APPARATUS FOR IDENTIFYING SUBSTANCES

Abstract

The invention relates to an apparatus for identifying one or more substances in a materialin particular, wherein the material is present in granulate formcomprising at least one light sourcepreferably in the form of a laserfor irradiating a sample of the material with light of at least one wavelength, a detector for detecting the light re-emitted and/or transmitted by the sample, and an analysis device for the spectroscopic analysis of the detected light, wherein

the analysis device interacts with the detector, and these are embodied in such a way for analyzing the detected light by means of a) UV/VIS spectroscopy and/or b) fluorescence spectroscopy and/or c) Raman spectroscopy and/or d) absorption spectroscopy,
and for creating a first identification result for at least one substance of the sample and, in the case of an ambiguous first identification result, creating a second identification result for the at least one substance on the basis of a e) fluorescence light decay time analysis, FLZA,
wherein the at least one substance is then at least partly identified on the basis of the first identification result or on the basis of the first and second identification results.

Claims

1. An apparatus for identifying one or more substances in a material, comprising: at least one light source configured to irradiate a sample of the material with light of at least one wavelength; a detector configured to detect light re-emitted or transmitted by the sample; and an analysis device configured to: analyze the detected light by at least one of: UV/VIS spectroscopy fluorescence spectroscopy, Raman spectroscopy, or absorption spectroscopy; generate a first identification result for at least one substance of the sample; and generate a second identification result for the at least one substance in response to the first identification result being an ambiguous identification result, wherein the analysis device configured to generate the second identification result by fluorescence light decay time analysis (FLZA), wherein the analysis device is configured to at least partially identify the at least one substance based on the first identification result or based on the first and second identification results.

2. The apparatus according to claim 1, wherein the at least one light source is configured to emit light of at least two wavelengths.

3. The apparatus according to claim 2, wherein the at least two wavelengths includes a first wavelength having a fundamental frequency and a second wavelength having a frequency that is an integer multiple of the fundamental frequency.

4. The apparatus according to claim 2, wherein a first wavelength of the at least two wavelengths is in a range between 233 nm and 300 nm, and a second wavelength of the at least two wavelengths is double the first wavelength.

5. The apparatus according to claim 1, wherein the at least one light source is configured to irradiate the sample in a first direction, and the detector is configured to detect light re-emitted or transmitted by the sample in a second direction, wherein the first direction and the second direction are different.

6. The apparatus according to claim 1, wherein the analysis device is configured to examine the first identification result for plausibility based on the second identification result.

7. The apparatus according to claim 1, wherein the at least one light source is a pulsed light source having a pulse duration of more than 1 ns and less than 1 ms.

8. The apparatus according to claim 7, wherein the pulse duration of the at least one light source is adapted to a detection time of the re-emitted or transmitted light, and a distance between the light pulses corresponds to the detection time of the re-emitted or transmitted light by the detector.

9. The apparatus according to claim 1, wherein the detector is configured to detect the re-emitted or transmitted light in less than 10 nanoseconds.

10. The apparatus according to claim 1, wherein a transport device is arranged for feeding and discharging the sample.

11. The apparatus according to claim 1, wherein the detector is configured to divide a recorded spectrum into relevant and non-relevant areas for later analysis, discard the non-relevant areas, and insert FLZA-relevant data into the spectrum in place of the non-relevant areas.

12. The apparatus according to claim 1, wherein the analysis device is configured to determine and analyze multi-exponential fluorescence decay time constants by the FLZA.

13. The apparatus according to claim 1, wherein a measurement signal of the detected light, output by the detector, is integrated by the analysis device over a period of time to determine a fluorescence decay time constant.

14. The apparatus according to claim 13, wherein the measurement signal is integrated over a plurality of non-overlappingtime spans by the analysis device.

15. The apparatus according to claim 14, wherein the plurality of non-overlapping time spans have a same duration.

16. The apparatus according to claim 13, wherein the analysis device is connectible or connected to a storage device, the analysis device includes at least one integrator configured to integrate the measurement signal separately over two non-overlapping time spans, wherein values for the separately integrated measurement signal over two non-overlapping time spans are related to one another, and the analysis device is configured to at least partially identify the at least one substance based on the related values of the measurement signal and reference relation values stored in the storage device.

17. The apparatus according to claim 13, wherein the at least one time span corresponds to a falling edge of the measurement signal.

18. The apparatus according to claim 13, wherein the analysis device is configured to selectively integrate the measurement signal over at least one time span prior to or after a normal half-life of a fluorescence lifespan.

19. The apparatus according to claim 1, wherein a fluorescence decay time constant of the at least one substance is determinable, and a duration of light pulses emitted by the at least one light source is less than the fluorescence decay time constant by at least a factor of 5.

20. The apparatus according to claim 1, wherein a sorting device is arranged to separate the identified at least one substance from a material stream of substances for recycling.

