WO2008075297A2

WO2008075297A2 - Proton-conduct ing gel-membranes

Info

Publication number: WO2008075297A2
Application number: PCT/IB2007/055223
Authority: WO
Inventors: Bruno Scrosati; Fausto Croce
Original assignee: Universita Degli Studi Di Roma 'la Sapienza'; Universita Degli Studi 'g. D'annunzio' Di Chieti-Pescara
Priority date: 2006-12-21
Filing date: 2007-12-19
Publication date: 2008-06-26
Also published as: WO2008075297A3; ITRM20060687A1

Abstract

Composite, nanoporous, proton membrane, comprising a polymer matrix compatible with proton ambient with ceramic particles of micrometric and submicrometric dimension, and having a proton-conducting gelified solution immobilized therein is disclosed. In said membrane, said polymer is preferably a PVdF-CTFE copolymer and said ceramic consists of SiO2 particles. The membrane according to the invention provides high proton conductivity, low methanol dispersion, satisfactory stability and good behaviour of direct methanol fuel cells, using said system as separator at ambient temperature and especially at intermediate temperatures. A further advantage of this system is low cost, which is convenient in the fabrication of direct methanol fuel cells.

Description

Low cost gel proton membranes

The present invention relates the technical field of direct methanol fuel cells (DMFC). In particular, the present invention refers to a low cost, alternative proton-conducting system based on swelling of a sulfuric acid gel immobilized in nanoporous composite membranes as electrolytes for DMFC at low or intermediate temperature.

Field of the invention

Direct methanol fuel cells (DMFCs) that work at low temperatures (V. Baglio, A.

Di Blasi, E. Modica, P. Creti, V. Antonucci, A. S. Aricό, Int. J. Electrochem. Set, 1(2006), 71-79) or at intermediate temperatures (up to 150 ⁰C) and use solid proton electrolytes are suitable systems for the production of energy in the field of portable energetic applications (Jaesung Han, Eun-Sung Park, Journal of Power

Sources 112 (2002) 477-483) and electro traction (A.S. Aricό, S. Snirisavasan, V.

Antonucci, Fuel Cells, 2 (2001)). Similar to internal combustion engines, DMFC use continuous energy with a much higher utilization efficiency and low intrinsic emission of pollutants (A.S. Aricό, V. Baglio, P. Creti, A.D. Blasi, V. Antonucci, J.

Brunca, A. Chapot, A. Bozzi, J. Schoemans, J. Power Sources , 123 (2003) 107).

Since vehicle transportation represent a significant part of world energy consumption and considerably contribute to atmospheric pollution, the development of a system of an appropriate fuel cell would be an important result from both an economical and environmental point of view. In order to be competitive in the field of transportation, the DMFC must be reasonably cheap and capable of providing high power density.

At present, some problems hinder the development of these systems. These problems mainly consist in i) electrocatalysts that efficiently increase oxidation kinetics of methanol at electrodes ii) electrolyte membranes having high ion conductivity and low methanol diffusion and iii) methanol-tolerant electrocatalysts having high oxygen reduction activity.

At present, the cost of the whole system is mainly determined by the presence of noble metals in the catalyst and the use of Nafϊon^® membranes. The person skilled in this field still has to face some difficulties before DMFC become a power source available for vehicle traction.

For example, the development of alternatives to Nation® membranes no to only from the point of view of costs, but also from the minimization of methanol diffusion.

Moreover, even if DMFC are now rather far from being economically competitive with respect to internal combustion engines (ICE) and acid-lead batteries, they are reaching levels of power density such that, when combined with their design easiness, they become attractive for some special applications, like portable energy systems and remote power generators. A step forward in the development of low cost catalysts, membranes and gas diffusion layers (GDL) will be capable of filling the leap still existing between DMFC and ICE-drive vehicles (A.S. Aricό, P. Bruce, B. Scrosati, J-M. Tarascon, W. van Schalkwijk; Nature Materials, 4 (2005) 366-377).

Although attractive, the fuel cell technology still poses some technical problems to be solved, for example, those relating membranes for electron transport.

A number of patents and patent applications relate to new materials for making this kind of membranes, see for example WO 99/19930, WO 2006/040905, WO 2006/06008157, WO 2005/024989 and US 200409672. US 5,945,233 discloses the preparation of pastes or gels of polybenzoimidazole in acid solution and its use in fuel cells.

