US2768301A

US2768301A - Method of mass spectral analysis with negative ions

Info

Publication number: US2768301A
Application number: US240966A
Authority: US
Inventors: Willard H Bennett
Original assignee: Individual
Current assignee: Individual
Priority date: 1951-08-08
Filing date: 1951-08-08
Publication date: 1956-10-23
Anticipated expiration: 1973-10-23

Description

W. H. BENNETT Oct;

' METHOD OF MASS SPECTRAL. ANALYSI$ WITH NEGATIVE IONS- 3 Sheets-Sheet 1 Filed Aug. 8. 1951 INVENTOR ATTORNEYS INVENT OR 3 Sheets-Sheet 2 W. H. BENNETT METHOD OF- MASS SPECTRAL ANALYSIS WITH :NEGATIVE IONS Oct. 23,1956

F i led Aug. 8, 1951.

BY 77m ATTORNEYS m m 4 v H Gnu A .1I w W Qmm w v u. nun I. Eef A g M .5 55 y H 9o 4 \U uSQUk 1 Ii b 3) 1 hwbvektx g v V r V .-\Q AN? QWW\ 3.. Mk u wt 7 m 0W n n u NW tk Q N tu\kw w \b hQ 9 \Qvfig wW LL U WW I m 1 un\ 4 an H R P d L H A A 4 gm map of Pas/71w: 10m:

APAcn/m/wz Oct. 23, 1956 w. H. BENNETT- 3 5 METHOD MASS SPECTRAL ANALYSIS WITH NEGATIVE IONS Filed Aug. 8, 195; JSheets-Sheel YIELD 07 1 05074: lows FQR D/Ff'E/Pt/V? mlw z/zva POTENTIAL! f/aj ION 37:41:: flea/v cnnaazvMa/vax/pa F01? OIFFtPf/VT lO/V/Zl/VG' P07647741) NEGA77V Aral: OXYGEN lO/Y 7 40 ,a 2'0 3'4 4'0 $0 6'0 POTENTIAL (p) //v v04 7:

E E 101v YUELD from hoLCcuLA/P oxraz/v 6 F011 D/F/Efil'A/T m/wzm a FOTE/YT/AAJ W i. I INVENTOR E z lV/LL/I/PD HfiE/V/VE-TT L o 8: g 10 zo a0 an BY O M POTENTIAL (P) //Y VOLT-5 1 ATTORNEYS METHOD OF MASS SPECTRAL ANALYSIS WITH NEGATIVE IONS Willard H. Bennett, Fayetteville, Ark.

Application August 8, 1951, Serial No. 240,966

2 Claims. (Cl. 25.0-41.9)

This invention relates to methods of mass spectral analysis with negative ions.

States Patent An object of the invention is to provide a method of analysis whereby it is possible to distinguish between gases of the same molecular weight.

Other objects and further advantages will appear as this description proceeds.

Practically all, if not all, prior art successful mass spectral analysis has been with positive ions but I have discovered ways of rendering analysis with negative ions very useful in certain circumstances.

The invention is especially useful in cases where it is desired to analyze a mixture having two or more gases of the same molecular weight. Very often in such cases there is a critical bombardment velocity at which one of the two gases will produce negative ions and the others will not. critical velocity and determining the number of negative ions produced, it is possible to conclude that a particular gas is present in the mixture in a particular amount. For example, to take a very simple case, it is impossible with prior art mass spectrometers to distinguish between carbon monoxide and molecular nitrogen because both have molecular weights of 28. Hence, it has been impossible to detect small percentages of carbon monoxide in the presence of air. However, carbon monoxide, if bombarded with 11 volt electrons, will yield negative oxygen ions whereas nitrogen will not yield negative ions if bombarded with electrons of any bombarding energy. Hence, the presence of small percentages of carbon monoxide may be determined by ap- Hence, by bombarding the mixture at this plying 11 volts accelerating potential to the electrons used in the ion source and then determining the masses of .the negative ions produced. In that particular case, a complicating factor is that molecular oxygen also produces negative atomic ions, but such ions are produced .only with electron velocities in the vicinity of 7 volts and none are produced with 11 volt electrons. Hence, by maintaining the potential of the ion source at the critical potential of 11 volts the desired determination may be made.

