Background of the invention

The present invention relates to a porous proton-conducting material which comprises an inorganic-organic network

representing a metal-organic complex of a coordinating metal cation with an aromatic scaffold compound and a proton- conducting dopant provided in the pores of said network.

Fuel cells hold great promise for applications ranging from portable electronics, through automotive use to stationary- power generation. A vast number of polymers have been

evaluated as materials for proton exchange membrane (PEM) fuel cells. Currently as a benchmark, sulfonic acid functionalized perfluorinated polymers (PFSAs) such as Nafion ^® feature excellent proton conductivity together with mechanical and chemical stability. Despite the wide use of Nafion ^® and other state-of-the-art PFSAs, their maximum performance in a fuel cell is limited by the boiling point of water since proton conductivity strongly depends on hydration of the membrane. Therefore non-water-based PEMs providing high, constant proton conductivity at intermediate temperatures (110 - 150°C) and low relative humidity (RH) are one of the biggest challenges for new separator materials.

Candidate "solvents" alternative to water include nitrogen heterocycles and phosphonic acids. While initial results effectively showed that such solvents could be immobilized on to polymer backbones, the low concentration of functional groups led to rather low proton conductivity values (e.g. Li et al. in Power Sources 2007, 172, pp. 30-38). Unfortunately, these membranes swell in cases of higher ion exchange capacity, which compromises their mechanical stability.

In contrast to polymer membranes, mesoporous materials

comprising zeolites and organic-inorganic hybrid materials as electrolytes have certain advantages: They combine a high density of ionic groups as proton source in the mesopores with mechanical and structural durability of their inorganic pore scaffolds. This inorganic wall structure prohibits swelling when hydrated. Moreover, desirable water molecules in the pores are retained by tailoring the sturdy mesostructured channels. It was recently shown, that proton diffusivity and the water sorption process in mesoporous silica can be

significantly increased by introducing hydrophilic groups such as sulfonic acid into the framework's structure (Fujita et al. Chem. Mater. 2013, 25, pp. 1584-1591) . However, the proton conductivity of these materials is still quite low, typically in the range of from 1 -10 ^"11 S/cm to 5 ·10 ^~3 S/cm.

WO 2008/073901 describes metal-organic solids for use in proton exchange membranes. These proton exchange membranes may comprise a salt of a metal cation and a phenyl-based aromatic compound having at least one acidic proton. As evident from the experimental data shown therein, this reference does not contemplate the use of an under-stoichiometric amount of metal cations in relation to the acidic groups of the aromatic compound. The proton exchange membrane may also comprise a dopant, in particular selected from an N-heterocycle and an oxoacid. However, the exemplary materials actually prepared showed rather low conductivities, in particular above 100 °C, and the structures tended to collapse at elevated

temperatures . In view of the drawbacks of the prior art, the main object of the present invention was to provide improved materials with very high and stable proton conductivities over a wide

temperature range, including temperatures above 100°C, which are advantageously applicable in fuel cells, in particular for automotives, electrode materials and catalytic devices.

This objective has been achieved by providing the proton- conducting materials according to claim 1, the proton exchange membrane according to claim 11, the fuel cell according to claim 12 and the method of synthesis according to claim 13. Additional aspects and more specific embodiments of the invention are the subject of further claims.

Description of the invention

The porous proton-conducting material of the invention is based on an inorganic-organic network (NET) , which inorganic- organic network represents a metal-organic complex of a coordinating metal cation with an aromatic compound

corresponding to the following general structure

and comprising a) a rigid core consisting of at least one of the following building bloc

tetraphenylmethane, adamantyl, b) a spacer region comprising at least one rigid unit selected from the group consisting of phenylene, polycyclic aromatic hydrocarbons, and acetylene, linked to the rigid core, and c) a peripheric region comprising at least 2 acidic

substituents selected from the group consisting of phosphonic acid, hypophosphorous acid, sulfonic acid, carboxylic acid, arsenous acid and hypoarsenous acid, wherein the acidic substituents are either directly bonded to the rigid units of the spacer region or via a linker group R, wherein the molar ratio of the metal cation to the acidic groups of the aromatic compound is below 1 and a fraction of uncoordinated free acidic groups is still present in the complex, and wherein the porous proton-conducting material further comprises a dopant, which is an intrinsic proton conductor selected from the group of oxoacids and N- heterocycles , in the pores thereof.

The coordinating metal cation is not especially limited and may be principally any metal cation which is capable to form coordinating bonds with one or more acidic groups. Multivalent cations, in particular divalent or trivalent cations, are generally preferred.