21. The apparatus according to claim 12, further comprising an optical grating arranged in a detection beam path of the re-emitted or transmitted light, wherein a zero order of the light diffracted from the optical grating is used by the FLZA for determining the fluorescence decay time constants.

22. The apparatus according to claim 1, wherein the at least one light source is configured to irradiate the sample a plurality of times in succession with the light of the at least one wavelength, and the analysis device is configured to measure and analyze corresponding re-emitted or transmitted light a plurality of times.

23. The apparatus according to claim 16, wherein the two non-overlapping time spans are respectively assigned to successive irradiations of the sample with the light.

24. The apparatus according to claim 1, wherein the detection of the re-emitted or transmitted light is triggered by irradiating the sample with light by the at least one light source.

25. A method for identifying one or more substances in a material, comprising: irradiating, by at least one light source, a sample of the material with light of at least one wavelength; detecting, by a detector, light re-emitted or transmitted by the sample; analyzing the detected light by at least one of: UV/VIS spectroscopy, fluorescence spectroscopy, Raman spectroscopy, or absorption spectroscopy; generating a first identification result for at least one substance in the sample based on the analysis of the detected light; generating a second identification result for the at least one substance based on a fluorescence light decay time analysis (FLZA) in response to the first identification result being an ambiguous identification result; and at least partly identifying the at least one substance based on the first identification result or based on the first and second identification results.

26. The method of claim 25, further comprising: examining, based on a fluorescence light decay time determined by the FLZA, a plausibility of the first identification result.

Description

[0079] Preferred designs and embodiments of the invention are shown in the drawings and are explained in more detail in the following description, wherein like reference numerals refer to like or similar or functionally similar components or elements. The following is shown:

[0080] FIG. 1 in schematic form, an apparatus according to one embodiment of the present invention;

[0081] FIG. 2 in schematic form, parts of a method according to one embodiment of the present invention;

[0082] FIG. 3 in schematic form, parts of an apparatus according to one embodiment of the present invention;

[0083] FIG. 4 Fluorescence decay times for summed fluorescence of polystyrene, without and with different flame retardants;

[0084] FIG. 5 Fluorescence decay time constants of different substances, for summed fluorescence;

[0085] FIG. 6 an overview of a bi-exponential evaluation of fluorescence decay times of different materials;

[0086] FIG. 7 a measurement of a fluorescence decay time constant of polystyrene; and

[0087] FIG. 8 a measurement of a sample of high impact polystyrene.

[0088] FIG. 1 shows, in schematic form, an apparatus according to one embodiment of the present invention, and FIG. 2 shows, in schematic form, parts of a method according to one embodiment of the present invention.

[0089] The reference signs in FIG. 1 are described as follows: apparatus 1, pulse light source 2, sample 3, filter 4, lens 5, sensor 6, signal processor 7, differentiator 8, trigger generator 9, delay 10, timer integrator 11a, 11b, integrator 12a, 12b, quotient calculator 13, evaluator 14, raw signal 60, differentiated signal 80, and trigger signal 90.

[0090] FIG. 1 shows an apparatus for identifying plastic and/or, where appropriate, one or more of its additives 1. The apparatus 1 thereby comprises a pulse light source 2here in the form of a laserwith which a sample 3 of the plastic to be identified is irradiated. The light re-emitted by the sample 3 is detected by a filter 4, along with a lens 5, using a sensor 6. The raw signal received from sensor 6 is processed by means of a signal processor 7, and a differentiated signal is likewise generated by a differentiator 8. A trigger is also generated by means of a trigger generator 9, which then triggers a first timing element integrator 12a and, via a delay 10, a second timing element integrator 12b.

[0091] By means of the two integrators 12a and 12b, the signal (see FIG. 2) is integrated for different, non-overlapping time spans of equal duration on the falling edge of the processed signal. The two values, which are provided by the two integrators 12a and 12b, are set in relation to each other by means of a quotient calculator 13here, by means of the formation of a quotientand the number thus obtained is fed to an evaluator 14.

[0092] The evaluator 14 may consist in, among other things, storing a large number of reference key figures for various combinations of plastics and their additives in a memory of the apparatus, and then identifying the plastic and/or its additives on the basis of a comparison between such key figures and the key figure determined by the measurement. For the storage of such values/key figures, multiple repeated measurements of the same plastic with the same additives can, for example, be carried out, and such measurements can then be stored in the memory with, for example, an average value and a corresponding deviation. In the case of an identification that is ambiguous, such a result can be displayed to a user and, if the method is used for recycling, the plastic can be separately sorted out and then, if necessary, sent to another identification method. Using modern, ultra-fast, analog-to-digital converters, in an additional embodiment of FIG. 1, a digitization of the measured values of sensor 6 directly behind or after the sensor can occur, such that the subsequent evaluation can be carried out flexibly, i.e., digitally, with that application specific to the problem, by means of software.

[0093] FIG. 3 shows, in schematic form, parts of an apparatus according to one embodiment of the present invention.

