Mass Spectrometer

Abstract

An isotope ratio mass spectrometer has an ion source, a static field mass filter, a reaction cell to induce a mass shift reaction, and a sector field mass analyser for spatially separating ions from the reaction cell according to their m/z. A detector platform detects a plurality of different ion species separated by the sector field mass analyser. The static field mass filter has a first Wien filter that deflects ions away from a longitudinal symmetry axis of the spectrometer in accordance with the ions' m/z, and a second Wien filter that deflects ions back towards the longitudinal symmetry axis in accordance with the ions' m/z. An inverting lens is positioned along the longitudinal axis between the Wien filters to invert the direction of deflection of the ions from the first Wien filter. The static field mass filter provides high transmission and improved spectrometer sensitivity. The first and second Wien filters permit simple tuning.

Claims

1. A static field mass filter for a mass spectrometer comprising: an entrance aperture configured to receive an ion beam from an ion source; an exit aperture configured to eject ions, the static field mass filter defining a longitudinal symmetry axis between the entrance and exit apertures; a first Wien filter configured to deflect ions away from the longitudinal symmetry axis in accordance with their m/z; a second Wien filter configured to deflect the ions back towards the longitudinal symmetry axis in accordance with their m/z; an inverting lens positioned along the longitudinal symmetry axis between the first and second Wien filters, to invert the direction of deflection of the ions from the first Wien filter, characterized in that the static field mass filter further comprises a diaphragm located between the first and the second Wien filters, the diaphragm having an aperture configured to allow passage of at least some of the ion species exiting the first Wien filter.

2. The static field mass filter of claim 1, wherein the aperture of the diaphragm is of a dimension such that ions having an m/z higher than a threshold m/z are deflected by the first Wien filter a relatively smaller distance in a direction orthogonal to the longitudinal symmetry axis so that they are able to pass through the aperture, and ions having an m/z lower than a threshold m/z are deflected by the first Wien filter a relatively greater distance in a direction orthogonal to the longitudinal symmetry axis so that they are unable to pass through the aperture.

3. The static field mass filter of claim 1, wherein the aperture of the diaphragm is of a dimension such that ions having an m/z lower than a threshold m/z are deflected by the first Wien filter a relatively smaller distance in a direction orthogonal to the longitudinal symmetry axis so that they are able to pass through the aperture, and ions having an m/z higher than a threshold m/z are deflected by the first Wien filter a relatively greater distance in a direction orthogonal to the longitudinal symmetry axis so that they are unable to pass through the aperture.

4. The static field mass filter of claim 1, wherein the aperture of the diaphragm is of a dimension such that ions having an m/z in a m/z window are deflected by the first Wien filter a relatively smaller distance in a direction orthogonal to the longitudinal symmetry axis so that they are able to pass through the aperture, and ions having an m/z outside m/z window are deflected by the first Wien filter a relatively greater distance in a direction orthogonal to the longitudinal symmetry axis so that they are unable to pass through the aperture.

5. The static field mass filter of claim 1, wherein the inverting lens is asymmetric.

6. The static field mass filter of claim 5, wherein the inverting lens is a multipole lens.

7. The static field mass filter of claim 6, wherein the multipole lens comprises a plurality of crossed slit aperture lenses.

8. The static field mass filter of claim 1, further comprising a first magnetic shield between the first Wien filter and the inverting lens, and a second magnetic shield between the inverting lens and the second Wien filter, to reduce or remove the effect of stray magnetic fields of the first and second Wien filters, upon the inverting lens.

9. The static field mass filter of claim 1, further comprising an entrance lens positioned between the entrance aperture of the static field mass filter and the first Wien filter.

10. The static field mass filter of claim 9, wherein the entrance lens is formed as an Einzel lens.

11. The static field mass filter of claim 1, further comprising an exit lens positioned between the second Wien filter and the exit aperture of the static field mass filter.

12. The static field mass filter of claim 1, further comprising an electrostatic controller.

13. The static field mass filter of claim 1, wherein the first Wien filter comprises a central channel through which ions in the ion beam can pass wherein the central channel is bounded by at least one baffle for catching ions deflected away from the said longitudinal symmetry axis.

14. The static field mass filter of claim 1, wherein the first Wien filter includes a plurality of electrical conductors extending in a direction parallel with the longitudinal symmetry axis, for generating an electric field within a central channel of the first Wien filter through which ions in the ion beam can pass.

15. The static field mass filter of claim 1, wherein the second Wien filter includes a plurality of electrical conductors extending in a direction parallel with the longitudinal symmetry axis, for generating an electric field within a central channel of the second Wien filter through which ions in the ion beam can pass.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The invention may be put into practice in a number of ways, some of which will now be described by way of example only and with reference to the accompanying drawings in which:

[0048] FIG. 1 shows a schematic view of a prior art isotope ratio mass spectrometer (IRMS) including an RF quadrupole mass filter as a pre-filter and a collision cell.

[0049] FIG. 2 shows a plot of intensity versus mass to charge ratio using an IRMS in accordance with FIG. 1.

[0050] FIG. 3 shows a plot of the stability of isotope ratio measurements of two different isotopes of Neodymium, versus DC quadrupole pole bias, for different mass windows of the RF quadrupole of FIG. 1.

[0051] FIG. 4 shows a plot of the stability of isotope ratio measurements of various isotopes of Neodymium, versus DC quadrupole pole bias, for the RF quadrupole of FIG. 1, following internal normalisation.

[0052] FIG. 5 shows an IRMS in accordance with an embodiment of the present invention, including a static field mass filter as a prefilter.

[0053] FIG. 6 shows, highly schematically, an arrangement of a Wien filter employed as part of the static field mass filter of FIG. 5, to illustrate the effect of magnetic and electric fields on ions of velocity v.

