SE1100262A1 - Instrument and method for determining multiple coupled optical quantities in a measuring volume - Google Patents

Abstract

The invention describes an optical system for measuring optical quantities of a sample, which is preferably located in a transparent tube, which is surrounded by light sources and detectors in a pattern which allows the use of different measuring distances and scattering angles. Furthermore, the invention provides a method of utilizing the instrument for analysis and process control. Analytical methods are preferably of a multi-varied nature. Applications include blood and urine analysis, water quality measurements, analysis of dairy products and other foods, chemical process fluids etc.

Description

Measurement problems in applied optics Problems with cross-talk and bilateral interference arise in almost all measurements of an optical property in a measuring volume such as a cuvette or tube in an industrial plant.

This is particularly problematic in cases where the composition of the sample may change randomly, as is often the case with liquid samples from the process industry, medicine or environmental monitoring. To illustrate this, we give some examples of cases where an optical property cannot be measured correctly without measuring all other disturbing properties. Ju greater precision required, the more physical phenomena must be taken into account.

Survival index measurement ~ This category of measurements is often used in process control to estimate the salt or sugar content of a sample. The measurement is based on Snell's law for total internal reaction or Fresnel's equations, where the reacted intensity IOt) in a given angle and polarization are compared with the incident intensity IOOt). Fresnels however, equations apply only to surfaces between two different non-absorbent media.

Refractive index nOt) and absorption depend on each other according to Kramer-Kronig's relations.

Accurate measurements of both parameters can be achieved by ellipsometry which often is a time-consuming measurement method. When turbidity or dispersive media, which generate diffuse re ectancy, or fl uorescence are involved are none of the above mentioned techniques applicable. When all sample volumes are finite, there will be speculative reactions R (7t) = I () t) / I0 (7 »), to influence light measurements based on the ratio of the light enters and the light leaving the sample. Furthermore, R is also related to the electrical the field, U, if one is imposed, and the optical emissivity, sOt).

Additional exotic specimens such as photonic crystals, microstructural dyes, or surface plasmas can also generate speculative reactions.

Qip measurement of absorption - The absorption of a sample provides information about the sample chemical composition as different molecules or atoms may have different absorption spectra. In classical absorption measurements, the transmitted intensity is IOt) related to the incident intensity according to Beer-Lambert's law, and determined by the product of the absorption coefficient ptabsOt), the concentration C and the distance light travel through the medium, x. The transmittance TOt) is consequently obtained according to T (>,) = log / roa) = e '* ”*'“ <*> * C '*.

The measurement is performed usually according to the standard addition method, by adding known concentrations of a sample in, for example, a cuvette. It is a time consuming procedure, and as we have already noted, the amount of reflected light comes from the first surface of the cuvette depends on the absorption. Furthermore, the amount of light comes from the secondary the reaction when light leaves the sample also due to the absorption. There are, of course infinitely many reactions back and forth with decreasing significance. It is now it is obvious that a given measure of absorption will change if one adds to example non-absorbent sugar to the sample as this would change the refractive index.

The above is a classic example of spectroscopic cross-talk. The problem can be solved through the use of integrating spheres, where both T and R are measured. This requires that the sample can be measured twice in time or that two integrating spheres - relatively expensive equipment - used.

By adding light scattering to the problem, we realize that the attenuation also changes as a result of scattering processes that change the propagation of light and partially prevent it it from reaching the detector. Photon migration can to some extent be approximated with the diffusion equation, or better yet, through Monte-Carlo simulations. Again we find that the above methods are underdetermined. We now need an extra measuring point, namely collimated and non-collimated transmission to distinguish between refractive index, absorption and dispersion. The problem can to some extent be solved by k peak time spectroscopy, where the distance traveled by each received photon in the sample appreciated. The technology involves pulsed lasers and time-resolved detection electronics.

The cost of the instruments for time and frequency resolved spectroscopes is extremely high, and the technology in many cases requires considerable measurement time.