More, specifically, the coordinating metal cation is selected from the group comprising divalent cations such as Ca ²⁺ , Mg ²⁺ , Ba ²⁺ , Fe ²⁺ , Cu ²⁺ , Zn ²⁺ , Sn ²⁺ , Zr ²⁺ , Ni ²⁺ , Co ²⁺ , Sr ²⁺ , or trivalent cations such as Fe ³⁺ , Al ³⁺ , Ga ³⁺ , In ³⁺ , As ³⁺ , Sb ³⁺ , Bi ³⁺ , Cr ³⁺ , Ru ³⁺ , Rh ³⁺ , Sc ³⁺ , Y ³⁺ , and lanthanides, as well as Ti ⁴⁺ , Mn ^{2+ " 7+} .

The dopant may be any proton-conducting oxoacid or N-hetero- cycle known in the art which fits into the pores of the respective network.

More specifically, the dopant is selected from the group of oxoacids comprising sulphuric acid, sulphurous acid, nitric acid, nitrous acid, phosphoric acid, phosphorus acid,

hyphophosphorus acid, selenic acid, selenous acid, chromic acid, chromous acid, arsenic acid, arsenous acid, hypoarsenous acid, iodic acid, iodous acid, perbromic acid, bromic acid, perchloric acid, chloric acid and chlorous acid, and N- heterocycles comprising a substituted or unsubstituted 1,2,3- triazole, 1, 2 , 4-triazole, triazine, tetrazole, oxazole, benzoxazole, isoxazole, thiazole, benzothiazole, 1,3,4- thiadiazole, lactam, imidazolidone, oxazolidinone, hydantoin, pyrrole, imidazole, benzimidazole, pyrazole, indole, carbozole, oxindole, 7-azaindole, dihydropyridine, pyridine, quinoline, pyrazine, piperazine, purine or pyrimidine.

Preferably, the dopant is an oxoacid and in particular

selected from the group comprising H ₃ P0 ₂ , H3PO3, H3PO4, H ₂ SO ₄ and mixtures thereof.

In one preferred embodiment of the invention the rigid core consists of a relatively small aromatic ring system such as a phenyl ring, however, larger polycyclic rings systems are principally also suited as cores.

The core is linked to one or more rigid units which form a spacer region surrounding the core region.

The maximal number of linked rigid units depends i.a. on the available linking positions of the respective core. Typically, the number of rigid units will be in the range from 3 to 12 or more, such as 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12.

The rigid units which may be the same or different for one core molecule are selected from the group consisting of phenylene, polycyclic aromatic hydrocarbons, and acetylene, with phenylene especially preferred. It is also possible to combine an aromatic ring system such as phenylene with an acetylene group in a chain-like arrangement to form a combined rigid unit.

At least some of the rigid units of the spacer region are linked (substituted) either directly or via a linker group R with acidic substituents forming the peripheric region.

The term "rigid" as used herein means a fixed torsion angle of the bounded units to one's another prohibiting any flexibility except rotation about the carbon-carbon bond axis. The minimal number of acidic substituents of one aromatic scaffold compound is 2, in order to enable the formation of a network, however, a larger number of acidic substituents is generally preferred. A larger number of acidic substituents provides more protons and also more potential binding sites for coordination with metal cations.

The maximal number of acidic substituents depends on the available attachment positions on the respective rigid units.

Typically, at least 3 acidic substituents, preferably at least 4, 5, 6 or more acidic substituents will be present.

The acidic substituents are selected from the group consisting of phosphonic acid, hypophosphorous acid, sulfonic acid, carboxylic acid, arsenous acid and hypoarsenous acid, with phosphonic acid especially preferred.

The linker group (s) R in the peripheric region of the above aromatic compound may be principally any linker known in the art.

More specifically, the linker is a short-chain alkylen linker R = C _n H _2n (n = 1-6, preferably 1-4) or an ether linker R = 0- (CH) n (n = 1-6, preferably 1-4).

The molar ratio of metal cation to aromatic compound in the inorganic-organic network complex is in the range from

(2/charge cation) :1 to ( (n'A) /charge cation) :1 (n'A = number of acid groups), preferably from 1:1 to 8:1, in particular from 1:1 to 1:4.

