GB2499868A - Mass spectrometer including magnetic and electrostatic fields and a detector array - Google Patents

Abstract

A mass spectrometer is disclosed which is particularly useful in analysing ion beams over a wide range of ion masses. Individual beams of different mass ions are imaged on to an array of slits in a slit array plate 7, the spacing of the slits corresponding to the spacing of the set of ion beams 1 to 6 resulting from magnetically spreading a mixed ion beam. Individual detectors 8, 9, 10, 17, 18 and 19, e.g. Faraday buckets, are associated with each slit. In accordance with the invention, a number of electrostatic analysers 14, 15, 16 are positioned between some or all of the slits in the slit array plate 7 and the associated detector elements. The electrostatic analysers capture stray ions and electrons and accordingly markedly increase the sensitivity of the apparatus.

Description

1

MASS SPECTROMETERS INCLUDING DETECTOR ARRAYS

This invention relates to mass spectrometers, and, in particular, to such 5 instruments which use a detector array, and which are to be used to analyse ion beams over a wide range of ion masses.

Much can be learned by studying the isotope ratios of atoms and molecules. It has been shown recently that the isotopic ratio of doubly substituted 10 species such as the 13C160180 isotopologue of CO2 or the 13CH3D

isotopologue of CH4 can provide direct information on the temperature at which the gasses were formed, (see e.g. Eiler and Schauble, Geochimica et Cosmochimica Acta, 68 (2004) 4767-4777 (Ref 1) and J.M. Eiler, Earth and Planetary Science Letters 262 (2007) 309-327 (Ref 2)) As such they can be 15 used to provide varied applications such as the possibility of an absolute thermometer for studying the thermal history of our planet (required for climate studies) or could act as a marker for different sources of methane in mining / drilling operations. The technique relies on the fact that the strength of bonds between two heavy isotopes is greater than between lighter ones 20 (due to the zero point energy of the molecular bond), so there is a preference for such a distribution over a completely randomised population. As the temperature is raised, the system will move to complete randomisation, hence the use of the technique as a thermometer. The isotopes of carbon, oxygen and hydrogen cited here are conventionally 25 termed "stable isotopes", since they do not decay with time, and hence the proportion in a sample stays constant. Isotopes of elements which decay over time are termed "radiogenic", and are often used in such studies as the dating of rocks.

30 The multiply substituted isotopic species tend to have an extremely low abundance in natural samples. Thus, for the case of carbon dioxide, the natural abundance of the carbon 13 isotope is 1.1% of the major 12C, whilst the oxygen18 isotope is only 0.2% of the major. Thus the rare 13C160180

2

isotopologue, (of atomic mass 47) exist at about 44ppm of the major (44amu) molecular species. Theoretical studies have shown that temperature effects during the formation of the molecule mentioned above will result in a total variation of the absolute concentration of this species by 5 about one part in a thousand. Close to room temperature, a variation in absolute concentration of a few ppm (parts per million) can permit temperatures of a few degrees Celsius to be measured. Such variations can be measured by use of mass spectrometry.

10 In the current practice of mass spectrometry in such studies, gas of the species under study is introduced into the source of a mass spectrometer, where it is ionised, and the ions produced extracted using a potential field of 3 to 10kv. The ions then pass through a magnetic field, where the lighter species undergo a greater deflection than the heavier constituents of the 15 sample. The separated ions are then focused onto an image plane, where an array of detectors is placed, to record the intensity of the individual masses. Taking carbon dioxide as an example, ions of mass 44 correspond to the molecule 12C1602, which is the major isotopic constituent of the gas. Mass 45 ions can arise from 13C1602 (about 1.1%) and 12C160170 (about 20 0.08%). Mass 46 ions arise from the 12C18O10O isotopic species (about 0.4%) together with a minor contribution from 13C170160. Techniques for allowing for the presence of the minor isobaric interferences to determine the absolute concentration of 13C and 180 in the original gas sample are well known.

25

These absolute concentrations can be measured to precisions of a few ppm by two modifications of a more familiar mass spectrometer arrangement. Firstly all measurements are made relative to a reference gas of known isotopic composition by alternating the measurement of the reference and 30 sample. This alternation is achieved by means of dedicated valving just prior to the source, and the sequencing normally used allows for about a 10 seconds measurement of each gas stream, before the valves are switched

3

and the second gas stream measured. Such technique is well known, and allows for any slow drift of transmission or sensitivity of the instrument to be cancelled out.

5 To allow for any variation in transmission of the various isotopic species (different masses) it is standard practice to record all the masses simultaneously with separate detectors. Thus, for the studies of interest here (but not exclusively) it is conventional to have an array of Faraday buckets placed along the mass spectrometer image plane, whereby each bucket is 10 positioned to collect ions of different (integer) mass. In the case described above three detectors would be positioned to record the 44, 45 and 46amu ion beams. Since both the reference and sample gas streams are measured in this way, any instrumental variation in transmission or detection of the different masses is cancelled out, allowing high accuracy to be achieved. 15 Such an array is termed a multi-collector detector array.

