WO2007042081A1 - Optical sensor device for local analysis of a combustion process in a combustor of a thermal power plant - Google Patents

Abstract

The invention concerns to an optical sensor device for local analysis of a combustion process in a combustor of a thermal power plant, in particular a gas turbine plant, wherein at least one wavelength selective optical element is exposed directly or indirectly to hot combustion gases being produced by said combustion process, said optical element provides an array of nano- and/or microcrystalline fibres which are created by use of shear flow crystallization.

Description

Optical sensor device for local analysis of a combustion process in a combustor of a thermal power plant

Technical Field

The invention relates to an optical sensor device for local analysis of a combustion process in a combustor of a thermal power plant, in particular a gas turbine plant.

An important aspect of operating high advanced thermal power plants, especially gas turbine plants concerns the quality of combustion and the generation of hot gases for powering turbine stages. It is a known fact that analyzing hot gases shortly after the combustion process and before entering the turbine stages valuable information about burner quality and emission values can be derived for online optimizing procedures concerning the combustion process.

Due to the very high temperature level inside a combustor the choice of available sensor systems withstanding such temperatures beyond 1000⁰C is very limited. No durable sensor system is known for present which is applicable reliable for measuring burner parameters in a very direct manner. Optical remote sensing systems are known which however have to be secured against the high temperatures inside the combustor. This requirement confines reliability and place of installation.

Summary of the invention

It is therefore an object of the invention to provide a sensor system for local analysis of a combustion process in a combustor of a thermal power plant, in particular a gas turbine plant which withstand the high temperature level inside a combustor and which yields credible burner information especially information of hot gas consistency and physical behavior like temperature and pressure. The sensor device shall be of simple construction and inexpensive fabrication

The inventive sensor is described in claim 1. Preferential features are disclosed in subclaims as well as in the following description.

The inventive sensor is an optical sensor device for local analysis of a combustion process in a combustor of a thermal power plant, in particular a gas turbine plant providing at least one wavelength selective optical element exposed directly or indirectly to hot combustion gases being produced by said combustion process. The optical element provides an array of nano- and/or microcrystalline fibers which are thermally and chemically resistance and created by use of shear flow crystallization. For measurement purpose a light source is provided emanating a light beam favorably of a broad bandwidth for passing through said optical element directed onto a mirror which is arranged oppositely to the light source in respect to said optical element having a mirror surface onto which said light beam is reflected at least partly so that at least a reflected light beam fraction passes said optical element in opposite direction. A detector is positioned at the same site of said optical element like the light source for detecting the reflected light.

Since the mirror surface and the optical element are arranged apart by distance and the space between said components is flooded with hot gas twice the distance between optical element and mirror surface therefore serves as absorption length for the propagating light beam in said space.

The optical element is confectioned by means of shear flow crystallization which is explained in the following broadly and providing the function of a wavelength selective filter like a notch or Bandwidth-filter. The wavelength selectivity has to be chosen in dependence of the absorption bands of atoms or molecules of interests which shall be measured along the before described absorption length. The nano- and/or microcrystalline fibers which are created as so called primary array in course of the shear flow crystallization process are made of high temperature resistant material like metal oxides, metal or other inorganic nanoparticles deposited onto an amorphous, polycrystalline or single crystalline flat or curvature support using shear-flow crystallization method under distinct crystallization conditions and post- crystallization thermal treatment, leading to closely packed and distinct crystallographically rational orientation of the crystal packing of microfibers relative to their external faces and leading to distinct orientation of the microfibers relative to crystallization cell geometry.

The starting suspensions of the shear-flow crystallization can be monodisperse or can have different particle size. The substrates surfaces can be flat or possess curvature. The pre-sintering process is important for formation of nanopattems of required dimensionality. The nano- or microarray thickness and shape can be controlled by crystallization cell design. Controlled heat treatment can be applied for controlling the coalescence of nanoparticles and for creating the crystallinity of the low-dimensional nano- or microarrays, providing gradient of properties across the micro- or nanoarrays, or across the article, built of micro- or nanoarray. The low- dimensional nano- or microarrays with grated properties can be fabricated. The possible applications include waveguides, monochromators, focusing devices, tunable transmitting filters, mono- and multilayered mirrors. One of the applications is use of patterned photonic microfiber arrays as pattern for design of array of aligned nanotube array with adsorbed molecular iodine or other material for use as part of filter system in filtered Rayleigh scattering (FRS) laser diagnostic or other sensors. Suggested fabrication method belongs to enabling technology for the inventive applications in gas turbine: broad-band and narrow-band filters for suppression of absorption from molecules in distinct spectrum regions (UV, visible, IR), selective band photonic defect-induced band-pass filters, mesoporous framework structures for FRS super-narrow notch filters, molecular super-narrow-band notch filters based on for example molecular iodine encapsulated inside of mesoporous framework solids. Particularly promising for local gas turbine diagnostics (temperature, pressure, NO_x pollutant emissions, CO emissions, unbumed hydrocarbons, volatile organic compounds, nitrous oxides, sulphur oxides) is application of new nanoarrays in nanocrystalline photonic filters at high temperature.

This invention therefore uses for the optical element one-dimensional arrays as primary arrays (nano- or microfibers) and built of these one-dimensional arrays two- dimensional nano- and microarrays (secondary arrays) of controlled shape, thickness and microstructure on the planar and curvature surfaces for applications in patterning, in visible and UV optics as mono- or multilayered filters. The inventive sensor can also be applicable as chemical or biological sensor beside of a local sensor systems in a gas turbines for local diagnostic purposes. Application in local diagnostic systems in gas turbines concerns use of primary 2D arrays as patterns for alignment of nanotubes doped with molecular iodine or other materials to ensure tunable ultra-thin notch filter properties for application in filtered Rayleigh scattering (FRS) laser diagnostic.

