ATR sensor and method for determining the concentration of glucose in a fluid

Abstract

The present invention relates to an ATR sensor (10) for determining the concentration of glucose in a fluid, comprising a sensor housing (12) in which at least the following sensor components are accommodated: an infrared radiation source (22); an ATR element (24) designed to transmit, with total internal reflection on at least one boundary surface of the ATR element (24), infrared radiation emitted by the infrared radiation source (22); at least one infrared detector (40) designed to sense infrared radiation transmitted by the ATR element (24) and to output a corresponding infrared measurement signal; and at least two infrared bandpass filters (36a, 36b) which are disposed between the ATR element (24) and the infrared detector (40), each of the infrared bandpass filters (36a, 36b) being designed such that it allows only infrared radiation having a wave number within a predefined wave number range (F1, F2) to pass. At a temperature of 37°C, a first of the infrared bandpass filters (36a) allows infrared radiation having a wave number in a first wave number range (F1) from 1005 cm-1 to 1025 cm-1 to pass. At a temperature of 37°C, a second of the infrared bandpass filters (36b) allows infrared radiation having a wave number in a second wave number range (F2) from 1075 cm-1 to 1095 cm-1 to pass.

Description

[0001] Title: ATR sensor and method for determining the concentration of glucose in a fluid

[0002] The present invention relates to an ATR sensor for determining the concentration of glucose in a fluid. The present invention also relates to a corresponding method for determining the concentration of glucose in a fluid using an ATR sensor.

[0003] In a particular embodiment, an ATR sensor is to be used that has a multi-channel sensor for a plurality of predetermined wavenumber windows. These multiple wavenumber windows can be implemented, in particular, by a plurality of infrared filters with bandpass characteristics; in particular, the ATR sensor can have a four-channel sensor for four wavenumber windows.

[0004] ATR sensors operate according to the principle of attenuated total reflection. With the ATR measurement principle, the light beam is totally reflected at an interface between a reflective element such as a prism and a sample containing the fluid to be examined. However, at the point of reflection, it penetrates the fluid at a wavelength of about one wavelength and is partially absorbed by the fluid.

[0005] The sensor according to the invention and the method according to the invention are intended to be used for determining the concentration of certain organic molecules in fluids, in particular in liquids such as aqueous solutions containing a number of ingredients. For example, in bioprocess engineering, concentrations of feed (e.g., glucose, glycerol) and / or metabolites (e.g., lactate, ammonium, glutamine, glutamate) in fluids, in particular in liquids, for example, aqueous solutions containing such substances, are of particular interest for optimal process control. Electromagnetic radiation in the mid-infrared range (MIR) has a wavenumber in the range of 4000 to 400 cm - 1 or a wavelength in the range of 2.5 to 25 μm. Numerous vibrational and rotational bands of molecules that play a role in bioprocessing and therefore require monitoring lie in this range. This includes glucose, in particular.

[0006] The present invention allows for the determination of glucose in such fluids, especially liquids such as the aforementioned aqueous solutions, to be recorded directly and with high temporal resolution. This can be achieved with good accuracy even when the fluid in question contains a number of other components, trace substances, and / or biologically active components that, due to their structure, exhibit strong interactions with radiation in the mid-infrared (MIR) range. Therefore, the determination of glucose concentration using radiation in the mid-infrared (MIR) range, for example, using an ATR sensor, appears difficult to implement due to massive cross-sensitivities.

[0007] The device according to the invention for the spectrometric analysis of a fluid comprises a process probe with a reflection element referred to as an ATR (attenuated total reflection) element. MIR radiation generated by a corresponding source is coupled into the process probe in a suitable form, e.g., via optical fibers such as silver halide fibers or through a suitable passage as radiation in air, and directed onto the reflection element. After total reflection of the MIR radiation in the reflection element and subsequent interaction of the MIR radiation with the fluid to be analyzed, the MIR radiation is decoupled again and directed onto a spectrometric detection device.

[0008] The ATR element (which is also in contact with the fluid under investigation) can be a prism, a fiber without a cladding, or a special ATR crystal in which the radiation is guided in total internal reflection at an interface between the ATR element and the fluid. In total internal reflection, an evanescent electromagnetic wave forms behind the reflecting interface between the optically denser medium (ATR element) and the optically less dense medium (fluid under investigation). This wave, depending on the angle of incidence of the radiation at the interface, typically has a range of approximately one wavelength.If a sample containing the fluid to be examined is brought close to the surface of the reflection element, the molecules of a substance that has well-developed and excitable vibrational / rotational transitions in the MIR can interact with the evanescent electromagnetic wave, whereby the MIR radiation is attenuated by absorption in a wavelength-dependent manner that is characteristic for the respective substance.

[0009] With the help of the spectrometric detection device, the wavelength dependence (usually expressed as wavenumber dependence) of the absorption of MIR radiation is investigated and used to analyze components in the sample. Numerous structural elements of organic molecules always absorb with similar intensity and in a narrowly defined region of the MIR spectrum, relatively independently of the overall molecule. Since these groups are linked within a molecule, the entire resulting spectrum is characteristic of a particular molecule or substance containing a specific combination of organic molecules. In addition to qualitative analysis, quantitative information is also available via the Beer-Lamberts law.Since the spectra of the individual components overlap in fluids with complex compositions, such as solutions, dispersions, and / or suspensions of multiple substances in a carrier fluid, additional chemometric methods are generally required (H. Mark, J. Workman Jr.: Chemometrics in Spectroscopy; 2007 Elsevier Inc., ISBN 978-0-12-374024-3), such as bandpass filters or so-called "linear variable filters" in conjunction with a detection element array (see, for example, DE 10 316 514 A1) or an entire FTIR spectrometer (FTIR = Fourier transform infrared spectrometry). FTIR spectrometers are relatively large and expensive and therefore unsuitable for continuous, real-time process monitoring in bioprocess engineering. The known variants with linear variable filters are, however, not satisfactory in terms of sensitivity and selectivity.The object of the present invention is therefore to provide a compact ATR sensor for determining the concentration of glucose in a fluid and / or a corresponding method for determining the concentration of glucose in a fluid, by means of which the concentration of glucose in a sample to be examined can be determined sufficiently accurately and with good temporal resolution or even in real time, even in the presence of numerous other substances, such as trace substances or metabolites, which themselves interact to a high degree with MIR radiation and therefore lead to cross-sensitivities when determining the concentration of glucose using methods based on infrared absorption spectrometry, such as ATR spectrometry.In particularly preferred embodiments of the invention, this should be possible even if the temperature and / or pH value of the fluid to be examined does not remain constant but is subject to fluctuations.

[0010] This object is achieved according to the invention by an ATR sensor for determining the concentration of glucose in a fluid of the type proposed here and by a corresponding method for determining the concentration of glucose in a fluid.

