In situ high-temperature Raman spectroscopy with a remote fiber-optic Raman probe.

Abstract

The optical fiber sensor system for in situ high-temperature Raman spectroscopy includes an excitation light source, a spectrometer, and a Raman probe. The Raman probe comprises at least one optical fiber connected to the excitation light source and the spectrometer. The at least one optical fiber transmits laser excitation from the excitation light source to the sample and collects light scattered by the sample for analysis by the spectrometer. An external lens arrangement positioned at a distal end of the at least one optical fiber of the Raman probe optically connects the at least one optical fiber to the sample and physically separates the at least one optical fiber from the sample during sampling. The spectrometer performs Raman spectroscopy of the sample based on the light collected by the at least one optical fiber.

Description

[Technical Field]

[0001] (CROSS-REFERENCE TO RELATED APPLICATIONS) This application claims the benefit of U.S. Provisional Application No. 63 / 447,594 (filed February 22, 2023), the entire disclosure of which is incorporated herein by reference. (Statement of Government Interest) This invention was made with government support under Advanced Manufacturing Office (AMO) Grants DE-EE0009392 and DE-EE0009119 awarded by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy (EERE). The government has certain rights in this invention. [Background technology]

[0002] When photons of light interact with molecules, many of the photons are affected by small density fluctuations, changing their direction and forming Rayleigh scattering. However, a small number of photons change their frequency through Raman scattering. The intensity of the Raman scattered light is 10 times the total scattered light intensity. -6 ~10 -10 Raman scattering arises from the energy exchange between photons and molecules, which changes the photon's energy. These scattered photons enter various vibrational states specific to the molecule's chemical bonds and symmetry. Raman scattering is therefore known as a fingerprinting technique for diagnosing chemicals, providing a tool for investigating chemical structure and identifying molecular structures. With the introduction of lasers, a powerful supply of high-quality, stable, and intense monochromatic laser light became available, providing a major impetus for studying Raman scattering and its applications in various conditions.

[0003] Raman spectroscopy has been applied in various fields, including chemistry, physics, biology, and medicine. In civil engineering, Raman spectroscopy has also been applied to detect the chemical properties of concrete materials during hydration and carbonation reactions. Furthermore, Raman spectroscopy is useful for qualitative analysis, advanced quantitative analysis, and molecular structure determination. While conventional stationary Raman systems have the advantages of high precision and accuracy, they also have obvious drawbacks. First, the complex equipment is not suitable for performing real-time online measurements. Second, stationary Raman systems have specific requirements regarding sample size. As a result, large samples require special preparation, and nondestructive measurements are not possible. However, online and real-time diagnostics for biochemistry, medicine, and materials science applications are essential for scientific research. Therefore, a portable fiber optic Raman sensor capable of studying material properties in real time is needed.

[0004] Several fiber-optic Raman sensor designs have been reported for various applications. The low light-collection efficiency of fiber-optic Raman sensors led to the development of beveled fiber-optic confocal Raman probes. A spherical lens is attached to the end of the fiber to focus the light into a small spot. By selecting an appropriate fiber-optic spherical lens and the bevel angle of the focusing fiber, the confocal Raman probe design can be optimized to maximize Raman measurements of superficial epithelial tissues. Fiber-optic Raman probes are widely used in biochemistry and medicine. However, the properties of the fiber material limit the expansion of fiber-optic Raman probes in some materials science applications, such as in situ high-temperature studies. Novel materials are being developed through high-temperature thermal processing, and understanding how chemical composition changes in real time during high-temperature reactions could be a novel breakthrough for materials science research. Early studies reported in situ Raman spectra of silica glass in the glass transition temperature range from room temperature to 1950 K for supercooled liquids. These studies used a novel wire-loop heating technique to heat the samples to high temperatures. However, the microheating coils are only 0.5–0.8 mm in diameter, significantly limiting the size of experimental samples. Furthermore, Raman spectra are acquired using a stationary Raman system, which limits their applicability to other applications.

[0005] Scientific researchers have presented preliminary research applying Raman spectroscopy at extremely high temperatures to study the chemical properties of molten materials. The researchers conducted preliminary high-temperature Raman spectroscopy using a small electric wire heater and a stationary Raman system. The results demonstrated a correlation between high-temperature Raman spectra and the chemical structure of liquids, contributing to understanding the chemical properties of molten samples. However, the small electric wire heater employs a heated wire loop with a diameter of 0.5–0.8 mm to hold the glass sample, which significantly limits the volume and size of the test sample. Second, the lack of portability of the stationary Raman system also limits the application of high-temperature Raman technology to practical industrial and production applications.

[0006] There is a need for a probe head that is remote from the Raman spectrometer and can be positioned in close proximity to the hot molten material. Summary of the Invention [Means for solving the problem]

[0007] Aspects of the present disclosure relate to an in situ high-temperature fiber optic Raman probe that enables structural studies of glass and slag samples at temperatures up to 1400°C. A specially designed external telescope is incorporated into the portable fiber optic Raman probe to extend the optical working distance, enabling the probe to operate in high-temperature environments. Additionally, aspects of the present disclosure include a deconvolution algorithm configured to identify peaks in the spectrum and correlate them with the molecular structure of components in each sample. This flexible and reliable high-temperature Raman measurement method has great potential for a variety of applications, such as materials development, compositional and structural monitoring during high-temperature processing, chemical identification, and process monitoring in industrial manufacturing.

[0008] In one aspect, an optical fiber sensor system for in situ high-temperature Raman spectroscopy is provided. The system includes an excitation light source, a spectrometer, and a Raman probe. The Raman probe comprises at least one optical fiber connected to the excitation light source and the spectrometer. The at least one optical fiber transmits laser excitation from the excitation light source to the sample and collects light scattered by the sample for analysis by the spectrometer. An external lens arrangement positioned at the distal end of the at least one optical fiber of the Raman probe optically connects the at least one optical fiber to the sample and physically separates the at least one optical fiber from the sample during sampling. The spectrometer performs Raman spectroscopy of the sample based on the light collected by the at least one optical fiber.

