Patent Application
20070190384
This invention relates to electrolytic membranes used in direct methanol fuel cells and methods to produce such membranes, and more particularly to cross-linked proton conducting membranes including water-binding fullerenes.
Direct methanol fuel cells (DMFC) are increasingly important, becoming a choice for fuel cells for portable applications such as batteries for laptop computers and cell phones. Unlike H2PEFC in which hydrogen is fed to the anode, DMFC uses liquid methanol as the fuel. At the anode, methyl alcohol (MeOH) is oxidized in the presence of water. This oxidation generates electrons to power the circuit, hydrogen ions that travel through the electrolytic membrane of the fuel cell, and carbon dioxide as a by-product. At the cathode, the hydrogen ions react with oxygen and electrons from the circuit producing water as the only other by-product.
One of the most serious technical hurdles in development of DMFC is the MeOH permeation through a membrane, otherwise known as “the methanol crossover.” Inefficiencies arise since the methanol crossover (i) reduces the power when methanol reaches at the cathode to be oxidized by the oxygen, (ii) loses the fuel, thus decreasing the fuel efficiency, (iii) enlarges unnecessarily the dimension of the fuel cell since using a high concentration of methanol results in more methanol crossover, thus resorting to lower concentrations which thus require a larger fuel storage, and (iv) makes it difficult to operate at high temperatures which increases the catalytic activity, but in turn promotes more methanol permeation. Most membranes that are used in DMFC employ water as the principal proton conducting medium, and efforts to block methanol while allowing water to freely permeate the membrane have turned out to be extremely difficult. Most efforts to reduce methanol crossover come at the expense of the proton conductivity.
The higher the equivalent weight (EW) of the membrane, the higher the water drag coefficient, thus more water will permeate through the membrane. There is a linear correlation between the water drag coefficient and the methanol crossover. Hence, one way to reduce the methanol crossover is to use membranes with high EW, such as, for example, by reducing the degree of sulfonation to the polymer. This approach, however, also usually reduces the proton conductivity.
The methanol crossover can also be reduced by employing thicker membranes. However, thicker membranes also result in higher ohmic resistance when assembled in a fuel cell. Another approach would be to use methanol impermeable polymers as the membrane, such as, for example, poly(phosphazine). Yet, again, the cell performance also decreases as the methanol crossover is reduced. Still another approach has been to use inorganic fillers such as SiO2 or TiO2. This is effective in reducing the MeOH crossover, but it often leads to increasing the membrane resistance.
It has been also found that cross-linking of a proton conducting membrane is effective in reducing the MeOH crossover. However, cross-linking of a membrane through chemical bonds tends to cause membrane stiffness and brittleness as well as increase the membrane resistance.
Upon investigating the relationship between the state of water in proton conducting membranes and their MeOH crossover, there appears to be a correlation between the amount of free water in the membrane and the MeOH crossover. In the present invention, it has been determined that some fullerene derivatives can bind water molecules. These fullerene derivatives, when mixed in a host polymer or when chemically attached to the polymer, exhibit very small quantities of free water, and contrary to other approaches, increasing the fullerene content in the polymer reduces the MeOH crossover while in fact maintaining the high proton conductivity.
The present invention, therefore, is directed to a proton conducting membrane for use in a direct methanol fuel cell, where the membrane comprises a polymer material and water-binding fullerene derivatives. The polymer can be any polymer and can be MeOH permeable so long as the polymer is proton conductive. The membrane may further comprise cross-linking functional groups to further reduce the MeOH crossover.
Fullerenes with functional groups such as amino groups (—NH2) interact strongly with the acid groups of a proton conducting membrane through acid-base interactions, thus forming an ionic cross-link with the polymer. Ionic cross-linking gives rise to a more flexible, less brittle membrane, compared to chemically cross-linked membranes. The membrane may further comprise cross-linking fullerenes to further reduce the MeOH crossover.
