METHOD FOR PREPARING MACROSCOPIC FIBRES OF TiO2 BY CONTINUOUS ONE-WAY EXTRUSION, FIBRES OBTAINED AND USES

Abstract

The invention relates to a method for manufacturing macroscopic fibres of titanium dioxide (TiO.sub.2) by continuous extrusion in a one-way flow, to the macroscopic fibres of TiO.sub.2 that can be obtained by such a method, to the use of said fibres in heterogeneous photocatalysis for decontamination of organic pollutants from gaseous environments, and to a method for decontaminating gaseous environments, in particular air, using such fibres.

Claims

1. A process for preparing a titanium dioxide macroscopic fiber continuously comprising the following steps: i) the preparation of a dispersion of titanium dioxide nanoparticles in a solution of at least one vinyl alcohol polymer dissolved in a solvent, ii) the continuous and unidirectional extrusion of the dispersion obtained above in the preceding step in a coagulation bath suitable for giving rise to the solidification of said polymer, said extrusion being carried out by means of a, or a set of, cylindrical injection needle(s) having a diameter between 250 and 350 m, in order to form a pre-fiber made of a composite material comprising the titanium dioxide nanoparticles and the solidified vinyl alcohol polymer, iii) the continuous extraction of the pre-fiber formed above in step ii) out of the coagulation bath, said extraction being carried out coaxially relative to the axis of extrusion of the dispersion in said coagulation bath, iv) the continuous washing of the pre-fiber extracted from the coagulation bath, v) the continuous drying of the pre-fiber from the preceding step in order to obtain a dry composite material fiber, vi) the elimination of the vinyl alcohol polymer by calcination of the dry composite material pre-fiber from the preceding step, in order to obtain a titanium dioxide macroscopic fiber.

2. The process as claimed in claim 1, wherein the titanium dioxide nanoparticles used during step i) are spherical nanoparticles, the mean diameter of which varies from 2 to 15 nm.

3. The process as claimed in claim 1, wherein the amount of titanium dioxide nanoparticles within the dispersion prepared in step i) varies from 3% to 7% by weight relative to the total weight of the dispersion.

4. The process as claimed in claim 1, wherein the vinyl alcohol polymer used during step i) is a vinyl alcohol homopolymer or copolymer, the molecular weight of which varies from 5000 to 300 000 g.mol.sup.1, g.mol.sup.1, and has a degree of hydrolysis of greater than 95%.

5. The process as claimed in claim 1, wherein the amount of vinyl alcohol polymer within the dispersion prepared in step 1) varies from 2% to 11% by weight relative to the total weight of the dispersion.

6. The process as claimed in claim 1, wherein step i) comprises the following substeps: i1) the preparation of an aqueous dispersion of titanium dioxide nanoparticles, i2) the preparation of an aqueous dispersion of solid particles of a vinyl alcohol polymer, i3) the mixing of the dispersions obtained above in steps i1) and i2), in order to obtain a dispersion of titanium dioxide nanoparticles in a vinyl alcohol polymer solution.

7. The process as claimed in claim 1, wherein the dispersion of titanium dioxide nanoparticles prepared in step i) also contains spherical particles of a polymer material selected from nanoparticles of polystyrene, polymethacrylate, polyethyl methacrylate, polybutadiene and poly(styrene-divinylbenzene).

8. The process as claimed in claim 1, wherein the coagulation bath is a saturated aqueous solution of sodium sulfate.

9. The process as claimed in claim 1, wherein the injection rate of the dispersion into the coagulation bath varies from 1 to 1.9 m/min.

10. The process as claimed in claim 1, wherein the duration of the washing step iv) varies from 1 to 3 minutes.

11. The process as claimed in claim 1, wherein the step of drying the composite material pre-fibers during step v) is carried out by exposing said pre-fibers to a temperature varying from 65 C. to 90 C. for a duration of 1 to 3 minutes.

12. The process as claimed in claim 1, it wherein said process also comprises, between the drying step v) and step vi), a step of shaping the pre-fibers, in order to obtain an assembly of shaped pre-fibers.

