WO1995009355A1

WO1995009355A1 - A new method of detecting substances in a liquid

Info

Publication number: WO1995009355A1
Application number: PCT/SE1994/000887
Authority: WO
Inventors: Anders Hanning
Original assignee: Anders Hanning
Priority date: 1993-09-28
Filing date: 1994-09-26
Publication date: 1995-04-06
Also published as: JPH09503064A; EP0721580A1; SE503289C2; SE9303149D0; SE9303149L; AU7825994A

Abstract

In a method of determining the concentration of substances in a liquid, the substances are first separated through liquid chromatography or capillary electrophoresis, and subsequently detected by monitoring the refractive index of the liquid flow at more than one wavelength, and the detection step comprises quantitative evaluation of the variation with wavelength of the refractive index, said quantitative evaluation comprising calculation of the difference, differential, or derivative of the refractive index with respect to wavelength, or integration, principal component analysis, PLS (partial least squares), factor analysis, or curve fitting.

Description

A NEW METHOD OF DETECTING SUBSTANCES IN A LIQUID

The present invention relates to refractometric detection of a substance or substances in a liquid chromatographic or capillary electrophoretic separation.

Capillary electrophoresis and liquid chromatography are modern and well-established separation methods with very high performance. The ideal detector for liquid chromatography and capillary electrophoresis, for example, should (i) measure concentration and thereby be miniaturizable, (ii) exhibit a sufficiently low concentration detection limit (<1 μM) , (iii) be applicable on column to avoid zone broadening in couplings or special detection cells, (iv) be fast so that the time constant of the detection will not cause a reduced resolution, and (v) be simple, robust and inexpensive.

The presently most popular detection technique in connection with the above mentioned separation methods is UV or visual light absorption. The absorbance, though, is proportional to the amount of absorbing substance in the path of the light ray rather than to the concentration of the absorbing substance, which makes absorbance methods unsuitable for miniaturization. A consequence of this is that the technique shows poor sensitivity in miniaturized systems.

An alternative technique is fluorescence measurement. Also this technique measures quantity rather than concentration and would therefore be unsuitable for miniaturization. Due to its very high mass sensitivity, the fluorescence method is, however, in practice miniaturizable. On the other hand, the instrumentation is complicated and expensive and the method is also sensitive to disturbing phenomena like stray light, background fluorescence, quenching and variations in the chemical matrix.

A third detection alternative is refractometry. This technique is more universally applicable, and may be used also when the substances of interest neither absorb nor fluoresce. The sensitivity and signal to noise ratio of presently available refractometric detectors are, however, not satisfactory, since the refractive index measurement is influenced by a number of disturbing phenomena, e.g. temperature and pressure variations. This reduces the attraction of refractometry in capillary electrophoresis and liquid chromatography. The fact that refractometry is a universal (as opposed to selective) technique may be an advantage or a disadvantage in different applications. An advantage of refractometry is the fact that the refractive index of a liquid depends on the concentration of dissolved substances. The technique is thus truly miniaturizable, and it may be applied also to narrow separation columns or capillaries. Further, refractometry is a fast technique, and refractometers are in general simple, robust, and inexpensive instruments.

The above mentioned deficiencies regarding the sensitivity and selectivity of refractometry relate to conventional refractometry, in which the refractive index is measured at a single wavelength. Several suggestions have been presented in order to improve the refractometric technique and to overcome the above mentioned deficiencies by measuring the refractive index at more than one wavelength simultaneously. In US, A, 4 704 029, A. van Heuvelen described a method for determining the concentration of glucose directly in blood (without any preceding separation step) by measuring the refractive index of blood. In one embodiment of the invention, the refractive index was measured at two wavelengths, corresponding to the refractive index maximum and minimum within the anomalous dispersion region of glucose, simultaneously, and the refractive index difference between these two wavelengths was taken as a measure of the concentration of glucose. The advantage was that the measurement was made specific to glucose.

