Fast thermo-optical particle characterisation

Abstract

The present invention relates to a method and an apparatus for a fast thermo-optical characterisation of particles. In particular, the present invention relates to a method and a device to measure the stability of (bio)molecules, the interaction of molecules, in particular biomolecules, with, e.g. further (bio)molecules, particularly modified (bio)molecules, particles, beads, and/or the determination of the length/size (e.g. hydrodynamic radius) of individual (bio)molecules, particles, beads and/or the determination of length/size (e.g. hydrodynamic radius).

Claims

1. A method of thermo-optically measuring characteristics of a biological cell in a solution comprising: providing a sample comprising a marked biological cell in a solution; exciting fluorescently said marked biological cell and firstly detecting fluorescence of said excited biological cell; irradiating a laser light beam into the solution to obtain a spatial temperature distribution in the solution around the irradiated laser light beam; detecting secondly a fluorescence of the biological cell in the solution at a predetermined time within the range of 1 ms to 250 ms after irradiation of the laser into the solution has been started, and characterizing the biological cell based on said two detections.

2. The method according to claim 1, wherein the predetermined time is in the range of 1 ms to 50 ms.

3. The method according to claim 1, wherein the laser beam is defocused such that a temperature gradient within the temperature distribution is in the range of from 0.0 to 2K/μm.

4. The method according to claim 3, wherein the laser beam is irradiated through an optical element into the solution.

5. The method according to claim 3, wherein the optical element is a single lens.

6. The method according to claim 1, further comprising measuring the temperature distribution in said solution around the irradiated beam with a temperature sensitive dye.

7. The method according to claim 6, wherein the temperature distribution is determined based on detected fluorescence of the temperature sensitive dye, wherein the solution comprising said temperature sensitive dye is heated by the irradiated laser beam and the fluorescence spatial fluorescence intensity is measured substantially perpendicular around the laser beam.

8. The method according to claim 1, wherein the predetermined time is within the range of 0.5 s to 250 s.

9. The method according to claim 8, wherein in said predetermined time, concentration change(s) within the spatial temperature distribution in the solution due to thermophoretic effects and such (an) concentration change(s) is(are) detected by a change of the distribution of fluorescence.

10. The method according to claim 8, wherein the laser beam is focused such that a temperature gradient within the temperature distribution is achieved in the range of from 0.001 to 10K/μm.

11. The method according to claim 8, wherein said fluorescence is detected with a CCD camera.

12. The method according to claim 8, wherein the brightness of said fluorescence is detected with a photodiode or a single pixel with the CCD in the centre of the laser beam.

13. The method according to claim 1, wherein the laser light is within the range of from 1200 nm to 2000 nm.

14. The method according to claim 1, wherein the laser is a high power laser within the range of from 0.1 W to 10 W.

15. The method according to claim 1, wherein the solution is a saline solution with concentrations in the range of from 0 to 1M.

16. The method according to claim 15, wherein said temperature gradient is created within 0.1 μm to 500 μm in diameter around the laser beam.

17. The method according to claim 1, wherein the spatial temperature distribution is between 0.1° C. and 100° C.

18. The method according to claim 1, wherein the irradiation of the laser and the detection of the fluorescence is conducted from the same side with respect to the sample.

19. The method according to claim 1, wherein the solution is provided with a thickness in direction of the laser light beam from 1 μm to 500 μm.

20. The method according to claim 1, wherein the detection of the fluorescence is detected within a range of from 1 nm to 500 μm in direction of the laser beam.

21. The method according to claim 1, wherein the fluorescence is detected substantially perpendicular with respect to the laser light beam with a CCD camera.

22. The method according to claim 21, wherein the second fluorescence detection is spatial measurement of the fluorescence in dependence of the temperature distribution substantially perpendicular with respect to the laser light beam.

23. The method according to claim 1, wherein the laser beam is defocused such that a temperature gradient within the temperature distribution is in the range of from 0.0 to 5K/μm.

24. The method according to claim 1, wherein the laser is a high power laser within the range of from 4 W to 6 W.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) In the following description of examples of preferred embodiments of the invention, elements having the comparable technical or physical effect have the same reference numerals.

(2) FIG. 1A shows an fluorescence IR scanning microscope according to the present invention. The IR scanning microscope is based on a standard fluorescence microscopy setup (e.g. Zeiss AxioTech, Vario). Said device comprises: one or more light sources 32, preferably high power LED (e.g. V-Star, Luxeon) to excite the particles. The signal from the particles may be collected with an optical system 1, preferably a 40× oil objectiv and is seperated from the light of the light source by one or more light separation elements 4, preferably a dichroic mirror. The said signal is recorded with one or more detectors 31, preferably CCD Camera (e.g. SensiCam QE, PCO). The beam of an IR LASER 30 (e.g. IPG, Raman fibre RLD-5-1455) is coupled into microfluidic chamber 45 (e.g. a multiwell plate). The system may comprise other components as would ordinarily be found in fluorescence and wide field microscopes. Examples for means of excitation and detection of fluorescence may be found in: Lakowicz, J. R. Principles of Fluorescence Spectroscopy, Kluwer Academic/Plenum Publishers (1999).

(3) FIG. 1B shows a further embodiment of an fluorescence IR scanning microscope according to the present invention similar to the embodiments shown in FIG. 1Aa. However, the light source 32 is oriented in a different way with respect to the dichroic mirror 4. The testing sample 50 is sandwiched between two pieces of glass 51, preferably coverslips.

(4) FIG. 2 shows another embodiment of an IR scanning microscope according to the invention. According to this embodiment of the present invention, the testing sample 50 is provided in form of a single droplet with the marked particles. The volume (preferably some nanoliter up to some microliter) of such a droplet can be easily adjusted such that also the dimensions, i.e. the thickness of the droplet which is irradiated with the laser beam, is predictable. In this embodiment, the laserbeam, the fluorescently exciting light from the light source 32, preferably a LED as well as the measured fluorescent light are all focussed by a common optical system 1, preferably a microscope objective with high numerical aperture and more preferably a objectiv with high numerical aperture and high IR transmission. Therefore, the LED, the LASER 30 and the detector 31, preferably a CCD can be arragend at a common side with respect to the sample. The exciting light is separated from the fluorescent light by a light separation element 5, preferably a dichroic mirror, which preferably seperates different parts of the light spectrum than the light separation element 4.

(5) FIG. 3A-D show how melting curves with an radiation of 150 ms can be taken. (A) By measuring the fluorescence of a temperature sensitive dye, the temperature distribution in the microfluidic chamber can be measured. (B) shows the radial average of the temperatures measured by fluorescence. (C) Aprox. 150 ms after the IR LASER is turned on an image of the fluorescently labelled DNA is taken. The high intensity shows melted ds DNA. From (A) and (C) the melting curve can be determined very fast (D).

(6) FIG. 4 shows a fast SNP detection. A 16 mer of dsDNA with a single nucleotide mismatch in the center (blue) is compared with the wild type (black). Within 150 ms both species can be clearly discriminated. The ordinate describes the fraction of dissociated dsDNA. The melting point is defined by 50% dissociated molecules. It is shifted by 15° C. by a single mismatch.

(7) FIG. 5A-5B show the mobility in a temperature gradient. The figures show the change of concentration in the central pixel of a heat spot over time. A few seconds after the start of the measurement the laser is turned on and the concentration decreases until a steady state is reached. The signal allows to distinguish between a 20 mer and 50 mer dsDNA (A) as well as between 20 bases ssDNA and 20 basepair dsDNA (B).

(8) FIG. 6 is an example for a fluorescence dye to be used in the methods of the present invention (6-carboxy-2′,4,4′,5′,7,7′-hexachlorofluorescein (HEX, SE; C20091) from Invitrogen).

(9) FIGS. 7-14 show further information with regard to a detailed example according to the present invention. These figures show in particular:

(10) FIG. 7: illustrates how thermodiffusion manipulates the DNA concentration by small temperature differences within the bulk solution. A thin water film is heated by 2 K along the letters “DNA” with an infrared laser. For a cooled chamber at 3° C., fluorescently tagged DNA accumulates to the warm letters. However at room temperature DNA moves into the cold, showing reduced fluorescence. The chamber is 60 μm thin, containing 50 nM DNA in 1 mM TRIS buffer. Every 50th base pair is labelled with TOTO-1.

(11) FIG. 8A-B illustrates the salt dependency. (A) Thermodiffusion in water is dominated by ionic shielding and water hydration. (B) Soret coefficient ST versus Debye length for carboxyl modified polystyrene beads of diameter 1.1, 0.5 and 0.2 μm. Linear plot (left) and logarithmic plot (right). The Soret coefficients are described by equation (2) with an effective surface charge of σeff=4500 e/μm2 known from electrophoresis. The intercept S.sub.T (λ.sub.DH=0) is fitted with an hydration entropy per particle surface of Shyd=−1400 J/(mol Kμm2).

(12) FIG. 9A-B shows a temperature dependency. (A) The temperature dependence is dominated by the linear change in the hydration entropy Shyd. It shifts the salt dependent thermodiffusion ST(λDH) to lower values. The particle size is 1.1 μm. (B) The Soret coefficient ST increases linearly with the temperature as expected for a hydration entropy Shyd(T). It depends on the molecule species, not its size, as seen from the rescaled Soret coefficients for DNA with different lengths.

(13) FIG. 10A-D shows a size dependency. (A) For polystyrene beads, the Soret coefficient scales with the particle surface over four orders of magnitude. Measurements are described by equation (2) with an effective surface charge density of σeff=4500 e/μm2 and negligible hydration entropy. The deviation for the bead with 20 nm diameter can be understood from an increased effective charge due to the onset of charge normalization for a ≤λ.sub.DH. (B) Accordingly, the thermodiffusion coefficient DT scales linearly with bead diameter. (C) The Soret coefficient of DNA scales according to S.sub.T∝√{square root over (L)} with the length L of the DNA based equation (2) with an effective charge per base pair of 0.12 e. (D) Thermodiffusion coefficient DT decreases over DNA length with D.sub.T∝L.sup.−0.25, caused by the scaling of diffusion coefficient D∝L.sup.−0.75.

(14) FIG. 11A-B: shows an effective Charge from Thermodiffusion. Effective charge is inferred from thermodiffusion using equation (3). Polystyrene beads (20 . . . 2000 nm) (A) and DNA (50-50,000 bp) (B) are measured over a large size range, impossible with electrophoresis. As expected, the effective charge of the beads scales with particle surface and linearly with length of DNA.

(15) FIG. 12A to 12D: shows the dependency of thermodiffusion from concentration over time. (A) Diffusion coefficient D was obtained from back diffusion after switching off the heat source. (B) D is varied until the finite element simulation matches the experiment. (C) Radial depletion of DNA from a focussed 2K heat spot is monitored over time. (D) Comparison with simulation with known D yields DT and ST.

(16) FIG. 13 shows a scaling of DNA diffusion coefficients. The diffusion coefficients as measured in this study at room temperature. The scaling over DNA length matches literature values with two scaling regimes with exponent −1 for short and −0.6 for long DNA33. As approximation, diffusion across the two scaling regimes is well described with an overall exponent of −0.75.

(17) FIG. 14A-B: shows a simulation of Microfluidic Heating. (A) A 10 μm thin water film is enclosed between PS walls. Low thermal conduction of the chamber walls allow a thickness independent temperature profile, confirmed by the shown finite element calculation. (B) Convection is slow at maximal velocities of 5 nm/s due to thin chamber and comparable broad heating focus.

(18) FIG. 15 shows the temperature dependence of a fluorescent dye measured by a fluorimeter with temperature control (Peltier element).

(19) FIG. 16 illustrates a particular embodiment of a device accordingly to the present invention. The device may have an substantially arbitrary orientation with respect to the direction of gravitation, preferably the device is oriented perpendicular with respect to the direction of gravitation, more preferably the device is oriented substantially parallel or anti-parallel with respect to the direction of gravitation. Preferably the orientation of the device with respect to the testing sample or chamber may be adjusted as shown in FIG. 1a, FIG. 1b and FIG. 2. The device comprises: 1: Objective (e.g. 40×, NA 1.3, oil immersion, ZEISS “Fluar”); 20: Scanning module, may be Galvano scanning mirror or may be acoustic optic deflector (AOD); 3: cold mirror, preferably high IR-transmission and preferably >90% reflection 350 nm-650 nm; 11: beam shaping module to determine laser beam diameter and focusing, may be a lens system which may comprise one, two or more lenses; 16: Laser fibre coupler w/o collimator; 15: Laser fibre (single mode or multimode); 30: IR-Laser (e.g. 1455 nm, 1480 nm, 0.1 W-10 W); 4: Dichroic mirror/beam splitter reflecting short wavelength (R>80%), transmitting long wavelength (T>80%); 7: Emission Filter (band pass/long pass); 31: detector, may be a CCD-Camera, Line-Camera, Photomultiplier Tube (PMT), Avalanche Photodiode (APD), CMOS-Camera; 6: Excitation Filter (e.g. band pass/long pass); 10: lens system to determine the beam properties of the excitation light source, may comprise one, two or more lenses; 32: Excitation light source, may be Laser, Fibre Laser, Diode-Laser, LED, HXP, Halogen, LED-Array, HBO. Preferably the listed parts are enclosed in a housing; a cold mirror is a specialized dielectric mirror, a dichromatic interference filter that operates over a very wide temperature range to reflect the entire visible light spectrum while very efficiently transmitting infrared wavelengths.

(20) FIG. 17 shows an embodiment of the invention according to FIG. 16, wherein the scanning module 20 is replaced by a fix mirror 21, preferably a silver mirror.