[0054] FIG. 7 shows a plot of mass dispersion against mass to charge ratio for different applied magnetic field strengths, in a Wien filter.

[0055] FIG. 8 shows a schematic view, in the X-Z plane, of a first embodiment of the static field mass filter of FIG. 5, comprising first and second Wien filters separated by an inversion lens, along with simulated ion trajectories in the X-Z plane.

[0056] FIGS. 9a, 9b and 9c show simulated ion trajectories in the X-Z plane, in the arrangement of FIG. 8, with first, second and third mass spreads respectively.

[0057] FIG. 10a shows a schematic side perspective view and FIGS. 10b and 10c show schematic views in the X-Z and Y-Z planes respectively of a second embodiment of the static field mass filter of FIG. 5, comprising first and second Wien filters separated by an inversion lens, along with simulated ion trajectories in those planes.

[0058] FIG. 11 shows a partial cutaway of a part of an Einzel lens forming a preferred embodiment of the inversion lens of FIGS. 8, 10a, 10b and 10c.

[0059] FIGS. 12a and 12b show orthogonal sections through the Einzel lens of FIG. 11, along with equipotential surfaces.

[0060] FIG. 12c shows a simplified version of FIG. 12a and FIG. 12d shows a simplified version of FIG. 12b.

[0061] FIG. 13a shows a simulation, in the X-Z plane, of the magnetic potential lines generated by the first and second Wien filters of FIG. 8, and FIG. 13b shows the magnetic field lines thereof.

[0062] FIG. 14 shows a simulation, in the Y-Z plane, of the electric potential lines generated by the lenses of FIG. 8.

[0063] FIG. 15a shows a perspective close up side view of the first Wien filter of FIG. 10a, sliced in the X-Z plane, illustrating a preferred arrangement of the electrodes for generating an electric field within the first Wien filter.

[0064] FIG. 15b shows a perspective close up side view of the first Wien filter of FIG. 10a, sliced in the Y-Z plane illustrating a preferred arrangement of the electrodes for generating an electric field within the first Wien filter.

[0065] FIG. 15c shows a perspective close up end view of the first Wien filter of FIG. 10a, illustrating a preferred arrangement of the electrodes for generating an electric field within the first Wien filter.

[0066] FIG. 15d shows a view of the first Wien filter of FIG. 10a, along the line A-A of FIGS. 15a and 15b, to illustrate the configuration of the electrodes of FIGS. 15a-15c in further detail.

[0067] FIG. 16 shows a plan view of the electrodes of FIGS. 15a-15d, along with a potential divider.

[0068] FIG. 17 shows a section through the first Wien filter of FIGS. 10a and 15a-15d, in the X-Y plane, along with equipotential lines.

[0069] FIG. 18a shows a perspective close up side view of the second Wien filter of FIG. 10a, sliced in the X-Z plane, illustrating a preferred arrangement of the electrodes for generating an electric field within the second Wien filter.

[0070] FIG. 18b shows a perspective close up end view of the second Wien filter of FIG. 10a, illustrating a preferred arrangement of the electrodes for generating an electric field within the second Wien filter.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT

[0071] FIG. 1 shows a schematic view of a prior art isotope ratio mass spectrometer (IRMS) 1, of the type described in the above referenced WO-A-2017/029200. The IRMS 1 has a triaxial ICP torch 10, a sampler cone 11, one or more skimmer cones 12, an extraction lens 13 and/or a further skimmer cone 14 and/or another ion optical device. Together these components form an inductively coupled plasma (ICP) ion source 15 which generates a collimated ion beam.

[0072] Downstream of the ion source is a quadrupole (RF) mass filter 20 having a preliminary filter 21 upstream of the quadrupole (RF) mass filter 20. The mass filter 20 has an entrance aperture into which the ion beam is directed. Ions leaving the RF mass filter 20 pass through a post filter 22 and then enter a collision cell 30, such as an HCD (high energy dissociation) cell, which may be heated to around 100-200° C.

[0073] Following the collision cell 30, ions are accelerated by an accelerator 40 which accelerates the ions. The accelerated ions are focussed into the ion optics of a double focusing high resolution multicollector mass spectrometer so that multiple different ion species (eg different analyte isotopes and/or standard isotopes for calibration etc) can be detected simultaneously.

[0074] The double focusing high resolution multicollector mass spectrometer includes an electrostatic sector 41 and a magnetostatic sector 43 varying its static field, separated by a focussing lens 42. The electrostatic sector 41 disperses ions by their energy and thus provides focusing for ions of the same energy. The magnetostatic sector 43 disperses ions by mass (strictly, by mass to charge ratio m/z). The electrostatic and magnetostatic sectors can be arranged in a so called Nier-Johnson geometry with scanning of the magnetic flux density of the magnetostatic sector 32 in order to allow sequential focusing of ions having different m/z ratios.

[0075] Downstream of the magnetic sector 43 is a set of dispersion optics 44 whose purpose is to change the mass dispersion and improve peak detection. The IRMS 1 also contains a detector platform 50 such as that described, for example, in GB-A-2,541,391, with 9 Faraday cups and up to 8 ion counters.

[0076] In the IRMS 1 described above, by tuning the amplitude of the RF voltages and the DC voltage offsets of the RF voltages applied to the RF mass filter 20, different ion species can be selected for onward transmission to the collision cell 30. For example, the parameters of the RF mass filter 20 can be set so as to pass ion species across a full range of m/z of the ions generated by the ion source 15 of the IRMS 1 (full transmission). In that case the RF mass filter 20 acts as an RF lens to focus the ions before they enter the collision cell 30.

