WO1987006056A1

WO1987006056A1 - Device for sample introduction to a mass spectrometer

Info

Publication number: WO1987006056A1
Application number: PCT/SE1986/000135
Authority: WO
Inventors: Bengt Kasemo; Karl-Erik Keck
Original assignee: Bengt Kasemo; Keck Karl Erik
Priority date: 1986-03-26
Filing date: 1986-03-26
Publication date: 1987-10-08
Also published as: SE8404840D0; SE445403B; SE8404840L

Abstract

This device includes a flow channel (10) for the sample which has a constriction (13) at the inlet end, connectable to the sample source, and which in the other end is connected to the mass spectrometric analysis chamber. The inlet constriction is primarily designed for a sample flow which is 10-3 mbar . l/s.

Description

DEV ICE FOR SAMPLE INTRODUCT ION TO A MASS SPECTROMETER .

The invention consists of a device for sample introduction to a mass spectrometer.

The device according to the invention is primarily developed for analysis of relatively light gases with mass numbers below 1000 and especially below 300. There is no principal reason, however, to use it even with certain liquids (without intermediate vapourizing). The term "sample" is here used to include gas as well as liquid.

Durirfg mass spectrometric gas analysis the gas pressure must usually be reduced considerably, since the mass spectrometer can only

-4 operate at pressures below about 10 torr, and since the gas to be analyzed usually exists at considerably higher pressure. The pressure reduction is obtained with a gas inlet system, usually constructed in one of the following ways:

A. A leak valve with suitable, and variable, leak rate is con¬ nected between the gas to be analyzed and the mass spectromet system.

B. The gas is pumped through a narrow tubing or capillary, to obtain a certain pressure reduction, e.g. from 1 bar to 1 mba and the gas is then leaked through a second leak of constant leak rate, e.g. a porous disc.

C. Between the gas to be analyzed, in the following called the sample gas, and the mass spectrometer a number of (2-3) differentially pumped vacuum chambers are connected in series via small apertures positioned on a common axis.

In case A and B the mass spectrometer measures the isotropically established partial gas pressures in the analysis chamber, where the mass spectrometer is positioned. In case C the local partial pressures are measured in the molecular beam formed by the coaxially positioned apertures and the differential pumping. Alternative C produces by far the most sensitive measurement, but is more costly because of the cost of the differentially pumped vacuum systems. Alternative C is superior because the local partial pressures in the sample gas is measured befo the gas molecules have interacted with the walls of the analysis chamb and because the background pressures can be measured and subtracted by use of a moveable shutter between the differentially pumped stages and the analyzer. Because of the much higher cost of alternative C in comparison with A and B, it is usually only used in situations wher unstable molecules or radicals are to be detected. For simpler analyti mass spectroffletry either of alternatives A and B are used.

The purpose with the invention is to obtain a device for sample introduction to a mass spectrometer, which offers most of the advantag of alternative C above, but at a cost that does not need to exceed tha of A or B.

The mentioned purpose is arrived at by giving the device the characteristics of the patent claim 1.

The following advantages are obtained.

(i) Most of the advantages of a molecular beam mass spectro¬ meter (except radical detection) is obtained, without differential pumping, (ii) Negligable sample consumption (< 10~ bar • 1/s) (iii) Small time constant for response to changes in gas composi tion e.g. during continuous analysis of gaseous environmen (iv) Increased sensitivity at high mass numbers, in comparison with conventional gas inlet systems, (v) Simple mounting on most available mass spectrometer system (vi) The magnitude of the measured signal is independent at the pumping speed of the analysis chamber, but the background signal decreases in proportion to the pumping speed.

For a more detailed explanation of the invention, various version of it will be described in the following, with reference to enclosed drawings of which

FIG. 1 is a schematic view of the device, according to the invention, in a more advanced version,

FIG: 2 shows a connection for the high pressure end with a particle filter,

FIG. 3 shows a modified version of the device of Fig. 2,

FIG: 4 shows a version of the high pressure end for analysis of samples at different positions, FIG. 5 is a diagram showing ΔP? /P¹? as a function of mass number for different gases,

FIG. 6 is a diagram showing the relation between A _T^ /P^{1 •} Sⁿ and the square root of the mass number,

FIG. 7 is a schematic drawing of one additional version of the device according to the invention, and

FIG. 8- is a schematic drawing of one additional version of the device according to the invention.