In Anal. Chem. 1988, 60, 2609, G. Gauglitz, J. Krause- Bonte, H. Schlemmer and A. atthes described a method of measuring the refractive index at several wavelengths simultaneously in connection with liquid chromatography. The idea was to increase the sensitivity by measuring within the region of anomalous dispersion, and to obtain spectral refractive index information, which would make it possible to identify the analyzed substances.

Measurement of anomalous dispersion is also described by A. Harming in the international patent application PCT/SE92/00558 as a means to increase the sensitivity of a type of surface plasmon resonance (SPR) and related assays based upon the measurement of chemical interactions on a sensing surface as changes of the refractive index of the surface layer. These changes are caused by the analyte involving or influencing the binding or release of a refractive index enhancing species to or from, respectively, the sensing surface. More particularly, the sensitivity is increased by matching the measurement wavelength with the absorptivity maximum of the refractive index enhancing species used in the particular assays, preferably a dye or chromophoric molecule, and specifically such that the measurement wavelength substantially corresponds to the maximum of the negative derivative of the absorptivity with respect to the wavelength. This may be accomplished either by selecting the index enhancing species to conform with the measuring wavelength of a particular instrument or application, or by selecting the measuring wavelength to conform with a specific index enhancing species.

In the international patent application PCT/SE94/00045, A. Hanning describes a method of determining an analyte in a liquid chromatographic or capillary electrophoretic separation by measuring bulk refractive index, characterized in that the analyte is labeled with a species having a high refractive index at the or at least one measuring wavelength. Further, in the international patent application PCT/SE94/00762, A. Hanning describes a method of determining an analyte in a liquid chromatographic or capillary electrophoretic separation by measuring the displacement of a species in the mobile phase by analyte as a change in refractive index, characterized in that the displaced species have a high refractive index at the or at least one measuring wavelength.

In accordance with the present invention, it has now been found that the basic principle of measuring the refractive index at more than one wavelength simultaneously may be further improved and may be made more generally applicable.

For all homogeneous optical materials (i.e. the liquid solution in the present case) , there is a universal and unambiguous relation between the absorption spectrum and the refractive index spectrum. In its most general case, this relation is described by the Kramers-Kronig optical dispersion equations (see e.g. N. . Ashcroft, N. D. Mermin, "Solid state physics", . B. Saunders,

Philadelphia, 1976, Appendix K) . In a simplified and rough way, the refractive index variation in the vicinity of an absorption peak (resonance wavelength) may be approximated by the negative derivative of the absorptivity with respect to wavelength. The refractive index shows a maximum in the vicinity of the maximum value of the derivative of the absorptivity with respect to wavelength (i.e. at a slightly higher wavelength than the resonance wavelength) , and a minimum in the vicinity of the minimum value of the negative derivative of the absorptivity with respect to wavelength (i.e. at a slightly lower wavelength than the resonance wavelength) . The variation of refractive index with wavelength is called dispersion. The wavelength region between the minimum and the maximum in refractive index is called the region of anomalous dispersion, while the wavelength region outside the anomalous region is called the region of normal dispersion. The dispersion has different signs in the normal and anomalous regions, respectively. A simplified description of the relation between absorption and refractive index, based on a local approximation, is given in C. F. Bohren, D. R. Huffman, "Absorption and scattering of light by small particles", Wiley, New York, 1983, Chapter 9. With this local approximation as a starting point, it may be shown that the refractive index difference between the maximum and the minimum is proportional to the maximum value of the absorption coefficient. For a dissolved substance in a liquid, this implies that the refractive index difference between the maximum and the minimum (or the refractive index difference between two arbitrarily chosen wavelengths in the vicinity of the absorption peak) is proportional to the product of the absorptivity of the dissolved substance and its concentration. This is also experimentally shown in the Example below. Thus:

Δn a * c (1)

, where Δn = refractive index difference, a = absorptivity, and c = concentration. This expression is as universally valid as the well-known Lambert-Beer law for quantitative absorption spectroscopy:

A = a * b * c (2)

, where A = absorbance and b = the path length of the light beam through the material. These two expressions are analogous, with one important difference: the factor b is involved in the Lambert-Beer law, which implies that the absorbance decreases with decreasing path length, and the sensitivity of the absorbance measurement gets worse when miniaturizing the system. The factor b is not involved in the expression describing the refractive index difference, and the sensitivity of the refractive index measurement does not deteriorate on miniaturization.

