US2768303A - Mass spectrometer - Google Patents

Description

Oct. 23, 1956 z. L. BAY

MASS SPECTROMETER 3 Sheets-Sheet 1 Filed Oct. 26, 1954 5 .QQNIZATION R O T N E V m ZOLTAN L. BAY

ATTORNEYS United States Patent Ofiice Patented Oct. 23, 1956 MASS SPECTROMETER Zoltan L. Bay, Chevy Chase, Md.

Application October 26, 1954, Serial No. 464,789

16 Claims. or. 250-413) particular ion mass accelerated through an electric field of known potential.

Important scientific applications of the mass spectrometer are: to survey the mass composition of a given sample; technical mass determinations; the determination of molecular weights of complicated organic and inorganic compounds; and for gas analysis and the composition of hydrocarbon mixtures. For these applications and purposes, it is desirable to have a mass spectrometer of small size that is a simple, easy to handle and rugged instrument. The device should be capable of serving for the qualitative detection of the kind of mass present with speed and convenience, and also be satisfactorily workable for the more difficult task of quantitative mass determinations.

Production and the control of the ion flight within the spectrometer envelope and the detection of ions arriving at the receiving electrode have posed ditficult problems. Ions may be produced by electrons emitted by a hot cathode electrode or by infrared and visible radiations and these electron or radiation fields should have no appreciable effect on the ion detection means. Electron and ion paths within the spectrometer may be separated by combined electric and magnetic fields. But an important feature of this invention eliminates need for a magnetic field or magnetic apparatus. According to the invention, ions are detected by the electric pulses produced on their arrival at the plate of the spectrometer, the pulses produced by different ion species being selectively gated by means having a determined time delay with respect to the initiation of the ion travel to the plate. By selectively varying the gating delay times, only the ions corresponding to definite components of the compound will be detected, while the relative amplitudes of the gated pulses produced by these ions give an indication of the relative quantities of the different ion masses. This gating may be accomplished either internally of the spectrometer envelope or externally. When internal gating is used, ions are detected by secondary electrons given off from the plate electrode of the spectrometer by the impact of ions and an electrostatic barrier field is arranged to efiectively keep ion generating electrons on one side and secondary electrons on the other side. One form of external pulse gating will be shown in which a pulse coincidence technique is used to gate only pulses of selected time delay subsequent to initiation of the ion burst.

Other ideas of this invention serving to improve the action and performance of a spectrometer may be considered.

Combining the ideas of generating ions in bursts of very short time duration with short ion flight times and small flight distances serves to increase the resolving power of a spectrometer and differentiate between ions having masses within one percent of each other. The very short electric pulses which are necessary for the production of bursts of very short duration may be generated by pulse techniques and apparatus understood and used in the electronics field. Short ion flight times and distances have an important advantage in that any desired potential distribution along the path of ions can be obtained with good accuracy and a consequent efiectiveness of the mass reso lution. Further, intense ion beams can be obtained to give a high order of sensitivity to the spectrometer.

In accordance with the invention, measurement of short ion flight times having a high degree of accuracy has been found possible by a delayed coincidence method. Here, short voltage pulses which control the ion source are used again after a predetermined time delay to register the ions arriving at the plate electrode of the spectrometer. Only such ions as have a flight time coinciding with the pulse delay time will register. Since flight time is dependent upon the ion mass and accelerating voltage, this accelerating voltage can be controlled or varied as required for ions having differing masses.

For best measurement accuracy, ions should be time focused or behave as if they all start from a fixed point or area, which may be an inner cylindrical surface in a spectrometer of convenient cylindrical symmetry and con struction. A close approximation of such a focus effect may be realized in practice by simulating a field of parabolic potential distribution: in such a field, the ion time of flight between any point and the focus point is constant and ions arrive at the focus point at the same time independent of actual ion originating points. A good approximation of a field of a parabolic potential distribution may be set up within a spectrometer envelope by the use of a grid electrode properly spaced and biased.