With reference to Fig. 1 this version of the device according to the invention includes a flow channel, constituted by a tubing 10. In the left end, which is the high pressure end, i.e. the end which is positioned at or connected to the sample source, a smaller tubing 12 is connected to 10 by a vacuum tight seal 11. The smaller tubing 12 has a very small hole or channel 13 at its left end. The other, right end of the tubing 10 is terminated by a tubing 14, and with a welding 15 the tubing is connected to a vacuum flange 16, for mounting on the analysis chamber 17 of a mass spectrometer system, so that the tubing 14 is directed into the analysis chamber through the flange opening 18. This end of the tubing 10 is the low pressure end and the terminating tubing 14 of the low pressure end is directed into the center of the sensitive volume of the ion source 19 of the mass spectrometer. A moveable shutter or rotatable chopper 20 is positioned between the exit of the tubing 14 and the ion source.

A closing valve 21 is positioned on the tubing 10 between the high and low pressure ends. A side tubing 22 is attached to the tubing 10, which via a closing valve 23 is connected to a not shown vacuum pump. This evacuation line is not necessary but convenient for evacuation of the volume between the high pressure end and the valve 21, when the latter, has been closed for some time. In such cases the volume is evacuated via the side tubing 22, after which the valve 23 is closed and the valve 21 is opened.

The dimensions of the hole/channel 13 should be chosen so that a suitable sample amount is flowing through the hole or channel, respectively. Out of several tried versions, a quartz tubing of inner diameter 1-2 mm, whose tip is melted into a narrow channel 13 was foun most suitable for the device according to the invention. Channel dimensions of length 0.1 mm and diameter 7 μm produces a sample gas flow of about 10 mbar • 1/s at 1 atm. inlet pressure.

An advantage with a quartz tubing over other glasses is that it stands high temperature ^'(up to 1500 °C has been used intermittently) and is chemically stable in most environments. Alternative solutions with metal tubings are harder to manufacture and usually less stable thermally and chemically. The solution with the quartz tubing 12 is in the following named the quartz leak. The quartz leak is connected to the tubing 10 via a vacuum tight seal 11 e.g. an Viton 0-ring, a metal ring gasket, a high-vacuum adhesive or a glass metal seal.

Any hole or channel 13 producing a sufficiently small gas flow at the low pressure end (i.e. into the analysis chamber) is in princip appropriate for the device according to the invention. The gas flow

_3 should typically be less than or around 10 mbar 1/s, which at a pumping speed of 100 1/s in the analysis chamber gives a base pressure

-5 during operation of < 10 torr, i.e. a suitable operating pressure for the mass spectrometer. The dimensions of the hole/channel 13 depen on the ratio between the diameter and axial extension of the aperture.

A channel which is too long should be avoided since it will increase the system time constant.

The terminating end 14 in the low pressure end of the arrangement can suitably have an inner diameter around 0.5-3 mm. The optimum value depends on the properties of the ion source of the mass spectrometer. The terminating end should terminate as close to the outer edge of the ion source 19 as possible, but sufficient space, about 1-3 mm, must be left for the shutter or chopper 20.

When this device according to the invention is in function, the pressure at the low pressure end is usually much larger than at the low pressure end (the latter is ~ 10^" - 10^" mbar). The gas flow at the low pressure end is then molecular while at the small hole/channel at the high pressure end the flow is viscous at inlet pressures > 100 torr. The device is completely gas tight so that all gas entering through 13 at the high pressure end is passing through 14 at the low pressure end into the analysis chamber. Since the flow through 13 is viscous, mass separation is avoided. The described device is functioning as follows: During gas analysis the high pressure end of the tubing 10 i.e. the quartz leak 12, is positioned in the sample. A gas flow tubing system may be installed so that the sample gas is flowing by the high pressure end. Due to the strong pressure gradient that is established over the narrow hole or channel, due to the pumping

-3 -5 in the analysis chamber, a small fraction (typically 10 -10 mbar • of the sample is sucked into the tubing and is transported (without mass loss) through it to its low pressure end where the gas under molecular flow conditions is "sprayed" into the analysis chamber. By choosing a small ratio between the length and the inner diameter of the low pressure end tubing 14 (typically < 1:5), the divergence

(relative the tube axes) of the emitted gas becomes relatively small.

