WO2014198814A1

WO2014198814A1 - Method and device for detecting undissolved particles in a fluid

Info

Publication number: WO2014198814A1
Application number: PCT/EP2014/062201
Authority: WO
Inventors: Michael Bassler; Christian Sommer; Stephan Quint; Peter Spang
Original assignee: Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e. V.
Priority date: 2013-06-12
Filing date: 2014-06-12
Publication date: 2014-12-18
Also published as: DE102013210953A1

Abstract

The invention relates to a method and a device for detecting undissolved particles in a flowing fluid, which fluid has the density p and the viscosity µ. The fluid with the particles is transported through a measuring channel (60) and flows through an excitation zone (76), in which it is recorded by a radiation source, and a measuring zone (80), in which an emission emanating from the particles as a result of the excitation and/or scattered light emanating from the particles is recorded by a detector (82). A mask (84) arranged between the measuring channel (60) and the detector (82) modulates the emission or the scattered light, and the modulated signal is evaluated by means of a correlation analysis. The fluid is transported through the measuring channel (60) at an average flow velocity (a) at which the channel Reynolds number (b) assumes values between 0.1 and 500, wherein the particles assume a velocity distribution of which the half width is less than 4% of the average particle velocity.

Description

Method and apparatus for detecting undissolved particles in a fluid

description

The invention relates to a method and a device for detecting undissolved particles in a flowing fluid, which fluid has the density p and the viscosity μ. The method provides that the fluid is conveyed with the particles through a measuring channel, wherein the fluid with the particles flows through an excitation zone, in which it is detected by a radiation source, and a measuring path, in the of the particles as a result of the excitation Emission and / or scattered light emanating from the particles is recorded by a detector, wherein a mask arranged in the region of the measuring path between the measuring channel and the detector modulates the emission or the scattered light, and wherein the modulated signal is evaluated by means of a correlation analysis in order to obtain the Prove particles.

Accordingly, the device comprises a measuring channel, through which the fluid can be conveyed with the particles and which comprises a measuring section, a pump device for conveying the fluid with the particles, a radiation source for electromagnetic radiation, which is aligned with an excitation zone in the measuring channel detector arranged on the measuring path and configured to detect emission emanating from the particles as a consequence of the electromagnetic radiation and / or scattered light emanating from the particles, a mask arranged in the region of the measuring path between the measuring channel and the detector for spatially modulating the Emission and / or scattered light and an evaluation device with means for correlation analysis for detecting the particles based on the modulated signal. The invention has its origin in the field of flow cytometry. In simple terms, this is a process in which the fluid with the particles, in this case biological cells, is pumped through a flow cell in which they are excited to emit fluorescence by means of incident light. The cells are therefore usually previously labeled with fluorescent substances. The emitted light is picked up by a detector as the particle passes through the flow cell. Between the measuring channel of the flow cell and the detector is a mask, also shadow mask, with which the emitted light is modulated in time due to the relative movement of the particles to the mask. Such a signal is recorded at a certain sampling rate. The result is called sampling.

The scanning sequence of the modulated signal or the successive translucent or transparent and opaque sections of the mask can be arranged periodically in the simplest case. Preferably, however, they are arranged as a so-called "pseudo random binary sequence", that is to say in an order with irregular distances and lengths in order to allow unambiguous identification, in particular when measuring a plurality of particles which overlap the measurement path in a time-overlapping manner.

The evaluation by means of the correlation analysis takes place by the evaluation device computationally forming a reference sequence from the known pattern of the mask. It is fixed in its absolute length (scaled) and compared or folded with the sampling, whereby it is shifted over the signal. The process is iteratively repeated at other scales until a match is found. The correspondence is shown for each particle passing through the detection cell in the form of a correlation peak, by means of which the particles can be counted, for example. Moreover, it is known to use this arrangement to measure the intensity of the emitted light and / or additionally a scattering signal emanating from the particle, from which further information about the particle can be obtained. For example, such a statement about its size can be made.

Reference is made by way of example to the documents US Pat. No. 8,373,860 B2, US Pat. No. 2013/0016335 A1, US Pat. No. 2008/0181827 A1, US Pat. No. 2008/0183418 A1 and US Pat. No. 7,763,856 B2, which deal with this technology.

The present invention proposes an improved method of this type and an apparatus for its execution, with which particle properties can be detected in a flowing fluid in a shorter time or with less computational effort.

The inventive method for detecting undissolved particles in a flowing fluid of the type described above provides that the fluid is conveyed through the entire measuring channel with an average flow velocity v at which the channel Reynolds number

_ p - v ^■ d _hyd

Values between 0.1 and 500, where d _{hyd denotes} the hydraulic diameter of the measuring channel and wherein the particles assume a velocity distribution whose half- _{width is} less than 4% of the average particle velocity, preferably less than 2% and more preferably less than 1%. The invention further provides that the device of the aforementioned type has a pump device associated control device which is set to set a flow velocity v for the fluid through the entire measuring channel, in which the channel Reynolds number p ^■ v ^■ d hyd

μ

Values between 0.1 and 500, where d _{hyd denotes} the hydraulic diameter of the measuring channel and wherein the particles assume a velocity distribution whose half- _{width is} less than 4% of the average particle velocity, preferably less than 2% and more preferably less than 1%.