To sum up: 1/ Measurement of absorption spectrum or refractive index spectrum gives, in principle, the same information (as is expressed by the Kramers-Kronig equations) . 2/ Absorbance and refractive index difference are both linearly related to the concentration (as is expressed by equations (1) and (2) above) . 3/ The sensitivity of absorbance detection deteriorates on miniaturization, while the sensitivity of refractive index detection does not (as is expressed by equations (l) and (2) above) . In its broadest aspect the present invention provides a method of determining the concentration of substances in a liquid, characterized in that the substances are first separated through liquid chromatography or capillary electrophoresis, and subsequently detected by monitoring the refractive index of the liquid flow at more than one wavelength, and the detection step comprises quantitative evaluation of the variation with wavelength of the refractive index. The method puts no restrictions on the type of analytes, but the highest sensitivity is obtained for analytes having a high absorptivity close to the measurement wavelengths. The method may be used in the infrared, visible, or ultraviolet wavelength regions, but the highest sensitivity is obtained if the measurement wavelengths are chosen in the vicinity of a region in which the analytes have a high absorptivity. The method puts no restrictions on the chromatographic or electrophoretic separation mode that is utilized for the separation. Separation modes include, but are not restricted to, reversed-phase chromatography, adsorption chromatography, ion chromatography, ion pairing chromatography, size exclusion chromatography, affinity chromatography, capillary zone electrophoresis, capillary ion electrophoresis, micellar electrokinetic capillary chromatography (MECC) , isotachophoresis, capillary gel electrophoresis, and capillary isoelectric focusing. Other separation modes conceivable for the purposes of the invention will be apparent to the skilled person. The method puts no restrictions on the type of refractometer used for the measurement. Refractometer types include, but are not restricted to, deflection refractometers, interferometers, Fresnel refractometers, surface plasmon resonance refractometers, and optical waveguide refractometers. Other refractometers conceivable for the purposes of the invention will be apparent to the skilled person. In comparison with the method described by van Heuvelen in US, A, 4 704 029 (above) , the present inventive concept shows several essential technical differences. Firstly: van Heuvelen does not realize that it is practically impossible to determine the concentration of one single substance in a complex mixture of several other, identified or unidentified, substances simply by measuring the refractive index difference at two wavelengths. The chance of finding two wavelengths that are specific to this single substance is almost negligible, and especially so if the other substances in the mixture are unidentified. For example, when glucose is to be determined in blood, several other molecules, e.g. other kinds of sugars and carbohydrates, will absorb in the same region as glucose, and hence interfere with the measurement, in the present concept, this problem is solved by combining the refractive index measurement with a chemical separation step, namely liquid chromatography or capillary electrophoresis. In this way, all the different substances are first separated and then detected, one by one, and it becomes possible to analyze complex mixtures of unknown substances, like e.g. blood. Secondly: van Heuvelen utilizes two specific wavelengths in order to determine one specific substance, so the wavelengths are specific to that substance. In order to determine the concentration of several different substances in a complex mixture, two specific wavelengths would have to be identified for each substance, which is both theoretically and practically impossible. In the present concept, which relates to liquid chromatography and capillary electrophoresis, two, or more, fixed wavelengths are utilized, so the wavelengths are not specific to any single substance. All the eluting substances will be detected, although with varying sensitivities (some substances may even show close to zero sensitivity) . Thirdly: The method of van Heuvelen is explicitly limited to the region of anomalous dispersion, van Heuvelen's method is based on the fact that the dispersion has different signs in the normal and anomalous regions, respectively. The present concept is not limited to the anomalous region, and does not put any demands on the sign of the dispersion. It is the magnitude, whether positive or negative, of the dispersion that is monitored, so the present method may be applied in the normal region as well as in the anomalous region.

