Method for calibrating a device for raman spectroscopy with laser diodes

Abstract

The invention relates to a method for calibrating a measuring device (10) for Raman spectroscopy of a gas flow (66) with at least one laser diode (16) as an excitation source, comprising at least the following method steps: initially there is a flushing of the measuring device (10) with a non-Raman-active gas (22), in particular an atomic noble gas. A background signal spectrum (92) is recorded for predefined receiving times in the measuring of the gas flow (66), and stored. There is a measuring of the gas flow and a receiving of a raw measuring signal (90) at the predefined receiving times, and a subtraction of the background signal spectrum (92) corresponding with the respective receiving time from the raw measuring signal (90). An evaluable Raman spectrum (94) obtained in this way is stored and evaluated on the basis of a known gas concentration.

Description

[0001] Method for calibrating a device for Raman spectroscopy with laser diodes

[0002] Technical area

[0003] The invention relates to a method for calibrating a device for Raman spectroscopy using laser diodes as the excitation wave. Furthermore, the invention relates to a method for Raman signal analysis for quantitative gas concentration determination.

[0004] State of the art

[0005] DE 10 2021 107 229 concerns a simple and improved Raman signal analysis system for quantitative gas concentration determination. As can be seen from DE 10 2021 107 229 and DE 10 2009 026 744 A1, as well as EP 3 748 339 A2, high-power laser diodes, especially in multi-mode operation, are used as excitation light sources instead of the otherwise conventional solid-state or gas lasers. Raman spectroscopy of gases requires light sources or lasers with several watts of power. Until now, suitable solutions for Raman spectroscopy have only been available in the near-infrared range according to DE 10 2009 026 744 A1. Recently, however, multi-mode high-power laser diodes have come onto the market, which have been used in automotive headlights in the blue / VIS spectral range.

[0006] Laser diodes, by their very nature, exhibit spectrally broadband intrinsic fluorescence of up to 100 nm around the actual laser frequency. This means that the laser diodes emit light in the Raman spectral range typically evaluated, namely the Raman Stokes line range at relative wavenumbers up to approximately 4500 1 / cm. This diode fluorescence light is orders of magnitude stronger in intensity than the Raman signals. Raman spectroscopy using laser diodes is only possible if suitable additional filtering of the laser emission region and suitable laser beam or scattered light guidance are ensured, as described, for example, in DE 2021 107 229. These optical measures result in the actual Raman spectrum being influenced or underlain by a spectrally irregular background signal.A so-called fluorescence background arises here to a significant extent from the diode's own fluorescence and cannot be further reduced by optical measures. For highly accurate quantitative gas analysis, background subtraction is required to evaluate a spectrum. Raman spectroscopy with solid-state or gas lasers on gases does not normally have an irregular fluorescence-related background, but is not suitable for industrial use. The spectrally constant background due to dark and electronic noise can simply be subtracted. A spectrum calculated in this way can then be used for Raman analysis. A spectrally irregular background due to fluorescence from the actual sample to be examined is known from Raman spectroscopy of liquids, for example from A. Stratmann and G.Schweiger, Fluid Phase Equlibria of Binary Mixtures with Supercritical Solvents with in-situ Concentration Measurements by Raman Spectroscopy, textbook article: Chapter 1.4 in Supercritical Fluids as Solvents and Reaction Media (ISBN 0-444-51574-7), 2004, 85-120) or on solids (e.g. American Association for Aerosol Research (AAAR) Conference, Atlanta (US, Georgia), A. Stratmann and G. Schweiger, Fast Portable Black Carbon Analyzer Based on Raman Spectroscopy, 10 / 2004). For evaluation, a rather complex and generally only moderately automatable background subtraction is required in cases of unknown sample composition with many and / or possibly overlapping Raman signals.