When adding fl ourescens to the list of optical phenomena that occur at absorption measurements, we realize that absorbed light can be re-emitted at longer wavelengths.

Neglecting this fact can result in T and R above 100% or result in negative concentrations when Beer-Lambert's law is applied. None of the above available techniques solve this problem, as both incident and resulting light must dissolved in wavelength. In other words, only data on the diagonal of the emission the excitation matrix (EEM; a matrix where the elements EMU., m indicate the amount of the resultant light at km where the incident light has the wavelength Åex) of the sample should contribute to the absorption measurement. Most techniques sum up along rows or columns and neglects fl uorescence peaks not located on the diagonal. Next comes photon delays in high-speed spectroscopy now not only due to the distance traveled but also at fl uorescence lifetimes.

The dynamic range of an absorption measurement is usually limited. Measurements based on T assumes that T assumes a value between 0% and 100% and is not close any of the boundaries. The minimum amount of substance that can be detected is limited of when T can not be separated from 100% and the maximum concentration that can be measured limited by when T can not be distinguished from 0%. Therefore, the distance traveled must often be optimized for a specific scenario for T to vary around 50%.

Measurement of scattering - The probability of scattering ptScaOt), the anisotropic the scattering and polarization dependence provide information about the microstructure of a sample. Information on particle size, concentration and refractive index for particles often measured in, for example, dairy products. Dispersion processes in the optical area described by Rayleigh and Mie theory. Dispersion is strongly correlated with absorption and refractive index, which we discussed above. A constant refractive index is often assumed over different wavelengths and poorly conditioned iterative algorithms are needed to produce the spread from the measured properties. Assumptions are also made about the spectral dependence on spread.

Measurement of uorescence - Collection of one or fl era orescence spectra in an EEM gives detailed information on the chemical composition of a sample. In case there absorption can not be used to distinguish the substance are ores uorescence measurements o fi a one naturally the next step. Even in cases where you can not use EEM matrices, you can nader nn differences in the orescence lifetime for each individual EEM element. Chlorophyll, caffeine, Advanced Glycinated End products (AGEs), oils and soap products are typical fl uorophores. EEMs are traditionally measured with two scanning monochromators aimed at a cuvette in 90 ° observation configuration. Such measurements are very time consuming.

Lifespan is measured with pulsed lasers in combination with streak cameras or with frequency domain methods. Consequently, only a few columns in the EEM can be measured. Generally speaking, is EEM measurements are unitless, which reflects the fact that the measurement is affected by everyone above-mentioned optical properties not measured in a traditional EEM array. One EEM measurement is based on excitation light with the wavelength M being transported through the sample to a or uorophore where it is absorbed, re - emitted with the wavelength X2 and transported through the sample to the detector. Transport is controlled by nOt), uabsOt), uscaOt), g, as well as polarization effects in scattering and fluorescence for both M and 9.2. Aside from difficulties with modeling, we will find that we again have a subordinate equation to estimate the absolute ores uorescence in EEM. Even in laboratory environments with a chemically purified sample, cross-talk occurs in professional EEM measuring systems where 90 ° configuration is used: Fluorescence is a consequence of absorption and in those cases the absorption is too high, no excitation light will reach the field of view for it in 90 ° placed detector. If the absorption is too low, almost all excitation light will be cross the field of view and generate very little fl uorescence. This means that the measurements must be arranged for internal absorption, which can be difficult in a scenario with real samples.

In addition to being affected by photon transport, EEM measurements can also be due to excitation exposure such as bleaching and photokinetics, and usually EEM measurements also depend on sample temperature, Tsampl ...

