INTRODUCTION
Ceramics are often prepared with surface layers of different composition from the bulk,[1'2] to impart a specific functionality to the surface or to act as a protective layer for the bulk material.[3,4] Here I describe a general process by which functional surface layers with a nanometer-scale compositional gradient can be readily formed during the production of bulk ceramic components. The first concept regarding this process was established by Ishikawa et al. in 1998.[5] The basis of this approach is to incorporate selected low-molecular-mass additives into either the precursor polymer from which the ceramic forms, or the binder polymer used to prepare bulk components from ceramic powders. Thermal treatment of the resulting bodies leads to controlled phase separation (”bleed out”) of the additives, analogous to the normally undesirable outward loss of low-molecular-mass components from some plastics;[6-10] subsequent calcination stabilizes the compositionally changed surface region, generating a functional surface layer. This approach is applicable to a wide range of materials and morphologies, and should find use in catalysts, composites, and environmental barrier coatings.
PREVIOUS METHODS TO OBTAIN SURFACE GRADIENT STRUCTURES
To avoid the concentration of thermomechanical stress at the interface between the surface layer and the bulk material, many materials were developed that had gradually varying properties as the distance into the material increases.[11] Such materials can contain gradients in morphology or in composition. For example, gradients in morphology can result in materials that have a graded distribution of pore sizes on a monolith of silica aerogel, and a type of integral plastic. These materials were created by strictly controlling the vaporization of the volatile during the production process.[12,13] Gradients in chemical composition have been achieved, for example: 1) chemical vapour deposition;[14,15] 2) powder methods such as slip cast or dry processing;[16] 3) various coating methods;1-17-1 and 4) thermal chemical reaction.[2,18] Of these, 1) and 4) are relatively expensive, complicated and result in damage to bulk substrates. Items 2) and 3) produce stepped gradient structures, and it is difficult to control the thickness of each layer to less than 100 nm. Furthermore, most of these processes are not easily adapted to coating samples in the form of fiber bundles, fine powders, or other materials with complicated shapes.
A NEW PROCESS FOR CREATING SURFACE FUNCTIONAL LAYERS
We have addressed the issue of establishing an inexpensive and widely applicable process for creating a material with a compositional gradient and excellent functionality. A schematic representative of our new in situ formation process for functional surface layers, which have a gradient-like structure toward the surface, is shown in Fig. 1. The important feature of this method is that the surface layer of the ceramic is not deposited on the substrate, but is formed during the production of the bulk ceramic. We confirmed that our process is applicable to any type of system as long as, in the green-body (that is, not calcined) state, the system contains a resin and a low-molecular-mass additive that can be converted into a functional ceramic at high temperatures. Here the resin is a type of precursor polymer (polycarbosilane, polycarbosilazane, polysilastyrene, methylchloropolysilane, and so on) or binder polymer used for preparing green bodies from ceramic powders.[19] Although the former case (using precursor polymers) is explained in detail in this paper, the latter case, using binder polymers, was also confirmed by treating a Si3N4 body with a TiN surface layer. Si3N4 can exhibit excellent thermal stability and wear resistance in the high-speed machining of cast iron, but shows poor chemical wear resistance in the machining of steel.[20] To avoid this problem, TiN coating, via expensive chemical vapor deposition, was often performed on previously prepared Si3N4 substrates. But if our new process is appropriately applied, formation of the TiN surface layer could be achieved during the sintering process of the Si3N4 green body. In this case, titanium(IV) butoxide and polystyrene are used as the low-molecular-mass additive and binder polymer, respectively. By a combination of sufficient maturation (in air at 100°C) and subsequent sintering (in NH3+H2+N2 at 1200°C), Si3N4 covered with TiN is successfully produced. This technology would be very useful for producing ceramic materials with complicated shapes and various coating layers. Moreover, our process is advantageous for preparing precursor ceramics (particularly fine particles, thin fibrous ceramics and films). The systems to which our concept is applicable are shown in Fig. 1.
Fig. 1 Schematic diagram of a general process for in situ formation of functional surface layers on ceramics.
Here I give a detailed account of the results for the precursor ceramic obtained by using polycarbosilane. Poly-rbosilane (-SiH(CH3)-CH2-)n is a representative prece-ramic polymer for preparing SiC ceramics—for example, Hi-Nicalon fiber[21] and Tyranno SA fiber.[22] Furthermore, oxide or nitride can also be produced from the polycarbosilane by firing in air or ammonia, respectively. Our new technology makes full use of the bleed-out phenomenon[6-10] of additives intentionally mixed in the polycarbosilane. Here we treated a polycarbosilane with Ti(OC4H9)4 or Zr(OC4H9)4, and created a strong, fibrous photocatalyst with a surface TiO2 layer, or a highly alkali-resistant SiC-based fiber with a surface ZrO2 layer.
Fig. 2 A new process for producing strong photocatalytic fiber with gradient structure.
Fig. 3 Surface gradient compositions of the photocatalytic
Fig. 4 Surface appearances and cross sections of the TiO2/ SiO2 photocatalytic fiber.