(21) FIG. 18 shows an embodiment of the invention according to FIG. 16, wherein the scanning module 20 is replaced by a fix mirror 21, preferably a silver mirror, and wherein a shutter 33 for controlling the IR LASER radiation is added and wherein a line forming module 12, may be a cylinder lens system or preferably a Powell lens, are added.

(22) FIG. 19 shows a further embodiment of the device according to the present invention, in particular a confocal setup. The device may have an arbitrary orientation with respect to the direction of gravitation, preferably the device is oriented substantially perpendicular with respect to the direction of gravitation, more preferably the device is oriented substantially parallel or anti-parallel with respect to the direction of gravitation. Preferably the orientation of the device with respect to the testing sample or chamber may be adjusted as shown in FIG. 1A, FIG. 1B and FIG. 2. The device comprises: 1: Objective (e.g. 40×, NA 1.3, oil immersion, ZEISS “Fluar”); 2: hot mirror, high IR-reflection, visible light transmission>80%; 4: Dichroic mirror reflecting short wavelength (R>80%), transmitting long wavelength (T>80%); 7: Emission Filter (band pass/long pass); 31: Detector, may be Photomultiplier Tube (PMT), Avalanche Photodiode (APD); 13: Pinhole aperture; 10: lens system to determine the beam properties of the excitation light source, may comprise one, two or more lenses; 32: excitation light source preferably a laser, more preferably a fibre coupled laser 11: beam shaping module to determine laser beam diameter and focusing, preferably lens system which may comprise one, two or more lenses; 16: Laser fibre coupler w/o collimator; 15: Laser fibre (single mode or multimode); 30: IR-Laser (e.g. 1455 nm, 1480 nm, 0.1 W-10 W); 33: Shutter; 14: pinhole aperture in confocale position to laser pinhole 13 or to laser fibre coupler 17 whereas the pinhole 13 may not be needed, preferably if a fibre coupled laser is used as excitation light source 32; 18 Laser fibre may be single mode or may be multi mode.

(23) FIG. 20 shows a further embodiment of the device according to the present invention. The device may have an arbitrary orientation with respect to the direction of gravitation, preferably the device is oriented substantially perpendicular with respect to the direction of gravitation, more preferably the device is oriented substantially parallel or anti-parallel with respect to the direction of gravitation. Preferably the orientation of the device with respect to the testing sample or chamber may be adjusted as shown in FIG. 1A, FIG. 1B and FIG. 2. The device comprises: 1: Objective (e.g. 40×, NA 1.3, oil immersion, ZEISS “Fluar”); 2: hot mirror, high IR-reflection, visible light transmission>80%; 4: Dichroic mirror reflecting short wavelength (R>80%), transmitting long wavelength (T>80%); 6: Excitation Filter (e.g. band pass/long pass); 7: Emission Filter (band pass/long pass); 10: lens system to determine the beam properties of the excitation light source; 31: Detector, may be Photomultiplier Tube (PMT), Avalanche Photodiode (APD); 32: excitation light source 16: Laser fibre coupler w/o collimator; 15: Laser fibre (single mode or multimode); 30: IR-Laser (e.g. 1455 nm, 1480 nm, 0.1 W-10 W); a hot mirror is a specialized dielectric mirror, a dichromatic interference filter often employed to protect optical systems by reflecting heat back into the light source. Hot mirrors can be designed to be inserted into the optical system at an incidence angle varying between zero and 45 degrees, and are useful in a variety of applications where heat build-up can damage components or adversely affect spectral characteristics of the illumination source. By transmitting visible light wavelengths while reflecting infrared, hot mirrors can also serve as dichromatic beam splitters for specialized applications in fluorescence microscopy.

(24) FIG. 21 shows an embodiment of the invention according to FIG. 20, wherein a shutter 33 is added to control the IR laser radiation.

(25) FIG. 22 shows an embodiment of the invention according to FIG. 20, wherein a shutter 33 and a line forming module 12, preferably a lens system comprising one, two or more lenses or more preferably a Powell lens are added to control the IR laser radiation.

(26) FIG. 23 shows an embodiment of the invention according to FIG. 16, wherein the scanning module 20 is replaced by a fix mirror 21, preferably a silver mirror and wherein an additional light separation element 5, preferably a dichroic mirror, which preferably separates different parts of the light spectrum than the light separation element 4 is added and wherein an emission filter 8 is added, preferably transmitting another range of wavelength than the emission filter 7 and wherein a second detector 31 is added which detects the signal passing through the filter 8.

(27) FIG. 24 shows a further embodiment of the device; 31: CCD camera; 7: Emission filter (e.g. bandpass or longpass); 4: light separation element, preferably a dichroic mirror for splitting excitation light pathway and emission light pathway; 1: Objective; 45: Chamber; 10: Lens system for focusing IR-Laser onto sample; 20: scanning module, may be galvanometric mirrors or an acoustic optical deflector; 16: Laser fibre coupler with collimating optics; 15: Laser fibre; 30: IR-Laser; 6: Excitation Filter; 10: optical system for excitation, may be a lens system comprising one, two or more lenses; 32: Light source (HXP,LED); 46: xyz-Translation stage for laser positioning, may be automated, preferably may be automated for scanning the chamber. 47: optical system for beam shaping of the IR Laser, may be a lens system comprising one, two or more lenses or may be a objective, preferably with high IR transmission.

(28) FIG. 25A-B shows a quantification of interaction between biomolecules. 100 nM of a fluorescently labelled antibody (anti-Interleukin 4) are titrated with various amounts interleukin. (left) The spatial fluorescence distribution in steady state is measured. Three curves with 5 nM, 80 nM and 300 nM are shown exemplarily in FIG. 25A. The signal changes dramatically from fluorescence decrease to a fluorescence increase. Integration of the fluorescence profile up to 80 μm (distance from the heated centre) allows to determine the number of complexes in solution, as shown in FIG. 25B. The concentration of free Interleukin 4 can be calculated plotted versus the concentration formed complex. These data can be fitted to determine the K.sub.D.

(29) FIG. 26 shows a single molecule binding to nanoparticles. The Soret coefficient of nanocrystals in complex with PEG molecules is measured by evaluating the concentration profile in steady state. The Soret coefficient increases linearly with the number of PEG molecules covalently bound to the particle. PEG molecules with a higher molecular weight show a steeper increase in the Soret coefficient. The PEG molecules shown here are comparable in size to proteins or short DNA molecules, which can be detected in the same manner.

(30) FIG. 27 shows a further embodiment of a device according to the present invention. The receiving means for receiving the sample probe 50 is a capillary 40 with inner diameter 5 μm to 500 μm such that the thickness of the sample probe is small in the direction perpendicular to the laser beam. The first valve 41 and the second valve 42 are provided for the controlled input/output of the sample probe 50 in/from the capillary 40. The capillary is mounted on a solid support 43, preferably a material with good thermal conductivity, e.g. aluminium, copper. The Peltier-element 44 is mounted on the solid support 43 such that the capillary 40 can be cooled.

(31) FIG. 28A-B shows the characterization of protein conformation. Thermo-optical characterization provides the means to characterize the conformation of a protein in solution. In FIG. 28A the temperature of a solution containing Bovine Serum Albumin (BSA) is cooled down to 0° C. Starting from this temperature the Soret coefficient is measured at different temperatures which are increased in a stepwise manner up to 60° C. The Soret coefficient is negative, up to values close to thermal denaturation, where a sudden jump to positive Soret coefficients is observed. At physiological temperatures (30-40° C.) the Soret coefficient does not change much. In this temperature range the protein has to have similar properties to perform its tasks. Because of the tight relation between structure and function, the conformation is preserved in this temperature range. The results are confirmed by the experiment shown on the right (FIG. 28B), which starts at high temperatures. The Soret Coefficients are still positive below 50° C. since the measurements are faster than the time the protein needs for refolding. After a certain time span (i.e. 20 minutes) the values reach the negative Soret coefficients obtained in the measurements started at low temperatures. A following temperature increase reproduces the negative Soret coefficient measured in the experiment starting at low temperatures.

(32) FIG. 29 shows measurements with a sample of fluorescently labelled bovine serum albumin (BSA). A sample of fluorescently labelled bovine serum albumin (BSA) has been split in two parts. One is only exposed to ambient temperatures, while the other half is heated up to 100° C. for several minutes. The thermo-optical properties of both samples (native and denatured) are measured at different power of the infrared laser (i.e. maximal temperature increase of 5° or 10° C.). As can be seen from the figure, the fluorescence of the denatured protein is lower than the fluorescence of the native protein. This is explained as follows. The fluorescence dye of both samples shows the same decrease in fluorescence due to the increase in temperature (i.e. temperature sensitivity of the fluorescence). But the denatured protein shows a positive thermophoretic mobility (i.e. moves to the cold), while the native protein has a negative thermophoretic mobility (i.e. moves to the hot). The accumulation at elevated temperatures is the reason, why the decrease in fluorescence is lower for the native protein, while the denatured protein is, in addition to temperature dependency, depleted from the region of elevated temperature. The differences between both samples is further increases by raising the temperature (i.e. maximum temperature of 10° C.), since positive and negative thermophoresis is enhanced.

(33) FIG. 30 measurements with a sample of fluorescently labelled bovine serum albumin (BSA). A sample of fluorescently labelled bovine serum albumin (BSA) has been split in two parts. One is only exposed to ambient temperatures, while the other half is heated up to 100° C. for several minutes (i.e. irreversibly denatured). The thermo-optical properties of both samples (native and denatured) are measured at 800 mA power of the infrared laser (i.e. maximal temperature increase of 20° C.). As can be seen from the figure, the fluorescence of the denatured protein is lower than the fluorescence of the native protein. This is explained as follows. The fluorescence dye of both samples shows the same decrease in fluorescence due to the increase in temperature (i.e. temperature sensitivity of the fluorescence). But the denatured protein shows a positive thermophoretic mobility (i.e. moves to the cold), while the native protein has a negative thermophoretic mobility (i.e. moves to the hot). The accumulation at elevated temperatures is the reason, why the decrease in fluorescence is lower for the native protein, while the denatured protein is, in addition to temperature dependency, depleted from the region of elevated temperature. Interestingly, by approaching the denaturing temperature (i.e. 50° C.) of the protein the amplitudes of the native and denatured protein approach each other an are essentially the same. This means that by measuring the amplitude of the fluorescence change an comparison to the reference sample allows to detect the melting temperature of a protein and to discriminate between native and denatured form of a protein. And to detect a shift in melting temperature due to interactions of the protein with other biomolecules or small molecules (e.g. drug candidates).

(34) FIG. 31A-B shows the measurement of the thermo-optical properties of two samples with/of Green Fluorescent protein (GFP). The thermo-optical properties of two samples of Green Fluorescent protein (GFP) are measured. In the first sample (FIG. 31A) only GFP is present, while in the second sample (FIG. 31B) the GFP is mixed with a 2 fold excess of an antibody fragment, specifically binding to GFP. In both cases, first the fluorescence is recorded without laser heating. Then the fluorescence excitation is turned off and the IR-laser radiation is turned on. The laser is turned off after a few seconds of heating and the fluorescence excitation is turned on at the same time. The relaxation of the spatial fluorescence distribution (i.e. concentration distribution) to a homogeneous state is recorded for a few seconds. As can be observed from the figure, in the sample with the two interacting species (i.e. GFP and the antibody fragment) the fluorescence profile needs more time to relax. This is explained by slower diffusion of the larger complex. The time evolution of the fluorescence profile is analyzed via a software tool to determine the diffusion constant. By using the Stokes-Einstein relation, an hydrodynamic radius is attributed to the diffusion constant. In case of the free GFP this is 5 nm and the complex has an radius of 10 nm.

(35) FIG. 32A-C shows a measurement of a particle which is trapped in a potential well created by a spatial temperature distribution. (A) A particle is trapped in a potential well created by a spatial temperature distribution. For silica particles the well is deepest at high temperatures. The fluctuations are recorded via a CCD camera (at t=1 s, 2 s, 3 s, 4 s, 5 s, 6 s, 7 s) and (B) the positions are tracked by Software with nanometer resolution. (C) A histogram is calculated from the positional information. The width of the distribution is very sensitive to the thermo-optical properties of the particle. If molecules bind to the surface of the particle, the effective potential for the bead changes and the amplitude of the fluctuations increases or decreases. By observing the amplitude change over time, a kinetic binding curve can be measured.

(36) FIG. 33A-D shows a time series of the thermophoretic motion of silica beads in a microfluidic chamber. Time series of the thermophoretic motion of silica beads in a microfluidic chamber. In the beginning (image A), without laser heating, the beads are almost equally distributed. The black circle shows the position of the laser focus. The following images (B-D) show the development of the particle distribution in the next three seconds after the heating laser is turned on. The particles are attracted by the heat source and accumulate at the point of highest temperature. The accumulation if observed because these particles have a negative thermophoretic mobility. A Particle with positive thermophoretic mobility can be trapped by heating e.g. a circle around it.

(37) FIG. 34 shows another example for the “Optothermal Trap”. Another Example for the “Optothermal Trap”: Several 1 μm beads are trapped at the bright spot at the centre of the image. The chamber is moved whereas the laser focus was kept fixed. The image is an addition of about 30 single images. As one can see, all beads were moved with the chamber, so the addition of the single images results in lines for the single particles. The trapped beads were hold on one position. No movement of the trapped beads was detected. The halo and the high intensity of the trapped beads is the result of the addition of the single images.