According to the above discussion, one needs to use an integrated measurement method such as includes everything. A solution to the problems of spectroscopic cross-talk and many others difficulties that arise when optical spectroscopy is implemented on many samples from process control in real life is presented in this patent application. LEDs for commercial use is a field that is developing rapidly with better emission yields, faster rise times as well greater spectral range each year. LEDs are cost-effective and stable light sources for spectroscopy. Use of LEDs in combination with detector pairs for transmission, turbidity or ores uorescence measurements have been presented. These instruments usually exist of a rectangular cuvette with a 0 °, 90 ° or 180 ° observation angle relative to the light source and the spectral resolution are usually determined by the LEDs used the units' emission band. Each detector measures a value typically associated with it categories; transmission, lateral spread or fl uoroscence. From the above the discussion, we understand that we need as many values as possible to be able to appreciate all the processes involved. According to traditional thinking, we therefore need one very large number of detectors in all possible configurations. Alternatively, we can bring in ideas regarding combinatorial light beam measurement from interstitial photodynamic therapy (IPDT) where the number of values corresponds to all possible combinations between sources and detectors rather than just the number of source-detector pairs. That angle of attack would give so as many measured values as possible. The measured values then serve as a basis for estimation of a number of "orthogonal" optical properties. We have also seen fl era examples of fast time and frequency domain methods with LEDs, where the measurement time can come down on a time scale of under a nanosecond. By applying such methods to the above-mentioned LED multiplexing, we would increase the number of measured values for it optical characterization of the sample, and fl uorescence lifetime values, scattering delay as well as photoinduced kinetics would be possible to appreciate.

We will now present a possible but non-limiting implementation of combinatorial light beam multiplexing with LEDs and explain why each physically phenomena are incorporated by that measurement method. Measured values are entered into multivariates algorithms and calibrated with known training sets for orthogonalization, to do parameters independently. We assume that the sample fl surfaces in a tube in the direction of Z (Fig. 1). The tube can be transparent or of blasted spreading frosted glass and is optical transparent to the entire spectral range covered by the instrument. Rows of LEDs are arranged along the Z axis and illuminates the tube containing the sample. In each position along the Z-axis there is a spectral wavelength band which is av niered by the band gap of the LED device at that Z position. For a given period of time, only one LED will flash to, fed well above its continuous dri fl level. The intensity of the light is measured by everyone detectors. After a sequence where all the LEDs flash, the total number comes measured values to be the product of the number of detectors and sources.

At each z-position there will be at least two sources and two detectors (Fig. 2): a detector with wide sensitivity that covers ÄZ at the z position and a detector with a blocking filter which allows long-wave radiation to pass and which prevents hl detected at the z position. It follows that at each z position four nos are measured identical light distances; transmitted light, laterally scattered light, laterally detected fl uorescence and fl uorescence detected in the forward direction. The filter may be of absorption, interference or wedge type.

When we observe light emitted in X1 at position z; in the XZ domain (Fig. 3) we understand how the light is scattered along the Z axis partly by the walls of the tube or the sample itself. This illustrates how each transmittance coefficient is measured with different distances in the sample. This is appropriate to expand the dynamics of the spectrometer (see discussion above). We also understand that we lead each wavelength is given a spatial profile along the x-axis for both transmitted light, laterally scattered light, laterally detected fl uorescence and fl uorescence detected in the forward direction. These spatial profiles will have different shapes depending on the main optical of the sample character (reflective, absorbent, cloudy or uorescent). Then we can gradually change the detector type from UV-optimized Si, to visible, to InAsP, to InGaAs or to microbolometers, the instrument can cover a much wider spectral range than any of the instruments using a single detector type.

Then any ores uorescence light is not only measured by the detector at one and the same z position but from all z positions (Fig. 4) it is possible to measure not only the total ores uorescence of the columns in the upper triangle of the EEM matrix, but also each element.

The resolution of the EEM will be the number of detectors multiplied by the number sources. We can also note that the cut-off filters at position z are selected so that they break in the middle of the band at z + 1 (Fig. 5). The spectral resolution of transmittance, reflctance and the spread will thus be higher than the number of sources then the transmission bands they are av driven by the product of the emission and sensitivity of the source and detector, respectively.