[0011] The sensor according to the invention comprises a sensor housing in which at least the following sensor components are accommodated:

[0012] -- an infrared radiation source;

[0013] -- an ATR element which is designed to transmit infrared radiation emitted by the infrared radiation source under total reflection at at least one boundary surface of the ATR element;

[0014] -- at least one infrared detector designed to detect infrared radiation transmitted by the ATR element and output a corresponding infrared measurement signal; and

[0015] -- at least two infrared bandpass filters arranged between the ATR element and the infrared detector, each of the infrared bandpass filters being configured to transmit only infrared radiation having a wavenumber within a predetermined wavenumber window. According to the invention, a first of the infrared bandpass filters is permeable, preferably at a temperature of 37 °C, to infrared radiation having a wavenumber within a first wavenumber window of 1005 cm - 1 up to 1025 cm - 1 transparent and a second of the infrared bandpass filters is, preferably at a temperature of 37 °C, for infrared radiation with a wavenumber in a second wavenumber window of 1075 cm - 1 up to 1095 cm - 1 permeable..

[0016] The method according to the invention for determining the concentration of glucose in a fluid using an ATR sensor comprises:

[0017] Providing a glucose-containing fluid,

[0018] Interaction of the glucose-containing fluid with infrared radiation, the wavelength of which lies in the mid-infrared range,

[0019] Detecting absorption of infrared radiation after interaction with the glucose-containing fluid in at least two different wavenumber windows within the mid-infrared range, wherein a first of the wavenumber windows, preferably at a temperature of 37 °C, infrared radiation with a wavenumber of 1005 cm - 1 up to 1025 cm - 1 and wherein a second of the wavenumber windows, preferably at a temperature of 37 °C, emits infrared radiation with a wavenumber of 1075 cm - 1 up to 1095 cm - 1 includes.

[0020] Each infrared bandpass filter has a wavenumber window within which the infrared bandpass filter is transparent to MIR radiation. For each infrared bandpass filter, the corresponding wavenumber window is characterized by a central wavenumber and an effective half-width. At the central wavenumber, the wavenumber window is maximally transparent. At the half-width, the bandpass filter still exhibits 50% of its maximum transmittance.

[0021] The wavenumbers given for the first and second wavenumber windows refer to the central wavenumber. For infrared radiation with a wavenumber in a first wavenumber window of 1005 cm - 1 up to 1025 cm - 1 transmittance means that the central wavenumber of the first wavenumber window is 1005 cm - 1 up to 1025 cm - 1 For infrared radiation with a wavenumber in a second wavenumber window of 1075 cm - 1 up to 1095 cm- 1 transparent means that the central wavenumber of the second wavenumber window is 1075 cm - 1 up to 1095 cm - 1 The values ​​given preferably refer to a temperature of 37 °C.

[0022] The full width at half maximum transmittance (FWHM) refers to the width at half maximum transmittance and thus expresses the difference between the upper and lower wavenumber limits of the wavenumber window at 50% of the maximum transmittance of the wavenumber window. The effective half-width refers to the half-width achieved by the respective infrared bandpass filter under real-world ambient conditions, particularly taking into account the divergence of the incident angle of infrared radiation and temperature.

[0023] The first wavenumber window can be particularly between 1010 cm - 1 and 1020 cm' 1preferably between 1012 cm - 1 and 1018 cm - 1 and particularly preferably between 1014 cm - 1 and 1016 cm - 1 preferably at a temperature of 37 °C.

[0024] The second wavenumber window can be particularly between 1080 cm - 1 and 1190 cm - 1 preferably between 1082 cm - 1 and 1087 cm - 1, and especially preferred between 1084 cm - 1 and 1086 cm - 1 , preferably at a temperature of 37 °C.

[0025] The respective effective half-widths for the first and second wavenumber windows can in particular be a maximum of 20 cm - 1 for the first wavenumber window and a maximum of 20 cm - 1 for the second wavenumber window. In some designs, it is advantageous if the effective half-widths are at most 15 cm - 1 for the first wavenumber window and a maximum of 15 cm - 1for the second wavenumber window. In further designs, it is advantageous if the effective half-widths are at most 10 cm - 1 for the first wavenumber window and a maximum of 10 cm - 1 for the second wavenumber window. These values ​​also preferably refer to a temperature of 37 °C.

[0026] The first wavenumber window particularly preferably has a central wavenumber of 1015 cm - 1 with an effective half-width of 20 cm - 1 preferably at a temperature of 37 °C.

[0027] The second wavenumber window particularly preferably has a central wavenumber of 1085 cm - 1 with an effective half-width of 20 cm - 1 preferably at a temperature of 37 °C.

[0028] The ATR sensor can be designed, in particular, to determine the concentration of glucose in a liquid, in particular to determine the concentration of glucose in an aqueous solution, suspension, or dispersion. For example, the glucose-containing liquid can contain CHO cell lines.

[0029] In the context of the present disclosure, the term "liquid" refers to all substances whose main phase exhibits a liquid state at temperatures between 20 and 50 °C and standard pressure. Liquids within the meaning of the application can be both pure liquids and single-phase or multiphase compositions consisting of several components, in particular solutions, suspensions, or dispersions.

[0030] The liquid can, in particular, be a medium for cell cultivation, such as a liquid suspension of cell cultures, nutrients, metabolites formed by the interaction of cell cultures with nutrients, and other substances. The cell cultures, nutrients, metabolites, and / or other substances can be present in a liquid phase, particularly in an aqueous phase, as a suspension, dispersion, or solution. Examples of such liquids in bioprocess engineering include feed (e.g., glucose, glycerol) and metabolites (e.g., lactate, ammonium, glutamine, glutamate), the quantitative observation and control of which are of particular interest for optimal process control. Monitoring and controlling the glucose concentration is often particularly important in this case.To achieve this, the concentration must be able to be determined reliably and quickly even if the concentration of other ingredients changes and / or environmental conditions such as temperature and / or pH change.

[0031] Examples of media for cell culture include RPMI media, available and described in more detail at: https: / / www.sigmaaldrich.com / CH / en / products / cell-culture-and-analysis / cell-culture-media-and-buffers / classical-media-and-buffers.

[0032] An ATR sensor, as used for the inventive determination of glucose concentration, is designed such that a sample beam generated by an infrared radiation source is totally reflected at an interface between a fluid to be analyzed and the ATR sensor, so that an evanescent field of infrared radiation propagates into the fluid on the fluid side. The evanescent field of infrared radiation attenuates the incident infrared radiation. To determine the absorption, the intensity of the totally reflected infrared radiation is recorded in relation to the intensity of the infrared radiation radiated by the infrared radiation source.

[0033] The present invention essentially proposes using a specific combination of two infrared bandpass filters with respective wavenumber windows to determine the concentration of glucose. The position of these two wavenumber windows is of primary importance, which are defined by the wavenumber ranges of 1005 cm -1 claimed in claim 1. - 1 up to 1025 cm - 1 for the central wavenumber of the first wavenumber window and of 1075 cm - 1 up to 1095 cm - 1for the central wavenumber of the second wavenumber window. The infrared absorption spectrum for glucose does indeed show a certain absorption of infrared radiation at wavenumbers lying within the first and second wavenumber windows. However, the known absorption maxima for glucose, which are caused by vibrational and rotational processes of functional groups of the glucose molecule, lie neither in the first wavenumber window nor in the second wavenumber window. Rather, the first and second wavenumber windows are chosen such that absorption maxima of glucose lie outside these wavenumber windows. At first glance, therefore, the selection of the first and second wavenumber windows for determining the concentration of glucose may seem counterintuitive and contrary to the usual rules for selecting suitable wavenumber ranges for determining the concentration of a specific substance.of a specific molecule in a fluid using absorption spectrometry. The general procedure for determining the concentration of a substance, such as glucose, using absorption spectrometry is to first determine the wavenumbers or wavenumber windows at which the substance's absorption of infrared radiation maxima occurs. One would therefore first determine the absorption spectrum of the substance, such as glucose, in the mid-infrared range and from this determine the wavenumber windows in which the relative maxima of the absorption coefficient for the substance lie. Once the wavenumber windows with absorption maxima are known, one would select one of these wavenumber windows, or possibly several wavenumber windows with absorption maxima for glucose, and from then on determine the glucose concentration using this selected wavenumber window or these selected wavenumber windows.