In view of these general ideas and features pertinent to a time of flight mass spectrometer design, the objects of this invention are:

To provide a mass spectrometer design using short times of ion flight in dimensions of small space;

To provide a design adapted for a measuring method using secondary electrons;

To provide a simple time measurement means;

To provide a means for time focusing ion beams;

To provide a design having cylindrical symmetry adapted for easy construction by present radio tube manufacturing techniques;

To provide a construction having electrical elements adapted for operation by simple electronic circuits;

To provide detail improvements increasing accuracy and resolving power of a spectrometer.

Further objects of this invention will be understood from the following specification considered with relation to the drawings.

In the drawings:

Fig. 1 is a vertical sectional view of a mass spectrometer simplified to show essential ion control elements of this invention;

Fig. 2 is a sectional view taken along the line 22 of Fig. 1;

Fig. 3 is a schematic diagram of a pentode electron ion generator suitable for construction within the inner portion of the mass spectrometer of Figs. 1 and 2;

Fig. 4 is a diagram of voltages representative of those applied to the ion and electron control elements or electrodes of the spectrometer of Figs. 1 and 2;

Fig. 5 is an enlargement of a part of the diagram of Fig. 4;

Fig. 6 is a diagram of apparatus for operating the spectrometer of Figs. 1 and 2;

Fig. 7 is a diagram illustrating a form of the time focusing idea of this invention;

Fig. 8 is a voltage diagram similar to that of Figs. 4 and 5 but for a spectrometer having fewer electrodes;

Fig. 9 is a schematic sectional view of a structure combining the elements of Figs. 1 and 3;

Fig. 10 is a schematic diagram of an arrangement corresponding to Fig. 8;

Fig. 11 is an idealized graphical representatipn of the results of a typical analysis; i

'Fig. 12 is a schematic diagram illustrating the principle of external gating; and

Fig. 13 is a circuit diagram, corresponding to 6, but employing external gating.

With reference to'Figs. 1, 2 and 9 of the drawings, 1 is the spectrometer envelope or tube which may be made of glass and, in a typical embodiment, may have, a diameter of 7 centimeters andalngth of 5 centimeters. Details of construction of the spectrometer tube will closely accord with practices commonly employed the manufacture of radio tubes wherein a pluralityof grids, a plate and a cathode having a heater are mountedwithin an envelope which can be evacuated. Constr uction of this spectrometer tube differs from that of a ra 0. tube primarily in that: the spectrometer requires. an opening 2.

which can be closed or opened readily for introduction and removal of test samples; and, it is desirable to provide a ground glass or other suitable impernianent seal 3 to permit replacement of a cathode or other central elements of the spectrometer in case of failurewithout destruction of the major part of the structure. Functioning electrodes of the.spectrometenexeept, for

the ion generating source, include grids 4,5, 6 and 15 and plate 7 which are suitably spaced and supported as by mica discs 8 and 9 and have electrodeeleads (not:

shown) for external electrical connections brought through the base of the spectrometer tube. These grids and plate are coaxial and of cylindrical shape and are accurately spaced as will be explained hereinafter indetail. Plate cylinder 7 preferably has a diameter of 6.4 centimeters and grid 4 may have a diameter off2 centimeters. Exact length of the cylindrical electrodes 4-7 inclusive is not critical and the proposed length here is about 4 centimeters.

The ion generating source of the spectrometer is essentially an electron tube such as the pentode structure represented schematically in Fig. 3. The only changes in the ion generating source from the construction 'of a typical electron or radio tube are: plate 4 of Fig. 3 is made in the form of a grid to permit electrons to pass therethrough; and part of the structure including at least the heater and cathode is preferably made removable and replaceable in case of failure. This ion generating source may be a pentode tube structure with elements including a cathode 10, a heater 11, a control grid 12, a screen 13, a suppressor grid 14, and a screen plate 4 which is idcn tical with grid 4 of Figsfl and 2. All or" these elements should have a cylindrical symmetry. Replaceable parts of this pentrode electron tube structure may be based on a suitable ground glass plug fitting the ground glass seal 3 of envelope 1 of Figs. 1 and 2.

