RU2402099C1 - Method for structural chemical analysis of organic and bioorganic compounds based on mass-spectrometric and kinetic separation of ions of said compounds - Google Patents

Abstract

FIELD: chemistry.

SUBSTANCE: method involves separation of ions in a linear radio-frequency trap with a gas stream along the axis of this trap based on differences in resistance of ions to collision-induced dissociation. The important feature is detection of induced signals from rotating or oscillating ions under the effect of a non-resonance rotating field which brings the analysed ions out of the gas stream, as well as under the effect of fields which are close to resonant fields on frequency, which excite harmonic motion of ions with selected m/z values. The mass spectrum of product ions during collision-induced dissociation can be recorded using a mass analyser which is interfaced with the trap, particularly a time-of-flight mass analyser with orthogonal ion input.

EFFECT: mass-spectrometric separation and obtaining quantitative information on analysed compounds.

21 cl, 10 dwg

Description

FIELD OF THE INVENTION

The present invention relates to methods and techniques for the chemical analysis of organic and bioorganic compounds based on a combination of the separation of the ions of these compounds by mass-to-charge ratios, mobility, resistance to collisional fragmentation of ions, and mass spectrometric analysis of ion products of this fragmentation. In particular, we are talking about the preliminary separation of ions under the combined action of electric fields and gas flows in a linear radio-frequency trap according to the magnitudes of charges and masses, collision cross sections, and decay resistance. Subsequent analysis of product ions by mass-to-charge ratios can be carried out using a time-of-flight mass spectrometer with orthogonal ion input (ortho-VMS) or on some other mass analyzer or on the basis of registration of induced signals from rotating or oscillating ions in the same radio frequency quadrupole.

The decay or death of ions can be caused both by heating of rotating ions due to their collisions with atoms or gas molecules, and by collisions of ions with electrode surfaces. The use of such decays or death for the separation and identification of the analyzed compounds is one of the distinguishing features of the present invention, it greatly increases the separation ability of the method.

The registration of the induced signal from rotating or oscillating ions before mass analysis of product ions, similar to how it was done in an ion-cyclotron resonance mass spectrometer (ICR) with fast Fourier transform, is another important feature of the present invention. It allows one to obtain additional information about the studied ions in the form of estimates of the cross sections for collisions with atoms or molecules of a buffer gas and the energies of bond cleavage during collisional dissociation. The total time required for measurements when solving structural-analytical problems is also reduced, and the dynamic range of the measured intensities of ion fluxes is expanded, which is a weak point of ortho-VPMS, especially when using time-to-digital conversion as a registration method.

Among the tasks for which, in addition to sensitivity, both separation ability and the dynamic range of measurements are important, express analysis of microimpurities in the air can be mentioned with respect to use in security systems, customs and environmental control. The second task, where separation ability and “information performance” are decisive, is the study of the structure of biomolecule ions, which may be important for proteomics, biomedical and biotechnological applications.

BACKGROUND

After the development and creation at our institute of the first time-of-flight mass spectrometers with orthogonal ion injection (ortho-VPMS) [1, 2], devices of this type were widely used both in solving analytical problems and in studying the structure of biomolecules [3-5]. The convenience of combining such devices with various devices for preliminary separation of ions, producing a continuous or quasicontinuous ion flow, with pulsed time-of-flight mass analysis, which is record-breaking in speed among all known types of mass analyzers, has made these combinations highly effective and attractive for solving a variety of analytical and structural problems . At the same time, there are important structural and analytical problems for which the separation ability and “information performance” of known instrument systems, including the ortho-VPMS, is insufficient. To overcome these limitations, it is natural to strive to introduce into the mass spectrometric experiment additional measurement dimensions associated with the controlled transformations of the studied ions and the recording of data during these transformations. To carry out such measurements, it is desirable to have a sufficiently large number (or sufficiently intense flux) of the studied ions in the reactor and to separate these ions or the signal from them from other interfering ions.

In the past 10-15 years, gas-filled radio frequency multipoles, devices containing a set of rods usually parallel to each other and symmetrically located around the axis of the device, have become widespread in mass spectrometry. Radio frequency voltages are most often in antiphase applied to adjacent rods. These devices are usually used as a means of focusing and efficiently transporting ions or for accumulating ions (in this case they are called linear radio frequency traps or linear ion traps) with the possible isolation of selected ions and conducting controlled dissociation and other structural transformations [6-8]. In these devices, the property of high-frequency force fields described in Mechanics of Landau and Lifshitz [9] is used to cause the particles to be ejected in such fields to reduce the strength of these fields. More specifically, the average particle motion in such (electric) fields is described, to a first approximation, by the effective potential, which is directly proportional to the square of the high-frequency field intensity multiplied by the particle’s charge and inversely proportional to its mass. For a special case of an ideal radio-frequency quadrupole, the effective potential depends quadratically on both coordinates (in a rectangular coordinate system), reaching a minimum value on the axis of the quadrupole, and the average free movement of ions in such a field is independent harmonic oscillations in both coordinates. As indicated there [9], the idea of such an effect of high-frequency fields belongs to P.L. Kapitza, and it was expressed by him in 1951. It took about 40 years before this idea found practical application in such effective means of manipulating ions as gas-filled radio frequency quadrupoles or multipoles. In these devices, used as ion storage rings and reactors, two important properties in this case - the ability to accumulate ions and the ability to separate these ions can conflict with each other. In order to effectively stop ions inside a multipole, a sufficiently high gas density is needed, and for high selectivity of isolation of selected ions or excitation of resonant oscillations of ions and their heating (for fragmentation and other transformations), the gas density should be relatively low.

One of the objectives of the present invention is to overcome this contradiction. The second goal is to create effective methods for carrying out sequential kinetic measurements of the transformations of several types of selected ions present in incompletely separated mixtures, to develop methods for analyzing experimental data for additional separation of the studied ions on this basis and to obtain quantitative information on the structure, thermochemistry and kinetic properties of these ions.

One of the most important prerequisites for the present invention is the creation by us of a technique for resonant excitation of rotation of selected ions around the axis of a radio frequency quadrupole and the fragmentation of these ions due to collisions with buffer gas molecules [10-12]. This technique was new, previously not proposed by anyone. In contrast to the present invention, the rotation of the ions in this case is excited during their movement along the quadrupole without prior accumulation. This limits the possibilities for kinetic measurements and provides a limited ability to detun from signals of interfering ions. In addition, this method of resonant rotation imposes very severe restrictions on the manufacturing quality of the quadrupole. Small deviations in the diameter of the rods or in the distances between them leads to significant losses in the resolving power of the method, which turned out to be no more than 100 during actual measurements in our case.

In the proposed embodiment, ions rotate in a relatively narrow zone (several mm), while they perform quasi-chaotic oscillations along this zone with an average transit time of this zone comparable to the rotation period and much shorter than the measurement time. Thus, the field inhomogeneities are significantly averaged, and their influence on the width of the resonance curves is weakened. In this case, the resolution of the resonant excitation for given ions and a given buffer gas will be mainly determined by the density of this gas in the region of rotation. At a reasonable residual pressure of 0.1 mTorr for helium, the expected mass resolution at half maximum peaks for organic ions with a mass of about 1000 D will be about or even more than 5000.

One of the difficulties in conducting kinetic studies in the realized case was some uncertainty in the actual speed of rotation of the ions and in the associated internal temperature of the decaying ions. This was a consequence of the fact that ion mobility measurements were carried out at electric field strengths lower than those at which fragmentation processes begin. It was believed that these mobilities remain the same in large fields. This assumption is not always true. In the proposed method, this difficulty is overcome by measuring the phase shift of the induced signal from rotating and decaying ions from the rotating field. This shift is determined by the deviation of the frequency of the rotating field from the resonant frequency and the relaxation time of the ion velocity, which is uniquely related to the mobility of these ions.

We developed a calculation model and conducted experiments on the formation of a supersonic gas stream at relatively low buffer gas pressures with ion transfer by this stream and reflection of ions from surfaces coated with a thin charged dielectric film [13-15]. These methods were also new, unknown in the literature.

A gas-dynamic interface has been created for the ortho-VPMS at our disposal with a gas flow former and a sectioned radio-frequency quadrupole. This interface configuration is new. In general terms, it is described in our US patent No. 7547878 of June 16, 2009 [16]. Unlike the present invention, there is no “locking” diaphragm in the quadrupole, and instead of a quasihomogeneous field inside the quadrupole, a parabolic potential distribution is created with a minimum located near the beginning of the quadrupole. Preliminary experiments have demonstrated the operability of this equipment, however, an effective capture of impurity ions in a helium buffer gas into such a trap has not yet been possible. The excitation of non-resonant rotation, as proposed in the present invention, the ions are removed from the area near the axis of the radio-frequency linear trap, which will deprive them of the possibility of a return exit through the input diaphragm of the trap. This should be guaranteed to ensure effective capture of ions in the trap if measures are taken (as is done in the present invention) to prevent the death of ions on the inner surface of the inlet diaphragm. The possible replacement of the buffer gas with a heavier one and the introduction of a locking diaphragm will reduce the time of thermalization of ions in the trap.

A theoretical model for heating ions moving in a gas under the influence of an electric field was described by us in [17]. The model is original, its predictions are somewhat different from the known models. In particular, it predicts a slightly higher value of the internal temperature of the ion compared to the temperature of its translational motion. This model is confirmed by the available experimental data, and its use to obtain quantitative thermochemical data for the studied ions is possible.

The software for the analysis of experimental data should include software packages that basically implement the original methods developed by us, described in [18–22, 32]. Among these methods, the most important are the following.

1. A correction method for the effects of saturation and “dead” time when using time-to-digital conversion to register data from the IMSS [20, 32].

2. A method for identifying exponential contributions to a decaying induced signal from decaying ions [18, p. 192] with finding the roots of the characteristic polynomial using the procedure described in [22].

3. A method for identifying exponential contributions to a set of ion current curves, developed earlier for the analysis of a set of effusiometric curves [21].

Existing methods for the implementation of collisional dissociation of ions or for kinetic mass spectrometric measurements usually involve preliminary isolation of one type of ion with the loss of all the others, thereby requiring the use of a large volume of the initial sample and a large time-consuming experiment. One of the exceptions is the “multi-reflective” ortho-VPMS by A.N. sufficiently far apart in exit time (for a time longer than the ion drift time in the secondary time-of-flight mass spectrometer). This approach is much more technically complex than in our case, of course, eliminates any kinetic measurements and produces primary ions for dissociation only in m / z. He also envisages, in contrast to the proposed method, the possibility of further collisional investigation of the ion-products of primary dissociation.