The aromatic ring systems of the core or of the rigid units may also comprise, additionally to the substituents

specifically indicated in the text and structural formulae herein, further substituents which do not interfere with the relevant functional features of the inorganic-organic network. Such substituents may be for example lower alkyl substituents (with 1- 4 carbon atoms) , halogens, in particular F, Cl, nitro, cyano, ester or ether.

In preferred embodiments of the proton-conducting material of the invention, the aromatic compound has one of the following basic structural formulae:

la ;

IIIA;

IVa;

Vb;

Vila;

Villa;

IXa;

Xa;

Xb;

with A in each of the above formulae I - X being an acidic group independently selected from the group consisting of phosphonic acid, hypophosphorous acid, sulfonic acid,

carboxylic acid, arsenous acid and hypoarsenous acid, B = H, F or CI, n being the number of the respective substituents

(n = 1-3 for A; n = 2-4 for B) ; R being an alkylene, in particular C1-C4 alkylene, or an ether linker group 0-(CH) _n

(n = 1-6) .

More specifically, the aromatic compound is selected from the group of compounds listed below:

Compound la

Q = F, CI

Compound la'

Compound la'" ,

or other partly halogenated analogues of compound la,

Compound lb;

Compound lc;

PA = P(0)(OH) ₂

Compound Id;

Compound le or structural analogues of compounds lb-le above, derived from said compounds lb-le above by completely or partially

halogenating aromatic rings in the same manner as indicated for compound la and/or by varying the substitution pattern of the PA substituents;

Compound 2a with R being selected from CH ₂ , C ₂ H ₄ , C ₃ H ₆ and C ₄ H _g ;

Compound 2b with R being selected from CH ₂ , C ₂ H ₄ , C ₃ H ₆ and C ₄ H ₈ ;

Compound 3a;

with PA = P(O) (OH) 2

Compound 3b;

Compound 4 ^N ;

Compound 6 ;

Compound 9 or structural analogues of compounds 2-9 above, derived from said compounds 2-9 above by completely or partially

halogenating aromatic rings in the same manner as indicated for compound la or compound 4 and/or by varying the

substitution pattern of the PA substituents .

In an especially preferred embodiment of the invention, the aromatic compound of the inorganic-organic network is phosphonic acid functionalized hexakis (phenyl ) benzene (HPB) .

The 6 phenyl rings of the spacer region of this molecule may comprise 6 phosphonic acid substituents in para-position

(hexakis (p-phosphonato (phenyl ) benzene, Compound la), in meta- position (Compound lb) or in mixed positions (e.g. Compound lc) . Also, the degree of substitution may be increased up to 12 by di-substituting one ore more of the phenyl rings (e.g. Compound Id, Compound le) .

In a further preferred variation of this basic molecule, one or more of the phosphonic acid substituents are linked with the phenyl rings via an alkylene or ether linker group R as defined above (e.g. Compounds 2a and 2b).

In another specific embodiment of the invention, the phenyl rings are replaced by polycylic ring systems such as biphenyl or larger entities. These rings systems may also have varying substitution patterns (e.g. Compounds 3a-3c, 4, 4 ).

As evidenced by the data shown in the experimental section below, the proton-conducting materials of the invention exhibit excellent conductivities which are comparable or under some conditions even superior to the conductivities obtained with the benchmark materials of the prior art such as Nafion and doped PBI. Typically, a conductivity of at least 1 x 10 ^~3 S/cm at a temperature in the range from 120°C to 140°C is reached and some materials even yield conductivities of at least 1 x 10 ^"2 S/cm in this temperature range.

These inorganic-organic networks with very high proton

conductivity may be advantageously used in various

applications involving the transport and/or exchange of protons. Specific applications of interest are for example in the fields of proton exchange membranes, fuel cells, electrode materials and catalytic devices.

Therefore, closely related aspects of the invention pertain to such uses and in particular to a proton exchange membrane and a fuel cell comprising the proton-conducting materials of the present invention.

A further related aspect of the invention pertains to methods for preparing inorganic-organic networks (NETs), also called metal-organic frameworks (MOFs) , as defined above and doping the same to obtain such proton-conducting materials.

A simple and effective method for preparing an inorganic- organic network as defined above comprises the following steps :

The desired aromatic scaffold compound (e.g. p-6PA-HPB) and a salt comprising the desired metal cation (e.g. Al ³⁺ ) are separately dissolved in a suitable solvent (e.g. H ₂ 0, alcohols, DMSO, DMF, formamides, nitrobenzene or nitromethane) ,

preferably by means of ultrasonication, and the solution of metal salt is added to the solution of the aromatic compound (preferably dropwise) . The reaction mixture is stirred for a given time, typically 1-7 days, while maintaining a

predetermined temperature (typically an elevated temperature if the solvent is H ₂ 0) . The precipitated product is separated from the reaction mixture, e.g. by means of centrifugation, and dried, preferably under vacuum and/or an elevated

temperature (e.g. in the range from 100°C to 150°C, more specifically 110°C to 130°C, such as about 120°C) .