[0034] The invention takes a completely different approach. Instead of searching for maxima of the absorption of radiation in the mid-infrared range (MIR radiation) by glucose, the present invention pursues the approach of finding a combination of wavenumber windows in which the absorption of radiation in the infrared range by glucose is influenced as little as possible by the presence of other substances in the fluid under investigation, such as glycerol and / or metabolites such as lactate, ammonium, glutamine, or glutamate, and furthermore remains largely stable even when certain environmental parameters such as the temperature and / or pH of the fluid under investigation change. A particular difficulty lies in the complexity of all these possible factors influencing the absorption of MIR radiation, compared to the absorption of MIR radiation by glucose.It has been found that it is possible to obtain a measurement signal which is largely stable with respect to the above-mentioned influencing factors and which can be used to determine the concentration of glucose in a fluid to be examined, in particular a liquid such as an aqueous solution, by using the absorption signals recorded in two wavenumber windows in the mid-infrared range and combining them, namely a first absorption signal in a first wavenumber window of 1005 cm. - 1 up to 1025 cm -1 and in a second wavenumber window of 1075 cm - 1 up to 1095 cm - 1 (preferably based on a temperature of 37 °C).

[0035] If we consider the absorption of MIR radiation by glucose in the first and second wavenumber windows separately, we can see that neither wavenumber window represents a region where the absorption of MIR radiation is particularly pronounced. The first wavenumber window, with a central wavenumber between 1005 cm - 1 and 1025 cm - 1 lies in a range in which the absorption coefficient for glucose has a rather medium value and increases to a well-defined maximum of the absorption coefficient, which is approximately at 1035 cm - 1 The second wavenumber window with a central wavenumber between 1075 cm - 1 and 1095 cm - 1 lies in a range where the absorption coefficient for glucose has a rather medium value. Unlike in the first wavenumber window, in the second wavenumber window the absorption coefficient for glucose drops from a slightly lower relative maximum at 1080 cm - 1to a minimum of the absorption coefficient at 1095 cm - 1 down.

[0036] This combination of a first wavenumber window with a central wavenumber between 1005 cm - 1 and 1025 cm - 1 and a second wavenumber window with a central wavenumber between 1075 cm - 1 and 1095 cm - 1has never been considered in the prior art for determining glucose concentration. Glucose possesses absorption bands in the mid-infrared range due to vibrational and rotational processes of some of its functional molecular groups. The maxima of these absorption bands lie outside the first and second wavenumber windows. Therefore, there was no reason to attribute any particular importance to the first and / or second wavenumber windows for determining glucose concentration. Rather, these two wavenumber windows appeared to be of secondary importance for determining glucose concentration compared to the maxima of the glucose absorption bands.

[0037] It was therefore surprising that the combination of wavenumber windows chosen according to the invention actually has some advantages over other – initially much closer – wavenumber windows or combinations of wavenumber windows with regard to determining glucose concentration. Further investigations by the applicant to understand why this particular combination of two wavenumber windows between 1005 cm - 1 and 1025 cm - 1 and between 1075 cm - 1 and 1095 cm - 1indicate that, precisely in these two wavenumber windows, the ratio of the absorption of MIR radiation behind the total reflection interface by glucose to the absorption of MIR radiation by other trace substances is extraordinarily large. Therefore, an absorption signal that is largely robust against cross-sensitivity from other trace substances can be obtained by combining the absorption signals from these two wavenumber windows. In this way, the signal-to-noise ratio in determining the concentration of glucose can be significantly improved, even though the absolute intensity of the absorption signals recorded is not particularly high.The combination of both wavenumber windows appears to play a role here, as it is not possible to achieve a comparable effect if only one of these two wavenumber windows is used to determine the glucose concentration. The advantages identified by the invention were not to be expected from the outset. There was no priori indication of the finding established by the invention that this method can suppress cross-sensitivities to other substances, for example, metabolites like lactate, and that the measurement signals also become significantly more stable against interference such as temperature fluctuations or changing pH values ​​of the fluid in which the glucose concentration is to be determined.

[0038] If the temperature of the fluid changes, particularly in the case of a liquid, e.g., a medium for cell cultivation, the temperature of the infrared bandpass filter used to determine the glucose concentration also changes accordingly, because the ATR sensor is in contact with the fluid at least at its interface, where total internal reflection occurs. This temperature change results in the central wavelength of a respective wavenumber window, in which the infrared bandpass filter is permeable, shifting towards larger wavenumbers as the temperature of the respective infrared bandpass filter increases. Therefore, the position of the respective wavenumber window, in which the infrared filter is permeable, also changes compared to the absorption spectrum of glucose, because the absorption bands of glucose are only slightly temperature-dependent. Furthermore, the intensity values ​​in the absorption spectrum are also temperature-dependent.As the fluid temperature increases, the maximum intensity of the absorption signal increases. The first wavenumber window, which is preferably at 37 °C, is at 1015 cm-1. - 1 is located in a region of the absorption spectrum of glucose where the absorption increases with increasing wavenumber. The second wavenumber window, which is preferably at 37 °C, is at 1085 cm 1is located in a region of the absorption spectrum of glucose where the absorption decreases with increasing wavenumber. For this reason, these effects counteract each other and can compensate each other to a large extent. If the temperature of the fluid (and thus also the temperature of the infrared bandpass filters) increases, the influence of the absorption signal obtained from the first wavenumber window increases and the influence of the absorption signal obtained from the second wavenumber window decreases. If the temperature of the fluid decreases, the situation is exactly the opposite: Now the influence of the absorption signal obtained from the first wavenumber window increases and the influence of the absorption signal obtained from the second wavenumber window decreases.By combining the two absorption signals from the first and second wavenumber windows, these effects can be approximately compensated, resulting in an absorption signal that is approximately independent of the fluid temperature, at least in a range between 20 °C and 45 °C. Therefore, by appropriately combining the absorption signals from the first and second wavenumber windows, a sufficiently temperature-independent measurement signal for the absorption of infrared radiation after interaction with the glucose-containing fluid can be derived. The term "combination of absorption signals" refers to any type of combination that has a sufficiently compensating effect on the absorption signals when the temperature changes.

[0039] The ATR sensor and the method for determining the concentration of glucose in a fluid can be designed, in particular, to determine the concentration of glucose in a glucose-containing fluid, in particular in a glucose-containing liquid, at a temperature in a range between 20°C and 45°C. The wavenumber windows are selected such that the absorption signal, which represents the detected absorption of infrared radiation after interaction with the fluid, is only slightly affected when the temperature changes within the selected temperature range. Thus, a glucose concentration in this range can be detected largely independently of temperature changes.