Operation of the electron tube of Fig. 3 is effected by means of a suitable power supply apparatus and circuits of a type familiar in electronic equipment. Provision is made for pulsing control grid 12 by a circuit connection with a pulse generator as shown in Fig. 6. Details of the pulse generator and circuitry of the ion generator of Fig. 3 are not of immediate concern in this invention since these matters are well understood in the electronic field. For the purposes of this invention, the ion generator must be capable of delivering short ionizing electron pulses of a time duration of the order of 10 second. Sufiicient ionizing electrons may be delivered at a current of 1 to 30 milliamperes which is within the plate current range of a pentode tube structure similar to that of the familiar radio tube known as the type 6AC7.

With the insertion of the ion generating means of Fig. 3

in the spectrometer structure of Figs. 1 and 2 as shown in Fig. 9, essential operating elements of the spectrometer device of this invention will be completed. Further consideration will be given to the purpose and function of individual elements, their relative size or spacing, relative applied potentials, and other details or accessory apparatus required to make the spectrometer serve effectively for its intended purposes.

Desirable small size in a spectrometer device is possible only where the ion flight distance is short and of the order of a few centimeters in length. For satisfactory resolution or the ability to recognize ions having differing mass numbers, the time duration of ion bursts. must be quite short relative to the ion flight time over the ion flight distance. Ion bursts of a very short time duration can be produced by similarly short ionizing electron bursts produced by suitable pulses applied to the control grid of the electronic ion generator. Means for producing pulses of a very short time duration havebeen developed in the electronic field apdparticularlyinthe: technique of photo multiplier tubes. For the purposes of this invention, pulses, electron bursts and ion bursts preferably have a time dura tion of, theorder, of 10- second. Such a pulse gcperatpr isschematically shown at 22 in Fig. 6. Any. suitable repetition rate of thesepulses, for example, 10 p l e e ec nd m y e sqd,

In the spectrometer of Fig. 9, ions are accelerated through a pote ntial appliedhetween grid 4 and plate 7 and theLiqn flight distance is the distance from grid 4 tol plate 7 which may have a length of 2.2 centimeters.

Ions having a mass M and an energy eVo (where ais.

When time t, distance L and the potential V0 are known or measured, the mass M can be determined.

For a preferred operating simplicity as atforded by application of the ideas of this invention, distance L of (l) is fixed by the positions of grid 4 and plate 7; and time t is also fixedby a time delay coincidence method whereby the ion.detecting means is activated by a pulse of fixed time delay following the ion generating pulse-as shown by pulsedelay network 24. in Fig. 6. Potential V0 is readily controlled and measured and the determination of ion mass is then. simplifiedto a measurement in terms of the value of potential Va.

Ion and electron control elements or electrodes 4, 5, 6,. 15 and 7 of Figs. 1 and 2 have applied potentials as indicated in Figs. 4 and 5 and these applied potentials or voltagesarederived from a suitable power supply unit of conventional type schematically indicated at 26 in'Fig. 6. In theinterest of clarity, grids 13 and 14 are omitted inFig. 6,. as their action is conventional. In view ofthe electric fieldscreated by these applied potentials, ionizing electrons from the generator within grid 4 are forced back, by thefield between .grids4 and 5 whilethe ions are accelerated toward electrodes. 5 and 7. The important ionizing region, of the mass spectrometer is in the spacebetween grids 4 and 5. and quite close to grid 4. Grid 5 and plate .7 are. operated at-a predetermined fixed potential difference.