[0094] FIG. 3 shows a spectrometer with an additive for measuring the fluorescence decay time. In the detection beam pathin particular, between lens 5, sensor 6, and sensor system 22 in FIG. 1an optical grating 21 is additionally arranged; this diffracts the light bundled by the lens 5. The sensor 6, which is intended for the re-emitted light for detecting the fluorescence decay time constant, is arranged in the 0th order of the light diffracted by the optical grating 21. Beside this, the sensor system 22, e.g., in the form of a photomultiplier array, is arranged; it serves to receive the higher orders of diffracted light from the optical grating 21. This is here particularly sensitive in the wavelength range between 250 nm and 750 nmpreferably, between 285 nm and 715 nm. Thereby, the light re-emitted and/or transmitted by the plastic and/or its additives can then also be analyzed in this spectral range.

[0095] FIG. 4 shows fluorescence decay times with a summed fluorescence decay time of polystyrene, without and with different flame retardants.

[0096] FIG. 4 shows different fluorescence decay times for a summed fluorescence with an excitation at a wavelength of 266 nm of impact resistant polystyrene, HIPS (high impact polystyrene), without flame retardants and with different flame retardants. The wavelength of 266 nm can be generated, for example, with a Nd:YAG laser, which features a fundamental wavelength of 1,064 nm, by means of frequency multiplication in an already known manner. FIG. 4 shows significant differences in the fluorescence decay time with different flame retardants. For example, the fluorescence decay time amounts to 10.2 ns for HIPS without flame retardancy, wherein, however, with the addition of the TBBPA and Sb.sub.2O.sub.3 flame retardants, it is smaller by more than halfspecifically, here, 3.9 ns.

[0097] As a whole, FIG. 4 shows that the addition of flame retardants in impact resistant polystyrene has significant effects on the duration of the fluorescence decay time and depends upon the flame retardants used. In particular, the decay time is reduced by flame retardants present in the plasticprobably due to quench effects.

[0098] FIG. 5 shows fluorescence decay time constants of different substances with summed fluorescence.

[0099] FIG. 5 shows the mono-exponential fluorescence decay times T with summed fluorescence for different materialsin particular, plasticsat an excitation wavelength of 266 nm. The following abbreviations are used: [0100] PTFE (polytetrafluoroethylene) [0101] PFA (perfluoroalkoxy polymer) [0102] PLA (polylactic acid/polylactide) [0103] polyamide [0104] POM (polyoxymethylene) [0105] PE-HD (high density polyethylene) [0106] PE-LD (low density polyethylene) [0107] PP (polypropylene, polypropylene copo) [0108] PVC (polyvinyl chloride (,S: soft; ,H: hard)) [0109] PC (polycarbonate (ABS: acrylonitrile butadiene styrene)) [0110] PS (polystyrene) [0111] HIPS (high impact polystyrene) [0112] FR (various flame retardants).

[0113] As FIG. 5 shows, rubber and polyurethane always have very short decay times. These are thus particularly well suited, on the whole, for being identified or distinguished from other plasticsalso due to the possibly weak spectral information of such substances. HIPS and PS have a very long decay time, wherein this is reduced by the presence of FR flame retardantsprobably due to quench effects: without flame retardants, HIPS features a decay time of approximately 11 ns. This is reduced to 4-7.5 ns due to flame retardants. PVC shows a very short decay time compared to other plastics. In this respect, the fluorescence decay constant can be used as additional information for identifying PVC. FIG. 5 also shows that pure ABS always features a longer fluorescence decay time constant than PC or ABS mixed with PC. PC alone features a short decay time. In this respect, the decay behavior can be used to separate ABS from PC or PC-ABS. In contrast, POM has the longest decay time of all the substances listed in FIG. 5. The fluorescence decay time constant of POM can be used to uniquely identify POM.

[0114] FIG. 6 shows an overview of a bi-exponential evaluation of fluorescence decay times of different materials.

[0115] FIG. 6 reproduces the bi-exponential decay times of different plastic materials. In principle, however, a mono-exponential decay time determination can also already be sufficient for certain plastics:

TABLE-US-00001 TABLE 1 Plastic material T T.sub.1 T.sub.2 No. Polymethyl methacrylate (PMMA) 0.841 0.124 3.669 1 Polystyrene (PS) 3.290 0.171 4.457 2 Polycarbonate (PC) 1.038 0.077 4.379 3 Polyethylene terephthalate (PET) bottle.sup.a) 1.840 1.176 4.205 4 Polyethylene terephthalate (PET) plate 4.466 1.387 8.933 5 Polyethylene LDPE 2.19 0.456 4.655 6 Polyethylene HDPE <0.2 0.155 4.238 7 Polyethylene UHDPE 1.58 0.217 4.932 8 Silicone Tectosil granulate 0.132 7.709 9 Silicone Tectosil film 0.084 8.572 10 Silicone dehesive Sn 3.078 1.432 6.825 11 Silicone dehesive Pt (1) 3.162 1.473 6.149 12 Silicone dehesive Pt (2) 3.114 1.707 6.106 13 Silicone tubing 4.333 1.793 8.180 14 .sup.a)PET beverage bottle of a known manufacturer of lemonade drinks.

