EP1430289A1

EP1430289A1 - Laser-absorption spectroscopy method and devices for carrying out said method

Info

Publication number: EP1430289A1
Application number: EP02774144A
Authority: EP
Inventors: Gerhard Totschnig; Franz Winter
Original assignee: Innovationsagentur GmbH
Current assignee: Innovationsagentur GmbH
Priority date: 2001-09-18
Filing date: 2002-09-18
Publication date: 2004-06-23
Also published as: WO2003025552A8; AT411298B; WO2003025552A1; ATA14682001A

Abstract

The invention relates to a laser-absorption spectroscopy method. According to said method, laser illumination (2) is directed through a sample volume (4) and the absorption of at least one target substance contained in the sample volume is measured by means of the transmission of said illumination through the sample volume. The measurement of the absorption is carried out over a short period of time compared to typical durations of significant adverse effects, such as fluctuations in the laser intensity and electronic or experimental noise, said adverse effects only causing a stationary, constant attenuation or amplification of the signal and a stationary, constant offset displacement (K).

Description

Process for laser absorption spectroscopy and devices for carrying out this process

The invention relates to a method for laser absorption spectroscopy, wherein laser radiation is passed through a sample volume and the absorption of at least one target substance contained in the sample volume is measured by measuring the transmission through the sample volume.

The invention further relates to devices for carrying out such a method, with a laser source for emitting laser radiation and with a detector for measuring the intensity of the laser radiation after passing through a sample volume.

The invention can be used with particular advantage for species-selective absorption measurement, and it enables the quantitative determination of concentrations of gaseous or liquid substances, in particular in mixtures with other substances. The absorption measurement according to the invention is particularly suitable for the detection of various substances in chemical reactors, for process control, for kinetic investigations and for determining concentrations, both at high pressures and temperatures and at low pressures and temperatures and with very high time resolution (μs) ,

In laser absorption spectroscopy, laser radiation is guided directly through a sample volume, and the concentration of this target substance is inferred from the absorption of the target substance at specific wavelengths. The basis for this is the well-known law of Lambert Beer. The concentration is determined taking into account other experimental parameters such as temperature, pressure and also the path length of the laser radiation in the sample volume. To achieve reference spectra, it is also common to measure the substances to be examined in the process in the laboratory under precisely known conditions or to calculate reference spectra using databases.

Problems with absorption measurement by laser spectroscopy are random fluctuations in the intensity of the laser radiation over time, electronic noise in the detection and signal processing components and experimental noise. examples for example, such experimental noise occurs in the case of combustion measurements by particles, such as ash, soot or sand, which fly through the beam path and partially block the laser radiation before it strikes the detector. This blockage appears on the detector as a disturbing additional absorption signal. Furthermore, experimental noise can be attributed to flickering light emissions from flames, which the detector receives as an interference signal, or also to flickering of the flame and the resulting (partial) deflection of the laser radiation from the active detector surface. Finally, a flame in combustion measurements leads to an undesired focusing of the laser radiation, which can lead to an undesirable increase in the intensity of the signal in the case of small detectors.

These factors thus prevent or limit the sensitivity of absorption measurements, whereby there are basically two groups of disturbing effects that cause noise:

1. Effects that lead to an amplification or attenuation of the signal, e.g. in the case of particles, focusing by gases with strong temperature gradients, e.g. a flame, deflection of the beam by flickering for gases with high temperature gradients, noise in the laser intensity or detector gain fluctuations;

2. Effects that lead to an additional signal, e.g. Illumination of a flame during combustion measurements or detector offset fluctuations.

In summary, the various disturbing effects that cause noise, some of which are mentioned above as examples, are also referred to as "fluctuations in laser intensity and electronic or experimental noise". Experimental and electronic noise is also characterized in that Comparison to the laser wavelengths used is independent of the wave, ie they have the same effect for all laser wavelengths used.

In order to eliminate these interference effects and to achieve a high detection sensitivity in laser absorption spectroscopy, a wide variety of techniques have already been proposed, such as in the articles Sonnenfroh et al. , "Ultrasensitive, visible tunable diode laser detection of N0 ₂ ", Applied Optics, Vol.35, No.21, July 20, 1996, pp. 4053-4058; and Pavone et al. , "Frequency and Wavelength Modulation Spectroscopies: Comparison of Experimental Methods using an AlGaAs Diode Laser", Appl.Phys. B56, 1993, pp.118-122. These proposals are based on the elimination of the disturbing noise in the useful signal with the aid of relatively complex experimental arrangements and circuits, with special modulation techniques or noise subtraction techniques being proposed, among other things.

It is an object of the invention to remedy this situation and to provide a technique for laser absorption spectroscopy in procedural as well as in device terms, with which highly sensitive absorption measurements are made possible in a comparatively simple manner.

To achieve this object, the invention provides a method as defined in claim 1. Furthermore, the invention provides devices for carrying out this method as specified in claims 14, 23, 25 and 26.

Advantageous embodiments and developments of the invention are specified in the subclaims.

The invention is based on the knowledge that the fluctuations in the laser intensity as well as electronic and experimental noise typically have a very pronounced dependence on the reciprocal of the frequency, i.e. that this noise, which limits the sensitivity, has a large proportion of low-frequency components (i.e. components below a few 10 kHz), but has only small high-frequency components (100 kHz to the megahertz range). The invention now makes use of this fact and is based on such a rapid measurement of the absorption that the fluctuations in the laser intensity and the electronic or experimental noise only have an effect as quasi-stationary, constant attenuation / amplification and offset shift and therefore do not affect the absorption measurement. Typical periods of time within which the absorption measurement is carried out according to the invention are of the order of magnitude of at most a few μs, preferably these periods are even less than one μs. With this rapid absorption measurement, the invention thus avoids the noise or the effect of noise and laser intensity fluctuations on the measurement by carrying out the measurement in a time period which is very short in comparison to the time. scale of the fastest disturbing fluctuations.

This is also a fundamental difference from the prior art, according to which the noise is filtered out of the measurement signal by complex, complicated arrangements and circuits. In contrast, the absorption measurement technology according to the invention is based on the fact that the noise is avoided by a quick measurement and does not appear in the measurement signal at all.

Various basic embodiments are possible for the rapid absorption measurement according to the invention. According to a first advantageous basic embodiment, a tunable laser is used which can be tuned sufficiently quickly (as mentioned, for example, on the order of a few μs or less than one μs) over a sufficiently wide range of the absorption spectrum (for example 0.2-1 cm ^{- 1} ). If the laser has a sufficiently wide tuning range, it is also possible to measure the transmission both at a wavelength at which there is greater absorption of the target substance and at a wavelength with weaker absorption of the substance to be examined. It is then possible, for example with knowledge of reference spectra, to determine the absorption of the substance to be measured, and with the inclusion of further experimental parameters, such as temperature, pressure and path length of the laser beam in the sample volume, as is known per se, the concentration of the target substance can subsequently be calculated.