(38) FIG. 35A shows the determination of the Soret Coefficient of complexes of nanocrytals (=quantum dot QD) and biomolecules. The Soret Coefficient of complexes of nanocrytals (=quantum dot QD) and biomolecules is determined by relating the spatial concentration distribution to a spatial temperature distribution. Three different samples have been analyzed. First a nanocrystal without protein modification is measured (QD), followed by a sample nanocrystals modified with the protein streptavidin (QD+Strep.)(approximately 5 proteins per nanocrystal). By binding the protein to the nanocrystal, the Soret-Coefficient is strongly increased. By adding a single stranded DNA to the sample (one DNA per Particle), the Soret coefficient is increased further (QD+Strep.+DNA). FIGS. 35B-D show the configurations of the complexes.

(39) FIG. 36 shows a further embodiment of the invention according to FIG. 20, wherein a stage 43 carrying a temperature control element 44 and the chamber 45 is connected to the optical system via connectors 48. The optical system 1 may also be also comprise a TIRF (total internal reflection fluorescence) objective so that Thermophoresis can be measured in direction of the Laser beam.

(40) FIG. 37 shows a further embodiment of the invention according to FIG. 20, wherein a shutter 33 and a line forming module 12, preferably a lens system comprising one, two or more lenses or more preferably a Powell lens are added to control the IR laser radiation and wherein the emission filter 7 is replaced by an optical instrument 22 which may be a spectrograph, polychromator or monochromator or combinations of one or more of these, e.g. an optical instrument which transforms different intervals of wavelength/frequency of light into different intervals of angles/distances or different places for example on a CCD.

(41) FIG. 38 shows an example for a lipid bilayer model system. A fraction of the layer constituting lipids is coupled to a surface (e.g. via a sulfhydryl-peptide to a gold surface) Transmembrane proteins or membrane associated proteins are inserted into the lipid bilayer. In addition also soluble proteins may be present in the aqueous solution on top of the membrane. By infrared laser absorption of the aqueous solution a temperature gradient can be established within the membrane. This way the thermo-optical properties like stability, interaction and conformation may can be measured for a fluorescently labelled compound (i.e. lipid, membrane protein or soluble protein).

(42) FIG. 39A-B shows thermophoresis and thermophoretic trapping of lipid vesicles. The images (200×200 μm) show a solution of unilamelar vesicles, without (A) and after 10 seconds of infrared laser heating. (A) shows a uniform distribution of the vesicles. The infrared laser heats the solution locally to a maximum temperature of 15° C. above room temperature of 20° C. As (B) shows the local temperature increase attracts the vesicles (i.e. negative thermophoresis) and confines their position to a region close to the center of the heat spot. The region around the heat spot is depleted of vesicles. Vesicles closer to the edge of the field of view experience only a small gradient an are not attracted within the time span of 10 seconds. Broadening of the temperature profile would also attracted these particles much faster.

EXAMPLES

(43) The following detailed example illustrates the invention without being limiting.

Example 1

Thermodiffusion

(44) Molecules drift along temperature gradients, an effect called thermophoresis, Soret-effect or thermodiffusion. In liquids, its theoretical foundation is subject of a long standing debate. Using a new all-optical microfluidic fluorescence method, we present experimental results for DNA and polystyrene beads over a large range of particle size, salt concentration and temperature. The data supports a unifying theory based on the solvation entropy. Stated in simple terms, the Soret coefficient is given by the negative solvation entropy, divided by kT. The theory predicts the thermodiffusion of polystyrene beads and DNA without any free parameters. We assume a local thermodynamic equilibrium of the solvent molecules around the molecule. This assumption is fulfilled for moderate temperature gradients below the fluctuation criterion. Above this criterion, thermodiffusion becomes non-linear. For both DNA and polystyrene beads, thermophoretic motion changes sign at lower temperatures. This thermophilicity towards lower temperatures is attributed to an increasing positive entropy of hydration, whereas the generally dominating thermophobicity is explained by the negative entropy of ionic shielding. The understanding of thermodiffusion sets the stage for detailed probing of solvation properties of colloids and biomolecules. For example, we successfully determine the effective charge of DNA and beads over a size range which is not accessible with electrophoresis.

(45) Introduction. Thermodiffusion has been known for a long time, but its theoretical explanation for molecules in liquids is still under debate. The search for the theoretical understanding is motivated by the fact that thermodiffusion in water might lead to powerful all-optical screening methods for biomolecules and colloids. Equally well, thermodiffusion handles and moves molecules all-optically and therefore can complement well established methods as for example electrophoresis or optical-tweezers. For the latter, forces of optical tweezers scale with particle volume and limit this method to particles only larger than 500 nm. Electrophoresis does not suffer from force limitations, but is difficult to miniaturize due to electrochemical reactions at the electrodes.

(46) On the other hand, thermodiffusion allows the microscale manipulation of even small particles and molecules. For example, 1000 bp DNA can be patterned arbitrarily in bulk water (FIG. 7). The temperature pattern “DNA”, heated by 2 K, was written into a water film with an infrared laser scanning microscope. The concentration of 1000 bp DNA was imaged using a fluorescent DNA tag. In an overall cooled chamber at 3° C., DNA accumulates towards the heated letters “DNA” (negative Soret effect) whereas at room temperature DNA is thermophobic (positive Soret effect) as seen by the dark letters.

(47) In the past, the apparent complexity of thermodiffusion prevented a full theoretical description. As seen for DNA in FIG. 13, molecules characteristically deplete from regions with an increased temperature, but they can also show the inverted effect and accumulate.sup.3. Moreover, the size scaling of thermodiffusion recorded by thermal field flow fractionation (ThFFF) showed fractional power laws with a variety of exponents which are hard to interpret.sup.4. The latter effect was resolved recently by revealing nonlinear thermophoretic drift for the strong thermal gradients used in ThFFF.

(48) A variety of methods were used to measure thermodiffusion, mostly in the nonaqueous regime. They range from beam deflection.sup.3,7, holographic scattering.sup.8,9, electrical heating to thermal lensing. Recently we have developed a fluorescence microfluidic imaging technique.sup.13,14 which allows the measurement of thermodiffusion over a wide molecule size range without artifacts induced by thermal convection. Highly diluted suspensions can be measured and therefore particle-particle interactions do not have an influence. We only apply moderate temperature gradients. In the following we used this method to confirm a straightforward theoretical explanation of thermodiffusion.

(49) Theoretical Approach. For diluted concentrations, it is generally assumed that the thermodiffusive drift velocity {right arrow over (V)} depends linearly on the temperature gradient ∇T with a proportionality constant which equals the thermodiffusion coefficient D.sub.T: {right arrow over (v)}=−D.sub.T∇T. In steady state, thermodiffusion is balanced by ordinary diffusion. Constant diffusion and thermodiffusion coefficients both lead to an exponential depletion law.sup.16 c/c.sub.0=exp[−(D.sub.T/D)(T-T.sub.0], with the concentration c depending on the temperature difference T-T.sub.0 only. The concentration c is normalized by the boundary condition of the concentration c.sub.0 with temperature T.sub.0. The Soret coefficient is defined as ratio S.sub.T=D.sub.T/D which determines the magnitude of thermodiffusion in the steady state. While the above exponential distribution could motivate an approach based on Boltzmann equilibrium statistics, it is commonly argued that thermodiffusion without exception is a local non-equilibrium effect that requires fluid dynamics, force fields or particle-solvent potentials.sup.17-20. However, in two previous papers.sup.16 we demonstrated that for moderate temperature gradients, the thermal fluctuations of the particle are the basis for a local equilibrium. This allows the description of the thermodiffusive steady state by a succession of local Boltzmann laws, yielding c/c.sub.0=exp[(G(T.sub.0)—G(T))/kT] with G the Gibbs-free enthalpy of the single particle-solvent system. Such an approach is only valid if the temperature gradient ∇T is below a threshold ∇T<(aS.sub.T).sup.−1 which is given by the particle fluctuations with the hydrodynamic radius a and Soret coefficient S.sub.T, as shown recently. For larger temperature gradients, thermodiffusive drift is nonlinearly dependent on the temperature gradient. In the present study, temperature gradients below this limit were used so that thermodiffusion is measured at local thermodynamic equilibrium conditions.

(50) Local thermodynamic equilibrium allows the derivation of a thermodynamic foundation of the Soret coefficient. The local Boltzmann distribution relates small concentration changes δc with small Gibbs-Free Energy differences: δc/c=−δG/kT. We equate this relation with a locally, linearized thermodiffusion steady state given by δc/c=−S.sub.TδT and thus find the Soret coefficient by the temperature derivative of G
S.sub.T=D.sub.T/D=(kT).sup.−1×∂G/∂T  (1)

(51) Whereas above relation is sufficient for the following derivation, it can be generalized by locally applying the thermodynamic relation dG=−SdT+Vdp+μdN. For single particles at a constant pressure we find that the Soret coefficient equals the negative entropy of the particle-solvent system S according to S.sub.T=−ΔS/kT. This relation is not surprising since the entropy is by definition related with the temperature derivative of the free enthalpy.

(52) The above general energetic treatment is inherent in previously described approaches based on local equilibrium.sup.22, including the successful interpretation of thermoelectric voltages of diluted electrolytes.sup.24,25 which are described by energies of transfer. Recently, the non-equilibrium approach by Ruckenstein was applied to colloids.sup.27 with the characteristic length 1 assigned to the Debye length λ.sub.DH. If instead it would assigned the characteristic length according to 1=2a/3 with the particle radius a, the Ruckenstein approach would actually confirm the above local equilibrium relation (1) for the Soret coefficient. Measurements on SDS micelles.sup.27 appeared to confirm this non-equilibrium approach, but for the chosen particles the competing parameter choices 1=2a/3 and 1=λ.sub.DH yielded comparable values. Thus the experiments could not distinguish between the competing theories.

(53) Here, we will use the above local equilibrium relations to derive the Soret coefficient for particles larger than the Debye length in aqueous solutions and put the results to rigorous experimental tests. Two contributions dominate the particle entropy S in water (FIG. 8a): the entropy of ionic shielding and the temperature sensitive entropy of water hydration. The contribution from the entropy of ionic shielding is calculated with the temperature derivative of the Gibbs-free enthal.sup.27,28 G.sub.ionic=Q.sub.eff.sup.2λ.sub.DH/[2Aεε.sub.0] with the effective charge Q.sub.eff and particle surface A. Alternatively, this enthalpy can be interpreted as an electrical field energy G.sub.ionic=Q.sub.eff.sup.2/[2C] in the ionic shielding capacitor C. We neglect the particle-particle interactions since the fluorescence approach allows the measurement of highly diluted systems. To obtain the Soret coefficient, temperature derivatives consider the Debye length λ.sub.DH(T)=√{square root over (ε(T)ε.sub.0kT/(2e.sup.2c.sub.S))} and the dielectric constant δ(T). Both temperature derivatives give rise to a factor β=1−(T/ε)∂ε/∂T. The effective charge Q.sub.eff is largely temperature insensitive which was confirmed by electrophoresis independently.sup.29. Such a dependence would be unexpected as the strongly adsorbed ions dominate the value of the effective charge. Experimentally, we deal with colloids exhibiting flat surfaces, i.e. the particle radius is larger than λ.sub.DH. In this case charge renormalization does not play a role and we can introduce an effective surface charge density σ.sub.eff=Q.sub.eff/A per molecule area A. From the temperature derivation according to equation (1), the ionic contribution to the Soret coefficient is S.sub.T.sup.(ionic)=(Aβσ.sub.eff.sup.2λ.sub.DH)/(4εε.sub.0kT.sup.2). A similar relation was derived for charged micelles recently.sup.23, however without considering the temperature dependence of the dielectric coefficient ε. Next, the contribution to the Soret coefficient from the hydration entropy of water can be directly inferred from the particle area specific hydration entropy S.sub.hyd=S.sub.hyd/A, namely S.sub.T.sup.(hyd)=−As.sub.hyd(T)/kT. Finally, the contribution from the Brownian motion is derived as S.sub.T=1/T by inserting the kinetic energy of the particle G=kT into equation (1). However this contribution is very small (S.sub.T=0.0034/K) and can be neglected for the molecules under consideration. The contributions from ionic shielding and hydration entropy add up to:

(54) $\begin{array}{cc}{S}_T=\frac{A}{\mathrm{kT}}\left(-s_{\mathrm{hyd}}+\frac{{\mathrm{\beta \sigma }}_{\mathrm{eff}}^2}{4{\mathrm{.Math..Math.}}_{0}T}\times {\lambda }_{\mathrm{DH}}\right)& \left(2\right)\end{array}$

(55) The Soret coefficient S.sub.T scales linearly with particle surface A and Debye length λ.sub.DH. We test equation (2) by measuring S.sub.T versus the salt concentration, temperature and molecule size. In all cases thermodiffusion is quantitatively predicted without any free parameters. We used fluorescence single particle tracking to follow carboxyl modified polystyrene (PS) beads (Molecular Probes F-8888) of 1.1 μm and 0.5 μm diameter at 25 attomolar concentration, dialyzed into 0.5 mM Tris-HCl at pH 7.6. Thermodiffusion of particles ≤0.2 μm is measured by the fluorescence decrease that reflects the bulk depletion of the particles.sup.13. The chamber thickness of 20 μm damped the thermal convection to negligible speeds.sup.16. The experimental design also excludes thermal lensing and optical trapping.sup.16. Debye lengths λ.sub.DH were titrated with KCl (see supplementary materials).