To measure the doubly substituted isotopologues the above approach has been modified to provide extra detectors. For the case of carbon dioxide, extra detectors placed to record the extra masses 47, 48 and 49 amu are 20 conventionally employed. The 47amu detector is used to record the required 13C160180 ion beam. However it was suspected, and since shown to be the case, that at the required level of measurement precision, isobaric interferences from other species are often present in the gases. Even following extreme "cleanup" of the samples, it has proved virtually impossible 25 to remove all nitrogen and chlorine species from the gas. Thus, for example, 15N1602 also has a nominal mass of 47amu, as does 35CI12C, and neither species can be totally removed from the sample or reference gas. To allow for these interferences it has been shown that the 48amu ion beam can be used to compensate for the nitrogen impurity (since the 47amu nitrogen 30 signal has been shown to be a constant fraction of the 48amu ion beam), whilst the 49amu ion beam can be similarly used to compensate for the chlorine contamination.

4

Unfortunately other interferences of these minor signals are also present in the present generation of instrument. These manifest themselves most obviously in the variation of the baseline around the peaks of interest, and 5 are due to scattered ions and electrons. Thus, for example, as the major ion beam passes through the flight tube of the mass spectrometer, some of the ions are incident on the containing walls, and produce secondary electrons and ions on collision with the metal surface. This effect can produce a smoothly varying background of scattered species, on top of which the 10 recorded beams of interest are monitored. For the case of scattered electrons the baseline is depressed (becomes negative in value), whilst for ions the baseline is raised. In theory this effect can be allowed for by measuring the baseline at "half-mass", where no peaks should be present, only this scattering. In some cases this "smooth" baseline depression is, in 15 absolute terms, many times larger than actual measured ion beam signal.

However, on top of this "smooth" effect is the collection of scattered ions and electrons resulting from the fact that not all the ions incident in a Faraday bucket are captured to impart the recorded signal, and some do escape. 20 Thus the assumption is conventionally made that once an ion passes through the entrance of a Faraday bucket it can be ignored from further discussion. This is an incorrect assumption, since no Faraday bucket is exactly 100% efficient. When an ion is incident on the base of the bucket, there is a non-zero possibility of secondary ions and electrons being 25 produced. Since the bucket is not infinitely deep, there is then a finite chance that some of these ions and electrons can escape from the bucket, and a proportion of these secondaries may be deflected by stray magnetic and electric fields near the detector array, and enter and be recorded by adjacent collectors. This type of scattered species results in an offset of the recorded 30 ion beam signals of the minor constituents, but is (approximately)

proportional to the major ion beam intensity. Such scattering results in an apparent non-linearity of the observed signals, and is conventionally allowed

5

for by careful comparison with standards prepared at known temperatures. (These are not the "reference" gases mentioned above used to allow for short term variation of mass spectrometer transmission but gasses prepared in the same way as the samples, but at different temperatures, so as to 5 provide the absolute temperature scaling.)

A further background contribution can occur when the ion beam is scanned across the entrance of the collector slits. As it is incident on the mechanical edge of the defining slit, large numbers of secondaries are produced. This 10 can manifest itself with severe distortion of the recorded minor signals around the start and end of the recorded ion beam peak. In itself this effect should not affect the instrumental performance of the measurement process, since measurement is not performed when the ion beam is not centred in the collector. However this effect can make the "smooth" variation of baseline 15 difficult to quantify, since the ion beam is not infinitely narrow relative to the collector width and spacing, and a significant contribution can still be present at the "half mass" position. Often such effects manifest themselves with asymmetric baselines between peaks, and the user must "guess" the best place to take the baseline measurement, making the assumption that the 20 position chosen provides a good representation of the general scattering.

Thus although there are numerical methods to allow for these baseline errors, they rely on the assumption that they are stable within the measurement period. However this can be quite long in practice. Due to the 25 small absolute concentrations of these minor, doubly substituted species, and the high precision required for their measurement in order to produce meaningful data, each sample preparation and measurement sequence often takes a period of several hours. Further, one has to also measure the standards (used for the absolute temperature scaling) on a similar frequency 30 if the final error is not to be determined by the uncertainty of standard measurement. It has been shown that even minor changes in the source tuning conditions can cause gross changes in the absolute values of these

6

scattering effects. These place extreme requirements on the long-term instrument stability which, in turn, means that one is restricted to dedicate one instrument to these studies to the exclusion of all else, rather than to provide a universal solution which can be used for such work as well as the 5 more conventional studies.