According to another basic embodiment, it is advantageous if a laser that can be switched between different wavelengths is used, which is switched quickly between the wavelengths compared to the fluctuations in the laser intensity and noise (for example within a few μs or less than a μs), the wavelengths are in the range of absorption wavelengths of the target substance, which has a stronger absorption at one of these wavelengths and a weaker absorption at another. Switching the laser between the different wavelengths (typically between two wavelengths) is again faster than the fluctuation of the laser radiation and the noise. From the difference in the transmissions at the two wavelengths, the absorption can be determined again with the aid of reference spectra and relation to other experimental parameters, such as temperature, pressure and path length of the laser beam in the sample volume, as a result of which the concentration of the target substance is determined. It may also make sense to use more than two wavelengths, for example when a target substance is to be determined in a mixture with several absorbing substances with partially overlapping spectra, or when several substances are to be detected and quantified at the same time.

It should be mentioned here that a gas analysis technique is known per se from DE 37 41 026 A, in which a relationship is formed between two discrete absorption lines, the wavelength being switched over. However, this switchover takes place slowly, for example at 100 Hz or 2000 Hz, so that pressure and temperature fluctuations in a chimney of a power plant are not incorporated into the measurement result. How this should work concretely is not explained. In fact, this known technique with the specific switching between two wavelengths, namely between a wavelength in the absorption maximum and a wavelength in the absorption minimum, is only useful and feasible in very special cases, since the absorption of a fixed frequency does not only depend on the concentration of the gas sought, but also from the shape of the absorption line. Therefore, in order to measure the concentration of a gas with the known technique, the exact shape of the absorption line must be known, but this is usually not the case. The shape of the absorption line depends on the pressure, the gas composition and the temperature, with different gases, for example, causing different pressure widenings in the measuring range. If, for example, gas concentrations in the area of a gas nozzle are to be measured on a high-pressure burner, high-frequency spatially varying pressure fluctuations occur. The shape of the absorption line for this gas measurement results from the integral along the laser beam over the locally and temporally different absorption line shapes. With the technology described in DE 37 41 026 A, it is not possible to measure the concentration of a gas under these conditions. The same applies to the influence of the gas compositions on the absorption lines and thus on the absorption at a fixed wavelength, for example in the case a reactive gas flow, e.g. a flame: A large amount of water vapor is generated in such a flame. However, compared to normal air, water vapor is a very effective collision partner (it is typically five to ten times more effective than normal air), and thus the absorption of a gas at a fixed wavelength is up to five to ten times weaker if there is a lot of water vapor, since the Absorption line is less high, but is wider. If gas concentrations are to be measured in a flame, there are high-frequency and locally varying concentration fluctuations of the various gases in the flame, which in turn lead to a local and temporal variation of the absorption line shape. The shape of the absorption line for this gas measurement is again the integral along the laser beam over the absorption lines, which differ in location and time. The measurement of the concentration of a gas under such conditions is also not possible with the gas analysis technique according to DE 37 41 026 A. Similarly, the effect of temperature in general, inhomogeneous gas flows on the absorption line shapes and thus on the absorption at a fixed wavelength is similar. A wavelength change at 100 Hz or 2000 Hz, as addressed in DE 37 41 026 A, would be much too slow for the area in question, especially for measurement in gas streams heavily loaded with particles. The system described, with the CC laser, is also not suitable for switching between predetermined wavelengths more quickly. In contrast, the present technique, with measurement times in the order of one μs and below, results in a frequency of 1 MHz or more, and with this technique, the shape of the absorption line can be determined with each rapid measurement, the aforementioned inhomogeneous effects not interfering with the absorption measurement , It is therefore also possible to measure in inhomogeneous systems without any problems.

According to a third embodiment principle of the invention, it is advantageous if at least two lasers with different wavelengths are used and switched relatively quickly from one laser to the other (for example within a few μs or less than one μs), the one laser having a wavelength at which absorbs the target substance more strongly and the other laser has a wavelength at which the target substance is weaker absorbed. Here, too, the switching between the two lasers has to be carried out correspondingly quickly, in order to make fluctuations in the laser intensity and, as mentioned, noise appear quasi "stationary" as a result of signal fluctuations. The absorption measurement can advantageously be carried out here in two ways, namely using time-division multiplexing or wavelength division multiplexing technology for signal detection. Such TDM or WDM techniques are known per se, cf. for example Totschnig et al. , "Multiplex continuous-wave diode-laser cavity ringdown measurements of multiple species", Applied Optics, Vol.39, No.12, 2000, S.2009-2016. From the difference in the transmission at the two wavelengths, the absorption and finally the substance concentration can again be determined on the basis of reference spectra. Here, too, it can be expedient to use more than just two wavelengths, that is to say more than just two lasers, for example in the case of a plurality of absorbing target substances with partially overlapping spectra or in order to be able to detect and quantify a number of substances simultaneously.

In special measuring applications, for example in combustion measurement, where particles pass through the beam path, it is advantageous if all laser beams take exactly the same optical path through the sample and have the same dimension, and in order to make this possible, it is advantageous to measure the different laser beams in to couple a single optical fiber (single mode fiber). All rays coupled into the optical fiber pass through the sample volume after leaving the fiber, and they have identical optical paths therein. Thus, experimental disturbances have the same effect on both laser beams or wavelengths and therefore do not affect the measurement.

In order to achieve particularly precise measurements, it is advantageous if changes in the emitted laser power that occur during the absorption measurement process are detected by a characterization preceding the measurement and then taken into account in the measurement. In a corresponding manner, it is favorable if changes in the emitted laser power occurring during the absorption measurement process are taken into account simultaneously during the measurement by measuring a reference laser beam.

There are already lasers, e.g. VCSEL lasers (= Vertical Surface Emitting lasers) that - as studies have shown - can be tuned at the required high speed (tuning rates> 1 cnrVμs), which is typically not possible with other diode lasers. These VCSEL lasers now advantageously enable the use of the rapid absorption measurement technology according to the invention.

In addition, the VCSELs examined can also be tuned with the laser current over a very wide range (eg over a range> 8 cm ^-1 ). The line width of gases increases with increasing pressure (approximately 0.2 cπrVbar), and accordingly with conventional diode lasers that can only be tuned about 1 cm ' ¹ , the tuning range at pressures> 10 bar may be too small. The large tuning range with the laser current (for example in the case of a VCSEL laser) now advantageously allows the use of the inventive technique of rapid absorption measurements even with samples under high pressure, for example in a motor. Conversely, the invention also makes it possible to estimate the pressure or temperature in the sample volume by calculating back the pressure or temperature from the line broadening when the target substance is absorbed; this measurement is particularly useful for combustion processes where there are high pressures and high temperatures. The line width ΔvP due to the pressure broadening is given by ΔvP = P * ∑ xi * 2 * γi (where P is the pressure, xi is the mole fraction and γi is the pressure broadening coefficient) and the line width ΔvD due to the temperature T (Doppler broadening) is given by ΔvD = vO * 7.17 * 10 ^"7 (T / M) *, (where M is the molar mass), where the absorption line profile for gases, for example, is a Voigt profile which is determined by the two parameters ΔvP, ΔvD (see Alan C. Eckbreth," Laser Diagnostics for Combustion Temperature and Species ", Gordon and Breach Publishers, 1996).

