Method for analyzing a gas by mass spectrometry, and mass spectrometer

Abstract

A method for analyzing a gas by mass spectrometry includes exciting ions of the gas to be analyzed in an FT ion trap, and recording a first frequency spectrum in a first measurement time interval during or after the excitation of the ions. The first frequency spectrum contains ion frequencies of the excited ions and interference frequencies. The method also includes recording a second frequency spectrum in a second measurement time interval. The second frequency spectrum contains the interference frequencies, but not the ion frequencies of the first frequency spectrum. The method further includes comparing the first frequency spectrum with the second frequency spectrum to identify the interference frequencies in the first frequency spectrum. The disclosure also relates to a mass spectrometer which is suitable for carrying out the method for analyzing the gas by mass spectrometry.

Claims

1. A method, comprising: exciting ions of a gas in an FT ion trap; recording a first frequency spectrum in a first measuring time interval during or after the excitation of the ions, the first frequency spectrum comprising ion frequencies of the excited ions and interference frequencies; recording a second frequency spectrum in a second measuring time interval, the second frequency spectrum comprising the interference frequencies of the first frequency spectrum but not the ion frequencies of the first frequency spectrum; removing the excited ions from the FT ion trap at a time selected from the group consisting of the beginning of the second measuring time interval and before the second measuring time interval; and comparing the first and second frequency spectra to identify the interference frequencies of the first frequency spectrum.

2. The method of claim 1, wherein the excited ions are removed from the FT ion trap at the end of the first measuring time interval.

3. The method of claim 1, wherein, before the second measuring time interval, the ions are excited with a degree of excitation of at least 100%.

4. The method of claim 1, wherein, before or in the first measuring time interval, the ions are excited with a degree of excitation of at least 100%.

5. The method of claim 1, further comprising, during the removal of the excited ions from the FT ion trap, determining an overall amount of ions of the removed excited ions.

6. The method of claim 5, further comprising assigning the ion frequencies in the first frequency spectrum an individual amount of ions on the basis of the determined overall amount of ions.

7. The method of claim 1, wherein the ions are excited before or in the first measuring time interval with a degree of excitation of less than 100%.

8. The method of claim 1, wherein the first measuring time interval and the second measuring time interval follow one another with a time difference of less than 10 ms.

9. The method of claim 1, wherein one of the following holds: a time period from the beginning of the first measuring time interval to the end of the second measuring time interval is less than 500 ms; and a time period from the beginning of the second measuring time interval to the end of the first measuring time interval is less than 500 ms.

10. The method of claim 1, wherein IFT excitation is used to selectively excite the ions depending on a mass-to-charge ratio of the ions.

11. The method of claim 1, further comprising identifying as interference frequencies of the first frequency spectrum: frequencies contained in the first frequency spectrum that lie in a frequency range in which no excitation of ions takes place; or an excitation of ions with a degree of excitation of more than 100% takes place.

12. A method, comprising: exciting ions of a gas in an FT ion trap; recording a first frequency spectrum comprising: a) peaks corresponding to ion frequencies for ion signals generated by the excited ions; and b) peaks corresponding to interference frequencies; recording a second frequency spectrum comprising peaks corresponding to the interference frequencies of the first frequency spectrum but not peaks corresponding to ion frequencies for ion signals generated by the excited ions; and comparing the first and second frequency spectra to identify the interference frequencies of the first frequency spectrum.

13. A mass spectrometer, comprising: an FT ion trap; a detector configured to record: a first frequency spectrum in a first measuring time interval during or after excitation of ions in the ion trap, the first frequency spectrum comprising peaks corresponding to ion frequencies for ion signals generated by the excited ions and peaks corresponding to interference frequencies; and a second frequency spectrum in a second measuring time interval, the second frequency spectrum comprising the peaks corresponding to the interference frequencies of the first frequency spectrum but not peaks corresponding to ion frequencies for ion signals generated by the excited ions, wherein the detector is configured to identify the interference frequencies of the first frequency spectrum by comparing the first and second frequency spectra.

14. The mass spectrometer of claim 13, wherein excited ions are removed from the FT ion trap at the beginning of the second measuring time interval or before the second measuring time interval.

15. The mass spectrometer of claim 14, wherein excited ions are removed from the FT ion trap at the end of the first measuring time interval.