Attempts to minimise the effects of scattered ion beams have been reported, but these have been limited to trying to record minor isotopic beams in the presence of intense majors, where the scattering is a result of gas collisions 10 within the mass spectrometer vacuum envelope. These gas collisions result in ions losing energy, and producing a tail in the mass scan to the low-mass side of the peak. The common example is the measurement of the minor isotopes of uranium which are often "swamped" by the low mass "abundance" tailing of the major 238amu ion. To overcome this tailing, two 15 approaches have been successfully applied, either the provision of a large (physical size) energy dispersive element or the use of a device to provide a potential barrier. The first device can physically separate in space the ions which have not undergone a collision (and hence still have the full acceleration energy of the original ion beam) and those which lost energy 20 due to the collision. The latter will experience a greater deflection by the (usually) electrostatic field of the dispersive element (an electrostatic analyser or ESA), and this can be used to reject, or at least minimise, the scattered tail from the major. In order to physically separate the unwanted scattered ions from main beam of interest, it is known to be necessary to use 25 an ESA with a mechanical radius similar to or greater than the radius of the magnetic analyser. The second type of device, which tends to be more prevalent nowadays due to its smaller physical size effectively provides a potential barrier close to the acceleration energy of the original ion beam. Thus ions which still have their original energy can just pass through the 30 device, but any ions which have suffered collisions, and hence lost energy, no longer have enough energy to pass the barrier, and are rejected. Both approaches only work for tailing to the low-mass side of the major peak,

7

since energy loss (collision induced) tailing is purely asymmetric to the low mass side.

These types of device are designed for the rejection of ions with energy 5 close to that of the original ion beam, since the energy lost in the collision tends to be small. In terms of the scattering problem referred to above, there are a number of major differences to note: the energy difference between the scattered, interfering species, and the required ions is large; the minor ions are to the high-mass side of the major ion beam present in the mass 10 spectrometer; much of the scattered contribution is due to electrons, not ions.

As explained in more detail below, in accordance with the invention, small electrostatic analysers are employed to reject these species, permitting a 15 simple solution to their removal from the recorded spectra. Since the electrostatic analysis devices are small, it is possible to fit a number of them adjacent to each other, enabling the background rejection to be undertaken for many adjacent masses.

20 In accordance with the present invention, there is provided a mass spectrometer including an ion beam source, means to generate a magnetic field to deflect ions in the ion beam from the source by amounts differing dependent on the mass of each ion, and to form the image of the source on to a primary defining slit array having a plurality of slits corresponding to 25 different deflection paths of the ions in the beam under the influence of the magnetic field, and a plurality of ion beam detecting elements constituting a detector array, each element being positioned to detect ions passing through a respective slit in the slit array, wherein the separation of the ion paths transverse to the direction of the beam is greater than the width of the ion 30 beam detecting elements, and wherein located on at least some of the deflection paths of the ions of masses higher than the predominant species of ions in the beam emerging from the ion beam source, are one or more

8

electrostatic analysers each located between a slit in the slit array and its corresponding ion beam detecting element.

In such a spectrometer, the addition of the ESAs detector array minimises 5 the unwanted scattering effects described above, thus providing a more stable and hence user-friendly instrument for studies over such large dynamic ranges, and accordingly removing the restriction of the use of a dedicated instrument for the studies of multiple substituted isotopic species.

10 The use of small ESAs in multi-collector arrays has been suggested before (see e.g. EP-A 0509887) but specifically to enable a detector array using large, discrete dynode detectors used in an arrangement with an oblique collector image plane. By rotating the beams after they have passed through slits in a plate set obliquely to the direction of the beams, the spacing 15 between adjacent detectors may be increased, allowing larger detectors to be used compared to the case if they had to lie on the paths of adjacent but still closely spaced separated beams of different mass ions, for example with a separation of only a few mm between beams of adjacent amu values. In the arrangement disclosed in EP-A-0509887, the small ESAs are used 20 merely to rotate the beams to lie approximately perpendicular to the incident ion beam direction so as to permit the large width detectors to be used and, because of this size constraint, such an arrangement needs to be used for every detector as illustrated in EP-A-0509887. EP-A-0611169 discloses an alternative approach to solving the problem of the spacing between adjacent 25 separated ion beams being less than the width of the detectors used, by providing secondary-emissive elements in the paths of the ion beams and detecting the strength of each ion beam by detecting emissions from those elements using detectors of width greater than the spacing between the separated ion beams.

30

In contrast, in the arrangement according to the present invention, there is enough room to place the detectors along the collector image plane. This

9

may be done if desired by retrofitting the ESAs and moving the detectors in instruments presently in use in such studies, especially the isotopic ratio studies mentioned above where the main beam following separation is many times stronger than the very weak beams to one or either side of the main 5 beam, and in respect of which scattering effects can easily lead to counting inaccuracies in the case of the numbers of ions in these very weak beams.