A laser with an external resonator can also be used as the laser source, the external resonator with its components acting as a quickly tunable wavelength filter. The wavelength filter means that only a defined wavelength is amplified in the resonator and that the laser only emits light with this wavelength. The laser can be tuned quickly thanks to the fast-tuning wavelength filter. One way of realizing a wavelength filter that can be tuned quickly is to provide a "fast Phase modulator "located, wherein an optical path length modulator and a beam deflector are received in the beam path between a laser amplifier and a diffraction grating in the resonator, and separate control electrodes for applying an electrical high voltage, for changing the optical properties of the path length modulator and the beam deflector.

In the case of an optical path length modulator, the optical path length is changed, for example, by changing the refractive index in an electro-optical crystal when an electrical voltage is applied. The deflector changes the beam angle in the laser resonator when an electrical voltage is applied. Preferably, the path length modulator and the deflector are placed on a single crystal, such as on a crystal of LiNb03 and LiTa03. Such a combined crystal for path length modulator and deflector can be produced, for example, by polarization in an electrical field. The method of polarizing a crystal in an electric field is known per se, cf. eg: MM Fejer et al. "Quasi-Phase-Matched Second Harmony Generation: Tuning and Tolerances", IEEE Journal of Quantum Electronics, 28, 2631-2654 (1992). With such a wavelength filter, it is possible to achieve a tuning range for the laser of approximately 0.2 to 1 cm ^-1 (wave number) in a very short time, typically in less than 1 μs. In this way, mode jumps in the laser can also be avoided, as would otherwise occur when tuning through a wide range.

In addition to the options mentioned above for integrating the path length modulator together with the deflector on an electro-optical crystal, these components can of course also be carried out separately. An electro-optic crystal (electro-optic modulator) or a phase shifter integrated on the laser amplifier, a conventional laser chip, can be used as the path length modulator. An electro-optic crystal which has the shape of a prism can be used as the beam deflector, the refractive index of the prism changing when the control voltage is applied, and thus also the deflection angle, or an electro-optic crystal with polarity in an electric field or possibly also an acousto-optic deflector.

The laser with an external resonator can be in a Littrow configuration as well as in a Littman configuration. Another possibility for realizing a penetrable wavelength filter is the combination of an electro-optical path length modulator and an etalon made of electro-optical material, control electrodes for applying an electrical high voltage, for changing the optical properties of the etalon and the path length modulator.

If the absorption is very weak, it can be problematic that this absorption, with regard to the slope of the characteristic curve of the detector, voltage versus optical power, can still be resolved with sufficient accuracy with a data acquisition. This is especially the case when the slope is relatively steep. In this case, it is expedient if the slope in the characteristic curve is removed. For this purpose, a differentiator signal processing circuit can be used which generates a signal which corresponds to the second derivative of the detector signal. The resulting modified measurement signal has a horizontal profile, which can then be amplified without problems.

The invention is explained in more detail below on the basis of preferred exemplary embodiments, to which it is not restricted, however, and with reference to the drawing. The drawing shows in detail:

Fig.l schematically a device for laser absorption spectroscopy;

2 schematically shows a modified device for laser absorption spectroscopy;

3 shows another device for laser absorption spectroscopy in a schematic representation;

4 shows the laser power as a function of the laser current in a diagram;

5 shows an example of a detector transmission signal;

6 shows an example of a processed detector signal over the laser current;

7 shows an associated diagram of an absorption coefficient over the wave number, derived from the detector signal according to FIG. 6; 8a and 8b in comparison a detector signal as a function of the laser current in the case of a fault-free measurement (FIG. 7) and in the case of a measurement in the case of sand particles and a glowing flame obstructs the same laser beam (Fig. 8);

9 shows a diagram showing the detector signal as a function of time to illustrate an absorption measurement disturbed by sand particles;

10 shows an enlarged detail according to X in FIG. 9 from this diagram;

11 shows a diagram of the absorption over the wave number for two different substances, to illustrate a measurement of the absorption with several lasers; 12 and 13 show diagrams of the detector signal over the laser current in the event that absorption is difficult to detect compared to the baseline of the detector signal with a corresponding slope (FIG. 12), but with appropriate treatment of the detector signal when the linear slope is removed in the current-voltage characteristic, a clearly recognizable absorption signal is obtained, which can be easily amplified and measured (Fig. 13);

14 schematically shows a laser with an external resonator in a Littrow configuration, with a fast phase modulator in the resonator;

14a schematically shows a laser with an external resonator in a Littrow configuration, with a fast phase modulator in the resonator, the path length modulator and the deflector being separate units;

15 shows the electrical contacts on the fast phase modulator;

16 shows a corresponding representation of a laser in a Littman configuration, likewise with a fast phase modulator;

Fig.17 in the partial figures Fig.17a, 17b, 17c, 17d and 17e, in which the amplification over the wavelength of the laser is shown, the principle of operation of a laser with external resonator and fast phase modulator according to Fig. 14 or 14a or 16 at tuning; 18 and 19 in a schematic representation two exemplary embodiments for the generation of high-frequency high-voltage control signals for the fast phase modulator according to FIGS. 14 and 14a or 16;

20 schematically shows another example of a wavelength-tunable filter; and Fig. 21 shows the absorption spectrum (absorption over the wave number) for water vapor in air at different pressures.

FIG. 1 shows a type of block diagram of a device for laser absorption spectroscopy, which has a laser 1 that emits a light beam 2, which is possibly collimated via a lens 3 before it enters a sample or measurement volume 4 this runs through and strikes a detector 5, which measures the intensity of the incident light beam.

The laser source 1 contains, for example, a laser that can be quickly tuned over a sufficiently wide range, such as a VCSEL laser, which has, for example, a tuning range of more than 8 cm ⁻¹ , with a tuning speed of about 1 crrrVμs. A laser with a fast phase modulator can also be used in an external resonator. A control unit 6 is assigned to the laser 1 for this tuning. With regard to the mode of action of this rapid tuning and the associated rapid absorption measurement, reference is made to the following description with reference to FIGS. 4 to 10.

From a certain threshold value of the laser current, the detector signal which is received by the detector 5 has an essentially linear dependence on the laser current, with a more or less large slope, depending inter alia on the attenuation in the beam path, for example due to the sample volume 4. As below 12 and 13 will be explained in more detail, due to the given slope of the detector voltage / laser current characteristic, a weak absorption of a target substance can under certain circumstances only be insufficiently resolved with the data acquisition, but the data acquisition is facilitated if that Detector signal above the laser current would have no such slope. To remedy this situation, the fast absorption measurement can be combined with techniques for removing this slope and possibly for subsequently amplifying the signal. This can be an electronic circuit that results in a signal that corresponds to the first or second derivative (eg, a differentiator operational amplifier, described on page 224 by P. Horowitz and W. Hill "The Art of Electronics", Cambridge University. Press (1989)), and which amplifies the resulting signal; an electronic circuit is also conceivable that the inverted slope to Added measurement signal and amplified the resulting signal; or an electronic circuit that divides the measurement signal by the baseline and amplifies the resulting signal.