[0007] Description of the invention

[0008] According to the invention, a method for calibrating a measuring device for Raman spectroscopy of a gas flow is proposed, with at least one laser diode as excitation source, wherein at least the following

[0009] The following process steps are carried out: a) purging the measuring device with a non-Raman active gas, in particular an atomic noble gas, b) recording and storing at least one background spectrum for predetermined recording times for measuring the gas flow, c) measuring the gas flow and recording a raw measurement signal according to the predetermined recording times and subtracting the background spectrum corresponding to the respective predetermined recording time from the raw measurement signal and d) storing the evaluable Raman spectrum determined according to c) and evaluating it based on a known gas concentration.

[0010] In an advantageous development of the method proposed according to the invention, at least one background spectrum is determined and stored according to b) for a short and a long recording period. The background spectrum is determined, for example, for the short and long recording periods or, alternatively, for other recording periods using suitable interpolation and extrapolation functions.

[0011] In the method proposed according to the invention, according to method step a), the measuring device is flushed with an atomic noble gas, such as helium, neon or argon as background calibration gas.

[0012] In the method proposed according to the invention, the calibration gas according to a) is passed through the measuring device in a pressure and temperature range within which the gas flow is measured according to c).

[0013] The method proposed according to the invention is further characterized in that the background spectra according to b) are recorded in a pressure and temperature range within which the gas flow is measured according to c).

[0014] In the method according to the invention, the measuring device records different background spectra when operating in different pressure and temperature ranges and at different pressures and different temperatures, whereby these are recorded and stored for each pressure and temperature range, or are recorded for selected pressure and temperature ranges and further background spectra are calculated for non-selected pressure and temperature ranges.

[0015] In a further development of the aforementioned embodiment of the method according to the invention, the additional background spectra are determined according to the ideal gas law, and a linear correction is performed. Alternatively, a real gas correction can be performed, for example, for gas mixtures as described in DIN EN ISO 14912.

[0016] Alternatively, a method for calibrating a measuring device for Raman spectroscopy of a gas flow with at least one laser diode as the output source is proposed, wherein at least the following method steps are carried out: a) generating a vacuum in the measuring device, b) determining an interference scattered light spectrum from components of the measuring device based on the diode's own fluorescence, c) measuring the gas flow and recording a raw measurement signal according to the specified recording times and subtracting the background spectrum corresponding to the respective specified recording time from the raw measurement signal and d) storing the evaluable Raman spectrum determined according to c) and evaluating it based on a known gas concentration.

[0017] Advantageously, according to the invention, the background, i.e. the background spectrum, is removed or subtracted from the actually recorded spectrum, i.e. which represents the raw spectrum or raw measurement signal, in order to obtain the actually evaluable Raman spectrum. To obtain this background spectrum, either the noble gas or vacuum method is used. For both, the method proposed according to the invention achieves a reduction in detector-based noise through detector calibration. For this purpose, an interference scattered light spectrum from components of the measuring device is determined based on the diode's own fluorescence. In a further development of the method proposed according to the invention, each receiver chip of signal detectors is calibrated with regard to its sensitivity, which is also achieved by determining an interference scattered light spectrum from components of the measuring device based on the diode's own fluorescence.

[0018] Furthermore, the invention relates to the use of the method for calibrating a measuring device for Raman spectroscopy, which comprises at least one laser diode as excitation source.

[0019] Advantages of the invention

[0020] The non-Raman signal background is advantageously determined by measuring non-Raman-active gases or, alternatively, by using a vacuum. Atomic gases used in background calibration, i.e., the noble gases helium, neon, and argon, do not generate a Raman signal. The interfering scattered light is caused by diode intrinsic fluorescence and by elastic, i.e., non-frequency-shifted Rayleigh scattered light components of the laser / laser diodes and their secondary frequency-shifted scattered light influences. If the spectrum of the noble gas is now measured and subtracted from the raw measurement signal, the actual Raman spectrum can be evaluated, in particular using the Voigt profile method, for example, and ultimately the gas concentration or the molecular particle concentration can be calculated by comparison with a known gas concentration measurement.Background calibration can be performed using an atomic noble gas, such as helium or argon, as the calibration gas. Argon is preferred because it is the most cost-effective and, due to its high molecular mass, very effectively removes interfering, Raman-signal-generating foreign gases, such as moisture, from the measuring device. The calibration gas used is preferably passed through the measuring device at the pressure and temperature range of the intended gas detection, and the corresponding spectrum is recorded.However, if the measuring system is used at different pressure and / or temperature ranges, resulting in different background calibration spectra, calibration spectra must either be recorded and saved for each pressure and temperature range, or, in a more advantageous embodiment, at specific pressure and temperature values, for example, at high and low pressure and temperature values, and then mathematically converted to the changes for the respective pressure and temperature ranges. For typical pressure and temperature measurement ranges, a linear correction will be sufficient due to the easily applicable ideal gas law. Alternatively, a real gas correction can be performed according to DIN EN ISO 14912.