(56) Salt dependence. FIG. 8b shows the Soret coefficients of polystyrene beads with different sizes versus λ.sub.DH. The Soret coefficients scale linearly with a small intercept at λ.sub.DH=0 and confirm the λ.sub.DH-dependence of equation (2). For smaller diameters of the beads the Soret coefficients scale with the particle surface area A (FIG. 8) as expected from equation (2). To check whether equation (2) also quantitatively explains the measured Soret coefficients, we inferred the effective charge of the beads by electrophoresis (see supplementary material). Using 40 nm beads with identical carboxyl surface modifications at λ.sub.DH=9.6 nm, we fluorescently observed free-flow electrophoresis and corrected for electroosmosis, finding an effective surface charge density of σ.sub.eff=4500±2000 e/μm.sup.2. This value is virtually independent from the used salt concentrations.sup.29. Using this inferred effective charge, equation (2) fits the Soret coefficient for various bead sizes and salt concentrations well (FIG. 8b, solid lines).

(57) The intercept S.sub.T(λ.sub.DH=0), where ionic contributions are zero, also scales with particle surface and is described by a hydration entropy per particle surface of S.sub.hyd=1400 J/(molKμm.sup.2). The value matches the literature values for similar surfaces reasonably well.sup.30,31. For example, Dansyl-Alanine, a molecule with surface groups comparable with polystyrene beads, was measured to have a hydration entropy.sup.30 of −0.13 J/(molK) at a comparable temperature. Linear scaling with its surface area by using a radius of a=2 nm results in a value of s.sub.hyd=−2500 J/(molKμm.sup.2), in qualitative agreement with our result. The hydration entropy is a highly informative molecule parameter which is notoriously difficult to measure, yielding an interesting application for thermodiffusion.

(58) Temperature dependence. Hydration entropies S.sub.hyd in water are known to increase linearly with decreasing temperatures.sup.30-32. Since the slope of the ionic contribution of S.sub.T versus λ.sub.DH is with high precision temperature insensitive for water (β(T)/(εT.sup.2)≅const), only the intercept is expected to decrease as the overall temperature of the chamber is reduced. This is indeed the case, as seen from the temperature dependence of beads with 1.1 μm diameter (FIG. 9a, T=6 . . . 29° C.). We infer from the intercept S.sub.T(λ.sub.DH=0) that the hydration entropy changes sign at about 20° C. As seen for DNA in FIG. 7, hydration entropy can dominate thermodiffusion at low temperatures and move molecules to the hot (D.sub.T<0).

(59) The properties of hydration entropy lead to a linear increase of S.sub.T over temperatures at fixed salt concentration as measured for 1.1 μm beads and DNA (FIG. 9b). We normalize S.sub.T by dividing with S.sub.T(30° C.) to compensate for molecule surface area. The slopes of S.sub.T over temperature differ between beads and DNA. However the slope does not differ between DNA of different size (50 base pairs versus 10000 base pairs). Based on equation (2), this is to be expected since the temperature dependence of the hydration entropy only depends on the type of surface of the molecule, not its size. We measured the diffusion coefficients of the DNA species at the respective temperature independently. Within experimental error, changes in the diffusion coefficient D match with the change of the water viscosity without the need to assume conformational changes of DNA over the temperature range. Please note that the change of the sign of the DNA Soret coefficient is situated near the point of maximal water density only by chance. There, the two entropic contributions balance. For polystyrene beads at λ.sub.DH=2 nm for example, the sign change is observed at 15° C. (FIG. 9a). An increased Soret coefficient over temperature was reported for aqueous solutions before.sup.3, however with a distinct nonlinearity which we attribute to remnant particle-particle interactions.

(60) Size dependence of the beads. The Soret coefficient was measured for carboxyl modified polystyrene beads in diameter ranging from 20 nm to 2 μm (Molecular Probes, F-8888, F-8795, F-8823, F-8827). Beads of diameter 0.2 μm, 0.1 μm, 0.04 μm and 0.02 μm were diluted to concentrations of 10 pM, 15 pM, 250 pM and 2 nM, and its bulk fluorescence was imaged over time to derive D.sub.T and D.sup.13,16 from the depletion and subsequent back-diffusion. Larger beads with a diameter of 1.9 μm, 1.1 μm and 0.5 μm were diluted to concentrations of 3.3 aM, 25 aM and 0.2 pM, and measured with single particle tracking.sup.6. The solutions were buffered in 1 mM Tris pH 7.6 with λ.sub.DH=9.6 nm. In all cases interactions between particles can be excluded. Care was taken to keep the temperature gradient in the local equilibrium regime.

(61) We find that the Soret coefficient scales with particle surface over four orders of magnitude (FIG. 10a). The data is described well with equation (2) with an effective surface charge density of σeff=4500 e/μm.sup.2 and neglected hydration entropy contribution. The 5-fold too low prediction for the smallest particle (20 nm diameter) can be explained by charge renormalization since its radius is smaller than λ.sub.DH.

(62) The diffusion coefficient D for spheres is given by the Einstein relation and scales inversely with radius D∝1/a. Inserting equation (2) into S.sub.T=D.sub.T/D, the thermodiffusion coefficient D.sub.T is expected to scale with particle radius a. This is experimentally confirmed over two orders of magnitude (FIG. 10b). These findings of ours contradict several theoretical studies claiming that D.sub.T should be independent of particle size.sup.17-20,27, based on ambiguous experimental results from thermal field flow fractionation (ThFFF).sup.4 which were probably biased by nonlinear thermodiffusion in large thermal gradients.

(63) Size dependence of DNA. Whereas polystyrene beads share a very narrow size distribution as a common feature with DNA molecules, beads are a much less complicated model system. Beads are rigid spheres which interact with the solvent only at its surface. In addition, the charges reside on the surface where the screening takes place. Thus the finding that thermodiffusion of flexible and homogeneously charged DNA is described equally well described with equation (2) is not readily expected and quite interesting (FIG. 10c,d).

(64) We measured DNA sizes with 50 base pairs to 48502 base pairs in 1 mM TRIS buffer (λ.sub.DH=9.6 nm) at low molecule concentrations between 1 (50 bp) and 1 nM (48502 bp). Only every 50th base pair was stained with the TOTO-1 fluorescent dye. The diffusion coefficient was measured by back-diffusion after the laser was turned off and depends on the length L of the DNA in a non-trivial way. The data is well fitted with a hydrodynamic radius scaling a∝L.sup.0.75. This scaling represents an effective average over two DNA length regimes. For DNA molecules longer than approximately 1000 bp, a scaling of 0.6 is found.sup.33 whereas shorter DNA scales with an exponent of ≈1 (see supplementary material).

(65) We can describe the measured Soret coefficient over three orders of magnitude of DNA lengths with equation (2) if we assume that effective charge of the DNA is shielded at the surface of a sphere with the hydrodynamic radius a. Due to the low salt concentration (λ.sub.DH=96 nm), such globular shielding is reasonable. Not only is the experimentally observed scaling of the Soret coefficient with the square root of its length correctly predicted based on equation (2) (S.sub.T∝Q.sub.eff.sup.2/a.sup.2∝L.sup.2/L.sup.1.5∝L.sup.0.5), also the Soret coefficient is fully described in a quantitative manner (FIG. 10c, solid line), with an effective charge of 0.12 e per base, matching well with literature values.sup.34 ranging from 0.05 e/bp to 0.3 e/bp.

(66) As shown in FIG. 10d, the thermodiffusion coefficient for DNA drops with DNA length according to D.sub.T=DS.sub.T∝Q.sub.eff.sup.2/a.sup.3∝L.sup.2/L.sup.225∝L.sup.0.25. Thus, shorter DNA actually drifts faster in a temperature gradient than longer DNA. It is important to point out that this finding is in no contradiction to experimental findings of a constant D.sub.T over polymer length in non-aqueous settings.sup.9. According to equation (1), the thermodynamic relevant parameter is the Soret coefficient, determined by the solvation energetics. The argument.sup.20 that polymers have to decouple into monomers to show a constant D.sub.T merely becomes the special case where the solvation energetics determine both S.sub.T and D with equal but inverted size scaling. In accordance with our local energetic equilibrium argument, S.sub.T and not D.sub.T dominates thermodiffusion also for non-aqueous polymers near a glass transition.sup.35. Here, S.sub.T is constant whereas D.sub.T and D scale according to an increased friction. However for a system of DNA in solution, where long ranging shielding couples the monomers, a constant D.sub.T over polymer length cannot be assumed a priori (FIG. 10d).

(67) Effective charge. The effective charge Q.sub.eff is a highly relevant parameter for colloid science, biology and biotechnology. So far it only could be inferred from electrophoresis, restricted to particles smaller than the Debye length (a≤3λ.sub.DH).sup.36. Unfortunately, many colloids are outside this regime. As shown before, a similar size restriction does not hold for thermodiffusion. In many cases, the hydration entropy S.sub.hyd contributes less than 15% (FIG. 8) and can be neglected at moderate salt levels. Thus we can invert equation (2) to obtain the effective charge Q.sub.eff for spherical molecules from

(68) $\begin{array}{cc}{Q}_{\mathrm{eff}}=\frac{2T^2}{3\eta D}\sqrt{\frac{{\mathrm{.Math..Math.}}_0{k}^3{S}_T}{{\mathrm{\beta \pi \lambda }}_{\mathrm{DH}}}}& \left(3\right)\end{array}$

(69) The effective charge derived from thermodiffusion measurements of polystyrene beads and DNA is plotted in FIG. 11 over several orders of magnitude in size. The effective charge of beads scales linearly with particle surface with a slope confirming the effective surface charge density of σ.sub.eff=4500 e/μm.sup.2 which was inferred from electrophoresis only for small particles. Average deviations from linear scaling are below 8% (FIG. 11a). The effective charge inferred from thermodiffusion measurements of DNA using equation (3) scales linearly with DNA length with an effective charge of 0.12 e per base pair. The length scaling is confirmed over four orders of magnitude with an average error of 12% (FIG. 11b). Thus thermodiffusion can be used to infer the effective charge with low errors for a wide range of particle sizes. This is even more interesting for biomolecule characterization since measurements of thermodiffusion can be performed all-optically in picoliter volumes.

(70) Conclusion. We describe thermodiffusion, the molecule drift along temperature gradients, in liquids with a general, microscopic theory. Applied to aqueous solutions, this theory predicts thermodiffusion of DNA and polystyrene beads with an average accuracy of 20%. We experimentally validate major parameter dependencies of the theory: linearity against screening length μ.sub.DH and molecule hydrodynamic area A, quadratic dependence on effective charge and linearity against temperature. Measurements of thermodiffusion can be miniaturized to micron scale with the used all-optical fluorescence technique and permits microscopic temperature differences to manipulate molecules based on their surface properties (FIG. 7). The theoretical description allows to extract the solvation entropy and the effective charge of molecules and particles over a wide size range.

(71) Infrared temperature control. The temperature gradients used to induce thermodiffusive motions were created by aqueous absorption of an infrared laser at 1480 nm wavelength and 25 mW power (Furukawa). Water strongly absorbs at this wavelength with an attenuation length of κ=320 μm. The laser beam was moderately focussed with a lens of 8 mm focal distance. Typically, the temperature in the solution was raised by 1-2 K in the beam center with a 1/e.sup.2 diameter of 25 μm, measured with the temperature dependent fluorescence signal of the dye BCECF.sup.13. Thin chamber heights of 10 μm to 20 μm and moderate focussing removed possible artifacts from optical trapping, thermal lensing and thermal convection.sup.13. For temperature dependent measurements both the objective and the microfluidic chip were tempered with a thermal bath. Imaging was provided from an AxioTech Vario fluorescence microscope (Zeiss), illuminated with a high power LED (Luxeon) and recorded with the CCD Camera SensiCam QE (PCO).

(72) Molecules. Highly monodisperse and protein-free DNA of 50 bp, 100 bp, 1000 bp, 4000 bp, 10000 bp and 48502 bp (Fast Ruler, fragments and λ-DNA, Fermentas) were diluted to 50 μM base pair concentration, i.e. the molecule concentration was between 1 μM (50 bp) and 1 nM (48502 bp). DNA was fluorescently labeled by the intercalating TOTO-1 fluorescent dye (Molecular Probes, Oreg.) with a low dye/base-pair ratio of 1/50. Carboxyl modified polystyrene beads of diameter 2 μm, 1 μm, 0.5 μm, 0.2 μm, 0.1 μm, 0.04 μm and 0.02 μm (F-8888, F-8823, F-8827, F-8888, F-8795, F-8823, F-8827, Molecular Probes) were dialyzed (Eluta Tube mini, Fermentas) in aq. dest. and diluted in 1 mM Tris pH 7.6 to concentrations between 3.3 aM (2 μm) and 2 nM (0.02 μm).

(73) Concentration imaging over time. Either the method of concentration imaging.sup.13 or single particle tracking were used to measure thermodiffusion at low concentrations, namely below 0.03 g/l for DNA and 10.sup.−5 g/l for beads. At higher concentrations, we found profound changes of thermodiffusion coefficients. DNA and polystyrene beads smaller than 0.5 μm diameter concentration were imaged over time.sup.13 by bright field fluorescence with a 40× oil immersion objective. Concentrations inferred after correcting for bleaching, inhomogeneous illumination and temperature dependent fluorescence.sup.13 were fitted with a finite element theory. The model captures all details of both thermodiffusive depletion and backdiffusion to measure D.sub.T and D independently (see supplementary material). Measurements were performed in microfluidic chips 10 μm in height with PDMS on both sides.sup.13.

(74) Single particle tracking. Polystyrene particles larger than 0.5 μm in diameter were measured by single particle tracking due to the slow equilibration time and risk that steady state depletion is disturbed by thermal convection. The thermodiffusive drift was imaged with a 32× air objective at 4 Hz at an initial stage of depletion in a 20 μm thick chamber. Averaging over the z-position of the particles removed effects from thermal convection. The drift velocity versus temperature gradient of 400 tracks were linearly fitted by v=−D.sub.T∇T to infer D.sub.T. The diffusion coefficients D of the particles were evaluated based on their squared displacement, matching within 10% the Einstein relationship.