[0077] Alternative voltages and frequencies can be applied to the RF mass filter 20 so as to permit selection of a m/z window, that is to say, a range of ion species across a window of m/z values contained within the wider range of m/z generated by the ion source. The width (that is, the range between the ions of the highest and lowest m/z) and the location (that is, the m/z value of the centre of the mass window) within the full mass range generated by the ion source 15 can then be set.

[0078] The advantage of the RF mass filter 20 is that the all ions of the ion beam of the selected m/z window exiting the RF mass filter 20 follow the same ion optical path and are not separated in space, like in magnetic sector instruments when exiting the quadrupole mass filter. The selected ions always exit the RF mass filter 20 through the same RF mass filter exit aperture, and have almost no lateral mass dispersion. This is fundamentally different from magnetic sector instruments, where the mass discrimination is based on separating different masses in space. The fact that there is no lateral mass dispersion makes the RF mass filter 20 ideally suited to be coupled with the collision cell 30, since it ensures that all ions leaving the RF mass filter 20 follow the same ion optical path. This in turn means that, at least in a first order there is only limited mass discrimination at the small entrance aperture (typical 2 mm diameter) to the high pressure collision cell 30. Thus an RF-based pre filter seems to be the appropriate solution for many applications.

[0079] However, a major limitation of this setup for high precision and accurate isotope ratio analysis is the “noding effect” which occurs more or less with every RF lens. RF mass lenses rely on alternating strong focusing and defocusing RF-cycles inside the RF lens. For those ions which are rejected by the RF mass filter 20, the defocusing action of the RF lens dominates and forces the ion trajectories to instability. As a result of this instability, the ions do not leave the quadrupole filter through the exit aperture, but rather contact the quadrupoles of the quadrupole filter. These ions are discharged, adsorbed or bonded at the struck quadrupole rod. For those ions which are transmitted by the RF mass filter, however, the focusing action prevails, and the ion trajectories are stable, focusing the transmitted ions to the exit aperture.

[0080] FIG. 2 shows the results from an experiment in which the mass window of a quadrupole mass filter has been selected to be about 8 u wide and there is only one peak major peak at mass 80 in the mass spectrum. In FIG. 2, the peak is a result of an Ar dimer (.sup.40Ar.sup.40Ar) which is formed in the ICP torch 10 of the ion source. The centre of the mass window has been scanned from mass 75u to mass 85u, in order to determine the uniformity of the mass response of the mass window, when the RF mass filter is scanned. The two lines that can be discerned in FIG. 2 are the result of carrying out the experiment twice, at different RF mass filter 20 tunings.

[0081] FIG. 2 shows that the peak steepness is less than 1 u on the rising and the falling edges. The flatness of the peak in the mass range from 78u to 83u appears to be flat on a linear scale, which is good enough for many applications.

[0082] In order to explore the effect of noding in the apparatus of FIG. 1, a Nd standard was introduced into the ICP torch 10. The isotope ratios of 142Nd/146Nd and 150Nd/146Nd were then measured using the detector platform 50, while the DC pole bias of the RF mass filter 20 was changed slightly.

[0083] FIG. 3 shows a plot of the stability of the isotope ratios, with the width of the mass window selected by the RF mass filter 20 selected to be 10 u wide, 16 u wide and finally full width. The solid lines in FIG. 3 represent a plot of the ratio of 142Nd/146Nd and the dashed lines in FIG. 3 represent a plot of the ratio of 150Nd/146Nd. The lines labelled A and A′ represent a wide window of 10 u, the lines labelled B and B′ represent a wider window of 16 u, and the lines labelled C and C′ represent a full mass window (no filtering).

[0084] Oscillations of the measured isotope ratio are clearly visible. These are dependent upon instrument tuning. They indicate that the transmission for the different masses depends on the DC pole bias of the quadrupole, and the size of the oscillations also depends on the mass window which is selected. The pole bias determines the energy of the ions as they travel through the RF mass filter 20. The oscillations of the transmissions are mass dependent and do not occur with the same phase for each mass. Depending on the pole bias adjustment, some ions will have a higher transmission rate and others will have a lower transmission rate. The systematics of this strongly depends upon the tuning settings, and these oscillations disturb the simple use of the internal normalization approach described in the introduction above, to correct for mass discriminations.

[0085] Without any pre filtering (full mass window, the lines labelled C and C′ in FIG. 3), the stability of the measured isotope ratios at different DC pole biases is substantially flat in the range of −1V to −3V. In case of the 16 u wide mass window (lines B and B′), the situation is similar but the curvature of the curve is greater. For the narrowest of the three mass windows (10 u; lines A and A′), the deviation increases to several percent, depending on tuning. This is a result of effects caused by the rounding of the transmission curve characteristics at the edges of the transmission window. Noding effects are magnified at the edges of the transmission profile and result in a periodic oscillation of the expected plateau shown in FIG. 2 of transmitted ions. Due to this noding effect, the precision of the IRMS results can be reduced.

[0086] A similar experiment was carried out using the apparatus of FIG. 1, to investigate the change of all Nd isotope ratios for one mass window setting at different DC pole biases. The results are shown in FIG. 4. The noding of the isotope ratios at different pole biases is clearly visible. The displayed isotope ratios have been internally normalized to the normalizing ratio of 146/144 Nd. The deviations plotted on the y-axis describe the deviation of the measured ratio versus the true ratio. In this case, the measured isotope ratios deviate from the true isotope ratios by several percent, even after internal normalization. This is an extreme example of how the tuning of the RF mass filter 20 can influence the accuracy of the measured isotope ratios. The noding effect of the RF mass filter 20 can result into inaccuracies in the range of several percent, even with internal normalization.

[0087] To address these problems of noding, an IRMS in accordance with an embodiment of the present invention is shown in FIG. 5. The arrangement of FIG. 5 includes a number of components shared with those in FIG. 1, which are labelled with like reference numbers.