The detailed description set forth below in connection with the appended drawings is intended as a description of presently-preferred embodiments of the invention and is not intended to represent the only forms in which the present invention may be constructed or utilized. The description sets forth the functions and the sequence of steps for constructing and operating the invention in connection with the illustrated embodiments. However, it is to be understood that the same or equivalent functions and sequences may be accomplished by different embodiments that are also intended to be encompassed within the spirit and scope of the invention.
A first aspect of the present invention is to use in the proton conducting membrane of a direct methanol fuel cell those fullerene derivatives that hold a large amount of bound water, referred to as “water-binding fullerenes.” Water-binding fullerenes are those fullerenes chemically attached by functional groups, including C60 (with no functional group) itself, where the functional groups strongly bind water molecules to themselves. The water binding by the functional groups appears to be primarily due to the electron transfer between the water molecule and the functional group, such as hydrogen bonds, which are caused by either hydrogen acceptance or hydrogen donation by the functional group. Thus, those functional groups with either a hydrogen acceptor or a hydrogen donor are good candidates for the invention. Thus, one embodiment of the present invention involves increasing the efficiency of water-binding agents by attaching the functional groups to fullerene, which reduces MeOH crossover.
Fullerenes can have the functional groups in a large surface density as well as in an extremely high volumetric density. The present invention takes advantage of this unique property of fullerenes to maximize the effect by the water-binding functional groups. The inventors have demonstrated this effect using, among other agents, polyhydroxy fullerene C60(OH)n, where n is within the range of more than 2 and less than 60, or more preferably, more than 2 and less than 48. Some of the examples discussed below involve C60(OH)12. Polyhydroxy fullerene (referred herein as PHF) was synthesized according to methods known in the art (e.g., Long Y. Chiang, et al., Efficient Synthesis of Polyhydroxylated Fullerene Derivatives via Hydrolysis of Polycyclosulfated Precursors, 59 J. Org. Chem. 3960 (1994)) through sulfonation of C60 and subsequent hydrolysis.
A second aspect of the present invention is based on cross-linking of proton conducting membranes (PCM) to reduce MeOH crossover, an example of which is given in
Aminofullerene, C60[NH(CH2)nNH2]m, where 1<n<50 and 2<m<60, was synthesized as follows: a gram of fullerene was added to 50 mL of freshly distilled ethylenediamine to form a solution. The fullerene dissolved in ethylenediamine and formed a dark solution. The dissolution of the hydrophobic fullerene in a hydrophilic solvent such as ethylenediamine was a strong indication that a reaction (neuclophilic addition) took place. Excess ethylenediamine was removed using a rotary evaporator, such as the Rotavapor®, which is commercially available from BUCHI Laboratory Equipment (BÜCHI Labortechnik AG), Switzerland. The product was a black solid, which was insoluble in water but was soluble in ethylenediamine. Based on elemental analysis, an average of five ethylenediamine molecules were attached to each molecule of fullerene.
The preparation of the solution cast fullerene-Nafion® composites was as follows: a 5% Nafion® solution was dried at 80° C. overnight. Since the present invention reduces MeOH crossover, the membrane may be any number of substrates, including even MeOH permeable polymers. Nafion® currently is the industry standard membrane for fuel cells generally, but alone Nafion® is highly methanol permeable and therefore previously of limited use in DMFC. It is a copolymer of tetrafluoroethylene and perfluoro-3,6-dioxa-4-methyl-7-octenesulfonyl fluoride and is commercially available from DuPont in either acid or ionomer form. More broadly, other perfluoro polymer sulfonic acids may be used as well. For the composites used in the present research, the dried Nafion® film was dissolved in dimethylacetamine (DMAc). This Nafion®-DMAc solution, was stirred for several hours and cast on a glass substrate. The film was dried at 120° C. overnight and then annealed at 170° C. (referred herein as Recast Nafion®). Subsequently, a predetermined amount of C60 was dissolved in o-dichlorobenzene, and the solution was added to the Nafion®-DMAc solution while stirring. The solution was cast on a glass substrate and dried overnight. Then, the cast membrane was annealed at 170° C. (to be referred to as C60-Nafion®). The weight percentage (wt %) of C60 in Nafion® was 1%. Separately, PHF was dissolved in DMAc, which was added to the Nafion®-DMAc solution. The mixed solution was cast on a glass substrate and dried in an oven at 120° C. overnight, followed by annealing at 170° C. (to be denoted as, e.g., PHF-Nafion®). The weight percentage of PHF is preferably less than about 20%, more preferably less than about 5%, and most preferably within the range of about 1% to 3%.
