JP2002303613A - Method and system for high-precision high-pressure gradient - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the gradient elusion accuracy of a high-precision high- pressure gradient liquid feeding system. SOLUTION: A parameter value F is finely adjusted by repeating a process 2 (measurement of pressure values at the time of elusion according to gradient), a process 3 (division of the pressure values into blocks and calculation of the mean pressure of each block), a process 4 (calculation of parameter values corresponding to the mean pressures of the blocks), and a process 5 (execution of a elusion program according to gradient through flow rate control and macro- step control) after a process 1 (execution of elusion according to gradient through flow rate control only), the processes 2, 3, 4, and 5, and a process 6 (execution of the elusion program according to gradient) are successively executed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a liquid sending system in chromatography, and more particularly to a high-precision high-pressure gradient method for a solvent by a micro twin pump suitable for high-precision liquid chromatography and a system thereof.

[0002]

2. Description of the Related Art High performance liquid chromatography (hereinafter, also referred to as HPLC) for separating unknown components uses a solvent (SO
(LVENT: solvent) An analytical sample (sample, hereinafter simply referred to as a sample) held in a sample injection valve is sent from a storage container to a detecting means including a separation column together with a solvent by a precision liquid sending pump, and a detection result is recorded. Or output to the monitor screen.

[0003] In particular, in high-precision chromatography, it is necessary to smoothly send a solvent at a high pressure and gradient.

As a disclosure of this type of prior art, there is, for example, US Pat. No. 5,472,598.

[0005]

FIG. 10 is a schematic diagram for explaining an example of the system configuration of high performance liquid chromatography, in which a solvent 20 stored in a first container (reservoir for solvent) 10 is connected to a pipe 30. After being degassed through the deaerator 40 by pumping through the pump 50, the gas passes through the sample injection valve (autosampler) 60 → the separation column 70 → the detector 80, and the solvent from the detector 80 becomes the second liquid 100 as a waste liquid 100. Discarded in container 90. The arrow in the figure indicates the liquid feeding direction.

[0006] The data detected by the detector 80 is transferred to the data processing device 110 and subjected to required data processing.
It is provided and stored as visible or computer-processable data.

[0007] The separation column 70 is housed in a constant temperature bath 70A to prevent the influence of the external temperature.
The system controller 120 controls the liquid sending pump 50 and the sample injection valve 60.

[0008] The deaerator 40 installed in front of the liquid sending pump 50 for precise liquid sending removes gas dissolved in the solution 20 sucked up from the first container 10 by the liquid sending pump 50, thereby providing a stable solution. It is designed to perform liquid transfer and accurate analysis.

The above-mentioned pump for precision liquid feeding (hereinafter also simply referred to as a pump) may be used alone or in plural units. In particular, in order to perform gradient elution by HPLC, it is necessary to perform an operation of sending liquid while simultaneously changing the flow rate of a plurality of pumps at high pressure.

The flow must be controlled precisely and smoothly by about two orders of magnitude under high pressure and in a situation where the pressure changes. The difficulty is that a single pump can be controlled at a constant pressure and a constant pressure. It can be said that this is a much more difficult technology than a single pump that sends a flow rate.

[0011] Although many analysts understand that gradient elution is good, choosing isocratic elution, which elutes the substance in as simple a single solvent as possible,
Gradient elution is nothing less than the reliability of the elution method compared to isocratic elution.

Although the description of the chemical contents of the gradient elution method is omitted here, the current problems of the gradient elution and a new gradient method using a micro twin pump developed by the present inventors will be described. [Present state and problems of gradient elution method] "Pump accuracy" In HPLC analysis, the elution time (retention time) of a substance is used as an identification parameter, so high repeatability of the pump flow rate used requires high accuracy. Is done.

GPC (gel permeation chromatography) using a single pump can be said to be a case requiring the strictest precision among HPLC analysis methods. The reproducibility of the flow rate here is a retention time of 1000 seconds, the fluctuation is within 3 seconds, and a high accuracy of 0.3% or less is required.

The reason why GPC requires high precision is that the retention time determines the molecular weight of the polymer, and it can be said that it is slightly different from a general HPLC method.

[Accuracy in General Analysis] In GPC, the pump and its peripheral units are sufficiently managed and maintained to maintain highly accurate retention time reproducibility.

On the other hand, the average reproducibility of the retention time in the isocratic analysis using a single solvent in general HPLC is about 0.6%. 10
Since the fluctuation of the peak eluted in minutes (600 seconds) is about 4 seconds, it can be said that the range is sufficient for identification of the eluted peak.

However, in gradient elution, since two or more pumps are used, it is expected that the reproducibility of the retention time is easily reduced by a factor of almost two.

[Problems of the gradient elution method] The gradient elution method has a specific problem which is not present in the isocratic elution method, but the following three factors are considered to cause the reduction in repetition accuracy.

Decrease in pump accuracy at low flow rate Decrease in pump accuracy due to pressure change Disturbance in pump delivery due to air bubbles "Decrease in pump accuracy at low flow rate" FIG. 11 shows a general elution program for two-liquid gradient elution. FIG. 3 is an explanatory diagram of a two-liquid gradient program using a micro twin pump, in which the horizontal axis represents time (TIME: Sec), and the vertical axis represents solvent strength (CONC:%). Curve α is a composition curve.
Note that a micro twin pump is a pump that uses two pumps in parallel.

As shown by the curve α in FIG.
B (solvent A in FIG. 11: solvent A = solution A = first solvent,
Of the solvent B: solvent B = solution B = second solvent), the solution A having the smaller initial flow rate has a higher solvent strength, and the retention time of the eluted component is mainly determined by the amount of the solution A.

The problem is that the liquid A is generated at a low flow rate. The accuracy of the flow rate of the liquid feed pump is not constant over the entire flow rate range, and is relatively lower at low flow rates than at high flow rates.