For the purposes of this document, the term "particles" includes extended solids as well as droplets of a liquid, meaning that the fluid with particles may be a suspension or also an emulsion flowable medium are understood and in particular rewrite gases and liquids.

The finding underlying the invention is that identical particles have a significant flow behavior in the fluid only under the aforementioned conditions, as will be explained in more detail below with reference to FIGS. Namely, by observing the above experimental conditions, the particles in the flowing fluid arrange on certain flow paths in the measurement channel due to an interaction with the fluid, thereby entraining the fluid at a significantly narrower velocity distribution in the fluid than outside the aforementioned conditions is. This is attributed to the hydrodynamic properties of the particles, where as "hydrodynamic particle property" in the For the purposes of this specification, the particle size, the elasticity, the shape and the density of the particles are understood in particular. The effect occurs depending on the hydrodynamic properties at just that channel Reynolds number between 0.1 and 500, preferably between 0.5 and 200, and more preferably between 1, 0 and 50, which at a given fluid and hydraulic diameter through the flow velocity v is adjusted. Below R _K = 0.1, the flow profile of the fluid is too shallow, so that the forces acting on the particles in the fluid flow are insufficient to arrange them within an acceptable period of time. In cases of some particles having particular hydrodynamic properties, a lower limit of R _K = 1.0 may be preferred. On the other hand, above the limit of R _K = 500, turbulence is expected on the particle, so that an uncontrollable behavior of the particles is to be expected. Depending on the shape of the measuring channel, more precisely on its cross-sectional dimension, cross-sectional shape, and length, already a channel Reynolds number of more than 200 or even more than 50 to a very high pressure in the measuring channel or the entire line system of the device not every basically suitable material (eg plastic, glass or steel) can withstand. In addition, also depending on the channel shape, the flow velocity in the measuring channel rises above values which require a considerable higher technical outlay for the detection of the particles or even a measurement of their velocity in the measuring channel. For these reasons, it is advantageous in some cases to limit the channel Reynolds number to 200 or even 50. Due to the significantly narrower velocity distribution of the particles, the evaluation by correlation analysis can be made considerably more targeted because the previously described scaling of the reference sequence only has to be iteratively changed within a narrow range around an average value corresponding to the mean velocity, which considerably shortens the entire evaluation process. The mentioned "flow behavior" can as a generic term at least one of the parameters - average migration velocity of a particle population

- Velocity distribution of a particle population

Change in migration rate depending on a change in flow conditions, i. in particular the flow velocity and thus the channel Reynolds number

Change in the distribution of the migration speed of the particles in

Depending on the change in the flow conditions are understood, of which not only the speed distribution used so far, but typically each parameter behaves in dependence on the hydrodynamic particle properties in compliance with the flow conditions in a significant manner. Due to the significance of the flow behavior, it is possible to use the effect for the detection of individual particles, because their flow behavior can be predicted more accurately. In addition to the mere detection of the particles, information about the hydrodynamic properties of the investigated particles can therefore be obtained with the method and the device according to the invention and the particles are determined on the basis of this information or if different hydrodynamic particle properties lead quantitatively to the same effect, ie to the same flow behavior , at least be classified. This is not possible without observing the experimental conditions, because a broad velocity distribution does not allow an assignment of the determined parameters to the particles. This will be clarified below by means of an example. For example, from the scaling of the reference signal in a calibrated sampling, the residence time of the particles in the measuring channel over the distance of the measuring section or directly the migration speed of the particles in the measuring section can be determined. Due to the narrow velocity distribution, discrete distribution maxima can form for particles with different hydrodynamic properties, which allow a differentiation of different particle populations or even allow a classification or a determination of the particles based on a specific average migration rate or its change as a result of a change in the flow conditions. At least for differentiation, not even a calibration is needed.

The differentiation, classification or determination of the particles is preferably carried out based on at least one of the above parameters. The parameter or parameters are obtained from the data of the determined migration speed in an analysis step. For this purpose, the device advantageously has an analysis device which is set up to analyze the flow behavior of the particles on the basis of the determined migration speed and to output them for differentiation, classification or determination of the particles. For example, the analysis also includes calibrating recorded sampling data, i. the conversion into concrete physical quantities. The analyzing may include evaluating one or a plurality of events to determine, for example, a distribution or a mean. At the end of the analysis, the analysis device outputs the result of the analysis, ie the parameter (s) characterizing the flow behavior, depending on the user's preference for image representation, for expression, in tabular form, as graphic or as electronic data record.

"Classification" means, for example, if only one or more of the hydrodynamic properties of the particles, but not the particle itself clearly can be determined, the classification of the particle in a particle class with just those hydrodynamic properties.