A possible approach that reduces primary ion loss is described in A.V. Loboda's Application No. 20070120053 for a US patent [24]. This application proposes, after the accumulation of ions in the quadrupole at a buffer gas pressure of about 0.1 Torr, to carry out dipole excitation of ion vibrations with the selected m / z, so that these ions in the plane of the dipole excitation deviate far enough from the quadrupole axis on average. During such excitation or after its completion, a quadrupole field constant in time is created that varies linearly along the quadrupole. The potentials of this field are chosen so that in the plane of excitation of the vibrations of the selected ions, an average electric field is generated that moves the ions to the exit of the quadrupole (on the axis of the quadrupole this field is 0, and in the perpendicular plane it moves the ions in the opposite direction). In this case, unexcited ions having, on average, a smaller deviation from the axis of the quadrupole in this plane, will be less affected by this pulling field. Thus, the packet of ions of interest can be moved into the collision chamber, and the remaining ions will remain in the storage quadrupole. After completing the first packet in the same way, the next packet can be delivered to the collision chamber. This approach is quite interesting and, apparently, will work. However, its resolution should be low enough (it is unlikely to be more than 10) for several reasons. The main one is the rather high density of the buffer gas needed to trap ions in the trap. Thus, ion packets transferred to the collision chamber will contain many ions in a rather wide mass range, and for the collisional dissociation of “individual” ions, all other ions from this packet must be removed. The relatively high gas density in the radio frequency multipole during the accumulation of ions in existing systems either leads to low selectivity of the ions during their isolation, or requires additional time to download the "excess" gas. Another possible solution is to create complex multi-tambour systems, where the functions of accumulation, isolation, and collisional dissociation are performed in different parts of the system with very different buffer gas densities. Such a design leads to additional ion losses and an increase in the cost of the instrument complex. It is such a construction that is proposed in the just described US patent application [24].

Dynamic methods of capturing ions in a quadrupole trap, when the reverse ion exit is blocked by switching on the corresponding potential until the launched ion packet returns from the pivot point, only a small part of the initial ion flux can be used if subsequent manipulations with the ions require a relatively long time. The initial ion flow should be locked at this time, and the corresponding ions are usually lost.

Fourier transform ion cyclotron resonance mass spectrometry (ICR FP) currently provides the most accurate measurement of the ratio of ion mass to charge (m / z) with a resolution of more than 100,000. In ICR FP, ions are either introduced externally into the cell or are created inside the cell and held in it by a combination of static magnetic and electric fields (Penning trap). Static magnetic and electric fields determine the cyclotron frequency of motion, which is dependent on m / z. This motion is excited by an oscillating electric field. After a short application time, the field turns off. The amplification and registration of weak stresses induced on the plate of a cell by the movement of ions allows us to determine the frequency of movement of ions and, thus, m / z ions. Ions are selectively isolated or fragmented by changing the magnitude and frequency of the applied transverse radio-frequency electric field and the residual gas pressure. Repeated sequences of ion isolation and fragmentation (MS ⁿ -operations) can be performed in a single cell. MS ICR FP is a "large" instrument complex that occupies a large area and is also quite expensive, mainly due to the cost of creating a magnetic field using a superconducting solenoid and ultrahigh vacuum. In addition, the MS ICR FP demonstrates weak ion retention in the MS ⁿ operation (compared to a three-dimensional ion trap). High resolution requires an extremely high vacuum (below 10 ^-9 Torr).

In addition to ICR mass spectrometry, another application of ion rotation has been proposed in traps of a constant electric field, similar to the Orbitrap instrument (described by Makarov in 1999, US Patent 5886346 [25]). Orbitrap has an important advantage over the ICR method due to the absence of a magnetic field. Both methods can provide extremely high resolution for small and medium ions. For large ions, the resolution decreases for two main reasons: lower ion rotation frequencies (or oscillations in Orbitrap) for large m / z and a decrease in the coherent motion of ions due to the increased probability of collisions with atoms or molecules of the residual gas due to large collision cross sections. In addition, Orbitrap has an additional restriction for introducing large ions into the trap of a constant electric field, since these ions can have a rather high probability of collision with atoms or molecules of residual gas in the storage quadrupole during their acceleration to a relatively high energy (about 1 keV). Thus, in the process of introducing these large ions into Orbitrap, a substantial part may be lost due to dissociation or have less energy than optimal during the capture of an ion in a trap of a constant electric field. These devices, like ICR devices, require a sufficiently long measurement time (up to several seconds) and a little longer for the preliminary accumulation of ions.

Registration of the induced signal during the free movement of ions in the radio-frequency quadrupole is described in US patent 6784421, M.A. Park [26]. The difference between our version of such registration consists in the use in our case of a predetermined type of ion motion in a quadrupole in the form of a superposition of two ion rotations: non-resonant, removing ions from a relatively dense region of the gas flow, and resonant or intrinsic rotation, the signal from which is recorded. In addition, in our case, it is supposed to organize the registration of purely forced rotations or ion oscillations, including in the presence of frequencies of a rotating or oscillating field close to the resonant frequency of the ions. It is in this case that important additional information is expected to be obtained about the studied ions. To reduce the level of induced signals from a rotating or oscillating field, in our case it is proposed to use the measurement of the potential difference between the electrodes located between the rods of the quadrupole so that the induced signals from the rods are close, the signal from the ions on the electrodes located closer to the center of the quadrupole would be prevailing. In the patent in question, it is proposed, after the accumulation of ions, which is carried out at a relatively high pressure of the buffer gas, to download most of the gas. Then, the intrinsic motions of the accumulated ions in the radio-frequency quadrupole are excited by a broadband voltage pulse applied in the quadrupole mode to the rods of the quadrupole, and then the induced signal from their free movement is recorded. Such an approach increases the time during measurements, requires a longer interruption of the ion flux, and, consequently, leads to their additional losses. In addition, the increased pressure of the buffer gas in our case would create difficulties for the operation of the electron ionization source, which would require differential pumping of the ion source region and a separate interrupted gas input into the quadrupole volume. This would lead to additional losses of the analyzed ions, inevitable when they are introduced into the quadrupole against the gas flow. The ion storage method of the present invention is devoid of these disadvantages.

The use of resonant rotational motion of ions, as well as their resonant one-dimensional vibrations in a radio frequency quadrupole, to eliminate unnecessary ions that interfere with the measurement of less intense analytical ions or cause saturation phenomena in the measuring system of a time-of-flight mass spectrometer, is described in US patent application No. 20080149825 to Kozlovsky B. AND. et al. [27]. In our case, similar goals can be achieved by appropriate resonant spin-up of nonresonantly rotating ions in the storage part of the radio-frequency quadrupole, which will increase the selectivity of this elimination in the case of a noticeable gas flow along the axis of the quadrupole. The use for this purpose of the death of untwisted ions on the conducting part of the output diaphragm will make it possible to get rid not only of ions with the selected m / z, but also to selectively attenuate the population of accumulated ions, which significantly differ in mobility or charge size. Periodic excitation of free additional (to non-resonant) rotation of the ions and registration of the induced signals will make it possible in our case to carry out these operations in a timely manner and prevent undesirable effects associated with excessive accumulation of certain ions.

The use of a rotational field for the selective dissociation of ions accumulated in a quadrupole linear trap in collisions with atoms or molecules of a buffer gas is described in US Pat. No. 7,351,965 B2 [28]. It is proposed that the registration of product ions, as well as the removal of unwanted ions, be made through gaps along the vertices of the main electrodes of a hyperbolic shape. It is proposed to compensate for violations of the quadrupole field near these slots with the help of thin electrodes located along the middle at the exit of these slots. During dissociation, it is proposed to intentionally distort the quadrupole field by setting potentials on these auxiliary electrodes that are different from the potentials of the main electrodes. This is useful for shifting the resonance frequencies of highly unwound ions to prevent their death on the quadrupole electrodes. In our case, the use of round rods of a quadrupole (which is technologically much easier to use hyperbolic rods) will lead to the same effect in the dissociation of selected ions. When registering the induced signals, there is no need to force the ions to move very close to the surfaces of the rods of the quadrupole, so that the deviation of the field from the ideally quadrupole in this case is less important than the registration of ions proposed in this patent. The patent proposes to capture ions dynamically in the trap, raising the voltage at the input diaphragm, because the buffer gas pressure in the quadrupole is not enough to stop the ions reflected from the blocking potential in the last section of the quadrupole. In this case, ions remain in the trap, which have reflected from this potential and did not have time to go back through the output diaphragm of the quadrupole until a blocking voltage is established on it. To ensure the capture of a sufficiently large number of analyzed ions, it is assumed that a relatively long quadrupole (1000 mm) is used. This length not only increases the dimensions of the device, but also imposes more stringent requirements on the parallelism of the rods of the quadrupole and other conditions for its manufacture to ensure uniformity of the resonant frequencies of free movements of ions in different places of the quadrupole. The ion accumulation method of the present invention is expected to allow accumulation of a sufficient number of ions in a quadrupole, an order of magnitude shorter, with a comparable residual pressure of the buffer gas.

Apparently, the closest method in its capabilities, which can be considered as a prototype of the invention, is claimed in US patent No. 7507953 [29]. In general terms, this is the development of the approach described in the previous paragraph, if we do not take into account the compensation of violations of the quadrupole field near the output slits along the rods of the linear quadrupole trap for ion transport to the registration system. In the case under consideration, the selected ions in the form of a “ribbon” beam can also enter a flat collision chamber and then be transported directly or after collisional dissociation to a time-of-flight mass spectrometer with orthogonal ion injection. The possibility of autonomous registration (as a separate system, as in the previous patent) of the initial ions is also provided. In this method, in contrast to generally accepted approaches, the use for collisional dissociation is achieved not of one but of any desired number of species of parent ions.

The same opportunity is realized in the present invention. The standard use of a linear quadrupole trap for sequential isolation and collisional ion dissociation for the implementation of the MS ⁿ method, including at the last stage the registration of product ions using a time-of-flight mass spectrometer, is also envisaged. The same possibility can be realized in the present invention.

In the future, it is planned to use the device described in the patent in combination with a liquid chromatograph for preliminary separation of complex samples of biological origin. In this case, this approach, in comparison with other known ones, seems to be the most effective in terms of the amount of information received, the analysis time and the amount of sample used.

The disadvantages of the described system include the loss of resolution and sensitivity during the orthogonal introduction of parent ions into the collision cell, including due to deviations of the field near the exit gap from the quadrupole one. In addition, the use of dipole excitation of ions, in contrast to our approach, where rotational fields are mainly used, leads to a stronger effect of the space charge of the remaining accumulated ions on the resolution of the release of parent ions. In our case, these ions are dispersed in different orbits of nonresonant rotation (depending on m / z and ion mobility). In the case of resonant dipole excitation, the remaining ions are concentrated near the axis of the quadrupole, and with a sufficiently large number they create a noticeable additional field that distorts the harmonic nature of the effective potential of the quadrupole field near the axis of the quadrupole.

To achieve a more efficient use of the source ion stream in one of the options proposed in the patent, it is provided that the linear quadrupole trap is divided into two parts. In the first part, operating at elevated pressure, all ions are accumulated, preferably continuously. In the second part, which performs the resonant selection of the parent ions under reduced pressure (for better selectivity), the ions in the selected m / z interval are translated by excitation of longitudinal vibrations in the first part to communicate enough energy to overcome the potential barrier between the first and second parts. In our case, rather efficient accumulation of ions, the separation of parent ions and their fragmentation is supposed to be carried out in a single quadrupole pumped out by one turbomolecular pump.

