EP0708976A1

EP0708976A1 - Process for operating a time-of-flight secondary ion mass spectrometer

Info

Publication number: EP0708976A1
Application number: EP95920015A
Authority: EP
Inventors: Ewald Niehuis
Original assignee: ION-TOF GmbH; ION TOF GmbH
Current assignee: ION-TOF GmbH; ION TOF GmbH
Priority date: 1994-05-10
Filing date: 1995-05-10
Publication date: 1996-05-01
Anticipated expiration: 2015-05-10
Also published as: JPH09500486A; DE4416413C2; DE4416413A1; WO1995031000A1; EP0708976B1; US5633495A; JP3358065B2

Abstract

The invention pertains to a process for operating a time-of-flight secondary ion mass spectrometer for analysis of mass spectra, wherein a number of finely structured mass ranges appear in isolation at major intervals, involving the following steps: a) a surface of a material sample is bombarded with primary ion pulses that follow each other at regular time intervals tz (cycle time), b) the secondary ions of various masses m released from the material sample surface by the primary ions are accelerated to the same energy, c) the mas-dependent time of flight t is measured over a path 1 and the mass is determined therefrom. To increase the resolution and the signal-to-noise ratio the process is characterized in that: d) each primary ion pulse consists of a number of subpulses, e) each subpulse is so narrow that it allows for resolution of the finely structured mass ranges, f) the interval tB between the subpulses is greater than the width of the finely structured mass ranges, g) the number n of subpulses is selected so that n.tB is smaller than the intervals between the finely structured mass ranges, h) the n subpulse spectra of each finely structured mass range are added up.

Description

Method of operating a time-of-flight secondary ion mass spectrometer

The invention relates to a method for operating a time-of-flight secondary ion mass spectrometer for the analysis of mass spectra, in which several finely structured mass areas occur isolated at larger distances, with the following steps: a) a material sample surface is t at regular intervals t (cycle time ) bombarding successive primary ion pulses, b) the secondary ions of different masses m released from the primary ions from the material sample surface are accelerated to the same energy, c) the mass-dependent flight time t is measured via a path 1 and the mass is determined therefrom.

In the above method, the flight time t is proportional to the root of the mass (t proportional «/ in). The number of secondary ions which correspond to a specific mass m result in fine structure maxima in "nominal mass ranges" within a specified cycle time t at specific time intervals, the latter corresponding to an integer atomic or molecular weight of element or molecular ions. The amplitudes of the fine structure maxima allow a qualitative and quantitative analysis of the composition of the material sample surface. Time-of-flight secondary ion mass spectrometry, also known as TOF-SIMS (time-of-flight secondary ion mass spectrometry), is known (cf. Analytical Chemistry, 1992, vol. 64, p. 1027 ff and 1993, vol. 65 , P. 630 A ff.). It is used for the chemical analysis of solid surfaces.

The sample surface is bombarded with a pulsed primary ion beam of pulse duration t. Secondary ions are detached from the sample surface by the primary ion beam. The extracted secondary ions are accelerated to the same energy E in an extraction field (a few KeV). Then they pass through a flight path 1, at the end of which they are detected by means of a time-resolving detector. The majority of the secondary ions are simply charged.

For the flight time of the secondary ions: 1 t = 1 / v = • im = const • - / ÜT (1)

Thus, the exact mass ra of a secondary ion can be determined from the flight time t thus determined with the same energy E.

Depending on the desired mass range, the secondary ions are registered within a time interval, called the cycle time tz, after the time at which the primary ion pulse strikes, with the following relationship (1):

tz = const • - / m '(2)

mm -_- = maximum mass within the desired

Mass range After this cycle time has elapsed, the next pri ion pulse can hit the sample. The flight times are therefore measured with a repetition frequency f = 1 / t Z. Only very few secondary ions are generated and detected per cycle, typically 0.1 to 10. A mass spectrum with sufficient dynamics over several orders of magnitude, ie one Adequate ratio of highest to lowest intensity arises from the accumulation of the count events over a large number of cycles. The measuring time is then typically about 100 to 1000 s.