In order to remedy this situation, as an example, the detector 5 according to FIG. 1 is followed by a signal processing circuit 7 in which a differentiator signal processing circuit (d ² / dx ² ) 8 is used which generates a signal which corresponds to the second derivative of the detector signal , The resulting modified measurement signal, which is obtained at output 9, has a horizontal profile, which can then be amplified without problems. On the other hand, according to FIG. 1, a DC output signal is obtained at an output 10 in the usual way.

FIG. 2 shows a device for laser absorption spectroscopy according to FIG. 1 which is modified compared to FIG. 1, a laser 11, which can be switched between two wavelengths or frequencies, with an associated control unit 16 for switching over being provided, for example, which has a parallel laser beam 12 emits, which in turn runs through a sample volume 14 and strikes a detector 15 for measuring the laser intensity.

To avoid any fluctuations in the laser power emitted during the measurement, e.g. When switching the wavelength to be taken into account, a reference beam 12 ′ is derived from the laser beam 12 with the aid of a semitransparent mirror 17, which is arranged in the beam path of the laser beam 12 between the laser 11 and the sample volume 14. The reference beam 12 'is fed to a detector 15' which measures the intensity of the incident reference beam 12 'during the absorption measurement. The semitransparent mirror 17, which acts as a beam splitter, is designed such that the intensity of the measuring beam, which runs through the sample volume 14, and of the reference beam 12 'have a time-independent constant ratio. The intensity of the measuring beam before entering the sample volume 14 can thus be calculated from the measurement of the intensity of the reference beam 12 ′, and the transmission of the sample volume and the absorption in the sample volume 14 can be determined from the difference in the intensity of the measuring beam before and after the sample volume 14 be calculated.

However, an etalon or a reference (gas) cell indicated with a block 18 'can also be installed in the reference beam 12'. or several reference beams can be generated with additional beam splitters in order to use the direct intensity measurement, the etalon and the reference (gas) cell in one reference beam each. The etalon allows the wavelength change of the laser 11 to be tracked and calibrated (cf. RM Mihalcea et al. "Tunable diode-laser absorption measurements of NO2 near 670 and 395 nm", App. Opt. 35, p 4059-4064, 1996). In a reference (gas) cell, the target substance is typically filled in in an easily measurable concentration, and by measuring the transmission through the reference (gas) cell, it is possible to reliably feed the laser 11 around the absorption peak at the desired wavelength for a long time center. The width of the absorption peak can also be used to calibrate the wavelength tuning rate and to correct laser aging processes.

A comparable reference beam generation can of course also be present in the device according to FIG. 1 or in the device according to FIG. 3 to be explained in more detail below. On the other hand, a signal processing circuit 7 as shown in FIG. 1 can also be provided in the devices according to FIG. 2 or 3. Furthermore, a wavelength filter 19 for suppressing undesired light emission on the detector 5 or 15 etc. can be attached to all devices in the beam path, for example between the sample volume 14 and the detector 15. With this e.g. the disruptive flame emission can be suppressed. Another way of eliminating the detector offset by undesired light emission is to install diaphragms in the beam path in front of the detector, or to operate the laser once per measurement below the threshold or to switch it off and the detector signal measured in this way (generated only by the unwanted light emission, because at this moment the laser is not lit) is stored and then subtracted from the detector signal during the absorption measurement.

According to FIG. 3 there are two lasers 21, 21 'with different wavelengths λi,% 2, the laser beams 22, 22' of which are coupled into two single-mode optical fibers 23a, 23b, which in turn are coupled into a single-mode optical fiber 23, and so on ensure an identical optical path through the sample volume 24. Then the respective laser beam 22 or 22 'in turn on the detection system. The detection can again be carried out according to the known time division multiplex method with one detector 25a or according to the wavelength division multiplex method with several detectors 25b, 25c. In order to separate the laser beams with the two wavelengths λi, λ ₂ , a diffraction grating 28 or an interference filter 29 can also be used, which is illustrated schematically in FIG. 3 in association with the detectors 25b, 25c.

In detail, in the device according to FIG. 3, switching between the two lasers 21, 21 'takes place comparatively quickly (<1 μs) with the aid of a control unit 26, which either does this by directly controlling the lasers 21, 21' or but opens and blocks this downstream diaphragm 27, 27 '(or electro-optical or acousto-optical switch).

All three devices, according to FIGS. 1 to 3, have in common that the tuning or switching of the lasers for the purpose of changing the wavelengths of the laser radiation takes place comparatively quickly, in particular in the μs range or in times <1 μs, so that during this time the change in the wavelength fluctuations in the signal due to noise, for example due to fluctuations in the laser intensity, or electronic noise or experimental noise in the measuring arrangement, are negligible, since they can be regarded as stationary during the measurement and because the experimental noise is compared is independent of the wave to the laser wavelengths used. From the point of view of the rapid change in the wavelength of the laser radiation, these fluctuations thus appear to be "frozen", since they are generally at least two orders of magnitude slower.

For a more detailed explanation of two options for evaluating transmission measurement, reference is made to FIGS. 4 to 10. As a first option, broadband absorption is dealt with using FIGS. 4 to 7. _{4 shows the} detector voltage L ₀ to be regarded as a measure of the optical power of a diode laser as a function of the laser current I, as can be measured with the reference beam. This performance characteristic Lo (I) is used as the baseline for the absorption calculations. As can be seen, from a certain threshold value (1.2 mA) of the laser current, the laser power has an essentially linear dependence on the laser current. Similarly, the wavelength is tuned from a threshold value (1.2 mA) approximately linearly with the laser current. The baseline can therefore also be given as a function of the wavelength L ₀ (λ). In addition to the laser current, however, a diode laser can also be tuned in the wavelength via the laser temperature; usually the laser temperature is constantly regulated to one millikelvin using a Peltier element built into the optical laser amplifier. According to FIG. 5, the laser temperature is selected such that the absorption line S to be measured comes to lie approximately in the middle of the wavelength tuning range with the laser current.

If the laser can now be tuned very quickly over the absorption line compared to laser noise, detector-related noise and experimental noise, these fluctuations are quasi constant during the short measuring time, i.e. the detector signal L _D can be written as a function of the laser current L _D (I) = L ₀ (I) * TrZ (I) * RT + R0, where TrZ is the wavelength-dependent transmission of the target substance in the sample volume and RT or RO is the constant attenuation / amplification or amplification during the short measuring time Offset shift of the detector signal caused by the fluctuations in electronic and experimental noise.

FIG. 5 shows the case in which, in addition to absorption by the target substance in the sample volume, there is an additional weakening of the laser intensity (for example by sooting a flame) and an offset shift (for example by lighting up a flame). Since the offset caused by experimental noise is constant within a short measurement time (μs) and the laser does not light below the threshold value of 1.2 mA, the offset RO can be measured (detector signal for laser current <1.2 mA) and from the measurement signal L. _D (I) are subtracted. The offset-corrected measurement signal then also contains the constant attenuation due to experimental noise (RT) (eg due to soot, particles) during the short measurement time (μs). If one now divides the offset-corrected measurement signal by the baseline and then takes the natural logarithm and divides it by the path length W of the laser beam in the sample volume, the following results: ln ((detector signal offset) / baseline) / path length = = ln (L ₀ (I) * TrZ (I) * RT / Lo (I) / W = ln (TrZ (I) * RT) / W = ln (TrZ (I)) / W + ln (RT) / = Absorption coefficient -K, whereby according to the Lambert Beer law, the absorption coefficient of the target substance in the sample volume is defined as -ln (TrZ (I)) / W and K = -ln (RT) / W is an unknown constant -In ((detector signal offset) / baseline) / W up to a constant offset K the absorption coefficient.