For example, in a pump having a flow rate accuracy of 0.6% in isocratic elution, a flow rate of 1/10 decreases about 2-3 times at a flow rate of 1/10. Therefore, in a region where the solution A has a low flow rate in the gradient elution, the accuracy is further lowered than expected when two pumps with a precision of 0.6% are simply used.

"Decrease in accuracy of liquid feed pump due to pressure change"
The pulsating flow of the pump depends in principle on the magnitude of the discharge pressure.
This is due to the volume loss due to the compression of the liquid inside the pump. Therefore, when the discharge pressure of the liquid sending pump changes, the pulsating flow also changes.

When a single solvent is flowed, the pulsating flow merely has a fluctuation in pressure, and the retention time of the elution peak is constant as long as the pump has a constant average flow rate. On the other hand, in gradient elution, two liquids (first solvent and second solvent) which are sent by different liquid sending pumps are mixed. Therefore, if there is a pulsating flow in two liquid sending pumps, the composition ratio of the two liquids Will change.

The result leads to a change in the retention time or a distortion of the peak. As a countermeasure against such a pulsating flow problem, current liquid feed pumps often employ a control method based on discharge pressure feedback in order to reduce the pulsating flow. And this pressure feedback method becomes a problem in gradient elution.

FIG. 12 is an explanatory diagram of the relationship between gradient elution and column pressure. As in FIG. 11, the horizontal axis represents time (T
IME: Sec), and the vertical axis indicates the solvent strength (CONC:%). As indicated by the bold line in the figure, one of the causes of the problem with the pressure feedback control method is that the composition of the two liquids changes under the gradient elution conditions, and as a result, the column caused by the change in the viscosity coefficient of the mixed liquid. Some are due to changes in pressure.

Although the column pressure changes due to the change in the viscosity coefficient, the change in the pressure is not similar to the solvent composition program (line α) as shown by the curve β in FIG. It changes complicatedly.

FIG. 13 is an explanatory diagram showing the relationship between the solvent composition and the column pressure for two liquids of water (H ₂ O) and methanol (MeOH). The horizontal axis (lower) shows the water composition ratio (%), The axis (top) is the composition ratio of methanol (%), and the vertical axis is the column pressure (MPa).

For example, when mixing water and methanol,
As shown by a curve γ in FIG. 13, the column pressure becomes maximum when the composition ratio of water and methanol is around 1: 1.
It will be lower elsewhere. The phenomenon is more complicated because the magnitude of the viscosity coefficient varies greatly depending on the type of solvent.

Another problem with column pressure changes is that there is a large system response delay between the solvent composition function due to the elution program and the solvent composition pressure change phenomenon occurring inside the analytical column.

The elution program referred to here means a procedure for controlling the amount of liquid sent by two pumps for sending the first solvent and the second solvent to the analytical column by the respective liquid sending pumps. The concentration of the two liquids is controlled by continuously changing the flow rates of the two pumps over time.

Further, the filler packed in the column is not a simple particle, but includes types such as ODS (a filler having a silica surface having an octadecyl group chemically modified), silica, ion exchange, and the like. For this reason, the change in pressure occurring inside the column further involves the relationship between the mixed solvent and the chemically modified solid in addition to the type of the solvent, which complicates the microscopic pressure change.

For the above-described reason, in the gradient elution system, the change in the discharge pressure of the liquid sending pump is poorly reproducible in a strict sense, the reproducibility is poor, and the conventional pressure value using the poor reproducibility is used for feedback control. It is difficult to obtain high-precision reproducibility with the gradient liquid feeding system described above, and solving this problem has been an issue.

An object of the present invention is to provide a high-precision high-pressure gradient method and a system thereof that solve the above-mentioned problems and improve the precision of the gradient elution method.

[0035]

In the gradient elution method, the composition of the two solutions changes with time, and the pressure in the analytical column changes according to the change. Then, the liquid feed pump performs feedback control based on the changing pressure. Considering that this would reduce the accuracy of the reproducibility of the analysis, the problem could be solved by separating and processing the flow control part that changes the flow rate and the part that performs feedback control according to the pressure. .

Now, assuming that there is no change in the analysis system pressure, it can be said that the whole control is sufficient only by controlling the flow rate. Therefore, the flow rate control part that is not related to the pressure at all is called flow control. This part is a simple control part that simply changes the flow rate with respect to time.

The problem is the control of the part corresponding to the pressure change. Actually, the solution of this part is almost determined by the control method and control system of the two liquid feed pumps used. In conclusion, it can be said that the liquid feed pump using the pressure feedback method is not the optimal pump for the gradient elution method.

Then, "What kind of pump is best?"
It can be said that the applicant of the present application
It can be said that the micro twin pump developed with the subsidy of the creation method in 1975 is a liquid sending pump suitable for gradient elution (see JP-A-2000-39427).

This liquid feed pump is completely separated from flow rate control and pulsating flow control, and has a liquid feed system in which pressure feedback is not used for liquid feed control. The control of the pulsating flow in the liquid sending pump is performed by giving an F value as a pulsating flow correction parameter.

The F value as a parameter is such that when the pressure on the discharge side of the pump becomes high, the pulsating flow of the pump is eliminated by giving the F value corresponding to the discharge pressure to the pump.
At the same time, the flow rate of the pump becomes the flow rate of the set value. The meaning of the F value is the volume loss caused by the compression of the liquid generated inside the pump when the pressure on the discharge side of the pump becomes high. For this reason, the pump will have a pulsating flow when the pressure is high, and at the same time the flow will be lower than the set value.

Since the F value is basically a pressure-dependent amount, the F value is a linear function of the discharge pressure. The slope of the linear function is determined by the type of liquid and the internal structure of the pump.
This gradient is measured from the drop in discharge pressure and flow.