In the preferred but not limited set, we demonstrate spectral coverage from 350 to 1700 nm (Fig. 5). The Gaussian problems indicate spectral emission bands from the selected LED units. The transmission of the ores uorescence filters is marked for edge fi ltren, which releases longer wavelengths. Several of the bands are divided so that the whole, respectively, the upper part of the band is isolated, which provides additional spectral information.

Dashed lines indicate the broad sensitivity curves of the selected detectors. The thick line indicates transmission for 5 mm H20. Dotted lines indicate transmission for chlorophyll and milk.

If the sample consists of air, water or mercury, it is obvious that we will get different reactions at the surface of the quarter tube (Fig. 6). In the case of Hg, which has zero transmission we understand that we will still get a signal as the photons move in all directions in the test tube itself. If you take the shapes of the four spatial profiles along the Z-axis in taking into account, one can assume a first pure reflectance profile (the signals from the detectors with cut-off vid lter at zn, zn + 1, Zn + 2. .. will essentially consist of fi lter fl uorescence from a retesting test). At a later stage when measurements are made on arbitrary samples, we can use the above-mentioned problems to estimate pure reflectance and refractive index for the tests. Additional parameters for the sample can be estimated by measuring the reaction at different electric fields under the influence of the Kerr effect.

Assuming we have a sample with absorption for a given spectral band (Fig. 7) will the transmitted intensity to decrease with the distance of the sample measured.

Reactions and the diffuse test tube walls will create a so-called White-cell and scatter the light in all three space dimensions. Again, we will have four spatial profiles along the Z axis. The shape of these will be different from those reflected pro fi lema. By standardized addition of liquids with known absorption properties as well as arbitrarily varying refractive index, pure absorption profiles can be achieved to train the system to later characterize absorption independent of refractive index.

For a scattering sample (Fig. 8) not only the walls of the tube but also the sample itself will contribute to the photon transport along the Z axis. For high scatter samples, the light comes on along the Z axis to be described by a Green function determined by uabsOt) and pscaOt). That in the forward direction and laterally observed light will be almost identical to strong disseminating media. For samples with smaller scatter, the anisotropy factor g (which can directly related to particle size) to play a larger role, so that most of the light is scattered forward and a smaller part can be observed laterally. It is a well known fact that it is one difficult problem to solve the pure absorption and spread of the Green- Actions. However, the profile will be measured from two angles, and further we will have the ability to make fl your measurements by modulating the light source on fl your radio frequencies (RF). By detecting phase shift between the drive currents and the detected signals, you can calculate the operating time. This is another time domain in frequency domain equivalent method.

When an absorbing sample secondary emits fl uorescence light at a longer wavelength can the emitted light is detected by the cut-off filter at the same z-position (Fig. 9).

Fluorescence light can be detected laterally or straight ahead. Even the cut-off itself comes to ores uoresc slightly, which is why at least two measured values must contribute to a pure sample ores uorescence. As in the previous example, the walls of the tube or the scattering of the sample to send light along the Z axis. Fluorescence light is thus detected through a number of cut-off filters, suggesting that ideally all elements are in the upper the triangle of the EEM matrix within the range of the instrument. We note that the transmission from z to z + l are measured both with and without filters. As in the case of proliferation can additional information regarding ores uorescence lifetimes is extracted from the sample by repeating the measurement at your RF modulated frequencies and detecting the phase ski.

As with the other demonstrations, calibration can be performed by a known fl uorophore with known absorption and fl uorescence spectra as well as known lifetimes are added.

Additional information regarding photoinduced kinetics and bleaching can be obtained by that the temperature and fate of the sample are checked.