[0021] Alternatively, a background calibration can be calibrated in a vacuum instead of using an atomic noble gas. This eliminates the need for a Raman spectrum, and stray light from the measuring device's components is detected using diode fluorescence. In certain applications, a vacuum can be applied more easily than an atomic noble gas. In contrast to the first embodiment using a noble gas calibration mentioned above, the stray light from the actual gas sample itself, which is partly dependent on density and thus pressure and temperature, cannot be detected with a vacuum calibration. However, this influence is not significant and can be neglected in certain applications, particularly with short measurement times and gas concentrations to be measured that are not too small.

[0022] A further advantage of the two background calibration methods proposed according to the invention, i.e., noble gas calibration and vacuum calibration, is the reduction of detector-based noise. The background calibration method proposed according to the invention includes any additional detector calibration that might otherwise be necessary. Each receiver chip of the signal detectors used, e.g., CCD / CMOS, has a slightly different sensitivity. Calibration adjusts this for all. In today's applications, a CCD or CMOS detector usually comprises a two-dimensional pixel array, e.g., 1024 x 128 pixels. The 1024 pixels correspond to the spectral wavelength axis, while the 128 pixels correspond to the height and represent each spectral point (1 - 1024). The 128 pixels add up to form a signal. The resulting 1024 signals, each of which consists of 128 added together, then form the spectrum.Each of the 1024 x 128 pixels has a slightly different light sensitivity, meaning that each pixel emits a slightly different signal at the same light intensity. This leads to increased noise and thus to a poorer signal-to-noise ratio during the actual Raman signal analysis. Using the method proposed by the invention—i.e., the noble gas and vacuum methods—the different pixel sensitivities can be automatically calibrated. This significantly reduces background noise. The analysis of a Raman signal depends significantly on the signal-to-noise ratio, so that with reduced noise, even the smallest signals can be detected and analyzed.

[0023] If the gas stream to be measured condenses due to thermodynamic boundary conditions or gas components condense from it, particularly in humid conditions, the correct gas concentration will no longer be measured for the condensed gas components, i.e. the gas concentration will be too low. For example, condensation is detected by the presence of an increased background measurement signal. An increased background measurement signal arises because the condensate adhering to surfaces, such as optical windows or measuring cell surfaces near the optical measuring range, generates interfering scattered light. The background monitoring / calibration proposed by the invention automatically monitors this as well and thus prevents incorrect evaluations and / or misuse due to the formation of condensate. However, the condensate must precipitate out within the optical measuring range.This can be achieved by designing the measuring device by integrating a cold trap in the optical measuring area so that any condensate that occurs falls here and not at any arbitrary point, ie the coldest point, in the gas flow pipe.

[0024] The cold trap can, for example, be a type of cooling rod that is thermally decoupled from the measuring cell but not from the gas stream, and that is cooled to a slightly lower temperature, e.g., a few degrees Kelvin lower than the gas temperature in the gas stream. One possible design is, for example, a pelletizing cooling element controlled by a gas temperature sensor.

[0025] Reactive gases, such as those with elevated levels of ammonia, sulfur compounds, acids, and the like, can attack the surfaces of the measuring cell of the measuring device in the gas path with optical windows and also generate the aforementioned background measurement signal. Defined background signal monitoring or calibration thus also safeguards against defects in the measuring unit.

[0026] Short description of the drawings

[0027] Embodiments of the invention are explained in more detail with reference to the drawings and the following description.

