal.

A compressible gasket 31 is typically placed between the housing cap 15 and the diffusion film 19 to create a seal. In this way, oxygen substantially only enters the sensor through the diffusion film 19.

The invention also comprises an anode 29 which is substantially lead and is formed so that part of the anode 29 passes directly through the housing wall 35 and is in contact with the electrolyte 25. It is also possible to pass a current collector (not shown) through the housing wall 35 to make contact with the anode 29. In such an embodiment, it is preferable to choose a contact material that is more noble than the material of the anode 29. It is also preferable to either match the size of the portion of the anode 29 which is in the electrolyte 25 to the amount of protons in the electrolyte 25, or to have up to 30% excess lead. In this way, when the anode 29 is oxidized, it liberates substantially the same coulombs as the coulombs resulting from the reaction of protons in the electrolyte with oxygen. For example, the lead and electrolyte described in the hereinafter described EXAMPLE I, theoretically would have a volume ratio of 1 to 6.67. Unlike the anodes utilized in alkaline electrolyte galvanic sensors, the anode of the current invention does not require high surface area since it does not passivate. However, an excess amount of lead is generally preferable because the lead must maintain an electrical continuity and some excess lead is required to serve as a current collector.

Another gasket 33 may be utilized to form a seal where the anode 29 enters the housing wall 35 to prevent the leakage of electrolyte outward or surrounding gas inward. To prevent the formation of hydrogen at the anode 29, the pH of the electrolyte must be greater than, or equal to, the value calculated in the formula set forth in the aforementioned Equation 1.

The electrolyte used generally comprises concentrated organic acid which has been buffered with alkali organic acid salt. The electrolyte may include a mixed solution of organic acid, a lead compound, and a buffer salt, (optionally including water) wherein the pH value of the electrolyte meets the relationship:

pH> 42.3/7+2.27 -0.2 \7ln{[Pb ]P}

where pH is negative log of the disassociated proton molarity concentration, T is the absolute temperature of the system, in degrees Kelvin, [Pb²⁺] is the molarity concentration of lead ions, and P is pressure in atmospheres. A preferred electrolyte 25 is comprised in part of acetic acid, formic acid, or a combination thereof.

Optionally, a cathode wire 37 and an anode wire 39 respectively connect the cathode 21 and anode 29 to a circuit board 41. The circuit board 41 is preferably configured so that an electrical current may pass from the anode 29 to the cathode 21 though a resistor; more preferably through a network of a thermistor and one or more resistors connected in series and in parallel. A connector 47 is preferably integrated into the network so that the voltage drop across the network can be measured. The connector 47 provides an interface to a device which can be utilized to convert a voltage signal to an oxygen concentration reading after calibration. Again, the circuit board 41 and connector 47 are optional since said cathode wire 37 and anode wire 39 can be attached directly to a device which enables the passing of an electric current from the anode 29 to the cathode 21. The anode 29 and cathode 21 are spatially and electrically separated except through the circuit.

In practice, at constant pressure and temperature, an electrical current increases in direct proportion to the oxygen concentration exposed to the diffusion film 19. The rate of oxygen diffusion through the diffusion film 19 is directly proportional to the partial pressure of oxygen. Thus, at constant temperature, the electrical current is directly proportional to the pressure. Electrical current passing through the circuit increases exponentially with rise in absolute temperature. For this reason, it is preferable to use a thermistor as part of the circuit so the conductance of the thermistor increases exponentially with absolute temperature resulting in a voltage drop across the network which does not significantly change within a temperature range.

An optional flexible diaphragm 49 is preferably integrated into the housing 11 to allow the sensor to accommodate expansion during temperature changes and to substantially equate the internal pressure to the external pressure. An optional hydrophobic film 51, which is also preferably gas permeable, may be adjoined to the exterior of the diffusion film 19. This hydrophobic film 51, which is typically a porous fluoropolymer, allows oxygen to quickly diffuse to the diffusion film 19 but prevents condensation or debris from directly occluding a portion of the diffusion film's 19 effective area. The following illustrative EX.AMPLES of electrolytes satisfy the minimum pH requirements at a given lead ion concentration, contain no water, and have been successfully utilized in high proton density galvanic oxygen sensors:

EXAMPLE I

Constituents: Concentrated acetic acid, 1.0 M potassium acetate, 0.1 M lead acetate. The resulting solution is 16.5 M acetic acid with a pH of 3.2 and a conductivity of 2.2 mS cm^'1 at room temperature. Assuming 20% excess lead, and not including housing and cathode, the minimum volume requirements of the sensor are 2.8 cm³ / Ah current required. Although the conductivity is two orders of magnitude lower than what Fujita et al. asserts is preferable for satisfactory sensor performance, sensors with this electrolyte performed satisfactory with regard to linearity, stability, and response time.