Then the selected detectors reach the shortwave infrared (SWIR) 2.4um region and since the sensitivity covers fl your bands, it is realistic to expect some black body radiation from the sample itself (Fig. 10). The measurement will to a large extent be depend on the sample temperature as well as the black body properties such as the emissivity 50.) which is associated with absorption and re ectancy. The calibration can be performed by measurements of samples with fl your different transmittances in the SWIR region and with different temperatures.

As we have previously discussed, polarization plays a major role in reactions according to F resnels formula. Even in scattering and anisotropic ores uorescence, the polarization will affect. Since the source only contributes to a negligible part of the instrument cost can the concept of combinatorial measurement of light distances is expanded as far as possible handled polarization phenomena (Fig. 12). In addition to reaching out polarization-dependent parameters, such an expansion can also be used to improve the conditions for orthogonalizing non-polarization dependent properties. Then few polarizing filter provides contrast throughout the spectral width of the instrument presented the type of filter is varied along the Z axis. Just as for polarization kanlter can be spectral filters are also used on the sources to suppress the emission of long wavelengths.

The sources can be miniaturized with surface mounted devices (SMD) (Fig. 12). This would make the instrument considerably more compact. The detectors can be multi-elements linear arrays of CCD, CMOS or PMT type. This would increase the number of light distances substantially. Instead of a number of absorption altars, a kilter with varying cutter off-wavelengths along the Z-axis are used for uorescence detection. The concept can too combined with the polarization expansion.

Alternatively, elastic and orescence light detectors can be replaced with a dispersive one elements such as a concave diffraction grating in combination with a 2D image detector. More image chips can be used to expand the spectral range. This concept is also compatible with excitation in several polarization directions.

Several electronics modules are involved in the data acquisition process. A preferred, but not limiting design is presented in Fig. 11. The sample is sucked into the quarter tube through a plastic tube for analysis. A pump regulates the flow. Both the plastic pipe and the quarter tube can be replaced to meet special requirements such as for medical materials used in dialysis system. A voltage can be applied to both sides of the quarter tube to examine Kerr coefficient theme. A metal bar on which the sources and detectors are mounted can be temperature stabilized at both ends with a thermoelectric cooler and a thermistor.

This provides consistent measurements over time as well as changed external conditions, especially as dark current and sensitivity change with temperature. The two perpendicularly mounted The LED panels are given a binary address to activate an individual LED. This address is obtained from a counter circuit operating at a frequency given by a DAQ card. The LED panel gets also the power of the activated LED. The current may exceed the static limit specified for each LED as the connection fraction is usually a few percent. The this type of flash operation increases the signal-to-noise ratio in transmission measurements.

Furthermore, the current can be varied over a spectrum of operating currents, which gives ability to measure the current-voltage (U-I) characteristic of each LED. With it the information can the temperature as well as the band gap in the education layer for each LED calculated. Using this information, the light output for each LED can be predicted and stable measurements can be obtained. Finally, the current can be modulated in fl your frequency steps in RF for the purpose of performing frequency domain spectroscopy for scattering and fl uorescence measurements.

Gain signal amplification can be linear or logarithmic with phase shift detection in in relation to the modulated drive current. Thus, no RF sampling is required for measurement of the phase shift.

Analogously, signals from the LED unit's U-I characteristics can detect detected light intensities and phase shifts are digitized by a DAQ card. The information is transferred from there to one computer. As discussed above, the instrument covers all of the aforementioned physical phenomena.

However, we would like to point out that the task in the computer interpretation is not to explain the observation from an exactly correct physical model but rather to give a linear or non-linear is predictive model based on training sets and multivariate analysis.