FIG. 14 is an explanatory diagram of pulsation flow correction based on the parameter F value. In the figure, is the pulsation-corrected pump discharge data, “6.4 MPa” is the pressure value of the pulsation-corrected data, and “0.16 MPa” is the Y-axis (pressure value: MP
The scale of a) is a data showing the change of the F value and the pulsation flow when methanol is flown at 1 ml / min.
At 00, the pulsating flow is large and the pressure fluctuation shows about 0.3 Mpa, and when F = 3.5, the pulsating flow becomes small,
Data.

The above F value is a simple linear function of only the pressure, and is a value that does not depend at all on the flow velocity. The point that it does not depend on the flow rate is a big difference from the conventional pressure feedback system.

The pressure feedback control has two disadvantages. The first disadvantage occurs at low flow rates. The pump plunger volume was 100 μl and the flow rate was 100 μl /
min, the operation time of one cycle of the pump is 1 m
in. It takes several cycles for normalizing a slightly large pressure fluctuation by pressure feedback.
This takes several minutes. Thus, the pressure feedback method has a disadvantage that it takes a long time to optimize at a low flow rate.

The second disadvantage occurs at high flow rates. Short-period pressure fluctuations measure pressure at regular sampling intervals,
A method of reducing pressure fluctuation by feedback control is employed. When the flow velocity increases, the pressure change at a constant sampling interval becomes relatively large, and the accuracy of the feedback control decreases.

A typical analysis time in HPLC is 20 to 3
Since the time is about 0 minutes, it can be said that the first problem is a larger problem than the above-mentioned problems.

On the other hand, in the new type liquid feed pump, the optimum operation is instantaneously performed by giving the parameter F value, so that there is no time delay as seen in the feedback control even at a low flow rate. Further, even under the condition of a high flow velocity, the parameter F value is determined only by the pressure without depending on the flow rate, so that the precision of the pressure control does not decrease even at a high flow velocity.

The pressure control using the parameter F value has a great advantage in quantitative terms in addition to the above advantages in control. Since the parameter F value is grasped in all steps of the gradient elution, the pulsatile flow control portion can be considered to be quantitative and reproducible. Although the discussion on this part will be described later, the control method described above completely separates the simple flow control part and the complicated pulsating flow control part of the phenomenon, thereby improving the accuracy of the gradient elution method. Can be improved. "Removal of air bubbles and blocking of response to external pressure fluctuation" The liquid feed pump used in HPLC is quite weak against inhalation of air bubbles, and the flow rate varies greatly due to inhalation of air bubbles. As a countermeasure against air bubbles, a deaerator (hereinafter, also referred to as a degasser) will be used, and a method of removing dissolved oxygen in the solvent will be adopted, but it is not always perfect. Because the amount of oxygen in the solvent removed by the degasser is several p
Since it is on the order of pm, if the degasser sucks air bubbles such as air bubbles for some reason, air particles may directly enter the pump. With two pumps, as in a high-pressure gradient system, the probability of troubles caused by air bubbles is doubled. The solution to this problem was solved with a new pump (micro twin pump) developed by our company.

FIG. 15 is an explanatory view of the suction of air bubbles and the pumping characteristics of the conventional dual plunger type pump and the twin head type pump according to the present invention. The horizontal axis is time (min, described as MIN in FIG. 15), and the vertical axis is pressure. In the figure, (a) is pressure data showing the liquid sending characteristics of a conventional dual plunger type pump, and shows the pressure change when 25 μl and 100 μl of air bubbles are injected from the suction side of the pump with a syringe. (B) is the data on the discharge side showing the liquid feeding characteristics of the micro twin head pump according to the present invention.
This shows that there is no drop in pressure when 100 μl of bubbles are injected with a syringe. This indicates that the micro twin-head type pump does not undergo any fluctuation in the flow rate even when the bubbles are inhaled.

To explain the outline of the pump, in the conventional dual plunger type pump, air bubbles that have entered the pump suction port from the suction line always enter the inside of the pump body. When air bubbles are sucked into the pump body, the flow rate always fluctuates (decreases).

On the other hand, in the pump according to the present invention, the whole pump system is composed of a first pump and a second pump, and a phase separator mechanism for removing dissolved oxygen and bubbles is provided between the two pumps. ing.

The liquid sending capacity of the first pump is set to be larger than that of the second pump. Under such circumstances, even if the first pump sucks the liquid containing air bubbles, it is possible to send liquid exceeding the amount of liquid sent by the second pump because the first pump has a large liquid sending capacity. .

The bubbles and liquid sent by the first pump are separated into gas and liquid by a phase separator, and the gas is discharged to the outside of the pump. The second pump sends only the liquid component inside the phase separator. In such a pump according to the present invention, even if bubbles are sucked in, no fluctuation of the liquid sending amount occurs.

Here, the meaning of FIG. 15 will be described in more detail. Data A shows data of the discharge pressure when a certain amount of air bubbles is injected from the suction line of the conventional pump. The flow rate decreases due to the inhalation of bubbles, and the pressure decreases. It can be seen that the flow rate does not become zero when the amount of air bubbles is 25 μl, but decreases to almost zero when the amount is 100 μl. Data B is pressure data under the same conditions, but shows that there is no decrease in the flow rate due to the influence of bubbles.

By adopting a new pump system called a twin-head type pump (micro twin pump) using a deformed coaxial plunger, a function of degassing the liquid sending pump itself and a function of automatically removing sucked air bubbles are provided. It can be said that the pump system is provided with. In the conventional pump, even if small bubbles are sucked into the pump, the pressure always drops. On the other hand, the new pump can send a liquid without reducing the pressure at all even if a considerably large amount (several tens of μl) of air bubbles are sucked.

Next, the following is a brief description of the liquid-supplying state of the conventional pump when the pump sucks air bubbles in gradient elution.

A and B are two pumps for performing a gradient.
Now, it is assumed that the pump A sucks air bubbles, the flow rate is reduced, and the pressure is reduced. At this time, the pressure of the pressure sensor of the pump B decreases even if the pump B normally sends the liquid. This is because the outlets of the two pumps are connected at a high pressure and connected by a three-way joint, as shown in the embodiment described later. Under such conditions, the pressure feedback control type pump is controlled in response to pressure fluctuations due to external factors even though the pump itself is normal.