[0040] Further particular embodiments of the ATR sensor and / or method for determining the concentration of glucose according to the invention can have at least one of the optional features specified below. It is understood that each of these optional features can be added individually to one of the previously described embodiments. The features specified below can also be combined with one another, and the combination of features can be added to one of the previously described embodiments, unless it is explicitly stated that certain of these features are mutually exclusive.

[0041] In particular embodiments, the ATR sensor described herein may comprise a third infrared bandpass filter, which is preferably at a temperature of 37 °C for infrared radiation with a wavenumber in a third wavenumber window of 945 cm - 1 up to 965 cm- 1 is transparent. Accordingly, in particular embodiments, the method according to the invention can comprise infrared radiation in the third wavenumber window with a wavenumber of 945 cm - 1 up to 965 cm - 1 to capture. For the third infrared bandpass filter, the wavenumber window is also characterized by a central wavenumber and an associated effective half-width. As previously described, at the central wavenumber, the wavenumber window is maximally permeable, and the half-width indicates the wavenumber range in which the bandpass filter still has at least 50% of its maximum transmittance. The wavenumbers specified for the third wavenumber window refer to the central wavenumber, and the half-width indicates the width of the wavenumber window at at least 50% transmittance (as FWHM). In other words: For infrared radiation with a wavenumber in a third wavenumber window of 945 cm- 1 up to 965 cm - 1 transparent means that the central wavenumber of the third wavenumber window is 945 cm - 1 up to 965 cm - 1 can amount to.

[0042] The third wavenumber window can be particularly between 950 cm - 1 and 960 cm - 1 and particularly preferably between 954 cm - 1 and 956 cm - 1 lay.

[0043] The half-width for the third wavenumber window can be 20 cm - 1 In special embodiments, the half-width for the third wavenumber window can be 15 cm - 1 More preferably, the half-width for the third wavenumber window can be 10 cm - 1 be.

[0044] The third wavenumber window particularly preferably has a central wavenumber of 955 cm - 1 with a half-width of 20 cm - 1 .

[0045] The absorption spectrum of glucose has a third wavenumber window between 945 cm - 1 and 965 cm - 1no special structures are found that can be assigned to vibrational or rotational bands of glucose. Since this finding also applies to most of the substances considered as possible cross-sensitivities, the absorption signal from the third wavenumber window can nevertheless serve as a reference signal, indicating a "background" or "noise" in the absorption signal detected by the ATR sensor caused by various influences. If the absorption signals from the first and second wavenumber windows are combined to determine the glucose concentration and related to the absorption signal from the third wavenumber window, then an absorption signal caused by glucose can be determined with greater accuracy than if the glucose concentration were determined without taking the third wavenumber window into account.This is because the third wavenumber window was specifically selected in a range in which no pronounced vibrational or rotational bands are known for any of the substances considered as potentially relevant components in the fluids under investigation. Therefore, the absorption signal from the third wavenumber window can be used as a reference signal to which the absorption signal derived from the first and / or second wavenumber window can be related. In this way, the concentration of glucose can be determined based on a relative measurement of absorption signals from the first and / or second wavenumber window and the third wavenumber window, eliminating the need for absolute calibration of the absorption signals from the first and / or second wavenumber window.

[0046] In a further embodiment, the ATR sensor may comprise a fourth infrared bandpass filter which is designed for infrared radiation with a wavenumber in a fourth wavenumber window of 1300 cm - 1 up to 1320 cm - 1 is transparent. Accordingly, in the method for determining the concentration of glucose in a fluid, infrared radiation can be detected in a fourth wavenumber window in the mid-infrared range, wherein the fourth wavenumber window detects infrared radiation with a wavenumber of 1300 cm - 1 up to 1320 cm - 1 includes.

[0047] For the fourth infrared bandpass filter, the wavenumber window is also characterized by a central wavenumber and a corresponding effective half-width. As previously described, at the central wavenumber, the wavenumber window is maximally permeable, and the half-width indicates the wavenumber range in which the bandpass filter still has at least 50% of its maximum transmittance. The wavenumbers specified for the fourth wavenumber window refer to the central wavenumber, and the half-width indicates the width of the wavenumber window at at least 50% transmittance (as FWHM). In other words: For infrared radiation with a wavenumber in a fourth wavenumber window of 1300 cm - 1 up to 1320 cm - 1 transparent means that the central wave number of the fourth wave number window is 1300 cm - 1 up to 1320 cm - 1 can amount to.

[0048] The fourth wavenumber window can be particularly between 1305 cm - 1 and 1315 cm - 1 and particularly preferably between 1309 cm - 1 and 1311 cm - 1 lay.

[0049] The half-width for the fourth wavenumber window can be 20 cm - 1 In special embodiments, the half-width for the fourth wavenumber window can be 15 cm - 1 In further embodiments, the half-width for the fourth wavenumber window can be 10 cm - 1 be.

[0050] The fourth wavenumber window particularly preferably has a central wavenumber of 1310 cm - 1 with a half-width of 20 cm - 1 .

[0051] Additional use of the fourth wavenumber window of 1300 cm - 1 up to 1320 cm - 1for determining the concentration of glucose enables even more effective suppression of cross-sensitivities of the absorption signal due to other trace substances in the fluid. Therefore, the signal-to-noise ratio of the absorption signal used to determine the glucose concentration can be further improved, even though the fourth wavenumber window also contains no significant vibrational / rotational bands that could be attributed to functional groups of glucose. However, the additional consideration of the fourth wavenumber window allows for a highly effective correction of the absorption signal supplied as "raw data" from the first, second, and, if applicable, third wavenumber windows for the proportions of important substances prone to cross-sensitivities, such as lactate in particular, but also glutamine, glutamate, and / or asparagine.For lactate in particular, it has been shown that the signal for determining glucose concentration, formed by combining the absorption signals from the first and second wavenumber windows, can be significantly influenced if a strong lactate component is present in the fluid. The influence of this lactate component on the combined absorption signal from the first and second wavenumber windows can be eliminated very directly and clearly by using the absorption signal from the fourth wavenumber window for correction.

[0052] The absorption spectrum for lactate shows a distinct structure of 1300 cm' 1 up to 1320 cm - 1 but also other structures at 1025 cm - 1 and 1100 cm' 1 . Therefore, it is assumed that the fourth wavenumber window (of 1300 cm - 1 up to 1320 cm - 1) primarily allows a conclusion to be drawn about the concentration of lactate and, based on the absorption of MIR radiation in the fourth wavenumber window, the absorption signal used to determine the concentration of glucose can be corrected on the basis of the first and second wavenumber windows in order to suppress cross-sensitivities.

[0053] Expressed in wavelengths, the first, second, third and fourth wavenumber windows result in the following:

[0054] A wave number of 955 cm - 1 with a half-width of 20 cm - 1 corresponds approximately to a wavelength of 10470 nm with a half-width of 220 nm.

[0055] A wavenumber of 1015 cm-m 1 with a half-width of 20 cm - 1 corresponds approximately to a wavelength of 9850 nm with a half-width of 200 nm.

