EP0919067B1

EP0919067B1 - Charged particle velocity analyser

Info

Publication number: EP0919067B1
Application number: EP97936783A
Authority: EP
Inventors: Timothy Andrew 227 Avenue Road Extension STEELE; Ronald Charles Unwin; Adrian John Eccles
Original assignee: Millbrook Instruments Ltd
Current assignee: Millbrook Instruments Ltd
Priority date: 1996-08-17
Filing date: 1997-08-18
Publication date: 2000-11-29
Anticipated expiration: 2017-08-18
Also published as: GB9617312D0; WO1998008244A2; EP0919067A2; WO1998008244A3; DE69703624D1; DE69703624T2

Description

This invention relates to a charged particle velocity analyser and to a method of analysing the velocity of charged particles.

The velocity of a charged particle is determined by its mass and its energy. For a mono-energetic beam of particles the results of the velocity analysis can be inverted to give a mass spectrum. Mass spectroscopy is a widely used analysis technique. For a beam of particles of one mass eg. electrons, the results of the velocity analysis can be used to determine the energy spectrum. Electron energy spectroscopy is a widely used technique.

One type of mass analysis is known as time-of-flight (ToF) analysis. Other methods of analysis are quadrupole analysis and magnetic sector analysis.

Quadrupole analysis has the disadvantage that very precise mechanical and electronic components are required. A particular disadvantage is that particle transmission in the apparatus decreases with increasing mass.

Magnetic sector mass analysis apparatus requires magnets which are often large and heavy. The size and weight of the magnets detracts from the utility of this type of apparatus.

Quadrupole and magnetic sector instruments have the further disadvantage of processing information on a particular mass sequentially, rather than all masses simultaneously, which is very time consuming.

Previous time-of-flight instruments have the benefit of being mechanically simple but have the disadvantage of having a low duty cycle. The duty cycle is the proportion of time of operation of the instrument during which data is recorded. A typical duty cycle for previous time-of-flight instruments is less than 0.1 %. The undesirable result is that long overall collection times are necessary and a large proportion of particles from a continuous source are wasted.

Two previous types of time-of-flight instruments have used a frequency modulation of the ion beam. In a first of these instruments many frequencies were encoded in one sequence. The data from the detector were deconvoluted by gating the detector. Gating selects a particular time-of-flight from the many times-of-flight present in the detected signal. Consequently, much of the signal is not used and is wasted. Also, the procedure must be repeated for each time of flight, which may take an undesirably long time.

A second of the previous time-of-flight instruments modulated the signal with a sine wave. The frequency was changed in a stepwise fashion and the detector was gated for a particular frequency, as with the last mentioned instrument. This second time-of-flight instrument also had the disadvantage of processing a single time-of-flight at a time.

It is an object of the present invention to address the disadvantages mentioned above.

According to one aspect of the present invention a time-of-flight velocity analyser comprises modulation means for imposing a binary sequence on a sample of charged particles, a detector for detecting the arrival of said charged particles thereat and producing an output signal, and transformation means for transforming the signal from the detector with the aid of the binary sequence to provide time-of-flight values for a spectrum of velocities; characterised in that the binary sequence is a substantially random binary sequence and that the detector is arranged to detect substantially all of the modulated charged particles and the transformation means is arranged to transform and deconvolute the detector signal to provide time of flight values for a spectrum of velocities substantially simultaneously.

The substantially random binary sequence may be a pseudo random binary sequence or a random binary sequence.

The transformation means may be deconvoluting means for deconvoluting the detector signal or may be cross-correlation means.

The cross-correlation means may be arranged to provide time-of-flight values for a spectrum of velocities substantially simultaneously.

The charged particles may be ions. The charged particles may be electrons.

The velocity analyser may be a mass spectrometer. The mass spectrometer may be a secondary ion mass spectrometer.

The velocity analyser may be an electron energy analyser.

In the case where the charged particles are ions, the mass spectrometer may include a drift tube having a drift potential in the range 5eV to 800eV. Preferably, the drift potential in the drift tube is in the range 20eV to 200eV.

The mass spectrometer may comprise a drift tube for which the time of flight values are in the range 0.1µs to 1000µs. Preferably, the time of flight values are in the range 1µs to 500µs.

The velocity analyser may comprise a drift tube for which the velocity of the particles is in the range 100 m/s to 1,000,000 m/s. Preferably the velocity of the ions is in the range 1,000 m/s to 500,000 m/s.