The first problem is that the composition ratio of the two liquids deviates from the programmed set value, and the second is that the total flow rate of the two liquids fluctuates. If the pump B can determine that the pressure drop of the pump A is caused by air bubbles, it is better that the pump does not perform the feedback control of the pressure, but the current pump does not have such a determination function.

In the new type pump, since control is performed by parameters, even if bubbles enter the pump A and the pressure drops, the pump B can completely ignore the pump A and send the liquid. As described above, the gradient elution by the new type pump has an extremely small trouble caused by bubbles and is an elution system which is not affected by disturbance pressure. In the "flow rate control and microstep control" gradient elution, the flow rate is changed at fixed time intervals, but the minimum unit of flow rate (flow rate cut) at the time of change is closely related to the gradient elution precision.

In the case of the conventional liquid feed pump, the amount of flow
μl, but the flow rate used in the analysis area is 1 ml
/ Min or less, the error due to the flow incision between the two pumps becomes 20 μl, and the flow rate becomes 1 ml / mi.
If it is n, the set value will fluctuate by 2%, resulting in insufficient control accuracy. Although the fluctuation of 2% is a value with respect to the total flow rate, in the gradient program as shown in FIG. 10, the fluctuation becomes relatively large in a region where the liquid A is small.

In the gradient elution of the two liquids, the liquid A is used as a solvent for determining the solvent strength of the mixed solvent.
When the flow rate is 10 ml at a flow rate of ml / min, the error of the liquid A itself becomes 10%, and the fluctuation of the liquid A in the low flow rate region appears in an enlarged form.

As can be seen from the above, when the flow rate is 10 μl, the flow rate precision of the gradient elution is insufficient. On the other hand, it can be said that the flow rate of the new pump developed by our company is 1 μl, which satisfies the conditions required for gradient elution.

According to the present invention, high-precision high-pressure gradient liquid feeding is executed by pulsating flow control and macro step control.

As described above, in the gradient elution method, the discharge pressure of the column changes due to a change in the solvent composition, and the change is poor in reproducibility. As described above, as long as the conventional pump adopts the pressure feedback control method, it becomes one of the causes of the decrease in the repetition accuracy. In the new pump according to the present invention, the pulsation flow is corrected by a parameter value corresponding to the magnitude of the pressure (the symbol of the parameter value is F value).
Will be performed by using.

In the feedback control of the conventional pump, a method is adopted in which feedback is performed every time the plunger of the pump operates. In contrast, the new pump divides the entire gradient elution time into relatively large blocks. The size of the block is, for example, 3 minutes. The basis of the block division is based on a pressure change of 0.5 Mpa in the block. The average pressure of each divided block is used as a control pressure value. The pulsation is controlled by using the pulsation correction parameter F value corresponding to the macro step pressure value.

With this method, it is possible to avoid the disadvantages of the conventional control, such as the time delay of the control and the response to disturbance pressure fluctuation. This new pulsating flow control method will be referred to as a macro step control method. On the other hand, the conventional method is called a micro-step control method. The process of the new control method according to the present invention will be briefly described as follows.

FIG. 1 is an explanatory view of a high-precision high-pressure gradient liquid feeding method according to the present invention. The liquid sending method of the present invention is executed by the following process.

Process 1: Execution of gradient elution only by flow control Process 2: Measurement of pressure value during gradient elution Process 3: Blocking of pressure value and calculation of average pressure of each block Process 4: Average pressure value of block Process 5: Execution of gradient elution program by flow rate control and macro step control Process 6: Execution of gradient elution program
(Process 2) → (Process 3) → (Process 4) →
(Process 5) is repeated to fine-tune the parameter F value (Fi
ne tuning).

FIG. 16 is an explanatory diagram of the process of the new control method according to the present invention described above. In the figure, similarly to FIG. 12, the horizontal axis represents time (TIME: Sec), and the vertical axis represents solvent strength (CONC:%).

First, when two liquids A and B are passed through a column while being mixed under high pressure by two pumps along a gradient elution program curve α in FIG. 16, a pressure curve β due to the mixed fluid is measured.

The characteristics of the pressure curve β are that the pressure change causes a time lag with respect to the program curve, and that the shape of the curve changes depending on the type of the solvent as already described with reference to FIG. For example, in a water / methanol system, as shown in FIG. 13, the pressure becomes maximum (MAX) near water / methanol (1: 1), and the pressure decreases on both sides. It can be seen that the pressure is twice as different in the case of water alone and the mixed solvent of 1: 1.

Next, gradient elution is performed again using the parameter F value obtained by learning in the above process, the parameter F value is obtained based on the pressure curve obtained again, and the old F value is updated. The optimum parameters can be obtained by learning the gradient program twice. The operation of the operator can be said to be an extremely simple operation by simply executing the gradient program twice in succession. "Size of Error in Macro Step Control" It is assumed that the flow error in the new control method does not exceed the upper limit of the flow error in each block. For example, assuming that the upper limit of the predicted flow rate error of each block is 0.5%, the error of the entire block is 0.5% or less.

The size of the block at the time of blocking is an important factor in determining the flow rate accuracy. Here, the width of the pressure value of the actually measured value in one block section is 5 kg / cm ² or less. Is set to the section width of the block.

The error caused by the flow rate control is considered to be a flow rate error of the pump under a certain condition.

FIG. 17 is an explanatory diagram of the flow rate error in the macro section, and shows the relationship between the pressure at the time of analysis in a certain block and the average pressure obtained by learning as a deviation width between the actual pressure value and the learning value. is there.

In FIG. 17, P (C) is a learning value of pressure (average value of pressure), P (RAV) is an average value of measured pressure values,
(R) is the measured value of the pressure. Note that ER (%) means a deviation of the execution value from the learning value by comparing the average value of the pressure of the block obtained by the learning with the average value of the pressure when the analysis is actually performed. Eventually, a value obtained by converting the deviation of the pressure into the deviation (%) of the flow rate is ER (%).