Basically methods of nanotechnology are extensively used for creating of new materials for optics, microelectronics and sensorics. Here the shear-flow crystallization of organic, metal oxide and metallic nanocolloids is used for obtaining different functional structures for the inventive application. Waveguides are important components of sensors and switches and composed of a core, surrounding by a cladding, which acts as a guide of electromagnetic radiation. The mechanism of wave guiding is total internal reflection of the radiation within the core. Generally, creating a waveguide requires changing the refractive index in the device. By using flexible substrates carrying many waveguides, flexible waveguides could be fabricated. In the waveguides, the core and cladding can be formed of the same material, for example of the polymer with different degree of polymerization, which depend on the curing time. The refraction index contrast in the micro- or nanoarrays, described in this invention disclosure, is created by the ordered alignment of one- dimensional arrays, controlled by the drying process. Brief description of the Drawings

The present invention is made more apparent, by way of example withput the intention of limiting the spirit or scope of the inventive idea, using preferred embodiments with reference to the accompanying drawings. Shown are in:

Fig. 1 High-temperature stable gas sensor for local combustion gas detection

Fig. 2 Patterns, showing structures NCA1 and NCA2,

Fig. 3 Preferential crystallographic orientation of colloidal arrays, imposed by the growth conditions. Fig. 4 Optical photograph, showing microfiber morphology of the NCA2 after drying,

Fig. 5 Possible geometries of two-dimensional microcrystalline arrays,

Fig. 6 Schematic illustrating location of carbon nanotubes inside of the channels in NCA1 , Fig. 7 Possible gratings built of patterned arrays of nanocrystalline building blocks, Fig. 8 Schematic of possible geometry of article for one of the optical applications, Fig. 9 Optical photographs, showing three different zones in the quartz capillary,

Fig. 10 Evolution of X-ray diffraction pattern of the colloidal array NCA2 as a function of temperature in high-temperature zone, Fig. 11 SEM patterns, illustrating change of the structure of the partially transparent crystal NCA3,

Fig. 12 TEM patterns taken from powderized crystals NCA3,

Fig. 13 SEM patterns of the crystals NCA4, consisting of the mixture of β- crystobalite and tridimite, Fig. 14 Transmission spectrum of NCA1 ,

Fig. 15 Schematic of remotely operated filter-bank and

Fig. 16 Detailed section view of a filter-bank. Ways to Carry out the Invention, Commercial Applicability

Figure 1 shows a temperature stable gas sensor for local combustion gas detection which is insertable in the wall 1 of a combustion chamber of a gas turbine arrangement. The wall 1 encloses the combustion chamber 2 in which hot gases 3 are produced in a combustion process. The optical sensor 4 is positioned at a location downstream the flame of the combustor. At this location an opening 6 in the combustors wall 1 is provided at which a measuring volume is encapsulated like a little chamber 5 which is positioned outside of the wall bordering the combustion chamber. The opening 6 is covered with a high temperature resistant selective porous membrane 7 through which selective constituents of hot gases can pass. To enhance the temperature resistance of the porous membrane 7 a protection coating 8 covers the porous membrane 7 at a side turning to the combustion chamber which withstands temperatures up to 1600⁰C.

Inside the little chamber 5 the optical sensor 4 is arranged for detecting important parameters of the combustion process, like temperature, pressure, NO_x pollutant emissions, CO emissions, unbumed hydrocarbons, volatile organic compounds, nitrous oxides and/or sulphur oxides inside the hot gases. The sensor provides a light source 9 which is a broad-band light source preferably, a waveguide element 10, an optical element 11 which is wavelength selective and a mirror 12. All components inside the little chamber 5 are of temperature resistance material and fabricated by shear flow crystallization at least the optical element 11 as described below.

Since the optical element 11 and the mirror 12 are placed inside said little chamber 5 by distance the so called absorption length 14 a light beam which emanates from the light source 9 and passes through the waveguide element 10, the optical filter element 11 to the mirror 12 at which the light beam will be reflected mainly and is redirected through the optical components 11 and 10 until entering a detector 13 which is provided like the light source 9 outside the little chamber 5, traverses the absorption length 14 twice. In present of absorbing particles like atoms and/molecules inside said little chamber 5 the light beam of the light source will be attenuated which is detected by the detector 13. The amount of attenuation can be brought into relation of special combustion parameters which are well known for a man who is skilled in the art. To ensure that parts of hot gases enter the little chamber 5 through the porous membrane 7 inside said little chamber 5 less pressure P1 is applied than the combustion pressure P2 inside the combustion chamber 3.

A main advantage of the inventive optical sensor is its temperature resistance which allows to measure very close to the combustion process to get pristine burner information. The basis of the temperature resistance is the way of production of preferably all sensor elements at least of the optical element by means of shear flow crystallization. The optical element 11 can be built as a single filter element or a filter bank as it is disclosed in preferred embodiment in Figure 15 and 16.

So in a preferred embodiment of the inventive optical sensor means for wave guiding is provided at least between said light source and said optical element for guiding light from said light source to said optical element and/or for guiding reflected light from said optical element to a detector which is made also by means of shear flow crystallization.

The necessity to optimize combustion operations, monitor the combustion processes to avoid instabilities and theirs severe consequences explain grown interest to control of combustion. Filtered Rayleigh scattering (FRS) is a new class of laser diagnostics permitting the measurement of single and multiple properties simultaneously, see papers by G. S. Elliott et al. in Measurement Science and Technology, 2001 , 12, 452-466 or D. Most and A. Leipertz in Applied Optics, 2001 , 40, 5379-5387. The FRS method employs molecular iodine filter in conjunction with an injection-seeded, frequency doubled Nd:YAG laser. In this technique, the different spectral broadening of the particle Mie and molecular Rayleigh scattering is used, to separate these two contributions by blocking out the Mie signal with appropriate ultra-thin molecular absorption filter, usually using molecular absorption line of iodine. The laser frequency can be tuned to transitions of iodine at 18 788 cm^"1. Comparison of measured signals with theoretical transmission (including Doppler shift and broadening) allows for the measurement of the average velocity, density, temperature and pressure can be determined.