[0009] Other objects and features of the present disclosure will be in part apparent and in part pointed out herein. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic diagram of an optical fiber Raman probe with an external telescope, according to one embodiment. [Figure 2] FIG. 2 is a schematic diagram of an in situ high-temperature fiber optic Raman system including the Raman probe of FIG. 1 according to one embodiment. [Figure 3] 1 is an example of Raman spectra of an electric arc furnace slag sample at room temperature and at a temperature of 1400° C., according to one embodiment. [Figure 4A] 4 shows an exemplary deconvolution result of the Raman spectrum of FIG. 3 for a sample at room temperature, according to one embodiment. [Figure 4B] 4 shows an exemplary deconvolution result of the Raman spectrum of FIG. 3 for a sample at 1400° C., according to one embodiment. [Figure 5] 1 is an example of Raman spectra of mold flux at room temperature, 800° C., and 1400° C., according to one embodiment. [Figure 6A] 6 shows an exemplary deconvolution result of the Raman spectrum of FIG. 5 for mold flux at room temperature, according to one embodiment. [Figure 6B] 6 shows an exemplary deconvolution result of the Raman spectrum of FIG. 5 for mold flux at 1400° C., according to one embodiment. [Figure 7] 1 is an example of Raman spectra of bioglass at room temperature, 100° C., 700° C., and 1300° C., according to one embodiment. [Figure 8A] 8 shows an exemplary deconvolution result of the Raman spectrum of FIG. 7 for bioglass at room temperature, according to one embodiment. [Figure 8B] 8 shows an exemplary deconvolution result of the Raman spectrum of FIG. 7 for bioglass at 1300° C., according to one embodiment. [Figure 9]2 is a schematic diagram of an in situ high-temperature fiber optic Raman system including the Raman probe of FIG. 1 according to another embodiment. [Figure 10] Examples of Raman spectra from 400 to 1200 cm-1 showing the Raman region of four synthetic flux samples at a temperature condition of 1400 °C are shown. [Figure 11A] For Sample 1 in Figure 10, an example of the deconvoluted Raman spectrum of the 1400°C synthesis flux in the Q region of 850 to 1100 cm −1 is shown. [Figure 11B] For sample 2 in Figure 10, an example of the deconvoluted Raman spectrum of the 1400°C synthesis flux in the Q region of 850-1100 cm-1 is shown. [Figure 11C] For sample 3 in Figure 10, an example of the deconvoluted Raman spectrum of the 1400°C synthesis flux in the Q region of 850-1100 cm-1 is shown. [Figure 11D] For sample 4 in Figure 10, an example of the deconvoluted Raman spectrum of the 1400°C synthesis flux in the Q region of 850-1100 cm-1 is shown. [Figure 12A] For Samples 1 to 4 in FIG. 10, examples of correlations obtained from the Q region peak area ratios with the chemical composition and basicity of the flux are shown. [Figure 12B] For Samples 1 to 4 in FIG. 10, examples of correlations obtained from the Q region peak area ratios with the chemical composition and basicity of the flux are shown. [Figure 12C] For Samples 1 to 4 in FIG. 10, examples of correlations obtained from the Q region peak area ratios with the chemical composition and basicity of the flux are shown. [Figure 13] 1 shows example Raman spectra for industrial sample A at various temperature conditions, according to one embodiment. [Figure 14A] FIG. 1 shows an example of a deconvoluted Raman spectrum of industrial flux sample A at 1350° C. for the Q region from 850 to 1100 cm, according to one embodiment. [Figure 14B]FIG. 1 shows an example of a deconvoluted Raman spectrum of industrial flux sample B at 1350° C. for the Q region from 850 to 1100 cm, according to one embodiment. [Figure 14C] FIG. 1 shows an example of a deconvoluted Raman spectrum of industrial flux sample C at 1350° C. for the Q region from 850 to 1100 cm, according to one embodiment. [Figure 15] 1 shows an example of Q-region relative Raman peak area ratios compared to industrial sample viscosity values, according to one embodiment. Corresponding reference numbers indicate corresponding parts throughout the drawings. DETAILED DESCRIPTION OF THE INVENTION

[0011] According to an embodiment of the present disclosure, a fiber optic Raman probe is configured for high-temperature Raman spectroscopy to perform in situ high-temperature Raman analysis. A specially designed external telescope is incorporated at the end of the Raman probe to extend the working distance of the probe, e.g., from 0.8 cm to 3 cm, thereby protecting the probe from the high-temperature environment (e.g., 1400°C) for in situ studies. A deconvolution algorithm is applied to analyze and deconvolve the Raman spectrum to identify the unique chemical bonds that contribute to the overall Raman spectral response.

[0012] Referring to FIG. 1, the fiber optic Raman probe 101 has a main optical body 103 and an extended tip 107 that holds a final collection lens 109. The probe 101 is connected to an excitation light source (see FIG. 2) and a spectrometer (see FIG. 2) via two optical fibers 111 and 113, respectively, to enable remote measurement of the sample. In one embodiment, as shown in FIG. 1, the probe 101 employs an excitation fiber 111 with a core diameter of 105 μm and a collection fiber 113 with a core diameter of 100 μm. It should be understood that the Raman probe 101 can include multiple collection fibers 113 within the scope of the present disclosure.

[0013] Micro-optics are used within the main optical body 103 of the fiber optic Raman probe 101 to deliver laser excitation light to the sample via fiber 111 and collect scattered light via fiber 113. This results in a compact probe head optically connected to the laser source and spectrometer. The efficient use of one or more filters, such as a bandpass filter 117, a dichroic filter 119, and an edge filter (not shown), to separate the excitation and scattered light results in a 180° sampling geometry. The backscattered light collection geometry allows for easy sample alignment and provides optimal throughput due to the overlap between the excitation and collection cones. Another advantage of using the second fiber 113 for signal collection is the elimination of inelastic background signals from the excitation fiber 111 itself. Based on the lens design of the Raman probe in the illustrated embodiment, the default working distance is, for example, 8 mm. Furthermore, the laser spot size can be varied based on the transmission characteristics of the sample being measured. For example, a minimum spot size of approximately 100 μm may be used with a sample-to-probe distance of approximately 8 mm. Each optical fiber 111, 113 is mounted within a protective polyurethane jacket 123. The Raman probe 101 has an outer jacket 125 formed of metal (e.g., 316 stainless steel), which provides a maximum temperature threshold of 650° C. for high-temperature use.

[0014] Common glass materials, slags, and fluxes are processed at high temperatures, around 1300–1400°C. Therefore, the Raman probe 101 employs an external telescope 129 to extend its working distance by isolating the probe 101 from the heated environment. Protected by a stainless steel jacket 125 (operable up to 900°C), the external telescope 129 is specifically designed to extend the high-temperature capabilities of the Raman probe 101. As shown in FIG. 1 , the excitation light can be focused to, for example, 0.8 cm after passing through lens 109 (e.g., a sapphire lens). In an alternative embodiment, the external telescope 129 includes two additional optical lenses 133 that first collimate the light and then focus it, thereby increasing the focal length of the optical path to, for example, 3 cm without energy loss. The increased working distance of the Raman probe 101 provided by the external telescope 129 enables even higher-temperature experiments.