In comparison with the method presented by G. Gauglitz et al. (above), the present inventive concept constitutes a significantly improved method. Gauglitz measures the absolute value of the refractive index at several wavelengths simultaneously, but he does not at all utilize, or even mention the possibility to utilize, the dispersion (in the wide meaning variation of refractive index with wavelength, like e.g. difference, differential, or derivative of the refractive index with respect to wavelength) for quantitative purposes. Clearly, the use of the dispersion for quantitative purposes is not even implicitly apparent to Gauglitz. On the contrary, he clearly points out, that temperature variations, pressure variations, and mechanical instability influence the absolute value of the refractive index, but he does not realize that all these sources of noise will be largely cancelled out, if the dispersion is monitored instead of the absolute value of the refractive index. Further, Gauglitz explicitly defines "the present state of the art" for his presented method as a chromatogram where the absolute value of the refractive index is plotted against time, so neither in this case does he make use of the dispersion, i.e. the variation of refractive index with wavelength, for quantitative purposes.

In comparison with the international patent application PCT/SE92/00558 (above) , the present method does not relate to the detection of surface interactions, neither does it make use of any refractive index enhancing species. In comparison with the international patent application

PCT/SE94/00045 (above) , the present method does not utilize labeling of the analyte. In comparison to the international patent application PCT/SE94/00762 (above) , the present method does not utilize detection of a substance that is being displaced by the analyte.

In the present inventive method, the variation of the refractive index with wavelength is related to the concentration of analyte in the liquid. In the simplest embodiment of the method, the refractive index is monitored at two different wavelengths, and the measured refractive index difference is related to the concentration of analyte in the liquid. The refractive index difference is linearly related to the concentration, as stated in equation (1) above. In other embodiments, the refractive index is continuously monitored at more than two discrete wavelengths, or in a continuous wavelength interval. The quantification may again be performed as a simple difference between two values, or the quantification procedure may comprise e.g. differentiation, derivatization, or integration of the obtained refractive index spectrum, or the quantification procedure may include any of the common techniques, including multivariate techniques, for quantitative evaluation of spectra, like e.g. principal component analysis (PCA) , partial least squares (PLS) and factor analysis. If the separation of different substances is insufficient (i.e. overlap of eluting peaks) , or if the baseline is not stable (e.g. sloping or fluctuating baseline) , the resolution may be improved through the use of e.g. curve fitting or the above mentioned multivariate techniques.

In another aspect of the present method, the information concerning the variation of the refractive index with wavelength may be used to determine the structure of or to identify the analytes. In one embodiment of this aspect, the refractive index is measured at two wavelengths, and the information concerning the sign and magnitude of the dispersion is used for qualitative analysis. This is analogous to measurement of the absorption at a single wavelength, which also gives a limited structural information. In other embodiments, the refractive index is continuously monitored at more than two discrete wavelengths, or in a continuous wavelength interval. The spectral information may also in this case be used to determine the structure of or to identify the analytes. A large number of well-established methods for the evaluation of spectra may be used for the qualitative analysis in this case, e.g. differentiation, derivatization, integration, PCA, PLS, factor analysis, curve fitting, or comparison with collections of refractive index values, or reference spectra. The different measurements of the refractive indices at different wavelengths must be performed simultaneously or in very rapid succession, so that the different sources of noise due to fluctuations with time will not exert any influence. When measuring at two wavelengths, it is possible to utilize e.g. a light source with two emission wavelengths, that strike two different detector elements (as in the Example below) , or one detection element in combination with a rapidly rotating filter, or two different laser wavelengths, that are alternatively turned on and off in rapid succession. When measuring over an extended wavelength interval, it is possible to utilize e.g. a lamp with several emission lines, or with a broad emission spectrum, in combination with a monochromator (sometimes termed a polychromator, since all the different wavelengths are allowed to pass through the sample) and a diode array or CCD detector. Other conceivable technical solutions to perform the measurements simultaneously or in rapid succession will be apparent to the skilled person. As for refractometric detection in general, the absolute value of the refractive index need not actually be measured. The important thing is to monitor changes with time, so it is sufficient to measure the relative refractive index or some property that is related to the refractive index, like e.g. deflection angle, light intensity, or phase difference.