In particular, the present invention finds application in so-called microfluidics in which minute volumes of liquid in the range of μΙ are handled and analyzed. In particular, the device according to the invention for miniaturization is suitable for a so-called microfluidic chip or lab-on-a-chip system with which, for example, blood samples or other organic cell material can be analyzed.

The term "measuring path" in the sense of the patent claims is the section of the measuring channel in which the migration speed of the particles takes place.Furthermore, the measuring channel comprises said "excitation zone" and can additionally comprise a "starting distance". While the start-up section and the measuring section are arranged spatially one behind the other in the flow direction, the excitation zone and the measuring section or the excitation zone and the start-up section may coincide spatially or overlap.

It should be noted that the adjustment of the significant flow behavior due to the hydrodynamic properties of the particles does not necessarily require a constant channel cross-section. All that is required is that the Reynolds number R _K lies in the claimed range in all sections, that is to say the start-up section and the measuring section of the measuring channel. Nevertheless, a consistently constant channel cross-section - and thus both the cross-sectional geometry and the cross-sectional area is addressed - for reasons of simplicity and process reliability to prefer, because then due to constant flow conditions over a longer Kanalab- cut can form a flow equilibrium within the flowing fluid, which increases the measurement precision.

Preferably, the method provides that the particles have a maximum particle diameter a _p and that the fluid with the particles in the measuring channel in the flow direction in front of the measuring section a start-up path with a length L of at least

1 h ⁴

L = 3π- _g

° K, a ^a p flows through, wherein R _{K a} denotes the channel Reynolds number in the start-up section and wherein the measuring channel has a height h at least in the region of the start-up section, but preferably in the entire measuring channel.

In a corresponding manner, the device according to the invention preferably provides that the measuring channel in the flow direction upstream of the measuring section has a start-up section with this length L, wherein the measuring channel has a height h at least in the region of the starting section, but preferably in the entire measuring channel. The particle diameter a _p is assumed to be the largest extent in the case of non-spherical particles, that is to say the main axis in the case of a rotational ellipsoid.

For a measuring channel with unequal extent of the cross section in two spatial directions, the term "height h" is understood to mean the dimension of the lesser extent, for example, for a channel with a rectangular cross section, the length of the short side of the cross section width w. For a circle, height h and width w are the same each the diameter of the cross section. Any changes in the cross section should generally and in particular in the direction of the height h not abruptly but only continuously, so that the flow within the channel always remains laminar. In general, however, the height h and the width w are constant throughout the measurement channel to avoid accelerating forces which can disturb the arrangement of the particles on certain flow paths in the measurement channel and thus widen their velocity distribution.

Compliance with the above condition ensures that an equilibrium flow has formed at the end of the starting section L and thus at the beginning of the measuring section, in which the particles or individual particle populations with uniform hydrodynamic properties in the flowing fluid are arranged in their equilibrium configuration and therefore the velocity distribution of the Particle has the mentioned significance.

The method and the apparatus preferably further provide that the measuring channel has a cross-section with a width w and a height h at least in the region of the measuring path, but preferably over the entire length of the measuring channel, the measuring channel always limiting the aspect ratio e = ^ <0.1.

To be able to enforce a sufficiently high amount of fluid, so that in special cases in a reasonable time frame sufficient statistical significance can be achieved, the cross section of the measuring channel in the second spatial direction, the width w, a much greater extension than the height h and even preferably at least ten times the extent.

The inventors have found that a particularly narrow velocity distribution of the particles occurs when its height h or minimum elongation of 200 times, preferably 50 times, does not exceed the particle diameter.

Accordingly, it is preferred for the method as well as for the device if the height h of the measuring channel is at least in the region of the measuring section, but preferably over the entire length of the measuring channel, 200 times, particularly preferably 50 times, the maximum particle diameter does not exceed a _p . It is furthermore advantageous if, at least in the region of the measuring path, but preferably over the entire length of the measuring channel, the ratio γ = <0.75, particularly preferably γ = <0.5.

This requirement reflects the fact that particles larger than half the minimum channel cross-section no longer have sufficient headroom to position themselves in the height h direction, which can eventually lead to ambiguities in the velocity distribution. While the method yields deformable particles still up to a γ of 0.75 clear readings, this is not ensured for rigid particles. Therefore, the compliance with a γ of the highest 0.5 is to be preferred.

Also, the velocity distribution of the particles can be adjusted if, according to a further aspect of the invention, the particle Reynolds number

R _P = R _K ■ (^) assumes values between 2.5 10 ^"6 and 0.1, where a _p turn, the maximum particle diameter and the height h of the measuring channel at least in the area of the measurement section, but preferably over its entire length, specify. This requirement includes the insight that not only a given range of the channel Reynolds number and a specific ratio of the particle diameter to the height of the measurement channel lead to a significant particle flow, but also there is a non-linear relationship between the parameters.

According to an advantageous development of the method, the velocity distribution and / or a mean velocity value is obtained from the determined migration speed and used for differentiation, for classifying or for determining the particles.