[0088] The IRMS 100 of FIG. 5 includes an ion source 15 which is configured in similar manner to the ion source of FIG. 1 and will not therefore be described in detail. The ion source 15 includes a triaxial ICP torch 10, a sampler cone 11, one or more skimmer cones 12, an extraction lens 13 and/or a further skimmer cone 14 and/or another ion optical device 14. This results in a collimated ion beam.

[0089] Downstream of the ion source 15, instead of a quadrupole (RF) mass filter, is positioned a static field mass filter 120 which will be described in further detail below. The static field mass filter 120 maintains constant electric and magnetic fields, so that transmission of ions through the static field mass filter has a flat response across the selected m/z range. A quadrupole mass filter does not provide such a flat response. This is because the ions are only influenced by static fields. In a quadrupole mass filter, the electromagnetic fields change with time according to the applied frequency. This results in a zig zag trajectory of the ions which are pushed back and forth. Moreover, small deviations in system tuning of the static field mass filter 120 do not change the measured isotope ratio in an unpredictable way. Nevertheless, the static field mass filter 120 of FIG. 5 does not introduce a lateral mass discrimination (as would happen in, for example, a magnetic sector analyser) so that the ion beam exiting the static field mass filter 120 can be focussed onto the relatively small (c. 2 mm) entrance aperture of a collision cell 30, across the width of the mass window selected for transmission by the static field mass filter 120.

[0090] As in the arrangement of FIG. 1, following the collision cell 30, ions are accelerated by an accelerator 40 and focussed into the ion optics of a double focusing high resolution multicollector mass spectrometer for simultaneous detection of different isotopes (of the sample or standards). Further, the double focusing high resolution multicollector mass spectrometer again includes an electrostatic sector 41 and a magnetostatic sector 43, separated by a focussing lens 42, and since these were already described in respect of FIG. 1, their specific arrangement and purpose will not be repeated here. Downstream of the high resolution multicollector mass spectrometer, the arrangement of FIG. 5 contains dispersion optics 44 and finally a detector platform 50 again, for example, such as that described in GB-A-2,541,391.

[0091] The preferred arrangement of a static field mass filter 120 in the arrangement of FIG. 5 is a double Wien filter. Wien filters employ an arrangement of crossed electrostatic and magnetostatic fields. Ions passing through this arrangement are subject to the magnetic Lorentz force and the electric field strength in accordance with the following equations:

{right arrow over (F)}.sub.electric=q.Math.{right arrow over (E)}

{right arrow over (F)}.sub.Lorentz=q.Math.{right arrow over (v)}.Math.{right arrow over (B)}

where q is the electric charge of the ion, v is the velocity of the ion, {right arrow over (E)} is the electric field and {right arrow over (B)} is the magnetic flux density. In the case that the initial ion beam velocity is perpendicular to both fields and both fields are perpendicular to each other, this gives, for the axial trajectory, the following condition:

[00001]${\stackrel{.fwdarw.}{F}}_{\mathrm{electric}}={\stackrel{.fwdarw.}{F}}_{\mathrm{Lorentz}}$$q.Math.\stackrel{.fwdarw.}{E}=q.Math.v.Math.\stackrel{.fwdarw.}{B}$$v=\frac{.Math.\stackrel{.fwdarw.}{E}.Math.}{.Math.\stackrel{.fwdarw.}{B}.Math.}$

[0092] In other words, the Wien filter is a velocity filter. It deflects charged particles according to their velocity. Ions with the same ion energy are separated by the square root of their mass (because the kinetic energy of an ion is ½ mv.sup.2). Lighter ions travel faster, for a given ion energy, than heavier ions, and, as such, lighter ions are deflected more in the crossed electric and magnetic fields than heavier ions. This is why the Wien filter also acts as a mass separator. So, in a Wien filter, masses are separated in space. The principle is illustrated in highly schematic form in FIG. 6, where those ions having a velocity v equal to E/B are allowed to pass through undeviated, whereas ions of different velocities not equal to E/B are caused to bend under the influence of the crossed E and B fields so that the ions are separated out. Ions may be shown to deviate from the Z axis in an orthogonal direction X, by a distance ΔX that is given by

[00002]$\begin{array}{cc}\Delta X=\frac{1}{4}\frac{q{l}^2}{E_{\mathrm{kin}}}\left(E-\sqrt{\frac{2E_{\mathrm{kin}}}{m}}B\right)& \left(1\right)\end{array}$

where q is the charge of the ion, l is the length of the Wien filter in the Z direction, E.sub.kin is the kinetic energy of the ion, m is the mass of the ion, and E and B are the electric field and magnetic flux density respectively. For singly charged ions of the same incident energy, the expression above can be simplified to

ΔX=C.sub.1−C.sub.2Bm.sup.−1/2

[0093] In other words, for given crossed electric and magnetic field strengths, the deviation from the Z axis in a Wien filter of fixed length l is related to the inverse square root of the mass of an ion travelling through it. The magnetic flux density and electric fields are kept constant during isotope ratio measurement. In order to adjust the Wien filter so as to select ions of different masses, either {right arrow over (E)} or {right arrow over (B)} may be adjusted. In practice, for small changes in the selected mass, the magnetic flux density is kept constant (eg at 0.5T) and the electric field is adjusted. For analysis of larger masses—such as Uranium for example—it is generally desirable to adjust the magnetic flux density to a higher values such as 1 or 1.5T. One of the reasons for this is that a higher magnetic flux density results in a higher mass dispersion to counteract the decrease of the mass dispersion with higher m/z values (which happens for a constant magnetic flux density). Thus, adjusting the value of magnetic flux density allows the desired mass dispersion in turn to be selected.