[0116] Table 1 above shows the fluorescence decay constants in ns of various technical polymers averaged over 10 s measurement time. Fluorescence excitation took place at a wavelength of 403 nm, mono-exponential evaluation: , bi-exponential evaluation: .sub.1 and .sub.2.

[0117] It can be seen from Table 1 that a classification is basically already possible via the mono-exponential decay time (nos. 1 through 8); for silicones, this is difficult simply due to the partially very short decay time (nos. 9 through 10). If both decay constants (.sub.1 and .sub.2) are taken into account, a classification is much simpler and more reliable, and it is possible to assign not only the plastic itself, but also special batches, such as silicone dehesives (nos. 11 through 13) or special processing forms (nos. 4 and 5 and nos. 9 and 10). This can be clearly seen in the two-dimensional application of .sub.2 against .sub.1, as shown in FIG. 6; clusters are then found, such as, for example, for the various dehesive films. Through such an analysis, together with the time constant of the mono-exponential evaluation, more difficult classifications can be madefor example, for the various types of polyethylene.

[0118] The fluorescence decay times were thereby determined over a period of 10 s by integrating the fluorescence, and resulted in highly reproducible values, even with different plastic samples. The measuring time could also be reduced to 1 ms without any problems, and the measured values were only slightly more scattered.

[0119] For fluorescence excitation, the usual light sources for short light flashes can be used, e.g., gas discharge lamps (flash lamps)preferably with hydrogen-containing gas fillingsor semiconductor lasers, which are available in a variety of forms and can be operated without any complications. The pulse duration can be in the range of nanoseconds, and the temporal progression of the fluorescent light in response to the excitation pulse can be obtained in a known mannerfor example, by means of deconvolution. For this purpose, it is not necessary, but advantageous, if the duration of the excitation pulse is considerably shorter than the fluorescence lifespan; this can be achieved efficiently with semiconductor lasers.

[0120] The fluorescence lifespan spectra were recorded with a PicoQuant FluoTime 300. The light source was a PC-405 laser, controlled with a PicoQuant PicoHarp 300, with 0.4 mW power at a pulse frequency of 20 MHz and an excitation wavelength of 403 nm.

TABLE-US-00002 TABLE 2 Determination of the detection wavelength. Selected detection Sample t.sub.Fluo1 t.sub.Fluo2 wavelength PMMA 2-3 ns 4-100 ns 440 nm PS 4-6 ns 6-100 ns 490 nm PC 2-3 ns 4-100 ns 440 nm Tectosil granulate 2-3 ns 4-100 ns 480 nm Tectosil film 2-3 ns 4-100 ns 480 nm PET 2-5 ns 8-100 ns 450 nm PE (HDPE) 2-3 ns 4-100 ns 500 nm

[0121] The detection wavelength was determined by recording a lifespan-dependent fluorescence spectrum. In the process, at the corresponding spectral wavelengths, the respective fluorescence intensities of the fluorescence components with predominantly short (t.sub.Fluo1) and predominantly long (t.sub.Fluo2) fluorescence lifespans were recorded. The selected ranges result from the nature of the sample measured and are not necessarily representative of the individual fluorophores contained in the sample. Detection was subsequently performed at the wavelength corresponding to the fluorescence maximum of the components with long fluorescence lifespans (Table 2).

[0122] The measurement time of the fluorescence lifespan determination amounted to 1.0 ms or 10 s. The decay curves thus obtained were evaluated with the FluoFit software program from PicoQuant. For this purpose, the maximum of the measurement curve was determined, and considered with respect to two abscissa sections. For one, that between the curve maximum (t.sub.max) and the corresponding time point x ns after the detected maximum intensity (t.sub.xns), resulting in the fluorescence lifespan .sub.1. The second abscissa section comprises, beginning with t.sub.yns, a range of 40 ns (up to t.sub.zns),from which the fluorescence lifespan .sub.2 results (Table 3). The fluorescence lifespans are obtained through an exponential tail fit of the respective curve section according to the formula I.sub.rel=A*e.sup.t/.

TABLE-US-00003 TABLE 3 Selection of time ranges as a function of the decay curve for a bi-exponential tail fit. Sample t.sub.max-t.sub.xns t.sub.yns-t.sub.zns PMMA 0-0.5 ns 3-43 ns PS 0-0.5 ns 3-43 ns PC 0-0.5 ns 3-43 ns Tectosil granulate 0-0.5 ns 2-42 ns Tectosil film 0-0.5 ns 2-42 ns PET 0-3 ns 5-45 ns PE 0-3 ns 3-43 ns

[0123] The excitation structure of the laser was not taken into account, due to its small halt-width. Thus, the values obtained are not absolute, but must be adapted to the specific device. In the case of the remaining silicones (dehesive, tubing), the deconvoluted data were evaluated bi-exponentially for the purpose of better reproducibility. This ensures a higher level of reliability during identification.