If, for example, two particle populations are investigated, with the particle populations each assuming a different average equilibrium velocity owing to at least one different hydrodynamic particle property, this manifests itself in the velocity distribution, for example by two maxima, which at least one differentiation, one classification or in a system then sufficiently determined (The dispersion or emulsion consisting of fluid and particles is also referred to as a system) just even a determination of the two Teilchenpo- pulations can be based. This presupposes that the velocity distributions of the particles are respectively sufficiently narrow and far enough apart from each other so that the maxima can be unambiguously identified, which is ensured in particular by compliance with the conditions according to the invention. A distribution according to the experience of the inventors is understood to be sufficiently narrow if the half-width of the velocity distribution of a particle population is less than 4%, preferably less than 2%, of the mean particle velocity. A preferred form of analysis therefore comprises, in particular, the determination of the average particle velocity and / or the distribution width, particularly preferably the half-width of the determined velocity distribution. This is useful in connection with a further development of the method in which a change in the measured migration speed and / or a change in the distribution of the migration rate as a result of the change in the flow velocity v is analyzed and used for differentiation, classification or determination of the particles.

The inventors have found that not only the absolute or relative particle velocity or the velocity distribution can be used to classify the particles, but also that the change of one of these two parameters allows a statement about the hydrodynamic properties of the particles. By a combination of both analyzes, under certain circumstances, a further degree of freedom or a further hydrodynamic particle property can be determined and thus the classification of the particles can be carried out more accurately or even a determination of the particles becomes possible.

As previously discussed, determining the migration rate preferably involves measuring the time that the particles take to pass the measurement path.

According to an advantageous development of the device, the detector is set up to quantitatively detect the intensity of the emission emanating from the particles and / or the scattered light emanating from the particles. Correspondingly, the analysis device is set up to output the intensity for differentiation, for classification or for the determination of the particles.

Further principles, features and advantages of the invention are explained in more detail below with reference to drawings and exemplary embodiments. In the figures show: Figure 1 is a schematic representation of the particle concentration in a cylindrical channel after Segre and Silberberg; Figure 2 shows four distributions of the particles in the cylindrical channel in the radial direction at different axial positions;

FIG. 3 shows velocity distributions of the same particles in one

Fluid at different flow rates;

Figure 4 shows two velocity distributions of different particles in the same fluid at constant flow rate;

Figure 5 is a schematic representation of the grouping different

Particles along the velocity profile of a fluid in a channel;

FIG. 6 a schematic representation of a measuring channel with a time measuring device for a modulated fluorescence or scattered light signal;

FIG. 7 shows the measuring channel according to FIG. 6 in plan view;

FIG. 8 shows a schematic measurement signal of a single particle taken with that of a measuring device according to FIG. 6;

FIG. 9 the signal of a particle after correlation analysis; Figure 10 derivative of the correlation peak of Figure 9; Figure 1 1 shows a typical signal of several particles taken with a measuring device according to Figure 6;

FIG. 12 shows the measurement signal according to FIG. 11 after passing through the correlation analysis and derivation;

Figure 13 is a graph of three velocity distributions of different particles measured at a first flow rate;

Figure 14 is a graph of three velocity distributions of the same three particles measured at a second flow rate;

FIG. 15 shows a two-dimensional intensity-velocity

Diagram of two particle populations; and

FIG. 16 shows a projection of the diagram according to FIG

Intensity axis.

In a fundamental scientific paper from 1961, Segre and Silberberg have shown that solid particles in a dispersion flowing through a cylindrical tube assume an inhomogeneous distribution after a certain distance, see G. Segre and A. Silberberg, "Radial particle displacement in poiseuille flow of suspensions" , Nature, Vol. 189, 21 January 1961. This behavior is shown schematically in FIGS.

1 shows a longitudinal section through such a cylindrical tube 10 with the central axis 12 and the radius r, in which the fluid with the particles, ie the dispersion, flows in the direction of arrow 14. It is assumed that the particles in the dispersion at the beginning of the tube 10 at position a) are still homogeneous Distributed with the fluid distributed over the entire cross-sectional area. The constant distribution at the position a) is shown schematically in the corresponding diagram a) in FIG. 2, the distribution from the center or symmetry axis 12 to the radius r being plotted there.

As the dwell time, or distance, in the channel increases, interactions between the channel walls, the fluid, and the particles cause a change in the spatial distribution in the fluid. Accordingly, the distribution is radially constricted downstream, so that the particle concentration near the peripheral walls rapidly decreases to zero. At the same time, however, forces also occur which displace the particles from the middle of the channel towards the edge. This effect is only slightly weaker so that the distribution first decreases slightly from the center or symmetry axis "0." As a result of both processes, a distribution maximum tends to form closer to the edge of the channel The distance in the direction of flow increases until the distribution of the particles is limited to a narrow annular section in the channel This development is shown for three further positions b), c) and d) by the associated diagrams in FIG.

The behavior can also be read in FIG. 1 on the basis of the contour lines 16 and 18. These show that the constriction of the spatial particle distribution with the distance covered in the channel is continuous and parabolic.