To prepare a composite membrane with 1% loading of amino-fullerene, 75.8 mg of amino-fullerene was dissolved in 1 mL of ethylenediamine to form a solution. The solution of amino-fullerene was then mixed with a solution of 750 mg of Nafion® in 10 mL of DMAc. The mixture was cast onto a Teflon® dish at 80° C. for 12 hours to form a membrane with thickness of about 0.1 mm. The membrane was acidified in 1 M H2SO4 at 80° C. for one hour. The ion-exchange capacity (IEC) of the composite membrane was much lower than expected, due to residual solvent in the membrane. Thus, it was re-acidified in a more acidic medium (3 M H2SO4) and at a longer time (8 hours). The solvent was completely removed after 2 cycles of re-acidification, as judged by the transparency of the acid solution used to wash the membrane. After re-acidification, the membrane was washed repeatedly in de-ionized water at 80° C. for 8 hours until the pH of the water was neutral. The weight percentage of amino-fullerene is preferably less than about 60%, more preferably less than about 20%, and most preferably between about 1% and 5%.
The membrane was pretreated as follows. Nafion® (117 or 115 as needed) was cut to required sizes and stirred in 3 vol % H2O2 solution at 90° C. for one hour. The as-received membrane was yellowish-brown in color. The treated membrane was stirred in distilled water for about one hour. After peroxide treatment, the membrane was completely colorless. The cut membrane was then stirred in 0.1M H2SO4 at 90° C. for one hour. The cut membrane was further stirred in distilled water for 1 hour to remove the excess acid. The treated clear membrane was stored in distilled water till use. Before making the membrane electrode assembly (MEA), the membrane was patted dry and allowed to air dry for about an hour.
The electrode was then prepared as follows. The electrode was purchased from E-TEK, and then punched out using 0.75-inch diameter punch (resulting in an area of approximately 2.85 cm2). Six to eight drops of diluted Nafion® solution (2:1 by volume Methanol:5% Nafion® solution in Isopropanol [from Aldrich]) were added on the catalyst side ensuring complete coverage. The electrodes were dried in an oven at 70° C. for about 30 minutes. Iron heating-blocks were then stacked on one another with the thermocouple in between them and preheated on a hot-plate to a temperature of 125° C.
The MEA was assembled as follows. A 150 micron Teflon coated fiber-glass tape gasket (TFG) was placed flat and a 250 micron TFG gasket was placed on top. One electrode was placed with the catalyst side up and the Nafion® membrane was placed on top. The second electrode was placed with the catalyst side down making sure that the electrodes were exactly aligned. A 250 micron TFG gasket was placed on top of the sandwich, aligned with the bottom gaskets. Finally, a 150 micron gasket was placed at the top, aligned with the rest of the gaskets.
After the blocks reach the desired temperature, the top block was lifted and the thermocouple removed. The MEA assembly was immediately placed on the bottom block and the top block was carefully placed on the MEA assembly. The hot blocks were transferred to the press, and the shield was closed. The press was pumped up rapidly until the plates contact each other, then gradually ramped to a load of 1200 lb-f. Once at 1200 lb-f, the load was maintained for 90 seconds. Then, the pressure was released by turning the release knob about 45 to 60 degrees counterclockwise until the plates were separated. The MEA was immediately taken off the blocks and allowed to gradually cool.