Solid material with absorbed non-bonded molecular iodide could be used as ultra-thin notch filter. But attempts to absorb molecular iodine in solid matrix usually lead to breakage of the l-l bond and formation of the bond matrix-l. For example the formation of ad-layers on the reconstructed surfaces of Si(111 ) and Ge(111 ) leads to formation of Si-I and correspondingly Ge-I bonds, though with the conservation of strong l-l interaction. For Ge(100) the formation of a layer of molecular iodine on the top of iodine atoms, connected to Ge(100) (2x2)(c(2x2) surface was reported. The sorption of iodine by fully Cd²⁺-exchanged zeolite X results in disproportion and formation of cyclo-U and cyclo-U products. Especially interesting are results on absorption of iodine by mesoporous materials. In iodine-doping complexes of activated carbon fibers (ACF) consisting of three-dimensional disordered network of nano-graphites with a mean in-plane size of about 30 A with many neutral b molecules present in micropores, the charge-transfer rate of 0.008. For the intercalation of I₂ by nanographite (each nanoparticle of the size 7-8 nm made of 3-7 graphene sheet has polyhedral shape with hollow inside) was also reported weak degree charge-transfer of 0.024 (Raman spectra). Products of reversible intercalation of b by single-walled nanotubes (SWNT) exhibit moderate charge transfer of 0.018, which could mean formation of I₃ ^" and I₅ ^" in interstitial channels of SWNT bundle. For double-walled carbon nanotubes (DWNT), poly-iodine anions and the charge transfer between iodine and DWNT were also identified. This means that carbon nanotubes, ACFs and nanographite have similar affinity to iodine.

The inventive optical sensor relies on a method for fabrication and use of nano- or microcrystalline metal-, metal oxide or other inorganic nano- or microparticles arrays for sensors in gas turbines or combustors, for example for analysis of burner gas composition locally, near to the burner (temperature, pressure, NO_x pollutant emissions, CO emissions, unbumed hydrocarbons, volatile organic compounds, nitrous oxides, sulphur oxides). Such sensor systems allow for the immediate monitoring of combustion process.

The inventive optical sensor is based on a method of obtaining two-dimensional arrays (secondary arrays) built of aligned parallel or more complicated geometrical pattern, built of one-dimensional arrays (microfibers), which consist of metal oxide, metal or other inorganic nanoparticles, deposited onto amorphous, polycrystalline or single crystal support. Using shear-flow crystallization of nanoparticles with subsequent coalescence of the particles by external treatment, e.g. by heat treatment it is also possible to assemble one-, two- or three-dimensional nano- or microcrystalline arrays with the gradient of microstructure and electron/hat transport properties along the arrays or articles, built of these arrays.

Beside of producing the wavelength selective optical element in way of shear-flow crystallization it is also possible to obtain a multilayer mirror using said technology, built of grated overcoated surfaces with a matching multilayer structure, which reflects within the wavelength around the Bragg peaks, with reflectivity, tunable by changing the angle of incidence. Also mirrors with very broad bandwidth so called supermirrors can obtained by depositing multilayers of two-dimensionally micro- or nanoordered arrays with different periods, one for each desired wavelength band. Main limit is absorption. In dependence of process parameters multilayers are also obtainable, which reflect only within around Bragg peak with reflectivity in λ, which can be tuned by changing of angle of incidence can also be obtained. Two- dimensional arrays of aligned nanotubes, doped or not doped with molecular iodine, which could be used for waveguiding or for design of solid-state ultra-thin notch filter systems for application in Filtered Rayleigh scattering laser diagnostics of combustion operation are also possible to produce by using shear flow crystallization. Finally means for waveguiding by creation of one-dimensional or two-dimensional arrays possessing diffraction index contrast across the array for waveguides are obtainable with the before cited technique. So the inventive optical sensor device which preferably provides a light source, means for waveguiding, an optical wavelength selective filter element, a mirror and finally a light detector can mainly of its components produced by using shear flow crystallization.

Description of the experimental procedure of fabrication of some particular Siθ2 nanoarrays can be found in the article by V. Shklover in Chemistry of Materials, 2005, 17, 608-614. The gaskets for the packing cell were constructed form micro slides (Superiors, Marienfeld, Germany) and Mylar film with the thickness of 25 μm (Fralock Div. Lockwood Ind. Inc., 21054 Osborne St., Canoga Park, CA 91304, USA). All the gasket elements were carefully cleaned. First of three discussed in paper by Y. Lu et al. in. Langmuir, 2001 , 17, 6344-5350 non-lithographic methods of generation of the channel structure was applied, namely, the use of piece of soft paper to wipe the surface of the Mylar film. For the crystallization of the colloids, the sonicator Branson 1510 was used (Bransonic®) operating at frequency of 42 kHz, maximum power 80 W, RF-power 80 W. The fabrication process was continued with interruptions about five days. The controlled drying of the samples was done at 90 ⁰C and 60 ⁰C after the assembling was accomplished.

For detection purpose of the processed layers a scanning electron microscopy (SEM) LEO 1530 microscope with software LEO 32 V02.03 was used (LEO Elektronenmikroskopie GmbH), accelerating voltage was 3 kV, in-lens detector, samples were coated with 3 nm Au to avoid charging problem. Further a transmission electron microscopy (TEM), Tecnai G² F30 microscope with ultra high atomic resolution was used (FEI Company, Eindhoven, The Netherlands), operating at accelerating voltage of 300 kV. The SiO₂ particles for check of their amorphous character were prepared and the nano-crystalline arrays (NCA3) crystals were powderized in the mixture of ethanol and acetone and were brought onto the Cu grid, coated with carbon film (Plane GmbH, D-35578 Germany). Optical microscopy. Leica MZ 16 optical microscope in both transmission and reflection modes with software IM 1000 (Leica Microsystems) was used.