The sample of charged particles may comprise a beam of charged particles, which may be produced by collimation and focusing of the charged particles. The charged particles may be focused on the detector. The charged particles may be focused by one or more lenses. The lenses may comprise at least one reflecting lens. The lenses may comprise at least one electrostatic lens.

The modulation means may comprise means for deflecting said sample of charged particles from the detector. The sample beam direction may be changed by electro-static means, which electro-static means preferably comprise a parallel plate deflector, but may comprise a single plate deflector or an electro-static lens or a grid of fine wires. The sample may alternatively be deflected by magnetic means.

The modulation means may comprise an array of charge carrying members, which may be wires. The charge carrying members may be substantially parallel to each other. The modulation means may be short in the direction of travel of the sample so as not to broaden the pulse length imparted to the beam. The charge carrying members may, in use, have a polarity of charge with respect to the charged particles which is arranged to result in substantially no electric field from the charge carrying members outside the modulation means. The array may comprise a plurality of layers of charge carrying members, which may be transversely offset relative to one another, to thereby form a lens.

The sample may comprise a pulsed beam of charged particles. The length of each pulse may be determined by the substantially random binary sequence. The length of each pulse may be a multiple of a time unit of the binary sequence. The length of each pulse may be determined by a trigger from a timing means.

The timing means may provide a trigger to a sequence generator, which sequence generator may provide the substantially random binary sequence. The timing means may be arranged to pulse with a period of between 0.01µs and 30µs. Preferably the timing means is arranged to pulse with a period of between 0.1µs and 3µs.

The sequence generator may be arranged to generate, in use, a sequence having M digits, where M is an integer greater than 3. Preferably, the sequence will be a binary sequence having 2^N digits, where N may be an integer. Preferably, the sequence generator is arranged to generate, in use, a binary sequence of one or more of 128 digits, 256 digits, 512 digits, 1,024 digits, 2,048 digits and/or 4,096 digits. Most preferably, the sequence generator is arranged, in use, to generate a sequence of 1,024 digits.

The velocity analyser may include a transient recorder for recording the signal produced by the detector.

The detector may be arranged to pass a signal to the transient recorder on direct detection of the sample. Preferably the detector may comprise a Faraday cup.

Alternatively the detector may be arranged to pass a signal to the transient recorder on detecting a secondary cascade of charged particles caused by the sample. Suitably, the detector may comprise a channel electron multiplier or an electron multiplier array/multi-channel plate.

The transient recorder may record the signal from the detector in a plurality of time channels. The size of the time channels may be determined by the timing means. Each time channel may be defined by the time between successive digits of the substantially random binary sequence. Alternatively, the time channel may be defined by other means. The transient recorder may be arranged to add signals from the detector resulting from a repeat of the substantially random binary sequence to the values recorded in previous cycles of the substantially random binary sequence. The transient recorder may be arranged to integrate the signals recorded in each channel for a plurality of repeats of the substantially random binary sequence. The transient recorder may comprise a digital signal processing (DSP) chip. Said DSP chip may be operatively linked to a computer. The transient recorder may comprise a peripheral component interconnect (PCI) card, which may be operatively linked to a computer.

The cross-correlation means may be arranged to use a mathematical function to separate the substantially random binary sequence from a time-of-flight spectrum of the sample of charged particles. Said mathematical function may be performed by computing means, which computing means may be a personal computer. The mathematical function may comprise involving a transformation of a signal from the frequency domain to time domain. Said transformation may be a Fourier transformation. Said transformation may be a fast Fourier transformation. Said transformation may be a Hadamard transformation.

A method of analysing the velocity of a sample of charged particles comprises modulating the sample with a binary sequence, detecting the arrival of the modulated sample of charged particles at a detector and cross-correlating a signal produced by the detector with the binary sequence to provide time-of-flight values for a spectrum of velocities; characterised in that the binary sequence is a substantially random binary sequence and that the detector is arranged to detect substantially all of the modulated charged particles and a transformation means is arranged to transform and deconvolute the detector signal to provide time of flight values for a spectrum of velocities substantially simultaneously.

The method may include cross-correlating the signal produced by the detector with the substantially random binary sequence to provide time-of-flight values for a spectrum of velocities substantially simultaneously.

The method may include the charged particles beam being accelerated along a drift tube, which acceleration may be caused by a drift potential in the range 5eV to 800eV. Preferably the charged particle sample is accelerated with a potential of 20eV to 200eV.

The sample of charged particles may be produced by secondary ionisations.

All of the above aspects may be combined with any of the features disclosed herein in any combination.