[0094] FIG. 7 shows an example of a plot of the dispersion coefficient (X,G) versus mass to charge ratio m/z in amu, for a Wien filter of length 70 mm, and an ion energy of 2 keV. The dispersion coefficient (X,G) links the ion displacement X of an ion of mass M relative to a central mass M.sub.0 as X=(X,G)*(M−M.sub.0)/M.sub.0.

[0095] The solid line represents an applied magnetic flux density of 0.5T and the broken (dashed) line represents an applied magnetic flux density of 1.0T. To calculate the actual deflection of the ion beam, the dispersion coefficient must be multiplied by the relative mass deviation. So for a relative mass deviation of 10%, the actual displacement X=[(X,G)*0.1] mm. For an applied magnetic field strength of 0.5T, ions at m/z around 90 amu will be deflected about 1 mm (10*0.1), whilst ions of m/z around 60 amu will be deflected around 2 mm when an applied magnetic field strength of 1.0T is applied. As will be understood from FIG. 6, the dispersion at a point downstream of the Wien filter will be greater than that at the exit to the Wien filter, because the ions are at that point travelling along a trajectory that diverges from the central axis.

[0096] FIG. 8 shows a more detailed schematic view of the static field mass filter 120 of FIG. 5, in the plane X-Z of FIG. 5 wherein the vector of the magnetic flux density is pointing into the Y-direction which is perpendicular to the plane of the drawing. The static field mass filter 120 comprises entrance and exit apertures 200, 210. The entrance aperture 200 has a typical diameter of around 2 mm. Adjacent to the entrance aperture 200, within the static field mass filter 120, is located a first Wien filter 220. A second Wien filter 230 is located downstream of the first Wien filter 220, within the static field mass filter 120, and adjacent to the exit aperture 210.

[0097] An inversion lens 240 is positioned between the first and second Wien filters 220, 230.

[0098] The trajectories of ions passing from left to right through the double Wien filter arrangement have been simulated and are also shown in FIG. 8. For simplicity, angular divergence (assuming a collimated ion beam at the entrance aperture) and energy spread are ignored. Three parallel beam paths (no divergence) are shown to the left of FIG. 8, entering the entrance aperture 200 of the static field mass filter. The ion beam energy is 2 KeV and three different ion mass to charge ratios are employed, with masses M.sub.0, 0.9 M.sub.0 and 1.1 M.sub.0. This reflects a mass spread of +/−10%. The orientation of the magnetic flux density is perpendicular to the plane of the paper (that is, in the Y direction of FIG. 5), whilst the direction of the electric field is in the X direction (parallel to the plane of the paper).

[0099] The lighter ions of m/z=0.9M.sub.0 are deflected in the negative X direction by the first Wien filter 220, whilst the heavier ions of m/z=1.1 M.sub.0 are deflected in the positive X direction by the first Wien filter 220. The electric field and magnetic flux density are adjusted until ions of m/z M.sub.0 travel along the symmetry axis (the Z axis in FIGS. 5 and 8) of the static field mass filter 120.

[0100] The inversion lens 240 inverts the deflection angles introduced by the first Wien filter 220. In other words, the lighter ions of m/z=0.9M.sub.0 and the heavier ions of m/z=1.1 M.sub.0 are both bent back towards the central symmetric axis Z, in proportion to the amount of deviation introduced by the first Wien filter 220. Ions of m/z=M.sub.0 pass straight through the inversion lens 240, substantially without deviation in the X direction.

[0101] The ions of the various m/z thus converge downstream of the inversion lens 240 into the second Wien filter 230. The electric field and magnetic flux density of the second Wien filter are orientated identically to those in the first Wien filter, so once again, the lighter ions (m/z=0.9M.sub.0) are deflected in the negative X direction by the second Wien filter 230, whilst the heavier ions of m/z=1.1 M.sub.0 are deflected in the positive X direction by the second Wien filter 220. As a consequence of the configuration of the static field mass filter with first and second Wien filters separated by an inverting lens, the spatial distribution of ions at the entrance aperture 200 is imaged onto the exit aperture 210, with a collimated (nearly parallel) beam of ions at the plane of the exit aperture 210 of the static field mass filter 120. Thus, angular focusing of a collimated ion beam at the entrance aperture 200 is preserved at the exit plane of the static field mass filter 120 of FIG. 8, regardless of mass. In mathematical terms, the spatial focusing (at least in first order), along with the mass dispersion focusing, may be expressed as

[00003]$\frac{dx}{d\gamma }=\frac{d\alpha }{d\gamma }=0$

where

[0102] x is the spatial displacement from the Z axis;

[0103] α is the angular divergence from the Z axis; and

[0104] γ is the relative mass difference of the ion beams.

[0105] In the ideal case of an optimized collimated ion beam at the entrance aperture, preferably the values

[00004]$\frac{dx}{d\gamma }\mathrm{and}\frac{d\alpha }{d\gamma }$

[0106] of the ion beam at the entrance aperture 200 and exit aperture 210 remain unchanged or nearly unchanged. So the relative difference between one or both of these values at the entrance aperture 200 and exit aperture 210 is more than 10%, preferably not more than 2% and in particular preferably not more than 0.5%.

[0107] Typically the value of

[00005]$\frac{dx}{d\gamma }$

[0108] at the exit aperture 210 is smaller than 0.2 mm, preferably smaller than 0.1 mm and particular preferably smaller than 0.05 mm for a filtered m/z window.

[0109] Typically the value of

[00006]$\frac{d\alpha }{d\gamma }$

[0110] at the exit aperture 210 is smaller than 4°, preferably smaller than 2° and particular preferably smaller than 1° for a filtered m/z window.