In a more recent work, scientists have found that the average radius of maximum particle concentration in a dispersion flowing through a tube depends inter alia on the diameter of the particles, see Bhagat, Kuntaegowdanahalli, Papautsky, "Inertial Microfluidics for Continuous Particle Filtration and Extraction". , Microfluid Nanofluid, Vol. 7, 217-226, 2009. They lead to the parametrization of this behavior the already mentioned particle Reynolds number

a, which indicates the smallest channel diameter in a rectangular channel h. The scientists suggest in their work to take advantage of this effect for particle extraction and filtration. These findings have led the inventors to study the rate at which the particles travel in a fluid in a channel. A first part of their findings is illustrated in FIG. This shows a three-dimensional diagram in which the velocity distribution of the particles measured at six different flow rates in the same channel is plotted. Specifically, the particle velocity in the x-axis direction, the flow rate in the y-axis direction, and the particle number in the z-axis direction are plotted therein. The particle velocity is normalized to the mean flow velocity of the fluid so that the change in flow rate does not result in a shift in the (absolute) particle velocity and the results are easier to compare.

It could be established so that // set at a flow rate of 10 min, a velocity distribution having a maximum which has close to the maximum flow rate of the fluid v _max a steep edge and on the other hand a small spur to a smaller particle out See graph 20. With increasing flow rate and thus flow velocity of the fluid, this maximum decreases and shifts slightly towards a lower particle velocity. At the same time, the initially small spur of slower particles increases until see above curves 21 and 22. As the flow rate continues to increase, the distribution shifts further and further towards slower particles, with a very narrow and high distribution maximum beginning to form as shown in curves 23 and 24 until, finally, at a flow rate of 200 //// min, as can be seen in curve 25, all particles are very much within a very high narrow, dominant velocity distribution peaks are focused, which is at a relative speed just above the average flow velocity v of the fluid. It is noteworthy that the distribution width decreases significantly over the initial distribution width at a flow rate of 10 / min. The inventors further made another observation, which is illustrated in FIG. While the diagram in FIG. 3 shows the behavior of a single particle population in a fluid, in a further experiment, whose result is shown in FIG. 4, the inventors have described the behavior of two different particle populations in a fluid under the same experimental conditions (constant channel cross-section and flow rate / Flow velocity).

FIG. 4 shows in a two-dimensional diagram a velocity distribution in which a particle counting rate in the vertical direction against the particle velocity in the horizontal direction is plotted. There are two distinct maxima at an average particle velocity of about 432 and 446 mm / ssc, with the first maximum having a half width of about 5.1 mm / ssc and the second maximum having a half width of about 1.9 mm / sQc , Thus, the two maxima are clearly separated and can each be clearly assigned to one of the two particle populations. This makes it obvious that different particles due to different hydrodynamic particle properties not only in their spatial but also in your velocity distribution in a flowing medium significantly different order and therefore differentiable, classifiable and can also be determined by appropriate behavior with a suitable calibration.

An explanation of the different dynamic behavior of the particles is made with reference to FIG. Shown is a cross section through a flat rectangular channel 30 which is formed symmetrically about the center line 32 and the height h, wherein the height h denotes the smallest extent of the rectangular channel. The channel 30 is traversed by a fluid having two different particle populations in the direction of the arrow 34. In this case, a parabolic-shaped flow profile 36 of the fluid is formed, which has a _maximum velocity v _max in the middle of the channel and decreases symmetrically towards the two edges of the channel 30.

Furthermore, two particle populations are present, of which the particles of a first population are schematized as small circles 37 and those of a second population as large triangles 38. It is assumed that the particles 37 of the first population have a smaller mean particle diameter than the particles 38 of the second population.

The spatial focussing of the particles explained with reference to FIGS. 1 and 2 causes them to be arranged on different equilibrium paths in the channel, the larger particles 38 being oriented slightly closer to the middle 32 of the channel 30 than the smaller particles 37 in that the larger particles 42 occupy a higher equilibrium velocity v ₂ corresponding to the flow profile 36 than the smaller particles 40 whose equilibrium velocity lies at v ₁ . Thus, it is understandable that Particles of different sizes are differentiable based on their average speed, if the flow velocity of the fluid is so high that the velocity distribution of all particle populations are sufficiently focused, see. 3, and the differences in the hydrodynamic properties are significant enough for the velocity distributions to have maxima which are far enough apart from each other, cf. FIG. 4.

The inventors have also found that in addition to the size of the particles and their elasticity, shape and density as hydrodynamic properties have an influence on the arrangement in the velocity profile. These properties can cancel each other out so that particles with different hydrodynamic properties still quantitatively result in very similar or even average velocities and similar or equal velocity distribution profiles. Under these circumstances, no definite determination but only a classification of the particles is possible. For example, the hydrodynamic focusing of bacteria with a certain diameter and elastic deformability takes place even at a lower flow velocity than the rigid spherical particles of the same size. In other words, a bacterium may assume the same average velocity as a much larger but rigid sphere.