In this patent, to control the number of accumulated ions, it is supposed to carry out preliminary experiments to determine the rate of accumulation of ions with given values of m / z, using the described methods for recording selected ions, or by measuring the total ion current accumulation rate of all ions. For such measurements, it is necessary to excite the vibrations of the corresponding ions in the plane of the exit slits to transport them to the registration system and the collision chamber. In our case, the measurement of the induced signal from rotating ions allows such control over the accumulation of ions in the process of accumulation without the need for preliminary experiments and without noticeable loss of accumulated ions.

In the patent, kinetic measurements are not provided and it is hardly possible in the described construction. On the other hand, such measurements as proposed in the present invention, in combination with preliminary chromatographic separation due to time limitations, are possible only for individual chromatographic peaks separated from each other by sufficient time intervals. At the same time, these measurements can provide additional ion separation, which can compensate for the absence or even exceed the separation capabilities of the chromatograph with a shorter total analysis time. A significant increase in the resulting separation is also possible when combined with a chromatograph, because resistance to collisional dissociation is unlikely to strongly correlate with retention times of compounds in chromatographic columns.

The advantage of recording the induced signal from the forcedly rotating ions, as proposed in the present invention and is absent in the discussed and other analogs, is to obtain additional information about the studied ions. In particular, these can be estimates of collision cross sections with atoms or molecules of a buffer gas and bond break energies leading to the formation of ion products of collisional dissociation.

A standard method for estimating the cross sections for collisions of ions moving in a gas is one or another type of measurement of the mobility of an ion or the proportionality coefficient between the stationary velocity of an ion and the electric field strength that causes this movement. Often this movement is used for preliminary separation of ions. Since in ordinary versions of the method, the time of movement of ions in a drift tube is relatively small, the most appropriate combination of separation of ions by mobility with a time-of-flight analyzer with orthogonal ion input.

A serious problem of this combination is the provision of high ion transmission through the drift tube in the HPMC. One of the possible solutions was proposed by us in US patent No. 6992284 [31], which provides a fairly detailed overview of the work on the separation of ions by mobility. Patent 6,992,284 refers to the use of a sequence of alternating sections of a strong and a weak field instead of a uniform electric field in a drift tube at a gas pressure of several Torr. This leads to the focusing of ions to the axis of the quadrupole and allows you to slightly increase the total voltage along the tube, which favorably affects the resolution of ion packets in mobility. Nevertheless, in all realized variants of the separation of ions by mobility, a sufficiently high resolution cannot be obtained. Even for ion drift at atmospheric pressure, a resolution of more than 100 is not achieved.

In the present invention, the mobility of ions in their separation acts indirectly. The greater the mobility of ions with a given m / z, the greater the speed of movement of the ions will be under the action of a rotating field and, therefore, a higher internal temperature is proportional to the square of the speed. If the resistance to decay of the ions under consideration is close, then ions with greater mobility will decay faster. Moreover, in our case, by choosing the frequency of the rotating field, it is possible to select m / z ions, and by choosing the amplitude of the rotating field, you can adjust the decay rate. In the case of the classical separation of ions by mobility, there is a rather weak possibility of influencing the exit time and the width of the packet of ions of interest. By decreasing the field strength along the pipe, the ion exit time can be increased, but the mobility resolution decreases. In our case, the measurement of the mobility is achieved without the presence of a drift tube and without the need for a high DC voltage.

SUMMARY OF THE INVENTION

The features of one of the possible implementations of the proposed methods are as follows.

At a buffer gas pressure (it can be helium, argon, nitrogen or other gases) several Torr at the inlet and less than mTorra at the exit from the cylindrical channel, a narrowly directed gas flow is formed. A relatively small admixture of the analyzed sample in the buffer gas stream is transported in the form of a focused molecular beam to the ion source of electron ionization.

The second possibility of forming a flow of the studied ions is the transportation of ions, including multiply charged ions of biomolecules, from an electrospray source with their focusing by a system of aperture diaphragms to the input of the gas flow formation channel. To minimize the loss of ions inside the channel, it is heated - this is one of the standard methods for transferring ions from an electrospray source to the region of relatively low pressure at the entrance to the mass spectrometer. Instead of an electrospray source, this ion injection can also use any other ion source operating at atmospheric or relatively high pressure.

Ionization of gases in a molecular beam is carried out by a source of electron ionization with a variable ionization energy, allowing direct ionization of the analyzed compounds. In the presence of multiply charged ions of biomolecules in the flow, for example, from an electrospray source, capture of slow electrons with subsequent dissociation processes is possible.

Accumulation, preliminary separation, controlled fragmentation and focusing of ions are carried out in a sectioned radio-frequency quadrupole, the axis of which coincides with the axis of the gas stream. At the beginning of the quadrupole, in the region limited by the “locking” diaphragm, a relatively weak quasihomogeneous electric field is created by setting the corresponding voltages on the sections of the rods of the quadrupole. The field between the input diaphragm and the first section of the quadrupole may have a slightly higher intensity to provide better transmission of ions through the input diaphragm. The inner surfaces of the diaphragms are covered with a thin (not more than 1 μm) dielectric film charged with buffer gas ions of the same sign as the analyzed ions. At a sufficient density of such charges, as we showed earlier [13, 14], ions with a relatively small kinetic energy (units of eV) will be reflected from such a surface. In the case of positive ions, buffer gas ions are obtained at increased energy of ionizing electrons in an ion source with electron ionization. These ions in the considered region of the quadrupole are unwound by a rotating field, the frequency of which is shifted towards low frequencies from the resonant frequency of the buffer gas ions. The shift of the frequency from the resonance is needed to, on the one hand, ensure the promotion of ions in the region of relatively high gas density near the axis of the quadrupole, and, on the other hand, to prevent the death of ions on the rods of the quadrupole with too large radii of rotation. Getting on the surface of the “locking” diaphragm, the ions charge the dielectric film on it to a potential that prevents further charging, causing the reflection of ions. A part of the reflected ions reaches the inner surface of the input diaphragm and charges a dielectric film on it, thereby forming a trap for rotating ions with radii larger than the radii of the openings of the output and input diaphragms. In this case, the longitudinal electric field and gas flow will ensure the transport of ions located near the axis of the quadrupole to the output part of the quadrupole and then to the time-of-flight mass analyzer.

Possessing larger masses than the buffer gas ions, the analyzed ions are effectively focused by the quadrupole at high radio frequency voltages, at which the buffer gas ions can lose stability in the quadrupole and die on its rods. This may be important, since otherwise the accumulation of a large space charge of buffer gas ions can lead to the death of the analyzed ions. Alternatively, by lowering the energy of ionizing electrons, the formation of buffer gas ions can be stopped. By the formation of the corresponding non-resonant rotating field (with a frequency shifted from the minimum resonant frequency of the analyzed ions to lower frequencies), the ions of a given range of mass to charge ratios begin to rotate and are held further in stable rotation orbits, if possible inside the flow and outside the openings of the input and output diaphragms. If necessary, the voltage at the output (“locking”) diaphragm can be increased to create a potential barrier to the accumulated ions.

Induced by the movement of ions, changes in the potentials of the recording electrodes inside the quadrupole are recorded and are used to determine the presence of the corresponding rotating ions and a rough estimate of their number. The beginning of the recording signal being locked up at the main frequency of the forced rotation of the ions will be an indication of the noticeable death of the accumulated ions due to repulsion by the space charge, indicating the unproductive further accumulation of ions. By applying rotational or dipole voltages with frequencies close to resonant, it is possible to remove ions among the accumulated ones that are not of interest for analysis. This will slightly increase the number of “useful” ions that are trapped as much as possible.

When a sufficient number of rotating ions is reached, or after a predetermined accumulation time has passed, the amplitude of the quadrupole's rotational voltage increases to the optimum value for holding and resonant spinning with possible fragmentation of the analyzed ions. In this case, the voltage of the input diaphragm increases in order to stop the entry of the analyzed ions into the quadrupole. If during the accumulation the voltage of the locking diaphragm was increased, then it decreases to ensure the free exit of ions moving near the axis of the quadrupole. By including a package of rotating fields for a given range of m / z for a short time, the accumulated ions are unwound into suitable orbits, and the induced signal from them is recorded during the lifetime of a significant signal at frequencies in the interval of the master field packet. After the inverse Fourier transform, the mass spectrum of the accumulated ions is obtained. Computer simulations show that such “free” rotation of ions at a residual pressure of a buffer gas (helium) of about 0.1 mTorr occurs within about ten milliseconds and allows one to obtain a mass spectrum with a resolution of about 1000 at half maximum peaks. In order to obtain a similar resolution for heavy buffer gases (more suitable for conducting stock ion dissociation), the residual pressure should be reduced approximately in proportion to the square root of the molecular weight of these gases. At the same time, during modeling, it was noted that the phenomenon of merging or coalescence of freely rotating sufficiently dense clouds of ions with close m / z similar to the same effect in ICR mass spectrometry [34] significantly reduces the dynamic range of such registration.

Alternatively, the accumulated ions can be subjected to mass analysis on ortho-VMS or other devices. In this case, with a decrease in the rotational voltage or an increase in the focusing radio frequency voltage, ions with lower m / z and lower mobility will gradually enter the zone close to the axis of the quadrupole. Under the influence of the longitudinal field and gas flow, they will pass through the opening of the “locking” diaphragm and ultimately end up in the ortho-VMS or another mass analyzer. When registering peaks of ions of interest, the rotational voltage can be slightly increased to stop the release of these ions. These ions can be subjected to additional resonant spinning, so that as a result of their heating during collisions with molecules or atoms of a buffer gas, processes of their fragmentation begin to occur. In this case, product ions with lower m / z values than the initial ions will be less susceptible to both spins (total non-resonant and resonant for parent ions) and will be recorded in the ortho-VMS or in other subsequent mass analyzers.

Under stationary conditions of heating of individual ions (in the absence of a noticeable loss of ions due to repulsion by a space charge), exponentially decaying signals of product ions should be observed, according to the characteristic decay time of the signals and amplitude ratios of which the corresponding decay rate constants can be determined. In the presence of several types of decaying ions with a given value of m / z, it becomes possible, based on the analysis of the peaks of all products, to identify the number of such types and to determine the decay constants for each channel of all these types of ions. The characteristic times of incidence of the recorded peaks will be determined by the total decay constants of the initial ions, which may make it possible to separate the corresponding signals for different initial ions, although their values of m / z and mobilities may coincide. Captures of some product ions are possible (especially with m / z larger than m / z of the starting ions) into stationary orbits. With the inclusion of the corresponding resonant rotating fields or a package of such fields, it becomes possible to register them without the use of ortho-VPMS and other mass analyzers. After their accumulation, these ions can also be detected by ortho-VPMS or other mass analyzers, as well as the initial ions, and their further fragmentation can be organized, including by the method described above for the initial ions.