References S. Svanberg, Atomic and Molecular Spectroscopy, Springer, Heidelberg Berlin (2004) J. Räty, K. E. Peiponen, T. Asukura, UV-Visible Rejection Spectroscopy of Liquids, Springer, Heidelberg Berlin, (2003) J. R. Lakowicz, Principles of Fluorescence Spectroscopy, 3'd ed. Springer, Heidelberg (2006) M. Brydegaard, Z., and S. Svanberg, “Broad-band multispectral microscope for imaging transmission spectroscopy employing an array of light-emitting diodes ”, Am. J. Phys., 77, 1, (2009) P. K. Dasgupta, In-Yong Eom, K. J. Morris, J. Li, “Light emitting diode-based detectors Absorbance, fluorescence and spectroelectrochemical measurements in a planar flow- through cell ”, Anal. Chim. Acta. 500, (2003) J. G. Schnabel, P. J. Grochowski, L. Wilhelm, C. Harding, M. Kiefer, R. S. Orr. Portable LED array VIS-NIR Spectrometer Nephelometer, Field. Anal. Chem. Tech. 21-28, (1998) P. C. Hauser, T. W. T. Rupasinghe, N. E. Cates, A Multi-wavelength photometer based on light-emitting diode, s Talanta 42 (4) pp. 605-612 (1995) A. Johansson, J. Axelsson, S. Andersson-Engels, J. Swartling, “Realtime light dosimetijy software tools for interstitial photodynamic therapy of the human prostate. ”, Med. Phys. 34 (11), pp. 4309-4321, (2007) B. Barbieri, E. Terpetschnig, D. M. Jameson ”F requency-domain fl uorescence spectroscopy using 280-nm and 300-nm light-emitting diodes: Measurement of proteins and protein-related fl uorophores ", Anal. Biochem. 344, (2005) 298-300 B. Montcel, R. Chabrier, P. Poulet, T ime-resolved absorption and hemoglobin concentration di fl erence maps: a method to retrieve depth-related information on cerebral hemodynamics. Opt. Expr. 14 pp.298-300 (2006) M. Boukadoum, A. Bensaoula, David Starikav, A portable Multiband Optoelectric System for Identifying and Measuring the Concentration of Fluorophore Substances, 1 13- 1 16, IEEE 2004 H. Peng, E. Makarona, Y. He, Y. K. Song, A. V. Nurrnikko, Ultraviolet light-emitting diodes operating in the 340 nm wavelength range and application to time resolved jluorescence spectroscopy Appl. Phys. Easy. 85 (8) pp.1436-l438 (2004) Q. Li, K. J. Morris, P. K. dasgupta, I. M. Raimondo Jr., H. Temkin, Portable fl ow- injection analyzer with liquid-core waveguide based fl uorescence, luminescence and long path length absorbance detector, Anal. Chem Acta. 479 pp 151-165 (2003) P. Herman, J. Vecer Frequency Domain F fluorometry with Pulsed Light-Emitting Diodes Ann. N. Y. Acad. Sci. 1130 pp. 56-61 (2008) P. Herman, B.P. Maliwal. H. J. Lin,. R. Lakowicz, Frequency-domain ores uorescence microscopy with the LED as a light source. J. Microsp 203 pp. 176-181 (2001).

K. J. Jeon, S. J. Kim, K. K. Park, Noninvasive total hemoglobin measurement, J.

Biomed. Opt. 7 pp.45-50 (2005) B. Standish, Bookham Inc. LEDs for Bioanalytical and Medical Instruments Bio. Phot.

Int. (2007) Jobin Yvon, Horiba 3D F fluorescence J obin Yvon, Horiba Linearization of the Spex-3D® Spectroflorometer Application notes S. J. Hart, R, D. J iji, Light Emitting diode excitation emission matrix fl uorescence spectroscopy, Analyst 127 pp. 1693-1699, (2002) U. Utzinger, A. J. Drukin, H. Fuchs, A. M. Gillenwater, D. L. Heintzelman, R. R.