X-ray powder diffraction. Mar300 imaging-plate detector system (Marresearch GmbH, 1999) equipped with house-made furnace was used for powder diffraction measurements. In this design, the original base of the Mar300 was modified to allow the additional translation along the spindle axis to accommodate the furnace. The furnace consists of a housing of stainless steel with integrated water cooling, X-ray entry and exit windows (Kapton), window for in situ observation of the specimen by external CCD-camera. The furnace is filled with helium to provide an inert atmosphere and good thermal stability. Calibration of the furnace in the range from RT to 900 ⁰C was performed before the measurements. The quartz-glass capillary of 0.5 mm diameter and 0.01 mm wall thickness was filled up to ca. 30 mm length with rod-like crystals and sealed in the air. The sample was rotation in the range of φ = 0- 180° during data collection, 1800 sec per exposition with A(MoKaI )=0.7093 A (quartz monochromator), collimator diameter of 0.5 mm, and sample to detector distance of 200 mm. The measurement of standard Si sample was measured for precise determination of the sample to detector distance and x,y-pixel coordinates of the direct beam. The measurements at 20, 200, 400, 500, 550, 600, 700, 750, 800, 850, 900, 950 (20 successive measurements were performed at this temperature 950 ⁰C), 800, 500, 20 ⁰C were performed, the heating/cooling rate was 10 °C/min, the holding time at each predetermined temperature was equal to the duration of the X-ray exposition (30 min). All the X-ray experiments on Mar300 were performed using the mar345(dtb) software package (Marresearch GmbH, 2003), the conversion of X-ray results from 2D to 1 D data was done with FIT2D software, published by A. P. Hammersley as FIT2D V10.3 Reference Manual V4.0, ESRF98HA01T, ESRF, August 26, 1998. The STOE automated powder diffractometer system was used to check the phase composition of the sample NCA3, formed in the "cold" part of the capillary, filled with NCA2, during the annealing (Debye-Scherrer scan mode, CuKa, Ge(111 ) monochromator, a linear position sensitive detector).

The described in e.g. papers by Y. Xia et al. in Adv. Mater. 2000, 12, 693-713 and \r\ Aust. J. Chem., 2001 , 54, 287-290 shear-flow assembling method, combining hydrodynamic flow and physical confinement, was used for assembling of nanocolloidal arrays. The shear-flow crystallization is convenient assembling method, but its understanding requires the account of several complex processes: (a) sedimentation in a gravitational field, (b) hydrodynamic shear-flow with very small gradient of velocities G (G = 2v_mr/R², v_m is maximum velocity, r and R are radii of particle and channel, (c) Brownian motion and particles diffusion, (d) local fluctuations caused by ultrasonic sound waves radiation (-42 kHz, output 0.15 W cm^{" 2}) and (e) capillary stresses (can lead to cracking the film during the drying). To ensure the use of shear-flow method, the reference experiment on the assembling spherical 255 nm Seradyn colloidal polystyrene (PS) nanoparticles into ordered 3D nanocrystalline arrays NCA1 was performed. Then the 80 nm Klebosol^® colloidal silica particles were assembled into NCA2 colloidal crystals. The thickness of NCA1 and NCA2 of 25 μm was controlled by the thickness of used Mylar film. The scanning electron microscopy was used for the characterization of the obtained crystalline arrays NCA1 and NCA2. Figure 2 hereto shows SEM patterns of structures NCA1 and NCA2. Figure 2a shows a projection of the structure NCA1 , showing termination of the crystal by the faces {111} (plane ABC) and {110} (plane ABD). The crack parallel to one of the faces {110} could be also seen (plane CBD). Figure 2 b shows projection of the structure NCA2. The rational crystallographic orientation of the termination face {110} could be seen. The examination of the SEM patterns clearly indicates ...ABC... stacking sequence of PS or correspondingly silica particles and fee structure of NCA1 and NCA2 (space group Fm3m, packing density 74.05 %), observed in many colloidal crystals. The structures of the NCA1 and especially NCA2, which is built of the less monodisperse Siθ2 nanoparticles therefore have many defects. Disordered structure of nanoarrays seems to be problematic for device fabrication for e.g. photonic or microelectronic. The rational crystallographic orientation of the faces and major cracks of the resulted NCA1 and NCA2 draws attention to Fig. 2.

The observed character of growth of nanocrystalline arrays is the result of a combination of preferential growth mechanism of the primary nanocrystallites (small arrays formed by several primary nanoparticles) and geometrical constraints imposed on the nanocrystallites by cell geometry. The observed orientation of one of the crystallographic directions of resulted colloidal arrays NCA1 and NCA2 is along the shear directions, second one is parallel to the template surfaces, Fig. 3a and b. It means, that the shear flow direction (at small gradient of velocities) and template geometry can be used for the control of the film crystallography, what is beneficial for practical applications. Preferential crystallographic orientation of colloidal arrays, imposed by the growth conditions is shown in Figures 3 a and b. One of the crystallographic directions of the nanoarray of spherical particles oriented preferentially along the shear directions as indicated in Figure 3a. Template geometry dictates crystallographic orientation of the external top and bottom faces of the nanoarray parallel to the substrate is shown in Figure 3b. Arrows indicate direction of the shear flow, which is very slow and comparable to the sedimentation rate. Possibility to control kinetically the crystallographic orientation of the nanoarray is beneficial for practical applications.

The parallel assembling of nanowires (NW) and nanotubes (NT) on the chemically patterned substrates is another example of the combined use of fluidic alignment (shear flow) for hierarchical assembly of 1 D nanomaterials into functional network of controlled periodicity of the several μm-size. The dependence of the NW angular distribution on the shear-flow rate was also established.

The drying leads to very characteristic microfiber morphology of resulted NCA2 with the microfibers directions approximately parallel to the shear in homogeneous part of the film, microfibers width ranging from 50 to 200 μm and microfibers length up to 1 cm, is illustrated on Fig. 4. In Figure 4 optical photographs shows microfiber morphology of the NCA2 after drying. Fig. 4a shows a photograph, recorded in reflectance mode. Fig. 4 b shows photograph, recorded in transmission mode. The arrows indicate one of the pattern lines on the surface of NCA2 across the microfiber, confirming the appearance of the microfiber morphology due to drying. The observed line pattern on the microfiber surface could result from the oblique orientation of the packing cell during the crystallization. Fig. 4 c shows an isolated microfiber, built of colloidal NCA2 with remarkable mechanical stability.