The metal salt may be any salt which is solvable in the desired solvent, in particular a solvent in which the aromatic compound is solvable as well.

In particular in the case of an aqueous solvent, the metal salt may be a nitrate, halide, such as a iodide, bromide or chloride, an anhydrous or hydrated sulphate etc. Other

suitable metal salts can be readily identified by a skilled person with routine experiments.

A preferred method for obtaining a doped proton-conducting material of the invention comprises at least the following steps :

providing an inorganic-organic network which is a complex of an aromatic compound as defined in claim 1 and of a

coordinating metal cation;

drying the inorganic-organic network (NET) under vacuum and/or at elevated temperature;

incubating the inorganic-organic network (NET) in an aqueous solution of a dopant for a predetermined period of time and at a predetermined temperature; and

separating the product having dopant molecules

incorporated in the pores thereof from the aqueous solution.

In a more specific embodiment, the dopant is H3PO3 and the aromatic compound of the metal-organic complex has one of the basic structural formulae I - X of claim 5, and the method comprises the steps drying of the inorganic-organic network under vacuum and/or elevated temperature;

incubating the dried inorganic-organic network (NET) in an aqueous solution of a dopant for a period of time at a temperature in the range from -80°C to 200°C, preferably in the range from room temperature to 120 °C;

separating the product having dopant molecules

incorporated in the pores thereof from the aqueous solution by filtration or, preferably, centrifugation

optionally drying the doped product under vacuum and/or at elevated temperature.

More specifically, the aromatic compound of the inorganic- organic network is one of the compounds of claim 7. which is complexed with an under-stoichiometric amount of the

coordinating metal cation Al ³⁺ .

In a further specific embodiment, the dopant is H ₃ P0 ₃ and the dry inorganic-organic network is incubated in an 0.1 N to 1.5 N, preferably 0.4 N to 1.2 N, aqueous solution of H ₃ P0 ₃ for a time period of 1-7 days, preferably a time period of 2-3 days, at a temperature in the range from 20°C to 120°C, preferably in the range from room temperature (ca. 24°C) to 50°C.

In this manner, a proton conducting material can be obtained which comprises a high amount of the dopant, typically in an amount of at least 5 wt.-%, such as from 5 wt.-% to 11 wt.-% for p-6PA-HPB or even more for larger aromatic compounds, with respect to the weight of the inorganic-organic network (NET) .

Brief Description of the Figures

Fig. 1 shows schematically the materials used and synthetic conditions applied for the synthesis of 3 exemplary inorganic- organic network materials of the invention (Al-HPB-NET 3:1, Al-HPB-NET 2:1 and Al-HPB-NET 1:1).

Fig. 2 shows SEM micrographs of a) Al-HPB-NET 3:1, b) Al-HPB- NET 2:1 and c) Al-HPB-NET 1:1.

Fig. 3 shows the pore size distribution obtained by the BJH method for a) Al-HPB-NET 3:1, b) Al-HPB-NET 2:1 and c) Al-HPB- NET 1:1.

Fig. 4 shows the maximum RH as a function of temperature at 1 bar H ₂ 0 pressure.

Fig. 5 shows plots of proton conductivity vs. temperature under 1 bar H ₂ 0 atmosphere for Al-HPB-NET 3:1, Al-HPB-NET 2:1 and Al-HPB-NET 1 : 1 in comparison with Nafion 117.

Fig. 6 shows plots of proton conductivity vs. temperature for doped (H ₃ P0 ₃ ) Al-HPB-NET 3:1, doped Al-HPB-NET 2:1, doped Al-HPB-NET 1:1 and ext. doped Al-HPB-NET 1:1 in comparison with Nafion 117.

Fig. 7 shows Arrhenius plots of proton conductivities and activation energies (E _a ) of doped Al-HPB-NET 1:1 at various relative humidities (RH) .

Fig. 8 shows plots of proton conductivities of doped Al-HPB- NET 1:1 and Nafion 117 vs. RH at 55°C.