The result of this evaluation is given in FIG. The signal in Fig.6 is

= -In ((detector signal offset) / baseline) / path length = = -In ((signal Fig. 5 - offset) / signal Fig. 4) / W. The size A is proportional to the concentration of the target substance and can be used to determine the concentration after suitable calibration. The quantity B, which corresponds to the absolute value of the absorption coefficient of the absorption peak S, and the offset K = -In (RT) / W, which results from the weakening of the laser intensity e.g. caused by experimental noise.

A reference spectrum of the target substance is given in FIG. 7, the pressure and the temperature corresponding to those of the sample volume. The ratio of the two quantities A '(difference between the absorption coefficients of the absorption peak S and the adjacent absorption valley T) and B' (absolute value of the absorption coefficient of the absorption peak S) has a fixed value for the given temperature and the given pressure. With the knowledge of a reference spectrum, therefore, the quantity B from FIG. 6 can be calculated as B = A * B '/ A', the unknown constant K = -In (RT) can be eliminated and an absorption spectrum can be calculated.

As already mentioned above, optical filters or diaphragms can also be used to suppress the flame immission on the detector.

Another possibility for isolated absorption lines is given in Fig.8a and 8b. The detector signal (detector voltage over the laser current) is shown in FIG. 8a if there are no faults. In comparison, the laser radiation in FIG. 8b is weakened by soot or particles, and the detector signal shows an additional offset due to the presence of a glowing flame in the sample volume 4 or 14 or 24 (FIGS. 1 to 3). Since these disturbing effects are constant during the short time in which the measurement takes place, they have an effect the detector signal only as a linear transformation in the sense of the relationship x → a * x + b. If, in the case of the device according to FIG. 1, the laser 1 is tuned from 0 mA upwards as in the representations of FIGS. 4 to 8, then the laser does not light below the threshold value, here at 2 mA, and one can thereby calculate the Determine the offset (value b in the linear transformation above) (signal at detector 5 if the laser current is <2 mA), and this offset value b can be subtracted from the signal. It can be seen from FIGS. 8a and 8b that no further absorption occurs neither at 3 mA nor at 9 mA. This makes it possible to adapt the baseline to the detector signal (detector voltage in volts), since the shape of the baseline is known from the reference measurement. So you can calculate the absorption and the absorption coefficient as usual:

Absorption ^ (1-offset_corrected_signal / baseline) absorption coefficient = -ln (offset_corrected_signal / baseline) / path length

9 and 10 show measurement results of an O2 absorption measurement as an example for the present technique of "fast" absorption measurement, a continuous stream of sand particles passing through the laser beam during the measurement and the laser beam passing through a strongly flickering blue liquid gas flame (propane / Butane) in the sample volume, the average transmission being 6%. The laser was tuned with a current ramp with a repetition rate of 40 kHz over several Ü2 lines. These 40 kHz ramps can just be seen in Fig. 9 as a dense vertical line pattern. Can be seen in Figure 9 at the height of the ramps that the sand and the flickering flame lead to strong ^{'fluctuations} in the transmission of the sample volume. The fastest fluctuations in Fig. 9 have a duration of about 200 μs.

An enlarged section X is shown in FIG. 10, whereby it can be seen that the laser passes through 5 absorption lines of O2 in a rapid sequence (40 kHz). In the example shown, the laser needs approximately 1 μs for tuning the wavelength over a single absorption line in the spectrum. In this short period of 1 μs, which is necessary for the measurement of an absorption line, the effect of the sand particles and the flame on the transmission signal changes not, there is a quasi-stationary attenuation and offset shift, and therefore the absorption measurement is not impaired, as has been explained with reference to FIGS. 7 and 8.

A device has already been explained above with reference to FIG. 3, in which a switch is quickly made between several, for example two, lasers 21, 21 'in order to carry out the desired absorption measurement. Of course, depending on the application, more than two lasers can be used for switching, and in Fig. 11 is a diagram showing the spectra of the methanol and isopropanol vapor, the usefulness of switching between three lasers or generally between three laser wavelengths ( illustrated with a device according to FIG. As shown in Fig.11, if three lasers A, B and C with wavelength ranges as shown are used in one system (device according to Fig.3), it is possible to use mixtures of methanol and isopropanol as vapor (both compounds have continuous spectra without discrete ones isolated lines and mixtures are therefore practically impossible to measure in a conventional manner) sensitive to measure with high repetition rates and thereby avoid the wavelength-independent experimental and electronic interference effects.

If several switchable lasers are used, a time or wavelength division multiplexing scheme is also suitable for signal detection, as is described, for example, in Totschnig et al. (mentioned above).

FIG. 12 shows a detector signal of a laser diode absorption measurement, the optical laser power increasing linearly with the current; accordingly, the A / D converter present in the detector circuit must be designed for a wide voltage range (0 to 5 volts, for example), regardless of the strength of the absorption. If the absorption is weak as shown in FIG. 12, the converter or the data acquisition card with a fixed bit resolution can no longer measure the absorption line 30 in FIG. 12 with sufficient accuracy.

If, as has already been explained in the device according to FIG. 1 with reference to the signal processing circuit 7, the linear slope is eliminated from the measurement signal, a horizontal measurement signal results as shown in FIG. 13. This means that the voltage range (eg 0 volts to 0.4 volts) Detector signal only depends on the strength of the absorption, ie the A / D converter can be designed for a much smaller voltage range (eg 0 volts to 0.4 volts) or vice versa, the measurement signal can be amplified to match the detector signal to the voltage range of the A / D converter. This makes it possible overall to measure even a very weak absorption line with the full resolution of the A / D converter, as shown in FIG. 13 at 30 '.

FIG. 14 shows a quickly tunable laser 1 in a Littrow configuration in a preferred embodiment, the path length modulator and the deflector being accommodated on a single crystal, such as on a crystal made of LiNb03 and LiTa03.

In this case, an optical amplifier 31 is provided which is designed to be highly reflective on one side 32, whereas the opposite side 33 is coated with an anti-reflective coating. The laser beam 33 emerges from this anti-reflecting side 33, which then passes through a lens 34 and then a path length modulator 35 with the length LP and a beam deflector 36 with the length LD before it reaches one for the Littrow configuration of a laser usual diffraction grating 37 arrives. The laser beam is partially reflected back from this diffraction grating 37, i.e. the laser resonator R is formed between the elements 32 and 37. The useful beam 38 is the beam reflected by the diffraction grating 37; the diffracted beam is identical to the incident beam.

Such a combined crystal for the path length modulator 35 and the deflector 36 can be produced, for example, by polarization in an electrical field. The method of polarizing a crystal in an electric field is known per se, e.g .: M.M. Fejer et al. "Quasi-Phase-Matched Second Harmony Generation: Tuning and Tolerances", IEEE Journal of Quantum Electronics, 28, 2631-2654 (1992).