The same inventors and colleagues disclosed in Electrochemistry Communications 8 (2006) 1125-1131 new types of porous membranes prepared rearranging synthetic processes tried in the 90's by Bellcore Laboratories with the aim of using them successfully in lithium battery technology (J. M. Tarascon, A.S. Gozdz, C. Schumtz, F. Shokoohi, P.C. Warren, Solid State Ionics, 86-88 (1996) 49). These membranes involve a first cast of a semiliquid slurry formed by a copolymer of poly (vinylidene)fluoride-chloro tetrafluoroethylene, PVdF-CTFE, and by a dispersed ceramic filler with the addition of dibutylphthalate, DBP. DBP is then removed by extraction with diethyl ether in order to increase flexibility and porosity. At the end, this porous composite membrane is activated by swelling it with an aqueous acid solution.

Clearly, these membranes have a structure and a conduction mechanism totally different from the one of Nafion^® membranes, where in these latter proton transport is assisted by sulphate groups, while in the former is provided by the acid solution entrapped in the polymer matrix (S. Panero, F. Ciuffa, A. D'Epifanio, and B. Scrosati, Electrochem. Acta, 48 (2003) 2009). Besides, these PVdF-based porous membranes have a predicted cost which is lower by and order of magnitude with respect to the one of Nafion^® membranes.

The importance of the use of these membranes as separators of fuel cells was proved by Peled and co-workers (E. Peled, T. Duvdevani, A. Ahoron, A. Melman, Electrochem. & Solid-State Lett., 3 (2000), 525) and confirmed by our previous works (F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006), 1125-1131).

However, further improvements of proton conductivity and thermal stability are desirable and the present invention provides a technical solution thereto.

Summary of the invention

It has now been found that an immobilized proton conducting solution which is gelified in a suitable polymer matrix containing dispersed ceramic powder having dimensions of micro- or nanoparticles, provides high proton conductivity, low dispersion of methanol, satisfactory thermal stability, and good behaviour of direct methanol fuel cells using this system as separator at ambient temperature and especially at intermediate temperatures.

A further advantage of this system is the low cost, which is convenient in the fabrication of direct methanol fuel cells.

The advantages of the present invention derive from the stability of the proton- conducting immobilized gel combined with the swelling effect of the high porosity membrane. Therefore, it is an object of the present invention a nanoporous, composite, proton membrane comprising a polymer matrix compatible with proton environment, comprising ceramic, micrometric and submicrometric dispersed particle and having a gelified, proton-conducting solution immobilized in it.

Another object of the present invention is a method for preparing said membrane and its use in the manufacture of fuel cells.

A further object of the present invention is a fuel cell, in particular a direct methanol fuel cell, comprising a membrane as disclosed in the present invention.

Thanks to the properties of the membranes of the present invention, fuel cells using said membranes achieve optima performances with respect to the performances of fuel cells using commercial Nafion^® 117. The major potential advantages offered by the present invention are with no doubt the low cost membrane and the simpler manufacture.

The value of the diffusion of methanol plays a fundamental role in the selection of membrane the in DMFC applications. In fact, the diffusion of methanol influences different parameters of the cell, including energy efficiency (Bogdan Guraua, Eugene S. Smotkinb, Journal of Power Sources 112 (2002) 339-352). The present invention provides a significant improvement in the solution of the problem of the diffusion of methanol.

These and other objects of the present invention will be described in detail in the following section also by means of Figures and examples.

In the Figures:

Figure 1 represents SEM pictures of samples of PVdF-O membranes (A), PVdF- 10(B) membranes and EDS analysis of samples of PVdF-IO(C) membranes. For the identification of the samples, see Table 1.

Figure 2 shows Differential Scanning Calorimetry DSC (A) and Thermal Gravimetric Analysis (B) of the immobilized acid gel based on silica Ludox HS40, Bindzil. Figure 3 shows time evolution of conductivity of different samples of membranes at 25 °C (A), and related Arrhenius diagrams (B). For the identification of the samples, see Table 1.

Figure 4 shows the comparison between current-voltage and current-power curves of the laboratory prototypes of DMFC obtained at ambient temperature using the sample of immobilized acid gel PVdF-30 and using commercial Nation^® 117 as electrolyte separator (A), and between the current-voltage and current-power curves of laboratory prototypes of DMFC obtained at ambient temperature using the sample of immobilized acid gel PVdF-30 at ambient temperature and at temperature of 50 °C (B). For the identification of the samples, see Table 1.