Fig. 9 shows plots of proton conductivity vs. temperature under 1 bar H ₂ 0 atmosphere for doped PBI (H ₃ P0 ₄ doping level of 5.6) and ext. doped Al-HPB-NET 1:1.

The following non-limiting examples are provided to illustrate the present invention in more detail, however, without limiting the same to the specific features and parameters thereof .

EXAMPLE 1

Preparation and characterization of aluminum-based inorganic-organic-networks

All used chemicals were of commercial reagent grade and were used as received. The experimental procedure for the synthesis of hexakis (p-phosphonatophenyl ) benzene (p-6PA-HPB) is

described in the literature (e.g. Miillen et al., Adv. Funct . Mater. 2011, 21, 2216 - 2224). Hexakis (p-phosphonatophenyl ) - benzene is also disclosed (as formula XX) in EP 2 264 040 Al .

The necessary amounts of HPB and salt were dissolved

separately and mixed together in an Ace Glass pressure tube. The formed suspensions were heated at 90°C for three days. The precipitated products were filtered off, washed with water followed by centrifugation . Finally, the obtained products were dried under vacuum at 120°C. Synthetic details and yields are given in Table 2 for 3 materials.

The same general method is applicable (and has been applied) for linking other aromatic compounds, such as triphenylbenzene (TPB) , with a coordinating metal cation.

Table 2: Synthetic details and yields of aluminium based organic-inorganic hybrid materials.

Al-HPB-NET Al-HPB-NET Al-HPB-NET 3:1 2:1 1:1

(HPB) / itimol 0.28 0.28 0.46

(HPB) / mg 150 150 250 n (A1(N0 ₃ ) ₃ ·9Η ₂ 0)/ 0.84 0.56 0.46 mmol

m (A1(N0 ₃ ) ₃ ·9Η ₂ 0)/ 315 210 173 mg

Solvent H ₂ 0 H ₂ 0 H ₂ 0

V (solvent HPB) / 3 3 5

ml

V (solvent salt) / 3 3 5

ml

Time/ d 3 3 3

Temperature/ °C 90 90 90

Yield/ % 50 46 30

All synthesized materials are insoluble in polar as well as in nonpolar solvents. They exhibit high thermal stability as evidenced by thermogravimetric analysis ( T _dec > 550°C). An initial mass loss at about 100°C is assigned to evaporation of entrapped residual water.

Synthesis of a hybrid material as well as retention of

functional acidic groups, initially present in the linker, were confirmed by Fourier transform infrared spectroscopy (FTIR) . All aluminum-based hybrid materials showed the typical OH-stretching vibrations (~ 2700 cm ^-1 ) of phosphonic acid groups and of the P=0 at 1138 cm ^-1 . The characteristic

stretching of P-phenyl linkage at 1389 cm ^-1 as well as other multiple modes involving the PC>3-group in both spectra are present .

X-ray powder diffraction patterns of the networks indicated a poor degree of crystallinity . They display two diffraction peaks in the small angle region (θχ = 13.3 - 13.6 A, θ ₂ = 7.5 - 7.8 A). In all cases, the diffraction peaks do not correspond to the characteristic peaks of the salt, deployed during synthesis. This proves complete wash-out of the latter. The various coordination modes and protonation states accessible to a RPC>3 ² ~ group of HPB with an under-stoichiometric amount of aluminum cations are challenging to control. Due to the strength of ligation, Al-HPB-NET readily precipitates in a partially crystalline state during mixing of linker and connector and contains hydrophilic proton-conducting pores distributed distinctly over the entire material.

The chemical structure was further confirmed by ¹ H, ³¹ P and ²⁷ A1 solid state NMR spectroscopy. The signal of the ¹ H MAS NMR spectrum for Al-HPB-NET 3:1 reveals no significant changes in chemical shift or signal width as compared to the already broad signal of the starting material HPB. Due to the under- stoichiometric usage of Al ³⁺ , Al-HPB-NET 3:1 points a

significantly broader chemical shift dispersion in the ³¹ P { ¹ H} CP MAS spectrum in direct comparison to the linker. The peak maximum at 18 ppm is shifted up-field to 8 ppm in the spectrum of Al-HPB-NET 3:1 indicating different phosphorous sites inside the network. The ²⁷ A1 MAS spectrum detects three broad peaks of different intensities at 10 ppm, 0 ppm and -25 ppm. Investigations of the other ionic networks Al-HPB-NET 2:1 and Al-HPB-NET 1:1 reveal identical information. These results, in combination with the previous X-ray study, confirm that a semi-crystalline material is formed after introducing Al ³⁺ to HPB.