The possible mechanism of cracking of primary 2D arrays was described by high capillary stresses and counteracting adherence to substrate by D. Bellet and L. Canham in Advanced Materials, 1998, 19, 487-490. The suggested methods of drying to avoid cracking comprise supercritical drying, drying with solvent of smaller surface tension, freeze-drying or slow evaporation. The formation of the microfibers during the drying process and not during the crystallization could be proved by detailed observation of the lines pattern on the microfibers surfaces, Fig. 4b. The NCA2 microfibers have remarkable mechanical stability. The microfiners in the central, most homogeneous part of the NCA2 can be approximately characterized by dimensions of 25 x 150 x 10000 μm. Of course, the 1 D nanoarrays (wires or fibers) can be used also as mats (both supporting and free-standing), but the observed controlled assembling of arrays with predicable crystallographic orientation brings many benefits, the nano- or microcrystalline planar patterns (films) could find more practical applications in optics and microelectronics than corresponding not oriented 2D or 3D bulk structures. The perfect alignment of the 1 D nano- or microarrays into planar high-density patterns is one of the challenges of nanotechnology.

Figures 5 a and b show possible geometries of two-dimensional microcrystalline arrays. The shown structure in Fig. 5a, consisting of parallel nearly equidistant planes, stabilized by corresponding annealing and sintering, could find application as integral components for VUV and soft X-rays optics, like monolayer and multilayer mirrors, plane and focusing gratings for focusing and monochromatization. The tunable parameters are: particles materials, particles diameter, gratings width, thickness, separations between the gratings, substrate material and substrate curvature. The possible substrate materials are: LiF, MgF₂, CaF₂, BaF₂, Al₂θ₃, quartz. Also non-uniformly spaced grooves can be used, especially variable-line spacings (VLS) gratings, see review by H. A. Padmore et al. in Vacuum Ultraviolet Spectroscopy II., Eds. J. A. Samson and D. L. Ederer. Academic Press, 1998. pp. 21-72.

The microfiber structures shown in Figures 6 and b can also find applications as templated surface for aligning of carbon nanotubes via self-assembling for application in field-emission displays and other microelectronic devices. The advantages of carbon nanotubes (CNT) as field-emission materials for displays and other vacuum microelectronic devices are low-threshold field for emission and sustainable high-emission current. Currently used CVD deposition at high temperatures > 800 ⁰C and reactive environment restrict application of CNT for devices with limited thermal and chemical stability, e.g. field-emission displays (FED). The disadvantage of screen-printing, one of the alternative approaches, is low resolution and inefficient use of materials. The channels width of 2-10 μm and separations between the channels, observed in the NCA1 , can be compared to self- assembled structure. Schematic illustration in Figure 6 shows locations of carbon nanotubes inside of the channels in NCA1 (cross-section view). The shear-flow method can be used for alignment of carbon nanotubes in the channels between Siθ2 microfibers. The values are of Di « 2 - 10 μm, D₂ « 50 - 200 μm.

The suggested "steric" assembling in Fig. 7 does not need pre-treatment of substrate for production of alternating hydrophobic and hydrophilic regions. The resulted pre- pattemed structure can be used, on the next step of nanostructure fabrication using self-assembling, for fabrication of 2D arrays of carbon nanotubes aligned within the channels between the microfibers. The aligned carbon nanotubes arrays can be used for waveguiding or for example after absorption of molecular iodine for design of ultra-thin notch filter systems in Filtered Rayleigh scattering (FRS), which is new class of laser diagnostics with important applications in optmization of combustion operation in gas turbines.

In order to check the sintering behavior of NCA2, the in situ X-ray study of the crystals NCA2 was carried out, placed into the quartz capillary of the diameter 0.5 mm and length ca. 30 mm. The X-ray image plate scanner MAR300, equipped with the high-temperature furnace was used. Three different zones can be distinguished in the capillary after the heating, see Fig. 7. Possible gratings built of patterned arrays of nanocrystalline building blocks like in Fig. 4a. Fig. 7 a and c show plane and focusing gratings. Fig. 7 b shows a nanocrystalline moiety, constituting e.g. black lines on a and c. The phase changes were checked in situ in the small part of the capillary, which was remaining in the X-ray beam during the measurements (X-ray spot size of 0.5 mm). Three structure transformations of NCA2 during the annealing can be observed, see Fig. 8 which shows schematic of possible geometry of article for one of the optical applications obtained by shear-flow crystallization of nanocolloids and subsequent heat-treatment curvature two-dimensional array, consisting of parallel aligned one-dimensional micro- or nanoarrays with crystallographically rational faces and controlled thickness (cross-sections of arrays are shown). One of the applications could be windowed photoemissive photodiodes.

The colloidal SiO₂ structure NCA2 transforms into cubic β-cristobalite (Fd3m, D_x=2.186 g cm^"3) between 750 ⁰C and 800 ⁰C. As a result of continuous heating at 950 ⁰C during 12 h, the β-cristobalite transforms into mixture of coexisting β - cristobalite (major phase) and hexagonal β -tridymite (P6₃/mmc, D_x=2.244 g cm^"3). Cooling to room temperature leads to formation of product, containing β-cristobalite, β-tridymite (major phases) and low quartz. According to ex situ X-ray study, performed ca. seven days after heating, the middle zone NCA4 contains mixture of tetragonal β-cristobalite (P4₃2-ι2, D_x=2.350 g cm^"3) and monoclininc α-tridimite (Cc, D_x=2.269 g cm^"3).