(75) The present invention is not limited by what has been particularly shown and described hereinabove. Rather the scope of the present invention includes both combinations and sub-combinations of the features described hereinabove as well as modifications and variations thereof which would occur to a person skilled in the art upon reading the foregoing description and which are not in the prior art.

(76) Infrared heating. The temperature gradients used to induce thermophoretic motions are created by aqueous absorption of a fiber coupled infrared solid state laser (Furukawa FOL1405-RTV-317), with a wavelength of 1480 nm and a maximum power of 320 mW typically used at 25 mW. Water strongly absorbs at this wavelength with an attenuation length of κ=320 μm. The infrared light is coupled out of the fiber to form a parallel beam with an 1/e.sup.2 diameter of 1 mm. The beam position in the x/y plane can be adjusted by two galvanoometrically controlled infrared mirrors (Cambridge Technology 6200-XY Scanner with Driver 67120). The laser beam is focussed from below the object stage by an infrared corrected aspheric lens with 8 mm focal distance (Thorlabs, C240TM-C). Typically, temperature was increased by only 2 K in the heated focus.

(77) Temperature measurement. The temperature gradient was measured via the temperature dependent fluorescence signal of the dye BCECF, diluted to 50 μM in 10 mM TRIS buffer. Details of bleaching correction and temperature extraction were described previously.sup.13. From the total temperature dependence of BCECF of −2.8%/K, only −1.3%/K stems from pH drift of the used TRIS buffer. The remaining −1.5%/K are the result of thermodiffusion of the dye itself, measured to be S.sub.T6=0.015/K with the concentration over time method described below.

(78) “DNA” image. Measurement of the “DNA” profile in FIG. 7 was performed in a 60 μm thick chamber between glass slides, imaged with a 10× objective and heated to 2 K along the letters “DNA” with laser scanning. The chamber was filled with a 50 nM solution of 1000 bp DNA stained with the intercalating fluorescent dye TOTO-1 (Molecular Probes) To switch from depletion to accumulation, the experiment was performed at room temperature or with the chamber cooled to 3° C., respectively.

(79) Fluorescence approaches. Historically, methods used to measure thermodiffusion in liquids are based on changes in refractive index upon change in solute concentration.sup.27. Inherently, this signal is small for low solute concentrations near the limit of non-interacting molecules, even for intricate detection methods like thermal lensing or holographic interference.sup.38. Although operating at much smaller volume, the used fluorescent microfluidic approach.sup.13 allows concentrations of 0.03 g/l for DNA and reaches 10.sup.−5 g/l for single particle tracking. This is necessary, since for example in thermodiffusion of DNA, we see profound changes at higher concentrations.

(80) Thermodiffusion from concentration over time. Both DNA and polystyrene beads smaller than 0.5 μm in diameter were measured by imaging molecule concentration over time by bright field fluorescence. A more basic steady state method was described previously.sup.3. Here, we refined it with a numerics theory to infer diffusion coefficient D and Soret coefficient S.sub.T independently.

(81) Highly monodisperse and protein-free DNA of 50 bp, 100 bp, 1000 bp, 4000 bp, 10000 bp and 48502 bp (Fast Ruler, fragments and λ-DNA, Fermentas) were used for the length dependent measurements. The DNA was fluorescently labeled by the intercalating TOTO-1 fluorescent dye (Molecular Probes, Oreg.) which shows 1000× fluorescence increase when bound to DNA. The dye/base-pair ratio was low (1/50) to avoid structural or charge artifacts from the bound dye. The fluorescence was observed with an 40× oil objective. DNA stock solutions were diluted to 50 μM base pair concentration, which corresponds to molecule concentrations between 1 μM (50 bp) and 1 nM (48502 bp), respectively. Polystyrene beads of diameter 0.2 μm, 0.1 μm, 0.04 μm and 0.02 μm (Molecular Probes, Oreg., F-8888, F-8795, F-8823, F-8827) were dialyzed (Fermentas, Eluta Tube mini) in aq. dest., and diluted in 1 mM Tris pH 7.6 to concentrations of 10 pM, 15 pM, 250 pM and 2 nM, respectively. DNA thermodiffusion measurements were performed in microfluidic chips 10 μm in height with PDMS on both sides.sup.13. They allow the measurement of small volumes in thin defined geometries. Polystyrene beads were sandwiched between a 1.25 mm thick polystyrene slide (Petri Dish, Roth) on the bottom and a plastic slide (170 μm thick, 1 cm×1 cm, Ibidi, Munich) on the top and sealed with nail polish. For temperature dependent measurements, both 40× oil objective and microfluidic chip were tempered from below with a thermal bath. Note that temperatures well below 0° C. can be achieved as the microfluidic geometry reduces the probability of water to freeze.

(82) Concentration of DNA was inferred from fluorescence images, that were measured with a 40× oil objective (NA=1.3) on an AxioTech Vario fluorescence microscope (Zeiss), illuminated with a high power LED (Luxeon) and imaged with the CCD Camera SensiCam QE (PCO). The time series allows to correct for inhomogenous illumination and bleaching.sup.13. A time series dependent photobleaching means that an individual bleaching factor is determined for every image. This correction is of advantage for high precision measurements for protein thermophoresis. If single molecules were visible in fluorescence, time averaging was used to average over particle positions.

(83) Radial profiles were taken over time and combined to a space-time image of both the thermophoretic depletion and back diffusion (FIG. 12a,c). Fluorescence was adjusted for the temperature dependence of the TOTO-1 dye determined independently with a fluorescence spectrometer as −0.5%/K. Typically, temperature in the solution was raised by 1-2 K in the beam center with a 1/e.sup.2 diameter of 25 μm.

(84) The Soret coefficient S.sub.T can be obtained from the steady state profile. Given the temperature at radius r obtained from temperature dependent fluorescence, the concentration c(r) can be fitted with the steady state thermophoretic profile.sup.13 given by
c(r)=c.sub.0e.sup.−S.sup.T.sup.(T(r)−T.sup.0.sup.)  (4)

(85) with chamber temperature T.sub.0 and bulk concentration c.sub.0.

(86) We can also obtain D and D.sub.T independently by analyzing the build up and flattening of concentration profile over time after turning the infrared laser beam on or off, respectively. Theory was provided from finite element model in radial coordinates (FEMLab, Comsol) over time with boundary conditions of concentration obtained from the experiment. Comparison with experiment of the time course of thermophoretic depletion reveals D.sub.T (FIG. 12c,d) and from the time course after switching off the heating source the diffusion coefficient D is obtained (FIG. 12a,b). Results for diffusion coefficients obtained for DNA molecules are shown in FIG. 13. Scaling of D for DNA larger than 1000 bp agree well with literature values and theoretical expectations.sup.33. However for DNA molecules in the order of the persistence length (about 150 bp) the power law exponent of −0.6 does not precisely fit the measured values and a different scaling with an exponent of −1 is necessary. A good description of DNA diffusion coefficients in the size range analyzed throughout this work is achieved with an intermediate exponent of −0.75.

(87) Screening length. Debye-Hückel length was titrated by adding c.sub.S=0 mM, 2 mM and 20 mM of KCl to c.sub.T=1 mM of TRIS buffer at pH 7.6 and calculated from

(88) $\begin{array}{cc}{\lambda }_{\mathrm{DH}}=\sqrt{\frac{{\mathrm{.Math..Math.}}_0\mathrm{kT}}{2e^2\left(c_S+c_T\right)}}& \left(5\right)\end{array}$

(89) Changes in the effective charge of the molecules can be excluded at these monovalent salt concentrations. For largest values λ.sub.DH=13.6 nm, solely 0.5 mM Tris-HCl pH 7.6 buffer was used.

(90) Thermodiffusion using single particle tracking. For fluorescent polystyrene particles of large size (2 μm, 1 μm, 0.5 μm, Molecular probes, Oreg., F-8888, F-8823 and F-8827) a different method had to be used due to the increasing visibility of single particles, slow equilibration time and the risk that steady state depletion is disturbed by thermal convection. Beads were dialyzed (Fermentas, Eluta Tube mini) in aq. dest., and diluted in 1 mM Tris pH 7.6 to concentrations of 3.3 aM, 25 aM and 0.2 pM, respectively. Thermodiffusion was measured in 20 μm thin chambers. A 1.25 mm thick polystyrene slide (Petri Dish, Roth, Karlsruhe) was chosen as for the bottom of the chamber, while a plastic slide (170 μm thick, 1 cm×1 cm, Ibidi, Munich) was taken as cover slip. The low thermal conductivity ensures constant temperature across the chamber. Chamber walls were made hydrophilic in a plasma cleaner (Harrick) for 10 min at 10 W electrical power. As a result, adsorption of polystyrene particles to the plastic is low even at high salt concentrations. Addition of 2 μl bead solution between the plastic sheets, followed by sealing the chamber with fast drying nail polish leads to reproducible chamber heights of 20 μm.

(91) Imaging was provided with an AxioTech Vario fluorescence microscope (Zeiss), illuminated with a high power LED (Luxeon) and imaged with the CCD Camera SensiCamQE (PCO). The center of the plain of view was heated 8 K above room temperature with maximal temperature gradient of 0.2 K/μm and 1/e.sup.2 spot radius of 50 μm Temperature was imaged with BCECF as described before in a separate chamber. In FIG. 14a finite element simulation of the experimental situation is shown using the commercial available Femlab software (Comsol). A 20 μm chamber is heated to 8 K in the center. Due to low heat conductivity of the PS walls the temperature profile is homogeneous throughout the chamber height (FIG. 14a). Due to the thin chamber the convection speed is suppressed to negligible speeds of 5 nm/s at maximum (FIG. 14b). The thermophoretic movement of the particles was imaged with a 32× air objective and recorded at 4 Hz. Particles all over the 20 μm chamber could be equally tracked with a custom-written Lab View program. Artefacts from toroidal thermal convection are averaged out to a high degree as convective attraction near the lower chamber wall cancelled with opposite convective repulsion near the top of the chamber. Typically the velocity of 400 tracks was plotted against radius and fitted with the drift velocity expected from thermodiffusion according to thermophoretic drift v=−D.sub.T∇T to find the thermophoretic mobility D.sub.T. Thermal fluctuations of the tracks were evaluated based on their squared displacement to obtain the diffusion coefficient D of the particles, which matched within 10% the Einstein relationship D=kT/(6πηa). This is expected since even for the worst case, the chamber is 20-fold thicker than the diameter of the measured particles.

(92) Electrophoresis. The effective surface charge density was measured for 40 nm diameter beads by electrophoretic drift in 400 μm thin and 5 cm long chambers (Ibidi, Germany). The velocity profile throughout the chamber height at 5 V was taken from single particle tracking of 2 μm beads. At 80 μm height the electroosmotic flow in a tightly sealed chamber is zero.sup.39. As expected the velocity of particles in this plane saturates for particles larger than 100 nm and is not related to effective charge.sup.36. A high numerical aperture oil objective has been used to analyze the velocity of 40 nm particles at the chamber surface at the same conditions. The constant velocity difference between chamber surface and plane of zero electroosmotic flow measured before has been used to calculate the purely electrophoretic velocity.

(93) Additional References Referred to Herein Above in Example 1: 3. S. J. Jeon, M. E. Schimpf and A. Nyborg, Anal. Chem. 69, 3442-3450 (1997) 4. P. M. Shiundu, G. Liu, and J. C. Giddings, Anal. Chem. 67, 2705-2713 (1995)6 7. B.-J. de Gans, R. Kita, B. Müller and S. Wiegand, J. Chem. Phys. 118, 8073 (2003). 8. J. Rauch and W. Köhler, Phys. Rev. Lett. 88, 185901 (2002) 9. S. Wiegand and W. Köhler, in Thermal Nonequilibrium Phenomena in Fluid Mixtures, Springer, Berlin, 189 (2002) 13. D. Braun and A. Libchaber, Physical Review Letters 89, 188103 (2002) 14. S. R. de Groot, P. Mazur, Non Equilibrium Thermodynamics (North-Holland, Amsterdam, 1969) 16. A. H. Jr. Emery and H. G. Drickhammer, J. Chem. Phys. 23, 2252 (1955) 17. J. S. Ham, J. Applied Physics, 31, 1853 (1960) 18. K. I. Morozov, J. Experim. and Theor. Phys. 88, 944 (1999) 19. M. E. Schimpf and S. N. Semenov, J. Phys. Chem. B 104, 9935 (2000) 20. A. Voit, A. Krekhov, W. Enge, L. Kramer, and W. Kühler, Phys. Rev. Lett. 92, 214501 (2005) 22. S. Fayolle, T. Bickel, S. Le Boiteux and A. WUrger, Phys. Rev. Lett. 95, 208301 (2005) 24. P. N. Snowdon, J. C. R. Turner, Trans. Faraday Soc. 56, 1409 (1960) 25. E. Ruckenstein, J. Colloid Interface Sci. 83, 77 (1981) 27. J. Israelachvili, Intermolecular & Surface Forces, 2nd edition, Academic Press, 1992 28. W. Lin, P. Galletto and M. Borkovec, Langmuir 20, 7465-7473 (2004) 29. D. Haidacher, A. Vailaya and C. Horvath, Proc. Natl. Acad. Sci. 93, 2290-2295 (1996) 30. N. T. Southall, K. A. Dill and A. D. J. Haymet, J. Phys. Chem B 106, 521-533 (2002) 31. B. Kronberg, M. Costas and R. Silveston, Pure & Applied Chemistry 67, 897-902 (1995)

Example 2

Determination of Hydrodynamic Radius and Interaction Between Proteins

(94) The thermo-optical characterization method of the present invention allows also to quantify the hydrodynamic radius of proteins and even more important of complexes of biomolecules which are not connected covalently to each other. Thermophoresis provides a comparably robust and precise way to measure the hydrodynamic radius of molecules from less than a nanometer up to a few microns. In comparison to the other thermo-optical characterization methods the precision of this method is not too sensitive on the geometry of measurement (e.g. height of the liquid layer) as it is the case for molecular interactions.

