WO1992012529A1

WO1992012529A1 - Mass-spectrometer

Info

Publication number: WO1992012529A1
Application number: PCT/GB1992/000010
Authority: WO
Inventors: Donald Young
Original assignee: Thorn Emi Plc
Priority date: 1991-01-05
Filing date: 1992-01-03
Publication date: 1992-07-23

Abstract

A mass-spectrometer comprises an ionisation chamber (4) and an electrode cell (10) for preferentially accelerating ions having a predetermined velocity towards a target electrode (12). The cell (10) comprises four serially arranged apertured electrodes (A, B, C, D) which are energized by an energization voltage generator (61). The generator (61) supplies a variable ion-attracting bias voltage to the first electrode (A) and periodic voltage waveforms, relative to the first electrode bias voltage, to the other electrodes (B, C, D), these waveforms all having the same period. At the start of each period the second and fourth electrodes (B and D) are supplied with a short ion-attracting voltage pulse. During each period the third electrode (C) is supplied with a voltage waveform which starts at an ion-repelling voltage V2 and ends at an ion-attracting voltage V2 and passes from one to the other via a step to an ion-repelling voltage V2/2, a linear ramp from this to an ion-attracting voltage V2/2, and a step from this to the ion-attracting voltage V2.

Description

MASS-SPECTROMETER

TECHNICAL FIELD

The present invention relates to mass spectrometers, and particularly but not exclusively to mass spectrometers which can be manufactured in miniature size, in comparison to known

spectrometers, using microengineering techniques.

BACKGROUND ART

The advantages of miniature mass spectrometers include light weight and small volume for portability and operation in experimental or other conditions in which such characteristics

are useful, (such as in a bell jar, or in space application) cheaper fabrication costs, geometrical construction accuracy afforded by the microengineering fabrication techniques used, which in turn provides enhanced ion collection efficiency and improved resolution.

Furthermore, the relatively miniature size

gives rise to a relatively short free ion path, enabling such devices to operate at higher pressures than conventionally sized mass spectrometers. This last characteristic is particularly advantageous for gas chromatograph-mass spectrometer combinations.

To enable a miniature mass spectrometer to be fabricated in standard type equipment normally used in microengineering

fabrication, the part thereof consisting of an ionisation chamber and ion selector is preferably no greater than 9cm in total length, and typically might be 2 cm in both breadth and

depth. In view of this total length restriction an ion selection method which gives optimum resolution is of major importance.

Ion selection in standard or full size mass spectrometers is usually carried out by one of a few recognised techniques.

Ion selection may be effected by deflection in a magnetic field such as described by A.J. Dempster in Physics Review 11, 1918, page 316. magnets to provide the deflecting magnetic field. Alternatively, ion selection may be effected by radio frequency focusing, such as in the Quadrupole Spectrometer Instrument as described by P.H. Dawson in "Quadrupole Mass Spectrometer and

Applications", published by E L Sevier of Amsterdam, Oxford, New York. The Quadrupole device, whilst being more compact than the magnetic sector instruments, is still large in comparison to the device envisaged by the present invention. It has also been proposed to effect ion selection by "time of flight techniques" as described by W.C. Wiley and J.H. McLaren in Review Science Institute No. 26 (1955) at page 1150. However, such instruments are typically 1 metre or longer in length, making them unsuitable for applications requiring a miniature device. Radio frequency mass spectrometers have also been proposed, such as that described by W.H. Bennett in Journal of Applied Physics 21 (1955) at page 143. These devices included a triode electrode cell to act as the ion filter, the electrodes all being fed with radio frequency sine waves. However, whilst devices of this kind can be made more compact than the above referred to spectrometers, the resolution of such a device is rather poor and alternative arrangements have been described which use a relatively large number of serially disposed electrodes, alternate electrodes being fed with radio frequency sine waves with each intermediate electrode being coupled to earth. However, it will be appreciated that, in view of the number of electrodes required to be employed as the ion filter, the device is relatively large.

Furthermore, because of the large number of electrodes required sensitivity is greatly reduced.

Lastly, linear accelerator arrangements have also been proposed, such as described by P.A. Redhead in Canadian Journal of Physics, 30 (1952) at page 1.