[0015] 2, an excitation light source 203 and a spectrometer 205 are connected to each end of the Raman probe 101 via optical fibers 111 and 113, respectively. In one embodiment, the excitation light source 203 comprises a 532 nm laser (green laser) configured to provide the best Raman performance for the test sample, and the spectrometer 205 comprises a 3 cm -1 The QE-Pro spectrometer is employed to collect Raman spectra with a spectral resolution of 100 kHz. An induction coil furnace 207 is used to provide high temperature continuous heating.

[0016] Referring further to FIG. 2 , the fiber optic Raman probe 101 is mounted on a stand 209 with an appropriate probe working distance. A K-type thermocouple 213 is fixed in close proximity to the Raman probe 101 to monitor its ambient temperature in real time. After repeated experiments, the temperature in the probe region is approximately 150°C at a controlled distance of 3 cm from the sample. Furthermore, real-time temperature data is sampled and recorded using a data logger 215. According to an embodiment of the present disclosure, two types of graphite crucibles can be used for high-temperature experiments. A base crucible 217 is surrounded by an aluminosilicate refractory and fixed within the induction coil 207. A working crucible or insert crucible 219 is embedded within the base crucible 217 as a container for heating the sample. An S-type thermocouple 223, placed in direct contact with the working crucible 219, is used to control the sample temperature. After repeated experimental measurements, it was found that the heating system could consistently heat the sample to 1400 °C in 10-15 minutes. Therefore, real-time high-temperature Raman spectra could be collected repeatedly under the same temperature conditions, improving the accuracy of the experiment.

[0017] It is well known that thermal radiation can affect Raman scattering at high temperatures. The energy level of thermal radiation is closely related to temperature. Therefore, a strong thermal radiation signal can significantly affect the Raman signal of a test sample. However, the energy level of thermal radiation is considered to be the same under the same temperature conditions. Therefore, if all Raman spectra are investigated at the same temperature, problems associated with the inherent temperature dependence of Raman scattering can be eliminated. Aspects of the present disclosure include a background subtraction method to eliminate background thermal radiation signals. In high-temperature experiments, for example, an empty furnace is first heated to and maintained at 1400°C to ensure thermal stability. Then, the fiber-optic Raman probe 101 is moved to the measurement area to acquire the background signal. Subsequently, the prepared sample is poured into a crucible for sintering and melting. Finally, when the sample is completely melted, a second Raman signal is acquired by the probe 101. To improve the signal-to-noise ratio of the actual sample Raman signal, the acquisition integration time is set to, for example, 2 to 5 seconds. Multiple sets of Raman spectra (eg, 10) for each temperature are acquired and averaged to further reduce noise signal interference and address uncertainties in signal collection.

[0018] Example 1 For example, the first sample examined by the in situ fiber optic Raman probe 101 was electric arc furnace (EAF) slag produced during the EAF steelmaking process. Previous studies have shown that the viscosity of this molten slag is between 800 and 1050 cm -1 This strongly correlates with the Raman signal in the Q region of EAF slag. Therefore, in this example, we investigated the Q region Raman spectrum of EAF slag. Raman spectra successfully collected at room temperature and 1400 °C show that the sample exhibits a strong carbonate peak (1083 cm) before heating, as shown in Figure 3. -1) were observed. However, after the slag was heated to a liquid state (1400 °C), the Raman peak carbonate intensity decreased due to carbonate decomposition in the high-temperature environment. The Raman signal spectrum in the Q region at room temperature changed dramatically at high temperatures. Multiple compounds and molecular bonds have close wavenumbers, often resulting in overlapping peaks in complex samples. Therefore, deconvolution techniques are widely used in Raman spectroscopy for further chemical analysis of wide-range Raman spectra.

[0019] In one embodiment, a deconvolution algorithm with a Gaussian function is applied and used for curve fitting. Figure 4A shows the deconvolution results of the Raman spectrum for the EAF slug at room temperature, and Figure 4B shows the deconvolution results of the Raman spectrum for the EAF slug at high temperature. Based on the Raman spectral data summarized in previous experiments, the signal in the Q region is estimated to be Q 0 (Si2O4) 4- ,Q 1 (Si2O7) 6- ,Q 2 (SiO3) 2- ,Q 3 (Si2O5) 2- The deconvolution results were obtained.

[0020] With further reference to Example 1, by comparing the deconvoluted results of the EAF slag sample at both temperatures, it is observed that the Raman signal in the Q region shifts to the right after heating. After comparing the Raman deconvolution data at room temperature and at elevated temperatures, the following findings were made: First, in the room temperature deconvoluted Raman results, the Q 3 The reason is that the high content of Fe2O3 (26 wt%) and high basicity (CaO / SiO2 = 2) in the slag resulted in a Q 3 Q 2 ,Q 1 ,Q 0 However, when the sample is heated to a molten state at 1400 °C, a small Q 3Next, the Q peak observed after heat treatment 2 The increase in binding is due to the n The burden of reduced bonding suggests that the crystalline phase is a cyclosilicate, a conclusion confirmed by previous studies on quenched and cooled samples.

[0021] Example 2 The second sample examined by the in situ fiber optic Raman probe 101 was a mold flux, Ca-Si-Al-oxyfluoride glass, used to control heat transfer and lubricate the mold in the continuous casting process. Raman spectra were collected for three different temperature conditions, as shown in Figure 5. Apart from the Q-region Raman signal, the Raman spectrum was also observed in the 520-540 cm -1 and 645 cm -1 Two prominent Raman peaks were observed at 520–540 cm, representing the vibrational bonds of Al–O–Al and Si–O. The high melting points of alumina and silicon are also observed. -1 and 645 cm -1 The peak at 800–1100 cm was extremely strong, and no significant change was observed with increasing temperature. -1 The changes in the Q-region Raman spectrum from are evident and represent the three states of the mold flux: solid, crystallized, and liquid or molten.

[0022] Again, deconvolution analysis was performed on the room temperature and high temperature mold flux Raman data, as shown in Figures 6A and 6B. Four distinct Raman peaks, Q, were identified for both Raman spectra. 0 ,Q 1 ,Q 2 ,Q 3 The most prominent Q in the Raman spectrum was 0 The peaks represent the majority of silicate monomers in the sample at room temperature. Upon heating, the majority of the monomers recombine with oxygen to form chains and sheets, i.e., Q 2 ,Q 3 Then, Q 0The Raman peaks are significantly reduced and Q 2 A Raman peak was found to be dominant in the Raman spectrum at 1400 °C, indicating that the molten flux sample contains silicate chains. Furthermore, an increased degree of polymerization is associated with an elevated melting point and higher mechanical strength. Previous studies have shown that Q 3 / Q 2 The Q ratio was found to be a good indicator of the degree of polymerization. After analyzing the data from this preliminary study, it was found that the Q ratio increased with increasing temperature. 3 / Q 2 The ratio decreased, which was in agreement with previous studies.