How the invention utilizes the found effects will be explained below.

FIGS. 6 and 7 schematically show somewhat simplified an embodiment of the device according to the invention, in which the measuring channel 60 with a flat rectangular profile is incorporated into the substrate of a microfluidic chip 62. The microfluidic chip is shown in Figure 6 in the side view in section and in Figure 7 in plan view. The side view shows the measuring nal 60 in the height direction in which has an extension or height h. In the width direction, it has an extension or width w. For the

Aspect ratio is preferably e = ^ <0.1. At the beginning, the measuring channel has a feed 64 and an outlet 66 for the fluid to be examined. The feed 64 and / or the discharge 66 can be connected to a pumping device, not shown, for conveying the fluid with the particles (this time from right to left) through the measuring channel 60, with which a pressure difference can be generated by means of a control device associated with the pumping device it can be set that the fluid flows through the measuring channel at a flow velocity v at which the channel Reynolds number

_ _ P ^■ V ^■ dhyd

κ _κ -

μ values between 0.1 and 500, preferably between 0.1 and 200 and particularly preferably between 0.1 and 50 assumes.

In Figure 7 it can be seen that the measuring channel 60 is further provided with two further terminals 68, 70, through which a sheath liquid viewed in the width direction can be supplied from both sides. The flow of the fluid through the feed 64 into the measuring channel 60 with the particles is hydrodynamically focused in the flow direction v behind an intersection 71 with the connections 68, 70 for the sheath liquid, so that it flows through the measuring channel 60 in the width direction w substantially in the center while only the sheath liquid added through the openings 68, 70 has contact with the channel wall in the width direction w. On the one hand, this causes the flow of the fluid with the particles to be adjusted to a desired width such that all the particles pass through the excitation zone described below. to run. On the other hand, this ensures that a substantially constant flow profile is formed in the width direction of the measuring channel and the flow profile varies substantially only in its vertical direction according to the parabolic shape shown in FIG.

Furthermore, the microfluidic chip 62 has an entrance window 72 for a laser beam 74 which is arranged to the direction of the laser beam so that it completely or predominantly in the substrate medium of the microfluidic chip 62, which preferably consists of plastic and more preferably of PMMA, broken and is passed under total internal reflection between the two interfaces of the microfluidic chip 62 to the environment. The laser beam passes through the measuring channel 60 and excites the particles in the fluid or irradiates them. Accordingly, section 76 within the meaning of the claims forms an excitation zone in the measurement channel, to which the radiation source (not shown) for electromagnetic radiation 74 is aligned in the sense of this invention.

In the flow direction v behind the channel intersection 71 starts a run-up path 78 of length L. In this section, the particles flowing in the fluid have an opportunity to locate within the airfoil as shown in Figure 5 and assume a corresponding migration speed. In the flow direction ^v behind the start-up section 78 is the measuring section 80, which is characterized in that a detector 82 is aligned thereon. In the region of the measuring section 80 is between see the measuring channel 60 and the detector 82 also has a mask 84. This so-called shadow mask has in the flow direction v spatially consecutive transparent and non-transparent sections. These sections block emission emanating from the particles as a result of the excitation and / or scattered light emanating from the particles on their way to detection. gate 82 or let this happen. This is exemplified at wavy lines 86 and 88, respectively, for two of the sections.

This measuring arrangement generates the time-modulated detector signal shown in FIG. 8 when passing through an emitting or scattering particle. It should be noted that the diagram of Figure 8 shows a schematically idealized representation and no real measurement. The duty cycle of the signal represents the sequence of the transparent and non-transparent portions of the shadow mask 84. The signal is recorded in a manner known per se with a sufficient sampling rate. The modulated signal of a particle has a time length (residence time) which results from its velocity and the length of the measuring channel. Time duration and speed are equivalent. From the residence time and the sampling rate, ie the number of measurement points per time unit, the total number of measurement points that make up the signal results. Also, this total number of measurement points occupied by the modulated signal thus depends on the velocity of the particle.

The scanning sequence or the successive sections can in the simplest case exist in the form of a periodic sequence of transparent and non-transparent sections. Preferably, however, they are arranged as a so-called "pseudo random binary sequence", that is to say in an order with irregular distances and lengths in order to enable unambiguous identification, in particular when measuring a plurality of particles which overlap the measuring path 80 in a time-overlapping manner.

This evaluation is done using a correlation analysis. The pattern of the shadow mask is known. From this, the evaluation device computationally forms a reference sequence. It is scaled in absolute length and compared to the sampled detector signal, shifting it over the entire sampling. Then the process is added iteratively repeats other scaling until a match is found. This process can also be referred to as a query of velocity channels or filtering and takes place at discrete values, whereby here too attention must be paid to a sufficient resolution.