The present invention, whilst seeking to provide a mass

spectrometer which can be of reduced size, also seeks to provide a device which can have increased resolution for a given sensitivity, sensitivity being defined as the percentage of total ions collected. DISCLOSURE OF INVENTION

The invention provides a mass spectrometer comprising an ionisation chamber, an ion detector electrode for receiving an ion-repelling bias voltage, an electrode cell for preferentially accelerating ions having a predetermined mass and velocity from said ionisation chamber towards said ion detector electrode, which cell comprises first, second, third and fourth apertured electrodes spaced in that order along a path for ions from said ionisation chamber to said ion detector electrode, and energization voltage generator means connected to said first, second, third and fourth electrodes for applying a first ion-attracting d.c. bias voltage to said first electrode, a first periodic voltage waveform to said second and fourth electrodes, and a second periodic voltage waveform to said third electrode, the first and second voltage waveforms having periods which are equal to each other, the first voltage waveform being ion-attracting during at least part of a first half of each said period, and the second voltage waveform being ion-repelling during at least part of said first half of each period and being ion-attracting during at least part of a second half of each period. BRIEF DESCRIPTION OF DRAWINGS

An embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings in which;

Figure 1 is a diagramatic representation of a mass spectrometer in accordance with the present invention;

Figure 2 is a schematic representation of energization voltages applied to apertured electrodes of the spectrometer shown in Figure 1;

Figure 3 shows a possible construction for .part of the

spectrometer of Figure 1 in more detail;

Figure 4 shows part of a possible construction for a two

electrode-cell version of the spectrometer shown in Figure 1; and

Figure 5 shows another part of the construction of Figure 4.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

In the following description like reference numerals have been used to designate like parts of the embodiments in the various figures.

Referring to Figure 1, a mass spectrometer 2 comprises an ionisation chamber 4 including, in the example shown, a filament 6 for providing a source of electrons, and an electrode 8 for receiving a positive potential, shown as +V volts, relative to ground for attracting the electrons emitted by the filament 6 which is at ground potential. The spectrometer 2 also includes an electrode cell 10 which comprises a minimum of four serially disposed apertured electrodes A,B,C,D having a mutual pitch S. A target electrode 12 is also provided to act as a detector for ions passing through the electrode cell 10.

A housing 14 is used to enclose the above components and may include an ion transparent window 16 for isolating the ionisation chamber 4 from the electrode cell 10. The filament may alternatively or additionally be separated by an electron transparent window 11, as shown in Figure 1, so as to isolate the filament from substances under investigation introduced into the ionisation chamber.

In use, a substance to be investigated is introduced into the ionisation chamber 4, usually in gaseous form, via a suitable aperture or pumping port 18. Electrons emitted from the filament 6 are accelerated into the ionisation chamber 4 by the potential on the electrode 8, where they ionise the neutral gas atoms of the substance under investigation, thereby producing ions. These ions may be positively or negatively charged. In the following description it is assumed that they are negatively charged. If positively charged ions are of interest the signs of the energization voltages of the

electrodes A, B, C and D and the target electrode 12 should be reversed.

The negatively charged ions so produced are accelerated towards the first apertured electrode A, which is under the influence of avariable positive DC potential V relative to ground. The ions atttain a velocity V dependent upon their mass , namely

where q is the charge on the ion, m is the mass of a unit ion, and M is the atomic mass number of the ion.

The spectrometer is designed around a selected ion velocity, and is arranged to ensure that ions with this velocity will attain greater energy than ions of any other velocity. By control of the DC potential V_a' typically between 20 volts and 200 volts, it can be ensured that ions of any specific ion mass Mm_o can attain this velocity at or near to the apertured electrode A. The potential V_a is produced at an output 60 of an energization voltage generator 61.

The energy gained by any ion between apertured electrodes is given by:

where s is the inter-electrode spacing and F is the force exerted on the ion for a time dt.

The apertured electrodes B and D are connected to an output 62 of generator 61 which supplies them with a first voltage waveform of period T, where T is. for example, 8 × 10^-7 secs. This waveform consists, at the start of each period T, of a positive pulse 20 of voltage magnitude V₁, typically 30 volts relative to the voltage V_a on electrode A, as shown in Figure 2.