16. The mass spectrometer of claim 14, wherein the detector is configured to determine an overall amount of ions of the removed excited ions during the removal of the excited ions from the FT ion trap.

17. The mass spectrometer of claim 16, wherein the detector is configured to assign a respective individual amount of ions to the ion frequencies in the first frequency spectrum on the basis of the overall amount of ions determined.

18. The mass spectrometer of claim 13, wherein the ions are excited via a selective IFT excitation dependent on a mass-to-charge ratio of the ions.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Exemplary embodiments are illustrated in the schematic drawing and are explained in the following description. In the drawing:

(2) FIG. 1 shows a schematic representation of a mass spectrometer with an electric FT-ICR ion trap and with an excitation device for exciting ions,

(3) FIG. 2 shows a schematic representation of a time sequence with a first measuring time interval for recording a first frequency spectrum, in which ion frequencies and interference frequencies are contained, and a second measuring time interval for recording a second frequency spectrum, in which interference frequencies but no ion frequencies are present,

(4) FIG. 3 shows a schematic representation analogous to FIG. 2, in which the ions are excited with a degree of excitation of more than 100% before a first measuring time interval for recording the first frequency spectrum,

(5) FIG. 4 shows a schematic representation analogous to FIG. 3, in which the second frequency spectrum is recorded in a second measuring time interval, which is at a time before the first measuring time interval,

(6) FIG. 5A, 5B show schematic representations of the first and second recorded frequency spectra,

(7) FIG. 6 shows a schematic representation of a first frequency spectrum with interference frequencies in frequency ranges in which no excitation or an excitation with a degree of excitation of more than 100% takes place, and

(8) FIG. 7a,b show schematic representations of the variation over time of an ion signal during the removal of ions from the ion trap for the determination of an overall amount of ions, and also a mass spectrum with ion frequencies or m/z ratios to which individual amounts of ions are assigned.

DESCRIPTION OF EXEMPLARY EMBODIMENTS

(9) In the following description of the drawings, identical reference signs are used for identical or functionally identical components, respectively.

(10) FIG. 1 schematically shows a mass spectrometer 1 which has an electric FT-ICR ion trap 2. The FT-ICR ion trap 2 has a ring electrode 3, applied to which is a high-frequency AC voltage V.sub.RF, which may have for example a frequency f.sub.RF of the order of magnitude of kHz to MHz, e.g. 1 MHz, and an amplitude V.sub.RF of several hundred volts. The high-frequency AC voltage V.sub.RF produces in the FT-ICR ion trap 2 a high-frequency alternating field in which ions 4a, 4b of a gas 4 to be analyzed are dynamically stored.

(11) From the high-frequency alternating field (E-field), there results an average restoring force that acts on the ions 4a, 4b all the more strongly the further away the ions 4a, 4b are from the middle or center of the FT-ICR ion trap 2. To measure the mass-to-charge ratio (m/z) of the ions 4a, 4b, the latter are excited by an excitation signal S1, S2 (stimulus) to carry out oscillations, the frequency f.sub.i of which is dependent on the ion mass and the ion charge and is typically in the frequency range of kHz to MHz orders of magnitude, e.g. from about 1 kHz to 200 kHz. The respective excitation signal S1, S2 is produced by a second excitation unit 5B and a third excitation unit 5c which forms an excitation device 5 together with a first excitation unit 5a, which serves for producing the high-frequency storage voltage V.sub.RF with the predetermined storage frequency f.sub.RF. The excitation device 5 also has a synchronization device 5d, which synchronizes the three excitation units 5a-c in time. Downstream of each excitation unit 5a-c is an amplifier, which is likewise part of the excitation device 5.

(12) For a non-reactive, non-destructive detection (i.e. the ions 4a, 4b are still present after the detection), the oscillation signals of the ions 4a, 4b are tapped in the form of induced mirror charges at the measuring electrodes 6a, 6b, as described for example in DE 10 2013 208 959 A, which was cited at the beginning, the entirety of which is incorporated into the content of this application by reference. As described in detail there, the respective measuring electrodes 6a, 6b are respectively connected to a low-noise charge amplifier 8a, 8b by way of a filter 7a, 7b. The charge amplifiers 8a, 8b on the one hand capture and amplify the ion signals from the two measuring electrodes 6a, 6b and on the other hand keep the measuring electrodes 6a, 6b at the virtual ground potential for the storage frequency f.sub.RF. By forming the difference from the signals supplied by the charge amplifiers 8a, 8b, an ion signal u.sub.i(t) is produced, the variation over time of which is shown at the bottom right in FIG. 1.