Formation of high-temperature polymorphic modification of SiO₂ cubic β-cristobalite at 750-800 ⁰C during annealing of colloidal crystal NCA1 draws attention. According to the generally accepted view of silica phase changes, the α-cristobalite is the stable form at room temperature. But the presence of β-cristobalite after heat treatment was observed in many other studies, see, for example by M. A. Butler et al. in J. Appl. Cryst. 1997, 30, 467-475. The transition between β- and β -cristobalite is strongly discontinuous and precise transition temperature can be altered by the presence of defects, like stacking faults, chemical impurities and there is a hysteresis in the measured value of T_tr on heating and cooling.

The β -cristobalite is more disordered than α-cristobalite and its presence in the product of annealing of colloidal nanocrystalline array immediately after heating could be very characteristic. Different models were developed to explain disordered structure of β -cristobalite, which has to possess unrealistic Si-O-Si bond angles of 180° and Si-O bond lengths of 1.51 A. A correlation between the domain theories of disorder nature or dynamic theories and observed in the present study formation of β-cristobalite from nanocrystalline material is of interest. Crystallization behavior and phase transitions using nanopowders may not follow the traditional phase transition routes. For example depending on synthesis route, nanocrystalline Zrθ2 starts to transform to the high-temperature stable tetragonal polymorph at essentially different temperatures: at ca. 1200 ⁰C for monoclinic crystalline ZrO₂ powder (size of cube-shaped crystallite ca. 14 nm) produced by forced hydrolysis, at. ca. 400 ⁰C for X-ray amorphous microspheres obtained by mixed-solvent precipitation, and at ca. 600 ⁰C for X-ray amorphous powder, obtained by alkoxide hydrolysis and condensation. The formation of tetragonal zirconia is probably related to the presence of amorphous zirconia because of their structural similarity, leading to possibility of topotactical crystallization on nuclei of amorphous zirconia as mechanism of crystallization of tetragonal zirconia. This special feature of nanocrystalline materials forms the basis of improvement of stability of thermally grown oxide by nano-controlling, reported in patent application by V. Shklover, et al. in patent Application WO 03/068673.

Three kinds of structures within one NCA3 crystal located between the heated and not heated zones can be observed having different degree of coalescence. Figure 9 show optical photographs hereto, showing three different zones in the quartz capillary, filled with microfiber colloidal crystals NCA2 and heated during in-situ X-ray measurements. Figure 9 a shows crystals NCA3 in low-temperature zone have transparency gradient along the crystal length with opaque part of the crystal in the right heated part. Figure 9 b shows opaque crystals NCA4 in medium-temperature zone and figure 9c shows opaque crystals NCA5 and products NCA6 of heating in high-temperature zone (smaller particles). The X-ray spot size 0.5 mm was focused on this part of the capillary during the in situ X-ray study. Capillary has diameter 0.5 mm, wall thickness 0.01 mm. The degree of ordering and porosity decreases when going from the "cold" to "hot" part of the NCA3 crystal. Remarkably, the structure NCA3 remains amorphous after the coalescence, according to the TEM data, Fig. 9c. Indeed, the Fast Fourier Transform (FFT) of the image on Fig. 9c shows weak diffraction spots of only photonic NCA3 crystal (its "cold" part), but not diffraction due to crystallinity of NCA3 grains. Diffusion of atoms on the cluster surface towards regions of lower curvature (less strongly bound to the neighboring atoms) was considered as driving force of coalescence of crystalline nanoparticles. It is possible, that coalescence of amorphous nanoparticles of NCA2 proceeds along other mechanism, including first crystallization stage. The formation of essentially disordered β -cristobalite structure as first observable crystalline product of annealing of NCA2, confirms this suggestion.

Figures 10 a, b and c show evolution of X-ray diffraction pattern of the colloidal array NCA2 as a function of temperature in high-temperature zone. Duration of every X-ray measurement at constant temperature was 30 min, heating or cooling rate was 10 °C/min, λ(MoKα1 ). Three structure transformations of CA2 during the annealing could be emphasized, (a) Heating-step. Formation of the β- crystobalite froms Siθ2 colloids during the heating at 750 - 800 ⁰C. (b) Isothermal annealing step. Formation of the mixture of coexisting cubic β-crystobalite (major phase) and hexagonal tridimite at prolonged isothermal heating at 950 ⁰C during ca. 12 h (pattern 950-20) and (c) Cooling step, final product contains β-crystobalite, tridimite and low quartz. Figures 11 a to d shows SEM patterns, illustrating change of the structure of the partially transparent crystal NCA3 as a function of the temperature gradient in the capillary during annealing. Figure 11a show crystal NCA3 used for SEM measurements. All the SEM patterns b - d were recorded from this crystal. Figure 11 b shows disordered closest fee packing in the initial NCA2 in the "cold" part of crystal NCA3. Figure 11c shows sintered structure in the middle part of crystal NCA3. The partial coalescence is observed, but the structure still exhibits ordering. Figure 11 d shows sintered "hot" part of the crystal NCA3 with local markers of the ordering still present.

Structure NCA3, exhibiting partial coalescence in Fig. 11c can be considered as model system for considering of phonon properties of nanomaterials. Opto- and microelectronic devices, like laser sources, systems for energy conversion, computers, may generate heat during their operation, but they have to remain within the very narrow temperature range to maintain possibility of frequency control, which defines circuit life-time, it means they need efficient heat-removal. On the contrary, current leads of high-T_c-superconductors have to have small heat conductivity and need thermal (e.g. metal oxide) buffer for directional heat removal. Electronic devices, comprising nano- and/or organic (or biological) components need especially careful heat management. Thermal conductivity of nanomaterials depends on both particle size and particles arrangement (nanoarrays) and may differ very much from corresponding bulk materials. It is known, that phonon transport in nano- and micrograin structures has to be reduced due to scattering at the grain boundaries. This is especially the case, when grain size becomes comparable to the phonon mean free part (MFP) at given temperature, which were approximately estimated as / = 5a(T_m/T), where a is the lattice constant, T_m is the melting point, and T is the temperature. The dependence of phonon transport in sintered structures depends on the diameter and transparency of the interface. The necks become transparent for phonons, if the necks size becomes comparable to the grain size. The anisotropy of

K S d thermal conductivity K equals approximately — = -¹^, where S₁₁, S₁ are the neck

K₁ S₁Ii₁₁ areas and d_p d_L are the grain sizes in the directions along and normal to the structure plane. This is the situation, we observe in NCA3, Fig. 12c. If the sintering leads to developing 1 D-, 2D- or 3D- network of grains, connected through transparent for phonons necks, we can expect a structure with anisotropy of thermal conductivity of corresponding dimensionality. Notice, that arrays NCA4 and NCA5, build of crystals with grains size much larger than phonon MFP, can be considered as bulk materials, what concerns phonon transport properties.