[0111] The entrance aperture of the collision cell 30 is, as noted above, relatively small (c. 2 mm) and it is important that any angular distribution of ions arriving at that entrance aperture is minim/zed. The preservation of both spatial and angular focusing of the ions in the static field mass filter 120 of embodiments of the present invention is thus highly beneficial in ensuring that ions are not mass discriminated at the entrance plane of the collision cell, in accordance with their mass.

[0112] FIGS. 9a, 9b and 9c illustrate simulated ion trajectories through the static field mass filter 120 of FIG. 9, again in the X-Z plane of FIG. 5, for different mass ranges. In FIG. 9a, the mass window is M.sub.0+/−10%, that is, FIGS. 9a and 8 are identical. FIG. 9b shows ion trajectories for a wider mass window M.sub.0+/−20%, and FIG. 9c shows ion trajectories for a still wider mass window of M.sub.0+/−30%.

[0113] As can be seen by comparison of FIGS. 9a, 9b and 9c, lateral mass dispersion (that is, dispersion in the X direction) is greatest at a point between the first and second Wien filters 220, 230, and, moreover, at that point the amount of ion dispersion increases with the width of the mass window. If a specific mass window is selected, only ions having a limited dispersion in the X direction between the two Wien filters should reach the second Wien filter.

[0114] FIGS. 10a, 10b and 10c show an arrangement of a static field mass filter 120 in accordance with an alternative embodiment. FIG. 10a in particular shows a perspective side view of the static field mass filter 120, whilst FIG. 10b shows a section through the central axis of symmetry in the X-Z plane (same plane as FIGS. 6, 8 and 9a, 9b and 9c) along with ion trajectory simulations. FIG. 10c shows the static field mass filter 120 in the orthogonal (Y-Z) plane, again with ion trajectory simulations. Features common to the embodiments of FIGS. 8 and 10a/10b/10c are labelled with like reference numbers. In FIGS. 10a, 10b and 10c, ions travel from left to right. The ion beam energy is 2 KeV, and the static field mass filter 120 has an entrance aperture 200 which is circular with a 2 mm diameter.

[0115] Ions travel through an entrance lens 300 and into a first Wien filter 220. They then pass through an aperture 245 in a diaphragm 255 positioned between the first Wien filter 220 and an inversion lens 240 (FIGS. 10b and 10c; not shown in FIG. 10a for clarity). Following transmission through the inversion lens 240, ions enter a second Wien filter 230. An exit lens 310 is located between the second Wien filter 230 and an exit aperture 210 of the static field mass filter 120.

[0116] The angular divergence of the ion beam is +−30 mrad relative to the Z axis. The mass range is set to M.sub.0+/−20%. The simulation (FIGS. 10b and 10c) shows that ions of the central mass M.sub.0 pass along the Z axis without deviation.

[0117] The mass range transmitted by the static field mass filter 120 may be controlled in a number of ways. For example, as noted above, the magnetic flux density may be adjusted since this in turn adjusts the mass dispersion. However, the magnetic flux density is limited in practice. As another (or different) means for adjusting the mass dispersion of the static field mass filter 120, therefore, the dimensions of the aperture 245 in the diaphragm 255 may be mechanically adjusted to open and close it. Mass dispersion occurs in the X direction (the direction perpendicular to the magnetic flux density). Thus in a simplest arrangement, the aperture 245 in the diaphragm 255 may be opened or closed using cooperating first and second parts that are moveable in the X direction only. Alternatively, the diaphragm may be in the form of an iris, with a circular aperture 245 of variable diameter. The diaphragm 255 may additionally or alternatively be moveable in the Z direction so that, for given dimensions of the aperture 245, ion species of smaller or larger masses may be able to pass through the aperture 245.

[0118] Referring briefly back to FIG. 7, it may be noted that the value for the dispersion X calculated at the exit to the Wien filter can be approximately half the dispersion at the aperture 245, because of the separation between the exit of the first Wien filter and the diaphragm 255, and the divergent trajectory taken by the ions as they leave the first Wien filter towards the diaphragm 255. With this information, the aperture diameter may be set for a given Wien filter dimensions and ion energy, at a specific magnetic flux density, so as to select specific ions. For example, if a magnetic field flux density of 0.5T is applied to a first Wien filter of length 70 mm and ion energy 2 keV, setting the aperture to a diameter of 1.6 mm will transmit all ions in a mass window of +1-10% of a central mass M.sub.0=100 amu: that is to say, in a mass window from 90 amu to 110 amu. Fine tuning may be achieved by adjustment of the magnetic flux density.

[0119] Because of the crossed uniform magnetic and electric fields, the Wien filter arrangement of FIGS. 10a, 10b and 10c has different symmetries in the X and Y directions. As a consequence, the ion optical characteristics are different in the X and Y directions respectively. In order to control the focus in both directions, an asymmetric lens design is desirable. This can be achieved by a multipole lens arrangement to which a static voltage is applied, such as a quadrupole lens, a set of crossed slit aperture lenses, or by an Einzel lens 240, for example. An Einzel lens 240 is shown in FIG. 10a, with outer electrodes 242, 246 between which is positioned a central electrode 244.

[0120] FIG. 11 shows a partial cutaway perspective view of the Einzel lens 240 of FIG. 10a, with the outer electrode 242 removed to reveal the structure of the central electrode 244. In the particularly preferred arrangement shown in FIG. 11, the centre electrode is cut into 4 segments 244a, 244b, 244c and 244d. Such quadrupole lenses can focus in a first direction whilst defocussing in a second direction orthogonal to it.