(13) The ion signal u.sub.i(t) is fed to a detector 9, which, in the example shown, has an analog-to-digital converter 9a and a spectrometer 9b for fast Fourier analysis (FFT) in order to produce a mass spectrum, which is shown by way of example at the top right in FIG. 1. In this case, the detector 9 or the spectrometer 9b firstly produces a frequency spectrum of the characteristic ion resonance frequencies f.sub.i of the ions 4a, 4b stored in the FT-ICR ion trap 2, which is converted into a mass spectrum on the basis of the dependence of the ion resonance frequencies f.sub.i on the mass and charge of the respective ions 4a, 4b. In the mass spectrum, the number of detected particles or charges in dependence on the mass-to-charge ratio m/z is shown.

(14) Consequently, the electric FT-ICR ion trap 2 allows a direct detection or the direct recording of a mass spectrum, as a result of which a quick gas analysis is made possible. However, the fast recording of a mass spectrum with the aid of Fourier spectrometry can be performed not only in the electric FT-ICR ion trap 2 described above, but also in variations of the type of trap shown in FIG. 1, for example in the case of a so-called toroidal trap or in the case of a differently shaped FT ion trap, such as, for example, in the case of a so-called Orbitrap.

(15) As described further above, all of the ions 4a, 4b in the FT-ICR ion trap 2 have an ion frequency f.sub.i which is proportional to their mass-to-charge ratio (m/z) and with which the stored ions 4a, 4b oscillate in the FT-ICR ion trap 2. If the ions 4a, 4b are excited at their respective ion frequency f.sub.i, they either can be excited in a targeted manner in this way or be thrown out of the FT-ICR ion trap 2 by a resonance step-up, i.e. by an excitation with a degree of excitation of 100% or more. Consequently, ions 4a, 4b with certain mass-to-m charge ratios m/z can be selectively excited or their storage in the FT-ICR ion trap 2 can be prevented/suppressed.

(16) The generalization of this principle leads to one or more regions (“windows”) in the ion resonance frequency range, in which ions 4a, 4b whose ion resonance frequency f.sub.i lies within the respective window can be excited or suppressed in a targeted manner. The inverse transformation of these regions by way of an inverse Fourier transform supplies the time signal used for the so-called IFT excitation. If these variations over time are calculated in advance, this is referred to as SWIFT excitation. The measuring electrodes 6a, 6b can be used for such a SWIFT excitation. By the SWIFT excitation, the ions 4a, 4b can be deflected in the direction of the measuring electrodes 6a, 6b in such a way that both during the ion production and ion storage and also immediately before the detection of the ion signals u.sub.i(t) certain ions 4a, 4b are on the one hand either stored or not stored and on the other hand are excited virtually continuously or not excited at all.

(17) In principle, there are two possibilities for producing the ions 4a, 4b by ionizing the gas 4: either the ions 4a, 4b are produced outside the FT-ICR ion trap 2 or the gas 4 is fed to the FT-ICR ion trap 2 in a charge-neutral form and the ionization takes place in the FT-ICR ion trap 2. For example, such an ionization in the FT-ICR ion trap 2 can be carried out in the way described in WO 2015/003819 A1, which was cited at the beginning and is incorporated into the content of this application with respect to this aspect by reference. Parasitic interference frequencies f.sub.n that lead to spectral lines in the recorded mass spectrum which are not attributable to the excited ions 4a, 4b stored in the FT-ICR ion trap 2 may occur in the FT-ICR ion trap 2 when recording mass spectra via the mass spectrometer 1. Such interference frequencies f.sub.n may lead to a misinterpretation of the mass spectrum.

(18) In order to identify, and possibly eliminate, interference frequencies f.sub.n in the mass spectrum, a method that is described in more detail below on the basis of FIG. 2 may be used:

(19) In a first step of the method, the ions 4a, 4b in the FT-ICR ion trap 2 are excited for a short time in an excitation time interval Δt.sub.S1 via a SWIFT excitation. The degree of excitation A in the SWIFT excitation is at a maximum of about 90%, so that the excited ions 4a, 4b remain in the FT-ICR ion trap 2 and are not removed from it. After the excitation, the excited ions 4a, 4b are detected in a first measuring time interval FFT1, in which the mirror charges induced at the measuring electrodes 6a, 6b are recorded via the detector 9. In this case, the ion oscillations or the ion signals decay with characteristic decay time constants, which depend inter alia on the average free path length of the molecules or of the ions 4a, 4b in the FT-ICR ion trap 2.