[0056] A wave number of 1085 cm - 1 with a half-width of 20 cm - 1corresponds approximately to a wavelength of 9215 nm with a half-width of 170 nm.

[0057] A wave number of 1310 cm - 1 with a half-width of 20 cm - 1 corresponds approximately to a wavelength of 7635 nm with a half-width of 120 nm.

[0058] The wavenumbers given preferably refer to a normal temperature of 37 °C.

[0059] In particular, a respective one of the first, second, possibly third and possibly fourth infrared bandpass filters can be designed such that a central wavenumber for the respective wavenumber window is determined with an accuracy of at least 5%.

[0060] In a further embodiment, the ATR sensor can have a multi-channel infrared bandpass filter, with one infrared bandpass filter each for the first, second, possibly third, and possibly fourth wavenumber window. Such multi-channel infrared bandpass filters can be constructed particularly compactly. In particular, a multi-channel infrared bandpass filter can be easily accommodated inside a tubular housing made of an inert material, such as stainless steel, with an outer diameter of 12 mm. In this way, it is possible to create a sensor equipped with an ATR sensor according to the invention for determining the concentration of glucose, which corresponds to the format conventionally used for analytical instruments in monitoring processes in biotechnology.Due to the associated high degree of standardization, the sensor can be easily inserted into corresponding mounting locations of conventional biotechnological devices, even those originally intended for recording other parameters. In other words, a multi-channel sensor has a housing in which at least one photodetector or infrared detector and at least two infrared bandpass filters are housed. If the photodetector or infrared detector is configured such that it can distinguish between two different spatial resolutions, a common photodetector is sufficient for the at least two bandpass filters. Alternatively, the multi-channel sensor can have a separate photodetector or infrared detector for each infrared bandpass filter.

[0061] In a further embodiment, the ATR sensor or the method for determining the concentration of glucose in a fluid can be designed to determine the concentration of glucose in a glucose-containing liquid which has a pH value in a range between 5 and 8, in particular in a range between 6.8 and 7.4.

[0062] The absorption signals from the aforementioned wavenumber windows, in particular from the first and second wavenumber windows, possibly related to the absorption signal from the third wavenumber window and / or corrected by the absorption signal from the fourth wavenumber window, can even be selected such that the ultimately used combined absorption signal, which represents the recorded absorption of infrared radiation after interaction with the liquid, fluctuates only slightly in the selected pH range. This allows

[0063] A glucose concentration can be measured in a given range, regardless of pH fluctuations. In some cases, the pH and / or temperature of a liquid, such as a nutrient solution in bioprocess technology, can fluctuate depending on the degree of fermentation.

[0064] The ATR sensor according to the invention and the method according to the invention also offer the possibility, in cases of the type described above, of selecting the two, three, or four wavenumber windows so that the detected absorption signal is largely unaffected by the pH value or temperature in the specified pH range and / or temperature range. In this way, the glucose concentration can be reliably monitored throughout the entire process.

[0065] In a further embodiment, the first, second, possibly third, and possibly fourth infrared bandpass filters can have a predetermined minimum quality, which is determined by the fact that a central wavenumber for the respective wavenumber window shifts with temperature within a temperature range of 20°C to 45°C by a maximum of 5 wavenumbers per 10 degrees Celsius. Furthermore, the ATR sensor can be designed to detect a reference signal based on the infrared radiation emitted by the infrared radiation source without transmission through the ATR element. In this way, the ATR sensor can detect a reference infrared radiation for a respective wavenumber window from the first, second, third, and possibly fourth wavenumber window. In particular, the reference infrared radiation is not guided through the ATR element and is therefore not influenced by absorption of infrared radiation in the fluid to be examined.In this way, the respective infrared radiation detected by the infrared detector can be detected after transmission through the ATR element in relation to the respectively assigned reference infrared radiation and thus the calibration of the ATR sensor can be simplified because fluctuations in the emission intensity of the infrared radiation source are eliminated by reference to the reference infrared radiation and in any case do not have a direct effect on the detected absorption signals.

[0066] The invention is explained in more detail below using exemplary embodiments with reference to the drawings.

[0067] Figure 1 shows a schematic longitudinal sectional view through an inventive embodiment of an ATR sensor of the present application.

[0068] Figure 2 a) and b) show a four-channel infrared measuring sensor with four-channel infrared bandpass filter for the ATR sensor in front view and perspective view.

[0069] Figure 3 shows simplified infrared absorption spectra of glucose and lactate together with the wavenumber windows chosen for the first, second, third and fourth infrared bandpass filters, preferably at 37 °C.

[0070] In Figure 1, an embodiment of an ATR sensor according to the invention is generally designated by 10. The ATR sensor 10 comprises a sensor housing 12 with a tube 13, which is preferably formed in one piece. The sensor housing 12, in particular its tube 13, can generally be made of plastic, but is preferably made of stainless steel in order to permanently withstand even aggressive chemical environments.

[0071] In the example shown, the sensor housing 12 is designed as a cylindrical housing with a cylindrical envelope 14. The cylindrical envelope 14 is indicated by a dash-dotted line in the region of a recess 16 of the otherwise cylindrical sensor housing 12. The sensor housing 12 extends along a sensor housing longitudinal axis L, which, as the sensor housing longitudinal axis L conceived to pass centrally through the sensor housing 12, is also the cylinder axis of the cylindrical sensor housing 12. Outside the recess 16, the cylindrical envelope 14 coincides with the outer surface 13a of the cylindrical tube 13.

[0072] A printed circuit board 18 is arranged in the sensor housing 12 as a carrier component 20 for an infrared radiation source 22. To accommodate the printed circuit board 18, an installation space available in the axial extension area of ​​an ATR element 24 is used. This installation space is available because the ATR element 24, with its flat outer surface 24a serving as the measuring surface 26, must be accessible to a fluid to be measured in the external environment U of the sensor housing 12. Therefore, the ATR element 24 is arranged in the sensor housing 12 at a distance from the sensor housing's longitudinal axis L.

[0073] The infrared radiation source 22 is soldered or welded onto the circuit board 18 as a surface-mounted device (SMD). The circuit board 18 is a printed circuit board 18 printed with conductor tracks in a conventional manner. It also carries a control device 27 that controls the operation of the infrared radiation source 22.

[0074] The circuit board 18 and thus the electronic components soldered, welded or otherwise connected thereto: infrared radiation source 22 and control device 27, can be connected by a ribbon cable 28 to a power source and / or to a higher-level control device of a laboratory equipment within which the ATR sensor 10 is used.

[0075] The ATR sensor 10 has a connection end 10a and a head end 10b axially opposite the sensor housing's longitudinal axis L. The ribbon cable 28 and a connection assembly 30 are electrically contactable at the connection end 10a. The connection assembly 30 serves to output measurement signals from an infrared measurement sensor 32 and an infrared reference sensor 34. At the head end 10b, the tube 13 is closed by a plug 35.

[0076] The infrared measurement sensor 32 and the infrared reference sensor 34 can be essentially identical sensor devices, which may differ, but need not differ, in the infrared bandpass filters 36 and 38 arranged therein. The output of reference detection signals by the infrared reference sensor 34 or their output from the sensor housing 12 is not shown in Figure 1 for the sake of simplicity, but does occur.