Figures 12 a to c shows TEM patterns taken from powderized crystals NCA3. Figure 12a show partially ordered structure in the middle part of the crystals NCA3. Figure 12b show essentially amorphous structure of the particles and interface "necks" in the middle and "hot" parts of the crystal NCA3. The absence of crystallinity was checked also by recording diffraction patterns. Figure 12c shows fragment of the NCA3 from the middle part. Diffraction due to only photonic structure of NCA3 could be seen, no diffraction due to crystallinity of SiO₂. Figures 13 a to c shows SEM patterns of the crystals NCA4, consisting of the mixture of β-cristobalite and tridimite. (a, b) Morphology of crystals, (c) Nanocrystals SiO₂, remaining on surface of sintered μm-sized crystals after annealing. The microcrystalline microfibers NCA5 after complete coalescence of nanoparticles are very porous and retain perfect microfiber morphology, Fig. 13a and 13b. The presence of not-transformed SiO₂ nanoparticles on the surface of μm-sized β- cristobalite crystals (Fig. 13c) after prolonged (around 12 h) heating may indicate possibility of the phase transition amorphous nanoparticles to crystalline nanoparticles at the first stage of annealing process with subsequent coalescence of small crystalline nanoparticles into μm-sized crystals.

An example of possibility of rejection/tuning of wavelength e.g. by the changing the angle between the incident beam and the normal to the surface of the filter is shown in Fig. 14 which shows transmission spectrums of NCA1 as a function of the angle φ between the primary monochromatic beam and normal to the NCA1 surface. The spectrum at φ = 35° was effected by the elements of cell design and can be excluded from discussion.

Schematic of remotely operated filter-bank for positioning the appropriate filter (A₁ to A₈) of UV, visible, NIR or IR spectra is shown in figure 15. M-M is the optical axis comprising direction to the sensor opening in gas turbine diagnostic system. Suggested fabrication method belongs to enabling technology for specific applications in gas turbine: broad-band and narrow-band filters for suppression of absorption by molecules in distinct spectrum regions (UV, visible, IR), selective band photonic defect-induced band-pass filters, mesoporous framework structures for FRS super-narrow notch filters, molecular super-narrow-band notch filters based on for example molecular iodine encapsulated inside of mesoporous framework solids. Particularly promising for gas turbine diagnostics is application of new nanoarrays in nanocrystalline photonic filters at high temperature (T< 600 ⁰C).

These kinds of different filter elements arranged all at one common substrate S are produced in one or several shear flow crystallisation processes using different flow conditions at each single filter element. To affect the flow conditions locally at each single filter location individually shaped flow barriers FB are provided between two neighbouring filters elements FE as depicted in Figure 16 showing a section of the disc shaped substrate. It will be assumed that the flow F of metal colloids while the shear flow step is directed from the centre C of the disc shaped substrate S radially outwards to its periphery P. Due to different shaped flow barriers FB there will be different formation of colloidal particles accumulation and at least different deposition at the filter element regions. All filter elements FE do provide the same distance D from the centre C of the substrate which is advantageously for integration in a sensor element so that the substrate S together with the multitude of filter elements can be arranged rotatable about its centre axis. Doing this different filter elements can be swung into the light beam if desired.

List of related patents

US patent No. 06 518 168 by N.-L. Jeon et al. entitled "Self-assembly monolayer directed patterning of surfaces", filed Feb. 11 , 2003, describes technique for creating self-assembly monolayer patterns of materials deposited on a surface and depositing via chemical vapor deposition of a pattern, complementary to the self-assembly monolayer.

US patent No. 06 355 198 by B. Enoch et al. entitled "Method of forming articles including waveguides via capillary micromolding and microtransfer molding", filed March 12, 2002, describes patterning by providing micromolding using fluid precursors for obtaining waveguides and claddings.

US 02 066 978 by E. Kim et al. entitled "Method of formation articles including waveguides via capillary micromolding and microtransfer molding", filed Jun. 06, 2002, describes formation of micropattemed articles from fluid precursors and mechanism of micro-scale positioning of biological objects on predetermined parts of surface.

US patent 6 180 239 by G. Whitesides et al. entitled "Derivatization and patterning of surfaces and more particularly to the formation of self-assembled monolayers on surfaces", filed Jan. 30, 2001 , describes the method of forming a patterned self- assembled monolayer on a surface and controlling the electrical properties of the surface.

US patent 6 304 364 by D. Qin et al. entitled "Elastomeric light valves", filed Nov. 16, 2001 , suggest electromagnetic radiation valves consisting of transparent elastomeric article having a contoured surface with protrusions and intervening indentations. CH patent application No. 03/068673 by V. Shklover, P. Bowen, K. Belaroui, H. Hofmann, M. Konter, entitled "Improvement of stability of thermally grown oxide by nano-controlling", filed July 16, 2003. The essence of this patent disclosure is the promotion of the growth of stable α-AI₂O₃ polymorph in the TGO (thermally grown oxide) by the introduction of thin (ca. 10 μm) pre-coating of nanocrystalline α-AI₂θ₃. The predominant content of the stable AI2O₃ phase and not complicated mixture of stable and metastable AI₂O₃ phases decreases amount of phase transitions and their associated volume changes in the TGO during the protected article service and by this mean increases stability and life-time of the coating.