Between electrodes 5 and 7 are grids 6 and 15; grid 6 serves as a collector of secondary electrons ejected by ion impacts on plate 7 while grid 15 serves as a time controlled gating grid for the passage of secondary electrons from plate 7 to grid 6. Grid 15 receives a delayed pulse voltage Pd in addition to the direct current potential indicated in Fig. 5. Since grid 15 has a steady potential normally negative with respect to the potential of plate 7, secondary electrons can pass through grid 15 only when the pulse voltage Pd raises the potential of grid 15.as, for example, from 2l5 volts to 12.5 volts. Secondary electrons from plate 7 cannot enter the important ionizing region between grids 4 and 5 of the spectrometer because grid 5 is at a low potential relative to the potential of grid 6. As an example, relative positive potentials of electrodes 6, 15 and 7 may be respectively 15 volts, 2.5 volts and volts above the potential of grid 5 which is at a zero reference potential. A comparatively high positive potential of a few hundred volts may be applied to grid 4.

As will be apparent from Fig. 6, production of ions occurs in very brief pulses whose timing is determined by operation of pulse generator 22, these pulses also appear, after a suitable delay determined by the characteristics of network 24, at grid 15, and raise its potential to the necessary gating level. If, at that time, secondary emission electrons have been produced by arrival at 7 of the ions released by the same pulse from the ion generator, then a charge appears at collector electrode 6 due to the arrival of the gated secondary emission electrons from plate 7 through pulse gating grid to collector electrode 6, the distributed capacity of which is represented by a dash-line condenser 16. The charge suitably amplified at 28 is impressed on oscilloscope 30. If the accelerating voltage between screen plate 4 (acting as cathode of the mass spectroscope) and plate 7 is modulated cyclically by the output of source 32, which may be a 60 cycle source, then for every particular mass of ion in the field the corresponding voltage will be reached at some phase of each cycle, which is necessary to give it just that acceleration which will cause secondary emission due to the ions of that mass reaching plate 7 at the correct time to pass gating grid 15, and so a pulse will be periodically reproduced for that mass at a particular point in the sweep of the oscilloscope. The resulting trace on the scope for a gas containing ions of three masses, A, B, and C'respectively, will appear as in Fig. 11, the relative heights of the respective pulses being related to the proportions of the respective quantities.

It will be apparent that vwhile a cathode ray oscilloscope is shown for registering the results, any other known means for measuring the transit times as related to the accelerating voltage may be employed. For example, if the device is used for testing for the presence or absence of a single element, as in a leak detector, then the system can obviously be simplified; a fixed accelerating voltage would then be employed as required for that particular element, and so related to the delay that only when that element is present is a pulse gated through. The oscilloscope may then be dispensed with and a simple pulse detector associated with grid 6 is sufiicient for the purpose. However, the method is fundamentally the same in both cases and it will be clear that many other variations'in the application of this method of determining the mass of ions can be employed as required by different types of research. It is also clear that instead of cyclic variation of the accelerated voltage, this could be varied manually in small steps and a reading taken at each step. The variable parameter need not be the accelerating voltage, as shown by way of example, but could be the delay time, as the delay can obviously by varied either cyclically or in definite steps while maintaining the accelerating voltage constant, to derive the same information.

An alternative arrangement for time controlling the passage of secondary electrons from plate 7 to grid 6 is represented in the voltage diagram of Fig. 8. Here the gating grid 15 is eliminated and plate 7 itself receives a delayed pulse voltage Pa of a value sufficient to lower the potential of plate 7 with respect to the potential of grid 6. When plate 7 is pulsed and grid 6 is at a higher potential, secondary electrons emitted from plate 7 are collected by grid 6. This alternative arrangement for controlling secondary electrons may be employed in a spectrometer tube design where the relative radii of the ion focusing elements are such as to make the space between grid 6 and plate 7 quite small. As a further alternative, in the arrangement of Fig. 8, the potential of 6'. plate 7 may be maintained constant and grid 6 would be positively pulsed to attract secondary electrons emitted from plate 7.