During each period T, the apertured electrode C is supplied from an output 63 of generator 61 with a second voltage waveform

comprising a train of equal-width voltage steps or pulses 22, also as shown in Figure 2. These pulses are also relative to the potentialV_a on electrode A. The first and last of pulses 22 have equal but opposite polarity magnitudes V₂, which may in certain circumstances be equal in magnitude to the pulses 20 applied to the electrodes B and D. As can be seen from Figure 2, the second and penultimate pulses of pulse train 22 have, respectively, a magnitude of half that of the first and last pulses of pulse train 22, and pulses

intermediate the second and penultimate pulses have amplitudes which change in a linear manner so as to effectively create a linear voltage ramp from the voltage -V₂/2 to the voltage +V₂/2. The leading edge of the first pulse of the pulse train 22 and the leading edge of the pulse 20 substantially coincide, and the trailing edge of the last pulse of the pulse train 22 substantially coincides with the leading edge of the next pulse 20. The pulses 20 and 22 give rise to pulses of electrostatic force within the electrode cell 10 so arranged in time and amplitude that they act upon the ions

accelerated by the DC potential V_a on apertured electrode A in such a way that ions of a selected velocity will attain a greater energy than ions of a different velocity. The pulse amplitudes and the positions of the pulses with respect to time are thus optimised with respect to ionic mass resolution and sensitivity. The pulse

amplitude distribution with respect to time on grid C is optimised to enable ions of the selected velocity to attain sufficient energy that the signal provided by these ions will be well above the noise level in the system, but to minimise the energy attained by faster ions.

Further discrimination between selected and non-selected ions is provided by the pulses 20 on apertured electrode B. The

electrostatic force provided by these pulses (either attracting or repelling) on the ions passing between apertured electrodes B and C is, at any time instant, given by the force due to the pulse 20 (if any) on electrode B, and the force due to the pulse train 22 on electrode C. This combination of forces discriminates against those ions which are faster, but arrive at electrode B later than the ions of the selected ion mass of interest. The pulse 20 on electrode D discriminates in a similar manner against those ions which are slightly slower than the ions of the selected ion mass of interest.

An accurate analysis of the spectrometer performance can be obtained by deriving mathematically the energy gained by the ions passing through the electrode cell.

To further improve resolution a number of tetrode cells may be placed in series. If this is done the distance d between cells should be chosen to satisfy the expression d = hs where h is unity or any odd integer greater than unity, and s is the mutual electrode spacing in cells, if the same energization signal generator 61 is to be used to energise each cell. (With other distances d separate generators would have to be used for the various cells, these generating energization signals which are appropriately phase-displaced with respect to their counterparts). Preferably the said odd integer is substantially greater than unity; it may have a value of, for example, 27. If this is the case ions travelling between two cells so spaced and which have slightly different velocities will arrive at the second cell at appreciably different times, so that only ions which have a predetermined velocity will be preferentially accelerated by the second cell. This can lead to improved resolution, the velocity filtering action of the second cell being enhanced because of the spacing. In one practical construction of a mass spectrometer including two electrode-cells spaced according to the above the same generator 61 was used to energise each cell so that the same voltages, with the exception of some modification due to the introduction of respective d.c. bias voltages, were applied to corresponding electrodes of the two cells. The voltage V applied to the first electrode A of the first cell (the cell nearest the ionisation chamber 4) was variable between +25 volts and +250 volts and the said respective d.c. bias voltages (relative to V_a) were: First cell electrode B - 1.3 volts

First cell electrode C - 6.6 volts

First cell electrode D -11.9 volts

Second cell electrode A -11.9 volts

Second cell electrode B -13.2 volts

Second cell electrode C -18.5 volts

Second cell electrode D -23.8 volts

Figure 3 shows a possible construction for the energization voltage generator 61 of Figure 1. In Figure 3 a digital counter 70 is driven by a clock pulse generator 71 of period t₁ so that counter 70 repeatedly cycles through all its possible count values. The parallel output 72 of counter 70 is connected both to the input 73 of a decoder 74 and to the digital signal input 75 of a

multiplying digital-to-analogue converter 76. Respective terminals 77A and 77B of the analogue signal input of converter 76 are supplied with voltages -V₂/2 and +V₂/2 relative to the voltage V_a by

means of voltage sources 78 and 79. The voltage V_a is generated by a voltage source 80 provided with means 81, for example a control knob controlling a potentiometer, for varying the voltage V_a applied to the terminal 60. The voltage V₁ relative to V_a is produced by a voltage source 82 and can be applied to the output 62 by control of a controllable change-over switch 83 to the position shown. In the other position switch 83 supplies the voltage V_a to output 62. A voltage -V₂ relative to V_a is produced with the aid of a V₂/2 voltage source 84 connected in series with V₂/2 voltage source 78, and can be applied to the output 63 by closure of a controllable switch 85. Similarly a voltage +V₂ relative to V _a is produced with the aid of a V₂/2 voltage source 86 connected in series with V₂/2 voltage source 79, and can be applied to the output 63 by closure of a controllable switch 87. The analogue output 88 of converter 76 can be connected to the output 63 by closure of a controllable switch 89.