In FIG. 17, (a) shows a case where the execution value is lower than the learning value. ER = + 0.25%, (b)
Is the case where the learning value and the execution value are equal, ER = 0%, (c)
Is the case where the execution value is larger than the learning value, and ER = −
0.25%.

That is, FIG. 17A shows that the average pressure value during the analysis is 2.5 kg / cm higher than the pressure value obtained by learning.
Low state (P (RAV) <P (C)), (b)
Is the case where the average pressure value during analysis is equal to the pressure value obtained by learning ((P (RAV) = P (C)), and (c) conversely the average pressure value during analysis is obtained by learning Pressure value above 2.
This represents a state in which the learning value is higher by 5 kg / cm ((P (RAV)> P (C)). The difference between the learning value and the actually measured value will be discussed later, but will not exceed the magnitude shown in FIG. Think.

FIG. 18 is an explanatory diagram showing the relationship between the parameter F value and the analytical column pressure. The explanation goes a little sideways, but Figure 1
8, methanol (M
In the data of (EOH), the F value per 1 kg / cm ² is 0.06. The parameter F value is substantially a value corresponding to the compression loss of the liquid inside the pump, and means that a volume increase of 0.06 μl is caused by a pressure increase of 1 kg / cm ² .

When the F value is 0.06, the model of the pump is specified, and if the solvent to be used is determined to be methanol, it becomes a unique value. In the case of water with the same type pump, this value is 0.04.

The volume inside the liquid sending pump according to the present invention is 64
Since the volume is μl, the volume loss of 0.06 μl is about 0.1%. In summary, for a new pump, a pressure increase of 1 kg / cm ² will result in a flow rate change of about 0.1%.

Returning to the original description, the condition shown in FIG.
When the pulsating flow is controlled in accordance with the two-liquid gradient program by the pump disclosed in Japanese Patent No. 39427 (see FIG. 3), the error in the flow rate is + 0.25% because the learned value is 2.5 kg / cm ² larger than the actually measured value. Become. In FIG. 13, on the contrary, -0.25
%, And becomes zero in FIG. As a result, when the analysis was performed under the conditions shown in FIG. 17, the maximum error was 0.5%.
Becomes

In the actual analysis, it is estimated that the error is smaller than 0.5%. Generally speaking, performing a learning operation,
Subsequently, the analysis operation is performed, so that ± 2.5 as shown in FIG.
There is no deviation of Kg / cm ² .
It is about 2 or less.

Since the deviation between the learning value and the measured value does not change in one direction in all blocks, it is considered that the total average error of the blocks is further reduced. Therefore, the error in the pulsating flow control part is estimated to be about 0.2%. Since the error in the flow control part is 0.3%, the overall error is estimated to be about 0.5%.

In the gradient elution according to the present invention based on the fact described above, the flow rate control and the pulsation flow control are separated, so that the respective control units can be grasped quantitatively. The flow rate control part controls the flow rate based on a simple calculation with respect to the set flow rate, and the cumulative flow rate of the entire analysis can be easily obtained by integrating the flow rate.

[0086]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of a high-precision high-pressure gradient liquid feeding system according to the present invention will be described below in detail with reference to the drawings.

FIG. 2 is a configuration diagram of one embodiment of a high-precision, high-pressure two-liquid gradient liquid sending system according to the present invention. In the figure,
1 is a first liquid sending pump (pump A), 2 is a first solvent sent by the first liquid sending pump 1, 2a is a storage container for storing the first solvent 2, 3 is a second liquid sending pump (Pump B), 4 is a second solvent to be fed by the second liquid sending pump 3, and 4a is a storage container for storing the second solvent 3. Reference numeral 5 denotes a gradient controller, which has a program creation unit 5a, a gradient control unit 5b, and a database 5c. Each pump has an input / output port that is controlled to open and close by a gradient controller.

Reference numeral 6 denotes a three-way branch joint, which includes a first solvent 2 and a second liquid pump 3 which are supplied from a first liquid supply pump 1.
The second solvent 4 sent from the above is merged. Reference numeral 7 denotes a mixing coil, which is a first coil joined by the three-way branch joint 6.
The solvent 2 and the second solvent 4 are mixed and supplied to a high-performance liquid chromatography system (HPLC) 8 equipped with an analytical column similar to the high-performance liquid chromatography of FIG.

The gradient controller 5 includes an arithmetic unit,
It is composed of a so-called computer (or personal computer) having a storage unit, an input / output unit, a control unit, and the like. In FIG. According to the prepared elution program (flow rate program), the first liquid sending pump 1 and the second liquid sending pump 1
Only the gradient control unit 5b and the database 5c for controlling the liquid feed pump 3 are shown. The database 5c
Includes a condition setting file for setting various operating conditions, a program file, a data table, and the like. Table 1 shows data contents stored in a condition setting file, a program file, a data table, and the like.

[0090]

[Table 1] Note that in Table 1, The system condition setting file is for setting the conditions of the gradient liquid sending system.

Plunger diameter: The diameter of the plunger of the pump.

A polarity, B polarity: the polarity of the solvent A,
Polarity of solvent B.

Average pressure time: Average pressure for one block time when a cycle is divided into a plurality of blocks.

Pressure sampling time: sampling interval when measuring pressure.

Upper pressure limit: The upper pressure limit at the time of emergency stop of the pump.

2. The execution condition setting file is a file for setting conditions for performing a gradient operation.

Mode (Normal / Learning): There are two types of gradient operations: a normal mode for performing analysis and a learning mode for performing learning. Choose which to choose.

Learning section time: In the learning mode, all sections constituting the program are executed at the same time, which means the section at this time.

Execution cycle number: The number of repetitions when gradient execution is performed in the learning mode and the normal mode.

3. The normal condition setting file is a file for setting which port of the personal computer to use for the RS232C communication of the program for performing the gradient.