By reason of the pulse timing coordination employed in this invention where a pulse Pi is applied to the ion generator and is delayed to appear as a delayed pulse Pd applied to the secondary electron sensing elements 6 of this spectrometer, it will be clear that only ions of a particular mass number will be registered on the oscilloscopes screen for a definite value of the ion accelerating voltage V0. This registration is accomplished by detecting the arrival of secondary electrons at grid 6, which produces a negative voltage pulse at this grid. Ions must travel the fixed flight distance in the fixed delay time t in order to register at the secondary electron collector 6 which is activated only at time intervals t. Where ion generating pulses having a time duration of l0 second are employed, these pulses may be repeated at a rate of 10 times per second. Ions of dilfering masses can be registered by changing the ion accelerating voltage V0. Since accelerating potential V0 must be changed between very great limits in order to scan a large mass range, the delay time t may be varied in suitable steps as by using any one of the steps t1, I2, is tn. These times t1, t2, etc. would of course be selected to correspond to the different masses which are being investigated, that is, they would correspond to the ratios of the masses of the successive elements.

Continuous operation of the time of flight mass spectrometer of this invention is effected by adding to the direct current accelerating potential V0 an alternating current potential Vt, which can be 60 cycles per second. Potentials V0 and Vt combine to give a slowly changing accelerating potential Va. which will be considered with relation to the pulses P1 and Pa which may be repeated at the rate of 10 per second or more. For every mass M1; there exists a definite voltage interval Vk--V k of the accelerating potential Va Within which secondary electrons ejected from plate 7 by ions of mass Mk arrive at collector grid 6. The potential Va will be within this potential interval during a time interval llr-l'k. If the total variation of potential of Va covers, for example, a mass range of 20 mass numbers as from M to M plus 20, the potential Va is in the voltage interval Vi V1; for approximately of the cycle period of of a second. During this short period or" about V1200 part of a second there are from 10 -10 pulses. This large number of pulses in a short time interval permits secondary electron collector grid 6 to pick up a relatively large number of secondary electrons with a resulting sizable value of signal current and an important improvement in the effective signal/noise ratio of the measurement.

Essential circuitry and accessory apparatus required for the operation of the spectrometer of this invention are represented in the simplified diagram of Fig. 6. A pulse generator of a suitable type delivers a pulse Pg to the electronic ion generator of the spectrometer and a similar pulse Pg to a delay network which may be a coaxial cable and the output pulse Pd is applied to grid 15 of the spec trometer through a capacitor 18. Resistor 19 supplies grid 15 with a potential as indicated in Fig. 5 from a suitable power unit which also supplies by conventional circuitry other electrodes of the spectrometer with required operating voltages and power. Grid 6, which is the collector of secondary electrons as has been described hereinbefore, is coupled with an amplifier whose output is connected with an oscilloscope for visual indications.

The charge of secondary electrons delivered by 10 -10 pulse periods received by grid 6 is discharged through resistor 17 where resistor 17 with capacitor 16, which represents the distributed capacitance of grid 6 may have a time constant 10 second. Since this time constant is in the audio frequency range, the pulses appearing at grid 6 as a result of the cumulative pickup and discharge of secondary electrons are readily amplified by an audio amplifierhaving a. cut-otf at a frequency which may be as low as. 1000 cycles per; second. Ionizing pulses P and delay pulses Pa repeated at the rate of 10 timesper; second, will not be passed by such an audio amplifier, but the pulses of secondary electrons group collected during a time interval such as tk-tk corresponding to masses Ms present, in the spectrometer. will beamplified fora showing on the oscilloscope or other indicator. By applying the alternating component Vt of the ion accelerating voltage Va as a sweep to an oscilloscope, different masses present in the spectrometer are represented separately on the screen by deflections proportional to the different ion mass concentrations in the gas sample under analysis.