13. The process as claimed in claim 1, wherein step vi) is carried out at a temperature varying from 350 C. to 500 C.

14. The process as claimed in claim 1, it wherein said process also comprises a step of winding and of hot drawing the composite material pre-fibers, said drawing step being carried out between the drying step and the winding step.

15. The process as claimed in claim 14, wherein said process is carried out on a production line comprising, in unidirectional alignment and in this order, a station for injecting the dispersion of nanoparticles into a coagulation bath, a washing station, a drying station and a winding station comprising a reel connected to a means that makes it possible to rotate the reel at constant speed, one or more intermediate rolls being able to be placed between the washing station and the drying station in order to support and convey the fibers to the winding station, and a calcining station.

16. Titanium dioxide macroscopic fibers obtained by the process of claim 1, said fiber comprising 85% to 90% of titanium dioxide in crystalline anatase form and of 10% to 15% of brookite, wherein said fiber has: two mean dimensions orthogonal to their longitudinal axis, the first mean dimension varying from 30 to 60 m and the second mean dimension varying from 10 to 40 m approximately; a macroscopic surface topology composed of a plurality of striations having a longitudinal axis parallel to the main axis of said fibers; and a structure comprising micropores and mesopores and the mean dimensions of which, as regards the mesopores, vary from 2 to 40 nm.

17. The fibers as claimed in claim 16, wherein said fiber has a specific surface area varying from 100 to 150 cm.sup.2/g.

18. A photocatalyst comprising: titanium dioxide macroscopic fibers as defined in claim 16.

19. The photocatalyst as claimed in claim 18, said photocatalyst configured to catalyzing the degradation of volatile organic compounds in a gaseous medium under the influence of UV irradiation.

20. The fibers as claimed in claim 16, wherein said titanium dioxide macroscopic fibers are configured to be implemented in a process for decontaminating a gaseous medium capable of containing one or more volatile organic compounds, where said gaseous medium comes into contact with titanium dioxide macroscopic fibers under light irradiation at a wavelength centered around 365 nm.

Description

EXAMPLES

[0074] The raw materials used in the following examples are listed below: [0075] titanium tetraisopropoxide at 98% (Ti(Oipr).sub.4), from Aldrich, [0076] hydrochloric acid at 37%, from Aldrich, [0077] nonylphenol ethoxylate containing 10 moles of ethylene oxide (nonionic surfactant), sold under the trade name Tergitol NP-10 by Aldrich; [0078] ammonia (NH.sub.4OH); sodium sulfate (Na.sub.2SO.sub.4), from Aldrich, [0079] hydrolyzed polyvinyl alcohol (PVA), having a molar mass of 195 000, 99% pure, from Fluka.

[0080] The other chemicals and solvents used in the examples were all of analytical grade or HPLC grade.

[0081] These raw materials were used as received from the manufacturers, with no additional purification.

Characterizations

[0082] The TiO.sub.2 macroscopic fibers prepared in the following examples were characterized by various analytical methods in order to demonstrate their mesoporous nature, and their microstructures (crystallinities):

[0083] Scanning electron microscopy (SEM) observations were carried out using a scanning electron microscope sold under the reference 6700F by the company JEOL, operating at 10 kV or at 5 kV.

[0084] Specific surface area measurements and characterizations on the mesoscopic scale were carried out by nitrogen adsorption-desorption techniques using a machine sold under the name Micromeritics ASAP 2010 after degassing of the TiO.sub.2 fibers at 150 C. under vacuum for 12 hours. The roughness of the TiO.sub.2 fibers was evaluated by determination of the fractal surface dimension (Ds), which can be deduced from the curves of the nitrogen adsorption isotherms. Ds was calculated according to the procedure described by Avnir, D. and Jaroniec, M. (An Isotherm Equation for Adsorption on Fractal Surfaces of Heterogeneous Porous Materials., Langmuir, 1989, 5, 1431-1433). The experimental data of the adsorption isotherms are plotted on a graph using equation (1) below:

=K[log(P.sub.0/P]v (1)

in which: [0085] v=3-Ds, [0086] is the relative adsorption calculated from the normalized curve with the greatest adsorption value, [0087] K is a constant, which represents the ordinate at the origin of the function Log =f(Log[log(P.sub.0/P].sup.v); K=0.166, [0088] P.sub.0 is atmospheric pressure (1 bar), [0089] P is the pressure at a time t; it is lower than the atmospheric pressure, [0090] Ds is the fractal surface dimension that it is desired to determine.

