JP2015224895A - Recycle isolation method in liquid chromatography - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a recycle isolation method in liquid chromatography, performing isolation treatment in an isolation possible time zone which is identified by predicting a number of recycle which allows isolation to be performed.SOLUTION: When a chromatogram which is obtained by recycle from first time to n-th time is displayed on a monitor screen which is provided on a data processing device, as a chart having a peak in which isolation is insufficient, simulation is performed so as to acquire a chromatogram in which a peak is isolated to a degree that isolation can be performed, in a relationship with a peak distribution state, in which the chart is a reference chart. Thus, a number of recycle capable of performing isolation on n-th time or following time, and an elution time zone thereof are predicted in advance, it is displayed on the monitor screen as an isolation possible prediction chart determined by the relationship between the isolation possible recycle number and the elution time zone, and a target component is isolated from a sample in an actual time zone corresponding to the predicted elution time zone.

Description

  The present invention relates to a method for recycling fractionation in liquid chromatography, which can predict a recyclable time zone in advance by predicting a suitable number of recycles when purifying and fractionating a target component from a sample in liquid chromatography. Technology.

  There are two types of liquid chromatographs: one for analyzing trace components in a sample, and one for separating and purifying target components from a sample. For example, there is also a recycle preparative liquid chromatograph disclosed in Patent Documents 1 and 2, which is repeatedly circulated through the same separation column.

  FIG. 6 is a reprint of FIG. 1 in Patent Document 1 with its extraction code changed. According to the figure, the entire recycle preparative liquid chromatograph 51 is composed of a solvent tank 52 in which a solvent S as a mobile phase is stored, and a three-way intervened in the flow path of the solvent S flowing down from the solvent tank 52. A valve 53, a liquid feed pump 54 interposed in the flow path to pump the solvent S flowing in through the three-way valve 53 downstream, and the solvent S fed by pressure through the liquid feed pump 54 An injector 55 for injecting a sample, a separation column 56 composed of a pre-column 57 and a main column 58 for flowing down the mobile phase after sample injection, and detecting a sample component in the mobile phase that has passed through the separation column 56, for example, a differential refractive index. A detector 59 such as a detector, a data processing device 60 with a monitor screen for recording and processing data detected by the detector 59, and a mobile phase flowing down through the detector 59 are introduced. And it includes at least a recycle box 61 which is capable of arranging the flow path switching to three directions of the liquid supply pump 54 direction and fraction collector 62 direction and preparative manual side partial vessel 63 direction.

Utility Model Registration No. 3176511 JP 2011-89778 A

  However, in the case of the conventional recycle sorting method including the disclosed techniques of Patent Documents 1 and 2 above, it is necessary to repeatedly recycle until a separation chart suitable for sorting is obtained. However, there is an inconvenience that it is necessary to wait for a long time while gazing at the chart on the monitor screen until reaching a recycle frequency suitable for preparative sorting.

  In addition, in the case of using the conventional recycling sorting method, continuous sorting can be automated for the same sample whose number of recycling and sorting time is determined by the sorting data acquired in the past. However, for the first mixed sample with no past data, the elution time of the target component could not be predicted, and therefore there was a problem that continuous fractionation could not be automated.

  In the present invention, in view of the above-mentioned problems found in the conventional method, the number of recycle times is predicted, so that the chart shows a preparable time zone in advance by predicting the separation state suitable for sorting. It is an object of the present invention to provide a recycle fractionation method in liquid chromatography that can perform fractionation processing in a time zone and can be automated.

  The present invention has been made to achieve the above-mentioned object, and its structural features can be obtained from the 0th to the nth recycling on the monitor screen provided in the data processing apparatus in the recycling sorting method in liquid chromatography. When a chromatogram is displayed as a chart having insufficiently separated peaks, a chromatogram is obtained in which the peaks are separated to such an extent that they can be sorted in relation to the distribution state of the peaks using the chart as a reference chart. As a simulation chart, the number of recyclable recycles after the nth time and the elution time zone are predicted in advance, and a preparable prediction chart determined by the relationship between the recyclable number of recycles and the elution time zone. The target component is displayed on the monitor screen and the target component is extracted from the sample in the actual time zone corresponding to the predicted elution time zone. It is to taken.