[0077] The infrared measuring sensor 32 is shown in Figure 2. Figure 2 a) shows the infrared measuring sensor 32 in a front view through its inlet window onto the four infrared bandpass filters 36a, 36b, 36c, 36d. Figure 2 b) shows a perspective view of the four-channel measuring sensor 32.

[0078] It is understood that the infrared reference sensor 24 can be constructed in the same way as the infrared measuring sensor 32, so that the illustrations in Figure 2 also apply accordingly to the infrared reference sensor 34. In the example shown, both the infrared measuring sensor 32 and the infrared reference sensor 34 each have four detectors 40a, 40b, 40c, 40d (designated in Figure 1 for the sake of simplicity by the reference numeral 40) and 42a, 42b, 42c, 42d (designated in Figure 1 for the sake of simplicity by the reference numeral 42), in front of which an infrared bandpass filter 36a, 36b, 36c, 36d or 38a, 38b, 38c, 38d is arranged in the beam path coming from the infrared radiation source 22. The infrared bandpass filter 36 has four infrared bandpass filters 36a, 36b, 36c, and 36d. Accordingly, the infrared bandpass filter 38 also has four infrared bandpass filters 38a, 38b, 38c, and 38d.The infrared measuring sensor 32 has a total of four detectors 40a, 40b, 40c, 40d, of which only the two detectors 40a and 40c located behind the longitudinal sectional plane containing the sensor housing longitudinal axis L can be seen in Figure 1 and are provided with the reference number 40 for simplicity. The other two detectors 40b and 40d are located orthogonally to the plane of the drawing in Figure 1 in front of this plane and in front of the two detectors 40 shown. The same applies to the infrared bandpass filters 36a, 36b, 36c, 36d, which are arranged in front of the incident side of the detectors 40 in order to allow only infrared radiation with a wavelength defined by the respective infrared bandpass filter 36a, 36b, 36c, 36d to be incident on a detector 40a, 40b, 40c, 40d.

[0079] The infrared bandpass filters 36a, 36b, 36c, 36d each have different wavenumber windows in which they are designed for radiation in the mid-infrared range (MIR radiation, wavenumber in the range of 4000 to 400 cm - 1 or a wavelength in the range of 2.5 to 25 μm). Each of the infrared bandpass filters 36 is opaque to MIR radiation outside its respective wavenumber window.

[0080] The individual infrared bandpass filters 36a, 36b, 36c, 36d and 38a, 38b, 38c, 38d are permeable to the following bandwidth of MIR radiation (also referred to as wavenumber window) specified by a respective central wavenumber and full width at half maximum (FWHM):

[0081] First infrared bandpass filter 36a or 38a: central wave number of 1015 cm - 1 with an effective half-width of 20 cm - 1 preferably at a temperature of 37 °C.

[0082] Second infrared bandpass filter 36b or 38b: central wave number of 1085 cm - 1 with an effective half-width of 20 cm - 1 preferably at a temperature of 37 °C.

[0083] Third infrared bandpass filter 36c or 38c: central wave number of 955 cm - 1 with a half-width of 20 cm - 1 .

[0084] Fourth infrared bandpass filter 36d or 38d: central wave number of 1310 cm - 1 with a half-width of 20 cm - 1 .

[0085] The infrared measuring sensor 32 is thus a four-channel measuring sensor. The infrared reference sensor 34 can be constructed identically, so that for the example presented here, the description given for the infrared measuring sensor 32 can also be used to explain the infrared reference sensor 34. The infrared measuring sensor 32 has a housing 44 which is cylindrical over at least 75% of its extension along the sensor housing's longitudinal axis L, with a cylinder axis Z32 that is coaxial with the sensor housing's longitudinal axis L. The housing 44 of the infrared measuring sensor 32 therefore also has a cylindrical shape over at least 75% of its axial extension along the cylinder axis Z32. What has been said about the housing 44 applies mutatis mutandis to the housing 46 of the infrared reference sensor 34 and its cylinder axis Z34.

[0086] The sensor housing 12 has an axial measuring section 48, which, as a cylindrical measuring section 48, is delimited by an axial section of the cylindrical outer surface 13a of the tube 13 of the sensor housing 12. The infrared measuring sensor 32 is accommodated in the measuring section 48.

[0087] At an axial distance from the measuring section 48, the sensor housing 12 has an axial reference section 50. This is also a cylindrical reference section 50 and is defined by an axial section of the cylindrical outer surface 13a of the tube 13 of the sensor housing 12. The infrared reference sensor 34 is housed in the reference section 50.

[0088] The respective infrared sensors, infrared measuring sensor 32 and infrared reference sensor 34, are accommodated in the respective sections, measuring section 48 and reference section 50, by means of a substantially identically designed positioning element 52. The positioning element 52 is an annular element which preferably has a groove formed on its radially outer surface for the passage of the ribbon cable 28. At the positioning element 52 of the reference section 50, the groove is empty in the illustration. The ribbon cable 28 is guided through the groove of the positioning element 52 of the measuring section 48. However, a conductor (not shown in Fig. 1) for conducting measurement signals from the infrared reference sensor 34 out of the sensor housing 12 can run through the groove of the positioning element 52 of the reference section 50 and, in its further course, also through the groove of the positioning element 52 of the measuring section 48.The annular positioning elements 52 can be arranged frictionally in the tube 13 and can frictionally hold the infrared sensor 32 or 34 held by them. However, the positioning elements 52 can also be arranged adhesively 25 by interposing an adhesive agent on the tube 13 and can be adhesively connected in a similar manner to the respective infrared sensor 32 or 34 positioned by them.

[0089] The support component 20 with the infrared radiation source 22 and the ATR element 24 are arranged axially between the infrared sensors 32 and 34, which are arranged with the detector surfaces of the detectors 40 and 42 facing one another. Additionally, an infrared measuring radiation reflector 54 is arranged axially between the infrared radiation source 22 and the ATR element 24. By means of its reflective surface 54a, the reflector deflects an infrared measuring radiation component 56 emitted by the infrared radiation source 22 and transmitted by the ATR element 24 with reflection at its parallel boundary surfaces 24a and 24b toward the bandpass filters 36 and the detectors 40 of the infrared measuring sensor 32.

[0090] The circuit board 18 has a recess 58 at the location where the infrared radiation source 22 is mounted. The infrared radiation source 22 is arranged above the recess 58, so that it can not only emit infrared radiation as the infrared measurement radiation component 56 toward the inclined surface of the ATR element 24, but in the example shown, can simultaneously emit an infrared reference radiation component 60 in the opposite direction through the recess 58 in the circuit board 18.

[0091] The infrared reference radiation component 60 reaches the infrared reference sensor 34 15 via an infrared reference radiation reflector 62, whose reflective surface 62a deflects the infrared reference radiation component 60 emitted by the infrared radiation source 22 toward the bandpass filters 38 and the detectors 42.