As an alternative, FIG. 14a illustrates an embodiment in which the path length modulator 35 and the beam deflector 36 are provided as separate components - for example in a laser in a Littrow configuration with an optical amplifier 31, lens 34 and diffraction grating 37 as in FIG. 14 ,

In the Littrow configuration shown, the emission wavelength of the laser 1 is determined by an angle of incidence θ on the grid 37 determined. The equation here is: nλ = 2sin (θ) * d.

In this equation, n is the diffraction order and d is the distance between the grating lines. Usually n = 1 for a laser with an external resonator.

In order to continuously tune the wavelength of the laser 1 without mode jump, the optical path length in the laser resonator R is tuned synchronously with the laser wavelength and the number of standing waves N in the resonator R is kept constant (R = N * λ), as follows 17 will be explained in more detail with reference to FIG. For fast tuning, a high-voltage signal UD is applied to the electro-optical deflector 36 and at the same time a corresponding other high-voltage signal U is applied to the electro-optical path length modulator 35, as is also illustrated schematically in FIG. 14 in addition to FIG. 14, the path length modulator 35 in FIG. 15 and the deflector 36 is schematically illustrated with electrodes connected to corresponding high-voltage control circuits 35 'and 36', respectively.

The operation of the deflector 36 as in Fig. 14 can be explained from the fact that the electro-optical deflector crystal has two different types of domains, the refractive index of domains of one type, e.g. when applying a voltage to the beam deflector 36. increased and the refractive index for domains of the other type e.g. reduced or less enlarged. The arrangement and shape of the domains are chosen so that this refractive index difference leads to a beam deflection which depends on the voltage applied. The domain form has e.g. Prism shape, as illustrated schematically in Fig. 14 and Fig. 16.

The deflection angle φ through the deflector 36 is proportional to the voltage UD which is applied to the electro-optical deflector 36; the following applies: φ [rad] = n _e ³ * r33 * UD / H * LD / T; where r33 is the electro-optical coefficient and n _{e is} the extraordinary refractive index.

This beam deflection φ causes a change in the angle of incidence θ on the diffraction grating 37, ie Δθ = φ, and thus also a change in the emission wavelength, see the above equation for nλ. By applying the voltage U to the Weglängenmodulator 35, the refractive index of the path modulator to .DELTA.n = l / 2 changes * n _e ³ * r33 * UL / H.

This results in a change ΔLP in the optical path length LP in the path length modulator 35 by ΔLP = ΔnL * LP = l / 2 * n _e ³ * r33 * UL / H * LP and therefore a corresponding change in the optical path length in the entire resonator R.

In practice, this continuous mode jump-free tuning is achieved by applying a high voltage signal (going from 0 to UD) to the deflector 36 and a corresponding voltage 0 to UL to the unit 35 for optical path length modulation, so that R = N * λ is retained.

This high voltage can be generated by amplifying a sine wave, a linear ramp or another type of function of a (programmable) function generator or a D / A converter with a high voltage amplifier, cf. also Fig. 18 and 19.

The modified laser 1 shown in FIG. 16 has a so-called Littman configuration, in which, in a construction that is otherwise analogous to the configuration according to Littrow (FIG. 14), a tuning mirror 39 is additionally present after the diffraction grating 37, so that here the resonator is formed by the elements 32 to 39, the laser beam being diffracted at the diffraction grating 37 to the mirror 39 and reflected back from there. The useful beam is in turn the beam 38 reflected at the diffraction grating 37. In a modification of the relationships for the laser with external resonator in the Littrow configuration, the emission wavelength length of the laser in the Littman configuration is nλ = (sin (θ) + sin (ß )) * d;

The relationship given above applies to the beam deflection φ, however this beam deflection φ in the laser in the Littman configuration leads to a change in the emission wavelength according to the following relationship:

Δλ = cos (θ) * d * n _e ³ * r33 * UD / H * LD / T.

Otherwise, the statements made above in connection with the Littrow laser from FIG. 14 apply correspondingly to the laser in the Littman configuration according to FIG. 16. Furthermore, in the case of the Littman configuration of FIG. 16, the path length modulator 35 and the deflector 36 can also be used as separate components. be realized.

The tuning mechanism in such a laser 1 with an external resonator R according to FIG. 14 or 16 will now be explained with reference to FIG. 17 with the sub-figures 17a-17e. The gain of a diode laser chip 31 as a function of the wavelength λ is indicated in FIG. 17a. The external resonator of laser 1 or 1 'forms an etalon. Only wavelengths λ at which the resonator length R (R = N * λ) corresponds to an integer multiple N of the wavelength can oscillate and amplify, so to speak form a standing wave, cf. Fig.17b. One speaks of longitudinal modes. The distance between two adjacent longitudinal modes is called the free spectral range FSB, with FSB = c / 2R, where c = speed of light and R = resonator length. If the length of the external resonator R is changed, the frequency comb shown in FIG. 17b shifts in a first approximation along the wavelength axis. All other wavelengths are attenuated. 17c shows the gain characteristic over the wavelength λ, which arises from the combination of Fig.17a and Fig.17b. By selecting the diffraction grating 37, the angle of incidence θ on the diffraction grating and the mirror angle β (in the case of a Littman configuration), the position of the wavelength range (Littrow: nλ = 2sin (θ) * d; Littman: nλ = (sin (θ ) + sin (ß)) * d), which is effectively reflected back into the diode laser chip 31, can be determined. This is shown in Fig.l7d. Only these wavelengths are effectively reflected back by the diffraction grating and can amplify. All other wavelengths are suppressed. Fig. 17e shows the gain characteristic that results from the combination of Fig. 17c and Fig. 17d.

If, in the case of a laser 1 with an external resonator R, only the resonator length changes (and therefore the frequency comb in FIG. 17b shifts along the wavelength axis), but not the wavelength range that is effectively reflected back into the diode laser chip 31 (FIG. 17d) , then a continuous wavelength tuning of more than the free spectral range FSB is not possible, since the frequency comb (Fig.17b) and the effective range of the diffraction grating (Fig.l7d) do not tune synchronously. The laser jumps from the longitudinal mode N to an adjacent mode (eg Nl or N + l), which has a greater gain, ie there is a mode jump. It also happens a mode jump if only the wavelength range that is effectively reflected back into the diode laser chip 31 (FIG. 17d) is tuned, but not the resonator length R. Only if both the length of the external resonator R (and thus the frequency comb according to FIG. 17b ) as well as the wavelength range λ (see Fig. 17d), which is effectively reflected back into the diode laser chip 31 (Littrow: nλ = 2sin (θ) * d; Littman: nλ = (sin (θ) + sin (ß))) * d), are tuned synchronously (ie R = N * λ), continuous mode jump-free tuning ranges of more than one FSB value can be achieved. The tuning of the wavelength range (see Fig. 17d) is accomplished by tuning the angle of incidence θ on the diffraction grating and / or the mirror angle ß.