To measure the water uptake, the membranes were first vacuum-dried at 100° C. overnight and weighed afterward, and then immersed in deionized water at room temperature for 24 hours. The wet water uptake was determined by the following equation:
where Wwet and Wdry are the weights of the wet and the dry membranes, respectively.
Subsequently, the membranes were equilibrated under 25% relative humidity (RH) overnight and weighed afterward as W25%. Dry water uptake was estimated from the following equation:
After all samples were equilibrated in water for 24 hrs, they were blotted on the surface before the measurement. Then, the sample was subject to dry N2 gas for 40 minutes at 30° C. and the weight loss was monitored. A Mettler Toledo TGA/SDTA 851e was used for the TGA measurements without heating.
MeOH crossover measurements were conducted by measuring the limiting current density for each membrane, using methods known in the art (e.g., Xiaoming Ren et al., Methanol Transport Through Nation Membranes: Electro-osmotic Drag Effects on Potential Step Measurements, 147 J. Electrochem. Soc. 466 (2000)). The cell was assembled using the platinum loading of 0.5 mg cm−2 and Nafion® loading of 0.8 mg cm−2 for the electrodes and the composite membrane dispersed by the water-binding fullerene. The anode catalyst was Pt/Ru, while the cathode used Pt only.
AC impedance measurements were performed for the films at 20° C. in the frequency range of 1 to 105 Hz by a Solariton spectrometer and potentiostat. The RH was controlled by adjusting the ratio of dry and wet N2 gas flow, and the internal RH of the membrane was monitored by humidity measurement. The membrane was equilibrated under a given RH for several hours prior to the impedance measurements. The resistance associated with the membrane at zero phase angle was used to estimate the proton conductivity of the membrane using the equation, σ=(1R) (L/A), where R is the bulk resistance of the membrane, L represents the membrane thickness, and A the membrane area. The thickness of the membranes varied from 178 to 195 microns.
Table 1 summarizes the water uptake in the fullerene composite membranes.
The water retention in the composites was examined by TGA measurements in which the weight loss of the composites was monitored under the dry N2 gas flow at 30° C. for 40 minutes, assuming the weight loss under this condition is primarily due to the water loss. Table 2 lists both water loss and the water remained in the membrane after subject to the dry N2 gas flow for 40 minutes.
Both Tables 1 and 2 confirm that the fullerene increases water retention in the composite membranes. The increased water retention helps reduce the water diffusion in the membrane, thus reducing MeOH crossover.
The methanol crossover is expressed in terms of the limiting current density which is proportional to the MeOH diffusion, as Jlim=k*6*F*D*C/l where k is the coefficient, F the Faraday constant, D the MeOH diffusion constant, C the concentration of MeOH, and l the membrane thickness. Since the limiting current depends on the membrane thickness, the limiting current density is normalized to the thickness:
It is clear that the fullerenes reduce the MeOH crossover of Nafion® with 3% inclusion of PHF achieving the lowest MeOH crossover, more than a 35% reduction, relative to a commercially available Nafion® membrane, in terms of the normalized measurements.
Table 4 summarizes the proton conductivity of various composites obtained from AC impedance measurements at 20° C. under 80% RH.
Thus, in contrast to many other attempts to reduce MeOH crossover, which result in lowered proton conductivity as well, the present invention reduces MeOH crossover while maintaining, and sometimes even enhancing, the conductivity.
The polarization curves, as shown in
While the present invention has been described with regards to particular embodiments, it is recognized that additional variations of the present invention may be devised without departing from the inventive concept.
This invention can be used to provide electrolytic membranes for use in direct methanol fuel cells, and more particularly to provide cross-linked proton conducting membranes including water-binding fullerenes in such applications.