As another example of an application of my invention, assume that there is amixture of three gases namely Cal-I4, CO and N2, all of mass 28. By bombarding with 11 volt electrons producing negative oxygen ions it is then possible to determine the quantity of CO in the mix ture by measuring the number of negative oxygen ions thereby produced. If the same analysis were attempted by prior art methods one would secure mass spectral lines for masses 1, 12, 13, 14, 15, 16, 24, 25, 26, 27 and 28 resulting in confusion and error.

While I am not limited to any particular apparatus so Fatented Oct. 23, 1956 an improvement on that disclosed in my published article entitled Radio Frequency Mass Spectrometer, Journal of Applied Physics, vol. 21, No. 2, pages 143 to 149, February 1950 and is also an improvement on that disclosed in my prior copending application S. N. 196,024 filed November 16, 1950, entitled Radio Frequency Mass Spectrometer.

In the drawings:

Figure 1 is a part cross-sectional and part schematic diagram of one form of tube that may be employed.

Figure 2 is a left hand end View of Figure 1.

Figure 3 is a right hand end view of the tube of Figure 1.

Figure 4 is a schematic diagram of a preferred hookup of the invention.

Figure 5 is a curve illustrating the yield of positive ions of any typical gas for different ionizing potentials.

Figure 6 illustrates the negative ion yields of carbon monoxide for difierent ionizing potentials.

Figure 7 illustrates the negative ion yield of oxygen for different ionizing potentials.

Referring to Figures 1 and 2, the tube employs a cathode 50 which has leads 50a. A metallic cylindrical shield 51 has lead 51a for connecting the shield 51 to a variable potential approximately the same as that of the cathode. Shield 51 has an aperture 51b which is about four millimeters in diameter. Cylindrical metal electrode 52 has an aperture at its lower end of about two millimeters in diameter and coaxial with aperture 5112. These parts are all mounted in an envelope 53 which has an inlet opening 54 through which gases to be analyzed are admitted and it also has exhaust outlets 55 and 56 adapted to be connected to exhaust pumps for evacuating the tube. As is apparent, the exhaust outlet 55 primarily exhausts the interior of cylindrical cup electrodes 51 and 52 for the following purpose. The gas to be analyzed may have its chemical composition somewhat changed if it strikes the very hot cathode. Therefore, in order to make sure that none of such gas with changed composition passes into the main analyzing chamber, the tube is continuously exhausted at outlet 55 whereby any gases of changed composition are promptly removed from the envelope 53. The gases to be analyzed are continuously fed into inlet 54 at a slow constant rate and are continuously exhausted by exhaust pumps connected to outlets 55 and 56.

The potential on accelerating electrode 52 is highly positive relative to the cathode, say by thirtyvolts. As the electrons emerge through aperture 52b they pass between the two parallel grids 61 and 63. The average potential of these grids relative to the cathode will be termed P, for simplicity. Assume that it is desired to effect a value of P of say eleven volts, which as has been said is the desirable potential used in detecting carbon monoxide. To carry out such an assumption the grid 61 would be made ten volts positive relative to cathode 50 and the grid 63 would be made twelve volts positive relative to the cathode 50. As a consequence the electrons from cathods 50 form a narrow beam, known as a pencil of electrons, as they emerge from aperture'52b at a potential of thirty volts. They slow down to a speed of P volts as they enter the rigion between grids 61 and 63. A magnet, or its equivalent in the form of suitable coils C, may be provided to produce a weak magnetic field which passes parallel to and in between the screens 61 and 63 and perpendicular to the electron stream emitted by the electron gun 50 52. In this case the magnetic field may have approximately thirty gauss. The magnetic field could equally well be applied coaxial with the electron stream. In the latter case it may have a field intensity of approximately gauss or more. When the field is perpendicular to the electron stream, as above stated, the potential diflerence between grids 61 and 63 must be set at just a sufficient amount to give an electrostatic field between grids 61 and 63 whose force on electrons traversing this space just equals the force in the opposite direction on them due to the magnetic field. If P is the potential difference in volts between the cathode 50 and the midpotential between grids 61 and 63, then the potential difference (in volts) between grids '61 and 63 should be equal to 0.0198 H.d. /P, where H equals the field gauss, and d equals the distance between grids 61 and 63.

in order that it can be determined how many electrons are being produced, an item which it is often desirable to know in calibrating and using the tube, a small metallic electrode 62 is connected in series with a meter 1, a battcry B, and the cathode 50.