(20) FIG. 5A shows a first frequency spectrum FS1, which is recorded in the first measuring time interval FFT1—corresponding to a first measuring time window of a Fast Fourier Transform. As can be seen in FIG. 5A, in the first frequency spectrum FS1 there are not only spectral lines or peaks that correspond to ion frequencies f.sub.i of the excited ions 4a, 4b but also spectral lines that correspond to parasitic interference frequencies f.sub.n in the FT-ICR ion trap 2. The interference frequencies f.sub.n are not produced by the excited ions 4a, 4b, and may therefore lead to a misinterpretation of the mass spectrum.

(21) In order to identify the interference frequencies f.sub.n in the first frequency spectrum FFT1, after the first measuring time interval FFT1 the ions 4a, 4b are excited once again in a second excitation interval Δt.sub.S2 via a SWIFT excitation, wherein a maximum degree of excitation A of about 150% is chosen in this case for the excitation. The high degree of excitation A causes the excited ions 4a, 4b to impinge on the measuring electrodes 6a, 6b and to be removed quickly from the FT-ICR ion trap 2. In a decay time interval Δt.sub.EX directly following the second excitation time interval Δt.sub.S2 and directly before the second measuring time interval FFT2, the (normalized) ion signal u.sub.i(t), to be more precise its component attributable to the ion frequencies f.sub.i, only the envelope of which is shown in FIG. 2, decays very quickly and steeply.

(22) In a frequency spectrum FS2 recorded in the second measuring time interval FFT2 and shown in FIG. 5B, the ion frequencies f.sub.1 can no longer be seen, since they have a much smaller line height in comparison with the interference frequencies f.sub.n that are present in the second frequency spectrum FS2, or have been eliminated virtually completely. By a comparison of the second frequency spectrum FS2, shown in FIG. 5B, with the first frequency spectrum FS1, shown in FIG. 5A, the interference frequencies f.sub.n in the first frequency spectrum FS1 can be identified and correspondingly marked, as can be seen in FIG. 5A, in which the interference frequencies f.sub.n are represented by broken lines and the ion frequencies f.sub.i are represented by solid lines. Optionally, the interference frequencies f.sub.n may be eliminated from the first frequency spectrum FS1 in the detector 9, to be more precise in the spectrometer 9b, so that only the “genuine” ion frequencies f.sub.1 can be seen in the representation or the display of the first frequency spectrum FS1. For example, a number of 8 kilo samples (kS), 16 kS, 32 kS or 64 kS may be recorded in the respective measuring time intervals FFT1, FFT2.

(23) FIG. 3 shows the time sequence of the recording of a first frequency spectrum FS1 and a second frequency spectrum FS2 for identifying interference frequencies f.sub.n in the first frequency spectrum FS1, which differs from the sequence shown in FIG. 2 in that, in an excitation time interval Δt.sub.S before the first measuring time interval FFT1, an excitation of the ions 4a, 4b with a “gentle” superexcitation takes place, with a maximum degree of excitation A of little more than 100%, to be specific of about 110%. In this way, the ions 4a, 4b are already removed from the FT ion trap 2 in the first measuring time interval FFT1, to be more precise at the beginning of the first measuring time interval FFT1.

(24) Correspondingly, the (normalized) ion signal u.sub.i(t), of which only the envelope is shown in FIG. 3, decays quickly and steeply at the beginning of the first measuring time interval FFT1, so that the stored ions 4a, 4b or the stored ion populations are removed completely from the FT ion trap 2. The first frequency spectrum FS1, which is shown in FIG. 5A and, as described in connection with FIG. 2, contains both ion frequencies f.sub.i and interference frequencies f.sub.n, is recorded on the basis of the mirror charges at the measuring electrodes 6a, 6b in the first measuring time interval FFT1.