Figure 5 shows a comparison between diffusion level of methanol of different samples of membrane studied in this work and the diffusion of methanol of the commercial membrane Nafion^® 117. For the identification of the samples, see Table 1.

Figure 6 represents time evolution of the voltage of the circuit of the laboratory prototype of the fuel cell using the sample of membrane with immobilized acid gel PVdF-IO as electrolyte separator at ambient temperature. For the identification of the samples, see Table 1.

Detailed description of the invention

For the purposes of the present invention, a polymer compatible with proton ambient is for example a polymer selected in the polyvinylidene family or a polymer selected in the Teflon^® family; a ceramic is an oxide of Zr, Ti, Al, Ce; a gelling agent is selected among the agents commonly known, which are compatible with proton ambient such as, for example, colloidal silica.

In a preferred embodiment of the present invention, nanoporous membranes are based on a matrix formed by a PVdF-CTFE copolymer with ceramic compounds SiO₂ dispersed therein.

The synthesis process is entirely disclosed in Electrochemistry Communications, 8 (2006) 1125-1131. Differently from what disclosed in this reference, sulfuric acid is gelified in the membrane by means of a suitable gelling agent.

In a first preferred form of embodiment, the agent gelling is colloidal silica, for example is 40% colloidal silica.

According to the present invention, the membrane, once formed according to the well-known method, is swelled at ambient temperature with the gelling agent for a suitable time, for example 12 hours. Then, an aqueous solution OfH₂SO₄ is added to the swelled membrane for gel formation.

The polymer is usually mixed with the required amount ceramic powder. Separately, a solution of porogen plasticizer component, such as dibutylphthalate, is prepared and subsequently added to the polymer-ceramic mixture in order to reach the complete dissolution of the polymer, so to obtain a semiliquid slurry homogeneous with the dispersed ceramic.

The slurry is then poured on substrate and cast in a film having a suitable thickness. After washing, in order to extract the porogen component, a highly porous and flexible membrane is obtained.

The materials and the working conditions are the conventional ones in this field and references can be found in F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006), 1125-1131.

According to the present invention, once the membrane is formed, it is swelled at ambient temperature with the gelling agent for a suitable time, for example 12 hours. Thereafter, a proton-conducting aqueous solution is added to the swelled membrane for the formation of the gel.

In a preferred form of embodiment of the present invention, the composite nanoporous membrane comprises a PVdF-CTFE polymer as polymer matrix in which SiO₂ is dispersed and the immobilized, proton-conducting gelified solution is H₂SO₄.

The following example further illustrates the invention. Example

PVdF-CTFE copolymer (Solef^® 32008) was intimately mixed in a ball mill with the suitable amount of ceramic powder (SiO₂ "fumed" silica, 99.8%, Cat. N. S5505 Aldrich, particle size 14 nm, surface area 200 m²/g ± 25 m²/g). The porogen, plasticizer component, namely dibutylphthalate (DBP, Aldrich) were dissolved in acetone, separately. The resulting solution was added to the mixed powder of PVdF-CTFE-SiO₂ and magnetically stirred for 16 hours at ambient temperature to obtain the complete dissolution of PVdF-CTFE and DBP, so to obtain a homogeneous, semiliquid slurry with SiO₂ dispersed therein. The semiliquid slurry was then poured on a glass substrate and cast in a 100 μm thin layer with a Doctor Blade. Membrane layers were repeatedly washed with diethyl ether to extract DBP, and to finally produce a highly porous, flexible membrane.

A number of samples with different SiO₂ content were prepared. Table 1 provides a list of samples and of their compositions.

Table 1

Sample SiO₂ Content (w/w%)

PVdF-O 0

PVdF-5 5

PVdF-IO 10

PVdF-20 20

PVdF-30 30

PVdF-40 40

The membrane samples were swelled at ambient temperature in Ludox HG 40 (40% silica colloidal Aldrich) for 12 hours. Then, a 6M H₂SO₄ aqueous solution was added to the swelled membrane for gel formation for 3-4 hours at ambient temperature.

The through-plane conductivity of the swelled membranes was determined by impedance spectroscopy, run on symmetric Pt/membrane sample/Pt cells, in a IHz - IMHz frequency range using a computer controlled Solarton 1260 FRA. The Differential Scanning Calorimetry, DSC, fu performed using a Mettler Toledo DSC821⁶ with a scanning range of 10 ⁰CmUi^"1, starting from 25°C and reaching the temperature of 160°, then cooling down to -4O⁰C and, at the end, heating up to 25°C.