[0023] Example 3 The third sample investigated using the in situ fiber-optic Raman probe 101 was a bioactive glass (45S5), which is used as an implantable device in the human body to repair and replace diseased or damaged bone. As shown in Figure 7, four Raman spectra were successfully observed at room temperature, 100°C, 700°C, and 1300°C. First, a room-temperature Raman spectrum was acquired before heating the sample. The sample was then heated to 1300°C to melt it, and a Raman spectrum was acquired. The furnace temperature was then adjusted to 700°C and maintained at a constant temperature. Once the temperature stabilized, a Raman spectrum was acquired at 700°C. Finally, the temperature was reduced to 100°C, and a Raman spectrum was acquired. This heating protocol was adopted to reduce the influence of metastable crystalline phases on the analysis. The room-temperature Raman spectrum exhibited two major bands: a PO bending band at 620 cm; and a PO bending band at 620 cm. -1 and 800-1100 cm as the Q region -1 However, the high-temperature Raman spectra at 100°C, 700°C, and 1300°C show the range of 620cm after heating. -1 The PO bending band disappeared at 585 cm. -1 A new Raman peak was found at v4 PO4, which probably includes contributions from acid phosphate (HPO4) and octacalcium phosphate (OCP). 3- Furthermore, the Raman signal in the Q region changes dramatically with temperature.

[0024] With further reference to Example 3, to further understand the relationship between the chemical bonds in bioglass and the Raman spectrum, peaks identified from previous studies on bioglass were applied to deconvolute the Raman spectra at room and elevated temperatures. Seven major Raman peaks were deconvoluted, and the attributes considered for peak fitting were: 864 cm -1 Q in 0 , 906cm -1 Q in 1 , 944cm -1 Q in 2 , 974cm -1 POP stretchable, 1008cm -1 OPO stretching of the P2O5 sheet unit at 1086 cm, asymmetric stretching of the bridging oxygen in all Q species, and -1 Q in 3 (See Figures 8A and 8B.) The room-temperature Raman results obtained by deconvolution also showed good agreement with those of previous studies.

[0025] Advantageously, the fiber optic Raman probe 101 has been demonstrated to perform real-time, in situ, high-temperature Raman spectroscopy. The external telescope 129 increases the working distance of the fiber optic Raman probe 101, enabling real-time Raman analysis to be performed on large samples at high temperatures. The extended working distance demonstrates the fiber optic Raman probe 101's capabilities for high-temperature environmental measurements and long-term detection. Aspects of the present disclosure serve as a roadmap for investigating material properties under high-temperature conditions, which can facilitate the use of high-temperature fiber optic Raman spectroscopy in metal processing, steelmaking, and other high-temperature-related material research. In other embodiments, aspects of the present disclosure provide an in situ high-temperature fiber optic Raman sensor for mold flux analysis in steel manufacturing applications.

[0026] The ability to remotely control molten flux composition through in situ chemical fingerprinting will have a significant impact on continuous casting in the steel industry. Continuous casting in steel production uses specially developed oxyfluoride glasses (mold fluxes) to lubricate the mold and control the solidification of steel within the mold. The composition of the flux influences properties such as basicity, viscosity, and crystallization rate, all of which affect the stability of the casting process and the quality of the solidified steel. However, as mold fluxes interact with the steel during the casting process, the flux chemistry changes, which must be considered in flux design. Currently, the chemical composition of mold fluxes must be determined by extracting flux samples from the mold during casting and processing these samples offline to estimate the actual chemical composition and predicted properties of the flux.

[0027] Raman spectroscopy offers an alternative method for conducting flux analysis with the potential to perform measurements online during the casting process. Raman spectroscopy uniquely identifies specific molecules in a vitreous flux by revealing peaks that are fingerprints of the vibrational modes of the molecules in the flux. The intensity of specific peaks in the Raman spectrum can be correlated with the chemical composition of the melt and related properties such as basicity and viscosity.

[0028] FIG. 9 illustrates an alternative embodiment of a high-temperature fiber-optic Raman system 901 configured to investigate the chemical composition of mold flux in a high-temperature environment. The probe system 901 includes a probe 101. In one embodiment, the main optical body 103 of the probe 101 includes a removable extension (e.g., 25 cm long) to further protect the probe from harsh environments. In this embodiment, the probe 101 has a 0.9 cm diameter tip, suitable for testing in the limited space and harsh environments associated with steel manufacturing. The Raman probe 101 includes a 105 μm excitation fiber 111 and a 100 μm collection fiber 113 for optical excitation and collection, respectively. Filters, such as a long-pass filter and a band-pass filter (e.g., center wavelength 532 nm), are installed inside the probe body to block laser reflections and eliminate inelastic background signals generated by the optical fiber. A concave sapphire window is employed at the tip of the extension probe, providing the Raman probe with a working distance of 7.5 mm. All optical fiber subunits are housed in an outer protective jacket 125 made of 316 stainless steel, giving a maximum operating temperature of 650° C. In the optical fiber Raman system shown in FIG. 9, the excitation source 203 comprises a 532 nm laser (green laser) and the spectrometer 205 comprises a 3 cm -1 It is equipped with a QE-Pro spectrometer that records Raman spectra with a spectral resolution of .

[0029] In the embodiment of FIG. 9, an induction furnace 207 provides heating for high-temperature Raman experiments. As shown, two thermocouples 213 (e.g., a K-type and an S-type) are installed to monitor the temperature of a crucible 219 containing molten flux. The K-type thermocouple is aligned with the tip of the Raman probe 101 and monitors the ambient temperature to prevent the probe head from overheating. Real-time temperature data is collected by a thermocouple data logger 215. The S-type thermocouple is placed in direct contact with a graphite crucible 219 containing a molten flux sample to monitor the flux temperature. The crucible 219 is, for example, approximately 10 mm in diameter and 20 mm deep, and the flux sample typically forms a 10 mm diameter sphere upon melting. Using this heating system, a synthetic mold flux sample is melted and heated to 1400°C, and high-temperature Raman spectra are collected in real time.