[0092] The infrared reference radiation reflector 62 is axially arranged between the support member

[0093] 20 with the infrared radiation source 22 and the ATR element 24 arranged thereon on the one hand, and the infrared reference sensor 34 on the other hand. The infrared reference radiation reflector 62 supports the support component 20. The support component is held by a mounting component 64, which is located behind the sectional plane of Figure 1. Another mounting component 64 can be located parallel to the mounting component 64 shown, in front of the sectional plane of Figure 1.

[0094] The ATR element 24, which is transparent to infrared radiation, is connected to the tube 13 by soldering through a circumferential solder joint 66. The measuring surface 26 is covered by a membrane 68, which is in direct contact with the measuring surface 26 and which is permeable to the fluid to be measured in the external environment U of the ATR sensor 10, but not to suspended matter contained in the fluid, such as cells, cell residues, and the like. This prevents falsification of the measurement result due to disruptive influences from solid particles. The membrane 68 is fixed to the measuring surface 26 by a frame 70 anchored positively in the tube 13 or in the sensor housing 12. A seal 72 between the frame 70 and the solder section 66 prevents fluid from the external environment U from entering the interior A of the ATR sensor 10 or its sensor housing 12.The membrane 68 is accessible for fluid in the external environment U via a recess 71 which completely penetrates the frame 70 in its thickness direction.

[0095] The design described above with components arranged successively in the axial direction, wherein the ATR element 24, the infrared radiation source 22 and a large part of the carrier component 20 axially overlap in the axial extension area of ​​the ATR element 24, allows a very slim dimensioning of the ATR sensor 20, which in its cylindrical and thus rotationally symmetrical sections 48 and 50 with respect to the sensor housing longitudinal axis L has a dimension D that is orthogonal to the sensor housing longitudinal axis L and does not exceed 12 mm.

[0096] In the area between the measuring section 48 and the reference section 50, namely in the area of ​​the recess 16, the sensor housing 12 is partially cylindrical and has a radial dimension starting from the sensor housing longitudinal axis L, which at least in the partially cylindrical area of ​​the tube 13 of the sensor housing 12 does not exceed a dimension of 6 mm.

[0097] Figure 3 shows simplified infrared absorption spectra of glucose and lactate in the mid-infrared radiation (MIR) range together with the wavenumber windows F1, F2, F3, F4 selected for the first (36a, 38a), second (36b, 38b), third (36c, 38c), and fourth (36d, 38d) infrared bandpass filters of the infrared measuring sensor 32 and the infrared reference sensor 34, respectively (preferably at a temperature of 37°C). The curve of the absorption coefficient for glucose versus wavenumber is designated by reference numeral 50 in Figure 3. The curve of the absorption coefficient for lactate versus wavenumber is designated by reference numeral 70 in Figure 3.

[0098] It can be seen in Figure 3 that the absorption coefficient 50 for glucose has absorption maxima at wavenumbers of 1035 cm - 1 (52), 1080 cm - 1 (54), 1110 cm - 1 (54) and 1150 cm - 1(58). The absorption coefficient 70 for lactate, on the other hand, shows absorption maxima at wavenumbers of 1040 cm - 1 (72), 1125 cm - 1 (74) and 1315 cm - 1 (76).

[0099] The first infrared bandpass filter 36a or 38a has a first wavenumber window F1 with a central wavenumber of 1015 cm - 1 and an effective half-width of 20 cm - 1 .

[0100] The second infrared bandpass filter 36b or 38b has a second wavenumber window F2 with a central wavenumber of 1085 cm - 1 and an effective half-width of 20 cm - 1 .

[0101] The third infrared bandpass filter 36c or 38c has a third wavenumber window with a central wavenumber of 955 cm - 1 and a half-width of 20 cm - 1 The fourth infrared bandpass filter 36d or 38d has a fourth wavenumber window with a central wavenumber of 1310 cm- 1 and a half-width of 20 cm - 1 .

[0102] If we consider the absorption coefficient 50 for glucose in the first wavenumber window F1 and the second wavenumber window F2 separately, we can see that neither wavenumber window represents a region where the absorption of MIR radiation is particularly pronounced. The first wavenumber window F1, with a central wavenumber of 1015 cm - 1 lies in a range in which the absorption coefficient 50 for glucose has a rather medium value and increases to a well-defined maximum 52 of the absorption coefficient, which is approximately at 1035 cm - 1 The second wavenumber window F2 with a central wavenumber of 1085 cm - 1lies in a range where the absorption coefficient 50 for glucose also has a rather medium value. Unlike in the first wavenumber window F1, in the second wavenumber window F2, the absorption coefficient 50 for glucose drops from a slightly lower relative maximum of 54 at 1080 cm - 1 to a minimum of the absorption coefficient at 1095 cm - 1 down.

[0103] However, the first wavenumber window F1 and the second wavenumber window F2 have in common that, precisely in these two wavenumber windows, the ratio of the absorption coefficient 50 for glucose to the absorption coefficient 70 for lactate is extraordinarily large. It has been shown that a similar relationship also applies to the absorption coefficient for other trace substances such as glutamine, glutamate, or asparagine. For this reason, an absorption signal for glucose that is largely robust against cross-sensitivities from other trace substances can be obtained by combining the absorption signals from these two wavenumber windows F1 and F2. In this way, the signal-to-noise ratio in determining the concentration of glucose can be significantly improved, even though the absolute intensity of the absorption signals recorded for glucose is not particularly high.

[0104] A further special feature of the selected combination of absorption signals from the first wavenumber window F1 and the second wavenumber window F2 results from the absorption coefficient 50 for glucose in the first wavenumber window F1 increasing with increasing wavenumber and the absorption coefficient 50 for glucose in the second wavenumber window F2 decreasing with increasing wavenumber. When the temperature of the fluid to be examined changes, the temperature of the infrared bandpass filters 36a, 36b, 38a, 38b used to determine the glucose concentration also changes accordingly because the ATR sensor 10 is in contact with the fluid to be examined at least with its interface, at which total reflection takes place.This temperature change results in the central wavelength of a respective wavenumber window F1, F2, in which the infrared bandpass filter 36a, 36b, 38a, 38b is permeable, shifting toward larger wavenumbers as the temperature of the respective infrared bandpass filter 36a, 36b, 38a, 38b increases. Therefore, the position of the respective wavenumber window F1, F2, in which the infrared bandpass filter is permeable, also changes relative to the absorption spectrum 50 for glucose, because the absorption bands 52, 54, 56, 58 for glucose are only slightly temperature-dependent.Since the first wavenumber window F1 is located in a region of the absorption spectrum of glucose where the absorption coefficient 50 increases with increasing wavenumber, and the second wavenumber window F2, in contrast, is located in a region of the absorption spectrum of glucose where the absorption coefficient 50 decreases with increasing wavenumber, these effects counteract each other and can thus at least partially compensate for each other. If the temperature of the fluid under investigation increases (and thus also the temperature of the infrared bandpass filters 36a, 36b, 38a, 38b), the influence of the absorption signal obtained from the first wavenumber window F1 increases, and the influence of the absorption signal obtained from the second wavenumber window F2 decreases.When the fluid temperature decreases, the opposite is true: Now the influence of the absorption signal obtained from the first wavenumber window F1 increases, while the influence of the absorption signal obtained from the second wavenumber window F2 decreases. Combining the two absorption signals from the first and second wavenumber windows F1 and F2 approximately balances these effects, yielding an absorption signal that is approximately independent of the fluid temperature, at least in a range between 20°C and 45°C. Therefore, by appropriately combining the absorption signals from the first and second wavenumber windows F1 and F2, a sufficiently temperature-independent measurement signal for the absorption of infrared radiation after interaction with the glucose-containing fluid can be derived.