By choosing the diameter of the tubing at the low pressure end smaller than the linear dimension of the effective volume of the ion source, and by positioning the low pressure end as close to the ion source as possible, a local pressure is established (local density of molecules) in the ion source, which is considerably larger than the isotropic pressure in the analysis chamber caused by the introduced gas.

The partial pressure y-γ of a given gas, n, in the ion source will therefore have two contributions. The first one is the isotropic partial pressure, , which is determined by the sampled gas flow,

Qⁿ, divided by the pumping speed, S^π, of the gas n in the analysis second contribution is the above mentioned local

pressure increase Δ -_joc, due to the directed gas flow. The total par¬ tial pressure is then

Pⁿ - Pⁿ + ΛP^Π ^PT0T ^" Ms ^{+ ΔF}loC

By equipping the gas inlet system with a moveable shutter 20, that can be moved in and out of the space between the low pressure end 14 and the ion source, two measurements can be performed:

When the shutter is positioned between the low pressure end and the ion source, the mass spectrometer signal I¹ for gas n is proportional to p|? , i.e. s ^" is Ms where c" is a sensitivity constant for gas n, that can easily be determined.

When the shutter is moved away a signal, I-._Q-., is measured, which is proportional to the sum of the two above mentioned contributions, i.e. ,

T-Π _ r^π Dⁿ _. rⁿ ΛD^Π T0T ^{= C}is ^{' P}is ^{+ C}loc ^{' ΔP}loc^> where C?lo,,c., is also a constant whose value can be determined,

By taking the difference between the two measurements a signal

ΔI?lo-c,. is obtained;

A T ¹"¹ Pⁿ A D⁰

I OC I OC 1 OC

which is representative for the sample gas above.

With the described device one can thus measure the local con¬ centration in the ion source of the incoming sample gas. This has several important valuable consequences for the sensitivity etc of the measurement. a. The same advantage as in a molecular beam mass spectrometer is obtained with respect to measurement of the sample gas content before it has interacted with the walls of the ana¬ lysis chamber. This can strongly reduce errors due to back¬ ground effects particularly at low concentrations. b. The background effects due to residual gases in the analysis chamber or due to wall interactions decrease proportionally to the pumping speed, which is often not the case when the gas is non-directed, but only the isotropic pressure is measured. c. The mass dependence of the measured signal is such that the sensitivity increases by the square root of the mass, in comparison with measurements of isotropic partial pressures. This is so because the gas density for a given gas in the ion source is (due to the directed beam) propor¬ tional to the amount of gas introduced per unit time, di¬ vided by the mean velocity of the gas molecules along the- beam axis. Since all molecules have a thermal velocity distribution, this mean velocity becomes proportional to the square root of the mass. This relation has been tested and verified for the present gas inlet system. d. By making the shutter a chopper that periodically chops the divergent sample gas beam, one can obtain ΔP? electroni¬ cally by phase sensitive, lock-in, detection techniques, or by synchronized pulse counting.

In a practical measurement situation it is of course necessary that the local pressure increase, ΔP? , is not too small in compariso with p . Fig. 5 shows ΔP? P^ as a function of mass number for dif¬ ferent gases. The measurements were performed in a turbo pumped system (pumping speed for air w 200 1/s) (base pressure < 10^" mbar), with a Balzers QMA 311 mass spectrometer system. Similar results have been obtained in a separate system with a Balzers QMA 112, with a different type of ion source.