The present inventive concept has a number of advantages as compared to conventional single wavelength refractometry. By measuring the variation of the refractive index with wavelength at constant time (a "snapshot") instead of measuring absolute refractive index, the influence of the above mentioned sources of noise, like e.g. temperature and pressure variations, and mechanical vibrations, will decrease, leading to increased signal-to- noise ratio and improved sensitivity. Measurement at several wavelengths gives spectral information, that may be used for purposes of qualitative analysis. The degree of universal versus selective detection may be controlled through the choice of measurement wavelengths. Still, all the advantages of conventional refractometry remain: miniaturizability, rapidity, simplicity, robustness, and low cost. Conventional single wavelength refractometry may also be performed with a refractometer designed for multi- wavelength measurements. Single wavelength and multi- wavelength measurements may thus be performed simultaneously in one single detection cell.

It is readily understood, that the criteria for an ideal detector for liquid chromatography or capillary electrophoresis stated further above, will be met by a refractometer utilizing the present inventive concept.

The method of the invention will now be illustrated in the following non-limiting Example, reference also being made to the accompanying drawings wherein: Fig. 1 is a schematic diagram showing the experimental set-up used in the Example;

Fig. 2 is a schematic diagram showing the liquid handling system used in the Example;

Fig. 3 is a diagram showing the refractive index spectrum for the dye HITC used in the Example; and

Fig. 4 is a plot of laser spot distances vs. HITC concentrations.

EXAMPLE Differential refractometer An experimental differential refractometer instrument was constructed as schematically illustrated in Fig. 1. This instrument consisted of two lasers 1 and 2, respectively, a prism-shaped flow cuvette 3, a CCD camera 4, a TV screen 5 and a Polaroid® camera 6.

Both lasers l and 2 were of diode type with collimating optics. Laser l had a wavelength of 660 nm (more precisely 658.5 nm) (Melles-Griot) and was driven by a voltage unit (Mascot Electronics Type 719) . The other laser 2 had a wavelength of 780 nm (Spindler & Hoyer) and was driven by a second voltage unit, Diode Laser DL 25 Control Unit (Spindler & Hoyer) . The two lasers 1, 2 were mounted at right angles to each other on a steel plate 7.

At the point of intersection of the two laser beams, a blackened brass tube (not shown; inner diameter 20 mm) was fastened to plate 7. The brass tube had slits in which a short wavelength pass filter 8 (Melles-Griot) having a cut- off at about 700 nm was mounted at an angle of 45° to the beam directions. This filter 8 transmitted the 660 nm beam of laser l but reflected the 780 nm beam of laser 2 with the resulting effect that the two beams were made to coincide. An aperture of l mm diameter served as exit slit. Flow cuvette 3, a commercial dual prism cell cuvette (consisting of two 45° prism cells, 1.5 x 7 mm, 8 μl volume, with their hypotenuses applied against each other) for a liquid chromatography refractive index detector (Pharmacia LKB Biotechnology AB, Uppsala, Sweden) , was then screwed to the outside of the brass tube in connection to the exit slit thereof. Only one of the two prism cells of cuvette 3 was used and connected by tubes (not shown) to a simple liquid handling system that will be described below (the other cell remained empty) . Steel plate 7 was turnably mounted to an aluminium plate 9 at one end thereof by a screw bolt 10. CCD camera 4 (Panasonic WV-CD50) , driven by a Panasonic Power Supply WV- CD52, was mounted at the other end of plate 9 at a distance of about 0.8 m (varying a little between the test series to be described) from prism cuvette 3 to be vertically adjustable by a micrometer screw. A bent black steel hood was placed over the CCD camera to screen stray light. The CCD camera picture was projected to TV screen 5 (Electrohome 10"), to which was taped a transparent cross- ruled pattern (OH film with a 1.4 X enlarged millimeter paper copied on it) . The transparent cross-ruled pattern was used for measuring distances on the TV screen 5 by counting squares in the pattern. The latter was photographed with camera 6 (Polaroid® 600 SE) , mounted on a tripod, at distance of about 0.5 m using a Polaroid® 611 Video Image Recording Film. The enlargement from CCD camera 4 to TV screen 5 was about 30 X.