The result thereof is shown in FIG. The correlation signal always grows approximately linearly from the two sides towards the middle, only exactly in the middle does a significant correlation peak form, which signals the agreement. For the exact determination of the position of the correlation peak, the correlation diagram according to FIG. 9 can be derived mathematically, from which the diagram according to FIG. 10 results. Therein, the zero crossing between the maximum and the minimum of the derivative signal marks the exact value of the correlation signal. The actual length of the measuring section is known. Therefore, by means of a calibration measurement with a known position of the correlation peak, i. the velocity of the particle is determined absolutely at the determined length of the reference sequence. Otherwise, the result is used for a relative comparison with the speed of other particles.

In reality, several particles will pass the measurement path 80 at the same time or at least in succession and partly in an overlapping time interval, which is why the time-modulated measurement signal at first glance does not allow a clear statement about the number of particles and their residence time within the measurement path 80. Such a measurement of eight partially time overlapping particles is shown in the sample of Figure 11. At the same time, for illustrative purposes, for each identified particle, the shadow mask 84 is shown, from which it can be seen that the events of passing particles partially overlap in the right half of the diagram. Nevertheless, the particles in the previously described Can be exactly determined, as can be seen in the figure 12 on the basis of the derivations of the correlation peaks, because the use of an unperiodic shadow mask despite significant overlap of the residence time of the particles in the region of the measuring section ensures significant discrete correlation results.

If the particles do not have an identical migration speed, which is basically the case in reality, the correlation analysis yields different scaling values for the reference sequence. This is preceded by a high computational effort, since the entire sampling must be correlated with a large number of scaled reference sequences, depending on the required resolution and width of the velocity interval to be scanned. To reduce this effort, the device according to the invention and the method according to the invention make it possible to select the velocity distribution so narrow that the scaling interval can be considerably reduced without loss of measurement data.

In addition, both velocity distributions of identical particles and velocities or their equivalents of particles of different populations can be determined. Both together is done by way of example in the diagram according to FIG.

The diagram shows the velocity distributions of three particle populations which were simultaneously conveyed through a measurement channel under the experimental conditions according to the invention at a channel Reynolds number R _K of 14. The solid line shows the velocity distribution of the smallest particles with an average particle diameter α = 0.84μ / «. The velocity of these particles is distributed from about 340 mm / sec to about 620 mm / sec over a large interval and has approximately the curve shape of the second distribution curve 21 in the three-dimensional diagram according to FIG. The curve is normalized to the maximum value of the right edge peak.

Furthermore, particles having an average particle diameter a _p of 2.1 μιι are contained in the fluid. At the same channel Reynolds number, these have a very sharp distribution in the range of about 410-415 mm / sec, as can be seen from the dashed curve. Finally, even larger particles with a mean particle diameter a _p of 6.42 μιι are contained in the fluid whose dotted-line distribution is narrower again and comprises a range of about 425 to about 428 mm / sec. The two latter curves are also normalized to the maximum value of the respective peaks. These both have the shape of the distribution according to curve 25 in the diagram in FIG. The diagram confirms the effects discussed above, namely that particles of different sizes concentrate in the fluid at different equilibrium velocities, having a very narrow velocity distribution at a sufficient flow velocity and thus being easily differentiable or identifiable. However, FIG. 13 shows yet another effect which has not been described in detail above. The fact that the smallest particles evidently have not yet reached their equilibrium position and therefore show a broad distribution in which a maximum value slowly begins to form at the left edge shows that the flow velocity and thus the parabolic distribution profile of the fluid velocity, cf. FIG 5, not yet steep enough, that the forces acting on the small particles are sufficient to move them completely into an equilibrium position. This shows that specifying an upper limit for the channel Reynolds number has no generality for achieving a narrow, significant velocity distribution. Therefore, the invention according to an advantageous embodiment, that in a system with particles having a maximum particle diameter a _p and a measuring channel, at least in the region of Measuring section, but preferably over its entire length, has a substantially constant height h, the flow rate is adjusted so that the particle Reynolds number R _P = R _K ^{■ is} greater than 2.5-10 ^"6th

The correctness of this assumption is confirmed by the diagram in Figure 14, which shows the velocity distributions of the same three particles in the same fluid and in the same device but at a mean flow velocity v of the fluid corresponding to a channel Reynolds number of 0.66. Under these conditions, neither particles with a mean particle diameter α = 0.84μ / "nor those with a mean particle diameter of 2.1 μιτι and not even the largest particles with a mean particle diameter of 6.42 μιτι arrange in the fluid.