FIG. 2 shows a relationship between sensor output voltage and oxygen concentration for a sensor built according to an embodiment of the invention. The sensor was constructed as depicted in FIG. 1 using an electrolyte as described in EXAMPLE I. The housing and diaphragm were constructed of polyvinylidene fluoride polymer. The diffusion film was fluorinated ethylene propylene. A linear regression fit of the data had an R² of 0.99996. The sensor including housing and circuit board had a volume of 11.6 cm³ with an expected life of 2.4 Ah yielding a ratio of 4.8 cmVAh. This volume requirement is substantially lower than possible in the prior art.

EXAMPLE II

Constituents: Concentrated formic acid, 6.2 M potassium formate, 0.1 M lead acetate.

The resulting solution is 19.9 M formic acid with a pH of 3.2 and a conductivity of 27.6 mS cm^"1 at room temperature. Assuming 20% excess lead, and not including housing and cathode, the minimum volume requirements of the sensor are 2.4 cm³/ Ah current required. These examples are over 3 times more volume efficient compared to the most concentrated example provided by Fujita et al. which had a higher pH and was diluted with water. Although the electrolyte conductivities are not as high as the electrolytes cited by Fujita et al. (e.g., 125 mS^'Vcm), they are sufficient for galvanic oxygen sensor applications. The relationship between sensor current and oxygen concentration is linear and the time required to stabilize between concentrations

(response time) is comparable to sensors utilizing electrolyte with significantly higher conductivities. FIG. 3 compares the specific volume requirement of acid-based galvanic oxygen sensors versus electrolyte concentration and the relative differences between the examples cited by Fujita et al. and the EX.AMPLES disclosed in this invention. The plot of FIG. 3 shows the advantage of using an electrolyte as taught in this invention where water is minimized, or excluded, and the electrolyte pH constraint is relaxed to the level required to avoid hydrogen evolution.

While the invention has been described with reference to certain preferred embodiments and examples, these are for illustrative purposes only, and the scope of the invention is to be determined in view of the appended claims.

Claims

ClaimsWhat is claimed is:

1. A galvanic cell oxygen sensor comprising: a galvanic cell comprising: a cathode comprising a metal effective for the electrolytic reduction of oxygen; an anode made up substantially of lead; an electrolyte comprising a mixed solution of organic acid, a lead compound, and a buffer salt, wherein the electrolyte's pH value is between as low as 2.4 and less than about 4 and meets the relationship:

pH>42.3lT+2.27-0.2\7ln{[Pb ]P}

wherein T is the absolute temperature of the system, in degrees Kelvin, is the molarity concentration of lead ions, and P is pressure in atmospheres; an electronic circuit connecting said anode to said cathode, said electronic circuit including at least a resistor; and a housing constructed such that the electrolyte is contained therein, said anode and cathode are maintained in spaced relationship, electronically separated except through said electronic circuit and wherein the galvanic cell is constructed such that oxygen diffusion to the cathode is primarily through an oxygen permeable polymer film in close proximity to said cathode.

2. The galvanic cell oxygen sensor of claim 1, wherein said organic acid is selected from the group consisting of formic acid, acetic acid, and mixtures thereof.

3. The galvanic cell oxygen sensor of claim 1 , wherein said buffer salt is an alkali salt of an acid selected from the group of acids consisting of formic acid, acetic acid, and mixtures thereof.

4. The galvamc cell oxygen sensor of claim 1, wherein said cathode is comprised of gold.

5. The galvanic cell oxygen sensor of claim 1, wherein said housing comprises a polyvinylidene fluoride polymer.

6. The galvanic cell oxygen sensor of claim 1, wherein the oxygen permeable polymer film is positioned adjacent said cathode.

7. The galvanic cell oxygen sensor of claim 6, wherein the polymer film is made of a polymer selected from the group consisting of polytetrafluoroethylene and fluorinated ethylene propylene.

8. The galvanic cell oxygen sensor of claim 1, wherein said housing includes a flexible section.

9. The galvanic cell oxygen sensor of claim 1, wherein a film or sheet, permeable to water and protons, in close proximity to the cathode, separates the cathode from the electrolyte.

10. The galvanic cell oxygen sensor of claim 1, wherein the electrolyte further comprises water.