  In the present invention, the simulation for obtaining a chromatogram in which peaks are separated to such an extent that they can be separated is the first peak that the recycling n-th chart displayed on the monitor screen is insufficiently separated. When this is the case, this is used as a reference chart, and the surface area from the peak starting point to the peak lodging point surrounded by the reference chart and the baseline is defined as the simulation range, and 2 for the peak with insufficient separation. Obtain the vertex and inflection point by performing the second derivative, find the number of single peaks and Gaussian curve from the obtained vertex and inflection point, and then make a temporary formula that combines each single peak, Each coefficient is slightly shifted so that this formula matches the actual insufficiently separated peak, and each difference in the direction in which the difference from the insufficiently separated peak becomes small at that time. Confirm the ratio of the number deviation, and shift each coefficient in the direction of decreasing the difference according to the ratio of the deviation. By shifting each coefficient a little so as to match the peak and confirming the difference from the unseparated peak at that time, it is repeated and finally the value of each coefficient is obtained. It is desirable to make it possible to predict the number of recycles and the elution time zone in advance.

  According to the present invention, the number of recyclable recycles can be predicted in advance, the elution time zone can be grasped, and the target component can be collected while waiting in that time zone. It is possible to eliminate the trouble of having to wait for a long time while observing the actual chart suitable for sorting while gazing at the actual chart.

  In addition, since the number of recycling cycles that can be sorted can be predicted in advance, data necessary for sorting must be input in advance to the data processing device even when sorting target components from different samples. This makes it possible to automate the sorting process.

Explanatory drawing which shows the example of an apparatus structure with which this invention method is applied. It is explanatory drawing which shows the relationship from the 0th recycle to the 3rd recycle chart, the recycle valve, and the fraction valve, (a) 0th recycle, (b) 1st recycle, (c) Recycle 2 The second time, (d) shows the third time of recycling. The graph which shows the relationship between the straight line which passes along three points, and a coefficient. The graph which shows the relationship between the measured value shown by a black dot, and the curve of the type | formula which combined two single peaks. It is explanatory drawing which shows the process sequence in the case where the simulation is performed in the first recycling as an example, and (a) shows the state in which the peak peak and the half width are determined based on the actual chart of the first recycling and the reference ( b) shows a state in which a single virtual peak obtained from the determined peak apex and half-value width is added together and a peak after deformation obtained by superimposing it on the first recycling real chart, and (c) shows a single virtual (D) shows the simulation result in relation to the simulation range a in FIG. 2 (b), the state of the peak after deformation when the peak is further slightly deformed. FIG. 1 is a reprint of FIG.

FIG. 1 is an explanatory diagram schematically showing a configuration example of a liquid chromatograph to which the method of the present invention is applied. According to the figure, the entire liquid chromatograph 11 is composed of a solvent tank 12 storing a solvent S as a mobile phase, a three-way valve 13 interposed in a flow path of the solvent S flowing down from the solvent tank 12, A sample is injected into the liquid feeding pump 14 interposed in the flow path so as to pump the solvent S flowing in through the three-way valve 13 downstream, and the solvent S pumped through the liquid feeding pump 14. A sample introduction device 15; a separation column 16 that causes the mobile phase after sample injection to flow down; a detector 17 such as a differential refractive index detector that detects a sample component in the mobile phase that has passed through the separation column 16; and the detection A data processing device 18 comprising a monitor screen 19 as output means such as a personal computer and a keyboard 20 as input means for recording and processing data detected by the device 17; Detector 17 and a control box 21 which is interposed between the data processing unit 18, the liquid feed passage L ₂ and fraction collector 24 by introducing a mobile phase flowing down through the detector 17 to the liquid feed pump 14 direction A recycle valve 22 disposed so as to be able to switch the flow path in two directions with the liquid feed path L ₃ in the direction, and a mobile phase interposed between the recycle valve 22 and the fraction collector 24 in the direction of the fraction collector 24. comprises at least a fraction Bal 23 to allow the flow path switching in the two directions of the liquid feed path L ₃ and the waste tank liquid feed path L ₄ to 25 directions to.