[0028] They show:

[0029] Figure 1 is a schematic representation of a measuring device for measuring a gas flow using Raman spectroscopy,

[0030] Figures 2.1 and 2.2 show spectra of air enriched in COj,

[0031] Figure 3 shows an evaluable Raman spectrum of the raw measurement signal,

[0032] Figure 4 a background spectrum,

[0033] Figure 5.1 an input signal with calibration values ​​and

[0034] Figure 5.2 A Raman spectrum with calibration

[0035] Embodiments of the invention

[0036] In the following description of the embodiments of the invention, identical or similar elements are designated by the same reference numerals, whereby a repeated description of these elements is omitted in individual cases. The figures only schematically illustrate the subject matter of the invention.

[0037] Figure 1 schematically shows the structure of a measuring device 10 proposed according to the invention. This device comprises a high-performance sensor 14 as radiation source 12, which in particular comprises at least one laser diode 16 as an excitation source. This sensor is operated within the visible spectral range, in particular within the blue spectral range. Following the radiation source 12 as shown in Figure 1 is a focusing optic 18, which is only indicated schematically here, which focuses the laser radiation generated by the at least one laser diode 16 onto a portion of a gas measurement chamber 20. The gas measurement chamber 20 contains a gas 22 or a gas mixture 24. The gas measurement chamber 20 can be part of a bypass line 26 through which a gas stream 66, for example gaseous H2, flows.

[0038] The gas measurement chamber 20 comprises at least one optical access 28 and at least one optical output 30 for the laser radiation generated by the at least one laser diode 16. Laser radiation exiting the at least one optical output 30 is guided into a radiation absorber 32 to prevent stray light influences and, after passing through the gas measurement chamber 20, is coupled into a further gas measurement chamber 56. The further gas measurement chamber 56, which serves as a calibration cell 58, contains a known gas concentration, for example, 100% N2. From the further gas measurement chamber 56, the laser radiation then enters the radiation absorber 32 to prevent interference, as explained above.

[0039] A parallel measurement 64 can be performed with the measuring device 10 shown in Figure 1. For this purpose, Raman scattered light 34 from the gas measurement chamber 20, on the one hand, and Raman scattered light 34 from the additional gas measurement chamber 56, which serves as a calibration cell 58, on the other hand, are fed in parallel to the Raman scattering intensity-enhancing optics 36. Thus, the Raman scattering intensity-enhancing optics 36 contains two scattered light components that can be measured parallel to one another within the Raman scattering intensity-enhancing optics 36. The Raman scattered light 62 from the additional gas measurement chamber 56, which serves as a calibration cell 58, is additionally guided into the Raman scattering intensity-enhancing optics 36 by means of the parallel measurement 64 and reaches the light detector 48 in the form of a CCD camera or a CMOS component or a number of receiver diodes.Depending on the application, for example, a parallel measurement 64 of a gas 22 not present in the gas stream 66 can be carried out directly. Thus, in addition to laser power fluctuations, signal fluctuations of the light detector 48 are also detected. From the measurement of the Raman scattered light 62 from the additional gas measurement chamber 56, the Raman scattered light measurements can then be recalibrated accordingly as part of the evaluation of the Raman scattered light 34 from the gas measurement chamber 20. With the structure of the measuring device 10 according to Figure 1, a background signal calibration can be carried out. Furthermore, the same gases can be kept in the gas measurement chambers 20, 56, and a separation 68 of the Raman scattered light components originating from the gas measurement chamber 20 and the additional gas measurement chamber 56 can be carried out within the scope of the Raman scattering intensity-enhancing optics 36.These can be detected in parallel within the framework of the parallel detection 64 in the light detector 48, which can be designed as a CCD camera, for example.

[0040] The illustration in Figure 1 further shows that the measuring device 10 contains a cold trap in the form of a cooling rod 74 within the optical measuring range, i.e., within the gas measurement chamber 20. The condensate therefore precipitates here and not at any arbitrary point in the flow path of the gas stream 66. In the illustration in Figure 1, the cold trap is designed as a cooling rod 74, which can be designed as a metal pin that is thermally decoupled from the measuring cell but not from the gas stream 66, and which is cooled to a slightly lower temperature, for example, 1 to 2 Kelvin lower than the gas temperature of the gas stream 66 flowing in the flow path. For example, the cooling rod 74 can be designed as a pelletizing element that can be controlled via the gas temperature sensor.