Richards Card, Combined Jluorescence and Rejection Spectroscopy, U.S. Pat. patent 6571 1 18, freepatentsonline.com H. Zeng, H. Lui, C. Macaulay, B. Palcic, D. I. Mclean, Apparatus and methods relating to optical systems for diagnosis of skin diseases, U.S. patent 6069689, freepatentsonlinecom M. Gouzman, N. Lifshitz, S. Luryi, O. Semyonov, D. Gavrilov, V. Kuzminskiy, Excitation-emission imeter uorimeter based on linear interference fi lters, Appl. Opt. 43 pp. 3066-3072, (2004) R. L. Martineau, B. C. Towe, V. Stout, An Optical Micro-Instrumentation System for Measurement of F fluorescent Proteins in Whole-cell Biosensors, IEEE pp. 90-93, (2006) M. Trtilek, D. M. Kramer, M. Koblizek, L. Nedbal, Dual-modulation LED kinetic fl uorometer, J. Luminescence pp. 72-74, (1997) C. Moser, T. Mayr, I. Klimant, Filter cubes with built-in ultrabright light-emitting diodes as exchangeable excitation light sources in ores uorescence microscopy, J. Microsp. 22 pp. 135-140, (2006) J. S. Kuo, C. L. Kuyper, P. B. Allen, G. S. Fiorini, D. T. Hiu, High-power blue / U V light- emitting diodes as excitation source for sensitive detection, Electrophoresis 25, pp. 3796- 3804, (2004) FORCE-A Real-Time Optical Solutions for Sustainable Agriculture, 2006 FORCE-A, Multiplex®2, 2007 G. Agati, Response of the in vivo chlorophyll ores uorescence spectrum to environmental factors and laser excitation wavelength, Pure Appl. Opt. 7 pp. 797-807, (1998) M. S. Kim, Y. Chen, S. Kang, I. Kim, A. M. Lefcourt, M. Kim. Fluorescence Characteristics of Wholesome and Unwholsome chicken Carcasses Appl. Spec. 60 (10), (2006) C. Sluszny, E. S. Yeung, One- and T wo-dimensional Miniaturized Electrophoresis of Proteins with Native F fluorescence Detection, Anal. Chem. 76 (5), pp.-1359-1356, (2004) Laser2000, EigenliteTM RS-5B Spectrally Programmable Light Source, Product Summary K. Davitt, Y. K. Song, W. Patterson, A. V. Nurmikko, Z. Ren, Q. Sun, J. Han, UVLED arrays at 280 and 3 40nm for spectroscopic biosensing, Phys. State. sol 204 pp. 2112- 2116, (2007) R. W. Cole, J. N. Tumer, Light-Emitting Diodes Are Better Illumination Sources for Biological Microscopy than Conventional Sources, Microsc. Microanal. 14 pp.243-250, (2008) J. Xu, B. Yangm H, Tian, Y. Guan, A windowless flow cell-based miniaturized fluorescence detector for capillary flow systems, Anal. Bional. Chem. 384 pp. 1590-1593, (2006) E. P. de Jong, C. A. Lucy, Spectral Filtering of Light-Emitting Diodes for Fluorescence detection, Anal. Chim. Acta 546 pp. 37-45, (2005) E. Moreno-García, C. E. Guerra, J. M de la Rosa-Vázquez Spectrometer to measure steady-state fl uorescence emitted by liquid and solid sample ,. IEEE (2008) S. Ek. B. Anderson, S. Svanberg, Compact fiber-Optic fluorosensor employing light- emitting ultraviolet diodes as excitation sources Spectro.Chim. Acta Part B 63 pp.349- 353, (2008) S. Landgraf, Use of ultrabright LEDs for the determination of static and time-resolved florescence information of liquid and crude oil samples, J. Biochem. Biophys. Methods 61 pp. 125-134, (2004) M. Yeary, R. Kelley, I. Meier, A. Snyder, A. Arul, T. Hicks, P.McCann, C. Roller, D.