[0124] In the following, the fluorescence decay time constants of PET drinking bottles are determinedin particular, PET water bottles and their shredder materialin order to distinguish them from PET material contaminated with oils and other lipophilic substances such as gasoline, diesel fuel, and lubricating oils (motor oil). Such contact may have occurred in accordance with the function or also through misusefor example, the unauthorized filling of fuels in drinking bottles. This resulted in the following fluorescence decay times:

TABLE-US-00004 TABLE 4 Fluorescence decay constants T, T.sub.1, T.sub.2 in ns of PET materials with various impurities averaged over a 10 s measurement duration. Fluorescence excitation at 403 nm, detection at 450 nm. Mono-exponential evaluation: T; bi-exponential evaluation: T.sub.1 and T.sub.2. PET material T T.sub.1 T.sub.2 No. Beverage bottle.sup.a) 1.840 1.176 4.205 1 Recycling flakes.sup.b) 1.867 0.980 5.301 2 Bottle, contact with diesel.sup.a,c) 0.971 0.937 3.479 3 Bottle, contact with diesel, washed.sup.a,c,d) 0.994 0.947 3.608 4 Bottle, contact with motor oil.sup.a,c) 1.020 0.973 4.166 5 Bottle, contact with motor oil, washed.sup.a,c,d) 1.060 1.028 3.662 6 PET plate 4.466 1.387 8.933 7 .sup.a)PET beverage bottle of a known manufacturer of lemonade drinks. .sup.b)commercial PET recycling flakes. .sup.c)contact with foreign matter after one week. .sup.d)washed material; see experimental part.

[0125] One can see from Table 4 that, for PET standard beverage bottles, decay constants of around 1.8 ns are obtained ( from nos. 1 and 2), which, surprisingly, become considerably shorter at 1 ns when contaminated with mineral oil products such as diesel oil or motor oil (nos. 3 and 5). Careful washing does little to change this (nos. 4 and 6). One PET plate (no. 7) showed significantly higher fluorescence decay times. If bi-exponential components (.sub.1 and .sub.2) are taken into account, high values .sub.1 for untreated material (nos. 1, 2, and 7) can be found in an analogous manner; these drop significantly when treated with mineral oil products (nos. 3 and 5), and, even after careful washing, no longer reach the original values (nos. 4 and 6), and a completely analogous picture arises for the decay time .sub.2 and thus allows a classification on the basis of two sizes. The shortening of the mono-exponential decay constant of contaminated material is partly due to the fact that the long-lived, bi-exponential component, characterized by .sub.2, turns out to be significantly smaller.

[0126] Here, the fluorescence decay times were determined over a period of 10 s by integrating the fluorescence, and resulted in highly reproducible values, even with different samples. The measuring time could be reduced to 1 ms without any problems, and the measured values were only slightly more scattered.

[0127] The detection wavelength was determined by recording a lifespan-dependent fluorescence spectrum from a PET derivative. In the process, at the corresponding spectral wavelengths, the respective fluorescence intensities of the fluorescence components with predominantly short (t.sub.Fluo1=2-5 ns) and predominantly long (t.sub.Fluo2=8-100 ns) fluorescence lifespans were recorded. The selected ranges result from the nature of the sample measured and are not necessarily representative of the individual fluorophores contained in the sample. Detection was subsequently performed at the wavelength corresponding to the fluorescence maximum of the components with long fluorescence lifespans (450 nm).

[0128] The measurement time of the fluorescence lifespan determination amounted to 1.0 ms or 10 s. The decay curves thus obtained were evaluated with the FluoFit software program from PicoQuant. For this purpose, the maximum of the measurement curve was determined, and considered with respect to two abscissa sections. For one, that between the curve maximum (t.sub.max) and the corresponding time point 3.0 ns after the detected maximum intensity (t.sub.3ns), resulting in the fluorescence lifespan .sub.1. The second abscissa section comprises, beginning with an abscissa section 5 ns behind the maximum (t.sub.5ns), a range of 40 ns (up to t.sub.45ns), from which the fluorescence lifespan .sub.2 results. The fluorescence lifespans are obtained through an exponential tail fit of the respective curve sections according to the formula I.sub.rel=A*e.sup.t/.

[0129] The excitation structure of the laser was not taken into account, due to its small half-width. Thus, the values obtained are not absolute, but must be adapted to the specific device. However, the described method explains the reliable differentiation of the different samples, and can be adapted to other experimental arrangements.

[0130] Cleaning was initially carried out by manual wiping with cloth cloths; compact PET parts were shredded into flakes. The PET flakes were then washed in a mixture composed of a 3% aqueous NaOH solution (100 mL) and a 15% aqueous sodium dodecyl sulfate solution (50 mL) at 85 C. for 2 h while stirring. Finally, the test specimens were dried with cloth paper, air pressure, and then for 16 h at 60 C.