[0030] Considering the possible influence of natural light on the Raman signal, high-temperature Raman signal collection is performed under darkroom conditions to improve the signal-to-noise ratio. Furthermore, thermal radiation generated at high temperatures is an important component of the Raman signal. Therefore, a background subtraction method is applied to remove the thermal radiation signal and minimize the interference from thermal radiation on the natural Raman spectrum from the flux sample. First, the furnace is heated to 1400 °C. Then, the Raman probe 101 is moved above the graphite crucible 219 to collect the background optical signal in a high-temperature environment. After acquiring the background spectrum, the mold flux material is added to the graphite crucible 219, completely melted at 1400 °C, and the Raman spectrum is collected again. The background subtraction method effectively subtracts the thermal radiation signal features from the acquired spectrum, leaving only the actual Raman signal from the sample. The integration time of the Raman signal is set, for example, to 2 to 5 seconds for data collection. Multiple spectra (e.g., five) are collected for each measurement condition and averaged to reduce measurement uncertainty.

[0031] Example 4 Four flux samples with different compositions were prepared. The melt structures of Samples 1 to 4 were analyzed by Raman spectroscopy. The real-time Raman spectrum at 1400 °C is shown in Figure 10. -1 and 590-740cm -1 The Raman peaks located at 800-1200 cm correspond to the Al-O-Al and Si-O-Si bending vibrations, respectively. -1 The Raman shift at 480-560 cm is related to the stretching of the Q(Si) bond. From the Raman spectrum shown in Figure 10, the peak in the Q region at 1400 °C changes with the chemical composition of the mold flux. -1 and 590-740cm -1 The Al-O-Al bond and Si-O-Si bending vibration characteristic peaks in the region of 640 cm do not show any significant shift or change in composition. However, a preliminary comparison of the silica peak intensities of Sample 1 and Sample 4 shows that the peaks at 640 cm -1 It can be seen that the characteristic peak of silica at is weaker for the S1 sample than for the S4 sample, indicating that the SiO2 content is reduced.

[0032] By identifying correlations between Raman spectroscopy and chemical composition, we demonstrate the relationship between Raman spectroscopy and flux properties. To more accurately quantify the obtained Raman spectra, we perform a typical deconvolution method on the peaks in the alumina and silica regions using, for example, Origin software. A Gaussian function was applied for curve fitting in the deconvolution algorithm. Based on previous studies, the deconvolution was performed on the peaks in the interval 480–560 cm. -1 and 590-740cm -1 It was carried out at 520cm -1 and 640 cm -1 The peak at was identified as the central peak after deconvolution. The cumulative fitting peaks in the figure were compared with the original spectral data, and an overall fitting greater than 96% R-squared value was achieved, resolving the uncertainty of the deconvolution.

[0033] Further referring to Example 4, the main components of the synthetic mold flux are SiO2, CaO, and Al2O3. Because Al2O3 is fixed at 4.69 wt% in all four samples, structural changes are strongly influenced by the CaO / Al2O3 or SiO2 / Al2O3 ratio. Quantitative comparisons using absolute peak intensities in Raman spectra for different samples are challenging. For example, unavoidable temperature fluctuations at high temperatures directly affect the background light in the Raman spectra and therefore the peak intensities. However, the energy from the background light affects the entire Raman spectrum uniformly. Therefore, examining the intensity ratios of characteristic peaks in Raman spectra for the same sample should provide a means of quantification. Completing the analysis of high-temperature Raman data involves the 520 cm -1 Al-O-Al at 640 cm -1 This involved collecting raw and deconvoluted Raman peaks representing Si-O-Si at 1000 sq m to establish relative Raman ratios that could be compared to the chemical composition ratios. The relative Raman peak ratios of the SiO2 / Al2O3 content were compared to the ratio of the SiO2 / Al2O3 content.

[0034] For both the raw peak data and the peak data extracted after deconvolution, the relative Raman peak intensity ratios were compared with the SiO2 / Al2O3 chemical content ratio. Sample S1 has the highest amount of SiO2 and the lowest amount of CaO. In contrast, sample S4 has the lowest amount of SiO2 and the highest amount of CaO. Sample S1 has the highest amount of SiO2, the highest SiO2 / Al2O3 content ratio, and the highest Raman peak intensity ratio. Sample S4 has the lowest amount of SiO2, the lowest SiO2 / Al2O3 content ratio, and the lowest Raman peak intensity ratio. In both data sets, the relative SiO2 / Al2O3 Raman peak intensity ratio is positively correlated with the SiO2 / Al2O3 chemical content ratio. Furthermore, these two data sets showed fitted R-squared values ​​greater than 92%. Next, the Raman peak intensity ratios were analyzed with the CaO / Al2O3 chemical content ratio. The S4 sample has the highest amount of CaO and the lowest amount of SiO2, which has the highest content ratio of CaO / Al2O3, but the lowest Raman peak intensity ratio of SiO2 / Al2O3 among all four flux samples.

[0035] The SiO2 / Al2O3 Raman relative peak intensity ratio of Example 4 increased with increasing SiO2 / Al2O3 content ratio. However, the CaO / Al2O3 content ratio decreased, indicating that the Raman spectra of silica and alumina in the flux samples were affected by the degree of polymerization of the silicate and calcium oxide networks. Furthermore, the fiber optic Raman system and deconvolution algorithm were reliable for further data analysis based on the high R-squared values. This data analysis indicated that correlations of material composition could be obtained using the deconvolution algorithm and the intensity ratios of characteristic Raman peaks. Studies on the viscous properties of mold flux materials can be further analyzed and correlated using the deconvolution algorithm.

[0036] Based on previous studies of mold flux, the Raman spectrum contains several overlapping peaks. Deconvolution analysis of the Raman spectrum is necessary to achieve a qualitative and quantitative description of the various structural units. The Raman curve is assumed to follow a Gaussian function and is fitted only in regions where a pronounced shoulder or peak is observed or rigorously demonstrated by previous studies. After deconvolution analysis, the Raman spectrum can reflect all possible structures, represented by peaks at various positions. Furthermore, because the SiO2 content in the melt flux samples is high relative to other components, silicates are thought to play an important role in the degree of polymerization of the melt. The silica mole fraction was further obtained by deconvolution of the Raman spectrum.

[0037] Regarding Example 4, the results of deconvolution of the Raman spectra by Gaussian fitting for each of the four samples in Example 4 are shown in Figures 11A to 11D. -1 In the Raman shift region, Q i The (Si) stretching band was deconvoluted into four typical peaks. These peaks were located at 850–895 cm, respectively. -1 (Q 0 ), 905~925cm -1 (Q 1 ), 945~985cm -1 (Q 2 ), and 1015–1100 cm -1 (Q 3 ) These peaks are located at [SiO4]Q 0 , [Si2O7]Q 1 , [SiO3]Q 2 , and [Si2O5]Q 3The identified peaks correspond to the stretching vibrations of . Details of the identified peaks are shown in Figures 11A-11D. A cumulative peak fit was generated by summing the four deconvoluted Raman peaks. The R-squared value was calculated by comparing the raw Raman data with the cumulative peak fit. The R-squared value of greater than 99.6% addresses any uncertainty in the deconvolution process. Compared to available literature, the flux sample at 1400 °C exhibited a high Q 0 The (monomer) peak shows a slight right shift, and Q 3 The (sheet) peak was found to exhibit a slight left-shift. This shift can be explained by the fact that the monomer and sheet content decreases slightly at higher temperatures compared to the results for rapidly cooled samples presented in the literature. This finding also demonstrates that for the cooling rates employed in this study, flux samples at high temperatures exhibit a significantly different Raman signal than the quenched flux samples. This further demonstrates that the intrinsic physical properties of the quenched flux samples change due to the cooling process. Therefore, it is of great interest to study the Raman signal to understand the property changes at high temperatures.