The invention is further explained with reference to the accompanying drawings in which:

10

Figure 1 illustrates the basic mass spectrometer arrangement used.

Figure 2 shows diagrammatically a collector arrangement used by instruments constructed using the present state of the art, (J.M. Eiler (Ref 15 2));

Figure 3 shows a collector arrangement as used in a spectrometer in accordance with the present invention.

20 Referring to Figures 1 and 2, in known prior art mass spectrometers, the gas of interest is ionised in a source 24 different masses in the beam are dispersed by a magnetic sector 25, of radius Rm, where the lighter ions undergo a greater deflection than heavier ions. The individual separated ion beams 1 to 6 (Figure 1 illustrates only the position of beams 1, 3 and 6) are 25 then incident on a mass spectrometer primary defining slit array 7, behind which are conventionally placed individual Faraday buckets 8 to 13. The primary defining slit array 7 is placed where the magnetic sector 25 images the mass dispersed image of the slit in source 24 from which the ion beam emerges. The arrangement shown in Figure 2 allows the simultaneous 30 measurement of six masses and corresponds to the case for the carbon dioxide studies mentioned above. In this example, ion beam 1 is the final path of the trajectory of the most deflected ions (mass 44amu) and ion beam

10

6 that of the ions least deflected (49amu). Other numbers of collectors are possible, and in the case of very weak ion beam signals, it is possible to replace one or more of the Faraday buckets with ion multipliers. This method of detecting very weak beams is well known to the art.

5

Figure 3 shows a detector array in a spectrometer in accordance with the present invention. In the arrangement illustrated in Figure 3, the three detectors 17, 18 and 19 for the three less deflected ion beams are positioned further away from the primary defining slit array 7. These three 10 ion beams, after passing through the respective slits in the mass spectrometer primary defining slit array 7 are deflected into the displaced Faraday buckets 17, 18, 19 by small electrostatic analysers 14, 15 and 16. In the example shown, as an example, the radius of the dispersing magnet is 250 mm, whilst the radii of these ESA assemblies is only 80 mm. The 15 separation between adjacent ion beams at the primary defining slit array 7 is about 12 mm. Because the physical width of the Faraday buckets 17,18, 19 is about 4 mm, there is adequate space to position the collectors behind the primary defining slit array 7. Voltages are applied to the inner and outer plates of the ESAs 14,15 and 16 to ensure the beams which have the 20 correct acceleration energy undergo the correct deflection to pass through a second defining slit in front of each of the Faraday buckets 17, 18, 19 and be collected by the corresponding detector. As shown, the trajectory 21 of beam 4 is deflected by the ESA 14 and then passed through a further slit 20 into Faraday bucket 17. The calculation of the values for the voltages which need 25 to be applied to the plates of the ESA is well known. Scattered ions in the beam 4 will have a much lower energy and undergo a greater deflection (trajectory 22), whilst electrons in the beam will be deflected in the opposite sense (trajectory 23) and, as a result, neither will enter the Faraday bucket 17 and be recorded. Although three small ESA assemblies are shown in the 30 example illustrated, the detector array may include more or fewer.

11

The present invention is directed not only to mass spectrometers as described above, but also to methods of increasing the sensitivity and accuracy of isotope spectrometry by the use of additional ESA assemblies to deflect the paths of weak ion beams after they have passed through a 5 respective slit in a primary defining slit array.

12

Claims (6)

1. A mass spectrometer including an ion beam source, means to generate 5 a magnetic field to deflect ions in the ion beam from the source by amounts differing dependent on the mass of each ion, and to form the image of the source on to a primary defining slit array having a plurality of slits corresponding to different deflection paths of the ions in the beam under the influence of the magnetic field, and a plurality of ion beam detecting 10 elements constituting a detector array, each element being positioned to detect ions passing through a respective slit in the slit array, wherein the separation of the ion paths transverse to the direction of the beam is greater than the width of the ion beam detecting elements, and wherein located on at least some of the deflection paths of the ions of masses higher than the 15 predominant species of ions in the beam emerging from the ion beam source, are one or more electrostatic analysers each located between a slit in the slit array and its corresponding ion beam detecting element.

2. A mass spectrometer according to Claim 1 wherein at least some of the 20 ion beam detecting elements are Faraday buckets.

3. A mass spectrometer according to Claim 1 or 2 wherein one or more of the ion beam detecting elements are ion multipliers.

25

4. A mass spectrometer according to any one of Claims 1 to 3 wherein an additional slit is located between each electronic analyser and its associated detector.

5. A mass spectrometer according to any one of Claims 1 to 4 wherein the 30 electrostatic analysers are configured to minimise the number of unwanted scattered ions reaching the ion beam detecting elements.

13

6. A mass spectrometer according to any one of Claims 1 to 5 wherein the electrostatic analysers are located exclusively in the deflection paths of ions of masses higher than the predominant species of ions in the beam.