One can see in FΣ§. 5 that ΔP? for all gases are larger than P¹ and that the difference increases with increasing mass number. A simple analysis (see above) shows that ΔP? /P|? • Sⁿ plotted vs the square root of the mass should give a straight line. Sⁿ is the pumping speed for gas n. FIG. 6 shows that this relation is obeyed very well. In the prototype system used so far several parameters have not been optimized, e.g. the diameter of the low pressure end terminating tubin 14, and its distance to the ion source 19. The ion source used is a standard cross-beam ion source. A simple analysis shows that the elec¬ tron beam in the ion source could be optimized to yield better sensiti with the present gas inlet system. Summarizing, these optimizations should allow a ratio of ΔP? /P¹? « 5 - 10 for hydrogen (n=2) and of 35 - 70 for mass number n = 100, at « 200 1/s pumping speed. The ratio will increase proportionally to an increase in pumping speed.

Required calibrations are in principle performed as those per¬ formed with any mass spectrometer system. Relative sensitivity factor are measured with calibration gases, with the gas directed and non- directed into the ion source, respectively. From ΔP? and P¹ the relative sensitivity factors for the measurements can be obtained. In practice usually only the sensitivity factor for ΔP? is of interest. If absolute sensitivity factors for both signals are desired, also the pumping speeds for different gases must be determined. It is usually practical to include such factors in the calibration constants, however, since it is ultimately only the relative concentrations of various gases in the sample gas that are of interest. When chopping and phase sensitive detection is used, the calibration is done directly with the measured signals utilizing known gas mixtures.

When there is a risk for plugging of the quartz leak 12 by particles a particle filter 25 made from e.g. porous, sintered metal or glass, is connected in a tubing 26 A or 26 B in front of the quartz leak. If a short response time is important the sample flow through the filter can be accelerated by a simple pump connected at one end of the tubing 26 A or at a side tubing of 26 B, respectively.

The very minute leak rate and the fact that no differential pumping is used, make the amount of sample required for analysis extremely small (e.g. much smaller than in the solutions B and C described earlier). This is a considerable advantage when the amount of available sample for analysis is limited, or when the sample con¬ sumption of the sampling site should be minimized .

In some cases gas analysis at different positions are desired e.g. in a room, in a gas container or in a reactor. An easy way of obtaining this with the described device is to connect the high pressur end 10 to a flexible vacuum tight bellows 28, which is in turn - connected to the closing valve according to FIG. 4. In this way the gas sampling position can be moved around manually or by a simple robot. The quartz leak T2 is attached at the sampling left end of the bellows tubing 28.

That part of the invention, which transports gas from the high pressure side to the low pressure side can for simpler gas analysis, where optimal sensitivity^"is not required, be connected to the analysis chamber, without directing the low pressure end terminating tubing 14 into the ion source. Then only the partial pressures P!|_s can be measured. Also for this type of less sensitive measurements the gas inlet system offers advantages over other existing constructions:

(i) The whole device can be mounted on a single flange and is very neat (superior inthis respect to B and C).

(ii) The device can be mounted on essentially any existing mass spectrometer equipment with a minimum of modifica¬ tions and requires no differential pumping. (iii) The response time for changes in gas composition can be very low (~ 0.05 s has been calculated and measured).

By minimizing the volume between the high pressure end tubing 10 with the quartz leak 12 and the closing value 21 (FIG. 1), the gas amount that must be pumped away, after the closing value 21 has been closed for a larger time, becomes so small that it can be leaked into the analysis chamber and pumped away by the pump of the analysis chamber. This version has been used in a prototype and the pressure rise in the main chamber can be kept less than 10 mbar, and the evaluation can be completed in less than 30 s. The advantage with this construction is that no pumping value 23 and no evacuation pump for this purpose is needed.