  In this case, the control box 21 is configured with an electronic circuit or the like for controlling the opening and closing of the flow path switching valve including the recycle valve 22 and the fraction valve 23 based on a control signal from the data processing device 18.

  Further, the data processing device 18 includes a chart recorder (not shown) for outputting the chromatogram 41 as a chart on recording paper as another output means.

  Next, when the method of the present invention (recycle fractionation method in liquid chromatography) applied to the liquid chromatograph 11 having the above configuration is applied, a peak 35 insufficiently separated is displayed on the monitor screen 19 in the first recycling. An example will be described below with reference to FIGS.

  First, the chromatogram 31 obtained at the 0th recycling is composed of unseparated peaks 33 that are not suitable for sorting on the monitor screen 19 provided in the data processing device 18, for example, as a chart 32 at the 0th recycling shown in FIG. Therefore, it cannot be confirmed whether or not it is a single peak.

Therefore, the mobile phase flowing down the detector 17 is performed first recycling flowing the liquid feed path L ₁ again through the three-way valve 13 by the open recycling valve 22 to the liquid feed path L ₂ side, The chromatogram 31 obtained at that time is displayed on the monitor screen 19 as an actual chart 34 for the first recycling composed of peaks 35 that are insufficiently separated as shown in FIG. Incidentally, unnecessary peak component appearing late in the negative direction in FIG. 2 (a), is discharged to the waste tank 25 by the open and recycling valve 22 and the fraction valve 23 to the liquid feed passage L ₄ side The

  Thus, when the actual chart 34 of the first recycling that is insufficiently separated is displayed on the monitor screen 19, the peak 34 is displayed to the extent that it can be sorted in relation to the distribution state of the peaks 35 using the chart 34 as a reference chart. 2B is predicted and displayed as a prediction chart 40 ′ for the third recycling corresponding to the actual chart 40 for the third recycling constituted by the peak 41 shown in FIG. A predetermined simulation required for the above is performed.

  In other words, by performing the above simulation, the monitor screen 19 has a recyclable prediction chart determined in relation to the recyclable third time that is the recyclable number of recycles for the nth recycle, in the illustrated example, the recycle from the first recycle. As a result of predicting and displaying the possible number of recycles as the prediction chart 40 ′ for the third recycle, the number of recyclable recycles can be known in advance.

  Therefore, when actually collecting the target component, the recyclable valve 22 and the fraction are kept waiting in accordance with the time zone that can be sorted as shown in the actual chart 40 of the third recycling in advance. By opening the valve 23 to the liquid feed path L3 side, the target component can be separated into the fraction collector 24.

Here, the specific process for the above-described simulation will be described in more detail. The chromatogram 31 obtained by liquid chromatography is displayed on the monitor screen 19 as a real chart 32 for the 0th recycling. That is, the sample introduced into the liquid feed path L ₁ via the sample introduction device 15 of the liquid chromatograph 11 illustrated in Figure 1, chromatogram 31 is detected by passing through the detector 17, to the chromatogram 31 For example, as shown in FIG. 2A, the actual chart 34 corresponding to the zeroth recycling is displayed on the monitor screen 19.

  In this case, the actual chart 32 of the zeroth recycling displayed on the monitor screen 19 does not know how many single peaks exist because the peaks 33 are overlapped without being separated, so at this stage The chromatogram 31 is not in a state suitable for sorting.

Therefore, the mobile phase is circulated from the liquid feed path L ₂ to the liquid feed path L ₁ through the recycle valve 22 and the three-way valve 13, and again passes through the separation column 16 and the detector 17. As shown in FIG. 2B, the actual chart 34 for the first recycling is displayed on the monitor screen 19 at a position (right side position in FIG. 2) that is twice as long as the retention time in the actual chart 32 for the zeroth recycling. Will be displayed in chronological order.