Registration of the induced signal from forcedly rotating and decaying ions can also allow one to determine, in favorable cases, the characteristic times of death of several types of ions. In addition, it is also possible to determine the phase shifts (with respect to the rotating field) of the induced signals from the rotation of each type of ion and thereby estimate the mobility of these ions, if the resonant frequencies of rotation of these ions are known. If these frequencies are not exactly known, then it is possible to spin the ions not by one but by a pair of frequencies close to the resonance. In this case, a pair of recorded phase shifts at these frequencies will allow us to find estimates of both the resonant frequencies and the mobilities of the ions.

Knowledge of ion mobilities, amplitudes of rotational stresses and phases of rotation of ions will make it possible to evaluate the internal temperatures of ions. A change in the amplitudes of the rotational stresses together with a change in the amplitude of the radio frequency field (in order to increase the speed and maintain the average orbit of the ion motion) will make it possible to carry out kinetic measurements of the processes of ion dissociation under conditions of an inhomogeneous density of the buffer gas in the quadrupole and to evaluate the activation energies of these processes, which may be close in favorable cases to the energies of broken bonds. With increasing amplitudes of the rotational stresses, it is also possible to include the decay processes of more stable ions and the elimination of relatively easily decaying ions from the “game”. Thus, all ions whose resonant frequencies are close to the frequencies of the rotating fields can be analyzed.

Since ions often differ quite significantly in their degree of resistance to decay, adequate estimates of the resonance frequencies and mobilities of ions are possible even when these values are quite close and could not be resolved by standard methods of mass analysis and separation of ions by mobility. This situation is similar to preliminary chromatographic separation of the analyzed compounds or separation based on electrophoresis. The difference is that in the present invention there is the possibility of choosing for kinetic study of ions in a relatively narrow range of m / z due to the selection of frequencies of the corresponding rotating fields.

The theoretical foundations of the method of separating exponential contributions in relation to the analysis of sets of effusiometric curves were developed by us earlier [21]. The essence of the approach is based on a complete analogy between systems of differential and finite-difference linear equations with constant coefficients. This analogy for one equation, for example, was used by us in the development of linear prediction methods for the effective estimation of the mass numbers of ions from their exit times for a magnetic static mass spectrometer [18, 33]. Having determined the coefficients of such a forecast, in particular, which is accurate for the sum of the exponential curves, the damping factors of the corresponding exponentials can be determined by finding the roots of the characteristic polynomial, as well as for the corresponding differential equation with constant coefficients. In this case, the computational procedure described by us in [22] can be used.

Based on this analogy, for a linear combination of exponentially decaying flows of product ions in the general case, the total intensity vector of these ions for subsequent registration without taking into account measurement errors can be expressed as the product of a certain transition matrix by a vector of such intensities for the previous registration. The matrix elements of this transition matrix can be found by minimizing the error of such an approximation, which reduces to solving the corresponding systems of linear algebraic equations for each row of this matrix. The eigenvectors of this matrix describe, up to normalization, the intensity vector of product ions of each type of ion present in the mixture being analyzed. The eigenvalues corresponding to these vectors characterize the exponential decay of the number of such ions. The coefficients describing the contributions of the eigenvectors of the transition matrix to the observed intensities of the product ions are found by the least squares method to ensure, on average, the minimum errors of the measured intensities of the mass spectrum peaks for all recordings. The quality of approximation compared with measurement errors is a criterion for the accuracy of the analysis. Knowing the number of types of the initial ions and the characteristic times of their exponential decay, one can find the contributions of each of these exponents to all samples in the recorded mass spectra by solving the corresponding systems of linear equations. Thereby, mass spectra of product ions of each type of starting ion will be obtained, as if they were completely separated before their registration. This is important because will allow to determine the “exact” masses of all product ions if the corresponding mass spectra were recorded on a high-resolution mass analyzer.

Alternatively, the death and decay of ions can be arranged upon their collision with the surface of electrodes not covered by a charged dielectric film. In particular, for this, a conductive flat ring adjacent to the “locking” diaphragm can be used, with an external diameter matching the external diameter of the diaphragm. In this case, the frequency of collisions of sufficiently strongly promoted ions on the surface of this ring will be controlled by the depth of the potential well created by the retarding potential of this ring and the longitudinal field of the quadrupole, which moves the ions to the locking diaphragm. A significant contribution to the shape of the potential well for rotating ions will be made by the relatively large non-resonant rotating field and the effective potential of the radio frequency field, the intensity of which near the ring on the locking diaphragm can be significantly greater than far away from it.

The effective ion temperature, which is the second factor for the collision frequency, will be determined for these ions and the buffer gas by the rotation speed of these ions. A small part of the ions, depending on their charge, whose energy is sufficient to overcome the difference in potential energies between the bottom of the well and the ring, will collide with the ring and die as a result of such a collision with a probability apparently close to 1. If this is not so, and the probability of death depends on the nature of the ion, this is an even more interesting case, which allows one to obtain additional ion separation. For multiply charged ions of biomolecules, during capture of an electron leaving the ring, dissociation processes are possible, leading to the formation of product ions with m / z leading them out of resonance rotation. Moreover, the differences in the frequency of all these processes can be many times higher than the differences in the effective temperature of the ions (up to 10 or more times at reasonable measurement times), because ions will collide with the surface of the ring at the edge of the Maxwell energy distribution of ions. If the efficiency of ion death in collisions with the surface of the rods is close for ions of mass and mobility, then the characteristic times of decrease in the number of rotating ions will be determined by these frequencies of collisions with the surface, which will ultimately be determined by the mobility of these ions. If the difference of 10% in the collision frequencies is sufficient for reliable separation of signals, then this means that differences in the mobility of the initial ions at the level of 1% may well be enough to separate the signals from the corresponding ions. It is possible to combine both of the described methods for organizing the death and decay of ions, when one of them is not enough for reliable separation of signals from close ions.

In the case where the probability of the death of an ion in a collision with a ring is close to 1, an estimate of their effective temperature can be obtained from the rate of decrease of such ions due to such collisions. The contribution of the death of ions on the ring to the total death during collision fragmentation can be found by varying the delay potential of the ring and determining the characteristic decay times of the signal from these ions. An estimate of the effective temperature of ions, obtained in this case not through a preliminary assessment of the mobility and average velocity of ions, is important for verifying the model used for heating ions moving in a gas [17] and, on this basis, to obtain adequate quantitative estimates of the activation energies of the corresponding processes of collision ion fragmentation. When studying the ion mixture in the manner described on the previous page, carrying out such an analysis at various amplitudes of the resonant rotating field can provide an additional criterion for the adequacy of the results. This can be indicated by the coincident number of components found for different rotating fields and the close to Arrhenius character of changing the formation constants of the corresponding product ions under the assumption that the internal temperature of the ions depends quadratically on the amplitude of the resonant rotating field [17]. After conducting an ion study for a given m / z, a further decrease in the total non-resonant rotational voltage can be continued, and thus all the ions of interest among the accumulated ones are investigated. Unlike many other existing methods, there is no loss in each measurement cycle of other accumulated ions other than those selected in this cycle. By increasing the radius of the non-resonant rotation of the starting ions, if necessary, it is possible to ensure the conservation of certain product ions, not very different in m / z from the starting ions, in stable orbits of non-resonant rotation around the axis of the quadrupole. In this case, after the accumulation of a sufficient amount, they can undergo collisional dissociation and study the kinetics of this dissociation along with the parent ions.

BRIEF DESCRIPTION OF ILLUSTRATIONS

For a more complete understanding of the present invention, the following description is related to the corresponding drawings, in which:

Figure 1. The general scheme of the gas-dynamic interface of ortho-VPMS for preliminary separation and registration of ions.

Figure 2. The calculated values of the minimum potential for reflection of ions from a flat surface of a conductor with a thin dielectric film (100 nm) and the minimum distances between charges on it (for a square network of charges): 1- (10 nm), 2- (20 nm), 3- (30 nm).

Figure 3. The calculated resonance curves for rotating ions with m / z = 400 and 401 inside and outside the argon flow for densities corresponding to 30 and 0.3 mTorr at room temperature.

Figure 4. Illustration of the possible mechanism of kinetic energy loss of ions colliding with a surface covered by a charged dielectric film.

Figure 5. The cross-sectional diagram (AA, Figure 1) of the ion rotation region.

6. Flowchart of the procedure for calculating the exponential contributions to the kinetic curve of decaying ions

7. Illustration for the method of separation of exponentially decaying superimposed peaks of ion-products of collisional fragmentation of the initial isobaric ions.

Fig. 8. Illustration for the processes of the death of rotating ions in collisions with the surface of the conducting ring at the edge of the “locking” diaphragm.

Fig.9. Screen shot of a program for simulating the movement of ions in a gas-filled radio-frequency quadrupole to obtain a mass spectrum of freely rotating ions without taking into account the effect of space charge

Figure 10. Screen shot of a program for simulating the motion of ions, taking into account the effect of a space charge, demonstrating the phenomenon of merging (coalescence) of ionic clouds with close m / z during their free rotation

All these drawings are explanatory in nature and do not impose any restrictions on the possible implementation of the invention.

DETAILED DESCRIPTION OF THE INVENTION

A new approach for transporting ions from the high-pressure region at the exit of the mobility cell to the vacuum part of the mass spectrometer through the formation of a supersonic gas flow is described in our US patent No. 7482582 dated January 27, 2009 [14]. It was further developed to provide additional analytical capabilities due to the resonant excitation of ion rotation around a supersonic flow in our next US patent No. 7547878 of June 16, 2009 [16]. The specific development of these approaches to ensure effective quantitative determination of the presence of impurities in gas mixtures and structural analysis of multiply charged ions of biomolecules is described in the present invention.

A feature of the proposed method, in contrast to the mentioned patents, is not the separation of ions from a helium stream with ions from a mobility cell, but the intentional formation of a gas stream, schematically shown in Figure 1, from a buffer gas, preferably heavier than helium (1) entering cylindrical channel (2). A stream of atmospheric air or other gas mixture containing microimpurities and main components directly or after enrichment, for example, by adsorption-desorption on a suitable carrier, is introduced into the stream (1) from the mixing chamber (3) or alternatively, ions of organic or bioorganic compounds from an electrospray source or other sources are introduced into the channel (2) together with the stream (1) by a system of focusing electrodes not shown in FIG. 1. In order to prevent the adsorption of ions of biomolecules on the inner wall of the channel (2), it is heated by current through the winding (4). The pumping of gases entering the stream (1) is carried out mainly by a pump (15). An insignificant part of the flow through the inlet diaphragm (17) enters the next stage of differential pumping of the ortho-VPMS or other mass analyzers.