Both element ions and molecular ions are detached from the sample surface. The exact mass of a secondary ion species, which can be an element ion or a molecular ion, results from the sum of the atomic weights. The individual atomic weights deviate slightly from integer values due to the binding energy of the atomic nuclei, so that the aforementioned nominal mass ranges result on both sides of an integer value. The exact masses of element and molecular ions differ only slightly. As an example of secondary ion species of 27 u: aluminum ^"*" : 26.99154 u; C 2H3 ^"*" : 27.023475 u. The different secondary ion species can be separated from one another by a sufficiently high mass resolution, ie they can be broken down into fine structure maxima and elements and compounds can thus be detected separately. The separation of such different species is an essential prerequisite for the detection of traces of compounds and elements.

With the TOF-SIMS, the mass resolution m / Δ m describes the mass difference Δ m with a mass m that can still be separated into two fine structure maxima. It hangs crucially depends on the primary ion pulse duration t. Further factors for the separation are the resolving power of the time-of-flight analyzer and the time resolution of the detector and registration electronics, which, however, are not the subject of the invention.

The TOF-SIMS method is not only used to analyze the surface composition, but also allows the recording of lateral distributions of the various elements and connections with high spatial resolution in the sub-μm range. For this purpose, the primary ion beam is focused on a very small point and scanned over the sample with the aid of a deflection device. A mass spectrum is recorded and evaluated for each grid point. A distribution image (imaging TOF-SIMS) can then be generated from the results for a large number of raster points (typically e.g. 256 x 256). Furthermore, a depth distribution of the different species can be measured (depth profile) by sample removal using the primary ion beam or using an additional ion source and analysis at different erosion depths.

The primary ion pulse duration required for a high mass resolution is only a few nanoseconds for a typical drift path 1 of approximately 2 m. The primary ion pulses are generated from the static beam of an ion source by a suitable beam pulsing method. The number of primary ions per pulse results from the static current of the ion source Ip and the pulse duration t:

N = 1 • t / e (e: elementary charge) (3)

_ »_> _»

From this it can be seen that when the primary pulse duration is reduced, the number of primary ions per pulse decreases.

The result of this is that more primary ion pulses are then required in order to achieve the same number of secondary ions. testify and prove. This means that the measuring time increases. This increase in the measurement time becomes a problem especially for finely focused ion sources for micro-range analysis, since only very small ion currents I are available here. The recording of spectra with high dynamics, the recording of lateral distribution images and the depth profiling with high dynamics then often lead to measurement times of greater than one hour to several hours.

A shortening of the measuring time is only possible in the prior art by extending the primary pulse duration t with a corresponding loss of mast resolution, or by increasing the repetition rate with a corresponding restriction of the mass range detected (see Eq. 2).

It is the task of specifying a method for operating a secondary ion time-of-flight mass spectrometer by means of which the measuring times can be shortened without loss of mass resolution and without reducing the mass range.

This object is achieved in a method of the type mentioned at the outset, which is characterized in that d) each primary ion pulse consists of a plurality of partial pulses, e) each partial pulse is so narrow that it allows the resolution of the finely structured mass areas, f) the distance t B of the partial pulses is greater than the width of the finely structured mass areas, g) the number n of partial pulses is selected such that is smaller than the distances between the finely structured mass ranges, ti) the n spectra belonging to the partial pulses of each finely structured mass range are added. In other words: in the time interval tz (see Eq. 2) the surface is not with a single short primary ion pulse, but within the cycle time t Z with a

Bombarded with a succession of several, essentially identical, partial pulses at short intervals. The distance between two partial pulses is greater than the time-of-flight difference of element and molecular ions of an integer nominal mass; in addition, the distance from the first to the last primary ion partial pulse is smaller than the time of flight difference between the nominal masses in the detected mass range. By adding the n resulting fine maxima, the signal-to-noise ratio of the measurement can be significantly improved without increasing the measurement time. The sequence preferably consists of n = 3 to n = 20 partial pulses.

A device for carrying out the method is accordingly a time-of-flight secondary ion mass spectrometer, in which a material sample surface is bombarded with primary ions in a pulsed manner (primary ion pulses) and the primary ions detach secondary ions of different masses m from the material sample surface. The same energy E is applied to them after they have been extracted. The mass-dependent flight time t is measured via a path 1, the flight time t being proportional to the root of the mass and the number of secondary ions which correspond to a specific mass m within a defined cycle time t Z in specific

Time intervals t result in fine structure maxima, the latter each corresponding approximately to an integer atomic or molecular weight of element or molecular ions.