In Figure 1 and 2 coils M are shown, as they may be conveniently used in lieu of the magnet above referred to.

In adjusting the apparatus for use, the magnet, or coils C as the case may be, is temporarily removed, the voltage on electrode 52 is adjusted to give maximum electron current, while the voltages on grids 61 and 63 are both set at the potential P (say eleven volts) initially in order to focus the beam. When the beam is in proper focus the magnet M is replaced and the potentials on grids 61 and 63 are rendered slightly different, according to the foregoing formula.

In view of the detailed disclosure in my published article, supra, and in my said prior copending application the theory of operation of the three stage mass spectrometer herein referred to need not be stated in detail. Adjacent grid 63, there are grids 64 and 65 respectively connected to

leads

64a and 65a, and these grids are charged positively at potentials intermediate those of grids 63 and 66. The grid 64 tends to draw the negative ions through grid 63 and grid 65 tends to further accelerate the negative ions. Either or both of grids 64 and 65 may be omitted in which event the negative ions attracted into the mesh of screen 63 will then be attracted by grid 66.

There are three groups of grids of three grids each, the grids of each group being closely spaced as compared to the spacing between groups.

Grids

66, 66 and .66" are connected to lead 66a. Grids 67, 67 and 67" are connected to lead67a.

Grids

68, 68 and 68 are connected to lead 63a. Leads 66:: and 68:: are connected to a source of positive polarity and lead 67a is fed with radio frequency potential. The distance between grids 67 and 67 is preferably different than that between grids 67 and .67 and the two distances preferably are according to the ratio of 7 to 5.

The grids 69 may be used as repelling electrodes by applying a negative potential thereto. This potential is selected as described in said article and in said prior copending application, and is of value in repelling background electrons and ions leaving only ions that attained more than a predetermined velocity. However, grids 69 may be omitted if desired. .Grid70 is connectedto cylindrical shield 70s and to lead 70a which in turn may be connected to lead 69a. The collector 71 is connectedto a source of positive potential through its lead 71a, and it will attract negative ions that passed all of the other grids. Since collector 71 is made as positive as any other electrode in the tube, or preferably more so, it will repel any positive ions that have passed the grid 70.

Figure 3 illustrates in more detail, the leads from the tube. The several gridsare knitted wire screens mounted in washer-draped disc holders, which holders are supported by horizontal rods 72. Instead of bringing the leads vertically throughthe side wall of the envelope they pass horizontally through the rods 72 and out the right hand end of the envelope as shown in Figure 3. Cylindrical wire screens 73 connect

grids

66 and 66" and grids 68 and 6S inside of the tube. It should be noted that the uppermost horizontal rod 72 is not directly between cathode 59 and thespace betweengrids 61 and 63, hence there is nothing to prevent electrons from the cathode from passing to electrode 62.

In Figure 4 a complete schematic diagram for operating the tube as a negative ion spectrometer. The amplifier and measuring instrument 8!) is connected between the collector 71 and ground, and measures the fiow of negative ions to the collector 71. Variable frequency oscillator 81 provides the variable radio frequency potentials, and for many types of work a suitable frequency range may be from kilocycles to 10 megacycles. The output of oscillator 81 is fed through condenser 33 and reactor 84. The drop across reactor 84 is fed to the middle grids 67, .67 and 67", and also the leads 66a and 66a through resistors 66b and 68b. Vacuum tube voltmeter 82 measures the potential of the oscillator 81 whereby the latter may be adjusted and held constant.

in order to fully understand the theory of this invention it is desirable to review some basic facts. I11 Figure 5 there is illustrated the yield of positive ions for a typical gas as the ionizing potential P is increased. This curve assumes that a typical gas is admitted in to a typical mass spectrometer tube. The ionizing potential P, in the apparatus shown in Figure l, is the average potential of grids 6163, relative to the cathode.