A molecular beam (5) containing (during analytical measurements) a small admixture of atmospheric air or other gas mixture or ions of organic or bioorganic compounds enters the electron ionization source (6) with a variable energy of ionizing electrons. To reduce the spread in electron energy, the source is equipped with a cathode (18), for example, of LaB ₆ with indirect heating (22). By choosing the appropriate electron energy and the composition of the gas mixture, various ionization modes can be realized, as a result of which either the formation of positive (or negative) ions of the analyzed compounds or the conversion of ions coming from other sources occurs. So, for example, when the energy of electrons exceeds the ionization threshold of atoms or molecules of a buffer gas (7), the main primary ions will be ions of this gas, the main component in the beam (5). Since the ionization potentials of organic compounds are usually less than the ionization potential of atoms or molecules of a buffer gas, such as Ar or N ₂ , then when the corresponding molecules collide with the ions of the buffer gas, charge transfer to these molecules will occur, and a significant fraction of these molecules will be ionized.

Before carrying out analytical measurements and further, as the need arises, charging is performed on the inner surfaces of the input (8) and locking (14) diaphragms coated with a thin dielectric film (9). For this, a buffer gas stream (5) without analyte impurities passing through an ion source (6) at an electron energy sufficient to ionize the buffer gas can be used. The ions of the buffer gas at the entrance to the partitioned radio-frequency quadrupole (10) are accelerated by the longitudinal field (13) created by the corresponding distribution of the potentials of the quadrupole sections (10) and the diaphragm potentials (8) and (14). To ensure better transmission of ions through the diaphragm (8), the field strength between it and the first section of the quadrupole (10) is slightly increased in comparison with the quasihomogeneous field inside the quadrupole. The high-frequency potentials of the quadrupole rods shifted in phase by π / 2 (10), with the angular frequency ω _rot , many times lower than the resonant frequency for the buffer gas ions, which can be calculated by the formula:

a rotating field is created (11). Under the action of this field, ions (12) with a mass to charge ratio m / z (with frequencies substantially higher ω _rot ) will rotate with an average radius:

For buffer gas ions, this radius should be large enough to prevent the passage of the “locking” diaphragm (14) through the central hole, but not sufficient for the death of these ions on the quadrupole rods (10) and on the surface of the conducting ring (19). In formulas (40) and (42), V _rf , ω, V _rot , ω _rot are the amplitudes and frequencies of the radio frequency and rotating fields, e is the elementary charge, r ₀ is the radius of the quadrupole - the minimum distance from the axis of the quadrupole to its rods. In this case, the buffer gas ions will charge the remaining accessible surface of the diaphragm (14), which is covered with a thin dielectric film (9).

After the buffer gas ions charge the dielectric film on the surface of the diaphragm (14) to a sufficient level, they will be reflected from this surface, and at least some of them will reach the reverse surface of the input diaphragm of the quadrupole (8) and charge the film on its surface (9). This will occur until the charge density reaches the maximum possible value under the given conditions. After this, ions will also be reflected from this diaphragm. Figure 2 shows the results of calculations of the potentials of the fields created by such charges. The calculated values of the minimum potential of reflection of ions from a flat surface of a conductor (35) with a thin dielectric film (36), a thickness of 100 nm and the minimum distance between charges on it (for a square grid of charges) are presented: 10 nm - (31), 20 nm - ( 32), 30 nm - (33). The calculation was made for a sufficiently large set of dipoles shown in the upper right part of the drawing. The potential calculation points were located along a straight line starting from the surface of the dielectric (36) in the shown location (34) in the middle between the charges and orthogonal to this surface. Positive charges of dipoles with a given step between them were placed inside a square centered at the shown point (34). The length of the side of the square was doubled until the result of the subsequent calculation did not coincide with the previous one with a given accuracy (for example, 10 ^-6 ). Since the contribution of each dipole to the calculated value of the potential is positive (the positive charge of any of these dipoles is closer to the calculation point than negative), the calculated potential curves are lower estimates of the real potential distributions for flat surfaces larger than the calculated square with the same dipole distribution. For simplicity of calculations, the dielectric constant of the dielectric (16) was set equal to 1, which can only underestimate the calculated value of the potential, compared with the expected value for large values of the dielectric constant.

After the capture of ions by the surfaces of both diaphragms (14) and (8) ceases, ions under the action of a weak field (13) will accumulate mainly in front of the surface of the diaphragm (14), forming a cloud (16). These ions will cause the appearance of charges of the opposite sign on the conductive surface of the diaphragm (14) in order to maintain its potential created by the corresponding power source and voltage divider, not shown in Fig. 1. According to the well-known “image” method, this is equivalent, from the point of view of calculating the field to the left of the diaphragm, to the presence of a virtual cloud of negative charges like (16) (21) symmetrically on the other side of the diaphragm surface. The presence of this cloud will attract real ions (16) to the surface of the diaphragm (14) and reduce the repulsive effect of the space charge (16). This should lead to the possibility of accumulation of a significantly larger number of ions in the quadrupole than it would be in the absence of a diaphragm (14).

The sectioning of the quadrupole rods (10) and to the right of the diaphragm (14) makes it possible to optimally create a radio-frequency and longitudinal electric field in this part of the quadrupole to ensure efficient registration of the analyzed ions (20) passing through the diaphragm (14) in the ortho-VMS or in other subsequent mass analyzers. To ensure optimal conditions for the passage of these ions through the diaphragm (14), controlled constant voltages can be applied to the rods of the quadrupole (10), which shift the ion beam in a plane perpendicular to the axis of the quadrupole. The process of charging the surface of the diaphragm (14) coated with a film (9) and the accumulation of an ion cloud (16) can be controlled by measuring the current of charges entering the diaphragm (14) through an appropriate system for measuring small slowly varying currents. The termination of such a current will mean the end of the process of charging the surface of the diaphragms (14) and (8) and the creation of a stationary rotating cloud of buffer gas ions near the diaphragm (14). Alternatively, the charging process can be monitored by the induced signal from the rotating ions (16), which after some growth should reach a stationary level. The organization of such a measurement channel will be described below.

After charging of the surfaces of the diaphragms (14) and (8) is completed, the radio frequency voltage of the quadrupole (10) increases so as to remove the buffer gas ions from the accumulation zone (causing loss of stability of their movement and death on the quadrupole electrodes) and ensure, if possible, close to optimal conditions of accumulation and research of the analyzed ions. For this, the amplitude of the radio frequency voltage must be high enough so that the focusing of the studied ions allows them to be kept in stable orbits of nonresonant rotation around the axis of the quadrupole without noticeable transmission of these ions through the diaphragm (14) and without significant loss of these ions on the rods of the quadrupole (10) and on the surface a conductive ring (19). In addition, the resonant frequencies of rotation of the studied ions and possible linear velocities of this rotation should, if necessary, cause collisional fragmentation of the studied ions with characteristic times convenient for recording series of the corresponding mass spectra on the ortho-VPMS or in other subsequent mass analyzers (in the range of a second up to tens of seconds).

Preliminary registration of the analyzed ions passing through the diaphragm (14) in the absence of additional exciting fields will provide information on the possible presence of ions of interest in the analyzed mixture and the m / z ranges of ions to be studied in detail. To accumulate the studied ions, a rotational voltage is applied to the rods of the quadrupole with a frequency significantly shifted towards low frequencies from the resonant frequencies of the studied ions. The amplitude of this voltage is large enough to unwind the ions in the region of relatively high gas density inside the flow, so that the ions could hardly pass through the holes of the diaphragms (8) and (14). At the same time, in the region of residual gas density, the influence of this rotating field should not be enough to significantly unwind the ions and lose them in decays or on the rods of the quadrupole.

That this can be done is illustrated in FIG. 3. Resonance curves are shown here for the relative average ion spin radii, based on formula (41), taken from our work [10]:

The curves are given for hypothetical ions with m / z 400 and 401 for the expected flux density in the ion spin region (48), (47) and 100 times lower residual gas density (50), (49). The widths of these curves at half maximum also differ 100 times, as predicted by formula (44):

The characteristic relaxation time of the ion velocity ^ was calculated by the formula (43):

proceeding from the mass of buffer gas atoms M = 40 D, the average thermal velocity V of their motion (ion velocities were neglected) at room temperature and a reasonable value of the cross section for collisions of ions with gas atoms σ = 200 Å ² . The gas density n was calculated from the equation of state of an ideal gas at pressures of 30 and 0.3 mTorr for the upper wide (47), (48) and lower narrow (49), (50) curves, respectively. It can be seen that with a sufficiently large deviation from the “resonance” point inside the stream, the average radius of rotation of the ions does not deviate significantly from the maximum value. At the same time, the resonance curves at a residual pressure of (49), (50) fall off quite sharply, and the spin radius with a sufficient deviation from the resonance even at relatively high rotational stresses can be quite acceptable. For large deviations from resonance, the rotation radius (42) is practically independent of the relaxation time of the ion velocity (43) or the gas density. A change in the gas density leads to a change in the phase shift (45) between the direction of the rotating field and the direction of the average velocity of the rotating ions.

The lower the density of the gas, the closer to π / 2 this shift (if the resonant frequency is greater than the rotation frequency) and the smaller the component of this field that rotates the ions, the linear velocity of which is determined by their mobility:

As the analyzed ions approach the surface of a thin dielectric film on a conductive surface charged with ions of the same sign, not only does the reflection of ions from such a surface occur, but some loss of their kinetic energy is also possible. 4 illustrates this feature. Approaching the surface (55), the ion (51) with kinetic energy higher than the thermal one, with its electric field (52), shifts the ions (53) located near the recesses on the surface (55) a little further (54) than the average thermal energy and attraction (57) to their images (56) behind the surface of the conductor. These ions on the film, having given part of their energy to the atomic-molecular structures of the film, are no longer able to return to the ion (51) all the energy that was received from it when it approached the surface (55). Therefore, the ion will bounce off the surface (55), but will have less energy than when approaching it. Thus, even at a very low pressure of the buffer gas in the zone of non-resonant rotation of the analyzed ions, they will lose their energy obtained in field (13), and they will accumulate like ions of the buffer gas (16) with the highest density near the diaphragm (14) - Figure 1.

To control the number of accumulated rotating ions, registration of induced voltage differences across thin electrodes, for example, (113) and (114) shown in FIG. 1 and FIG. 5, can be used. With the symmetrical position of these electrodes relative to the nearest rods (101) of the quadrupole, the induced signal from the rotating field (103) will be largely compensated and, with careful adjustment, reduced to a value comparable to the signal from rotating ions. This signal at the frequency of the rotating field with acceptable stability of the generator of the rotating field will be almost constant in amplitude. The signal from the accumulated ions will initially increase almost linearly with time, and when a significant space charge density is reached, the signal will begin to saturate due to the death of the ions caused by their repulsion. The presence of such saturation will signal the end of the accumulation process. The measurement of the phase shift of the registered variable signal from the signal of the rotating field will carry qualitative information on the composition of the accumulated ions - formula (45).