FIG. 15 shows a two-dimensional diagram in which the signal intensity in the vertical direction is plotted against the velocity distribution in the horizontal direction. Again, this measurement contains two particle populations that are clearly separated from each other in both projection on the velocity axis and projection onto the intensity axis. If the measurement is thus carried out in such a way that the intensity of the emitted or scattered radiation is registered and coincidentally measured in addition to the velocity distribution, a further particle property can be determined in a measurement and the particles can therefore be classified or determined more precisely. The projection on the velocity axis is shown in FIG. 4 described above. The projection onto the intensity axis is shown in FIG. LIST OF REFERENCE NUMBERS

10 cylindrical channel

12 midline

14 flow direction

16 contour line

18 contour line

20 speed distribution curve

21 Velocity distribution curve

22 Velocity distribution curve

23 Velocity distribution curve

24 Velocity distribution curve

25 Velocity distribution curve

30 measuring channel

32 midline

34 flow direction

36 flow profile of the fluid

37 first particle population

38 second particle population

60 measuring channel

62 microfluidic chip

64 access, inlet

66 outlet, outlet

68 connection

70 connection

71 channel intersection 72 entrance window

74 laser beam

76 excitation zone

78 start-up route

c

O

80 measuring distance

82 detector

84 mask

86 light emission or scattering

0 88 Light emission or dispersion b Width of the measuring channel

h Height of the measuring channel

L Length of the start-up route

5 r radius or half height of the channel

V flow velocity or direction

V mean flow velocity of the fluid

V max maximum flow velocity of the fluid

^V l Particle velocity of the first particle population 0 ^V 2 Particle velocity of the second particle population

Claims

Method for detecting undissolved particles in a flowing fluid, which fluid has the density p and the viscosity μ, in which

the fluid with the particles is conveyed through a measuring channel (60),

- wherein the fluid with the particles flows through an excitation zone (76) in which it is detected by a radiation source,

- wherein the fluid with the particles flows through a measuring path (80) in which the emission emanating from the particles as a result of the excitation and / or scattered light emanating from the particles is picked up by a detector (82),

- wherein a in the region of the measuring section (80) between the measuring channel (60) and the detector (82) arranged mask (84) modulates the emission or the scattered light, and

- wherein the modulated signal is evaluated by means of a correlation analysis to detect the particles, characterized in that the fluid is conveyed through the measuring channel (60) at a mean flow velocity v, wherein the channel Reynolds number

_ P ^{"v 'd-hyd}

μ _Assumes values between 0.1 and 500, where d _{hy denotes} the hydraulic diameter of the measuring channel (60), the particles assuming a velocity distribution whose half-width is less than 4% of the mean particle velocity.

2. The method according to claim 1,

characterized in that the particles have a maximum particle diameter a _p and that the fluid with the particles in the measuring channel (60) in the flow direction in front of the measuring section (80) has a start-up path (78) with a length L of at least h ⁴

L = 3π- 3

wherein R _{K a} denotes the channel Reynolds number in the start-up section (78) and wherein the measuring channel (60) has a substantially constant height h at least in the region of the start-up section (78).

Method according to claim 1 or 2,

characterized in that the measuring channel (60) at least in the region of the measuring section has a substantially constant cross-section with a width w and a height h and that the measuring channel (60) a

Aspect ratio e = ^ <0.1 has.

Method according to one of the preceding claims,

characterized in that the particles have a maximum particle diameter a _p and that the measuring channel (60) at least in the region of the measuring section (80) has a substantially constant height h, wherein the ratio γ = - <0.75. Method according to claim 4,

characterized in that the height h of the measuring channel (60) at least in the region of the measuring section (80) does not exceed 200 times, preferably 50 times, the maximum particle diameter a _p .

Device for detecting undissolved particles in a flowing fluid, said fluid having the density p and the viscosity μ

a measuring channel (60) through which the fluid can be conveyed with the particles and which comprises a measuring section (80),

a pumping device for conveying the fluid with the particles, a radiation source for electromagnetic radiation, which is aligned with an excitation zone (76) in the measuring channel (60),

a detector (82) aligned with the measuring path (80) and arranged to detect emission from the particles as a consequence of the electromagnetic radiation and / or scattered light emanating from the particles,

a mask (84) arranged in the region of the measuring path (80) between the measuring channel (60) and the detector (82) for spatially modulating the emission and / or the scattered light, and

an evaluation device with correlation analysis means for detecting the particles on the basis of the modulated signal, characterized by a pumping means associated with the control device which is adapted to set a flow velocity v for the fluid through the entire measuring channel (30, 40, 60), wherein the channel -Reynoldszahl _ p - v ^■ d _hyd

μ

_Assumes values between 0.1 and 500, where d _{hy denotes} the hydraulic diameter of the measuring channel (30, 40, 60), the particles assuming a velocity distribution whose half-width is less than 4% of the mean particle velocity.

Device according to claim 6,

characterized in that the measuring channel (60) in the flow direction in front of the measuring section (80) a start-up path (78) with a length L of at least

1 h ⁴

L = 3π- _g

° K, a ^a p, where R _{K a} denotes the channel Reynolds number in the start-up section (78) and wherein the measuring channel (60) has a substantially constant height h at least in the region of the start-up section (78).

Apparatus according to claim 6 or 7,

characterized in that the measuring channel (60) at least in the region of the measuring section (80) has a substantially constant cross-section with a width w and a height h and that the measuring channel (60) has an aspect ratio e = - <0.1.

Device according to one of claims 6 to 8,

characterized in that the detector (82) is arranged, the intensity of the emanating from the particles emission and / or of quantitatively detecting the scattered light emanating from the particles and that the analysis device is set up to output the intensity for differentiation, classification or comminution of the particles.