The lasers 1 and 2 were adjusted laterally and vertically and the brass tube laterally such that nice and symmetrical light pictures (1-1.5 mm spots) of the two respective laser beams were obtained above each other on the TV screen. As a general strategy, a low laser intensity, a high contrast on the screen and an exposure time of 1/4 second were used. Then the brightness was adjusted until a sharp picture for the eye was obtained, and finally the diaphragm was adjusted until a sharp picture was obtained on the photograph.

The position of the spots was determined by taking a photograph of the TV screen 5, and then counting the checks in a microscope to determine the positions of the left and right edges, respectively, of the spots, and the center of each spot was assumed to be halfway between them. Liquid handling system

The liquid handling system used is shown in Fig. 2 and consisted of a pump 11 of peristaltic type (Pi, Pharmacia LKB Biotechnology AB, Uppsala, Sweden) . The pump 11 was connected, on one hand, to a sample reservoir 12 via a tube 13, and, on the other hand, to a manual turn valve 14 via a tube 15. A tube 16 connected valve 14 with a drain. Via a 0.4 mm cannula 17, a tube 18 connected valve 14 with flow cuvette 3 described in connection with Fig. l. A tube 19 connected cuvette 3 with a cuvette drain. Manual valve 14 permitted liquid to be pumped either to the drain or to cuvette 3. During the measurements to be described below, liquid was pumped through the cuvette at about 23 μl/min. When the liquid was changed the valve 14 was turned to drain, and the flow was increased ten times to purge the pump tubes 13 and 15. Then the flow was decreased again, the valve was set in cuvette position and liquid was pumped for at least 15 minutes through the cuvette (due to the dead volume of the cannula 17) until measurements were conducted. Preparation of dye solutions A citrate buffer (pH 3, 0.1 M citrate, 0.4 M NaCl,

0.05% Tween® 20) was prepared by dissolving 21 g of citric acid (M & B p.a.) and 23 g of NaCl (Merck p.a.) in 1000 ml of purified water. 5 ml of Tween® 20 (Calbiochem 655206, 10%, protein grade) were added. 4 M NaOH (p.a.) was added to adjust the pH from 1.95 to 3.00, and the mixture was filtered through a 0.22 μm filter. 500 ml of the citrate buffer were then mixed with 500 ml of spectrographically pure ethanol and homogenized with ultrasonic sound for a couple of minutes. A 500 μM stock solution of the dye HITC

(1,1', 3, 3, 3', 3' -hexamethylindotricarbocyanine) was then prepared by mixing 14 mg of HITC iodide (Sigma H0387, 94% purity, M^, 537 g/mol, ; Sigma Chemical Co., St. Louis, Mo., U.S.A.) with 50 ml of the above prepared citrate/ethanol buffer. After homogenization with ultrasonic sound for a couple of minutes, the mixture was filtered through a 0.45 μm filter.