и

с частотами ω₁, ω₂ и амплитудами

,

, создающих два вращающих поля, направление первого, нерезонансного (108) в некоторый момент времени показано на чертеже. Условная граница зоны прохода ионов через диафрагму (14) показана пунктирной линией (104). Пунктирной линией с большим шагом (107) показана граница, после которой ионы начинают сталкиваться с поверхностями стержней квадруполя. Зона между двумя этими пунктирными линиями - это область возможного нерезонансного вращения ионов (105), (106), (109) без значительной вероятности их потери либо на поверхностях стержней, либо путем переноса в орто-ВПМС или в другие последующие масс-анализаторы. Усредненная траектория резонансного вращения (111) выбранных ионов вокруг центра (110), вращающегося вокруг оси квадруполя, показана точками. Эти ионы, нагреваясь за счет столкновений с атомами или молекулами буферного газа, с присущим им при заданных условиях характерным временем будут распадаться по различным каналам, давая соответствующий набор фрагментов, который при их регистрации может быть использован для идентификации известных соединений или для установления структуры новых или неизвестных соединений.By the formation of a resonant rotating electric field corresponding to the selected m / z, starting from the ions of interest with the smallest m / z accumulated, additional rotation of the ions of these compounds around their position in the packet of non-resonantly rotating ions is excited. Figure 5 schematically shows a cross section of a quadrupole in the region of rotation of the accumulated ions. Inside the image of the rods (101) of the quadrupole, expressions are given for the stresses applied to them. This is a focusing radio frequency voltage (102) ± V _rf cosωt with frequency ω and amplitude V _rf and two rotational voltages (103)

and

with frequencies ω ₁ , ω ₂ and amplitudes

,

creating two rotating fields, the direction of the first, non-resonant (108) at some point in time is shown in the drawing. The conditional boundary of the zone of passage of ions through the diaphragm (14) is shown by the dashed line (104). The dashed line with a large step (107) shows the boundary after which the ions begin to collide with the surfaces of the rods of the quadrupole. The zone between these two dashed lines is the region of possible nonresonant rotation of ions (105), (106), (109) without a significant probability of their loss either on the surfaces of the rods, or by transfer to the ortho-VPMS or other subsequent mass analyzers. The average trajectory of the resonant rotation (111) of the selected ions around the center (110), rotating around the axis of the quadrupole, is shown by points. These ions, heating up due to collisions with atoms or molecules of a buffer gas, with a characteristic time characteristic of them under given conditions, will decay through different channels, giving an appropriate set of fragments that, when registered, can be used to identify known compounds or to establish the structure of new or unknown compounds.

By changing the total ion velocity, one can change the characteristic ion fragmentation time and the ratio of the intensities of the fluxes of ion fragments. The ion velocity can be changed both due to the radii of non-resonant and resonant rotations, and due to the frequency of rotations, which in the case of resonant rotation is determined by the focusing properties of the radio frequency field of the quadrupole (formula (40)). These properties can be changed by selecting the appropriate amplitude V _rf or angular frequency and radio frequency voltage. As follows from our [17] and other models of heating ions in a gas under the influence of an electric field while maintaining the mean square of the ion velocity and maintaining other experimental conditions, the internal temperature of the ions does not change, and the decay rate and the ratio between the ion fragmentation channels do not change. If, under this condition, the radii of non-resonant and resonant rotation are changed so that the (111) path of ions passes near the center of the quadrupole, as shown in Fig. 5, then the maximum relative ion residence time in zone (104) of possible passage through the diaphragm (14) will be in the case of a sufficiently large radius of the trajectory (111) when it passes exactly through the center of the quadrupole. If we associate with this the maximum value of the measured flux of the corresponding ions on the ortho-VPMS or in other subsequent mass analyzers, it becomes possible to estimate the radius of the resonant rotation of the selected ions and, at a known frequency of this rotation, the corresponding linear velocity. This is due to the fact that in this case the radii of the resonant (111) and non-resonant (124) rotations coincide, and the radius of the non-resonant rotation with a sufficiently large shift from the resonance is almost exactly calculated regardless of the mobility of the ion and the density of the buffer gas [10] - formula (42 )

Moving near the surface of introduced electrodes (112), (113), (122), (123), ions (106), (105) and (109) cause the appearance of an induced charge of the opposite sign on these electrodes. The position and shape of these electrodes should be such as to ensure, on the one hand, the maximum possible level of induced signals, and, on the other hand, distortions of the quadrupole field that are inevitable when introduced, should not significantly distort the movement of the analyzed ions and lead to significant losses in resolution and sensitivity. The calculations of the induced signals and the simulation of the motion of ions using the Simion v.7 package for calculating electric fields showed that cylindrical electrodes with a diameter of 0.2 mm located symmetrically with respect to the nearest quadrupole rods with axes at a distance of the radius of the quadrupole r ₀ from the axis have acceptable characteristics quadrupole. The paired electrodes (113), (114), (117), (116), located symmetrically with respect to the planes passing through the axis of the rods (101) of the quadrupole, are significantly less affected by the ion field, while the induced voltages from the fields of the round rods of the quadrupole will be almost the same for the electrodes of the corresponding pairs. This allows, similarly to ICR spectrometry, to organize registration of rotating ions by measuring the signal between the electrodes of the corresponding pairs. Moreover, in this case, such registration can be carried out through four channels - for each pair of electrodes (112) - (115), (113) - (114), (122) - (117) and (123) - (116). By simply summing the amplitudes of the signals over all four channels, the signal-to-noise ratio can be doubled. More complex processing of signals from freely rotating ions, including the fast Fourier transform with allowance for phase shifts of the resulting complex amplitudes, can lead to a four-fold increase in the resolution of the resulting spectrum. A similar technique was proposed by E. N. Nikolayev and J. Franzen for ICR spectra during signal registration using a multi-electrode cell [30]. The resonant frequencies of the signals from the ions will be substantially less than the radio frequency ω of the voltage focusing the ions (102) and substantially greater than the frequency of the forced rotation (108) around the gas stream. Using amplifiers with an appropriate bandwidth connected between the electrodes (112) - (115), (113) - (114), (122) - (117) and (123) - (116), the high-frequency and low-frequency components of the recorded signal can be significantly weakened .

To determine the values of the resonant frequencies and amplitudes of the rotational stresses for the corresponding target compounds, preliminary experiments should be carried out with these compounds added to the stream in sufficient quantities. In addition to establishing the conditions for the motion of the corresponding ions in stable orbits, in the absence of their noticeable decrease due to decays and transfer to the ortho-VPMS or other subsequent mass analyzers, both during the accumulation of ions and their subsequent analysis, the optimal amplitudes of the radio frequency voltage, frequency and amplitudes of rotational stresses for observing the decay processes of rotating ions due to their heating in collisions with atoms or molecules of residual gases. If the decrease of these ions occurs only due to the processes of monomolecular decay (as well as any independent death processes described by a certain constant probability of death), then it will be exponentially decaying in time with a characteristic time specific for given conditions for each compound. If this time is significantly longer than the period of rotation of the initial ions, then the phases of rotation at which the decays occur will be practically uniformly distributed throughout the entire circle of rotation. If the resonant frequencies of rotation of the product ions noticeably differ from the frequency of rotation of the starting ions, the signal from these ions will be averaged to a large extent and will be close to 0. Thus, the change in time of the signal at the speed of rotation of the starting ions should follow the process of decreasing these ions due to their decays. On this basis, it is possible to measure the characteristic decay times of these ions and use them to identify the target compounds even in the presence of other ions having the same m / z value, but differing in the values of the characteristic decay times under given conditions of ion rotation.

As you know, the sum of the exponential curves satisfies the finite-difference equation with constant coefficients, or in other words, there is an accurate linear forecast of the subsequent value of such curves from the previous ones, measured with some (constant in this case) time step. Having determined the coefficients of such a forecast, for example, by the least squares method, the damping factors of the corresponding exponentials can be determined by finding the roots of the characteristic polynomial, as for the corresponding differential equation with constant coefficients. In this case, the computational procedure described by us in [22] can be used, the modified block diagram of which is shown in Fig. 6. The initial data of this procedure are the number and values of the linear prediction coefficients. To determine this number in this case, you can use the natural separation of the observed data at the frequency of rotation of the initial ions. Firstly, these are the curves proportional to the sinusoidal signal from the rotational voltage itself, and the curves shifted in phase from this signal by π / 2, which are caused only by rotating ions. Secondly, it can be data taken for different pairs of corresponding electrodes, for example, (112) - (115) and (113) -114). By calculating the Fourier coefficients for these two sinusoids (cosine and sine transforms) for some set of time intervals, we obtain data sets for forecasting. Consistently increasing the number of coefficients of such a forecast, we find them in the first array, requiring a minimum of forecast errors. Using the found coefficients, we find the forecast error for the second array. We leave such a number of forecast coefficients N, which provides a minimum of such an error. We find N refined coefficients for both sets of Fourier coefficients {C _k }, requiring a minimum of the average forecast error for both arrays or for all available data if multichannel registration is implemented.

If we write down the condition for accurate prediction for one exponent exp (-t / τ) for equally spaced measurements with a step Δt, we obtain a characteristic equation if we introduce the notation w = exp (-Δt / τ):

.

.

Its roots (positive numbers less than 1) are equal to the attenuation factors of all the exponentials included in the measured data. In the flowchart of FIG. 6, these desired roots are called exponential factors. It was the roots of the polynomial with constraints that were found using the procedure described by us in [22]. The procedure takes into account the errors of measurement or calculation of the coefficients of the polynomial, reducing the problem to finding the minimum of the functions of many variables, which describes the sum of the squares of the discrepancies of these coefficients with their calculated values through the desired roots. To increase the accuracy of finding such a minimum, a random choice of the initial descent point to a minimum is made, and the obtained search results are then averaged.

Apparently, one can use standard procedures for finding the roots of a polynomial. However, inaccuracies in setting the polynomial coefficients can lead to the appearance of roots that are outside the permissible range, or even complex. What to do in such cases is not entirely clear. It is better to use a method where restrictions on the sought quantities are laid down from the very beginning. Moreover, the accuracy of the approximation of the coefficients of the polynomial through the found roots in the presence of estimates of the error in the calculation of these coefficients will be a criterion for the adequacy of the obtained exponential attenuation factors.

Knowing these factors or the characteristic decay times of all the exponentials included in the signal measured at the ion rotation frequency close to the resonant one, by solving the corresponding system of normal equations, one can find the contributions of these exponentials and find phase shifts with respect to the rotating field, the signal from which will be undamped An “exponent” with a “damping” factor of 1. Formula (45) allows the ion to be measured from the measured phase shift φ (converted to the phase shift between the directions of the rotating field and the average linear velocity s at a given frequency) determine the characteristic relaxation time of the velocity τ _v . It is uniquely associated with the mobility of κ ions (46).

If the resonant frequency of ions dying with some characteristic time found is unknown, then it is possible to carry out measurements simultaneously for two close rotation frequencies. In this case, having obtained the corresponding phase shifts, we will have two equations (45) with two unknowns, of which both quantities, the resonance frequency and the characteristic relaxation time of the velocity of these ions can be found. Thus, if rotating ions significantly differ in the characteristic times of their death, then their resonant frequencies and mobilities can be determined even in cases where the corresponding ions cannot be resolved by any known method of mass analysis or ion separation by mobility.