Specific embodiments of the present invention will now be described with reference to the accompanying drawing, in which:

Figure 1 is a functional diagram of a charged particle velocity analyser;

Figure 2 is a schematic side view of a first embodiment of a drift tube of the charged particle velocity analyser;

Figure 3 is a schematic side view of a second embodiment of a drift tube of the charged particle velocity analyser;

Figure 4a is a partial schematic side view of an embodiment of modulation means;

Figure 4b is a partial schematic front view of the embodiment of modulation means; and

Figure 4c is a partial schematic top view of the embodiment of modulation means.

An ion velocity analyser 10 comprises a source 12 of ions, which source is in a vacuum. The ions are focused into a beam by a series of electro-static lenses in the focus section 14 of the ion velocity analyser 10. The beam is modulated by a deflector 16 to give a pseudo-random binary signal having pulses of variable length. The modulated beam travels down an evacuated drift tube 18 having a chosen electro-static potential to a detector 20. The signal from the detector 20 is amplified and the signal intensity for each of an array of time channels is measured and recorded by a multi-channel analyser 22. The resultant signal is a convolution of the pseudo-random binary signal and the time-of-flight spectrum of the ion sample. The signal from the detector 20 is cross-correlated with the modulated signal and de-convoluted to give a time-of-flight spectrum for the ion sample. The time-of-flight spectrum is then inverted to give a mass spectrum.

The operation and construction of the ion velocity analyser 10 will now be described in greater detail.

The ions in the source 12 are created by a high energy ion beam directed onto the surface of a sample for analysis. The high energy ion beam ionises the surface of the sample. The resulting ions are used in the analysis. Alternative methods may be used to create a source of ions, including electron impact ionisation for gases (in which the velocity analyser would be a residual gas analyser giving mass spectra of the vapour phase particles); plasma ionisation for gases, liquids and solutions; high energy neutral beam ionisation for solids; glow discharge ionisation for solids; laser irradiation for solids; radioactive ionisation; secondary ionisation from ion beams; and other related methods. The ions from the sample are then collimated and focused to form a beam by a series of electro-static lenses in the focus section 14. In this example the energy spread of the ions in the beam is controlled to be a small fraction of the drift tube potential by using an energy filter.

Next, the beam of ions passes to the deflector 16 which can take a number of various forms, such as a parallel plate deflector, a single plate deflector, an electro-static lens, a series of electro-static fields formed by a grid of fine wires or a deflector using magnetic fields. The ions are deflected from being focused towards the detector 20. The deflector 16 is controlled to create a beam which is given a pseudo-random binary modulation. A clock 24 is used to set the width of the time channels to between 0.1µs and 3µs, the clock pulses with this period. A pseudo-random digital sequence is provided by a sequence generator 26 on being triggered by the clock 24. The digital sequence is combined with the pulsed signal provided by the clock 24 to give a signal having a known sequence of ON/OFF pulses of known duration. This signal is amplified by an amplifier 28 and is used to control the deflector 16 by varying the voltage thereof according to the signal. The beam is deflected so that it misses the detector 20 or is left un-deflected so that the beam hits the detector 20. The beam of ions therefore comprises pulses of ions of varying lengths. The lengths of the pulses will be multiples of the channel width.

The number of ON/OFF digits is chosen to be 2^N, where N is an integer, so that the de-convolution can be performed efficiently. The cross-correlation function used in this example is a fast Fourier transform which can be performed considerably more quickly if the number of data points is equal to 2^N, where N is an integer. In this example the number of cycles is 1,024 but others could be used such as 128, 256, 512, 2,048 or 4,096.

Further focusing of the beam is then carried out after it has passed through the deflector 16, however this may not be necessary for different sources of ions. The further focusing may be carried out in a number of ways but two examples are given in figures 2 and 3.

In figure 2 three spaced lenses 36 are located in the drift tube 18. The lenses may be electrostatic and may be controlled separately to focus the beam of ions.

Figure 3 shows a second example of further focusing. In this example the drift tube 18 comprises first and

second sections

38, 40 which form an angle. A lens 42 is located at the entrance to the flight tube 18. At the apex of the angle between the first and

second sections

38, 40 of the flight tube 18 there is a reflecting lens 44. Ions travelling down the first section 38 enter the reflecting lens 44 and are focused and reflected down the second section 40 to a further lens 46 and on to the detector 20. This example has the benefit of reducing flight time broadening due to energy spread.