The SEM micrographs of all materials display a sponge-like morphology with different degree of surface roughness (Fig. 2). The micrograph of Al-HPB-NET 3:1 mainly reveals that there are small heterogeneous particles of about 200 nm width and also spheres of different sizes ranging from Ιμπι to 2μπι (Fig. 2a). The micrograph of Al-HPB-NET 2:1 shows homogeneous submicron particles of 20 nm that grow into bigger aggregates with holes in the range of 50 μιη to 200 μπι at the surface (Fig. 2b). The micrograph of Al-HPB-NET 1:1 shows a rough closed-cell surface. On higher magnification it becomes evident that the surface is built up by very small needles of about 100 nm length (Fig. 2c) .

The textural characterization of the Al-HPB-NETs reveals a mesoporous character of the materials with pore sizes ranging from 4 to 20 nm. The low surface areas might be due to the insertion of bulky non-complexed phosphonic acid groups in the obtained porous networks. Because of the very small amount of Al ³⁺ applied, Al-HPB-NET 1:1 better correlates with a chainlike arrangement than a dense framework which might be the reason for the very low surface area of about 20 m ² g ^"1 .

Proton conductivity studies were carried out under 1 bar H ₂ 0 atmosphere by increasing the temperature until 180°C, i.e. by decreasing RH until 20%. Fujita et al. {Chem. Mater. 2013, 25, 1584-1591) stated an improved proton diffusivity in mesopores by high densification of acidic groups even at very small amounts of adsorbed water, generally speaking at low RH values. These observations are in good agreement with. the present proton conductivity results which reveal high

conductivity values over a broad range of temperature (Fig. 5) . Even though proton conductivity of the Al-HPB-NETs is inferior to that observed for Nafion ^® 117, they present certain advantages since proton transport within the networks is less temperature-dependent. Nafion ^® 117 strongly suffers from the loss of water in its proton-conducting channels (vehicle mechanism) , whereas all Al-HPB-based materials show a flat temperature response.

The slope for the conductivity plots significantly decreases from Al-HPB-NET 3:1 to Al-HPB-NET 1:1 and exhibits a nearly constant proton conductivity for the latter. While the amount of Al ³⁺ is stepwise decreased for the three materials from Al- HPB-NET 3:1 to Al-HPB-NET 1:1, the number of free phosphonic acid groups able to participate in the conductivity mechanism goes up leading at the same time to an increase in hydro- philicity of the pores inside the materials.

EXAMPLE 2

Preparation and characterization of doped

aluminum based inorganic-organic-networks

In order to further improve proton conductivity of the networks prepared in Example 1, the embedded volatile H ₂ 0 molecules in the framework pores were replaced by phosphonic acid molecules. They were chosen as doping agent due to their high proton mobility and tendency to build up strong hydrogen bonded networks. The storing of H3 PO3 occurs by non-covalent , electrostatic interactions between linker bounded acidic groups and mobile phosphonic acid.

To empty the framework pores from incorporated H ₂ 0 molecules, all Al-HPB-NET samples were heated at 100°C for 36 h under high vacuum and were stirred in an 0.4 M aqueous solution of H3 PO3 for 2 days at RT . In view of its superior proton conductivity Al-HPB-NET 1:1 was additionally subjected to an extended (ext.) doping procedure (1.2 M aq. H3 PO3 solution, doping time: 7 days). The samples were separated from the doping agent solution by centrifugation and were finally heated in a vacuum oven at 50 °C for 12 h to get rid of the trapped volatile H ₂ 0 molecules.

An analogous doping procedure was used for Al-TPB networks (compare Table 6 below) .

Synthetic details and yields are given in Table 3 for 4 materials . Table 3: Synthetic details and yields for doping of aluminium based . organic-inorganic hybrid materials .

Al-HPB-NET Al-HPB-NET Al-HPB-NET Al-HPB-

3:1 2:1 1:1 NET 1:1 doped doped doped ext.

doped

M (NET) / mg 150 150 150 150

Solvent 0.4 M H3 PO3 0.4 M H3PO3 0 .4 M H3PO3 1.2 M

(aq) (aq) (aq) H3 PO3 (aq)

V (solvent)/ 10 10 10 30 ml

Time/ d 2 2 2 7

Temperature/ RT RT RT RT

°C

Yield/ % 43 46 30 22

The entire doping process might be envisioned as immersing the sponge-like NETs into a concentrated solution of the doping agent H3 PO3 . Due to its porous nature the sponge will absorb the intrinsic proton conductor and keep it sealed into its pores .