Another possibility is the identification of target compounds or the separation of signals from ions with the same m / z, but differing in the characteristic times of their fragmentation or death, is obtained by analyzing the mass spectra of product ions recorded at ortho-VPMS or other subsequent mass analyzers. When high resolution and sensitivity are not critical, recording of the induced signal from freely rotating product ions in the initial part of the radio frequency quadrupole may be sufficient. In this case, the signals from freely rotating ions, which are not products of dissociation, even if m / z coincide, can be separated from the signal-ions of the products, because the signal from the latter will be damped, and from the former it should be stationary in the absence of their noticeable death for any reason.

7 illustrates the idea of such an analysis of ions. Here, the change in time t of the intensities J of three hypothetical mass spectrum peaks is shown. One of them (150) with a minimum m / z, shown by thick lines, decays relatively slowly exponentially (151) with a characteristic time τ ₁ . The other, (153) with maximum m / z, shown by thin lines, exponentially disappears much faster (154) with characteristic time τ ₂ . The average peak (155) is the sum of the first two peaks and is composed respectively of two lines - thick and thin. These three peaks correspond to two hypothetical starting ions, disappearing with characteristic times τ ₁ and τ ₂ , giving two equally probable product ions during fragmentation, shown by thick and thin lines, respectively. These three peaks at consecutive equally spaced time instants with an interval Δt (152) can be expressed through each other using linear expressions, which are written out in matrix form in (156):

In a more general case, the time-following intensities of the mass spectrum peaks are written as the product of the transition matrix A _kl by the vector of previous intensities - equation (157):

и собственные числа (159):Here 1 numbers the measured mass spectral peaks (the total number is n), k - sets the number of initial ions (no more than n), decaying with different characteristic times and producing product ions, the intensities of which form a linearly independent system of vectors. The matrix elements of the transition matrix can be calculated using the least squares method from the condition of the best in the mean-square fulfillment of equality (157) for all moments of measurement. For this, the number of recorded mass spectra (measurement times) should be no less than the number of peaks of the mass spectrum (n). The eigenvectors are found to solve the eigenvalue problem (158) for the transition matrix

and eigenvalues (159):

The eigenvectors describe the intensity distributions of the product ions for each source ion, which corresponds to the eigenvalue (159), which is an exponential decay factor of the number of these ions. So, for example, for the transition matrix in expression (156), vectors with components (1, 1, 0) and (0, 1, 1) are eigenvalues with eigenvalues equal to the first exp (-Δt / τ ₁ ) and the second exp ( −Δt / τ ₂ ) exponential decay factors. After the number of initial ions and the exponential decay or relaxation factors of the corresponding signals are found, the contributions of the mass spectra of product ions from these initial ions in all recorded samples of the initial mass spectrum to successive equally spaced time instants can be calculated by solving the corresponding systems of linear algebraic equations based on the requirement of a minimum of the standard error of the approximation. In this way, mass spectra can be obtained for each source ion, as in the case of their complete separation before measurements. With previously known m / z, some product ions can be left in stable orbits of nonresonant rotation. This will make it possible, if necessary, after their accumulation, to carry out a more detailed analysis of these ions, including their separation according to characteristic times of death.

Product ions, in order not to interfere with an adequate analysis of other accumulated ions, must have m / z different from other ions rotating around the axis of the quadrupole. In the case of ordinary singly charged ions, this means that the most convenient are measurements of the decay of ions with the smallest m / z among all rotating ions. If the resonant frequency of rotation of such ions is sufficient to ensure the decay of these ions with a characteristic time convenient for measurement during rotation in orbit with the maximum possible radius, then the ion fragmentation mode is organized by switching on the corresponding rotational voltage. If the rotation speed is insufficient, the radio frequency voltage of the quadrupole must be increased or its frequency reduced to necessary increase the resonant frequency of this type of ions. In this case, the radii of non-resonant rotation of this and other ions will also change, so it may be necessary to change the frequency and / or amplitude of this rotating field. After times proportional to the relaxation times of the velocities of the corresponding ions, ion rotations will be established with a new frequency and rotation radii. This establishment process can be controlled by registering induced signals from rotating ions. Additional time, proportional to the characteristic time of establishing the internal temperature of these ions, is necessary for heating the ions selected for decay to a stationary level of internal temperature, at which the number of these ions will further decrease exponentially. The resulting product ions will enter the ion transport zone in the ortho-VMSS or other subsequent mass analyzers (104) and will fall into these parts of the device either with a relatively small radius of their non-resonant rotation or when the initial ion (111) intersects this zone with the motion path with their subsequent decay inside this zone or without such decay, if this trajectory passes sufficiently close to the opening of the locking diaphragm (14) - Figure 1.

The second possibility of organizing controlled death or possibly fragmentation of multiply charged ions is to use their collisions with the surface of the conducting ring (19). This case is illustrated in FIG. The potential well (215) near the ring (19) is qualitatively shown for the displacement of Z in arbitrary units from the minimum position. It is the result of field superposition from the potentials of the quadrupole sections (205), from the retaining potential of the ring (19), curved near the ring (19) and the diaphragm (14) of the rotating field, and the effective potential from the radio frequency field distorted by the ring (19) and the diaphragm (14) . The density distribution of rotating ions in the potential well (215) near the ring (19) in accordance with the Boltzmann principle under stationary conditions is given by the expression (221):

Two parameters of this distribution are determined by the nature of the rotating ions: the effective temperature T _eff and the ion charge ez. For given m / z ions and a rotating field, the effective temperature in the first approximation is ultimately determined by the mobility of the κ ions - formulas (222) and (223), where the proportionality coefficient α in different models of ion heating depends slightly on the properties of the ion and atoms or gas molecules:

The phase shift of the rotating ions φ arises due to the deviation of the frequency of the rotating field from the resonant frequency of the ions - formula (45). The graphs (216) show the relative densities of three hypothetical ion types for an ideal quadratic well (215) versus the displacement along the axis of the quadrupole Z. The cloud of maximally untwisted target ions (210) has a maximum rotation radius and, therefore, a maximum rotation speed at a given rotation frequency and accordingly the maximum effective temperature. It is maximally extended among other ions, and it is precisely the ions from the far right part of the cloud that have a chance to recombine or die when they collide with the wall (19). Apparently, the probability of such a death, at least for singly charged complex organic and bioorganic ions, is practically equal to 1. To bounce back and “move away” from the wall upon impact with the conductive surface, an ion that gained additional energy near the surface due to attraction to the counterion image , elastic collision is necessary, and this is extremely unlikely for a complex ion with its increased energy.

In the configuration shown in FIG. 8, ions (214) having an approximately 10% lower effective temperature than ions (210) will have a more than 2 times lower frequency of collisions with wall (19) than ions (210) , since in this case the right tails of the distributions “work” - the “cut-off” levels of the dying ions (218) and (220). If the probabilities of the death of these two types of ions in a collision with a wall are close to 1, then such a difference in the collision frequency will make it quite easy to separate the signals from such ions using known methods and described, for example, in our monograph [18]. If the rotation speed of these ions is large enough and they begin to fragment with appreciable efficiency, forming a spectrum of fragment ions, then using the methods described above, it is possible to separate the signals from different ions, while the characteristic decay times of these signals will be determined as the processes of the death of the initial ions by ring (19), and processes of their fragmentation. By comparing these characteristic times at different ring potentials, both components can be distinguished and the effective temperature (222) of fragmenting ions can be estimated on this basis.

A remarkable property of the separation of signals from different ions upon their death on the ring (19) is the possibility of almost complete separation of ions of different charge in those cases when they cannot be distinguished by m / z and resistance to collision fragmentation. The cloud of 4-charged ions (213) shown in Fig. 8, which coincide in m / z and are close in mobility to ions (210), is concentrated in a region remote from the surface (19), so that these ions practically will not collide with it . This property distinguishes this method from other mass spectrometric methods, where there is no direct separation of ions by the absolute value of the charge, and ions are separated by m / z. The presence of a conducting ring (19) when a potential repelling ions is applied to it, allows one to remove resonantly rotating ions (210), (213) and (214) from the bulk region of non-resonantly rotating ions (16) and thereby weaken the influence of the space charge on resonant rotational frequencies and the probability of death of ions on the ring (19). Ions participating only in nonresonant rotation (16) are assumed to have radii of rotation not leading them to the region of the ring (19) and rotating near the diaphragm (14) covered by a charged dielectric film (9). The influence of the image cloud of counterions (21) will make it possible to have a significantly larger than usual ion capacity of the proposed quadrupole ion trap.

Measurement of the induced signal from rotating ions is possible both in the case of their stationary rotation under the influence of a constant rotating field, and in the case of their free rotation with their own rotation frequencies, when their initial movement is excited either by a broadband pulse or by a package of selected rotational voltages. The latter method, for example, can allow measuring the mass spectra of ions in a given range of m / z. Figure 9 shows a screen shot of a program for simulating the movement of ions in a radio-frequency quadrupole for the case when two types of ions with m / z = 1000 T (304) and m / z = 998 T (305) are trapped in a quadratic potential trap and spontaneous rotation of these ions against the background of their nonresonant rotation around the helium flow (303). The cross section of the quadrupole is shown in the upper left window. Between the rods of the quadrupole (301) are shown the electrodes (302), from which the induced signal from the rotating ions can be removed. They are shown in Fig. 5 under the numbers (112), (113), (122), (123). The paired electrodes (115), (114), (117), (116) extend beyond the part of the quadrupole section shown in FIG. 9. The signal from one pair of electrodes (113) and (114) is shown in window (308). Window (307) shows the kinetic energy distributions of the rotating ions. The result of the inverse Fourier transform of signal (308) after recording for ~ 10 ms in the scale of rotation periods (proportional to m / z ions) is shown in window (309). This result shows that, with characteristic death times of the starting ions in the range of seconds, a sufficient number of mass spectra of product ions for separating isobaric ions can, in principle, be obtained on the basis of excitation of proper ion rotations and registration of induced signals. Data recording on the ortho-VPMS or in other subsequent mass analyzers can be important not only for cases of small signals, separation of superimposed peaks, and obtaining “exact” values of ion masses. It can provide a quantitative calibration (estimation of the number of rotating ions) of the induced signals in the radio-frequency quadrupole and control of the mass scale of the mass spectra obtained from the induced signals using the inverse Fourier transform, if the same mass spectrum is registered after the rotation and the induced signals are recorded on ortho-VPMS or other subsequent mass analyzers.