The actual chart 34 for the first recycling in FIG. 2B is displayed on the monitor screen 19 as an insufficiently separated peak 35 having two peak vertices 35a and 35b, for example, unlike the actual chart 32 for the zeroth recycling. cage, so that during the up peak lodging point P _2, which falls down from the peak starting point P ₁ to the peak 35 rises is defined as the simulation range a to be described later.

  3 (a) to 3 (d) show examples of simulation processes in the case where the actual chart 34 of the first recycle consisting of the insufficiently separated peak 35 located in the simulation range a shown in FIG. 2 (b) is used as a reference chart. It is explanatory drawing which shows.

  That is, the first recycle chromatogram 31 is, as shown in FIGS. 2B and 3A, for example, a recycle 1 consisting of an insufficiently separated peak 35 having two peak vertices 35a and 35b. It is displayed on the monitor screen 19 as the actual chart 34 for the second time.

  Here, the baseline BL and the retention time of the peak 35 having two peak vertices 35a and 35b are obtained. Here, the baseline BL is obtained if the heights of the two peak vertices 35a and 35b (the length of the perpendicular line from the baseline BL to each of the peak vertices 35a and 35b) and the shape are not accurately obtained, and an approximation of the peak waveform is obtained. This is because a curve cannot be obtained. Further, the retention time is obtained by the time until the detector 17 confirms the peak vertices 35a and 35b after the sample is introduced through the injection (not shown) constituting the sample introduction device 15. Since the time (retention time for the same sample when recycling) is equal, the time after doubling the retention time obtained in advance is the next time, that is, the position where the peak 39 of the second recycling appears. It comes from becoming.

Base line BL shown in FIG. 2 (a) obtains first the peak start point P ₁ and the peak lodging point P ₂ of the peak 35, can be obtained by drawing a straight line between them. This is the peak start point P ₁ and the peak lodging point P ₂ in the case, to remove the noise component from the waveform data, seeking second derivative can be confirmed from the maximum minimum value of the inflection point. The drawn straight line is adopted as the current baseline BL if it is determined that the height and inclination of the straight line before and after the straight line are substantially the same. The range of height and inclination necessary for determining the base line BL is automatically determined for those within a certain allowable range with reference to the base fluctuation at the position where the peak 35 is not present.

  In this case, the surface area surrounded by the actual chart 34 and the base line BL of the first recycling shown in FIG. 2B in which the target peak 35 is not sufficiently separated is set as the simulation range a. become.

  As shown in detail in FIG. 3A, the actual chart 34 for the first recycling, which is a reference chart for performing the simulation, is the height a1, of the peak vertices 35a and 35b within the simulation range a. a2 and the half-value widths 1σ and 2σ at half the heights a1 and a2 are determined.

  After that, it is obtained from one peak peak 35a and its half-value width 1σ and one single virtual peak 36 consisting of a Gaussian curve indicated by a broken line, the other peak vertex 35b and its half-value width 2σ. The other single virtual peak 37 composed of a Gaussian curve indicated by a broken line is superimposed on the actual chart 34 of the first recycling as shown in FIGS. 2 (b) and 3 (a). In FIG. 3A, 'time m1 is the retention time to one peak vertex 35a in the first recycling actual chart 34, and' time m2 is the other peak vertex 35b in the first recycling actual chart 34. Each retention time is shown.

  In addition, in this case, one single virtual peak 36 and the other single virtual peak 37 indicated by a broken line can be obtained by a predetermined simulation method given to the data processing device 18 as a program as shown in FIGS. Slightly deform as shown in That is, in the present invention, a single peak is assumed to be a Gaussian curve, and the coefficient of the Gaussian curve is first obtained so that the sum of the heights at the same position of each single peak included in the non-separable peak matches. Then, the position, height, and width of the single peak are grasped to some extent by obtaining the positions of the peak vertices 35a and 1σ from the differential equation of the unseparated peak, and the temporary peak of the unseparated peak is obtained from these data. Create an expression. In this case, since one Gaussian curve has three coefficients, it is necessary to obtain 3 × n coefficients if the provisional formula includes n peaks. Since the Gaussian curve is non-linear, for example, the Newton-Simpson method is used as a coefficient to obtain the non-linear least square method.