[0091] Ds may then easily be obtained by converting equation (1) to equation (2) below:

Log()=log(K)v log(log[P.sub.0/P]) (2)

[0092] Ds is deduced from the slope of the curve and may vary between 2 (non-rough surface) and 3. Thus, any value of Ds greater than 2 denotes an increase in the surface roughness. The adsorption range to be used for this analysis should be within the report range of the partial pressures Po/P ranging from 0.05 to 0.3 (Avnir, D. and Jaroniec, M., cited above).

[0093] The crystalline structure of the TiO.sub.2 was characterized by wide-angle X-ray diffraction via a diffractometer sold under the trade name PANalytical X'pert MPD in Bragg-Brentano - geometry, equipped with a graphite monochromator and an Anton-Paar HTK16 chamber (irradiation at the Cu K1 line, Cu, equal to 1.5418 (40 kV, 50 mA)).

Example 1

Preparation and Characterization of TiO.SUB.2 .Macroscopic Fibers According to the Process in Accordance with the Invention

[0094] In this example, TiO.sub.2 macroscopic fibers were prepared according to the continuous unidirectional extrusion process that is the subject of the invention.

1.1. Synthesis of the TiO.SUB.2 .Nanoparticles

[0095] 5 mL of a 37 vol % (i.e. 12 M) solution of HCl were added to 50 mL of a 15 wt % aqueous solution of Tergitol NP-10. The pH of the resulting solution was then adjusted to 1.8 by addition of NH.sub.4OH. 6.4 mL of Ti(Oipr)4 were then added slowly to said solution while stirring, the resulting mixture was left at room temperature. At the end of 12 hours approximately, a white precipitate of TiO.sub.2 nanoparticles was formed and was recovered by evaporation of the liquid phase of the mixture. The powder of TiO.sub.2 nanoparticles thus recovered was washed several times with deionized water so as to eliminate any trace of the nonionic surfactant.

1.2. Formulation of the Dispersions

[0096] An 8 wt % aqueous dispersion was prepared using the TiO.sub.2 nanoparticles prepared above in step 1.1.

[0097] 10 g of this aqueous dispersion were mixed with 10 g of a 12 wt % aqueous dispersion of PVA at a temperature of 80 C. in order to melt the PVA and obtain a PVA solution containing the TiO.sub.2 nanoparticles.

1.3. Generation of the TiO.SUB.2 .Macroscopic Fibers

[0098] A photograph of the various parts of the device used for carrying out the continuous unidirectional extrusion of the TiO.sub.2 macroscopic fibers is given by the appended FIG. 1. The device is composed of: [0099] a syringe pump (FIG. 1a) attached to which is a syringe with a 300 m diameter needle, filled with the solution of TiO.sub.2 nanoparticles prepared above in step 1.2.; [0100] a coagulation bath filled with a saturated aqueous solution of Na.sub.2SO.sub.4, (i.e. around 320 g.L.sup.1 of Na.sub.2SO.sub.4) maintained at 40 C. (FIG. 1b, 1c); [0101] a washing bath filled with a 0.1 mol.L.sup.1 aqueous solution of sodium tetraborate (FIG. 1d); [0102] a drying element comprising two infrared lamps facing each other and arranged on each side of a conveyor belt having a Teflon coating (FIG. 1e), [0103] a rotating take up roll that makes it possible to continuously wind the fibers leaving the drying element (FIG. 1f).

[0104] The syringe pump was used to inject the solution of TiO.sub.2 nanoparticles into the coagulation bath at a rate of 6 mL/h, corresponding to a calculated linear injection rate of 1.4 m/min. The solution of TiO.sub.2 nanoparticles coagulated immediately on leaving the syringe, as soon as it is injected into the coagulation bath, in the form of a continuous filament of a composite material formed of TiO.sub.2 nanoparticles trapped in a solid PVA matrix (TiO.sub.2-PVA). The minimum contact time of the TiO.sub.2-PVA filaments with the Na.sub.2SO.sub.4 solution in the coagulation bath was set at 30 seconds in order to increase their mechanical stability.