List of reference singes

Wall of combustion chamber

Combustion chamber

Hot Gas

Optical sensor

Little chamber

Opening

Selective porous membrane

Protection coating

Light source

Waveguide

Optical element

Mirror Detector

Claims

What is claimed:

1. Optical sensor device for local analysis of a combustion process in a combustor (3) of a thermal power plant, in particular a gas turbine plant, wherein at least one wavelength selective optical element (11 ) is exposed directly or indirectly to hot combustion gases (3) being produced by said combustion process, said optical element (11 ) provides an array of nano- and/or microcrystalline fibres which are created by use of shear flow crystallization.

2. Optical sensor device according to claim 1 , wherein a light source (9) is provided emanating a light beam for passing through said optical element (11 ) directed onto a mirror (12) which is arranged oppositely to the light source (9) in respect to said optical element (11 ) having a mirror surface onto which said light beam is reflected at least partially so that at least a reflected light beam fraction passes said optical element (11 ) in opposite direction and being detected by a detector (13) positioned at the same site of said optical element (11 ) like the light source (9).

3. Optical sensor according to claim 2, wherein said mirror surface is of plane or parabolically shape and separated by distance from said optical element.

4. Optical sensor according to claims 2 or 3, wherein means for waveguiding (10) is provided at least between said light source (9) and said optical element (11 ) at least for guiding light from said light source (9)to said optical element (11 ) and/or for guiding reflected light from said optical element (11 ) to a detector (13).

5. Optical sensor according to one of the claims 1 to 4, wherein said optical element (11 ) provides a function of a photonic filter-bank having a local maximum peak of photonic transmission.

6. Optical sensor according to one of the claims 2 to 5, wherein at least said optical element (11) and said mirror (12) are positioned locally in said combustor or in kind of a measuring volume (5) which borders the wall (1 ) of the combustor, said volume (5) being encapsulated and communicate through an opening (6) to said combustor (2) so that at least parts of hot gases enter said measuring volume (5).

7. Optical sensor according to claim 6, wherein said opening (6) is covered by a high temperature resistant membrane (7) which is selectively for chemical compounds.

8. Optical sensor according to one of claims 1 to 7, wherein said optical sensor is positioned downstream to a burner flame in said combustor (2).

9. Optical sensor according to one of claims 1 to 8, wherein temperature, pressure, NO_x pollutant emissions, CO emissions, unbumed hydrocarbons, volatile organic compounds, nitrous oxides and/or sulphur oxides inside said combustor is/are detectable.

10. Optical sensor according to one of claims 1 to 9, wherein said array of nano- and/or microcrystalline fibres of the optical element (11 ) are of controllable dimensions, consisting of colloidal organic, metal oxides, metal or other inorganic nanoparticles, deposited onto amorphous, polycrystalline or single crystalline flat or curvature support using shear-flow crystallization method under distinct crystallization conditions and post-crystallization thermal treatment, leading to closely packed and distinct crystallographically rational orientation of the crystal packing of microfibers relative to their external faces and leading to distinct orientation of the microfibers relative to crystallization cell geometry, so called primary array.

11. Optical Sensor according to claim 10, wherein the primary array of nano- and/or microcrystalline fibres are aligned, especially parallel into two-dimensional arrays so called secondary arrays.

12. Optical sensor according to claim 11 , wherein the primary or secondary arrays are treated after crystallization, especially by external thermal treatment, to create gradient of coalescence of the particles and resulted gradient of microstructure and electron/heat transport properties along the arrays or articles, built of these arrays.

13. Optical sensor according to claim 11 or 12, wherein secondary arrays are designed to provide curvature gratings on at least one surface of said optical element for application in focusing and monochromatization of a light beam passing through said optical element.

14. Optical sensor according to claim 13, wherein said at least one surface of said optical element (11) is of spherical, ellipsoidal, toroidal, paraboloidal or cylindrical shape.

15. Optical sensor according to one of claims 11 to 14 wherein the secondary array is crystallized as a window that transmits the radiation of interest for possible application in windowed photoemissive photodiodes.

16. Optical sensor according to claims 11 to 15, wherein said secondary arrays of the optical element (11 ) are aligned in multilayers with different periods, each period for a different wavelength band.

17. Optical sensor according to claims 11 to 16, wherein said secondary arrays provides self aligned nanotubes, nanorods, nanowires onto a patterned optically transparent and high-temperature stable substrate.

18. Optical sensor according to claim 17, wherein the patterned substrate provides channels for self alignment of said nanotubes, nanorods, nanowires having a channel width between 2 to 10 μm and being separated laterally by distance between 50 to 200 μm.

19. Optical sensor according to claims 17 or 18, wherein said nanotubes, nanorods, nanowires are treated by adsorption or other treatment with molecular iodine or other molecular or atomic media exhibiting no remarkable chemical interaction.

20. Optical sensor according to one of claims 1 to 19, wherein said at least two optical elements (4, FE) are provides on one and the same substrate (S) onto which each array of nano- and/or microcrystalline fibres are created by the use of shear flow crystallization while rotating said substrate (S) about an axis of rotation and said at least two optical elements are positioned on said substrate at different circular sections towards said axis.

21. Optical sensor according to claim 20, wherein at least two optical elements (4, FE) differ in properties of optical wavelength transmission.

22. Optical sensor according to claim 20 or 21 , wherein said optical elements are arranged with equal distance (D) from said axis of rotation (C).

23. Optical sensor according to claim 22, wherein said substrate is arranged pivotable about said axis of rotation in an manner, so that the optical elements (11 , FE) are alignable if required one after the other to said light beam emanating from said light source (9).

24. Optical sensor according to claims 4 to 23, wherein said for waveguiding means (10) is produced by means of shear flow crystallization.

25. Use of an optical sensor according to one of claims 1 to 24 for combined passive filtering of required parts of UV, visible, NIR or IR spectra and for active qualitative and quantitative sensing of change of gases inside the combustion chamber or inside other parts of a gas turbine due to change oxygen concentration and solid state conductivity inside the metal oxide ceramic or non-oxide ceramic working elements of sensors.