PHOTOCATALYTIC FIBER PRODUCED BY THIS NEW PROCESS
The preceramic polymer containing Ti(OC4H9)4 (50 wt.%) was shaped into a fiber by melt spinning. The fiber was then matured in air at 70°C for 100 h; during this process, the titanium compound oozed from the preceramic polymer. The bleed-out phenomenon was confirmed by using electron spectroscopy for chemical analysis (ESCA). After the maturation, the fiber material was cured in air at 200°C. The cured material was then calcined in air at 1200°C to obtain a silica fiber covered with titania. This production process is schematically shown in Fig. 2.
Fig. 5 Felt material of the photocatalytic fiber.
Fig. 3 shows Auger electron spectroscopy (AES) depth analysis of the surface layer of the fiber. The thickness of the surface layer can be controlled in the range 5-500 nm by changing the temperature (70-100°C) and time (10- 200 h) of the maturation.[23] Furthermore, the composition of the surface layer reflects the composition of the low-molecular-mass additives.
Fig. 6 Decomposition of acetaldehyde using the photo-catalytic fiber with UV irradiation.
Fig. 7 The results of extinction activity of coliform using the photocatalytic fiber with UV irradiation.
According to the results of X-ray diffraction analysis, these fibers were composed of anatase TiO2 and an amorphous SiO2 phase. Investigation by scanning electron microscopy (SEM) showed that the outer surface of the fiber prepared via maturation for 100 h appeared to be smooth (Fig. 4a) and was covered with titania particles (Fig. 4b). Furthermore, the transmission electron microscopy (TEM) image of the cross section near the surface showed a TiO2-sintered surface and particle-dispersed bulk structures (Fig. 4c). Most of the surface TiO2 crystals (crystalline size: 8 nm, Fig. 4d) were directly sintered (directly sintered structure), whereas the internal TiO2 crystals were bound with the amorphous SiO2 phase (liquid-phase sintered structure). This result corresponds to the gradient composition shown in the AES data. The gradient-like structure resulted in strong adhesion between the surface TiO2 layer and the bulk material, which is different from the behavior of other coating layers formed on substrates via conventional methods. Although in the case of the other TiO2-covered silica fiber, the TiO2 layer was easily peeled off, and our TiO2-sintered surface definitely did not drop off after heat-cycling, washing, or rubbing.
Fig. 8 The relationship between the photocatalytic activity and the size of a titania crystal.
Fig. 9 Circulation purifier for pollutants using photocatalytic fiber with a UV lamp.
Furthermore, a strengthening effect of the fine particles (<10 nm) dispersed in the bulk ceramics can be achieved at the same time by using our technology. Tensile testing of the monofilaments according to the ASTM D3379-75 standard with a 25-mm-gauge length demonstrated that the strength was markedly higher (>2.5 GPa) than that of ordinary sol-gel TiO2/SiO2 fibers (< 1 GPa).[24] At present, the reason for this is not clear, but it was assumed that this was caused by less stress concentration at the surface region because of the gradient structure toward the surface.
We subsequently confirmed the objective function (photocatalytic activity) of the above-mentioned fiber (titania fiber) as follows. We prepared a quartz tubular reactor (inner volume: 7 cm3) with an ultraviolet (UV) lamp. The tubular reactor was filled with 2 g of the felt material (Fig. 5) made of the titania fiber. The photocatalytic activity was confirmed by using air containing 140 ppm of acetaldehyde, at a flow rate of 1 L/min, and UV light at an intensity of 1-5 mW cm-2 (wavelength, 352 nm). Under these conditions (single pass), the acetaldehyde was effectively decomposed accompanied by the generation of CO2. The results are shown in Fig. 6.
It is well known that the catalytic effect of TiO2 is attributed to the generation of a strong oxidant, hydroxyl radicals. Following this theory, the quantum efficiency of the felt material prepared from the above titania fiber was calculated from a decrease in the acetaldehyde. In this case, if the number of molecules is significantly larger than the number of photon, acetaldehyde is oxidized to CH3COOH as follows:
Fig. 10 Purification test of the water of a circulation bath system using the circulation purifier.
In this case, the apparent quantum efficiency (QE) of the felt material prepared from the aforementioned titania fiber is calculated by the following equation:
The calculation result using these values showed an extremely high QE value even at room temperature (over 37%). It is believed that this higher value at high temperatures is attributable to the evaporation of the formed acetic acid adsorbed on the fiber surface. These excellent QE values could be realized by the dense existence of very fine anatase-TiO2 crystals (8 nm) on the surface. These fine crystals are considered to facilitate the diffusion of excited electrons and holes toward the surface before their recombination.