(25) Since the ions 4a, 4b have been completely removed from the FT ion trap 2 in the first measuring time interval FFT1, there are no longer any excited ions 4a, 4b in the FT ion trap 2 in the case of the time sequence shown in FIG. 3 in the second measuring time interval FFT2, which follows on directly after the first measuring time interval FFT1. Correspondingly, only interference frequencies f.sub.n but no longer any “genuine” ion frequencies f.sub.i are detected in the recording of the second frequency spectrum FS2 in the second measuring time interval FFT2.

(26) In the examples shown in FIG. 2 and in FIG. 3, the excited ions 4a, 4b are completely removed from the FT ion trap 2 directly before the second measuring time interval FFT2 or in the first measuring time interval FFT1. During the removal of the ions 4a, 4b, the comparatively steep ion signal u.sub.i(t) shown in FIG. 2 and in FIG. 3 is recorded, on the basis of the slope or decay rate of which, to be more precise on the basis of the signal strength u.sub.i,s of which at the beginning of the decay time interval Δt.sub.EX or the first measuring time interval FFT1, the overall amount of ions n.sub.i,ex (i.e. the overall number of ions) of the excited ions 4a, 4b removed from the FT ion trap 2 is determined.

(27) FIG. 7A shows an example of an ion signal u.sub.i(t) which is recorded during a measurement and contains both the ion frequencies f.sub.i and the interference frequencies f.sub.n. The ion signal u.sub.i(t) has an envelope in the form of a decaying exponential function, which is likewise shown in FIG. 7A. The signal component that contains the ion frequencies f.sub.1 has a linear decay curve, likewise shown in FIG. 7A, in the form of a straight line, which at the point in time t=0 forms a tangent to the exponential function and which at the point in time t=5 ms has a zero transition, i.e. at the point in time t=5 ms the ions have been removed virtually completely from the FT ion trap 2. Given suitable calibration of the FT ion trap 2, the amplitude u.sub.i,s of the envelope of the ion signal u.sub.i(t) at the point in time t=0 ms may be brought into a relationship with the overall amount of ions n.sub.i,ex in the FT ion trap 2, wherein the following applies to the ion signal u.sub.i(t) represented by way of example in FIG. 7A: n.sub.i,ex=24211, which altogether corresponds approximately to a population of twenty four thousand ions.

(28) In the case of the first frequency spectrum FS1 shown in FIG. 5A, in principle only the relative proportion of the individual ion frequencies f.sub.i,a in the overall amount or in the overall number of excited ions 4a, 4b in the FT ion trap 2 is known on the basis of the height of the spectral lines thereof. On the other hand, the absolute amount of the ions 4a, 4b at the respective individual ion frequencies f.sub.i,a in the FT ion trap 2 is not known in the case of conventional mass spectra. However, by the destructive measurement described in FIG. 2 and in FIG. 3, in which the ions 4a, 4b are removed from the FT ion trap 2, the overall amount n.sub.i,ex of the excited ions 4a, 4b can be determined in the way described further above. On the basis of the ratio of the respective heights of the spectral lines measured at the individual ion frequencies f.sub.i,a to the sum of the heights of all of the spectral lines present in the first frequency spectrum FS1, it is possible to determine for the respective ion frequencies f.sub.i,a the amount of ions n.sub.i,a or the number of ions individually for each individual ion frequency f.sub.i,a. FIG. 7B shows a mass spectrum belonging to the ion signal u.sub.i(t) from FIG. 7A (in dependence on the mass-to-charge ratio m/z in amu), in the case of which the respective ion frequencies f.sub.i,a or the mass-to-charge ratios (m/z).sub.a corresponding to them are assigned an individual number of ions n.sub.i,a, to be precise in the case of the example shown in FIG. 7B as follows: (m/z).sub.1=50.40, n.sub.i,1=827; (m/z).sub.2=50.20, n.sub.i,2=3652; (m/z).sub.3=49.59, n.sub.i,3=927; (m/z).sub.4=37.69, n.sub.i,4=6318; (m/z).sub.5=36.12, n.sub.i,5=1319; (m/z).sub.6=36.10, n.sub.i,6=1399; (m/z).sub.7=35.24, n.sub.i,7=2651; (m/z).sub.8=23.68, n.sub.i,8=3474; (m/z).sub.9=22.78; n.sub.i,9=3640.