The amount of H3PO3 absorbed by the respective NET structures was determined by a potentiometric titration of the doped samples against a 0.01 M NaOH solution. Doping levels for the above Al-HPB materials are given in Table 4.

Table 4. Doping levels of H ₃ P0 ₃ -doped Al-HPB-based networks

Al-HPB- Al-HPB- Al-HPB-NET

NET 3:1 NET 2:1 1:1

Doped Doped Ext.

Doped Doped

6.8% 6.2% 6.9% 10.6% Plots of proton conductivity for the materials of Table 4 above were measured above 100 °C under 1 bar H ₂ 0 atmosphere.

Fig. 6 shows the corresponding plots of proton conductivity versus temperature for doped Al-HPB-NET 3:1, doped

Al-HPB-NET 2:1, doped Al-HPB-NET 1:1, ext. doped Al-HPB-NET 1:1 and Nafion ^® 117.

All doped materials present a significant increase in

conductivity in comparison to their undoped analogues (compare Figures 5 and 6). Contrary to Nafion ^® 117, they have almost constant proton conductivity with increasing temperature. All networks, except doped Al-HPB-NET 2:1, exhibit higher values of proton conductivity than Nafion above 160°C. Conductivity data are given in Table 5.

Table 5. Conductivity data of H ₃ P0 ₃ -doped Al-HPB-NETs

Ext. doped Al-HPB-NET 1:1 features the best conductivity performance due to its superior acid take-up of about 11%. Fig. 9 demonstrates that the conductivity performance of this material is similar to that of doped polybenzamidazole (PBI) having a H ₃ P0 ₄ doping level of 5.6 which corresponds to an acid take-up of 177 wt.-%. Thus, the material of the present invention is advantageous in that considerable less dopant is required to achieve the same performance.

However, ext. doped Al-HPB-NET 1:1 suffers from a lack in mechanical stability under some conditions. The increased amount of incorporated acid leads to a material that is stable under dry experimental conditions. The solid becomes sticky and even shows some slight deformation under very high values of RH (> 85%). At present, doped Al-HPB-NET 1:1 (with the shorter time of exposition to the dopant) appears to represent a superior material which combines very high proton

conductivity with structural durability even under fully immersed conditions.

In order to confirm this high conductivity and to gain deeper insights into the proton conductivity mechanism of doped

Al-HPB-NET 1:1, investigations at 55°C as a function of RH haven been performed and compared (see Fig. 8). Contrary to the vapour measurement above 100 °C under 1 bar H ₂ 0 atmosphere, there is a constant increase in conductivity when starting with a well-dried material indicating that the conductivity mechanism is at least water-assisted, if H ₂ 0 molecules are present .

Conductivity measurements of doped Al-HPB-NET 1:1 at various RHs (80, 50 and 15%) above 25°C were performed and the

corresponding Arrhenius plots of the conductivities and activation energies (E _a ) calculated (Fig. 7).

Proton conductivity increases with increasing temperature and indicates high conductivity values of 9.6 · 10 ^~4 S/cm, 1.6-10 ^"2 S/cm and 2.9 ·10 ^"2 S/cm at 70°C for RH 80%, 50% and 15%, respectively. The Arrhenius plots of doped Al-HPB-NET 1:1 are almost linear in the experimental temperature range. It indicates that one dominant proton conducting mechanism, with constant activation energy, is present in the material.

However, ion conduction in solid electrolytes is affected by complex interactions of different parameters such as

temperature-dependent equilibrium of charge carriers present at various RH, changes in viscosity etc., which were not considered altogether in the performed measurements.

Due to these discrepancies to the concept of activation energy, the calculated E _a values should be more correctly indicated as "apparent E _a " . The values of apparent E _a for doped Al-HPB-NET 1:1 at RH 15%, 50% and 80% were estimated to be 38 kJ mol ^"1 , 12 kJ mol ^"1 and 10 kJ mol ^"1 , respectively. As a comparison, the E _a for Nafion ^® 117 has been calculated to 27 kJ mol ^"1 , 17 kJ mol ^"1 and 13 kJ mol ^"1 for RH 15%, 50% and 80%. It can be observed that the magnitude of E _a strongly decreases with increasing RH, because water molecules as mobile species are added. Doped Al-HPB-NET 1:1 shows significantly low E _a values under mediated and fully immersed conditions, even lower than those obtained for Nafion ^® 117.