Pump A connection port: Communication port between pump A and personal computer.

Pump B connection port: Communication port between pump B and personal computer.

4. The program file is a data file required when creating a gradient program.

P Time: Program time.

T Flow: Total flow rate of solution A and solution B.

Sol A: Solvent A flow rate value.

Sol B: Solvent B flow rate value.

P (ave) (initial value): In learning, the factor value for controlling the pump in the first cycle is calculated from the initial value (Pave).

Data Num: Number of steps in the program. 5. The sorbet data table is a table that stores data required to calculate factors for various solvents.

Offset: The value when the pressure is zero when the factor value is a linear function of the pressure.

Gain: The value of the slope of a straight line when the factor value is a linear function of pressure. 6. The pressure storage tables (A) and (B) are tables for storing data of pressure values obtained by the pumps A and B when the gradient program is executed.

The program file contains set values (P) for the gradient controller 5 to control the present system.
Time: Program time, TFlow: Full scale of flow rate = upper limit, SolA: Flow rate value of solvent A, SolB: Solvent B
, P (ave): Initial value and flow rate set value are stored.

The gradient controller 5 sends the flow rate (liquid flow rate) data “flow rate 1” of the first solvent 2 and the parameter F value data “value” to the first liquid feed pump 1 in accordance with the “program” created by the elution program generator 5a. 1 ”
The flow rate data “flow rate 2” and the parameter F value data “F value 2” of the second solvent 4 are given to the second liquid sending pump 3.

The gradient controller 5b of the gradient controller 5 transmits the liquid pressure data "pressure 1" from the first liquid pump 1 and the liquid pressure data "pressure 2" from the second liquid pump 3. take in.

A part of the solvent not supplied by the first liquid supply pump 1 is stored in the storage container 2a for the first solvent 2 in the second solvent not supplied by the second liquid supply pump 3. Some are second
The solvent 2 is returned to the storage container 4a. The solvent that has washed each of the liquid feed pumps is discharged to a discharge container (not shown).

The first liquid feed pump 1 and the second liquid feed pump 3 are micro twin head pumps and have a structure as disclosed in Japanese Patent Application No. 2000-39427, for example.

Here, a schematic configuration of the micro twin head pump will be described.

FIG. 3 is an explanatory diagram of a configuration example of a micro twin head pump constituting the first liquid supply pump and the second liquid supply pump. This liquid sending pump is composed of two heads operated by a coaxial plunger 51. The plunger 51 is referred to as a coaxial plunger of a different diameter, and a portion having a small diameter at the tip forms a second head, and a step portion between a large diameter portion and a small diameter portion constitutes a plunger of a first head. The plunger 51 reciprocates with a cam mechanism (not shown) as shown by the arrow in the figure.

50 is a main body block, 52 is a sealing material, 5
3A1 is a suction valve of the first head, 53A2 is a suction valve of the second head, 54A1 is a discharge valve of the first head, and 54A2 is a discharge valve of the second head. Reference numeral 3 denotes a thin tube through which the solvent flows, and reference numeral 34 denotes a discharge tube.

Each of the two heads is independently held by a plurality of seal members 52 so that they do not affect each other. The seal member 52 is exposed to the atmospheric pressure except for a portion that hermetically seals with the plunger in order to maintain the inside of the pump at a high pressure. A cleaning line 55 for cleaning the atmospheric pressure side of the sealing material of the second head is formed between the first head and the second head. The solvent used for cleaning is discharged through the discharge pipe 34.

When a liquid such as an organic solvent is caused to flow at a high pressure inside the main body block 50, the liquid slightly leaks out of the sealing material. Since the low-pressure side of the sealing material is normally in contact with the atmosphere, the organic solvent having a low boiling point is vaporized in advance, but the dissolved components are deposited and adhere to the sealing material.
For example, when an aqueous solution containing a salt is sent, moisture leaking from the sealing material evaporates, but the salt becomes crystals and adheres and remains on the sealing material in a powder state. The salt powder remaining in the sealing material penetrates into the sealing material itself, or the sealing material and the movable part of the pump, and damages or corrodes them, causing a pump trouble.

In this embodiment, the above-described pump trouble is prevented by arranging a washing line at this portion and constantly flowing the eluent.

Next, the control operation of the high-precision, high-pressure two-liquid gradient liquid feeding system of the present invention having the above-described configuration will be described in detail with reference to flowcharts.

FIGS. 4 and 5 and FIGS. 6 to 9 are flowcharts for explaining an example of the control operation of the high-precision high-pressure two-liquid gradient liquid supply system according to the present invention. FIGS. 4 and 5 show the overall control flow. 6 to 9 show the flow of a timer interrupt operation in the whole operation. 4 and FIG. 5 are connected by reference numeral A, and the operation flows of FIG. 6 to FIG.

The gradient controller 5 has a timer, and the timer controls the following operations. The function of this timer is to set a numerical value related to the control of the flow rate with respect to time and to calculate a factor value by sampling the pressure value.

First, the flow of the entire control operation will be described with reference to FIG. 4 and FIG. Each step in the operation flow is indicated by (S-xx).

When the high-speed gradient liquid supply starts, the gradient controller 5 takes in various setting values from each file or table of the database 5c (S-1).

The input data is system condition setting data, execution condition setting data, communication condition setting data, elution program, and solvent (solvent) data.

The contents of each data are as follows.

System condition setting data Plunger diameter of pump Polarity of solvent A Polarity of solvent B Time width when calculating average pressure for calculating factor value Pressure value is necessary when calculating average pressure value However, the sampling interval for measuring the pressure value, the minimum time when changing the flow rate by a program, the upper limit value of the pressure when the pump is stopped urgently ・ Execution condition setting data Whether the execution mode is the learning mode or not Mode data in communication mode ・ Communication condition setting data ・ Eluting program ・ Solvent (solvent) data 2), the data transmission / reception mode between each pump and the gradient controller 5 is set to the serial communication mode (S-3).