When a quartz leak 12 is used a very simple design of the device can be obtained, as shown in Fig. 7. Here the flow channel for the sample is not the tubing 10 but the quartz leak 12 whose two ends are the high and low pressure ends, respectively. A quiding support is mounted at 29. Closing of the leak is obtained by pushing a membrane 30 of e.g. Viton or Teflon agianst the quartz leak 12 at its high pressure end as shown in the figure. The membrane is attached to a shaft 31 quided by the tubing 32 on tubing 26A, with sealing establish by 0-rings 33. This construciton is convenient because it minimizes th wall area exposed to gas. It is also convenient when heating of the whole inlet system is desired (due to gas adsorption and exchange on the walls of the gas flow system). Heating is achieved by a heating spiral 34 on the outside of the tubing 10, covered by a shield 35, which is connected to 10 at the low pressure end at 36, and is further connected to the vacuum flange 16 by a welded seal 37. There is one drawback with this solution, however, in that the vacuum of the maan system must be broken for replacement of the quartz leak. This is not necessary with a closing valve mounted.

In some cases, e.g. when the gas pressure at the high pressure end (the sampling site) is varying over a large range, a constant leak rate in the form of a hole or channel 13 is impractical since the amount of sampled gas becomes -too small at the lower pressures. In such cases the fixed leak must be replaced by a variable leak e.g. a leak valve or a variable aperture. Even in this case the advantages of the construction at the low pressure end with a directed beam and shutter function as described above are preserved.

The shutter 20 can be replaced by a moveable bellow in analogy with the solution at the high pressure end (FIG. 4). The direction of the sample gas beam energing from the low pressure end can then be changed by a simple vacuum tight manipulator. This eliminates the need for a moveable shutter or chopper but can produce the same function. The shutter 20 can also be replaced with a piezo¬ electric shutter.

A cheap and elegant solution of the problem to measure the mass spectrometer signal both with the beam directed into the ion source and not directed into it, respectively is obtained with a magnetic ball. The principle is shown in FIG. 8. Chopping is achieved by electromagnetically coupled motion of the ball. In one position of the ball 39 the gas flow is directed into the ion source via the channel 38A and PI -. is measured. In the other position the gas is introduced via the channel 38B but not into the ion source. P¹ i?s is then measured since the bal l i s bl ocking the di rect beam path but al lows another non-di rected gas fl ow path into the analysis chamber.

Claims

π

CLAIMS 1. Device for sample introduction to a mass spectrometer, c h a r a c t e i z e d in that the device comprises a flow channel (10; 12) for the sample, at one end thereof connected to the analysis chamber of the mass spectrometer and having a constricted inlet opening (13) at the other end to be connected to the sample source preferably dimensioned for a sample flow which is - 10^" mbar ^• 1/s.

2. Device as in claim 1, c h a r a c t e r ¬ i z e d in that the constricted opening (13) is arranged in an inlet tubing (12), preferably a quartz tubing, which is connected by a vacuum tight seal to said other end of the flow channel (10).

3. Device as in claim 1 or 2, c h a r a c t e r ¬ i z e d in that the end of the flow channel (10) which is connected to the analysis chamber (17) is provided with a nozzle tubing (14) directed towards the ion source (19) of the mass spectrometer.

4. Device as in claim 2 or 3 wherein the inlet tubing (12) comprises a quartz tubing, c h a a c ¬ t e r i z e d in that the quartz tubing forms the flow channel and the nozzle tubing.

5. Device as in claim 4, c h a r a c t e r - i z e d in that the quartz tubing (12) is encompassed by heating means (34) .

6. Device as in claim 1 , c h a r a c t e r - i z e d in that said one end of the flow channel (10) is connected to a conduit (26A; 26B) for positive flow of the sample, said conduit being provided with a filter (25) for incoming sample.

7. Device as in claim 1, c h a r a c t e r - i z e d in that the flow channel (10) is provided with shut-off means (21) for the sample flowing therethrough.

8. Device as in claim 7, c h a r a c t e r - i z e d in that the flow passage (10) via shut-off means (23) is connected to a vacuum pump upstream of the shut-off means (21) in the flow channel .

9. Device as in claim 3, c h a r a c t e r ¬ i z e d in that means (20; 39) are provided to cut off al ternatingly the flow of sample' to the ion source (19) .

10. Device as in claim 9, c h a r a c t e r ¬ i z e d in that said means comprises a chopper (20) for cutting-off the flow at a predetermined frequency

11. Device as in claim 1, c h a r a c t e r ¬ i z e d in that said other end of the flow channel (10) can be adjusted to different positions.