[0131] Mono- and bi-exponential fluorescence decay times were also determined for polyethylene:

[0132] For LDPE (high pressure polyethylene), the longest fluorescence decay time of 2.19 ns was found for mono-exponential application, which differs so significantly from the decay time of other polyethylene species that easy detection is clearly possible. The fluorescence decay times of the low-pressure polyethylene types, HDPE and UHDPE, are considerably shorter. However, a distinction is also possible here, and, for UHDPE, 1.58 ns is found, and finally, for HDPE, the shortest decay time of less than 0.2 ns is found; for the latter material, a further supporting process may be desirable due to the very short decay time. The fluorescence decay curves of polyethylenes contain high bi-exponential components, and a bi-exponential evaluation of the curves shows 0.456 ns and 4.655 ns for LDPE, 0.155 ns and 4.238 ns for HDPE, and 0.217 ns and 4.923 ns for UHPE; see Table 5 below. Using the bi-exponential components, the classification of polyethylenes is considerably simplified and significantly more reliable.

TABLE-US-00005 TABLE 5 Fluorescence decay constants T in ns of various PE materials averaged over a 10 s measurement duration. Fluorescence excitation at 403 nm, detection at 450 nm. Mono-exponential evaluation: T, bi-exponential evaluation: T.sub.1 and T.sub.2. The standard deviation is given in parentheses, with an unfavorable integration time of 1 ms and 10 independent material samples. PE material T T.sub.1 T.sub.2 No. LDPE 2.19 0.456 (0.039) 4.655 (0.93) 1 HDPE <0.2 0.155 (0.013) 4.238 (0.77) 2 UHDPE 1.58 0.217 (0.022) 4.923 (0.91) 3

[0133] The values .sub.1 and .sub.2 in Table 5 have been determined at an integration time of 10 s and are to be regarded as reliable mean values. To estimate the effect of measurement errors, the standard deviations for the less favorable integration time of only 1 ms were determined from the measurements of 10 independent samples. Even considering the now unfavorable boundary conditions, a clear identification of the polymer materials is possible.

[0134] The detection wavelength for PE was here determined by recording a lifespan-dependent fluorescence spectrum of a PE derivative (HDPE). In the process, at the corresponding spectral wavelengths, the respective fluorescence intensities of the fluorescence components with predominantly short (t.sub.Fluo1=2-3 ns) and predominantly long (t.sub.Fluo2=4-100 ns) fluorescence lifespans were recorded. The selected ranges result from the nature of the sample measured and are not necessarily representative of the individual fluorophores contained in the sample. Detection was subsequently performed at the wavelength corresponding to the fluorescence maximum of the components with long fluorescence lifespans (500 nm).

[0135] The measurement time of the fluorescence lifespan determination again amounted to 1.0 ms or 10 s. The decay curves thus obtained were evaluated with the FluoFit software program from PicoQuant. For this purpose, the maximum of the measurement curve was determined, and considered with respect to two abscissa sections. For one, that between the curve maximum (t.sub.max) and the corresponding time point 3.0 ns after the detected maximum intensity (t.sub.3ns), resulting in the fluorescence lifespan .sub.1. The second abscissa section comprises, beginning with t.sub.3ns, a range of 40 ns (up to t.sub.43ns),from which the fluorescence lifespan .sub.2 results. The fluorescence lifespans are obtained through an exponential tail fit of the respective curve sections according to the formula I.sub.rel=A*e.sup.t/.

[0136] The excitation structure of the laser was again not taken into account, due to its small half-width. Thus, the values obtained are not absolute, but must be adapted to the specific device. However, the described method explains the reliable differentiation of the different samples, and can be adapted to other experimental arrangements.

[0137] In addition, mono- and bi-exponential fluorescence decay times were also determined for various silicone materials.

TABLE-US-00006 TABLE 6 Fluorescence decay constants T in ns of various silicone materials averaged over a 10 s measurement duration. Fluorescence excitation at 403 nm, detection at 450 nm. Bi-exponential evaluation: T.sub.1 and T.sub.2. Silicone material T T.sub.1 T.sub.2 No. Tectosil granulate 0.132 7.709 1 Tectosil film 0.084 8.572 2 Silicone dehesive Sn 3.078 1.432 6.825 3 Silicone dehesive Pt (1) 3.162 1.473 6.149 4 Silicone dehesive Pt (2) 3.114 1.707 6.106 5 Silicone tubing 4.333 1.793 8.180 6

[0138] One can see from Table 6 that the silicones can be clearly classified via the two decay constants (.sub.1 and .sub.2). Two clusters are obtained by obtaining, for the commercial silicone elastomer Tectosil (nos. 1 and 2), short decay constants .sub.1, and longer decay constants for dehesive materials (nos. 3 through 5). The processing of Tectosil has a smaller, but characteristic influence (nos. 1 and 2). The decay constants .sub.2 are very long compared to other polymers and can be used for the allocation to silicones, and also for fine allocation. This size also makes it possible to distinguish the manufacturing of dehesive materials, by finding a longer constant for tin catalysts (no. 3) and a shorter constant for platinum catalysts (nos. 4 and 5). In its data, a commercially available silicone tubing (no. 6) corresponds more to the dehesive materials, but can be clearly distinguished from these; materials of various manufacturers and for various purposes of use can be efficiently distinguished and classified. FIG. 5 shows this in two dimensions.