The amplitudes of the fine structure maxima allow a qualitative and quantitative analysis of the composition of the material sample surface. The device is characterized by a pulsed primary ion source, with which the material sample surface can be bombarded in the time interval t with a sequence of n essentially identical primary ions at a short time interval t B, the time interval t B of two primary ion pulses being greater than the time-of-flight difference between element and molecular ions one Nominal mass range and also the distance

t = n • t

from the first to the last primary ion pulse is shorter than the time-of-flight difference between the nominal masses in the detected mass range, the n fine structure maxima belonging to the same secondary ion species being addable.

An embodiment of the method and the device is explained in the drawing. The figures in the drawing show:

1 shows a schematic arrangement of a time-of-flight mass spectrometer

2 shows a mass spectrum that was obtained according to the state of the art. A) Overview spectrum in the mass range 1-50 u. B) Detail in the mass range 26.5-28.5 u

3 shows a mass spectrum which was obtained in the operating method according to the invention. A) Overview spectrum in the mass range 1-50 u. B) Detail in the mass range 26.5-28.5 u

Fig. 4 secondary ion distribution images obtained according to the prior art

5 secondary ion distribution images, which in the operational processes according to the invention were obtained.

Fig. 1 shows the measuring principle of the TOF-SIMS method. A continuously operating ion source IQ is pulsed with the aid of a suitable beam pulser PS, which results in the primary ion pulses already mentioned.

The pulsed primary ion beam, which is filtered with a mass filter MF, is then focused and positioned on the sample P (target) using a focusing unit FK and a raster unit RS. All simply charged secondary ions generated by the primary ion beam are accelerated to the same energy E with the aid of a suction voltage U. Their runtime is then measured in a time-of-flight analyzer FZA with spatial and temporal focusing properties. A suitable time-resolving ion detector ID is used for the detection. The output pulses of the ion detector ID are processed by the registration electronics, consisting of a discriminator DS and a time-to-digital converter TDC in connection with a fast memory.

A typical measurement result, which was recorded by the method according to the prior art, is shown in FIGS. 2a and 2b. Here, a single primary ion pulse was used per cycle time t Z with a pulse duration of 1.3 ns.

The triggered secondary ions are registered over the cycle time of 100 μs and all events are added over a total of 1695 · 10 ⁷ cycles. The measuring time here is 1695 s = 28 min. In the present case, the sample is a silicon wafer with an aluminum test structure. 2a shows the overview spectrum in the mass range (nominal masses) from 1 to 50 u.

Fig. 2b shows in detail the fine structure of the maxima Mass range from 26.5 to 28.5 u of the spectrum according to Figure 2a. Here the separation of various atomic and molecular ions becomes clear with the help of the high mass resolution. With the nominal mass 27 there is a separation into AI "^" and C 2H3 ^"*" , with the nominal mass 28 one

Separation into Si ^* , A1H ^* and C ZH4 ^"*" . As a result of the exact

Masses of the elements of the periodic table can be no further maxima between the maxima of the nominal masses 27 and 28, eg between CH ^"*" and Si ^"4" ).

In order to significantly reduce the measuring time or to increase the number of registered secondary ions and thus the dynamics in the case of a predetermined measuring time, an operating mode is selected which corresponds to the invention and which is expressed in FIGS. 3a and 3b. Instead of a single primary ion pulse, 12 partial pulses of the same pulse duration with a time interval of 25 ns were used with an unchanged one

Cycle time t of 100 μs.

3a shows the overview spectrum in the mass range 1 to 50 u. The separation of the different nominal masses is clearly recognizable. The flight time difference between m = 49 and = 50 is therefore greater than 12. 25 ns = 300 ns.

3b shows the fine structure of the same mass spectrum in the mass range from 26.5 to 28.5 u. The 12-fold superimposition of the peak structure from FIG. 2b can be clearly seen by using the sequence of 12

Partial pulses with a defined distance. By choosing the distance of 25 ns, a superposition of the maxima belonging to different primary ion pulses is avoided, so that an assignment of the peak series to a specific connection is possible. As in FIG. 2b, the maxima for AI ^"*" and C 2H3 are detected for the nominal mass 27 and the maxima for the nominal mass 28 for Si ^"1" , A1H ^* and C 2 H 4 ^"*" .