A series dilution to seven different concentrations of HITC was performed by diluting different volumes of the HITC stock solution with the citrate/ethanol buffer to 25 ml:

HITC (ml) HITC cone. (μM)

25 500

12.5 250 6.25 125

3.125 62.5

1.563 31.25 0.781 15.63

0 0

Measuring of refractive index

The refractive index spectrum for 1 mM HITC is shown in Fig. 3 (solid line: theoretically calculated curve, crosses: experimental data) . As appears therefrom, the measuring wavelength 780 nm (laser 2) is in the peak region of the refractive index on the high wavelength side thereof, whereas the second measuring wavelength 660 nm is on the refractive index minimum plateau.

With reference to the above described apparatus (Figs. 1 and 2) , the positioning of the liquid-filled cuvette cell (3) was adjusted by filling the cell with ethanol/water 50/50 (purified water and spectrographically pure ethanol) by means of a syringe and turning the steel plate (7) until the two laser light beams were centered above each other on the CCD camera. The distance between the cuvette and the CCD camera was 73 cm.

Refractive index measurements were then performed for the different HITC concentrations described above, the liquid handling being carried out as described above under "Liquid handling system". The laser intensity, TV brightness and camera exposure and diaphragm settings were adjusted for each different dye concentration to obtain sharp pictures of the light spots.

For each dye concentration, two photographs were taken with an interval of a couple of minutes. The distance between the two laser light spots was estimated as described above by counting squares. The results are presented in Figure 4 that shows the average values of the spot distance in millimeters as a function of the HITC concentration. The relation is rectilinear with a very good fitting, the determination coefficient being 0.9992. The slope of the straight line, i.e. the sensitivity with regard to HITC concentration, is 3.l μm/μM, or in angular units, 0.00024 °/μM, or, in refractive index units (RIU) , 2.0 μRIU/μM. The above described experiments clearly demonstrate the feasibility of the present inventive concept.

The invention is, of course, not restricted to the above specially described embodiments, but many changes and modifications may be made without departing from the general inventive concept as defined in the following claims.

Claims

1. A method of determining the concentration of substances in a liquid, characterized in that the substances are first separated through liquid chromatography or capillary electrophoresis, and subsequently detected by monitoring the refractive index of the liquid flow at more than one wavelength, and the detection step comprises quantitative evaluation of the variation with wavelength of the refractive index, said quantitative evaluation comprising calculation of the difference, differential, or derivative of the refractive index with respect to wavelength, or integration, principal component analysis, PLS (partial least squares) , factor analysis, or curve fitting.

2. The method according to claim l, characterized in that the determination comprises measurement of the refractive index at two different wavelengths.

3. The method according to claim 2, characterized in that at least one of the wavelengths is within, or in the vicinity of, a wavelength region of high absorptivity of the determined substances.

4. The method according to claim 2 or 3, characterized in that at least one of the wavelengths is outside the region of anomalous dispersion of the determined substances.

5. The method according to claim l, characterized in that the determination comprises measurement of the refractive index at more than two discrete wavelengths, or within a continuous range of wavelengths.

6. The method according to claim 5, characterized in that at least part of the discrete wavelengths, or part of the wavelength interval, is within, or in the vicinity of, a wavelength region of high absorptivity of the determined substances.

7. The method according to claim 5 or 6, characterized in that at least part of the discrete wavelengths, or part of the wavelength interval, is outside the region of anomalous dispersion of the determined substances.

8. The method according to any one of claims 1 to 7, characterized in that the optical measurement is based on deflection refractometry, interferometry, Fresnel refractometry, surface plasmon resonance refractometry, or optical waveguide refractometry.

9. The method according to any one of claims 1 to 8, characterized in that the separation step is based on reversed-phase chromatography, adsorption chromatography, ion chromatography, ion pairing chromatography, size exclusion chromatography, affinity chromatography, capillary zone electrophoresis, capillary ion electrophoresis, micellar electrokinetic capillary chromatography (MECC) , isotachophoresis, capillary gel electrophoresis, or capillary isoelectric focusing.