[0038] In mold fluxes in the CaO-Al2O3-SiO2 system, alumina is a conditional glass former (does not form glass without other network formers), and SiO2 is the primary network former at a low, fixed alumina content. In silica networks, four oxygen anions surround each Si cation center, and Si-O-Si bonds form all tetrahedral SiO4 4- Interconnecting complexes. NaO or CaO + or Ca 2+ In the presence of [AlO4] ions, some Si-O-Si bonds are broken. Therefore, when the melt has a high CaO / SiO2 ratio, as in sample 4, many network breakers in the melt are used to break the Si-O-Si bonds. However, these are [AlO4] 5-This indicates that at high CaO contents, the Q region of the Raman spectrum is most affected by the breaking of Si-O-Si bonds. 3 / Q 2 The ratio of Q to Q may be used as a polymerization index to quantify the effect of silicate structure on melt viscosity. 0 (monomer) was investigated as a factor affecting the properties of mold flux materials. Therefore, Q 3 / Q 2 and Q 3 / Q 0 The Raman peak coefficients of the flux samples were studied and correlated with CaO (wt%), SiO2 (wt%), and basicity values. The main components of the synthetic flux samples employed in this study are SiO2, CaO, and Al2O3. The Al2O3 content was fixed at 4.69% for all four samples. Therefore, the flux structure is strongly influenced by the CaO / Al2O3 or SiO2 / Al2O3 ratio.

[0039] Further referring to Example 4, as the CaO content decreases, Q 3 / Q 2 Ratio and Q 3 / Q 0 On the other hand, as the SiO2 content increases, the degree of polymerization of the silicate network increases, which is related to the Q 3 / Q 2 Ratio and Q 3 / Q 0 Finally, Q 3 / Q 2 Ratio and Q 3 / Q 0 The ratio correlates with the basicity ratio of the four mold flux compositions. 3 / Q 2 and Q 3 / Q 0 The scale factor of decreases linearly with increasing basicity ratio, and Q 3 / Q 0shows a linear correlation of 99.5%. Relative peak ratios based on Q-region peaks can be seen as a well-correlated basicity indicator, as large differences in intensity always exist. However, peak area ratios (mole fractions) are more robust than peak intensities and have been widely used in previous studies.

[0040] The combined area of ​​each deconvoluted peak is Q i It is a semi-quantitative evaluation of the number of units. As the SiO2 content decreases, Q 0 The content of (monomer) gradually increases, and Q 3 The content of (sheet) gradually decreases. To compare the correlation more effectively, Q 3 / Q 2 and Q 3 / Q 0 The peak areas of the Q region were also compared with the chemical composition and viscosity of the flux. Figures 12A and 12B show the relative peak area ratios of the Q region compared with SiO2 (wt%) and CaO (wt%), respectively. 3 / Q 2 and Q 3 / Q 0 The peak area ratio of Q is linearly and positively correlated with the SiO2 content. 3 / Q 2 and Q 3 / Q 0 The peak area ratio of Q also increased, and the fitting R-squared coefficient exceeded 96%. Conversely, the CaO content 3 / Q 2 and Q 3 / Q 0 The peak area ratio was negatively correlated with the Q value. 3 / Q 2 and Q 3 / Q 0 Finally, as shown in Figure 12C, the peak area ratio of Q 3 / Q 2 and Q 3 / Q 0 The peak area ratio of Q was compared with the basicity value of the sample. 3 / Q 2 and Q3 / Q 0 The peak area ratios of Q were negatively correlated with sample viscosity, with high R-squared values ​​of 94% and 95%, respectively. The available analytical data indicate that both peak intensity ratios and area ratios exhibit promising correlations with flux sample composition and properties. Furthermore, comparison of analytical results suggests that peak area ratios provided a higher correlation than peak intensity ratios for predicting flux chemical composition and viscosity. Previous researchers have reported that Q 3 / Q 2 It was pointed out that the area ratio can be used to estimate the viscosity of the flux, which is consistent with the results of this study. 3 / Q 0 It was also found that the area ratio appears to be an important parameter for the correlation between the properties of mold flux and the measured Raman spectra.

[0041] Example 5 After observing the relationship between Raman spectra and chemical composition of synthetic fluxes, we also employed in situ fiber-optic Raman spectroscopy on more complex commercial flux samples. Five samples with different chemical compositions were tested. Previous analysis showed that the Si-O-Si stretching band directly influences viscosity values. Therefore, we tested flux samples with reported viscosities of 2.8, 6.6, 1.5, 1, and 1.5 at 1300°C.

[0042] To clearly understand the effect of various temperature conditions on the Q-region Raman spectra of industrial flux samples, real-time Raman spectra were collected for sample A from room temperature to 1350 °C. The results are shown in Figure 13. As can be observed, these Raman spectra showed significant differences with temperature. -1 and 1125cm -1 The Raman spectral range between 500 and 700 cm -1 and 800-1100cm -1 It consists of two important areas:

[0043] In Example 5, 500 to 700 cm -1 In the Raman region of 540 cm -1 Peak at 640cm -1 The peaks at 520 cm correspond to the bending vibrations of the Al-O-Al bond and Si-O-Si, respectively. The alumina peaks of the industrial samples are higher than those of the synthetic flux samples due to the higher CaO / SiO ratio. -1 From 540cm -1 This appears to be due to the presence of more network breaks (non-bridging oxygen) in the Si-O-Si bonds, and at the same time, the [AlO4] 5- Promotes the formation of 500-700cm -1 The Raman spectrum of the alumina and silica regions at 1000 s is again stable at high temperatures without significant changes. The most significant change is observed at 800–1100 cm with increasing temperature. -1 A change in the Q region of the flux occurred between room temperature and 800°C, at which point the sample was completely solid and contained crystallites. Between 800°C and 1000°C, the flux sample remained solid, but there was no significant change in the Q region above the onset of crystallization. Between 1200°C and the melting point of the flux, a second significant change in the Q region appeared due to the scission of Si-O stretch bonds. At 1350°C, the sample became entirely liquid, and the Raman spectra showed additional changes in the Q region that were directly affected by increasing temperature. Monitoring these changes with temperature provides a way to track the viscosity-temperature dependence of industrial flux samples.