A significant drawback of the “free” rotation of ions in a quadrupole, as well as in the magnetic field of an ICR mass spectrometer, is a serious limitation on the number of rotating ions close to m / z due to the fusion or coalescence of ion clouds. A screen shot of the simulation program for the case of rotation of 84 50-charged ions with the same m / z as in FIG. 9, taking into account the Coulomb repulsion of ions, is shown in FIG. 10. In this case, the rotation frequencies of these ions turn out to be substantially closer than dictated by the difference in m / z values (409). The case shown is close to the limit of the onset of ion cloud fusion. During the rotation of ions with a large total number of charges, the frequencies of rotation of ions with close m / z cease to differ within the limits of calculation error. In ICR mass spectrometry, to reduce the effect of coalescence, substantially larger radii of rotation (about 1 cm or more) are used than is usually possible in a quadrupole, the radius of which in our case is 3.6 mm. Of course, it is possible to use a quadrupole of a larger radius, but in this case, to ensure close to the previous rotation frequencies of the studied ions, it would be necessary to use large voltages of the radio frequency field (proportional to the square of the radius - formula (40)). In addition, the linear velocity of ions will increase in proportion to the radius of rotation, which at frequencies of about 100 kHz will lead to too rapid ion decay in a gas-filled quadrupole. Apparently, this is one of the main reasons that requires ultra-high vacuum in ICR devices - any collision of an ion at speeds of about 6 km / s with a residual gas molecule (for a nitrogen molecule, the collision energy in this case is about 6 eV) will lead to the elimination of this ion from a packet of rotating ions. Thus, in order to preserve the possibility of sufficiently highly selective kinetic measurements of collision-induced ion dissociation in the radio-frequency quadrupole, the radii of rotation of the ions should not exceed 1-2 mm. With such radii of rotation, the registration of ions participating in free rotation is possible in a relatively narrow range in the number of charges of rotating ions. If these are accumulated product ions from a sufficiently large number of decaying initial ions, then the organization of their death on the conducting ring (19) - Fig. 8 is one of the possible solutions to the problem of limiting the number of rotating ions. Such a death can be carried out periodically by spinning up the necessary ions with a resonant rotating field, and then recording the induced signal after switching off this field. Alternatively, the induced signal can be recorded simultaneously with the excitation of stationary quasi-resonant ion rotation. In this case, the recorded signal will contain exponentially decaying contributions with a characteristic decay time of the initial ions and a characteristic death time on the ring (19). In some cases, in the presence of several types of product ions with different characteristic death times on the ring (19), their separation by these times is possible according to the method described above. Measuring also the phase shifts of the recorded signals with respect to the rotating field, it is possible to determine the mobilities of the product ions, even if they have indistinguishable values of m / z under the given measurement conditions.

When carrying out the measurements described above, it is important to limit the number of accumulated ions, although the influence of the space charge determines the natural blocking of this number, ensuring the departure of "excess" ions on the rods of the quadrupole. However, much earlier than this limit, undesirable effects are possible during the detection of these ions both during registration of induced signals from rotating ions in a quadrupole, and during registration of mass spectra in subsequent mass analyzers. Intentional release from unwanted or excessive ions can also be important in the accumulation of trace ions, which otherwise could be impossible or difficult due to the influence of the space charge. Periodic excitation of free rotation additional to non-resonant rotation of ions and registration of induced signals makes it possible in our case to carry out these operations in a timely manner and prevent undesirable effects associated with excessive accumulation of certain ions. The corresponding resonant spin-up of non-resonantly rotating ions in the storage part of the radio-frequency quadrupole allows one to obtain a sufficiently high selectivity of such elimination even in the case of a noticeable gas flow along the axis of the quadrupole. The use for this purpose of the death of untwisted ions on the conducting part of the locking diaphragm (19), Fig. 1, will make it possible to get rid not only of ions with the selected m / z, but also to selectively attenuate the population of accumulated ions, which significantly differ in mobility or charge size. This possibility is illustrated in Fig. 8, where the conductive part of the locking diaphragm is shown under the number (19), and the corresponding rotating ion clouds and the corresponding ion density distributions are shown under the numbers (210), (213) and (214). This illustration shows that the selective elimination of unwanted ions with a given m / z when using static voltages on the ring (19) and the diaphragm (16) is possible with resonant excitation of rotation of these ions, if they have maximum mobility and minimum charge. If it is necessary to eliminate ions with lower mobility, then this can be achieved by shifting the frequency of the ion spin in the direction that is less dangerous for the loss of any other ions of interest. So, for example, if ions with smaller m / z have already been studied, then a shift towards higher frequencies and, consequently, lower masses - formula (40) - will not lead to the loss of ions not yet studied. In this case, ions of different mobilities and coinciding m / z will have different phase shifts relative to the rotating field - formula (45). The repulsive electric potential of the ring (19) decreases for some time, corresponding to the passage of unwanted ions past the ring (19), and they die on its surface.

Claims (21)

1. The method of structural-chemical analysis of organic and bioorganic compounds based on the accumulation and sequential selective fragmentation of many parent ions in a linear radio-frequency trap, characterized in that, to increase the efficiency of accumulation, a non-resonant rotation of ions is excited around the axis of the trap by a rotating electric field that displays ion packets of a given range mass-to-charge ratios (m / z) from the near-axis zone into which ions under the action of a longitudinal electric field and gas flow, for example trajectories along the axis of the trap, they can enter through the inlet diaphragm and output in the forward and reverse directions (in the absence of a rotating field of sufficient strength), and the fragmentation induced by collisions of accumulated nonresonantly rotating ions with atoms or molecules of a buffer gas is carried out mainly in the region of residual density a buffer gas (which increases its selectivity) by exciting additional rotation or oscillation of ions with a selected m / z with a frequency close to the resonance for these ions; in this case, the time-varying fluxes of ions of fragmentation products and / or the number of selected parent ions disappearing with different characteristic times of the selected parent ions remaining in the trap are recorded, which makes it possible to further separate these parent ions with an estimate of m / z and the mobilities of the fractions of these ions. 2. The method according to claim 1, characterized in that to ensure higher selectivity of selective fragmentation, the linear radio frequency trap is a radio frequency quadrupole. 3. The method according to claim 1, characterized in that to create an ion stream of the compounds under study into a linear radio-frequency trap, the analyzed mixture is added to the buffer gas stream, and in the form of a well-collimated molecular beam, this stream is passed through an electron ionization source located in front of the linear linear input diaphragm radio frequency trap. 4. The method according to claim 3, characterized in that the electron ionization source is a variable ionization energy source. 5. The method according to claim 3, characterized in that a gas mixture is added to the gas stream containing ions of the analyzed compounds from an external ion source, for example, multiply charged ions of biomolecules from an electrospray ion source. 6. The method according to claim 2, characterized in that to prevent unauthorized death of resonantly rotating or oscillating ions on the rods of the quadrupole, these rods have a circular cross section, providing near the rods a significant deviation from the ideal quadrupole field and, accordingly, the shift of the resonant frequencies of the ions. 7. The method according to claim 1, characterized in that for the creation of a longitudinal electric field in a linear radio frequency trap and a possible correction of the longitudinal distribution of the focusing radio frequency field, its rods are sectioned. 8. The method according to claim 1, characterized in that for detecting ion-products of fragmentation induced by collisions with m / z less than m / z of the parent ions, a linear radio-frequency trap is coupled to a mass analyzer, in particular, it can be time span mass analyzer with orthogonal ion input. 9. The method according to claim 1, characterized in that the linear radio-frequency trap in its middle part contains a locking diaphragm, which bounds along with the input diaphragm the region of accumulation and fragmentation of ions. 10. The method according to claim 1, characterized in that to prevent the death of ions on the inner surfaces of the input diaphragm of the linear radio frequency trap and the locking diaphragm of the linear radio frequency trap according to claim 9, these surfaces are completely coated with a thin dielectric film and pre-charged with buffer gas ions having same polarity as the analyzed ions. 11. The method according to claim 10, characterized in that to create a stream of buffer gas ions in a linear radio frequency trap of the method according to claim 10, the molecular beam of this gas is exposed to the electron stream of the electron ionization source according to claim 3 or 4 with sufficient energy for effective ionization this gas. 12. The method according to claim 9, characterized in that by the initiation of rotation close to resonant, ions with selected values of m / z and mobility are periodically displayed in the zone of possible contact with the conductive part of the surface of the locking diaphragm; by changing the constant value or the amplitude and phase of the alternating potential of the aforementioned conducting part of the surface of the locking diaphragm, the amplitude and / or frequency of the rotating field, the characteristic time of the death of the ions changes, which is determined under the given stationary rotation conditions by the mobility of the ions and their charge. 13. The method according to claim 9, characterized in that to prevent unauthorized death of ions on the inner surface of the locking diaphragm of a linear radio frequency trap, this surface is partially covered with a thin dielectric film and pre-charged with buffer gas ions having the same polarity as the analyzed ions. 14. The method according to claim 1, characterized in that when registering successive mass spectra of collision fragmentation product ions, for example, according to the method of claim 8, separating the contributions of product ions to these mass spectra from the starting ions differing in characteristic times death, is based on the calculation of a statistically stable transition matrix between time-equilibrium sets of peak intensities of the mass spectra and estimates of its eigenvalues. 15. The method according to claim 1, characterized in that for recording the induced signals from ions moving in a linear radio frequency trap, a potential difference is measured periodically between two extended recording electrodes placed along the trap rods, so that the induced voltages from these rods to these the electrodes turn out to be close and, to a large extent, during measurement cancel each other at frequencies close to the natural frequencies of rotation or oscillations of these ions. 16. The method according to claim 1, characterized in that for the registration of induced signals from ions, the method of clause 15 is repeated for pairs of said recording electrodes located, inter alia, between pairs of the closest rods of a linear radio-frequency trap, thereby realizing multi-channel recording. 17. The method according to claim 1, characterized in that for controlling the accumulation of rotating ions in a linear radio-frequency trap by measuring the rate of increase of the induced signal at a speed recorded by the method of clause 15 or 16, a significant decrease in this speed can be a criterion for stopping the accumulation . 18. The method according to claim 1, characterized in that to obtain an estimate of the mobility of the selected fragmenting or dying ions at a known resonant frequency of vibration or rotation of these ions in a linear radio frequency trap, a measurement is made using the methods of claim 15 or 16 of the phase shift of the induced signal from of these ions (decaying exponentially for fragmenting or dying ions) from a signal from a rotating or oscillating field that is constant in amplitude for stationary measurement conditions. 19. The method according to claim 1, characterized in that in order to simultaneously obtain estimates of the resonant frequency and mobility of the selected ions, two rotating or oscillating fields with frequencies close to the resonant are exposed to the ions, and two phase shifts are measured at the frequencies of the mentioned fields of the induced signal from ions from the signals of these fields according to the method of claim 18. 20. The method according to claim 1, characterized in that in the presence of fragmenting or dying ions with different characteristic death times in a linear radio-frequency trap, the separation of signals from them obtained by the methods of clause 15 or 16 and phase shifts are determined based on the construction of the finite -difference equations statistically stably describing the observed signals, and finding the roots of the corresponding characteristic equations. 21. The method according to claim 1, characterized in that to obtain a mass spectrum of ions in a selected m / z range present in a linear radio frequency trap, free rotations or ion oscillations are excited by a short wave packet in a given frequency range, and after registration of the induced signals according to the methods of clause 15 or 16, a fast Fourier transform is performed for a given time.

Images