The pulse beam then passes into the drift tube 18, which has a drift potential in the range 20 to 200eV. The drift potential determines the flight time of the ion when combined with the tube length and the ion mass. The mass is proportional to the square of flight time. In this embodiment the drift tube 18 is approximately 50 cm long.

Un-deflected ions will arrive on the detector 20 and thus form an electric current flow in the detector 20. The detector output signal is passed to an amplifier 30 which, typically, provides an output voltage of 1V for a detected drift tube current of 1nA. The amplification may be varied by a factor of 1,000 though. Various detectors can be used with this ion velocity analyser, in this example a Faraday cup is a used. Alternatively, an electron multiplier or electron multiplier array could be used to pre-amplify the ion current before the amplifier 30 by secondary cascade in the vacuum from 1pA to 1µA or 1nA respectively. This current is then passed to the amplifier 30 which provides an output voltage of 10mV for a detected cascade current of 1nA. The signal is then passed to an analogue to digital converter 32 so that it can be analysed in the multi-channel analyser 22, which requires a digital input. The multi-channel analyser 22 in this example is a digital signal processing chip but this could be replaced by various equivalents such as a PCI card.

The multi-channel analyser 22 receives the pulses from the clock 24, which pulses are also received by the sequence generator 26, as mentioned above. For each pulse, which comprises one channel of the 1,024 channels, the intensity of the signal from the amplifier 30 is recorded by the multi-channel analyser 22. Consequently, after one cycle of the pseudo-random signal 1,024 values of the detector current will be stored.

The cycle of 1,024 pulses is repeated and the intensity of the signal is added to the previously stored signal for a particular channel. The cycle of 1,024 pulses is repeated typically 256 times. A cumulative signal is thus built up as the 1,024 pulse cycle is repeated. The cycle is repeated until the signal to noise ratio is sufficiently high for the system.

The cumulative data for the 1,024 channels are then cross-correlated with the output signal from the sequence generator 26 by a computer 34. The correlation function used in this example is a fast Fourier transform. A Hadamard transform may provide an alternative correlation function. The result is the intensity of the signal corresponding to each flight time divided into the clock timing units. This signal is then processed in the computer 34 to convert clock pulse units into time and convert the time-of-flight into its corresponding mass. As mentioned above, a time-of-flight corresponds to a particular mass for a given drift potential and length of drift tube 18.

If a particular mass range is of interest then the correlation may be carried out on only a subset of the 1,024 channels. This will reduce the amount of time taken to perform the cross-correlation.

Several transformed results, up to 100 or more, may be added together to produce a spectrum which shows low intensity peaks more easily than a single transformed result.

The modulation of the beam may be a frequency modulation or an amplitude modulation.

An alternative modulation means 50 is shown in figures 4a, b and c. The alternative modulation means 50 comprises a three layer array of wires 52. The wires are arranged transverse to the direction of particle travel, in this example in a vertical orientation. All of the wires are substantially parallel giving an increase in transmission over a grid of wires. Each layer 54 of the array 52 comprises wires with a spacing (x) of 0.2-1.5mm. The second layer 54b is offset transversely relative to the first layer 54a by a distance of approximately half the wire spacing (x). The distance between adjacent layers is 0.15-1mm. The wires of the third layer 54c are substantially aligned in the direction of particle travel with the wires of the first layer 54a. The wires of the array have the drift tube potential, to cause deflection the second layer 54b of the array has a potential difference to the drift potential.

A further alternative modulation means is to have a single layer of parallel wires in which the adjacent wires have opposite polarity but equal potential difference from drift tube potential to cause deflection.

The alternative modulation 50 means functions in the same manner as described above. Additional benefits, however, include the feature that the array of wires 52 functions as a micro lens and is uniform in effect over several mm² area of the beam. Also, the use of an array means that overall length of the modulation means can be shortened giving shortened transit times and pulse periods. The required voltage of such an array is also lower from that of a corresponding square grid, making it easier to manufacture and operate.

An alternative to the detector described above is to use multiple synchronous detectors. This is achieved by using a sequence generator to modulate the signal, which signal then arrives at the detector at various times depending on the velocity of the ions. A delay line is used to delay the modulation for a chosen flight time to be measured. The signal at the detector is multiplied by the modulated and delayed output of the delay line. The multiplication results in the signal being multiplied by 1 when the modulation is 1 and then by -1 when the modulation is zero. The average value of the output from the multiplication is the intensity of the signal at the delay time (time of flight of choice) determined by the delay line.