Decoder 74 has an output 90 at which it is arranged to produce a logic "1" each time the count in counter 70 is zero, and an output 91 at which it is arranged to produce a logic "1" each time the count in counter 70 is the maximum. Output 90 is connected to the control inputs 92 and 93 of switches 83 and 85 respectively so that switch 83 adopts the position shown and switch 85 closes each time the count in counter 70 is zero. Output 91 is connected to the control input 94 of switch 87 so that this switch closes each time the count in counter 70 is the maximum. The outputs 90 and 91 are connected to the control input 95 of switch 89 via a NOR-gate 96 so that switch 89 is closed at all times except when the count in counter 70 is zero or the maximum. Thus, each time the count in counter 70 is zero the outputs 62 and 63 receive the voltages V₁ and -V₂ respectively

(corresponding to the start of a period T in Figure 2). When the count in counter 70 subsequently increases towards the maximum output 62 receives zero volts relative to V_a whereas output 63 receives a voltage ramp which increases in a series of voltage steps

substantially linearly from substantially -V₂/2 to substantially +V₂/2 from the output 88 of converter 70, and when the count in counter 70 is the maximum outputs 62 and 63 receive the voltages zero and +V₂ relative to V_a, respectively, as required.

It should be noted that the duration of the V₁ pulse at output 62 is not necessarily the same as that of the -V₂ pulse at output 63. It may, for example, be longer, but in any case its leading edge preferably substantially coincides with that of the -V₂ pulse .

The hot filament 6 shown in Figure 1 may be replaced by a field emission cathode consisting of a silicon preform having a number of field emitting points. By the use of microengineering techniques a control grid is placed at a few microns distance from the field emitting points. Application of a positive potential to the control grid causes emission of electrons from the points. However,

application of a negative potential of sufficient value to the pointsto cause a field of approximately 10⁸ volts/cm, causes directly field ionisation of the gas molecules of the substance under investigation in the ionisation chamber 4.

Electron ionisation causes fragmentation of long chain molecules of the substance being investigated. Different molecules can end up with daughter products of similar mass to charge ratio as other parent molecules, which can give rise to ambiguities in the results obtained from the spectrometer. Field ionisation tends to remove only one electron from the molecules. Hence, by alternating the potential applied to the control grid between appropriate positive and negative values, the field emission cathode can be switched between the two ionisation modes, i.e electron and field ionisation, making it possible to obtain more unambiguous information about the parent molecules than would be provided by either mode alone.

The configuration of the mass spectrometer according to the present invention, enabling a microminiature device to be achieved, enables such field emission cathodes to be used, providing a further advantage over known forms of spectrometer of conventional size.

Normal methods of fabrication for spectrometers involve

machining of individual parts, welding and spot welding, soldering and either glass blowing or machining the outer container or housing for the ionisation chamber and electrode cell. Furthermore, normal methods of fabrication present additionally problems in grid

electrode alignment which affects both sensitivity and resolution.

A method of manufacturing a mass spectrometer according to the invention, which is described below for a two electrode-cell device, substantially alleviates the problems associated with normal

fabrication methods. It should be realised that although the

following description relates to a two cell device, longer devices incorporating a greater number of tetrode cells can be made by the same techniques.

Figure 4 is a longitudinal section through the electrode

structure of a mass spectrometer in accordance with the invention which is manufactured as follows. A silicon slice, for example of 2cm thickness and at least 7.5cm diameter, and whose major face is parallel to the (110) crystallographic plane is cut into the form of a rectangle, for example, 2cm × 5.4cm long orientated so that the end faces 31 of the resulting rectangular block 30 are parallel to the (111) crystallographic plane.

A quartz substrate or base plate 32 slightly larger than the silicon block and, for example, of 4mra thickness is prepared by cutting two series of four holes 34 through it. The holes 34, which are conveniently of circular section and the pitch of each series of which corresponds to the required pitch, e.g. 0.2 or 0.3cm, of the cell electrodes A, B, C and D (see Figure 1), can be formed, for example, by ultrasonic drilling, drilling with a rotary diamond drill, or etching using standard photo lithographic techniques.