Thermogravimetric Analysis TGA was run using a Perkin-Elmer at a scan rate of 5⁰C min^'1 in the 25 °C - 220 ⁰C temperature range.

For the fuel cell test at ambient temperature, a monolayer Membrane-Electrode assembly, ML_MEA, was fabricated following a procedure similar to the one adopted for the preparation of the electrolyte membrane of the sample.

The monolayer was obtained by first intimately mixing a blend of Super P carbon and Pt black (6:4 weight ratio) with PVdF powder (6020 Solvay-Solef Binder) in a 20% total weight. The mixture was dispersed in acetone and added with a Teflon emulsion in a 1:1 weight ratio. The resulting final suspension was mixed with DBP in a 1:2 weight ratio. The semiliquid slurry was dried for 15 minutes at 70 ⁰C. The procedure gave a highly viscous paste, which was pressed at 70°C and 1 ton/cm² to obtain a thin, homogeneous membrane. Lastly, this membrane was washed with diethyl ether to extract DBP and promote porosity. The Pt loading in this electrode porous membrane was 4 mg/cm². Two of these electrode monolayer membranes, one at the anode end and the other at the cathode end, were combined with the selected electrolyte membrane and pressed together to obtain the MEA.

For the tests on fuel cells at intermediate temperature (50 °C) a three layer, carbon-based electrolyte membrane assembly was fabricated, TL_MEA. The TL_MEA is formed by:

a) a commercial, treated or non treated, 100 micron PTFE porous support (Electrochem^® ) used as support layer both for cathode and for anode;

b) a diffusion layer to make reagent flow homogeneous on the catalytic layer, fabricated through the following steps: i) a water/isopropanol homogeneous suspension was prepared by mixing carbon powder (Super P) and a suitable amount of PTFE, and spread on carbon paper with a Doctor-Blade; ii) the diffusion layer was air-dried at a temperature of 120⁰C for 1 hour; iii) the diffusion layer was thermally treated at a temperature of 28O⁰C for 30 minutes; iv) the diffusion layer was thermally treated at a temperature of 350⁰C for 30 minutes (sintering temperature).

c) A catalysis layer was prepared from a homogeneous suspension formed by the amount of Pt/Ct catalyst (20 % w/w Pt Electrochem), solution of Nafion^® (5% Nafion, Aldrich), with ethanol as solvent and deposited by spraying on the diffusion layer and dried at 70°C for 30 minutes.

Two of these three layer electrolyte membranes, one on the anode end and the other on the cathode end, were combined with the selected electrolyte membrane and pressed together to obtain the final MEA.

Current- voltage curves of the cells were obtained with a PAR 273 A potentiostat under air flow at the cathode end and H₂Sθ₄-acidified 2M methanol aqueous solution flow, at the anode end.

The diffusion of methanol was determined using a U-shaped cell having two compartments separated by the given sample of membrane. One compartment was filled water and the other with a methanol aqueous solution. At fixed time intervals, the samples on the water side were analyzed with gas chromatography to monitor the diffusion of methanol through the membrane. Stationary phase was polyethylene glycol (Carbowax).

Figure 1 shows scanning electron microscope of the samples PVdF-O (A), PVdF- 10 (B) and the electron dispersion spectroscopy (EDS) of the sample PVdF-IO (C). High and uniformly distributed porosity is clearly visible in both the examples.

The presence of this extended porosity, involving a sequence of empty voids of nanometric dimensions (198 nm) is essential to favour immobilized acid gel entrapment (F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006) 1125-1131). Figure 1C shows also a uniform distribution of the ceramic filler, which induces a uniform absorption of the silica gel, consequently a uniform distribution of the immobilized acid gel and a good behaviour in the cell using this membrane as electrolyte. Figure 2 reports differential scanning calorimetry, DSC, (A) and Thermal Gravimetric Analysis (B) of the immobilized acid gel on silica Ludox HS40, Bindzil. By observing the DSC profile, Figure 2A₅ it is possible to see a broad endothermic peak, starting from the temperature of 115°C. This peak can be assigned to the gel thermal decomposition in crystalline phase and in an acid aqueous phase.

The absence of recrystallization peak, during the cooling phase, confirms the absence of the amorpho-crystalline transition and confirms the hypothesis of the decomposition. Thermal Gravimetric Analysis, TGA, of the immobilized acid gel, obtained by using silica Ludox HS40, Bindzil, reported in Figure 2B, shows a weight loss associated with the peak centred at the temperature of 100°C in the derivative curve, assigned to the water evaporation process.