[0105] The fibers were then passed through the washing bath in order to eliminate the Na.sub.2SO.sub.4 residues that may be present in the fibers after the coagulation. As the PVA is soluble in water, sodium tetraborate was added to the washing bath, the presence of which makes it possible to prevent the fibers from degrading and breaking during the washing step. The contact time of the TiO.sub.2-PVA fibers with the washing bath was set at one minute approximately.

[0106] On leaving the washing bath, the TiO.sub.2-PVA fibers were conveyed by the conveyor belt and dried at a temperature of 75 C. approximately by passing between the infrared lamps located on either side of the conveyor belt.

[0107] At the end of the drying zone, the fibers were stretched and raised from the conveyor belt and wound using the rotary take up roll set at one, the minimum speed of rotation of which is 3.2 m.min.sup.1. Insofar as the rate of injection leaving the syringe was 1.4 m.min.sup.1, it is possible to deduce therefrom that the corresponding minimum stretch factor during the whole of the process for manufacturing the TiO.sub.2 macroscopic fibers was 2.3.

[0108] According to this process, it was possible to continuously manufacture several hundreds of meters of fibers of a TiO.sub.2-PVA composite material.

[0109] A second washing of the TiO.sub.2-PVA fibers was then carried out by unwinding the fibers and passing them into a water bath, followed by a second drying under the same conditions as before, prior to them again being wound onto the rotary take up roll. The purpose of this second session of washing/drying/winding of the fibers was to eliminate any residue of Na.sub.2SO.sub.4 and sodium tetraborate in the TiO.sub.2-PVA composite material.

[0110] Finally, the TiO.sub.2-PVA fibers were then calcined for 6 hours at 450 C., by applying a temperature rise rate of 5 C./min in order to eliminate the PVA and induce the crystallization of the TiO.sub.2 in anatase form. TiO.sub.2 macroscopic fibers were thus obtained that were then characterized.

1.4. Characterisations of the TiO.SUB.2 .Macroscopic Fibers

[0111] The electron microscopy results of the fibers after calcination are given by the appended FIG. 2 at various magnifications (FIGS. 2a: 3000; 2b: 25000 and 2c: 2000).

[0112] Generally, the fibers have a flattened shape, with a width ranging from 30 to 60 m, and more particularly from 40 to 47 m, and a thickness ranging from 10 to 40 m.

[0113] Their topology on the macroscopic scale is clearly different from those obtained by rotary extrusion according to the process previously described by Kinadjian N. et al. (cited above). Indeed, whereas the fibers obtained by rotary extrusion had a surface topology composed of honeycomb surface roughness (juxtaposition of pores), the TiO.sub.2 fibers obtained by coaxial extrusion according to the process in accordance with the invention have a tree trunk topology, the fibers having a plurality of wide and deep, longitudinal surface striations oriented parallel to the main axis of the fibers. As the fibers are stretched, these striations result from the alignment of the PVA chains induced by the shear rate during the extrusion process.

[0114] Nitrogen adsorption/desorption measurements were used to characterize the porosity of the fibers and also their roughness on the mesoscopic scale. The corresponding results are given by appended FIGS. 3 and 4.

[0115] FIG. 3 represents the amount of nitrogen adsorbed (in cm.sup.3/g) as a function of the relative pressure (P/P.sub.0), the solid circles corresponding to the desorption curves and the solid squares to the adsorption curves.

[0116] FIG. 4 represents the dV/dlog(D) ratio (in cm.sup.3/g.) as a function of the pore diameter (in nm). In the dV/dlog(D) ratio, V is the volume of nitrogen adsorbed at the interfaces and D represents the pore diameter.

[0117] These results show that the TiO.sub.2 fibers obtained by the process in accordance with the invention have an adsorption profile representative of a type IV isotherm, and a distribution of the pore diameter in the mesopore range (between 2 and 37 nm approximately). The presence of the adsorption points at P/P.sub.0<0.05 indicates that the fibers are also microporous.