The rectangular silicon block 30 is then bonded to the quartz base plate 32 using known bonding techniques, after which it is cut, such as with a diamond impregnated saw blade, to provide a plurality of trenches 36 (seen end-on in Figure 4) extending through the silicon member and slightly into the quartz base plate 32. This cutting operation provides a number of mutually isolated silicon plate members 38 which, ultimately, form the grid electrodes A,B,C,D of the electrode cell 10 shown in Figure 1. The cuts in the silicon member 30 are arranged so that the silicon plate members 38 are placed centrally over the holes 34 in the quartz base member 32. The silicon remaining outside the cut areas, indicated as blocks 40 in Figure 4, is then removed by ultrasonic drilling to leave, in the two electrode cell embodiment shown, two groups of four silicon plate members 38 extending from and substantially normal to the quartz base member 32 (and normal to the plane of the paper). The assembly of silicon plate members 38 and quartz base member 32 is then etched in an anisotropic etch, such as EDP, which etches the relatively rough surfaces on the silicon plate members 38, resulting from the sawing operation, back to the (111) crystallographic plane.

The assembly is then ovidised; for reasons which will become apparent from the description below, such as by a low temperature CV (chemical vapour) deposition process. The silicon plates 38 are then drilled, such as by laser drilling, to provide an array of holes 42 in each of the silicon plates 38. Two holes 42 are shown in each plate 38 in the section of Figure 4 although many more are actually present outside the plane of the paper. Because the silicon plates are bonded to the quartz base member 32, the alignment of the holes 42 in each plate 38 will be very accurately defined. The perforated silicon plates act as the apertured electrodes of the two electrode cells provided in the spectrometer of the present example and, hence, the above procedure for defining the grid electrodes overcomes the previously referred to grid electrode alignment problems associated with known mass spectrometer arrangements. The oxide coating is then removed in buffered HF (hydrogen fluoride) which also removes any debris on the silicon plates 38 resulting from the laser drilling of the holes 42.

A cover member 48 as shown in perspective view in Figure 5 is machined from quartz, this having dimensions such that it co-operates with the quartz base member 32 to provide an enclosure for the silicon plates 38. It can be seen from Figure 5 that gold

electrodes, 8 and 12, are provided on selected inner surfaces of the quartz cover member 48. The gold electrodes 8 and 12 act,

respectively, as the electrode for the ionisation chamber 4 and the target electrode for detecting ions of the critical ion mass. It can be seen with reference to Figure 1 that the gold electrode 12 is arranged to receive a voltage potential from a suitable source 50, the potential being of appropriate polarity to ensure that the electrode 12 has a repelling effect on ions passing through the electrode cell 10. Because only those ions of the selected ion mass pass through the electrode cells 10 with maximum energy only these ions will have sufficient energy to overcome the repelling force of the electrode 12 and thus strike the electrode 12, the ions of allother masses in effect being trapped between the last grid electrodeD of electrode cell 10 and the detector electrode 12. The cover member 48, which is also drilled at one end to provide the pumping port for allowing the substance under investigation to be introduced into the ionisation chamber, is sealed to the quartz base member 32 using any suitable bonding process, such as a thermo electric technique. Contact wires 50, (Figure 4) are then connected to the silicon plate members 38 through the holes 34 in the quartz base member 32. Typically, the contact wires 50 will comprise gold and the electrical contact to the silicon plate members 38 will be achieved by embedding the end portions of the contact wires into small quantities of a suitable material, such as indium, which is then compressed into the holes 34 in the quartz base plate 32 to form contact members (not shown in Figure 4). As well as an electrical connection, the contact members also act to seal the holes 34. If an ion transparent window, such as shown by dotted line 16 in Figure 5, of typically, silicon oxide or silicon nitride, is positioned between the ionisation chamber 4 and the first electrode cell 10, the apertured electrodes A,B,C,D, constituted by the silicon plate members 38, can be maintained in a vacuum and, therefore, free of contaminating influences of the substances introduced into the ionisation chamber. The filament 6 can be sealed in the quartz base member 32 using any known technique.

It will be realised from the above description that the present invention provides a mass spectrometer including an electrode cell having at least four grid electrodes and which, by the use of pulses arranged appropriately in amplitude and time, optimises the energy gain attained by ions of a selected critical mass, thereby providing enhanced sensitivity and resolution. Furthermore, the spectrometer can be fabricated in a small size with low weight. Additionally the method of fabrication can provide cheaper construction, in comparison with known devices, and give rise to greater accuracy of aperture alignment, further enhancing performance.