[0121] The Einzel lens arrangement of FIG. 11 with a split central electrode 244 is particularly preferred because it permits superimposition of a quadrupole component onto the Einzel lens, between the first and second Wien filters. The segmenting of the central electrode 244 allows different focal strengths to be generated in the X and Y directions respectively. FIG. 12a shows a section along the line C-C of the Einzel lens of FIG. 11, whilst FIG. 12b shows a section along the orthogonal line B-B of FIG. 11. The second and fourth electrodes 244b, 244d, lying in the upper and lower quadrants of the lens as seen in FIG. 11, are supplied with a negative DC voltage of, for example, −400V. The first and third electrodes 244a and 244c, lying in the left and right quadrants of the lens as seen in FIG. 11, are grounded (voltages are indicated relative to the ion creation potential). The outer electrodes are supplied with −2000V, and the ion energy is 2000 eV.

[0122] The equipotential surfaces are shown in FIGS. 12a and 12b and also the ion trajectories. FIGS. 12c and 12d show, respectively, simplified versions of FIGS. 12a and 12b, with fewer contours and equipotential surfaces. The different focusing strengths in the orthogonal X and Y directions can be discerned from FIGS. 12c and 12d in particular.

[0123] In FIGS. 10b and 10c, the ion beam is decelerated from 2 KeV at the exit aperture 210 of the static field mass filter 120, down to 250 eV to be focused into the entrance aperture of the collision cell 30 (FIG. 5). The focal point on the right hand side of FIGS. 10b and 10c shows that all of the ions are focused into the entrance aperture of the collision cell 30, and no mass discrimination takes place.

[0124] FIG. 13a shows a simulation of the magnetic equipotential surfaces in the static field mass filter of FIGS. 10a, 10b and 10c, in the X-Z plane of FIG. 8, and FIG. 13b shows a simulation of the more usual magnetic flux density lines in that same static field mass filter. The arrangement of components in FIGS. 13a and 13b is the same as shown in FIGS. 10a and 10b, that is, ions travel through the static field mass filter 120 from left to right in FIGS. 13a and 13b. In addition to the components shown in FIGS. 10a, 10b and 10c, FIGS. 13a and 13b also show a first pair of magnetic shields 320a, 320b positioned at the entrance and exit of the first Wien filter 220, and a second pair of magnetic shields 320c and 320d positioned at the entrance and exit of the second Wien filter 230. The magnetic shields 320a, 320b, 320c and 320d are preferably iron shims, made of soft iron. These help to control the magnetic fringe fields in the static field mass filter 120, which is desirable in order to achieve the desired ion optical performance of the IRMS 100. As noted in connection with FIGS. 10a, 10b and 10c, the inversion lens 240 may comprise or include an Einzel lens. The entrance lens 300 and the exit lens 310 may also consist of or comprise Einzel lenses. These Einzel lenses may have multipole components in order to allow for the different focusing conditions in the X and Y direction of the system (as described above in connection with FIGS. 11, 12a and 12b).

[0125] The magnetic coils of the first and second Wien filters may be of similar construction and geometry. The coil current supplied to each may be the same, with the two coils of the double Wien filter arrangement of FIGS. 10a, 10b and 10c and 13a/13b coupled in series. Then, only a single magnetic field controller is needed. The two Wien filters may alternatively be connected to a single magnetic yoke powered by a single coil, as is shown schematically in the arrangement of FIG. 5.

[0126] Any small differences resulting from constructional tolerances can then be compensated by the electric fields. Although the electric fields may differ slightly, the symmetry of such an arrangement enhances the robustness of the ion optical setup, because instabilities in the magnetic field of the first Wien filter are then compensated in the magnetic field of the second Wien filter 230.

[0127] FIG. 14 shows the electric potentials in the static field mass filter of FIGS. 10a, 10b, 10c, 12a and 12b. Again, instabilities in the electric field in the first Wien filter 220 are compensated in the electric field of the second Wien filter 230.

[0128] The electric field parallel to the surface of the magnet pole piece of the first and second Wien filters 220, 230 also needs to be reasonable homogeneous. This is challenging since the width of the gap between the pole pieces is in the range of about 10 mm while the length of the surface of the pole pieces is about 100 mm. The challenge is to control the electric field in this small gap and to create a homogeneous electric field inside this gap; the usual technique of ensuring that the plates are much larger than the separation is not possible because of the specific geometric challenges in the present arrangement.

[0129] FIG. 15a shows a perspective slice through the first Wien filter 220 of FIG. 10a, in the X-Z plane. FIG. 15b shows a perspective slice through the first Wien filter 220 of FIG. 10a, in the orthogonal Y-Z plane. FIG. 15c shows an end perspective view of the entrance to the first Wien filter 220 of FIG. 10a, and FIG. 15d shows a view in the X-Z plane along the line A-A of FIGS. 15a, 15b and 15c.

[0130] As seen in FIGS. 15a, 15b and 15c, the first Wien filter 220 comprises first and second magnetic pole pieces 400a, 400b which are separated in the Y direction. The magnetic shields 320a, 320b are mounted adjacent to the first pole piece 400a and the second pole piece 400b (FIG. 15b in particular). Suitably, the shields comprise disks with a central hole close to the magnetic pole piece and an adjacent tube extending in the Z direction. The shields are open at the entrance and exit of the system. Between the two fields, the shields are connected to a closed cylinder having first and second holes.

[0131] Mounted upon the magnetic first pole piece 400a is a first plurality of thin electrically conducting lines 450a. A second plurality of thin electrically conducting lines 450b is formed upon the opposed second pole piece 400b. Each thin electrically conducting line in the first and the second plurality of thin electrically conducting lines 450a, 450b is electrically isolated from its adjacent conducting line(s), so that a different electrostatic potential may be applied to each conducting line, as will be further described in connection with FIG. 16 below. One suitable way of achieving this is by printing conductive lines onto an insulating substrate (not shown in FIG. 15a, 15b or 15c) so as to form a printed sheet, as best seen in FIG. 15d.