(95) Data acquisition: A typical measurement can be described as follows:

(96) Step 1:

(97) A solution of fluorescently labeled molecules is brought into a microfluidic measurement chamber (e.g. capillary, microfluidic chip). Fluorescence is excited and recorded with spatial resolution for less than 5 seconds on a CCD device with a frame rate between 40 Hz down to 0.2 Hz (i.e. for fast diffusing molecules, a high frame rate is chosen). These image(s) provides the necessary information about the fluorescence level at 100% concentration. Then fluorescence excitation is turned off.

(98) Step 2:

(99) The Infrared laser heating is turned on. The immediately established local spatial temperature distribution causes the molecule drift to lower or higher temperatures, depending on the particular molecule to be analyzed. The laser is focused in a way that temperature gradients between 0.0 and 5 K/μm are achieved. The temperature gradient has been calibrated once and it is not necessary to repeat this calibration every time an experiment is performed. The maximal temperature elevation is below the temperature which is known to cause damage to the molecules or disintegrate their interaction. Depending on the thermophoretic properties of the molecules in the solution (i.e. if they move fast in a thermal gradient or slow) the infrared laser heats the solution for 5 seconds up to 100 seconds. After this period of time the infrared laser is turned off

(100) Step 3:

(101) After the spatial temperature distribution has vanished (typically 2-50 ms) the fluorescence excitation is turned on and fluorescence is recorded with the same frame rate used in the first step of fluorescence imaging. This time the redistribution of the molecules is imaged for 5 seconds up to 50 seconds. The exact time depends on the velocity with which the molecules diffuse (i.e. the time it takes them to equalize 90% of the concentration gradient established by thermophoresis).

(102) Data processing—Photobleaching: The fluorescence images have to be corrected for photobleaching. Since there is no spatial temperature profile in the solution while fluorescence images are taken, the bleaching correction is possible with high precision (i.e. high precision is possible since the rate of photobleaching is temperature dependent).

(103) Therefore, the fluorescence at a edge of the measurement chamber (i.e. a spot as far away from the heated center as possible), were thermophoresis during step 3 was negligible (i.e. for a person skilled in the art, this is where the temperature gradient during laser heating was lower than 0.001 K/μm), is evaluated to determine the photobleaching from the image series taken in step 3. If photobleaching is present, the fluorescence will decrease from image to image. The individual factor for each image is used to correct all images for bleaching. Another possibility is to calculate the bleaching for every single pixel from the images taken in Step 1. The bleaching rate per pixel can be used to correct every pixel from step 3 images for the photobleaching effect.

(104) Data processing—Inhomogeneous Illumination correction and normalization to 100% concentration: All images taken in step 4 are divided by a single or all images taken in step 2 and multiplied by 100. This way a correction for inhomogeneous illumination is achieved and the fluorescence is normalized to 100% concentration.

(105) Data processing—Determining the hydrodynamic radius: From the first images of the step 4 image series the concentration distribution is extracted. A software tool evaluates the Diffusion coefficient (or multiple Diffusion coefficients in case of a mixture) which describes the experimentally measured relaxation of the concentration gradient. Using the Stokes-Einstein relation the hydrodynamic radius is inferred from the diffusion coefficient.

(106) In particular, the above described experiment was conducted for a sample of GFP as follows: The thermo-optical properties of two samples of Green Fluorescent protein (GFP) are measured with the devices of the present invention. 2 μl of GFP (5 μM, 1×PBS buffer) is pipetted on an object slide. The sample is sandwiched by pitting an cover slip (round 12 mm diameter) on top. The liquid spreads uniformly in between the glass surfaces an the chamber is sealed off by using nail polish. This prevents the liquid from rapid evaporation, which would in turn lead to a comparably strong flow of liquid in the chamber. This sample is palaced on a device shown in FIG. 1a. The measurement steps and data processing steps are performed as described above. The same experiment is performed with a second sample containing 5 μM GFP and 10 μM of a GFP binding antibody fragment, specifically binding to GFP. In both cases, first the fluorescence is recorded without laser heating. Then the fluorescence excitation is turned off and the IR-laser radiation is turned on (the maximum temperature is kept below 35° C. (i.e. about 15° C. above ambient temperature (about 20° C.) to avoid denaturation or damage to the protein). The laser is turned off after a few seconds of heating and the fluorescence excitation is turned on at the same time. The relaxation of the spatial fluorescence distribution (i.e. concentration distribution) to a homogeneous state is recorded for a few seconds. As can be observed from FIG. 31, in the sample with the two interacting species (i.e. GFP and the antibody fragment) the fluorescence profile needs more time to relax. This is explained by slower diffusion of the larger complex. The time evolution of the fluorescence profile is analyzed via a software tool (selfmade Software, Labview, National Instruments) to determine the diffusion constant. By using the Stokes-Einstein relation, a hydrodynamic radius is attributed to the diffusion constant. In case of the free GFP this is 5 nm and the complex has an radius of 10 nm.

Example 3

Detection of Interactions Between Biomolecules and Discrimination of Nucleic Acids by Size

(107) The thermo-optical characterization of the present invention provides the means for fast all optical biomolecule analysis. Present methods for detection and quantification of biomolecular interactions are very time consuming which means that the time needed for an analysis is on the order of 30 minutes up to hours. The present invention can detect and quantify biomolecular interactions within 1 second up to 50 seconds. The term interaction comprises interaction between biomolecules (e.g. protein, DNA, RNA, hyaluronic acids etc) but also between modified nanoparticles/micro beads and biomolecules. A typical experiment to detect/quantify interactions can be described as follows:

(108) Step 1a, Background Measurement:

(109) The sample buffer without fluorescently labelled sample molecules/particles is filled in the microfluidic chamber and the fluorescence is measured, while the excitation light source is turned on.

(110) Step 1b, Determination of Fluorescence Level Before Laser Heating:

(111) An aqueous solution of a fluorescently labelled sample (e.g. biomolecules, nanoparticles, microbeads whereas all of them have a specific affinity for other biomolecules) at a given concentration is filled in a microfluidic chamber (preferably a capillary) which preferably guarantees a defined height of the chamber. Fluorescence is excited and recorded with (CCD-Camera) or without (Photomultiplier tube, Avalanche Photodiode) spatial resolution for less than 10 seconds on a CCD device or photomultiplier with exposure times of 25 milliseconds up to 0.5 seconds. Then the fluorescence excitation is turned off.

(112) Step 2, Starting of Infrared Laser Heating:

(113) In the following the infrared heating laser is turned on and the spatial temperature distribution is established within a few milliseconds within the solution. The temperature gradient has been calibrated once and it is not necessary to repeat this calibration every time an experiment is performed. In particular a setup were infrared heating and fluorescence imaging are performed through the same optical element from one side is advantageous for the stability of the optical and infrared foci.

(114) It is of advantage that in the experiment the decrease of fluorescence due to photobleaching is lower than 5%.

(115) The maximal temperature elevation is below the temperature which is known to cause damage to the molecules in the solution or disturb their interaction (e.g. temperatures between 1 and 5° C. above ambient temperature).

(116) Depending on the thermophoretic properties of the molecules in the solution (i.e. if they move fast in a thermal gradient or slow) the infrared laser heats the solution for 5 seconds up to 100 seconds.

(117) Step 3, Recording of the Spatial Fluorescence (i.e. Concentration) Profile:

(118) After this period of time the fluorescence excitation is turned on and images are recorded with the same frame rate and length as described in step 1b. Step 3 is the last acquisition step necessary for evaluation of thermo-optical properties.

(119) For detection and quantification of interactions more measurements following the protocol described previously are necessary. Step 1a is repeated with sample buffer and in step 1b the aqueous solution of a fluorescently labelled sample is mixed with an amount of the biomolecule with which the interaction should be detected or quantified. For the detection of an interaction it is necessary to mix the fluorescently labelled sample with a sufficient amount of binding partner so that a substantial amount of the fluorescently labelled molecule is in the complex with the binding partner. If the strength of the interaction should be quantified in terms of a dissociation or association constant (Ka, Kd), than the procedure described previously has to be conducted with varying concentrations of binding partner (e.g. 0.1××10× the concentration of the fluorescently labelled binding partner). This means that a titration of binding partner should be performed.

(120) Processing the raw data: For a linear bleaching correction it is necessary to wait for the back-diffusion of all molecules following the end of step 3. This increases the time consumption of the analysis dramatically. For precise and fast measurements it is of advantage to determine the strength of bleaching from image to image and correct every individual image with its own bleaching factor. For a precise bleaching correction it is important that the temperature gradient at distance from the heat spot is low (e.g. below 0.001 K/μm). The images taken in step 1b are used to correct all images for inhomogeneous illumination. In case fluorescence is recorded without spatial resolution (e.g. avalanche photodiode or photomultiplier) photobleaching is corrected best by determining once the bleaching characteristic of a certain dye without heating laser in a control experiment.

(121) Data evaluation: Qualitative detection of interaction: From the image series the spatial fluorescence distribution of the reference experiment (i.e. fluorescently labelled molecule/particle without binding partner) and the second experiment (i.e. were the binding partner is present) is extracted. The fluorescence is plotted versus the distance from the heat spot. An averaging is only possible for pixels with the same temperature and same distance. The spatial concentration distribution is obtained by correcting the fluorescence intensities for the respective temperature dependence of the dye. With knowledge of temperature dependence of the fluorescent dye and the spatial temperature distribution, the effect of a decreasing fluorescence due to temperature increase can be corrected. For the qualitative detection of interaction as well as their quantification a correction for temperature dependency is not necessary, and the spatial fluorescence distribution is sufficient. This enables us to use any fluorescent dye on the market without characterization of its temperature dependency.

(122) The values of the fluorescence profile are integrated up to the distance were the temperature is below 10% of the maximum temperature (e.g. 70 μm). The integrated values are compared and a change give a precise indication if there is an affinity between the substances at the concentrations used, since the interaction changes the thermo-optic properties (e.g. thermophoretic mobility, surface size and chemical groups on surface). In most cases the interaction leads to higher fluorescence (concentration) at higher temperatures.

(123) In case the whole cross-section of a capillary is heated (i.e. using cylindrical lenses to give the IR laser beam a ellipsoidal shape, which heats a cross section of a capillary homogeneously), the intensity of a lot more pixels from the centred heat spot can be averaged since all pixels at same distance to the heated line have the same temperature. This is advantageous for high precision measurements. In case fluorescence is recorded without spatial resolution the fluorescence change in the centre of the heat spot/line is measured, whereas it is again advantageous to heat the whole cross section. In general if more than a single frame is recorded in step 1b and 3 an integration of multiple frames is possible.

(124) For a quantification of affinities the same procedure is performed for all experiments at various concentrations of non fluorescent binding partner. The result of the integration for the reference experiment (i.e. without binding partner) is subtracted from the integrated values obtained for the different concentrations of binding partners. From this evaluation on gets the amount of interacting complexes in arbitrary units. By dividing the values by the value were binding is saturated the relative amount of complexes at a certain concentration of binder is obtained. From these dataset also the concentration of free non fluorescent binding partner can be determined and the strength of the interaction can be quantified in terms of association or dissociation constant (see FIG. 25).

Example 3a

Interaction of Proteins

(125) FIG. 25 shows a quantification of interaction between biomolecules. 100 nM of a fluorescently labelled antibody in 1×PBS buffer (anti-Interleukin 4, Sigma-Aldrich) are titrated with various amounts interleukin 4 1×PBS buffer(0-300 nM IL4 Sigma-Aldrich). (left). Approximately 200 nl of the sample mix are soaked into a capillary of 40 nm inner diameter (World Precision Instruments). The capillary is situated on a device as shown in FIG. 27. Fluid drift is prevented by closing the valves on both sides of the capillary. The measurement is performed with the device shown in FIG. 20. The spatial fluorescence distribution in steady state is measured using the previously described protocol. Three results for 5 nM, 80 nM and 300 nM are shown exemplarily. After each measurement the capillary is flushed with approximately 5 μl of 1×PBS buffer. The binding of IL 4 to the antibody changes the signal dramatically from fluorescence decrease to a fluorescence increase. Integration of the fluorescence profile up to 80 μm (distance from the heated centre, see procedure described in this example herein above) allows to determine the number of complexes in solution. (right) The concentration of free Interleukin 4 can be calculated plotted versus the concentration formed complex. These data can be fitted to determine the K.sub.D.