[0044] High-temperature Raman spectra of five industrial flux samples were successfully collected at 1350 °C. The five spectra show the Al-O-Al bond region, Si-O-Si bending vibration, and Q region (Q i) showed prominent Raman peaks, which is consistent with previous results from synthetic flux samples. Samples D and E contained viscosity values ​​very similar to that of sample C. Three industrial flux samples from samples A to C were selected to study the relationship between Raman spectra and chemical composition (shown in Figures 14A-14C). Four individual Q values ​​were found from the original Raman spectra. i The peaks were deconvoluted. R-squared values ​​of greater than 98.7% indicate a matching coefficient of the cumulative fit of the deconvolution process, providing additional confidence in the approach.

[0045] According to the deconvolution results shown in Figures 14A-14C of Example 5, there is no significant shift in the Q region of the spectrum. However, based on the study of the synthetic flux samples detailed in the previous section, there is a significant shift in Q 3 / Q 2 and Q 3 / Q 0 The correlation between Raman spectra and viscosity was studied by applying the Raman peak intensity ratio of 3 / Q 2 and Q 3 / Q 0 The fitting curves of Q showed a negative linear correlation with the viscosity of the industrial flux samples, which was consistent with the results of the synthetic flux. 3 / Q 2 The ratio showed a 97.1% correlation with viscosity.

[0046] Based on the data analysis results for the synthetic flux samples detailed in the previous section, the area fraction of each Raman peak in the Q region was calculated. As the viscosity of the industrial flux samples decreased, the Q 0 (monomer) and Q 1 The mole fraction of (dimer) decreases, and Q 3 (sheet) increases. This conclusion is consistent with previous findings for the control sample. Finally, as shown in Figure 15, Q 3 / Q 2 and Q 3 / Q 0 The area ratio of Q was compared with the viscosity of industrial flux samples.3 / Q 2 and Q 3 / Q 0 The curve function of was found to be negatively correlated with the viscosity of the industrial flux samples and was in agreement with the results from the synthetic flux samples. The accuracy of the linear correlation for both parameters was close to 90%.

[0047] Advantageously, embodiments of the present disclosure provide real-time mold flux analysis using an in situ fiber optic Raman sensor that can be performed directly in a high-temperature environment at 1400°C. Raman spectroscopy uniquely identifies specific molecules through high-resolution detection of vibrational bands, providing insight into molecular structure. Direct online processing of flux samples to determine chemical composition and other properties is enabled. The advantages of using fiber optic Raman spectroscopy at high temperatures to evaluate the structure and chemical composition of mold flux are demonstrated. Furthermore, the Raman spectroscopy results demonstrate that a range of chemicals can be successfully captured using an in situ fiber optic Raman sensor at 1400°C. The experimental results also demonstrate differences between the Raman spectrum of mold flux in the high-temperature molten state and that previously reported for quenched samples. Deconvolution of the high-temperature Raman spectrum reveals the Q 3 / Q 2 Ratio (sheet / chain) and Q 3 / Q 0 It was found that both the sheet / monomer ratio and the Raman peak area ratio could be used to evaluate the chemical composition and physical properties of mold flux. Furthermore, relationships between sample viscosity, chemical composition, and specific Raman peak area ratios of mold flux samples were also identified, showing promise for online flux analysis.

[0048] According to one embodiment of the present disclosure, a method for in situ real-time analysis of mold flux composition during continuous steel casting comprises the following steps. a) Introducing a portable fiber optic Raman sensor into the molten flux environment at a temperature of 1400° C. during the continuous casting process. b) performing high temperature Raman spectroscopy measurements on the melt flux using the fiber optic Raman sensor; c) detecting a composition-dependent Raman signal shift in a high temperature environment. d) Correlating the intensity of specific peaks in the Raman spectrum with the chemical composition, basicity, and viscosity of the molten flux. e) Providing real-time insight into the molecular structure and composition of the mold flux during continuous casting operations.

[0049] According to another embodiment of the present disclosure, a remote fiber optic Raman sensor system for real-time measurement of materials at high temperatures comprises: a) A high temperature resistant fiber optic probe head designed for in situ installation in a molten flux environment. b) Raman spectroscopy unit integrated into the fiber optic probe head, allowing recording of Raman spectra at 1400 °C. c) A high-temperature Raman system configured for online flux analysis and showing great promise for real-time evaluation of mold flux composition. d) Deconvolution algorithm applied to the original Raman spectrum for accurate analysis. e) Judiciously selected Raman peak intensities and areas used to establish ratios that correlate with chemical composition and physical material properties.

[0050] According to yet another embodiment of the present disclosure, a system for in situ chemical fingerprinting of molten flux composition in the steel industry using remote fiber optic Raman technology comprises: a) A high-temperature fiber optic Raman sensor system that can be remotely deployed within a molten flux environment. b) A real-time Raman signal analysis module configured to remotely control the molten flux composition through chemical fingerprinting. c) The ability to influence in-line slag and flux compositional analysis during continuous casting in the steel industry. d) The correlations observed between the structure and thermophysical properties of silicate melts (including basicity, viscosity, and density). e) A system demonstrating applicability to both synthetic and complex industrial flux samples.

[0051] When introducing elements of aspects of the invention or embodiments thereof, the articles "a," "an," "the," and "said" are intended to mean that there are one or more of the elements. The terms "comprising," "including," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements.

[0052] Not all depicted components shown or described may be required. Furthermore, some implementations and embodiments may include additional components. Variations in the arrangement and type of components may be made without departing from the spirit or scope of the claims defined herein. Additional, different, or fewer components may be provided, or components may be combined. Alternatively, or additionally, a component may be implemented by multiple components.

[0053] The foregoing description describes aspects of the present invention by way of example and not limitation. This specification describes several embodiments, adaptations, variations, alternatives, and uses of aspects of the present invention that will enable those skilled in the art to make and use the invention, including what is currently contemplated to be the best mode of carrying out the invention. Furthermore, it is to be understood that aspects of the present invention are not limited in their application to the details of construction and the arrangements of components set forth in the following description or illustrated in the drawings. Aspects of the present invention are capable of other embodiments and of being practiced or carried out in various ways. It is also to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting.

[0054] It will be apparent that modifications and variations are possible without departing from the scope of the invention as defined in the appended claims. Because various changes may be made in the structure and methods described above without departing from the scope of the invention, it is intended that all matter contained in the above description and shown in the accompanying drawings be interpreted as illustrative and not in a limiting sense.