  In this case, if it is linear like a straight line as shown in FIG. 4, the coefficient can be easily obtained by the method of least squares. For example, when obtaining a straight line passing through three points, the coefficients a and b of the straight line are obtained as shown in FIG. Here, the least square method is to minimize the distance from each point to the straight line, and the coefficient is determined by differentiating the square of the distance and obtaining the minimum value 0 as in the following equation (Equation 1). To do.

  In the present invention, since each point is data of an actual peak 35 that is inseparable, a straight line is an equation obtained by adding the Gaussian curves. The following equation (Equation 2) is an equation of a Gaussian curve, the coefficients are a, m, and σ, and the vertex is found by second-order differentiation of the unseparated peak.

  The number of single peaks is also found from the number of vertices found, and as shown in FIG. 3A, the height of the vertices (for example, 35a and 35b) in the y-axis direction is a, The position of the vertex (for example, 35a, 35b) in the x-axis direction is substantially equal to m. Further, since the inflection point of the Gaussian curve coincides with the position of σ, the position (σ) is determined. Each single peak (Gaussian curve) consisting of one single virtual peak 36 and the other single virtual peak 37 thus obtained is an unseparated peak (first recycling) as shown in FIG. Although it can be superimposed on the actual peak 35), since the height of each point is the sum of the heights of each single peak, it is still not exactly as shown in FIG. 3 (b) as a thick broken line. Cannot match.

  The curve shown by the solid line in FIG. 5 is a curve derived from the following equation (Equation 3) that combines two of the single virtual peak 36 and the single virtual peak 37 (the peak 45 after deformation). '), And the points scattered along the curve are actually measured values.

  In this case, the six coefficients (x1, y1, xn, yn, xm, ym) of the curve (deformed peak 45 ′) are obtained by calculation using the least square method, but the equation is nonlinear. Therefore, it cannot be easily obtained. For this reason, for example, the Newton-Simpson method is used. Here, as in the case of obtaining the linear coefficient described above, first, the coefficient is a value used in the second derivative, each coefficient is slightly shifted, and then squared and divided by the numerical value shifted.

  Here, if U is a value obtained by subtracting each point and squaring, U can be obtained by the following equation (Equation 4).

  Accordingly, if a1 and m2 are shifted by Δa1 and Δm2, the following equation (Formula 5) is established.

  Here, an inflection point at which Z becomes 0 is performed on a1, m1, σ1, a2, m2, and σ2 to search for a value that becomes 0. That is, a peak after deformation in which a1, m1, σ1, a2, m2, and σ2 are slightly shifted and one single virtual peak 36, which is a single peak superposition, and the other single virtual peak 37 are superposed. Look for a direction where 35 'coincides with the unseparated peak 35. At this time, in the first calculation, a1 and a2 are greatly reduced, and σ1, m1, σ2, and m2 have almost the same numerical values.

  In the above equation (Equation 5), a small value is set for Δ, and the values of a1 and Δm2 are determined depending on how close Z is to 0. At this time, if Z is far from 0, the values of Δa1 and Δm2 are negative. In this calculation, once the above equation (Equation 5) is differentiated, that is, second-order differentiated, it is possible to know which coefficient change contributes to the ratio of Z to 0. In other words, here, the operation of repeatedly checking which coefficient is shifted to a true peak (the first recycle peak 35 as the reference chart) while shifting each coefficient slightly is repeated.