Although the present invention has been described with respect to a specific embodiment it should be realised that many modications can be effected without departing from the scope of the invention as defined by the claims.

For example, the spectrometer may include only one or more than two sets of silicon plate members, i.e. electrode cells. Also, any or all of the electrode cells may include more than four electrodes, appropriately pulsed to discriminate against those ions not of the selected critical mass.

As another example, the waveform applied to electrode C may be generated by analogue rather than digital circuitry.

Claims

1. A mass spectrometer comprising an ionisation chamber, an ion detector electrode for receiving an ion-repelling bias voltage, an electrode cell for preferentially accelerating ions having a predetermined mass and velocity from said ionisation chamber towards said ion detector electrode, which cell comprises first, second, third and fourth apertured electrodes spaced in that order along a path for ions from said ionisation chamber to said ion detector electrode, and energization voltage generator means connected to said first, second, third and fourth electrodes for applying a first ion-attracting d.c. bias voltage to said first electrode, a first periodic voltage waveform to said second and fourth electrodes, and a second periodic voltage waveform to said third electrode, the first and second voltage waveforms having periods which are equal to each other, the first voltage waveform being ion-attracting during at least part of a first half of each said period, and the second voltage waveform being ion-repelling during at least part of said first half of each period and being ion-attracting during at least part of a second half of each period.

2. A mass spectrometer according to Claim 1 wherein said second voltage waveform is shaped to progress during each said period from an ion-repelling voltage V₂ of duration t₁ at the beginning of said first half of the corresponding period to an ion-attracting voltage V₂ of duration t₁ at the end of said second half of the corresponding period via a step from the ion-repelling voltage V₂ to an ion-repelling voltage V₂/2 substantially linear change from the ion-repelling voltage V₂/2 to an ion-attracting voltage V₂/2, and a step from the ion-attracting voltage V₂/2 to the

ion-attracting voltage V₂.

3. A mass spectrometer according to Claim 2, wherein said second vo1tage was form is shaped as a succession of voltage Step .

4. A mass spectrometer according to Claim 3, wherein said voltage steps all have the duration t₁ .

5. A mass spectrometer according to Claim 1, wherein said first voltage waveform comprises, during each said period, an

ion-attracting voltage pulse having a leading edge which

substantially coincides with the beginning of the first half of the corresponding said period, said voltage pulse terminating before the end of said first half.

6. A mass spectrometer according to Claim 1, wherein said first and second voltage waveforms are relative to said d.c. bias voltage.

7. A mass spectrometer according to Claim 6, including means for varying said d.c. bias voltage.

8. A mass spectrometer according to Claim 1 comprising a housing which defines the ionisation chamber and houses the electrode call, the housing comprising a quartz base and a quartz cover, said apertured electrodes comprising apertured silicon plates which are bonded to and extend substantially normal to the quartz base and extend substantially normal to said path, the quartz base having a plurality of apertures which accommodate electrical conductors which connect the energization voltage generator means to said plates.

9. A mass spectrometer according to Claim 8, wherein the ion detector electrode is formed on a wall of the cover and the

ionisation chamber includes an electrode formed on part of the base.

10. A mass spectrometer according to Claim 1, including a second such electrode cell spaced from the cell specified in Claim 1 along said path, said second electrode cell having energization voltage generator means connected to its first, second, third and fourth electrodes for applying an ion-attracting d.c. bias voltage to its first electrode, a periodic voltage waveform to its second and fourth electrodes, and a periodic voltage waveform to its third electrode, these periodic voltage waveforms having periods which are identical to each other, the voltage waveform applied to its second and fourth electrodes being ion—attracting during at least part of a first half of each of these periods, and the voltage waveform applied to its third electrode being ion-repelling during at least part of the said first half of each of these periods and being ion-attracting during at least part of a second half of each of these periods.

11. A mass spectrometer according to Claim 10, wherein the first, second, third and fourth electrodes of the electrode cell specified in Claim 1 are positioned with a pitch d along said path, and the first, second, third and fourth electrodes of said second electrode cell are also positioned with the pitch d along said path, wherein the electrode cell specified in Claim 1 is spaced from the second electrode cell by an odd number, greater than unity, times the pitch d, and wherein the period of the said voltage waveforms for the second, third and fourth electrodes of the electrode cell specified in Claim 1 is the same as, and synchronized with, the period of the said voltage waveforms for the second, third and fourth electrodes of said second electrode cell.