(29) FIG. 4 shows the time sequence of a measurement in which the sequence over time of the first measuring time interval FFT1 and the second measuring time interval FFT2 is changed over, i.e. the first measuring time interval FFT1 follows the second measuring time interval FFT2 in time. In the example shown, the second frequency spectrum FS2 is recorded before the excitation of the ions 4a, 4b. To be more precise, in the example shown the excitation of the ions 4a, 4b takes place in an excitation time interval Δt.sub.s, which is directly before the end of the second measuring time interval FFT2 and therefore overlaps with it.

(30) The excitation of the ions 4a, 4b in the excitation time interval Δt.sub.s takes place with a degree of excitation A of less than 100%, so that the excited ions 4a, 4b are not removed from the FT ion trap 2. Therefore, in the first measuring time interval FFT1, following on directly after the second measuring time interval FFT2, the excited ions 4a, 4b can be detected in a non-reactive manner.

(31) As described further above in connection with FIG. 2 and FIG. 3, only interference frequencies f.sub.n, but no ion frequencies f.sub.i, are contained in the second frequency spectrum FS2, since the excitation of the ions 4a, 4b only takes place at the end of the second measuring time interval FFT2. It goes without saying that, in the case where signals or spectral lines with too great a height already occur during the excitation, the excitation time interval Δt.sub.s can only begin after the second measuring time interval FFT2, so that no excitation of ions 4a, 4b takes place in the second measuring time interval FFT2. In this case, the excitation time interval Δt.sub.s lies between the second measuring time interval FFT2 and the first measuring time interval FFT1.

(32) For a meaningful measurement of the interference frequencies f.sub.n or the interference signals, it is generally favorable if the time difference between the first measuring time interval FFT1 and the second measuring time interval FFT2 is as small as possible. Ideally, the second measuring time interval FFT2 follows the first measuring time interval FFT1 (or vice versa) at a time difference Δt.sub.s of less than about 10 ms, less than 5 ms or ideally of less than 1 ms. Altogether, it is advantageous if the entire measurement proceeds as quickly as possible, i.e. if the time period between the beginning of the first measuring time interval FFT1 and the end of the second measuring time interval FFT2 is as small as possible, for example if it is less than about 500 ms, or if the two measuring time intervals FFT1, FFT2 are as short as possible. The time period of the measurement may vary in dependence on the desired mass resolution or oscillation resolution, for example in an order of magnitude of between one millisecond and several hundred milliseconds. The time period of the measurement is made up of the time period of the two measuring time intervals FFT1, FFT2 and, in the case of the example shown in FIG. 2, additionally the time period of the excitation time interval Δt.sub.S2 as well as the decay time interval Δt.sub.EX, since these lie between the first and second measuring time intervals FFT1, FFT2. The time period of the first excitation time interval Δt.sub.S2 from FIG. 2 and the excitation time interval Δt.sub.S from FIG. 3 in each case lies before the first measuring time interval FFT1, and is therefore ignored in the determination of the overall time.

(33) Shown in FIG. 6 is a (first) frequency spectrum FS1, which has three frequency ranges f.sub.a, f.sub.b, f.sub.c. In the first frequency range f.sub.a and in the second frequency range f.sub.b, a SWIFT excitation that is frequency-dependent, and consequently dependent on the mass-to-charge ratio m/z, takes place in each case, wherein the degree of excitation A in the first frequency range f.sub.a is about 120% and in the second frequency range f.sub.b is about 90%. No excitation of the ions 4a, 4b takes place in the third frequency range f.sub.c.

(34) Since in the first frequency range the ions 4a, 4b are excited with a degree of excitation A of significantly more than 100%, no ion frequencies f.sub.i or ion signals can occur there. Therefore, the frequencies which, in FIG. 6, are present in the first frequency range f.sub.a can be clearly identified as interference frequencies f.sub.n. The same applies to the frequencies or spectral lines present in the third frequency range f.sub.c, which likewise cannot be ion frequencies f.sub.i because of the absent excitation, so that these can be clearly identified as interference frequencies f.sub.n. Ion oscillations or ion frequencies f.sub.i can only occur in the second frequency range f.sub.b, in which an excitation with a degree of excitation A of less than 100% takes place. It goes without saying that it cannot be ruled out that there are also interference frequencies f.sub.n in the second frequency range f.sub.b. These can however be identified in the way described further above in connection with FIG. 2 to FIG. 4, and if desired be eliminated from the first frequency spectrum FFT1.