  Since such processing is performed for all the coefficients, differentiation is performed using a determinant. In this way, the direction (ratio of deviation with respect to the coefficient) is obtained by differentiation, and the value of the coefficient is changed according to the deviation ratio. The calculation is repeated until Z becomes close to 0 while repeating this. Actually, if the deviation ratio is used as it is for the next coefficient as it is, oscillation may occur and the ultimate value may not be determined. Therefore, the deviation ratio is somewhat weighted. In other words, here, each coefficient is shifted a little and it is determined by checking which one of the coefficients greatly matches the true peak value and repeating this.

  Thus, although the height of the peak 35 ′ after the deformation can match the height of the non-separable peak 35 as shown in FIG. 3C, a single peak, that is, one of the single peaks The height of the virtual peak 36 and the other single virtual peak 37 cannot coincide with the height of the non-separable peak 35. Therefore, by repeating the second and third calculations, the numerical values of σ1, m1, σ2, and m2 change little by little, and finally, the waveform of the deformed peak 35 ′ as shown in FIG. Can be matched with the waveform of the non-separable peak 35. The tailing waveform can also be obtained by calculating in this way. In this case, the number of waveform calculations can be substantially matched by performing about 50 times. The waveform shown in FIG. 3 is an example, and the directions of all of a1, σ1, m1, a2, σ2, and m2 vary depending on the shape.

  To summarize the above, the above simulation method first obtains vertices and inflection points by performing second order differentiation on unseparated peaks, and determines the number of single peaks and Gaussian from the obtained vertices and inflection points. Find a curve. After that, make a temporary formula that combines each single peak, shift the coefficients slightly so that this formula matches the actual unseparated peak, and check the difference from the unseparated peak at that time. This is done by checking the rate of deviation of each coefficient in the direction in which the difference decreases. Thus, each coefficient is shifted in the direction of decreasing the difference according to the ratio of the shift, and each coefficient is set so that the provisional formula matches the actual unseparated peak until the difference becomes as close to 0 as possible. The process of confirming the difference from the unseparated peak at that time by slightly shifting is repeatedly performed, and finally the value of each coefficient can be obtained.

  FIG. 3D shows an example of the display on the monitor screen 19 after the predetermined simulation is performed in this way, after being transformed into an actual insufficiently separated peak 35 of the first recycling shown by a solid line. FIG. 2B is a diagram showing a state in which one peak 35 ′ and one single virtual peak 36 constituting the peak 35 ′ and the other single virtual peak 37 are overwritten. This is shown as a simulation range a in FIG. It is the chart 44 of the 1st recycling shown.

  That is, when the actual chart 34 of the first recycling is displayed on the monitor screen 19, the peak of insufficient separation of the actual first recycling as shown in FIG. 2B and FIG. 3D. 35 and the deformed peak 35 'are overwritten together with one single virtual peak 36 and the other single virtual peak 37 constituting the peak 35', as well as the second prediction chart 38 'and the recycling The third prediction chart 40 ′ is also displayed on the monitor screen 19 as shown in FIG. 2B, for example, under a display pattern with different broken lines and colors.

  Therefore, when the actual chart 34 for the first recycling is displayed on the monitor screen 19 as shown in FIG. 2B, the actual chart 38 for the second recycling and the actual chart for the third recycling shown in FIG. 2D. 40 can be predicted and displayed as a second prediction chart 38 ′ that is not yet sufficiently separated and a third recycling prediction chart 40 ′ that is sufficiently separated, and as a result, a sufficiently separated virtual peak 41 can be obtained. It can be known in advance that sorting can be performed with reference to the virtual peak vertices 41a 'and 41b' of the prediction chart 40 'for the third recycling shown as'. In addition, the actual chart 38 of the second recycling shown in FIG. 2C is based on the actual chromatogram 41 of the second recycling, and the actual chart of the third recycling shown in FIG. 40 is displayed on the monitor screen 19 in time series as the peak 39 having the peak vertices 39a and 39b and the peak 41 having the peak vertices 41a and 41b, respectively, based on the actual chromatogram 41 of the third recycling. become.

  Since the method of the present invention is configured as described above, the target component can be separated by predicting the number of recyclable cycles in advance and waiting in that time zone. It is possible to eliminate the trouble of having to wait for a long time while gazing at the time when a chart suitable for sorting is displayed while gazing at the displayed chart.