When a gradient application program is about to run, it is necessary to first set the type of communication inside the personal computer. Here, since the communication of RS232C is performed, the setting is performed. In this setting, the port rate, stop bit length, presence / absence of parity, and data length are set.

Then, the upper limit pressure is taken from the system condition setting file (S-4). The upper limit pressure is a value arbitrarily determined by the operator, and when the pump reaches this pressure value, the pump automatically stops.

Next, a timer interrupt for one second is set (S-5). The timer interrupt control process means that within one second after the program is started, one is to set the parameters related to flow rate control, the second is to take in the pressure value, and the third is a factor to be performed based on it. In the process for calculating the value, after 1 second elapses, 1
The above three operations are repeated every second. The timer interruption time is not limited to one second.

After setting the timer interrupt for one second, the control process is called (S-6). This timer interrupt control process will be described later with reference to FIGS.

The time of a segment of a certain program in the program is inputted from the program file, and the execution of the control (execution of the program) in the time (program segment time) S-6 is within the time (program segment time). Is determined (S-7). If this is within the time (TRUE), the next timer interrupt is awaited (S-8), and the process returns to step (S-6).

In S-7, if the execution of the control in S-6 exceeds the time (FALSE), the maximum number of programs to be repeatedly executed is input from the execution condition setting file, and
It is determined whether the designated number of cycles has been completed (S-
9) When the designated cycle number is completed (FALSE9)
Goes to S-8.

If the number of designated cycles has not ended in S-9 (TRUE), the timer is canceled (S-1).
0), the apparatus of this system is set to the “manual” mode (S-11), and the ports of the first liquid pump 1 and the second liquid pump 3 are closed and the process is terminated (S-12). .

Next, the flow of the timer interrupt control is shown in FIG.
This will be described with reference to FIGS. The calculation of the flow rate in the following flow rate setting is performed by the following equation.

Calculation of flow rate Flow rate A = TFlow × (CONC / 100) Flow rate B = TFlow− (TFlow × (CONC / 100) where TFlow is the full scale (upper limit) of the flow rate CONC is the flow rate value of pump A with respect to TFlow The above setting values are input from the program file.

When a timer interrupt is performed in step (S-5) of FIG. 4, it is determined whether or not this interrupt is the first time (S-21). If this is the first interrupt (TRU
E) outputs the flow rate to the pump A (S-22) and outputs the flow rate to the pump B (S-23). Then
Output parameter F value (initial setting value) to pump A (S
-24), and outputs the parameter F value (initial setting value) to the pump B (S-25).

In FIG. 7, it is determined whether or not the pressure sampling time after the previous pressure acquisition input from the system condition setting file has elapsed (S-26). If the sampling time has elapsed (TRUE), the pump A A pressure value is input (S-27). Step (S-2)
6) If the sampling time has not elapsed (FA
LSE) goes to the end.

The obtained pressure value of the pump A is stored in the pressure storage table (A) (S-28). Similarly, a pressure value is input from the pump B (S-29), and the obtained pressure value of the pump B is stored in the pressure storage table (B) (S-29).
30) End.

The determination in step (S-21) of FIG.
If the LSE, that is, the interrupt is not the first time, it is determined in step (S-31) in FIG. 8 whether or not the flow rate change time has elapsed since the previous flow rate setting. It is determined whether there is a change in the set value (S-32). If there is a change (TRUE), the flow rate is output to the pump A (S-33), and the pump B is output.
Is output (S-33), and step (S-33) in FIG.
Go to 35).

In step (S-31), FALSE
That is, when it is determined that the flow rate change time has not elapsed since the last time the flow rate was set, FALSE in step (S-32).
That is, when it is determined that there is a change in the previously set flow rate set value, the process directly proceeds to step (S-35) in FIG.

In FIG. 9, it is determined whether the average pressure time has elapsed since the previous average pressure calculation (S-35). If it is determined that the time has elapsed (TRUE), the pressure storage table (A) The F value of the pump A is calculated based on the pressure value of the pump A stored in the storage unit (S-36). The calculated F value is output to the pump A (S-37).

The formula for calculating the F value is as follows.

First, the average pressure P (ave) = (P1 + P2 +
P3 + P4 + P5 +... Pn) / n Calculate F value from solvent data table F value = A × P (ave) + B where A is gain (increase in factor value per 1 MPa) B is offset (pressure) Similarly, the F value of the pump B is calculated based on the pressure value of the pump A stored in the pressure storage table (A) (S-38). The calculated F value is output to pump B (S-
39), and end (FIG. 7).

If it is determined in step (S-35) that the average pressure time has not elapsed since the previous calculation of the average pressure, the process ends (FIG. 7).

By performing the above procedure, it is possible to obtain a high-precision high-pressure gradient liquid sending system in which the accuracy of the gradient elution method is improved.

In the gradient elution according to the present embodiment, since the flow rate control and the pulsation flow control are separated, it is possible to quantitatively grasp each control unit. The flow rate control part controls the flow rate based on a simple calculation with respect to the set flow rate, and the cumulative flow rate of the entire analysis can be easily obtained by integrating the flow rate. The cumulative flow rate is an amount calculated without depending on the type of the solvent.

Regarding the quantification of the pulsation control part, the parameter F value macro-stepped in the program is used as a fixed parameter in an analysis program that is repeatedly used under the same conditions. It will be quantitative.

The content of the parameter F value is a numerical value representing the corrected flow rate approximately. Therefore, the integrated value of the F value that has been macrostepped represents the cumulative flow rate used for the pulsating flow control.

Here, the amount of flow control and the amount of pulsation flow control will be briefly described. Now, suppose that the cumulative amount of flow control and the cumulative amount of pulsation control are 100 ml and 5 ml, respectively. At this time, the total amount required for the entire analysis is not 105 ml.