Example 3b

Discrimination of DNA by Size and Interaction of DNA Strands

(126) Qualitative detection of interaction, using a modified protocol according to this example 3, is shown in FIG. 5. Here the concentration profiles of 20 base pair DNA hast been compared to 50 base pair DNA at a maximum temperature increase of 10° C. (at ambient temperature of 20° C.). In a second experiment 20 bases single stranded DNA is compared with 20 base pair double stranded DNA. All four experiments were performed in 1×SSC buffer the following way: 2 μl of sample are pipetted on an object slide (Roth, 1 mm thick) an sandwiched between an cover slip of 12 mm diameter. The aqueous solution spreads uniformly in between the two glass surfaces yielding a thin sheet of water with a height of approximately 20 μm. The liquid sheet is sealed by using nail polish. This avoids rapid evaporation of the sample. The microfluidic chamber is placed on the object stage of the thermo-optical setup (e.g. FIG. 1a) and imaged via a 40× oil objective (NA 1.3, Zeiss). The laser focus is positioned so that it is approximately in the center of the field of view and has a half width of about 20 μm. Only a single pixel of a CCD camera used for detection of the fluorescence. This pixel measures the fluorescence in the center of the heat spot. The fluorescence is recorded for approx. 1 seconds without laser heating, then the IR Laser is turned on while fluorescence is still recorded. After 20 seconds of laser heating the measurement stopped. As can be seen from FIG. 5 single stranded DNA can be discriminated from double stranded DNA and DNA of different length can be discriminated within a time span of a few seconds. Bleach correction is not performed in this measurement strong change in concentration is observed.

Example 4

Detection of Binding of PEG Molecules to Nanoparticles

(127) As mentioned previously it is also possible to detect the binding of molecules to larger inorganic particles or nanocrystals using the procedure described previously. Inorganic CdSe particles (core diameter about 12 nm) have been modified with a varying numbers (1 up to 3) of Poly-ethylen-glycol (PEG) of different molecular weight. 2 μl of sample are pipetted on an object slide (Roth, 1 mm thick) and sandwiched between a cover slip of 12 mm diameter. The aqueous solution spreads uniformly in between the two glass surfaces yielding a thin sheet of water with a height of approximately 20 μm. The liquid sheet is sealed by using nail polish. This avoids rapid evaporation of the sample. The microfluidic chamber is placed on the object stage of the thermo-optical setup (e.g. FIG. 1a) and imaged via a 40× oil objective (NA 1.3, Zeiss). The laser focus is positioned so that it is approximately in the center of the field of view and has a half width of about 100 μm. The maximum temperature increase was determined to 5° C. above ambient temperature. The spatial fluorescence profile is recorded as described previously for the detection of biomolecular interactions. Also the raw data are processed as described previously. To measure the number or size of PEG molecules bound to the nanocrystals it is sufficient to compare the spatial fluorescence profiles obtained with the protocol described previously. However, a correction for the temperature dependent decrease of the fluorescence allows a quantification in terms of the Soret coefficient. FIG. 26 shows that the Soret coefficient increases linearly with the number of PEG molecules bound to the nanocrystals. The slope of the increase depends on the molecular weight of the PEG. FIG. 26 shows that the binding of single molecules of the size of a protein is detectable.

Example 5

Thermophoresis of Proteins

(128) An example how the conformation, structure and surface of a molecule effect the thermo-optical characteristic of said molecules and how these characteristics may be measured, detected or characterized is given in the following:

(129) A sample of fluorescently labelled BSA (Bovine Serum Albumin, Fermentas) is transferred in a microfluidic chamber (e.g. capillary). The temperature of the whole sample volume is adjusted by a Peltier element in thermal contact to the solution (the microfluidic chamber is placed on a device shown in FIG. 27). The Peltier element is used to regulate the “ambient temperature” of the solution. It does not create a spatial temperature distribution. The adjustment of temperature is important because the thermo-optic properties (e.g. surface properties, conformation) at varying ambient temperatures (i.e. protein conformations) should be measured. The thermo-optic properties are measured following the protocol described previously for biomolecule interactions (step 1 to step 3, without addition of binding partners since only intramolecular properties are measured). Also the processing of the raw data follows the procedure described for the detection of molecule interactions. The thermo-optical properties are evaluated by determining the concentration profile in steady state. The thermo-optic properties are plotted as Soret coefficient ST as shown in FIG. 28. The Soret Coefficient is obtained by correlating the concentration profile in an exponential fashion (c=c.sub.0e.sup.−S.sup.T.sup.(T-T.sup.0.sup.)) to the temperature distribution. The Soret-Coefficient is sensitive to changes in the interplay between the amino acids of the protein and the water molecules. At low temperatures the molecules are accumulated in a region of elevated temperature, which corresponds to a negative Soret coefficient. At increasing temperatures the accumulation of the molecules changes (i.e. the Soret coefficient increases). This can be readily explained by changes in the conformation of the molecule (e.g. hydrophobic groups or loops rearrange). As can be seen from FIG. 28 the sign of thermophoresis changes from negative values at lower temperatures to positive values (i.e. depletion) at higher temperatures. The sudden jump to positive Soret coefficients correlates very well with the temperature were thermal denaturation occurs (i.e. 50° C.). In the range of body temperature (i.e.30° C.-40° C.) the thermo-optical signal does not significantly change. An explanation for this unexpected behaviour is that the protein is evolutionary designed to be functional within this temperature range. Since there is a tight structure-function relationship in nature, the structure is preserved in this temperature range. The values shown in FIG. 28 are corrected for the temperature dependence of the fluorescence dye.

(130) The local temperature increase in the system causes a change in fluorescence which is not purely caused by changes in concentration due to thermophoresis. Since the temperature of transition from negative to positive values is important for measurements of absolute protein stability, a correction for the temperature dependence of fluorescence is advantageous. In application were differences in stability should be detected (e.g. small molecule binding to a protein) the correction for temperature dependence of the fluorescence is not necessary. The argument that protein structure/conformation is measured is supported by FIG. 28b where the experiment is started at high temperatures. Even at temperatures below the thermal denaturation temperature the Soret coefficient is positive. This is based on the slow refolding time compared to the speed of measurement, which is faster than 50 seconds. After a certain time span the thermo-optical ST values shown in FIG. 28a are also obtained again in the measurement shown in FIG. 28b. As a control the temperature of the system is increased again and the negative ST values are obtained as expected at temperatures of about 40° C.

Example 6

Detection of Conformational Changes, Like Denaturation of Proteins

(131) a) An example where the denatured form of a protein is distinguished from the native form without any correction for the temperature dependence of the fluorescence dye, given in FIG. 15. An aliquot of a fluorescently labelled Bovine Serum Albumin is heated up to 90° C. for 10 minutes, which is well above the temperature of denaturation and the protein cannot refold. Also the thermo-optic characteristics of a native aliquot (i.e. not heated to 90° C.) is measured. Starting with the native sample, approximately 200 nl of the sample are soaked into a capillary of 40 nm inner diameter (World Precision Instruments). The capillary is situated on a device as shown in FIG. 27. Fluid drift is prevented by closing the valves on both sides of the capillary. The measurement is performed with the device shown in FIG. 20. The spatial fluorescence distribution in steady state is measured using the previously described protocol. After each sample the capillary is flushed with approximately 5 μl of 1×PBS buffer. The experiment is performed as described for the measurement of biomolecular interactions (Step 1-Step 4). The experiment is performed twice with different infrared laser powers (i.e. maximal temperatures of 5° C. and 10° C. above ambient temperature are employed). Afterwards the sample with the denatured protein is measured. Again three experiments with different laser powers are used. In FIG. 29 the fluorescence is plotted as a function of the distance to the heat source (i.e. laser focus) is shown for the native and denatured form at two different laser powers (i.e. maximum temperatures of 5° C. and 10° C.). In both cases there are two contributions to the fluorescence change. First there is an increase or decrease in concentration (for the native or denatured form, respectively) and secondly there is a decrease in fluorescence due to the temperature dependence of the fluorescence dye. For a qualitative comparison (i.e. to distinguish between a native and denatured form.) a correction for the temperature dependence of the dye is not necessary. In all cases shown in FIG. 29 the fluorescence decreases. But as expected the decrease in fluorescence for the native protein is not as strong decrease observed for the denatured form. This is readily explained by the negative Soret coefficient of the native protein at 20° C. ambient temperature (see also FIG. 28), which leads to an accumulation of molecules at higher temperatures. This counteracts the decrease in fluorescence caused by the temperature dependence of the fluorescence. At higher laser powers (i.e. a maximum temperature increase of 10° C.) the difference between native and denatured form is even stronger because the thermophoretic accumulation and thermophoretic depletion, for the respective form, get stronger. Also smaller conformational changes can be detected on the basis of different strength of thermophoresis. A different direction of thermophoretic movement is advantageous but not necessary. b) A sample of fluorescently labelled bovine serum albumin (BSA) has been split in two parts. One is only exposed to ambient temperatures, while the other half is heated up to 100° C. for several minutes (i.e. irreversibly denatured). The thermo-optical properties of both samples (native and denatured) are measured at 800 mA power of the infrared laser (i.e. maximal temperature increase of 20° C.). As can be seen from FIG. 30, the fluorescence of the denatured protein is lower than the fluorescence of the native protein. This is explained as follows. The fluorescence dye of both samples shows the same decrease in fluorescence due to the increase in temperature (i.e. temperature sensitivity of the fluorescence). But the denatured protein shows a positive thermophoretic mobility (i.e. moves to the cold), while the native protein has a negative thermophoretic mobility (i.e. moves to the hot). The accumulation at elevated temperatures is the reason, why the decrease in fluorescence is lower for the native protein, while the denatured protein is, in addition to temperature dependency, depleted from the region of elevated temperature. Interestingly, by approaching the denaturing temperature (i.e. 50° C.) of the protein the amplitudes of the native and denatured protein approach each other an are essentially the same. This means that by measuring the amplitude of the fluorescence change an comparison to the reference sample allows to detect the melting temperature of a protein and to discriminate between native and denatured form of a protein. And to detect a shift in melting temperature due to interactions of the protein with other biomolecules or small molecules (e.g. drug candidates).

Example 7

Optothermal Trap/Thermooptical Trap

(132) In the following, silica particles are employed as illustrative particles/beads to be thermo-optically trapped by the method of the present invention. It is understood that the described method can also be employed for the thermo-optical trapping of other molecules, like biomolecules or lipid vesicles (as also illustrated in a further example).

(133) Silica particles (1 μm diameter, plain, Kisker Biotech) are diluted 1/100 in distilled water. 2 μl are pipetted on an object slide (Roth, 1 mm thick) an sandwiched between an cover slip of 12 mm diameter. In the following the term bead is used as a synonyme for particle. The aqueous bead-containing solution spreads uniformly in between the two glass surfaces, yielding a thin sheet of water with a height of approximately 20 μm. The liquid sheet is sealed by using nail polish. This avoids rapid evaporation of the sample. The microfluidic chamber is placed on the object stage of the thermo-optical setup (as e.g. illustrated in the appended FIG. 1a) and imaged via a 40× oil objective (NA 1.3, Zeiss). The laser focus is positioned so that it is approximately in the center of the field of view and has a half width of about 100 μm. Then the IR Laser is turned on and heats the solution to 10° C. above ambient temperature (20° C.) at the maximum of the spatial temperature distribution in the center of the IR laser focus. The images series shown in FIG. 33 illustrates the process of particle accumulation/trapping. In the beginning (first image of FIG. 33, top of the page), without laser heating, the beads are almost equally distributed. The black circle shows the position of the laser focus. The following images show the development of the particle distribution in the next three seconds after the heating laser is turned on. The particles show a directed movement to the region of elevated temperature at the laser focus, which can be also referred to as negative Soret effect or negative thermophoresis. Surprisingly, silica particle show negative thermophoresis at room temperature. The particles are trapped in the center of the temperature distribution at the laser focus. The particle experiences a potential well created by the spatial temperature distribution. The directed motion to the region of highest temperature is explained by the particle's tendency to minimize its energy of solvation. The position of the particle is not exactly in the center of the heat spot, since the thermal fluctuations push the particle out of its position. Appended FIG. 34 illustrates that the beads are trapped within the region of highest temperature even when the stage (i.e. sample) is moved with speeds of millimeters/second relative to the fixed laser focus. By using the thermo-optical trap, particle can be moved arbitrarily and can also be concentrated. After confining antibody modified beads to the heat spot, an interaction between the particles due to a binding of a single antigen in the solution to more than one bead can be detected.

(134) Another approach of a thermo-optical characterization in accordance with this invention is shown in FIG. 32. Silica particles (1 μm diameter, plain, Kisker Biotech) are diluted 1/1000 in distilled water. The dilution factor is empirical. The dilution is such that only a single particle is observed in a region of approximately 400 μm times 400 μm. 2 μl are pipetted on an object slide (Roth, 1 mm thick) an sandwiched between an cover slip of 12 mm diameter. The aqueous bead-containing solution spreads uniformly in between the two glass surfaces yielding a thin sheet of water with a height of approximately 20 μm. The liquid sheet is sealed by using nail polish. This avoids rapid evaporation of the sample. The microfluidic chamber is placed on the object stage of the thermo-optical setup (e.g. FIG. 1a). The laser focus is positioned so that it is approximately in the center of the field of view and has a half width of about 100 μm. Then the IR Laser is turned on and heats the solution to 10° C. above ambient temperature (20° C.) at the maximum in the center of the IR laser focus. Due to the high dilution, a single particle is trapped in a potential well created by a spatial temperature distribution (s. FIG. 32a). As silica particles show a negative thermophoresis the well is deepest at high temperatures (i.e the particles minimize their energy of solvation at high temperatures). The single silica particle fluctuates in the potential well since the thermal fluctuations push the particle out of its position. The fluctuations are recorded via a CCD camera (at t=1 s, 2 s, 3 s, 4 s, 5 s, 6 s, 7 s) and the positions are tracked by Software (selfmade Software, Labview National Instruments, detecting the pixel with highest intensity) with nanometer resolution (see FIG. 32b). A histogram is calculated from the positional information (see FIG. 32c). The width of the distribution is very sensitive to the thermo-optical properties of the particle. If molecules bind to the surface of the particle, the effective potential for the bead changes and the amplitude of the fluctuations increases or decreases. By observing the amplitude change over time, a kinetic binding curve can be measured. In a microfluidic system such an experiment is performed as follows: a solution containing a 1/1000 dilution (distilled water) is flushed or soaked into a capillary (see FIG. 27). The valves at the end of the capillary are closed. A single particle (modified (e.g. coated) with an antibody specific to a certain antigen, e.g. Interleukin 4) is trapped and the fluctuations of the particle are detected for 10 seconds to 100 seconds. The beads is surrounded by pure buffer solution. In a next step the buffer is exchanged with a buffer containing the respective antigen. While the buffer is exchanged the bead is still trapped. After exchange of buffer the fluctuation of the same bead as before are recorded. The change in fluctuation amplitude is used to detect interactions between antibody and antigen. Its change over time is used to measure binding kinetics.