All events were in turn added up over a total of 1695 "10 ^{" 7} cycles. The measurement time here, as in FIGS. 2a and 2b, is 1695 s = 28 min. The example shows that with the same measurement time the 12-fold secondary ion intensity can be registered without loss of mass resolution Accordingly, by adding the intensities of the respective secondary ion species, the same information as in Fig. 2a and 2b can be obtained in 1/12 of the measuring time, which means, for example, a shortening of the measuring time from 28 min to 2.3 min .

The mode of operation corresponding to the invention also shortens the recording time for secondary ion images. Here, an analysis as in FIG. 2 or FIG. 3 is carried out for each pixel (pixel) and then the distribution images of the different secondary ion species are constructed.

4 shows distribution images with an operating mode according to the prior art. A single primary ion pulse was used per cycle. The events for each pixel were added and evaluated over 200 cycles. The total measuring time for 256 x 256 pixels is 1310 s = 22 min.

5 shows distribution images of the same sample with an operating mode according to the invention. A pulse sequence of 12 partial pulses per cycle at intervals of 25 ns was used. The events over 200 cycles were added and evaluated. As in FIG. 5, the total measuring time is 1310 s = 22 min. By operating according to the invention, the intensity and the dynamics in the secondary ion distribution images are significantly greater with the same measuring time and the same information content. So in Fig. 4 only 47 secondary ions AI ^{"*" are registered} in the brightest pixel, while in FIG. 5 a total of 411 secondary ions are contained in the brightest pixel. The other distributions of CH ^"*" , Si ^"1" and A1H ^"*" also show similar improvements in image quality with the same recording time. For the same image quality, there is a corresponding reduction in the image acquisition time by a factor of 12.

The procedural use of a time pulse sequence instead of a single pulse can also be applied to analog methods, in particular to gas phase analysis using time-of-flight mass spectrometry. The ion generation takes place here with an electron pulse; the gas ions generated are accelerated and their masses determined by means of a time-of-flight measurement. If a sequence of electron pulses is used instead of a single electron pulse, a shortening of the measurement time in high-resolution time-of-flight mass spectrometry can be achieved, as with the TOT-SIMS - utatis mutandis.

Claims

Expectations :

Method for operating a time-of-flight secondary ion mass spectrometer for the analysis of mass spectra, in which several finely structured mass areas occur isolated at larger distances, with the following steps: a) a material sample surface is tz at regular intervals (Cycle time) successive primary ion pulses b) the secondary ions of different masses m released from the primary ions from the surface of the material sample are accelerated to the same energy, c) the mass-dependent flight time t is measured via a path 1 and the mass is determined therefrom, characterized in that d) each primary ion pulse consists of several partial pulses, e) each partial pulse is so narrow that it allows the resolution of the finely structured mass areas, f) the distance t B of the partial pulses is greater than the width of the finely structured mass areas, g) the Number n of partial pulses is chosen so that is smaller than the distances between the finely structured mass ranges, h) the n spectra belonging to the partial pulses of each finely structured mass range are added.

2. The method according to claim 1, characterized in that the sequence consists of n = 3 to n = 20 primary ion pulses.

3. Time-of-flight secondary ion mass spectrometer for performing the method according to claim 1 or 2, in which ⁵ a) a material sample surface with in regular

• intervals tz (cycle time) consecutive

Primary ion pulses are bombarded, b) the secondary ions released from the primary ions from the surface of the material sample

^ Masses m are accelerated to the same energy, c) the mass-dependent flight time t is measured via a path 1 and the mass is determined therefrom, characterized by a pulsed primary ion source, with which the - * 5 material sample surface with a time interval tz

Sequence of n essentially identical primary ions can be bombarded at a short time interval t B, the time interval t B of two primary ion pulses being greater than the time-of-flight difference between element and molecular ions 20 of a nominal mass range and also the distance

from the first to the last primary ion pulse, the flight time difference between the nominal masses in the detected mass range is shorter than 25, the n fine structure maxima belonging to the same secondary ion species being addable.

30

35