  In addition, since the number of recyclable recycles can be predicted in advance, data necessary for the separation is input in advance to the data processing device 18 even when the target component is separated from different samples. Thus, the sorting process can be automated.

  In addition, since the prediction process enabling the sorting is performed by detecting the peak via the data processing device 18, even if the sensitivity of the detector 17 or the chart speed is changed during the operation, the prediction process is performed retroactively to the start point. Since the contents of the change are reflected, it is possible to add time series. In addition, since a past chart can be simultaneously displayed on the monitor screen 19 provided in the data processing device 18 while recording the operation, it is possible to compare data smoothly.

  In the above, the present invention has been described by taking as an example the case where the peak 35 with insufficient separation as shown in FIG. 2B is displayed on the monitor screen 19 at the first recycling. In addition, an insufficiently separated peak corresponding to the actual chart 38 of the first recycling in FIG. 2B may be displayed on the monitor screen 19, and in this case, a predetermined simulation is performed using this as a reference chart. become.

DESCRIPTION OF SYMBOLS 11 Liquid chromatograph 12 Solvent tank 13 Three-way valve 14 Liquid feed pump 15 Sample introduction apparatus 16 Separation column 17 Detector 18 Data processing apparatus 19 Monitor screen 20 Input means 21 Control box 22 Recycle valve 23 Fraction valve 24 Fraction collector 25 Waste liquid tank 31 Chromatogram 32 Actual chart of the first recycling 33 Peak 33a Peak apex 34 Actual chart of the first recycling (reference chart)
35 peak 35 ′ deformed peak 35a, 35b peak apex 36 one single virtual peak 37 other single virtual peak 38 second recycling second chart 38 ′ second recycling prediction chart 39 peak 39a, 39b peak apex 40 recycling third real chart 40 'recycling third prediction chart 41 peak 41' virtual peak 41a, 41b peak apex 41a ', 41b' virtual peak apex BL baseline _{L 1, L 2, L 3} , L 4 liquid feeding passage P ₁ peak starting point P ₂ peak lodging point S solvent

Claims (2)

In the recycle fractionation method in liquid chromatography,
When the chromatogram obtained from the 0th recycling to the nth recycling is displayed as a chart having insufficiently separated peaks on the monitor screen of the data processing apparatus, the chart is used as a reference chart in relation to the distribution state of the peaks. By performing simulation to obtain a chromatogram in which the peak is separated to the extent that it can be separated, the number of recyclable recycles after the nth time and the elution time zone are predicted in advance, and the recyclable recycles are possible. It is displayed on the monitor screen as a preparable prediction chart determined by the relationship between the number of times and the elution time zone, and the target component is fractionated from the sample in the real time zone corresponding to the predicted elution time zone. Recycle fractionation method in liquid chromatography. In the simulation, when the recycle n-th chart displayed on the monitor screen has a first peak that is insufficiently separated, this is used as a reference chart, and the peak start point surrounded by the reference chart and the baseline The surface area from the peak point to the peak yield point is defined as the simulation range, and the peak of the insufficiently separated peak is differentiated to obtain its vertex and inflection point. After obtaining the number of Gaussian curves and the Gaussian curve, create a temporary formula that combines each single peak and shift each coefficient slightly so that this formula matches the actual peak with insufficient separation. The ratio of the deviation of each coefficient in the direction in which the difference from the insufficiently separated peak becomes smaller is confirmed, and each coefficient is shifted in the direction in which the difference becomes smaller in accordance with the percentage of the deviation. Until it is close to each other, the coefficients are slightly shifted so that the provisional formula matches the actual insufficiently separated peak, and repeatedly performed while confirming the difference from the unseparated peak. The recycling method for liquid chromatography according to claim 1, wherein the recycling value is obtained in advance by determining the value of the coefficient so that the number of recyclable recycles after the nth time and the elution time zone can be predicted in advance.

Images