(135) By using a device as illustrated in the appended FIG. 24, a temperature gradient can be generated in solution, by scanning lines (e.g. 10) perpendicular to each other in the solution. Where these lines cross, a temperature maximum is observed. Points on the scanned line have an intermediate temperature, while spaces in between the lines represent temperature minima (e.g. ambient temperature if spaces in between the lines are sufficiently wide). The silica particle described above will move to equilibrium positions on the crossing points of the heated lines. By moving the temperature grating all particles will move simultaneously. In addition the fluctuation of all beads may be measured simultaneously.

Example 8

DNA Melting Curves

(136) Standard protocol for the measurement of melting curves (using the device shown in e.g. FIG. 1a, 1b, 16 to 18, 20 to 24 or 37): Raman-Laser 1455 nm; coupled to galvanometric mirrors via fibre No collimator after fibre, Laser-beam hits mirrors divergently; mirrors are reflecting the laser-beam onto a lens; laser beam is focused to the chamber via the lens. Laser on/off is controlled via moving the laser in an out of the field of view via the mirrors. Preparations: Coverslips (170 μm) thick, one with a diameter of 12 mm the other quadratic 24×24 mm rinsed with deionized water, rinsed with ethanol, then again rinsed with deionized water. Dilution of solutions: +10 μM Tamra in 1×SSC + Hairpin from 100 μM in MilliQ-water diluted to 10 μM, 1 μM, 100 nM, or less in SSC-Buffer (1×, 0.5×, 0.1× or less). + add the detergent TWEEN20 to an end volume of 0.01% (only if there is unspecific adsorption). Adjustment: Check if everything is ok with the microscope (apertures, filters) 10 μM Tamra (tetramethylrhodamine) in 1×SSC (150 mM NaCl, 15 mM Na3-citrate, pH 8.1), with 0.01% TWEEN 20; volume of 2 μl into chamber built with 2 Coverslips (170 μm thick), sealed with nail polish Wait until nail polish is dry Add immersion oil onto the upper coverslip Put chamber onto the measurement stage Focussing fluorescence image by nearly closing the aperture Find laser spot Focus laser Defocus laser by increasing the distance between lens and chamber Adjust laser focus in a way that a broad and appropriate temperature distribution is established Move laser spot with galvanometric mirrors out of the field of view in such a way that the influence of the laser to the fluorescence image is minimal.fwdarw.save this settings to the measurement program (avoid to cross the zero point on the voltage scale of the galvanometric mirrors) Measure the room temperature Measurement: Use the following steps for the measurement of temperature as well as for the measurement of melting curves: Conduct the measurement with the trigger-programme. Settings: 40× oil immersion-objective, 8×8 Binning, 10 ms exposure time (−>28 Hz readout rate) To do manually: 1. Laser on 2. Light source on (open shutter in case of HXP), write down the settings of the light source 3. Start the trigger program 4. Laser off Light source on

(137) The measurement is based on the detection of fluorescence and therefore may be conducted in a device according to appended FIGS. 16-18 and/or 20-24. As the exact timing of measurements is preferable, the used devices like CCD, IR-Laser and light source are synchronised by the use of an electronic trigger signal. As the used CCD-camera (Andor Luca) has a trigger output port, the IR-Laser control element and the light source control element are synchronised with the trigger signal of the CDD-camera. In particular the second high level of the CCD trigger output signal is taken as the zero point of time of the measurement. In the following the exposure time of the CCD-camera is 10 ms and the minimal time span between two images is determined by the frame rate of the CCD-camera. As chamber a 2 μl solution of the fluorescently labelled sample molecules (e.g. 10 μM tetramethylrhodamine (TAMRA) in 75 mM NaCl, 7.5 mM Na3-citrate, 0.01% TWEEN 20, pH 8.1) is sandwiched between two 170 μm thick glass coverslips with diameter of 12 mm and sealed with nail polish. This results in a thickness of the chamber of about 20 μm. Then the chamber is moved to the measurement device and the optics are focussed to the chamber.

Brief Description of the Sequence of the Measurement

(138) Before the measurement starts, the fluorescence background of the measurement device is recorded in “Step 0”. As this background is characteristic for the used device and may not change during a long time period this step is preferably done only once to characterize the used device.

(139) TABLE-US-00002 Time t = 0: Step 1: a first fluorescence image of the spatial fluorescence distribution in the chamber is recorded. Time t = 20 ms Step 2: the IR-Laser is switched on Time t = 60 ms Step 3: a second fluorescence image of the spatial fluorescence distribution in the chamber is recorded.

(140) After these steps the measurement is finished and the raw data processing and the data evaluation is conducted.

Detailed Description of the Sequence of the Measurement

(141) Step 0, Background Measurement:

(142) A sample buffer without fluorescently labelled sample molecules/particles is filled into a microfluidic chamber and the spatial fluorescence distribution in the chamber is measured with the CCD-camera, while the excitation light source is turned on.

(143) Step 1, Determination of Fluorescence Level at Ambient Temperature Before Laser Heating:

(144) An aqueous solution of a temperature sensitive dye (e.g. TAMRA) at a given concentration (e.g. 10 μM) is filled in the chamber which preferably provides a defined height of the chamber. Fluorescence is excited with the light source (LED) and recorded with spatial resolution with the CCD-camera at ambient temperature.

(145) First the CCD-camera is started and the first high level of the camera trigger output signal is used to synchronise the IR-Laser control element and LED control element with the CCD-camera. For the synchronising a measurement card from e.g. National Instruments can be used.

(146) For good illumination the LED is turned on with the use of the camera trigger signal a short time before the CCD-camera records a first fluorescence image I.sub.0(x,y). Therefore during the exposure time of the CCD-camera of 10 ms the fluorescence excitation light source has reached its steady state level of light output. When the camera starts its recording the output trigger signal of the camera is the second time at the high level state. This second high level state of the output signal determines the zero point of time t=0.

(147) Step 2, Starting of Infrared Laser Heating:

(148) The infrared heating laser is turned on at time t=20 ms and the spatial temperature distribution is established within a few milliseconds within the solution. The temperature distribution has been calibrated once in a way that for example all temperatures between 30° C. and 90° C. are present in the recorded image and it is not necessary to repeat this calibration every time an experiment is performed.

(149) Step 3, Recording of the Spatial Fluorescence Profile with Infrared Laser Heating:

(150) At t=60 ms, 40 ms after IR-Laser irridation was started, a second fluorescence image I.sub.1(x,y) is recorded with the CCD-camera with an exposure time of 10 ms.

(151) After the CCD-camera has taken the second picture, the first and the second picture are saved to the hard disk of a PC for processing the raw data and for data evaluation. Then the measurement process is finished.

(152) Processing the Raw Data and Data Evaluation:

(153) Because of the short time of measurement no correction for bleaching may be necessary. The images I.sub.0 and I.sub.1 are corrected against camera background. Then calculating the ratio K(x,y)=I.sub.1(x,y)/I.sub.0(x,y) for each pixel of the fluorescence images (second image divided by the first image, both background corrected) ensures the removal of artefacts from an inhomogeneous illumination. As the temperature dependence F(T) (FIG. 15) of the dye TAMRA is known from a calibration experiment in a fluorimeter, the spatial temperature distribution T(x,y) (FIG. 3c) can be derived from to the ratio K(x,y).

(154) As the timing of measurement is the same for the temperature measurement and for the measurement of the melting curve, both measurements can be conducted at one time in one chamber if the emission spectrum of the dye for the temperature measurement (e.g. TAMRA) can be well separated from the emission spectrum of the fluorescent label of for example the molecular beacon (e.g. HEX as fluorophore (see appended FIG. 6) and Dabcyl as a quencher).

Example 9

Detection of Covalent and Non-Covalent Modifications of Nanoparticles

(155) An example for the detection of covalent and non-covalent modification is shown in FIG. 35. Nanoparticles (i.e. nanocrystals or quantum dots) have been obtained from Invitrogen. Particles were purchased which have charged polymere coating to be stable in aqueous solution (diameter 12 nm). In addition particles with covalently coupled strepavidin were purchased (diameter 21 nm), as well as a biotinylated 40 bases single stranded DNA. The following samples have been prepared: First unmodified nanoparticles diluted to 1 μM concentration in 1×SSC (Saline-Sodium Citrate) buffer. Secondly, streptavidin coated nanoparticles diluted in 1×SSC buffer to 1 μM concentration. Thirdly, streptavidin coated nanoparticles diluted in 1×SSC buffer to 2 μM concentration were mixed with 2 μM 40 bases biotinylated single stranded DNA (IBA GmbH, Gottingen) in 1×SSC buffer in a 1/1 ratio. All three experiments were prepared in 1×SSC buffer in accordance with the following protocol: 2 μl of sample are pipetted on an object slide (Roth, 1 mm thick) an sandwiched between an cover slip of 12 mm diameter. The aqueous solution spreads uniformly in between the two glass surfaces leading to a thin sheet of water with a height of approximately 20 μm. The liquid sheet is sealed with nail polish to avoid rapid evaporation of the sample. The microfluidic chamber is placed on the object stage of the thermo-optical setup (e.g. FIG. 1a) and imaged via a 40× oil objective (NA 1.3, Zeiss). The laser focus is positioned so that it is approximately in the center of the field of view and has a half width of about 20 μm. The experiments were conducted as described herein above for the detection of interactions, with a maximum temperature increase of 5° C. above room temperature. A correction for temperature dependent fluorescence has been performed for precise measurement of the Soret Coefficient from the spatial concentration distribution. (i.e. the temperature dependency of the fluorescence of the nanocrystal has been determined in an independent fluorimeter experiment). As can be seen from FIG. 35, the Soret coefficient is positive and larger for the streptavidin coated nanoparticle than for the nanoparticle without protein. Furthermore the binding of a single single-stranded DNA molecule to the particle is detected as a change in Soret coefficient. This is interesting, since the short flexible DNA molecule does not contribute substantially to the size of the nanoparticle (as the streptavidin does). Since thermophoresis is sensible to changes of the surface properties upon binding of the DNA molecule, the binding of the comparatively small DNA molecule to the particle can be detected.

Example 10

Thermophoresis and Thermophoretic Trapping of Lipid Vesicles

(156) Texas-Red DHPE (1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt), a fluorescently labelled phospholipid, is used for staining the vesicles. The lipid is added to the process of vesicle formation with approximately 1 mol percent with respect to the main constituting phospholipids. The following stock solutions were used for preparation of the vesicles by electro swelling:

(157) Lipid Stock Solution: 2.5 mg/ml DPhPC (1,2 Diphytanoyl-sn-Glycero-3-Phosphocholine) in Chloroform (CHCl3) 1-2% stearylamine 1% fluorescent lipid (1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine, triethylammonium salt (Mol. Probes)

(158) Sucrose Stock Solution: 300 mM Sucrose in distilled water

(159) Buffer Solution: 150 mM KCl, 20 mM MES, pH 5

(160) The final vesicle solution is diluted 1/100 in distilled water. 5 μl of the dilution were pipetted on an object slide. The droplet was sandwiched in between the object slide and a round 12 mm diameter coverslip and placed on the measurement apparatus. Fluorescence was observed through an oil objective using the setup from 1a, and alternatively the setup from FIG. 19. In the latter case, IR-Laser heating is performed through the same objective. The maximum temperature increase was about 15° C. The temperature profile had a half width of 20 μm. Images before and after infrared laser heating were taken. Two representative images before and 10 seconds after infrared laser heating are shown in FIG. 39. The vesicles are attracted by heat spot due to negative thermophoresis. They are accumulating in the centre of the heat spot. The catchment area is dependent on the width of the temperature profile (i.e. a sufficient strong temperature gradient is necessary to drive directed particle motion toward the heated centre within a finite time). The trapping of vesicles has several, important applications, e.g. vesicles (as well as cells) can be transported and moved within a solution. Also, the fluctuation of single vesicles in the potential well (i.e. created by temperature increase) can be observed. Since the amplitude of fluctuations is dependent on the properties of the vesicle, any changes, like protein binding to vesicles or the activity of a membrane protein, e.g. an ion-pumping membrane protein can be detected as a change in fluctuation amplitude. The sign of the thermophoretic motion (e.g. attracted to the heat centre or repelled by the heated centre) is dependent on the properties (e.g. charge, size, surface modification, protein binding). When a vesicle which is normally attracted by the region of highest temperature is repelled from the heated centre, this is indicative for a change in the properties of this vesicle. This behaviour may be observed after changing the buffer around the vesicle to a solution containing, e.g. a binding partner. Also the behaviour of two samples containing vesicles in buffer and vesicles in buffer with binding partner (or membrane protein activating substance, e.g. ATP) may be compared or observed. Finally, the differences in the thermophoretic properties can be used to sort vesicles or cells.