[0118] The specific surface areas, calculated according to the B.E.T. and B.J.H. methods, and also the total pore volume and the surface roughness (Ds) values are presented in table 1 below:

TABLE-US-00001 TABLE 1 B.E.T. specific B.J.H. specific surface area surface area Total pore volume Surface roughness (m.sup.2 .Math. g.sup.1) (m.sup.2 .Math. g.sup.1) (mL .Math. g.sup.1) (Ds) 136 166 0.31 2.56

[0119] The pore size distribution, calculated according to the Barret-Joyner-Halenda (B.J.H.) equation from the desorption curves, indicates a pore size distribution between 6 and 18 nm.

[0120] Furthermore, the Ds value obtained (2.56) shows that the preparation process in accordance with the present invention makes it possible to achieve TiO.sub.2 fibers that have a large surface roughness on the mesoscopic scale.

[0121] The structure of the fibers on the microscopic scale was studied by X-ray diffraction. The corresponding results are represented in the appended FIG. 5 on which the intensity (in arbitrary units) is a function of the 2 angle (in degrees). These results indicate that the fibers are mainly composed of TiO.sub.2 in the anatase phase (85-90%), with a small amount of brookite (10-15%). The amount of brookite present in the TiO.sub.2 fibers obtained according to the process in accordance with the invention is greater than that present in the fibers obtained according to the rotary extrusion preparation process (Kinadjian N. et at., cited above). The X-ray diffraction results furthermore confirm that the washing operations carried out have made it possible to eliminate all traces of the salts used during the manufacture (Na.sub.2SO.sub.4 and sodium tetraborate).

Example 2

Study of the Photocatalytic Properties of the TiO.SUB.2 .Macroscopic Fibers Obtained According to the Preparation Process in Accordance with the Invention

[0122] In this example, the photocatalytic properties of the TiO.sub.2 macroscopic fibers as prepared above in example 1 were studied.

2.1. Test Principle

[0123] In this example, the photocatalytic properties of the TiO.sub.2 macroscopic fibers as prepared above in example 1 according to the process in accordance with the invention of coaxial extrusion (which are hereinafter referred to as F-TiO.sub.2-extr.CoAx) were studied, in comparison, on the one hand, to those of a reference material, known for its excellent photocatalytic properties and consisting of anatase TiO.sub.2 nanoparticles, supported by quartz fibers and sold under the trade name Quartzel PCO by the company Saint Gobain (which are hereinafter referred to as F-TiO.sub.2-Quartzel), and, on other hand, to those of TiO.sub.2 macroscopic fibers obtained according to the rotary extrusion process as described by Kinadjian N. et al. (cited above), (which are hereinafter referred to as F-TiO.sub.2-Extr-ROT).

[0124] The reaction used for the study of the photocatalytic properties of the various materials tested in this example is the mineralization of acetone, according to the following equation:

##STR00001##

2.2. Experimental Device

[0125] The photocatalytic properties of the materials were studied on the device depicted in the appended FIG. 6 comprising in particular a cylindrical photoreactor 1 integrated into a circuit made of polytetrafluoroethylene (PTFE).

[0126] This device, represented in its entirety in FIG. 6a) is composed of a 17-liter thermoregulated gas (air) reservoir 2 connected to a circulation pump 3, the flow rate of which is controlled by a flowmeter 4. In order to limit the heating of the gas, thermoregulated columns 6,7 are positioned upstream and downstream of the circulation pump 3. The gas reservoir comprises an air inlet 9 and an air outlet 10 and is equipped with a manometer 11 and with a probe 12 that measures the temperature and the hygrometry in order to record the temperature and relative humidity of the gas circulating in the circuit throughout the duration of the experiment. The thermoregulation of the device was carried out by a water bath 5. The gas was sampled automatically and periodically through a pump (flow rate 10 L.min.sup.1) with the aid of a gas chromatograph 13 sold under the reference Varian 3800 GC with a flame ionization detector for the analysis of the volatile organic compounds (VOCs) and equipped with a katharometer detector for the analysis of the CO.sub.2. The cylindrical photoreactor 1 is presented in detail in FIG. 6b and is composed of a thermoregulated pyrex column 1a comprising a water inlet 1b and a water outlet 1c and containing a Pyrex capillary 1d having an internal diameter of 10 mm (known as a canister) containing the fibers to be tested for their photocatalytic properties. The flow rate of the gas in the circuit was around 0.7 m.s.sup.1 when the flowmeter 4 was set at 3.3 L.min.sup.1. Four fluorescent tubes 8 (Philips TLD8W) emitting at 366 nm are attached above the photoreactor 1, in order to deliver an irradiance of 3.2 mW/cm.sup.2 as UV-A rays inside the reactor. The irradiance actually received by the fibers to be tested placed in the capillary 1b of the photoreactor 1 was less than 3.2 mW/cm.sup.2 due to the absorbance of the Pyrex walls of the thermoregulated column 1a and of the capillary 1b.