18 and 19 show two examples for the generation of the high-frequency high-voltage signal UL and UD for the path length modulator 35 and the beam deflector 36. According to FIG. 18, two function generators 41 and 41 'are provided which generate voltage signals which are amplified by high-voltage amplifiers 42 and 42'; the resulting signals U _L and U _D are applied to the path length modulator 35 and to the deflector 36. The opposite electrodes of these components 35, 36 are grounded.

FIG. 19 shows a comparatively inexpensive embodiment, in which the circuits 42, 42 'can be simpler high-voltage amplifiers which are installed in a resonant circuit, each with an inductor 43 or 43'. The resonant circuit leads to an increase (resonance) in the voltage at the path length modulator 35 or deflector 36 and minimizes the requirements for the high-voltage amplifiers 42, 42 '. The signals fed to the amplifiers 42, 42 'are in turn generated by a function generator 41 or 41'. The respective resonant circuit also includes the capacitance of the controlled component, namely the path length modulator 35 and the deflector 36, these components each having a capacitance of approximately 2 pF to 60 pF. The exact value of the capacitance depends on the exact dimensions and on the material of the path length modulator 35 or deflector 36.

With the configuration according to Fig. 19, it is easily possible to achieve control voltages from 0 to 900 volts at a repetition rate of up to 1 MHz. The achievable Their continuous tuning ranges are from 0.33 cm ^-1 to 0.66 cm ^"1 , depending on the exact configuration of the laser with an external resonator. When using LiTaθ3 or LiNbθ3 crystals for components 35, 36 this results r ₃₃ of approximately 30.8 * 10 ^"12 m / V and a refractive index of approximately n = 2.1, depending on the exact wavelength.

For optimal operation, the resonator should be kept as short as possible. This results in a large free spectral range (FSB) and increased stability against mode changes. On the other hand, the optical path length modulator 35 needs a certain length LP in order to obtain the desired continuous tuning range with UL _/ max = 900V.

A favorable optical length of the resonator (without an electro-optical phase shifter) is approximately 20 mm for a structure according to Littrow and 45 mm for a structure according to Littman.

The following are examples of a structure of an electro-optical phase modulator (with the components 35, 36). The laser is e.g. a 1490-1550 nm laser with an external resonator with a lens 34 with a focal length f = 3.1 mm and with a 1200 g / mm diffraction grating 37. The following values according to FIGS. 14 to 16 are still to be specified (labeling according to FIG. 14 -16):

- Littrow configuration: LP = 20 mm, LD = 2 mm, D = 4 mm, H = 0.5 mm, N = 10, B = 0.2 mm, θ = 66 °. Continuous tuning range for UL, max = 900 V: approx. 0.5 cm ^"1 .

- Littman configuration: LP = 23 mm, LD = 7 mm, D = 2 mm, H = 0.5 mm, N = 10, B = 0.7 mm, θ-85-87 ⁰ . Continuous tuning range for UL, max = 900V: approx. 0.36 cm ^"1 .

- Modified Littrow configuration: LP = 6 mm, LD = 10 mm, T = 1 mm, H = 0.5 mm, N = 10, B = 1 mm, θ = 66 °. Approximately 0.2 cm " ¹ continuous tuning range without mode jumps and 15 cm ^-1 discontinuous tuning range with mode jumps for UL, max, UD, max = 900V.

The modified Littrow configuration can be used very well for a device according to FIG. 2. Since the laser can be electro-optically tuned quickly (μs) over 15 cm ^-1 , several molecular species can be measured simultaneously.

A further example of a wavelength-tunable filter is shown in FIG. 20, wherein a crystal 45 made of an electro-optical material (see FIG. 17b), for example a LiNbθ3 crystal, is used. This crystal 45 is coated on its two end faces 46, 47 in a highly reflective manner and on its sides 48, 49 provided with electrodes to apply a high-frequency high voltage. The path length modulator 50 is coated on both sides 51, 52 with an anti-reflection coating. Two etalons are formed: one etalon is formed by the highly reflective side of the laser amplifier 31 and the side 46 of the crystal 45 facing the laser. The second etalon is formed between the anti-reflective sides 46 and 47 of the crystal 45. These two etalons each form a frequency comb (see FIG. 17b), the two frequency combs being able to be selected by the dimensions of the crystal 45 and of the entire resonator in such a way that only one wavelength is amplified. The laser beam is in turn emitted by an optical amplifier 31 with a highly reflective surface 32 and an anti-reflective coated exit surface 33, and it passes through the path length modulator 50 and the crystal 45, where it is reflected or where the useful beam is coupled out; the wavelengths that the etalon or crystal 45 transmits depend on the etalon length. Therefore, such an etalon filter 45 can tune the wavelength when a high voltage is applied to the crystal 45. Simultaneously, the path length modulator 50 can also be used to tune the resonator length.

It is also possible to combine the various components, namely the optical amplifier 31, the waveguide phase shifter, consisting of the components 35 and 36, and the diffraction grating 37 or the etalon 45 in one component, the lens 34 also being able to be omitted ,

Finally, FIG. 21 shows an application of the technique according to the invention for measuring at high pressures or for recording pressures via the absorption spectra, with the diagram according to FIG. 21 showing in detail the absorption over the wave number for water vapor (10%) in air at different total pressures, namely 1 bar, 10 bar, 20 bar and 40 bar. The absorption path length is 1 cm. As such, it is difficult to measure trace gases at high pressure. The line width (half width) of an absorption line at a pressure of 1 bar and at 296 K is typically 0.1 to 0.4 cm ^-1 . However, the line width increases proportionally with the pressure. Therefore, the line width for 40 bar is already around 4 cm ^"1 to 16 cm ^{" 1} (at 296 K). With a standard DFB diode laser, only about 1 cm ^{"1 can be} tuned, and the widened absorption lines can therefore only be measured inadequately. The use of a laser which can be rapidly tuned over a wide range (> 8 cm" ¹ ) enables the desired absorption measurements under these conditions; An example of a laser that meets these conditions is the aforementioned VCSEL laser.

The present laser absorption spectroscopy can advantageously be used in chemical reactors, for example in the field of organic chemistry, inorganic chemistry, for example in the production of iron or steel, and in biochemistry, as well as in reactors which are used for energy generation, such as in furnaces, Heating plants, motors. The process can be monitored very quickly at the reaction site. It is also conceivable to use the devices described in connection with process control or regulation circuits in order to directly control or regulate the running processes. Furthermore, an application for the determination of kinetic aspects in chemical reactions as well as for the control of certain species concentrations (eg 0 ₂ content in exhaust air) is conceivable. Likewise, the pressure prevailing in an environment or the existing temperature can be measured by measuring the line width, which is of particular interest for motor examinations and experiments in and around flames.

The technique described is suitable for gas species, e.g. O2, NO, HCN, NH ₃ , H ₂ O, OH, NH ₂ , CH ₄ , or organic molecules, in concentration ranges from ppb to% in gases and in with particles (dust , Soot, aerosols) loaded gases (e.g. in the ambient air, troposphere, stratosphere) and gas channels (e.g. from power plants, SNCR and SCR systems, chimneys, gas outlet channels) to measure, from low temperatures (-100 ° C) up to high temperatures (2000 ° C), from low pressures (1 mbar) to high pressures (300 bar). It is also possible to measure alkali vapors, especially sodium vapor, in situ. This is technically relevant since, among other things, low-melting alloys are formed in turbines. IR-active species (eg tryptophan) can also be measured in the liquid phase and in the liquid phase loaded with particles.