Figure 6 shows that in the case of carbon monoxide there are also negative ions produced at an appearance potential in a narrow range of ten to twelve volts and again in a broad range at voltages in excess of twenty volts.

Figure 7 shows the yield of negative ions when oxygen is the only gas in the tube. As shown there is no yield of negative ions for potentials between about eight and sixteen volts.

Nitrogen gas and inert gases will not yield negative ions when bombarded with electrons.

'With the foregoing background information assume a mixture of gases that may contain nitrogen C2H4 and carbon monoxide. Further assume that it is desired to know if carbon monoxide is present, and if so, how much. First an analysis, may be made according to ordinary methods of mass spectrometry as contemplated in said article and in my said prior copending applications. This would give indications for mass lines numbers 1, 12, 13, 14, 15, 16, 24, 25, 26 27 and 28. It could be concluded that there was present a mass of value 28 but it would be impossible to say what it was or how much of it was present.

According to the present invention the unknown mixture is fed into inlet 54 after the focus is suitably adjusted and the potential P set for about ten to twelve volts. If the bombardment produced negative ions for mass line 16 it would be obvious that carbon monoxide was present. By adjusting P to give the maximum intensity and measuring the number of such ions it is possible to conclude how much carbon monoxide is in the mixture. This would be done with the connections shown in Figure 4.

-While I have illustrated my invention by showing how to detect carbon monoxide in the presence of nitrogen and ethylene, it is understood that the broader aspects of this invention are not limited to particular gases but cover the steps whenever applied to gases involving similar problems to those discussed herein.

To illustrate how the present method may be performed with other apparatus, I will describe now how the method may be practiced with the apparatus shown in Figures 1 and 2 of U. 5. Patent No. 2,387,786 to H. W. Washburn, granted October 30, 1945, entitled Analytical System. To practice my method, a change in the Washburn apparatus is required in that the polarity of electrodes 21 and 23 ,with respectto each other must be reversed. This may be done by reversing the polarity of the battery in the ion beam deflection control circuit. The electrode 25 will then be highly positive. With this modified apparatus one would practice my method by fixing the potentialbetween the cathode ,17 and the drift space between electrodes 21 and 23 at the critical value, which in the case of carbon monoxide is ten to twelve volts. A very sensitive recorder would be employed and would indicate the presence of negative oxygen ions and thus indicate the presence of carbon monoxide.

I claim to have invented:

1. The method of analyzing a gaseous mixture for the presence of a particular component thereof comprising the steps of bombarding the gaseous mixture with elec trons of lowest energy range known to produce negative ions from the particular component, providing a segregating field of intensity required to segregate the negative ions from said particular component from ions of said other components of said gaseous mixture, and indicating the amount of said segregated ions present.

2. The method of analyzing a gaseous mixture for the presence of a particular component thereof, said component being characterized in that it will produce negative ions when bombarded with electrons within either of two ranges of velocities one of which ranges is a narrow range of velocities as compared to the other and the ionizing electron velocities of the narrow range are lower than those of the wider range, comprising the steps of bombarding the gaseous mixture with electrons whose velocities are within said narrow range, producing a segregating field of intensity required to segregate the negative ions from said particular component from ions of other components of said gaseous mixture, and indicating the presence of said segregated ions.

References Cited in the file of this patent UNITED STATES PATENTS 2,221,467 Bleakney Nov. 12, 1940 2,490,278 Nier Dec. 6, 1949 2,535,032 Bennett Dec. 26, 1950 OTHER REFERENCES Radio Frequency Mass Spectrometer, National Bureau of Standards Tech. News Bulletin, vol. 32, September 1948, pages -108.

Radio Frequency Mass Spectrometer, by Bennett, published in Journal of Applied Physics, vol. 21, February 1950, pages 143-149.

A Mass Spectrum Analysis of the Products of Ionization by Electron Impact in Nitrogen Acetylene, Nitric Oxide Cynagen and Carbon Monoxide, by Tate et al., published in Physical Review, vol 48, September 15, 1935, pages 525-531.