[0127] The photocatalytic properties of the TiO.sub.2 fibers prepared according to example 1 above and those of the commercial Quartzel PCO fibers were evaluated under strictly comparative conditions, the same procedure being followed in all the experiments.

[0128] Once the material to be tested was inserted in the photoreactor 1 (5 and 40 mg of fibers in each experiment), synthetic air was injected into the circuit and the relative humidity was adjusted to 15% by injecting the necessary amount of water into the circuit. The flow rate of the circulation pump 3 was set at 3.3 L.min.sup.1, corresponding to a gas flow rate in the circuit of 0.7 m.s.sup.1, which corresponds to the optimal conditions for carrying out the photocatalytic conversion of the acetone. The temperature of the circulating gas being 22 C.2 C. The fibers are firstly irradiated for 20 hours under these conditions, without acetone in order to activate them and to evaluate the possible change in the amount of VOCs or CO.sub.2 adsorbed. The chromatograph did not record any adsorption peak during this activation period. The lamps 8 were then switched off.

[0129] 45 ppmv (2 l) of acetone were then injected into the gas reservoir 2. The adsorption of acetone by the fibers to be tested was then monitored by chromatography using the gas chromatograph 13. When the concentration of gaseous acetone had stabilized, the lamps were switched on (t=0). The concentrations of acetone and CO.sub.2 were determined and recorded every 7 minutes. Possible leaks of CO.sub.2 and acetone were furthermore determined by operating the device without irradiation (less than 10% loss over 10 hours) and subtracted from the results obtained in the analysis phase of the fibers to be tested.

2.3. Results

[0130] The results obtained are presented in table 2 below, the photocatalytic efficiency of the fibers being determined by measuring the first-order reaction kinetics for the degradation of acetone and expressed in min.sup.1.g.sup.1:

TABLE-US-00002 TABLE 2 FTiO.sub.2- FTiO.sub.2- FTiO.sub.2-extr. Quartzel Extr-ROT Fibers tested CoAx (*) (*) Diameter (m) 40 9 130 Macroscopic tree trunk TiO.sub.2 Juxtaposition surface striations nanoparticles of pores topology deposited on quartz fibers Photocatalytic 0.12 min.sup.1 .Math. g.sup.1 0.15 min.sup.1 .Math. g.sup.1 0.037 min.sup.1 .Math. g.sup.1 efficiency (*) Fibers not in accordance with the invention

[0131] These results show that the TiO.sub.2 macroscopic fibers obtained according to the coaxial extrusion preparation process in accordance with the invention (F-TiO.sub.2-extr.CoAx) have a mean efficiency which is around 3 times higher than that of the TiO.sub.2 macroscopic fibers obtained according to the rotary extrusion preparation process of the prior art (F-TiO.sub.2-Extr-ROT), this efficiency being of the same order of magnitude as those of the F-TiO.sub.2-Quartzel commercial fibers, which makes them competitive from an industrial and commercial viewpoint.

[0132] As is also indicated in table 2 above, the TiO.sub.2 macroscopic fibers obtained according to the coaxial extrusion preparation process in accordance with the invention are essentially distinguished from the TiO.sub.2 macroscopic fibers obtained according to the coaxial extrusion preparation process described in the document by Kinadjian N. et al. (cited above), by the fact that they have a diameter that is around 3 times smaller, of the order of 40 m approximately (instead of 130 m approximately) and a macroscopic surface topology that comprises longitudinal striations parallel to the longitudinal axis of the fibers and referred to as tree trunk topology (see appended FIG. 2).