[0139] The fluorescence decay times were determined over a period of 10 s by integrating the fluorescence, and resulted in highly reproducible values, even with different samples. The measuring time could be reduced to 1 ms without any problems, and the measured values were only slightly more scattered.

[0140] The fluorescence lifespan spectra were recorded with a PicoQuant FluoTime 300. The light source was a PC-405 laser, controlled with a PicoQuant PicoHarp 300, with 0.4 mW power at a pulse frequency of 20 MHz and an excitation wavelength of 403 nm.

[0141] The detection wavelength was determined by recording a lifespan-dependent fluorescence spectrum of Tectosil (granulate). In the process, at the corresponding spectral wavelengths, the respective fluorescence intensities of the fluorescence components with predominantly short (t.sub.Fluo1=2-3 ns), and predominantly long (t.sub.Fluo2=4-100 ns) fluorescence lifespans were recorded. The selected ranges result from the nature of the sample measured and are not necessarily representative of the individual fluorophores contained in the sample. Detection was subsequently performed at the wavelength corresponding to the fluorescence maximum of the components with long fluorescence lifespans (480 nm).

[0142] The measurement time of the fluorescence lifespan determination amounted to 1.0 ms or 10 s. The decay curves thus obtained were evaluated with the FluoFit software program from PicoQuant. For Tectosil, the maximum of the measurement curve was determined, and considered with respect to two abscissa sections. For one, that between the curve maximum (t.sub.max) and the corresponding time point 0.5 ns after the detected maximum intensity (t.sub.0.5ns), resulting in the fluorescence lifespan .sub.1. The second abscissa section comprises, beginning with an abscissa section 2 ns behind the maximum (t.sub.2ns), a range of 40 ns (up to t.sub.42ns),from which the fluorescence lifespan .sub.2 results. The fluorescence lifespans are obtained through an exponential tail fit of the respective curve sections according to the formula I.sub.rel=A*e.sup.t/. In the case of the remaining silicones, the deconvoluted data was evaluated bi-exponentially for better reproducibility. This ensures a higher sorting reliability.

[0143] In general, the detection of polymers via the fluorescence decay constant can be used to sort materials for recyclingfor example, thermoplastics, where reuse can be done easily. In addition, it is also advantageously applicable to plastics that are to be chemically processed, e.g., when using consumed Thermodurs, since it is then possible to feed the processes with a uniform starting material, with which they can then be operated in a manner that is more stable. Here, it is also possible to recover targeted valuable substances, such as platinum catalysts, if they are used in certain processes, because their products are then recognized. Finally, the method can also be used outside of recycling to optically detect plastics, e.g., during product inspectionparticularly for high-quality end products, with which various starting materials are combined.

[0144] To determine the fluorescence decay times, it is not necessary to record the entire exponential progression; rather, two or three punctiform or integrating intensity measurements (cumulationsummationof the individual measurements, each over a defined duration) at different times are sufficient. For bi-exponential progression, three intensity measurements are required. In principle, additional measurements are also possible. This can further improve accuracy.

[0145] In doing so, the integration over a defined time span (expediently, measurements are taken prior to the first half-life and after the first half-life) is particularly advantageous, because the signal-to-noise ratio is significantly improved (the fluorescent light of the sample is used more efficiently). Assuming normal fluorescence decay times of approximately 5 ns, the measuring processes can be performed at intervals of one to two nanoseconds, with integration times of approximately one to two nanoseconds. Measurements with such a temporal resolution pose no problem in terms of electronics. The measuring process can be further simplified if the plastic sample is optically excited, not only once, but periodically. It can be assumed that, after approximately ten half-lives, the optical excitation has decayed to such an extent that a new excitation can take place without any disturbance; if one assumes an unfavorable case with a fluorescence decay time of 10 ns, this is achieved after approximately 70 ns. As such, the plastic sample can be optically excited periodically with a pulse sequence of 70 ns, i.e., with a repetition frequency of approximately 15 MHz. The two measurements for the determination of the fluorescence decay times can be carried out with a time shiftin particular, with successive pulsesand the requirements on the electronic components for the evaluation are thus further reduced.