The delay line is arranged to have several taps each defining a different flight time. Each tap has its own multiplier and averager. Any number of taps can be used so that, in essence, the number of "tap, multiplier, averager" sets is the number of detector channels. The delay line can be a shift register and the multipliers can be simple +/- devices, the averager can be as simple as a resistor/capacitor pair (low pass filter). The multiple synchronous detector that works on all arrival time channels is simply carrying out a transformation by hardware rather than in a computer.

The ion velocity analyser disclosed herein provides significant advantages over prior art devices. The function of analysing a complete time-of-flight signal simultaneously, rather than one particular time-of-flight at a time, provides significant benefits in terms of the speed of analysis of a particular sample. Using a relatively low drift potential of 20 to 200eV compared to the more usual 3,000 to 10,000eV enables a variety of components to be used which, although decreasing the accuracy of the ion velocity analyser, result in considerable cost savings in the manufacture and maintenance of the device, as well as a more robust device, over prior art devices. For instance, a Faraday cup detector is cheaper than and has a considerably longer life span than the usual multiplier or multi-channel plate. The use of longer channel times and the resulting reduction of the amount of information enables the use of a PC to process the signal and provide a mass spectrum in a matter of seconds.

The ion velocity described herein offers a higher duty cycle than a previous time-of-flight analyser. Additionally, inexpensive components can be used.

Previous instruments have deconvoluted the signal by gating before the detector. The advantage of deconvoluting the signal after the detector, using a computer (as is the case here), is that a much higher duty cycle can be used together with parallel collection of flight times.

The above advantages may be exploited either by shorter collection time for a particular spectrum, collection of very weak spectra or collection of repetitive spectra with a high repetition rate.

Claims

A time-of-flight velocity analyser comprises modulation means (16) for imposing a binary sequence on a sample of charged particles, a detector (20) for detecting the arrival of said charged particles thereat and producing an output signal, and transformation means (34) for transforming the signal from the detector (20) with the aid of the binary sequence to provide time-of-flight values for a spectrum of velocities; characterised in that the binary sequence is a substantially random binary sequence and that the detector (20) is arranged to detect substantially all of the modulated charged particles and the transformation means (34) is arranged to transform and deconvolute the detector signal to provide time of flight values for a spectrum of velocities substantially simultaneously.
A time-of-flight velocity analyser as claimed in claim 1, in which the substantially random binary sequence is a pseudo random binary sequence.
A time-of-flight velocity analyser according to either claim 1 or claim 2, in which the transformation means (34) are cross-correlation means.
A time-of-flight velocity analyser as claimed in claim 3, in which the cross-correlation means are arranged to provide time-of-flight values for a spectrum of velocities substantially simultaneously.
A time-of-flight velocity analyser as claimed in any preceding claim, in which the time-of-flight velocity analyser is a mass spectrometer.
A time-of-flight velocity analyser according to any one of claims 1 to 4, which is an electron energy analyser.
A time-of-flight velocity analyser as claimed in either claim 5 or claim 6, which includes a drift tube (18) having a potential in the range 5eV to 800eV.
A time-of-flight velocity analyser as claimed in any preceding claim, in which the charged particles are focused by one or more lenses (36).
A time-of-flight velocity analyser as claimed in any preceding claim, in which the modulation means (16) comprises an array of charge carrying members (54) for deflecting the sample of charged particles from the detector (20).
A time-of-flight velocity analyser according to any one of the preceding claims, which includes a transient recorder (22) for recording the signal produced from the detector (20).
A time-of-flight velocity analyser according to claim 10, in which the detector (20) is arranged to pass a signal to a transient recorder (22) on detecting a secondary cascade of charged particles caused by the sample.
A time-of-flight velocity analyser according to claim 11, in which the detector (20) comprises a channel electron multiplier or an electron multiplier array (30).
A time-of-flight velocity analyser according to any one of claims 10 to 12, in which the transient recorder (22) is arranged to add signals from the detector resulting from a repeat of the substantially random binary sequence to the values recorded in previous cycles of the substantially random binary sequence.
A method of analysing the velocity of a sample of charged particles comprises modulating the sample with a binary sequence, detecting the arrival of the modulated sample of charged particles at a detector (20) and cross-correlating a signal produced by the detector (20) with the binary sequence to provide time-of-flight values for a spectrum of velocities; characterised in that the binary sequence is a substantially random binary sequence and that the detector (20) is arranged to detect substantially all of the modulated charged particles and a transformation means (34) is arranged to transform and deconvolute the detector signal to provide time of flight values for a spectrum of velocities substantially simultaneously.