[0132] A first printed sheet may then be bonded onto the first magnetic pole piece 400a whilst a second printed sheet may be bonded onto the second magnetic pole piece 400b. As may be seen in the Figures, the plurality of thin electrically conducting lines 450a and 450b are printed symmetrically along the Z axis about X=0, and extend outwardly (in the +/−X directions) across only a part of the full width of the first and second pole pieces 400a, 400b.

[0133] Between the magnetic shields 320a, 320b formed along the outer edges of the first magnetic pole piece 400a, and the outermost thin electrically conducting lines printed on the insulating substrates is are positioned a plurality of angled baffles. As may be seen best in FIGS. 15a and 15d, the angled baffles are positioned on both sides of the first plurality of thin electrically conducting lines 450a (and also, with reference to FIG. 15b, on both sides of the second plurality of thin electrically conducting lines 450b). In the illustrated embodiment, there are 6 angled baffles 430a-430f separated in the Z direction towards the edge of the first and second pole pieces 440a, 440b in the +X direction, and 6 corresponding angled baffles 440a-440f that are separated in the Z direction towards the edge of the first and second pole pieces 440a, 440b in the −X direction. The angled baffles 430a-430f and 440a-440f are arranged symmetrically both in the X and Z directions so that they form a herringbone shape. Such a shape helps to catch those lighter ions having a greater angle of deflection in the first Wien filter, which are to be filtered out. The herringbone arrangement thus avoids reflections of the ion beam along the side walls of the arrangement, as the ion beam traverses the magnet gap in the first Wien filter.

[0134] FIG. 16 shows a plan view of the printed sheet upon which the first plurality of thin electrically conducting lines 450a is printed, prior to bonding to the first pole piece 400a and the addition of the angled baffles 430a-f; 440a-f and the magnetic shields 320a, 320b. As may be seen in FIG. 16, each electrically conducting line of the plurality 450a is connected to a voltage divider 460 such that there is a linear drop in electrical potential, in the X direction transverse to the longitudinal direction of the first Wien filter 220.

[0135] Although not illustrated in the Figures, it will be understood that the printed sheet upon which is formed the second plurality of thin electrically conducting lines 450b for bonding with the second pole piece 400b of the first Wien filter 220, is also provided with an array of resistors forming a voltage divider. The voltage divider of each of the printed sheets (bonded to the first and second pole pieces 400a, 400b respectively) is preferably configured identically so that there is a linear drop in electric potential in the transverse (X) direction of the first Wien filter 220, whilst the electrostatic potential along the magnet gap of the first Wien filter (in the Z direction), and the electrostatic potential in the Y direction between the first and second pole pieces 400a, 400b, is constant. A simulation of the electric field generated by the arrangement of FIGS. 15a-15d and 16 is shown in FIG. 17. In particular, FIG. 17 shows a section in the X-Y plane through the first Wien filter 220, at a location along the Z axis which is towards the centre (in the longitudinal Z direction) of the first Wien filter 220 so that any fringe effects are negligible. The first plurality of thin electrically conducting lines 450a (bonded in use to the first magnetic pole piece 400a which is not shown in FIG. 17 for clarity) and the second plurality of thin electrically conducting lines 450b (bonded in use to the second magnetic pole piece 400b which is also not shown in FIG. 17 for clarity) are each shown in FIG. 17 and separated in the Y direction. Equipotential lines are shown in FIG. 17, illustrating the good homogeneity of the electric field in a direction orthogonal to the magnetic field.

[0136] FIG. 18a shows a perspective slice in the X-Z plane through the second Wien filter 230 in the arrangement of FIG. 10a in particular. The slice through the second Wien filter 230 of FIG. 18a is the same as the slice shown in FIG. 15a in respect of the first Wien filter 220. FIG. 18b shows an end perspective view of the second Wien filter 230.

[0137] In each of FIGS. 18a and 18b, it may be seen that a first plurality of thin electrically conducting lines 500a are positioned on a first magnetic pole piece 510 of the second Wien filter 230, whereas a second plurality of thin electrically conducting lines 500b are positioned on a second magnetic pole piece 520 of the second Wien filter 230. The construction and configuration of each of the two sets of thin electrically conducting lines 500a, 500b may be the same as that described in connection with FIGS. 15a-15d and 16 above, with lines printed onto an insulating substrate that is then bonded to the first and second pole pieces 510, 520. Each printed substrate may also be provided with a potential divider such as is shown in FIG. 16.

[0138] By contrast with the arrangement in the first Wien filter 220, however, the second Wien filter does not contain any angled baffles. The flat electrodes (FIG. 18) are accordingly positioned immediately adjacent the outermost lines of the plurality of electrically conducting lines 500a, 500b. The purpose of the baffles, desirably present in the first Wien filter 220, is to minim/ze charging effects at the electrodes by intense ion beams. In ICP, for example, there is a very intense argon beam. The effects of such a beam are already filtered out by the time the ions arrive at the second Wien filter 230, so that relatively very few ions strike the electrodes in that, and accordingly, baffles are not considered to be necessary in the second Wien filter 230.

[0139] Although some specific embodiments have been described, it will be understood that these are merely for the purposes of illustration. Various modifications are possible. For example, although preferred embodiments have been described in the context of IRMS, it is to be understood that the invention is in no way thus limited, and that other types of mass spectrometer could also benefit from the static field mass filter described herein. Moreover, it is to be understood that the collision cell described above, that induces a mass shift reaction, is merely for the purposes of illustration. More generally, the ions that are filtered by the static field mass filter may be provided to any type of reaction cell, which might induce any type of change to the filtered ions prior to analysis. Indeed, in other embodiment contemplated and considered to be within the scope of the present disclosure, the ions exiting the static field mass filter could be directed instead directly to an analysing unit such as a mass analyser.