[0041] Figure 2.1 shows a CO2 and air spectrum, and Figure 2.2 shows a corrected spectrum.

[0042] The representation according to Figure 2.1 shows a CO2 and air spectrum 78 with a CO2 peak 80 and an FbO peak 86. Furthermore, the CO2 and air spectrum 78 includes an O2 peak 82 and a pronounced IXh peak 84.

[0043] The CO2 and air spectrum 78 thus represents exhaled air enriched with CO2. In contrast, Figure 2.2 shows a corrected spectrum 88. The correction in the corrected spectrum 88 according to Figure 2.2 is made in such a way that the corrected spectrum 88 contains the same spectrum measurement, taking into account the removal of ambient air. In Figure 2.2, the Nj peak 84 and the Os peak 82 for air are thus eliminated. At the corresponding points in the corrected spectrum 88, as shown in Figure 2.2, harmonics and subharmonics remain, which, however, compensate for each other within the framework of an integral evaluation.

[0044] Figure 3 shows an evaluable Raman spectrum 94, which was calculated by subtracting a previously recorded helium spectrum from the raw measurement signal. It is flushed with an atomic noble gas, whereby a background signal spectrum 92 is recorded and saved for all recording times specified within the gas detection. A background signal spectrum is fundamentally composed of a bias and dark level of the detector pixels (including the influences of the evaluation electronics). Both calibration methods, i.e., calibration using a noble gas and calibration using a vacuum, eliminate these influences. When measuring the gas flow 66 in the measuring device 10, the noble gas spectra corresponding to the specified recording times, i.e., the background signal spectrum 92, are then subtracted from a raw gas measurement signal 90, so that a calculated, evaluable Raman spectrum 94 is obtained.This is saved and can then be evaluated accordingly to determine the gas concentration. Alternatively, it is possible to determine the background spectra 92 not for all gas detection recording times, but only for a few specific recording times. The intermediate gas detection points are calculated mathematically, whereby a linear relationship can be assumed for simplification.

[0045] In the calibration method according to the invention, atomic noble gases, such as helium or argon, are typically used as calibration gases. Argon is preferably used for cost reasons and because, due to its high molecular mass, it very effectively removes interfering, Raman-signal-generating foreign gases, such as moisture or water, from the measuring device 10. Preferably, the respective calibration gas is passed through the measuring device 10 in the gas stream 66 in the pressure and temperature range of the actually intended gas detection, and corresponding spectra are recorded.If the measuring device 10 is used in different pressure and / or temperature ranges / conditions, so that the background signal spectra 92 differ, corresponding background signal spectra 92 must either be recorded and stored for each pressure and temperature range or, in a preferred embodiment, recorded at specific pressure and temperature values, for example, at high and low pressure and temperature values, and then mathematically converted to the change for the respective pressure and temperature range. In typical, normal pressure and temperature measurement ranges, such as atmospheric pressure and room temperature, a linear correction will be sufficient due to the easily applicable ideal gas law.

[0046] As an alternative to purging the measuring device 10 with an atomic noble gas, it can also be calibrated in a vacuum. With this background calibration method, no Raman spectrum or stray light from the components of the device is generated due to the diode fluorescence of the at least one laser diode 16 used as the excitation source. In contrast to calibration with an atomic noble gas, the stray light from the actual gas sample itself, which is partly dependent on density and thus pressure and temperature, cannot be determined. This influence is negligible in certain applications, for example, with short measurement times and gas concentrations to be measured in the gas stream 66 that are not too small.