The membrane conductivity (F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006) 1125-1131) at ambient temperature remains unchanged for several days, as shown in Figure 3 A, reporting the time evolution of the conductivity of different samples of membrane at 25 ⁰C. Arrhenius diagrams, reported in Figure 3B, clearly demonstrate that the conductivity remains constant also going from ambient temperature to the temperature of 80 °C. This demonstrates that the membranes are stable and have a non detectable release of phase gel at intermediate temperature.

The best membranes have conductivity of the order of 10^"2S cm^"1, at high ceramic content, and are very fit for their application in fuel cells, designed to operate in the temperature range of 25-80⁰C.

The response of laboratory prototypes of fuel cells was examined using the best PVdF-based composite immobilized acid gel membrane as electrolyte separator.

A set of monolayer electrode membrane, MLJVlEA, was used at ambient temperature and a set of three- layer electrode membrane, TL_MEA, was used at intermediate temperature of 50°C. Figure 4A shows the comparison between the current- voltage and current-power curves of DMFC of laboratory type at ambient temperature using the PVdF-30 immobilized gel sample and commercial Nafion^®

117 as electrolyte separator. The cell based on PVdF-30 immobilized gel shows a much better response (a power density of about 2 Wcm^" and a current in the order of 33 mAcm^"2) with respect to the cell based on commercial Nafion^® 117 (a power density of about 1.1 mWcm^"2 and a current of the order of 18 mAcm^'2). The performances of the cell based on PVdF -30 immobilized acid gel substantially increase at the intermediate temperature of 50 °C (a power density of about 6.1 mWcrn^"2 and a current of the order of 71 mAcm^"2), as shown in Figure 4B. This increase can be referred to the high catalyst activity at this value of temperature and the positive effect of the three layer membrane-electrode assembly, TL_MEA, in terms of flow homogeneity and optimal diffusion of reagents.

The comparison between the diffusion level of methanol of different samples of membrane of the present invention and the diffusion of methanol of the commercial Nafion^® 117 membrane is shown in Figure 5. Other than the increase of the diffusion of methanol for the membranes having progressively increasing ceramic content, one can also observe for these membranes a methanol diffusion value comparable with the observed value for the commercial Nation^® 117 membrane. This unexpected result confirms the applicability of the membranes of the present invention in the field of DMFC.

An interesting approach for the evaluation of the diffusion degree is based on the observation of the evolution in time of the voltage of the open circuit of DMFC, using the composite membranes as electrolyte separator (F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006) 1125-1131). Figure 6 reports the evolution in time of the voltage of the open circuit at ambient temperature of the laboratory prototype of the fuel cell using the sample of immobilized acid gel membrane PVdF-IO as electrolyte separator (F. Croce, J. Hassoun, C. Tizzani, B. Scrosati, Electrochemistry Communications 8 (2006) 1125-1131). The presence of diffusion of methanol is revealed by OCV decay and the percent loss, for this membrane, has a value of 3.9%. This value can be used for the evaluation of the diffusion degree (Z. H. Wang and C. Y. Wang, Journal of Electrochemical Society, 150,4(2003) A508-A519).

Claims

1. Composite, nanoporous, proton membrane, comprising a polymer matrix compatible with proton ambient with ceramic particles of micrometric and submicrometric dimension, and having a proton-conducting gelified solution immobilized therein.

2. Membrane according to claim 1, wherein said polymer compatible with proton ambient is a polymer selected from the group consisting of: polyvinylidene family and Teflon^® family.

3. Membrane according to claim 1 or 2, wherein said ceramic is an oxide selected from the group consisting of Si, Zr, Ti, Al and Ce.

4. Membrane according to any one of claims 1 to 3 wherein said gelling agent is colloidal silica.

5. Membrane according to any one of claims 1 to 4 wherein said polymer is a PVdF- CTFE copolymer.

6. Membrane according to any one of claims 1 to 5 wherein said ceramic consists of SiO₂ particles.

7. Membrane according to any one of claims 1 to 6 wherein said proton conducting solution is H₂SO₄.

8. Method for the preparation of the membrane of any one of claims 1 to 7 comprising the preparation of a microporous membrane from said polymer matrix compatible with proton ambient and said ceramic particles followed by jellification of the proton conducting solution contained therein.

9. Use of the membrane of any one of claims 1 to 7 in the manufacture of fuel cell.

10. Fuel cell comprising a membrane of any one of claims 1 to 7.