[0105] The absorption spectrum of glucose shows the third wavenumber window F3 at a wavenumber of 955 cm - 1 no special structures are found that can be assigned to vibrational or rotational bands of glucose. Since this finding also applies to most of the substances considered as potential cross-sensitivities, the absorption signal from the third wavenumber window F3 can serve as a reference signal, indicating a "background" or "noise" of the absorption signal detected by the ATR sensor 10 caused by various influences.

[0106] Additional use of the fourth wavenumber window F4 at a wavenumber of 1310 cm - 1for determining the concentration of glucose enables even more effective suppression of cross-sensitivities of the absorption signal due to other trace substances in the fluid. Therefore, the signal-to-noise ratio of the absorption signal used to determine the glucose concentration can be further improved, even though the fourth wavenumber window F4 also contains no prominent vibrational / rotational bands that could be attributed to functional groups of glucose. However, the additional consideration of the fourth wavenumber window F4 allows for highly effective correction of the absorption signal supplied as "raw data" from the first, second, and possibly third wavenumber windows for the proportions of important substances prone to cross-sensitivities, such as lactate in particular, but possibly also glutamines, glutamates, and / or asparagines.

[0107] For lactate in particular, it has been shown that the signal for determining glucose concentration, formed by combining the absorption signals from the first and second wavenumber windows F1 and F2, can be significantly influenced if a strong lactate component is present in the fluid. The influence of this lactate component on the combined absorption signal from the first and second wavenumber windows F1 and F2 can be calculated very directly and clearly by using the absorption signal from the fourth wavenumber window F4 for correction.

[0108] The absorption coefficient 70 for lactate has a pronounced maximum 76 at a wavenumber of 1315 cm - 1 but also further maxima 72, 74 at wavenumbers of 1040 cm - 1 and 1125 cm - 1. Therefore, it can be assumed that the absorption signal provided by the fourth wavenumber window F4 primarily allows a conclusion to be drawn about the concentration of lactate and that, based on this absorption signal from the fourth wavenumber window F4, the absorption signal used to determine the concentration of glucose can be corrected on the basis of the first and second wavenumber windows F1 and F2 in order to suppress cross-sensitivities.

Claims

Claims:

1. ATR sensor (10) for determining the concentration of glucose in a fluid, comprising a sensor housing (12) in which at least the following sensor components are accommodated: -- an infrared radiation source (22); - an ATR element (24) which is designed to transmit infrared radiation emitted by the infrared radiation source (22) under total reflection at at least one boundary surface of the ATR element (24); - at least one infrared detector (40) designed to detect the infrared radiation transmitted by the ATR element (24) and to output a corresponding infrared measurement signal; and - at least two infrared bandpass filters (36a, 36b) arranged between the ATR element (24) and the infrared detector (40), each of the infrared bandpass filters (36a, 36b) being designed such that it is only permeable to infrared radiation with a wavenumber within a predetermined wavenumber window (F1, F2); characterized in that a first of the infrared bandpass filters (36a) is permeable to infrared radiation with a wavenumber in a first wavenumber window (F1) of 1005 cm - 1 up to 1025 cm - 1 is permeable. A second of the infrared bandpass filters (36b) for infrared radiation with a wavenumber in a second wavenumber window (F2) of 1075 cm - 1 up to 1095 cm - 1 is permeable.

2. ATR sensor (10) according to claim 1, which is designed to determine the concentration of glucose in a liquid.

3. ATR sensor (10) according to claim 1 or 2, which is designed to determine the concentration of glucose in an aqueous solution, suspension or dispersion.

4. ATR sensor (10) according to claim 2 or 3, wherein the glucose-containing liquid contains CHO cell lines.

5. ATR sensor (10) according to one of claims 1 to 4, which has a third infrared bandpass filter (36c) which is designed for infrared radiation with a wavenumber in a third wavenumber window (F3) of 945 cm - 1 up to 965 cm - 1 is permeable.

6. ATR sensor (10) according to one of claims 1 to 5, which has a fourth infrared bandpass filter (36d) which is designed for infrared radiation with a wavenumber in a fourth wavenumber window (F4) of 1300 cm - 1 up to 1320 cm - 1 is permeable, 7. ATR sensor (10) according to one of claims 1 to 6, wherein a respective one of the first, second, optionally third and optionally fourth infrared bandpass filters (36a, 36b, 36c, 36d) is designed such that a central wave number for the respective wave number window (F1, F2, F3, F4) is determined with an accuracy of at least 5%.

8. ATR sensor (10) according to one of claims 1 to 7, comprising a multi-channel infrared bandpass filter (36) with an infrared bandpass filter for each of the first (F1), second (F2), optionally third (F3) and optionally fourth (F4) wavenumber windows.

9. ATR sensor (10) according to one of claims 2 to 8, which is designed to determine the concentration of glucose in a glucose-containing liquid which has a pH value in a range between 5 and 8, in particular in a range between 6.8 and 7.

4.

10. ATR sensor (10) according to one of claims 1 to 9, wherein the first, second, possibly third and possibly fourth infrared bandpass filter (36a, 36b, 36c, 36d) has a predetermined minimum quality, which is determined by a central wave number for the respective wave number window (F1, F2, F3, F4) being within a Temperature range from 20 °C to 45 °C shifts by a maximum of 5 wavenumbers per 10 degrees Celsius with temperature.

11. ATR sensor (10) according to one of claims 1 to 10, which is designed to detect a reference signal based on infrared radiation emitted by the infrared radiation source (22) without transmission through the ATR element (24).

12. A method for determining the concentration of glucose in a fluid using an ATR sensor (10), comprising: Providing a glucose-containing fluid, Interaction of the glucose-containing fluid with infrared radiation whose wavelength lies in the mid-infrared range, Detecting absorption of infrared radiation after interaction with the glucose-containing fluid in at least two different wavenumber windows (F1, F2) within the mid-infrared range, wherein a first of the wavenumber windows (F1) detects infrared radiation with a wavenumber of 1005 cm - 1 up to 1025 cm - 1 and wherein a second of the wavenumber windows (F2) infrared radiation with a wavenumber of 1075 cm - 1 up to 1095 cm - 1 includes.

13. The method according to claim 12, wherein infrared radiation is detected in a third wavenumber window (F3) in the mid-infrared range, wherein the third wavenumber window (F3) detects infrared radiation with a wavenumber of 945 cm - 1 up to 965 cm - 1 includes.

14. The method according to claim 12 or 13, wherein infrared radiation is detected in a fourth wavenumber window (F4) in the mid-infrared range, wherein the fourth wavenumber window (F4) is for infrared radiation with a wavenumber of 1300 cm - 1 up to 1320 cm - 1 includes.

15. Method according to one of claims 12 to 14, wherein the glucose content of an aqueous solution containing in particular the CHO cell lines is determined.

Images