Claims

Claims:

1. A method for laser absorption spectroscopy, wherein laser radiation (2) is passed through a sample volume (4) and the absorption of at least one target substance contained in the sample volume is measured by measuring the transmission through the sample volume, characterized in that the measurement of the absorption in an im Compared to typical periods of significant, disturbing effects, such as fluctuations in laser intensity and electronic or experimental noise, a short period of time is carried out, these disturbing effects only for a steady, constant attenuation or amplification of the signal and for a steady, constant offset shift ( K) lead.

2. The method according to claim 1, characterized in that the wavelength of the laser radiation is changed by comparatively rapid tuning of the laser (1) over a predetermined wavelength range and the transmission is measured at different wavelengths, preferably both wavelengths with a stronger absorption and wavelengths, in which the target substance has a weaker absorption, are measured.

3. The method according to claim 2, characterized in that the laser (1) is tuned in a time period substantially in the order of magnitude of a maximum of a few μs.

4. The method according to claim 3, characterized in that the laser (1) is tuned in a period of less than 1 μs.

5. The method according to any one of claims 1 to 4, characterized in that a switchable between different wavelengths (λi, λ2> laser (11) is used, which is quickly switched between the wavelengths compared to the disruptive effects, and the transmission the different wavelengths are measured, the wavelengths preferably being in the range of absorption wavelengths of the target substance which have a high absorption at one of these wavelengths and at another has poor absorption.

6. The method according to any one of claims 1 to 5, characterized in that at least two lasers (21, 21 ') with different wavelengths (λi, \ ι) are used and compared to the disruptive effects quickly switched from one laser to the other where the transmission is measured at the different wavelengths and preferably one laser has a wavelength at which the target substance has a strong absorption, whereas the other laser has a wavelength at which the target substance has a weak absorption.

7. The method according to claim 6, characterized in that the detector signals to be traced back to the different lasers (21, 21 ') are processed in a time or wavelength multiplex method known per se.

8. The method according to any one of claims 1 to 7, characterized in that the or each laser radiation is coupled into an optical fiber (23) and fed to the sample volume (24) via this.

9. The method according to any one of claims 1 to 8, characterized in that changes occurring during the absorption measurement process and tuning the laser (1) in the emitted laser power are detected by a characterization preceding the measurement and then taken into account in the measurement.

10. The method according to any one of claims 1 to 9, characterized in that changes occurring in the emitted laser power during the measurement during the absorption measurement process and the tuning of the laser (11) are taken into account simultaneously by measuring a reference laser beam (12 ').

11. The method according to any one of claims 1 to 10, characterized in that based on the absorption measurement by back calculation from a broadening of the spectral line, pressure and / or temperature in the sample volume (4) are determined.

12. The method according to any one of claims 1 to 11, characterized in that the concentration of the target substance is subsequently calculated on the basis of the absorption measurement, including further experimental parameters, such as temperature, pressure and path length of the laser beam (2) in the sample volume (4) becomes.

13. The method according to any one of claims 1 to 12, characterized in that the absorption measurement at a high pressure, e.g. several bar to several 100 bar, is carried out in the sample volume (4).

14. Device for carrying out the method according to one of claims 1 to 13, with a laser source (1) for emitting laser radiation and with a detector (5) for measuring the intensity of the laser radiation after passing through a sample volume (4), characterized in that that the laser source (1) is a laser with an external resonator (R) and in the resonator a tunable wavelength filter (35, 36, 37, 39; 45, 35) is received in the beam path, with the help of which the wavelength of the single-mode laser ( 1) can be changed quickly.

15. The device according to claim 14, characterized in that the tunable wavelength filter (35, 36, 37, 39) in combination an optical path length modulator (35) and a beam deflector (36) in the beam path between a laser amplifier (31) and a diffraction grating (37 ), optionally also has a mirror (39).

16. The device according to claim 15, characterized in that the path length modulator (35) is formed by a separate electro-optical crystal (electro-optical modulator), electrical contacts being provided in order to apply the voltage to the refractive index of the electro-optical crystal and thus the optical path length of the To change the laser beam in the crystal.

17. The device according to claim 15, characterized in that the path length modulator (35) is formed by a phase shifter integrated on the laser amplifier (31).

18. Device according to one of claims 14 to 17, characterized in that the beam deflector (36) is formed by a separate electro-optical crystal, electrical contacts being provided in order to change the refractive index of the electro-optical crystal by applying a voltage and in combination with the shape of the electro-optical crystal, e.g. Prism shape to achieve a change in the refraction of the laser beam (2) and thus a change in the deflection angle of the laser beam.

19. Device according to one of claims 14 to 17, characterized in that the beam deflector (36) is formed by an electro-optical crystal which two different, e.g. has domain types which can be produced by the known method of "poling in an electric field", the refractive index for domains of one type changing differently when the voltage is applied to the beam deflector (36), and wherein the domains of the other type differ by this voltage-dependent difference in the refractive index and by the shape of the domains, for example Prism shape, the laser beam is refracted and deflected.

20. Device according to one of claims 1 to 14, characterized in that the beam deflector (36) is formed by an acousto-optical modulator.

21. Device according to one of claims 14 to 20, characterized in that the path length modulator (35) and the beam deflector (36) are integrated on a single electro-optical crystal, both parts (35, 36) having independent electrical contacts and separately are controllable.

22. Device according to one of claims 14 to 17, characterized in that the wavelength filter (45, 35) by an etalon made of electro-optical material (45) and a path length modulator (35) with corresponding control electrodes (48, 49) for applying a high voltage is formed.

23. Device for carrying out the method according to one of the Claims 1 to 13, with a laser source (1) for emitting laser radiation and with a detector for measuring the intensity of the laser radiation after passing through a sample volume, characterized in that the laser source (1) is formed by a VCSEL laser.

24. Device according to claim 23, characterized in that the VCSEL laser has a tuning rate of approximately 1 cm ^'1 /! has μs or greater.

25. Device for carrying out the method according to one of claims 1 to 13, with a laser source for emitting laser radiation (12) and with a detector (15) for measuring the intensity of the laser radiation after passing through a sample volume (14), characterized in that that the laser source is a laser (11) which can be switched between two wavelengths and whose wavelengths (λi, λ ₂ ) are in the range of absorption wavelengths of the target substance.

26. Device for carrying out the method according to one of claims 1 to 13, with a laser source for emitting laser radiation (22) and with a detector (25a, 25b, 25c) for measuring the intensity of the laser radiation after passing through a sample volume (24) , characterized in that at least two lasers (21, 21 ') are provided as laser sources, which are alternatively switchable.

27. Device according to one of claims 14 to 26, characterized in that the detector (5) is followed by a signal processing circuit (8) eliminating the slope in the current / voltage characteristic of the detector signal, e.g. a differentiator signal processing circuit that generates a signal corresponding to the second derivative of the detector signal.