In the flow rate control, although 100 ml of liquid was sent, the actual amount of liquid sent was 95 ml, and the remaining 5 ml was sent.
Should be considered as the amount corrected by the pulsation control. This 5 ml is due to the volume loss that occurred when the fluid was compressed inside the pump. And this amount will be different under different program conditions even with the same solvent combination,
If the combination of the solvents is different, it naturally differs.

As described above, in the new gradient elution method using the high-precision high-pressure gradient solution sending system according to the present invention, all steps of the gradient elution are quantitatively controlled, so that high-accuracy flow rate reproducibility is achieved. Will be.

It should be noted that the present invention is not limited to the embodiments described above, and it is needless to say that various modifications can be made without departing from the technical idea of the present invention.

[0157]

As described above, according to the present invention,
Since all steps of the gradient elution are quantitatively controlled, high precision flow reproducibility is achieved. In particular,
A high-precision high-pressure gradient liquid sending system using a micro twin pump suitable for high-precision liquid chromatography can be provided.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a high-precision high-pressure two-liquid gradient liquid feeding method according to the present invention.

FIG. 2 is a configuration diagram of one embodiment of a high-precision high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 3 is an explanatory diagram of a configuration example of a micro twin head pump used in a high-precision high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 4 is a flowchart illustrating an example of a control operation of the high-precision, high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 5 is a flowchart subsequent to FIG. 4 for explaining an example of a control operation of the high-precision high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 6 is a flowchart subsequent to FIG. 5 for explaining an example of the control operation of the high-precision, high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 7 is a flowchart illustrating an example of a control operation of the high-precision high-pressure two-liquid gradient liquid sending system according to the present invention, which is subsequent to FIG. 6;

FIG. 8 is a flowchart following FIG. 7 for explaining an example of the control operation of the high-precision, high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 9 is a flowchart subsequent to FIG. 8 illustrating an example of a control operation of the high-precision, high-pressure two-liquid gradient liquid sending system according to the present invention.

FIG. 10 is a schematic diagram illustrating a system configuration of high performance liquid chromatography.

FIG. 11 is an illustration of a two-part gradient program with a micro-twin pump showing general flow programming for two-part gradient elution.

FIG. 12 is an explanatory diagram showing the relationship between gradient elution and column pressure.

FIG. 13 shows the relationship between solvent composition and column pressure using water (H ₂ O).
FIG. 3 is an explanatory diagram showing two liquids, methanol and MeOH.

FIG. 14 is an explanatory diagram of pulsation flow correction using a parameter F value.

FIG. 15 is an explanatory diagram of air bubble suction and pump liquid feeding characteristics in a conventional dual plunger pump and a twin head pump according to the present invention.

FIG. 16 is an explanatory diagram of a process of a new control method according to the present invention.

FIG. 17 is an explanatory diagram of a flow rate error in a macro section.

FIG. 18 is an explanatory diagram of a relationship between a parameter F value and a column pressure.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 1st liquid-feeding pump (pump A) 2 1st solvent sent by 1st liquid-feeding pump 1 2a Storage container which stores 1st solvent 2 3 2nd liquid-feeding pump (pump B) 4 2nd The second solvent 4a to be sent by the liquid sending pump 3 of 4a The storage container for storing the second solvent 3 5 The gradient controller 5a The program creation unit 5b The gradient control unit 5c The database 6 The three-way branch joint 7 The mixing coil 8 High-performance liquid chromatography system (HPLC) .

Claims (4)

[Claims] 1. A high-precision high-pressure gradient method for gradient elution of a measurement sample with a solvent sent by a micro-twin pump and sending the sample to high-performance liquid chromatography, wherein the gradient is obtained by only controlling the flow rate of the micro-twin pump. A first process for performing elution; and a second process for measuring a pressure value during the gradient elution.
A third process of dividing the pressure value into a plurality of blocks and calculating an average pressure of each block; a fourth process of calculating a parameter value corresponding to an average pressure value of the block; a flow control and a macro step A fifth process for executing a gradient elution program by control; and a sixth process for finely adjusting parameter values by repeating the second process → the third process → the fourth process → the fifth process after the execution of the gradient elution program. And a high-precision high-pressure gradient method comprising: 2. A first liquid sending pump for sending a first solvent,
A second liquid pump for supplying the second solvent, a three-way branch joint for combining the first solvent and the second solvent supplied by the first liquid pump and the second liquid pump, A mixing coil for mixing the first solvent and the second solvent, wherein the first solvent 2 and the second solvent are mixed by the mixing coil.
A high-performance liquid chromatography having a separation column to which a sample eluted with the solvent 4 is supplied, a gradient elution program generation unit and a gradient control unit for generating a liquid sending procedure, and a database including a condition setting file, a program file, and a data table A high-precision high-pressure gradient system comprising: a gradient system controller comprising a gradient controller comprising: 3. The liquid supply pump according to claim 1, wherein the first liquid supply pump and the second liquid supply pump are micro twin pumps, and wherein the gradient control unit is configured to determine the liquid supply conditions stored in the condition setting file of the database and the first liquid supply pump. The first elution program according to the gradient elution program generated by the elution program generation unit based on the liquid sending pressure values of the liquid sending pump and the second liquid sending pump and a parameter value defining a decrease in flow rate due to a compression loss of the liquid. The high-precision high-pressure gradient system according to claim 2, wherein the liquid supply of the liquid supply pump and the second liquid supply pump is controlled. 4. The gradient elution program generating section includes a first process for executing a gradient elution only by controlling a flow rate of the micro twin pump, and a second process for measuring a pressure value at the time of the gradient elution.
A third process of dividing the pressure value into a plurality of stages determined by parameter values and calculating an average pressure of each block; and a fourth process of calculating a parameter value corresponding to the average pressure value of the block. A fifth process of executing a gradient elution program by the flow rate control and the macro step control; and, after the execution of the gradient elution program, the second process → the third process → the fourth process → the fifth process are repeated to obtain a parameter value. The high-precision high-pressure gradient system according to claim 3, wherein programming is performed to execute a sixth process of fine-tuning.