Guidry, Next-Generation Digital High-Bandwidth Spectroscopy Sensor Systems, IEEE Instr. Meas. Tech. Conf. (2007) J. B. Fishkin, P. T. C. So, A. E. Cerussi, S. Fantini, M. A. Franceschini, E. Gratton, Frequency-domain method for measuring spectral properties in multiple-scattering media: methemoglobin absorption spectrum in a tissuelike phantom, App. Opt. 34 pp. 1143-1155 (1995) S. Fantini, M. A. Franceschini, J. B. Fishkin, B. Barberi, E. Gratton, Quantitative determination of the absorption spectra of chromophores in strongly scattering media: a light-emitting diode based technique, App. Opt. 33 pp. 5204-5213 (1994) E. Gratton, S. Fantini, M. A. Franceschini, G. Gratton, M. Fabiani, Measurements of scattering and absorption changes in muscle and brain, Phil. Trans. R. Soc. London. 352, pp.727-735 (1997) B. Chance, M. Cope, E. Gratton, N. Ramanuj am, B. Tromberg, Phase measurement of light absorption and scatter in human tissue, Rev. Soi. Instr. 69 pp. 3457-3481 (1998)

Claims (13)

10 15 20 25 30 35 Patent claim Apparatus for spectroscopic characterization of a sample based on optical spectroscopy, comprising a channel containing said samples, light sources arranged around said channel along the longitudinal axis of the channel, and detectors arranged around said channel along the longitudinal axis of the channel characterized in that each light source radiation is adapted to detect from avs of said detectors along the longitudinal axis of said channel for an integrated estimate of at least three optical partners selected from the group comprising: absorption coefficient, scattering coefficient, anisoptropic scattering coefficient, refractive index, can affect each other during said estimation. The device of claim 1, wherein the light sources are selected from the group comprising: light emitting diodes (LEDs), laser diodes (LDs), resonant cavity diodes, quantum well-based LEDs, and optically pumped LEDs, having a spectral range from deep ultraviolet (UV) through visible (VIS), near infrared (N IR), infrared to the thermally infrared parts of the optical spectrum. The device of claim 1 or 2, wherein the light sources are selected from the group consisting of: photodiodes (PD), phototransistors, LEDs, microbolometers, avalanche photodiodes (APD), photomultipliers (PMT), linear PD arrays, Charged Coupled Devices (CCD ), linear CMOS (complementary metal oxide semiconductor) detectors, image intensifiers and multi-element PMTs. Device according to any one of the preceding claims, wherein each spectral band is measured from fl your different observation angles in relation to the light source, different distances between light source and detector, and by different fi filter configurations Device according to any one of the preceding claims, wherein a single instrument can measure concentrations in the sample over an extended dynamic range compared to instruments based on individual transmittance measurements, this as a result of the number of possible measured light path distances. 10 15 20 25 Patent claims Device according to any one of the preceding claims, wherein said light sources can be modulated up to radio frequencies for measuring different photon delays caused by ores uorescence and / or scattering. Device according to any one of the preceding claims, wherein said light sources are arranged to be stabilized by access to the local temperatures in the depletion layer of the light source via characterization of voltage and current of the light source. Device according to any one of the preceding claims, wherein the sample consists of a liquid which fl surfaces in a straight or curved transparent channel, of which at least a part is clear. Device according to any one of the preceding claims, wherein the sample consists of a liquid which fl surfaces in a straight or curved transparent channel, at least a part of which is diffusely dispersing. Analysis method using a device according to any one of claims 1-9 characterized by the use of multivariate statistical methods which may involve training sets. An analysis method using a device according to any one of claims 1-9, wherein optical properties are extracted individually from measured data. A method using a device according to any one of claims 1-9, comprising steps of process control, monitoring, quality control or classification, including possible warning systems. A method using a device according to any one of claims 1-9, wherein the sample is a substance selected from the group consisting of; blood, a blood solution, surgical fluids, milk or dairy products, soft drinks, beer, and alcoholic beverages, light-flowing foods, water of various purities, chemical process fluids, and urine.