[0133] It is therefore possible to conclude that the improvement in the catalytic properties of the TiO.sub.2 macroscopic fibers obtained according to the process in accordance with the invention results from the modification of these characteristics (diameter and macroscopic surface topology).

Example 3

Photocatalytic Degradation of Toluene

[0134] In this example, the photocatalytic properties of the TiO.sub.2 macroscopic fibers as prepared above in example 1 were studied with respect to the mineralization of toluene. By way of comparison, the F-TiO.sub.2-Quartzel commercial fibers were also tested.

3.1. Test Principle

[0135] The reaction used in this example is the mineralization of toluene, according to the following equation:

##STR00002##

3.2. Experimental Device

[0136] The experiment was carried out on the same device as that used above in example 2 and under the same conditions, apart from the differences explained in detail below: [0137] the TiO.sub.2 fibers prepared according to example 1 above (80 mg) and the Quartzel PCO commercial fibers (80 mg) were entangled in the form of a mat having a diameter of 72 mm and a thickness of 10 mm. It is these mats of fibers, respectively referred to as F-TiO.sub.2-extr.CoAx-Mat and F-TiO.sub.2-Quartzel-Mat that were then placed in the photoreactor; [0138] cylindrical photoreactor having a diameter of 80 mm approximately and a height of around 200 mm, equipped with a Teflon cover and two Swagelok fittings. Inside this photoreactor, a porous glass disk that is horizontal, i.e. placed transversely to the vertical axis of the photoreactor, supports the mat of fibers to be tested; [0139] irradiation device composed of 4 light-emitting diodes (LEDs) (H2A1-H365-E 350 mA Roithner Lasertechnik) with a maximum at 350 nm, delivering an irradiance that may vary between 1 and 10 mW/cm.sup.2 as UV-A rays inside the photoreactor; [0140] relative humidity adjusted to 20% by injecting the necessary amount of water into the circuit; [0141] injection of 100 ppmv of toluene into the gas reservoir 2 (i.e. 13 l).

3.3. Results

[0142] The results obtained are presented in table 3 below, the photocatalytic efficiency of the fibers being determined by measuring the first-order reaction kinetics for the degradation of toluene and expressed in ppmv.min.sup.1:

TABLE-US-00003 TABLE 3 Fibers tested FTiO.sub.2-extr. CoAx-Mat FTiO.sub.2-Quartzel-Mat (*) Photocatalytic 0.0011 ppmv .Math. min.sup.1 0.0004 ppmv .Math. min.sup.1 efficiency (*) Fibers not in accordance with the invention

[0143] These results show that the kinetics for the degradation of toluene are faster with the mat of entangled TiO.sub.2 fibers in accordance with the invention than with the mat of entangled Quartzel PCO commercial fibers not in accordance with the invention. Furthermore, around 30% only of the amount of toluene injected was mineralized after 500 minutes of irradiation with the mat of entangled Quartzel PCO commercial fibers that is not part of the invention whereas with the mat of entangled TiO.sub.2 fibers prepared according to example 1 and therefore in accordance with the invention, this percentage was 47%. Furthermore, the subsequent conversion of the toluene was virtually zero between 500 and 1200 minutes with the mat of entangled Quartzel PCO commercial fibers not in accordance with the invention, whereas it reached 94% with the mat of entangled TiO.sub.2 fibers prepared according to example 1.

[0144] Furthermore, observation of the fibers after the photocatalytic reaction (not represented) showed that the Quartzel PCO commercial fibers had a yellow coloration and that black lumps were present between the entangled fibers. An intensive purification of 48 hours under irradiation in the reactor without pollutant (20% RH, clean air) was necessary in order to restore the initial color and activity of the mat of fibers. A contrario, the mat of fibers in accordance with the invention had, at the end of the toluene mineralization experiment, no yellow coloration and no black lumps, and led to the complete degradation of the toluene in 1500 min, without any sign of deactivation (continuous increase in the amount of CO.sub.2 released).