[0047] A further advantage of the invention is that with the two background calibration methods presented according to the invention, i.e., noble gas-based and vacuum-based, a reduction in detector-based noise can be achieved, thus eliminating the otherwise required detector calibration. Each receiver chip of the signal detectors used (CCD, CMOS) has a slightly different sensitivity. Calibration adjusts the sensitivity for all receiver chips of the signal detectors used. This allows a significant reduction in background noise. A reduction in background noise, in turn, enables the evaluation of very small Raman signals, which can now be detected more effectively due to the significantly reduced signal-to-noise ratio, and therefore even these small signals can be reliably evaluated.

[0048] Figure 4 shows a background signal spectrum 92 comprising a bias level 100 and a dark level 102. The fluorescence range is represented by reference numeral 104. According to Figure 1, the Raman signals are designated by reference numeral 106, and their intensity 96 is plotted against the number of pixels 98. The graphs according to Figures 5.1 and 5.2 show that, ignoring the fluorescence 104, calibration values ​​112 are assigned to the input signals, i.e., the raw measurement signal 90. This results in an output signal 110, shown in Figure 5.2, which only comprises the Raman spectrum, i.e., the signals 106 and the fluorescence range 104.

[0049] For the evaluation of the Raman spectra, for example using the Voigt method, a Raman spectrum is essential, which requires a correct elimination of the background signal spectrum 92 or the subtraction of the background signal spectrum 92.

[0050] The invention is not limited to the embodiments described here and the aspects highlighted therein. Rather, numerous modifications are possible within the scope of the claims, which are within the scope of one skilled in the art.

Claims

Claims 1. Method for calibrating a measuring device (10) for Raman spectroscopy of a gas stream (66) with at least one laser diode (16) as excitation source, comprising at least the following steps: a) purging the measuring device (10) with a non-Raman-active gas (22), in particular an atomic noble gas, b) recording and storing at least one background signal spectrum (92) for predetermined recording times for measuring the gas stream (66), c) performing the measurement of the gas stream (66) and recording a raw measurement signal (90) according to the predetermined recording times and subtracting the background signal spectrum (92) corresponding to the respective predetermined recording time from the raw measurement signal (90), d) storing the evaluable Raman spectrum (94) determined according to c) and evaluating it based on a known gas concentration.

2. Method according to claim 1, characterized in that according to b) the at least one background signal spectrum (92) is determined and stored for a short and for a long recording time, and a complete gas detection is calculated under the assumption of a linear relationship.

3. Method according to claim 1, characterized in that, according to a), an atomic noble gas, such as helium, neon or argon, is used as a calibration gas for purging.

4. Method according to claim 3, characterized in that the calibration gas according to a) is passed through the measuring device (10) in a pressure and temperature range in which the measurement of the gas flow (66) takes place according to c). A method according to claim 3, characterized in that the background signal spectra (92) are recorded according to b) within a pressure and temperature range within which the measurement of the gas flow (66) is carried out according to c). A method according to claim 1, characterized in that the measuring device (10) records different background signal spectra (92) when operating in different pressure and temperature ranges and at different pressures and temperatures, records and stores these for each pressure and temperature range, or records them for selected pressure and temperature ranges and calculates further background signal spectra (92) for unselected pressure and temperature ranges. A method according to claim 6, characterized in that the further background signal spectra (92) are determined according to the ideal gas law and a linear correction is applied.Method for calibrating a measuring device (10) for Raman spectroscopy of a gas stream (66) with at least one laser diode (16) as excitation source and at least the following process steps: a) generating a vacuum in the measuring device (10), b) determining a stray light spectrum from components of the measuring device (10) due to diode fluorescence, c) measuring the gas stream (66) and recording a raw measurement signal (90) according to the specified recording times and subtracting the background signal spectrum (92) corresponding to the respective specified recording time from the raw measurement signal (90), d) storing the evaluable Raman spectrum (94) determined according to c) and evaluating it on the basis of a known gas concentration.

9. Method according to claims 1 to 7 or 8, characterized in that a reduction of detector-based noise is achieved by detector calibration.

10. The method according to claim 9, characterized in that each of the receiver chips of signal detectors is calibrated with respect to its sensitivity.

11. Use of the method according to any one of claims 1 to 10 for Calibration of a measuring device (10) for Raman spectroscopy comprising at least one laser diode (16) as an excitation source.

Images