From these impedance spectroscopy investigations it can be concluded that doped Al-HPB-NET 1:1 features a very high conductivity (3.6-10 ^"2 S/cm) under 1 bar H ₂ 0 atmosphere with a smooth decrease below RH 30% at 55°C which indicates that the mechanism for proton transport may be different from the vehicle mechanism. It furthermore exhibits an activation energy of about 40 kJ mol ^"1 at RH 15% that starts to diminish down to 10 kJ mol ^"1 at RH 80% being even smaller that the E _a value obtained for Nafion ^® 117. Summarizing, the claimed porous aluminum phosphonate networks can be easily prepared and optimized due to the broad

availability of the required building blocks. In contrast to state-of-the-art polymers, doped aluminum phosphonates provide high and furthermore temperature-independent proton

conductivity, thus satisfying one of the prerequisites for new separator materials in FC systems. With an super proton conductivity value of about 5 -10 ^"2 S/cm at 120°C and RH 50%, doped Al-HPB-NET 1:1 are matching the guideline of close to 1 -10 ^"1 S/cm for conductivity of a membrane established by the U.S. Department of Energy as target operating conditions for automotive applications.

The following Table 6 presents a compilation of the most relevant data and experimental results for a variety of proton-conducting materials of the invention.

Table 6

General information . Elemental analysis (EA) was carried out on a Foss Heraeus Vario EL. Thermal gravimetric analysis (TGA) data were acquired with Mettler instrument (TGA/SDTA 851e) at a heating rate of 10 K rnin ^"1 under nitrogen atmosphere.

Infrared spectroscopy was measured on a Nicolet 730 FT-IR spectrometer in the evanescence field of a diamond. The sample was deposited as pristine material on the diamond crystal and was pressed on it with a stamp. 64 measurements were recorded for each sample, the background was substracted.

Powder X-ray diffraction measurements were performed on usin a Siemens D 500 Kristalloflex diffractometer with a graphite monochromatized CuKa X-ray beam, emitted from a rotating

Rigaku RU-300 anode.

Scanning electron microscopy (SEM) measurements were performed on a LEO 1530 field emission scanning electron microscope.

Nitrogen sorption experiments and micropore analysis were conducted at -195.8 °C using an Autosorb-1 from Quantachrome Instruments. Before sorption measurements, the samples were degassed in vacuum overnight at 150 °C. NLDFT pore-size

distributions were determined using the carbon/slit- cylindrical pore model of the Quadrawin software. Pore

volumes at p/po = 0.1 and p/po = 0.8 were converted into the corresponding liquid volumes using a nitrogen density of

1.25 -10 ^"3 g (cm ³ ) ^"1 (gaseous) and 8.10 -10 ^"1 g (cm ³ ) ^"1 (liquid).

Through-plane proton conductivity was measured by impedance spectroscopy in a two-electrode geometry using an SI 1260 impedance/gain-phase analyzer. A 100 mg uniaxially pressed pellet of 5.5 mm diameter and 2.1 mm thickness was used.

Conductivity measurements in pure water vapor (p(H ₂ 0) = 10 ⁵ Pa) above 100°C were carried out in a temperature-controlled glass chamber with a water inlet and outlet, both heated by a small furnace. To constantly flush the sample with pure H ₂ 0

atmosphere, water was evaporated, the gas subsequently

adjusted to the desired temperature and piped through the heated inlet of the glass oven. A pressure of 10 ⁵ Pa adapts itself due to the small outlet of the oven against ambient.

It should be noted that the relative humidity (RH) , set by a H ₂ 0 atmosphere at 10 ⁵ Pa, decreases with increasing temperature according to the saturation vapor pressure. 120 °C corresponds to a RH of ~ 50%, at 150°C the RH is close to 20%. For

measurements below 90°C temperature and RH were adjusted using a Binder KBF240 climate chamber. Nafion ^® 117 membrane was pre-treated by boiling in deionized water for 1 h, boiling in 3 % H ₂ 0 ₂ for 1 h, rinsing in water repeatedly, boiling in 0.5 M H ₂ S0 ₄ for 1 h and rinsing again in water. The membrane was stored in deionized water.

Potentiometric titrations were conducted with a Metrohm

Titranda 836 at 25°C. The 10 mg compound was allowed to agitate during 24 h in 10 ml aqueous solution and was titrated trice against a 0.01 M NaOH solution. The average consumption of NaOH was taken to calculate the equivalent amount of absorbed H ₃ P0 ₃ in the sample.