[0146] The measurement can, advantageously, be carried out within a fluorescence decay process by the detection of the required integrated signals undertaken with a time-shifted periodic excitation, and in a manner triggered by the excitation pulse; here, for example, a separation can be carried out with several phase-sensitive detectors (PSD) working in parallel, via which the intensities can be measured by integrating over different time ranges of the decay curve. In doing so, it is not necessary to determine the absolute decay time. Device-specific raw data can also be used here, as long as it is sufficiently reproducible; with all devices used here, excellent reproducibility of the measured values has resulted, even as raw data (for example, not corrected by deconvolution). The unproblematic use of raw data simplifies the method even further.

[0147] Assuming that a recycled flake is 10 mm maximum in unfavorable cases, and that a 20 mm space is left between two flakes for good measure, more than 1,000 excitation pulses per flake would be available at a pulse sequence of 15 MHz and a feed rate of 140 m/s (for technological reasons, it should remain significantly below the speed of sound). If the pulses are averaged, the signal-to-noise ratio can be significantly improved, thus further increasing detection reliability. With a mass of 25 mg for a recycled flake (the value was obtained by averaging flakes from commercially available technical recycling material), one can, for example, then sort half a ton of material per hour with a sorting line without any problems. In many cases, such a high sorting capacity is not required; at lower sorting capacities, the demands on electronics and mechanics are considerably lower.

[0148] Using the fluorescence decay time of self-fluorescencemono- and bi-, tri-, or higher-exponentially evaluatedplastics and their batches can be clearly identified and, in this way, mechanically sorted accurately by type for recycling purposes. This applies to all macromolecular substances that can thus be reused. Due to the high speed of the recognition process, possibilities are opened up for sorting systems with high material throughput. By using phase-sensitive detectors and integral measurements, the electronic effort for detection units can be designed favorably. Finally, the method for the detection of macromolecular substances can be used not only for recycling, but also for applications such as product inspection.

[0149] Using the fluorescence decay time of self-fluorescencemono- and bi-, tri-, and/or higher-exponentially evaluatedPET materials can be clearly identified with respect to contamination by their previous use, and in this manner mechanically sorted accurately by type; this is of particular importance for distinguishing PET material contaminated with mineral oil products from non-contaminated material, above all, for use in the food industry. Using the fluorescence decay time of self-fluorescencein particular, bi-, tri-, and/or higher-exponentially evaluatedsilicone materials can be classified and distinguished, as shown here with silicone elastomers and dehesive films. A refined evaluation allows conclusions to be drawn regarding the processing of the respective silicone. The different catalysts for the manufacturing of silicone dehesive material are reflected, in particular, in the .sub.2 time constant, which can be used to efficiently detect auxiliary materials such as platinum catalysts and recover them by type. Finally, the method can also be used for routine product inspection, as it can be easily automated.

[0150] Using the fluorescence decay time of self-fluorescencemono-() and bi-(.sub.1, .sub.2), tri-, and/or higher-exponentially evaluatedthe PE materials, LDPE, HDPE, and UHDPE, for example, can also be clearly identified and sorted by type.

[0151] In summary, a FLZAthus, in particular, with the assistance of the fluorescence decay time constantallows substances and/or their additives in or on materials to be easily and reliably distinguished. In particular, the measurement of the fluorescence lifespanboth in the case of a mono-exponential evaluation and in the case of a bi-, tri-, or even higher exponential evaluationis possible in an easy manner with, at the same time, a high degree of reliability upon its determination. An additional advantage is that a FLZAin particular, the measurement of the fluorescence lifespancan be easily implemented, and can thus be used for the recycling of large quantities of plastics in particular. The FLZA and its evaluation requires only a few nanoseconds, such that, for example, plastic flakes can be reliably irradiated with light, e.g., on a conveyor belt, and the re-emitted light can then be used for the FLZA.

[0152] FIG. 7 and FIG. 8 show intensity distributions and fluorescence lifespan measurements of polystyrene (FIG. 7) and high impact polystyrene (HIPS) (FIG. 8). The upper left half shows a two-dimensional intensity distribution (continuous plot). At the bottom left, the corresponding one-dimensional intensity distribution is shown along a line in the spectrum from left to right. The intensity of fluorescence as a function of time is plotted on the upper right. This results in a decay time constant of 7.7 ns for polystyrene (FIG. 7) and, in FIG. 8, a decay time of 10.2 ns for high impact polystyrene.

[0153] In summary, the invention features the advantage, among others, that it enables a particularly reliable identification of substances and/or their additives. In addition, the invention has the advantage that it can be used to identify plastics in particularespecially in the field of recyclingquickly and in large quantities, and in an extremely reliable manner and, if necessary, to separate them from a material stream for further processing. An additional advantage is that substancesin particular, plastics or chemically similar substancescan be reliably distinguished.

[0154] With regard to other advantageous embodiments of the device according to the invention, to avoid repetition, reference is made to the general part of the description and also to the accompanying claims.

[0155] Finally, it must be expressly pointed out that the embodiments of the apparatus in accordance with the invention described above serve only as a discussion of the claimed teaching, but do not restrict it to the embodiments.