We also confirmed the coliform-sterilization ability of this fiber as follows. Our fiber (0.2 g) was placed in waste-water (20 mL) containing coliform at a concentration of 1 x 106 mL-1. Irradiation by UV light (wavelength: 352 nm, 2 mW cm- 2) was performed at room temperature, and a small amount of the wastewater was extracted. After cultivation using the extracted water, the amount of active coliform was calculated from the number of colonies formed. In this experiment, using the desirable fiber covered with very fine titania crystals (8 nm), all of the coli-form in the wastewater was completely sterilized within 5 hr, accompanied by the generation of CO2. In the comparative study, using undesirable fibers covered with large titania crystals (9-11 nm), sterilization of the coliform was markedly slow (Fig. 7). From the results, the size of the titania crystal is found to be closely related to the photo-catalytic activity. It is assumed that, in the case of large crystals, the recombination (inactivation) of the hole and excited electron generated by UV irradiation easily occurs (Fig. 8). To suppress the recombination and obtain the good photocatalytic activity, the creation of the smaller titania crystals is very important. The new process described in this paper is very desirable for controlling the size of fine crystals, because both the bleed-out phenomenon of the low-molecular-mass additive and the crystallization of the led functional material proceed competitively.
Fig. 11 The results of the purification test using the circulation purifier.
Fig. 12 Decomposition of dioxin using the circulation purifier.
INDUSTRIAL APPLICATION OF THE PHOTOCATALYTIC FIBER WITH A GRADIENT TITANIA LAYER
A circulation purifier for pollutants (Fig. 9) was developed by using the felt material made of the aforementioned photocatalytic fiber. This is a very simple purifier with a module composed of the cone-shaped felt material (made of the photocatalytic fiber) and UV lamp. Purification of the bath water of a circulation bath system was performed by using the above purifier (Fig. 10). Many bacteria (common bacterium, legionera germ, and coliform), which existed in the bath water before the purification, were perfectly decomposed into CO2 and H2O by using the above purifier (Fig. 11).
Furthermore, the photocatalytic fiber can be used for the purification of the many types of wastewater. Fig. 12 shows the result regarding a decomposition of dioxin contained in a wastewater. In this case, 95.1% of the dioxin was found to be decomposed after only 2 hr.
FORMATION OF THE ALKALI-RESISTANT SURFACE LAYER ON SiC
The next fiber was prepared from polycarbosilane containing Zr(OC4H9)4 by the same process as that used for the TiO2/SiO2 fiber material (the aforementioned photocatalytic fiber), except that the calcination was performed in Ar atmosphere at 1300°C. In this case, the polycarbosilane and Zr(OC4H9)4 were effectively converted into SiC-based bulk ceramic and zirconium oxide (cubic zirconia), respectively (X-ray diffraction results are given in Fig. 13). Before the conversion, bleed-out of the zirconium compound proceeded effectively. AES depth analysis of the fiber surface showed an increase in the concentration of zirconium toward the surface. This construction was confirmed by the TEM image of the cross section near the fiber surface (Fig. 14a and b). This indicates the direct production of an SiC-based fiber covered with a ZrO2 surface layer, which has a gradientlike composition toward the surface. In general, amorphous fibers covered with ceramic crystal do not show high strength.[25] However, this fiber showed relatively high strength (2.5 GPa) compared with other SiC fiber (2.1 GPa) coated with zirconia nanocrystals via the sol-gel method. The initial strength of the SiC fiber used for the comparative study was 3.1 GPa. The ZrO2 surface layer, a basic oxide material, can provide better alkali resistance for SiC ceramics.
Fig. 13 X-ray diffraction pattern of the SiC fiber covered with zirconia.
Fig. 14 Surface appearances and cross sections of the SiC fiber covered with ZrO2.
To confirm the better alkali resistance for our ZrO2-covered SiC fiber, the following experiment was performed. The fiber material was immersed for 15 min in deionized water saturated with potassium acetate and then annealed at 800°C for 100 hr in air after drying. Comparative studies were conducted by using the SiC-based fiber prepared from polycarbosilane, which did not contain zirconium (IV) butoxide, as well as commercial SiC fibers, namely, Hi-Nicalon and an alkali-resistant sintered SiC fiber[22] (Tyranno SA fiber). Fig. 15 shows the fractured surfaces of the tested fiber bundles, obtained by using field-emission scanning electron microscopy (FE-SEM). As can be seen from the micrographs, only the ZrO2-covered SiC fiber, which was obtained by using our method, retained its intact fibrous shape, whereas the other SiC fibers were extensively oxidized and then bonded together.
CONCLUSION
Fundamentally, this new process can be applied for preparing functional ceramics with gradient, nano-sized surface structures as long as in the green-body state, the system contains both a polymer component and a low-molecular-mass additive which can be converted into a functional ceramic by heat treatment at high temperatures. Namely, this process does not care about the shape of the precursor materials. Fine particles, thin fibrous ceramics and films (SiC, Si3N4, SiO2) covered with functional layers (BN, TiN, TiO2, ZrO2, Al2O3) could also be synthesized by sufficiently maturing and firing the precursor powders of polycarbosilane or polysilazane including the corresponding additives. The atmosphere and temperature during firing would need to be strictly controlled. This process is applicable to a wide range of materials and morphologies, and should find use in catalysts, composites, and environmental barrier coatings.
Fig. 15 Alkali resistance of the SiC fiber covered with ZrO2 with comparative results.