[0055] In view of the above, it will be seen that the several advantages of the inventive aspects are achieved and other advantageous results attained.

[0056] The Abstract and Overview are provided to allow the reader to quickly grasp the nature of the present technical disclosure. They are submitted with the understanding that they will not be used to interpret or limit the scope or meaning of the claims. The Overview is provided to introduce some of the concepts further described in the Detailed Description in a simplified form. The Overview is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the claimed subject matter.

Claims

1. 1. A fiber optic sensor system for in situ high temperature Raman spectroscopy, comprising: an excitation light source; a spectrometer; a Raman probe configured for in situ placement in a molten flux environment, the Raman probe including at least one optical fiber connected to an excitation light source and a spectrometer, the at least one optical fiber transmitting laser excitation from the excitation light source to the sample and collecting light scattered by the sample for analysis by the spectrometer; an external lens arrangement positioned at a distal end of the at least one optical fiber of the Raman probe, the external lens arrangement optically connecting the at least one optical fiber to the sample and physically isolating the at least one optical fiber from the sample during sampling; The spectrometer is configured to perform Raman spectroscopy of the sample based on light collected by the at least one optical fiber.

2. a processor connected to the spectrometer; a memory storage device coupled to the processor; The fiber optic sensor system of claim 1 , wherein the memory storage device stores instructions that, when executed by the processor, configure the processor to: receiving a Raman spectrum of the sample from the spectrometer; - Deconvolving peaks from the Raman spectrum; and - Identifying the composition of a sample as a function of the deconvoluted peaks.

3. The fiber optic sensor system of claim 2 , wherein the memory storage device stores instructions that, when executed by the processor, configure the processor to: Selecting the intensities and / or areas of the Raman peaks to establish ratios that correlate to one or more of the chemical composition and physical material properties.

4. 4. The optical fiber sensor system of claim 2 or 3, wherein the memory storage device stores instructions that, when executed by the processor, configure the processor to: - Performing chemical fingerprinting of molten flux compositions.

5. The optical fiber sensor system of any one of claims 1 to 4, wherein the memory storage device stores instructions that, when executed by the processor, configure the processor to: • Identifying observed correlations between the structure of the silicate melt under examination and its thermophysical properties, including one or more of basicity, viscosity, and density.

6. The optical fiber sensor system according to any one of claims 1 to 5, further comprising an induction coil system for heating the sample.

7. further comprising a protective sleeve on the at least one optical fiber; The optical fiber sensor system of any one of claims 1 to 6, wherein the protective sleeve is configured to withstand high temperatures.

8. 8. The fiber optic sensor system of claim 7, wherein the protective sleeve is polyurethane.

9. further comprising an outer protective jacket on the Raman probe; The optical fiber sensor system of any preceding claim, wherein the outer protective jacket is configured to withstand high temperatures.

10. 10. The fiber optic sensor system of claim 9, wherein the outer protective jacket is stainless steel.

11. The optical fiber sensor system of any of claims 1 to 10, wherein at least one optical fiber of the Raman probe comprises an excitation fiber and a collection fiber.

12. The excitation light source generates a laser signal; The fiber optic sensor system of claim 11 , wherein the excitation fiber transmits a laser signal into the sample.

13. The optical fiber sensor system of any preceding claim, further comprising one or more filters for separating the laser excitation from the collected light scattered by the sample.

14. 1. A Raman probe for in situ high temperature Raman spectroscopy, comprising: a probe body containing an excitation optical fiber and a collection optical fiber, the excitation optical fiber being connected to an excitation light source and configured at its distal end to transmit laser excitation from the excitation light source to the sample, and the collection optical fiber being connected to a spectrometer and configured at its distal end to collect light scattered by the sample for analysis by the spectrometer; an external lens arrangement positioned at the distal ends of the excitation and collection optical fibers, the external lens arrangement optically connecting the excitation and collection optical fibers to the sample and physically isolating the excitation and collection optical fibers from the sample during sampling; A Raman probe comprising an outer protective jacket on the probe body and an external lens arrangement configured to withstand high temperatures.

15. further comprising protective sleeves for the excitation and collection optical fibers; The Raman probe of claim 14 , wherein the protective sleeve is configured to withstand high temperatures.

16. The Raman probe of claim 15 , wherein the protective sleeve is polyurethane.

17. The Raman probe according to any one of claims 14 to 16, wherein the outer protective jacket is stainless steel.

18. The excitation light source generates a laser signal; A Raman probe according to any one of claims 14 to 17, wherein the excitation optical fibre transmits the laser signal into the sample.

19. A Raman probe according to any one of claims 14 to 18, further comprising one or more filters housed within the probe body for separating the laser excitation from collected light scattered by the sample.

20. 1. A method for in situ and real-time analysis of high temperature Raman spectroscopy, comprising: introducing a portable fiber optic Raman sensor into a high temperature environment during a continuous casting process; performing high-temperature Raman spectroscopy on the molten flux in a high-temperature environment using the optical fiber Raman sensor to obtain a Raman spectrum of the molten flux; detecting a composition-dependent Raman signal shift in the Raman spectrum in the high temperature environment in response to the Raman spectroscopy measurement; correlating the intensity of a particular peak in the composition-dependent Raman signal shift with at least one property of the molten flux; and identifying a real-time change in at least one property of the molten flux based on the correlated intensities.

21. The Raman sensor comprises a probe body housing an excitation optical fiber and a collection optical fiber; an excitation optical fiber connected to the excitation light source and configured at a distal end thereof to transmit laser excitation from the excitation light source to the sample; 21. The method of claim 20, wherein the collection optical fiber is connected to a spectrometer and configured to collect light scattered by the sample at its distal end for analysis by the spectrometer.

22. 22. The method of claim 20 or 21, wherein the at least one property of the molten flux includes at least one of chemical composition, basicity, and viscosity.

23. The step of performing high temperature Raman spectroscopy on the molten flux comprises: receiving a Raman spectrum for the melt flux from the spectrometer; Deconvolving peaks from the Raman spectrum; and A method according to any one of claims 20 to 22, comprising identifying the composition of the molten flux as a function of the deconvoluted peaks.

24. 24. The method of any of claims 20 to 23, wherein the step of correlating the intensities of specific peaks comprises selecting the intensities and / or areas of Raman peaks to establish ratios that correlate with one or more of chemical composition and physical material properties.

25. The method of any of claims 20 to 24, further comprising performing chemical fingerprinting of the molten flux composition based on the correlation strength.

26. 26. The method of any of claims 20 to 25, further comprising identifying observed correlations between the structure of the molten flux and thermophysical properties comprising one or more of basicity, viscosity, and density based on the correlation strength.

Images