It will be clear, then, that the method and means for registering ions by collection of secondary electrons has much latitude and practical adaptability. The cumulative charge of the collector grid and its discharge through an RC circuit gives voltage pulses of appreciable magnitude with a practical elimination of serious ditliculties associ ated with the handling of otherwise weak voltage changes both within the spectrometer itself and outside the spectrometer in the amplifier and indicator. No magnetic fields are needed. An audio amplifier of simple design and a standard type of oscilloscope serve satisfactorily for amplifying and indicating purposes. This audio frequency type of measurement eliminates compensating difficulties inherent in a direct current type of measurement such as: need for very stable voltage conditions and their repeated checking; and need for compensating the current of grid 6 for current caused by ions striking it and for the secondary electrons striking grids 5 and 15. Also, the audio type of measurement does not make it necessary to keep background current at a minimum as may limit grid area to using thin wires widely spaced. Limited grid areas give a poor electric field distribution detrimental to spectrometer performance.

The preceding description shows the manner of using internal gating to detect the pulses. It is also possible to detect the pulses by external gating, as will be shown below. This has the advantage of requiring a simpler tube construction, since secondary emission gating is not used, and, therefore, electrodes 6 and 15 of Fig. 9 can be eliminated; except for this the tube may be as shown in Fig. 9; alternatively, a linear tube construction may be employed. The detection will in this case take place directly at the plate 7, which is connected to the external gating circuit.

The original ion pulse appears at the plate as a group of n pulses, where n is the number of components of the mixture. All of these 11 pulses appear within a short time which is a fraction of the time of flight of the slowest ions of the mixture. It is clear that after indicating one of the 11 pulses the indicating device must rapidly come back to its zero position in order to be able to indicate separately the next ion pulse of the group. This requires an indicating device capable of high speed operations. Let the pulse period of the original ion pulse be denoted by T, then T is also the period of each of the separated ion pulses of the group arriving at the plate and the speed of operation of this indicating device must be such as to resolve signals separated by the time T or smaller. This means in the well known language of radio technique that the band width of the indicating device must be at least in the order l/T sec- On the other hand, in the new method, a gating is applied which is elfective only for one pulse of the group of 11 pulses and, therefore, the indicating device is not required to come back to its zero position rapidly after accepting the signal because in this method there is no need to indicate the next pulse of the group. This means that the output pulse of the indicating device can be several orders of magnitude longer than the input ion pulse and the frequency band width can be chosen several orders of magnitude smaller than l/T.

,- tional to the pressure.

Asis:we1l.kn,own;. the inherent noise of any indicating device: (including. amplifiers, instruments, oscilloscopes, etc.) is-.pr.oportional to the square root of the band width of the device. lower bandwidth for the detecting device which is equivalenttto a better'signal to noise ratio. Since the signal to noise ratio determinesthe lower limit ofsensitivity, it is clear'thatrnuch weaker signals can be detected with the method of the invention. In the language of'mass spectrometry, this means that much weaker components of a gaseous mixture canbe detected, even though the detecting device is l simpler and cheaper.

Another advantage-of the new method is that by repeating the original ion pulses the long time constant of the indicating deviceproduces a comulative effect. This increases the signal to noise ratio further and this increase'can amount to several orders of magnitude.

Another advantage of the new method is that the pulse period and time constant of the indicating device are independent quantities. This means that T can be chosen as small as permitted by the limits of known pulse techniques. (e. g., T=l0- sec. or less), while the amplifiers usedin the indicating can operate at, e. g., audio frequencies.

Another advantage of the new method is that, choosing T very small, the distance of flight of the ions can be very small1(e. g., if*T=1O' sec., the distance of flight can be as small as approximately one cm.). It should be noted that decreasing the distance of flight, i. e., decreasing the sizeof amass spectrometer, is not only a matter of convenience. The number of collisions between ions and neutral atoms of the gas inside the spectrometer. is proportionalto the path length and propor- Thus:a small path affords the possibility of operating at higher. pressures and thus detecting weaker components.

The methodillustrated by Figs. 1-10 utilize the gating ofsecondary electrons released by the impact of ions. In the following, description, there is described the useof external coincidence circuits for gating the ion pulses. Fig. 12 shows-the general blockdiagram of the equipment for utilizing the new method. P is the plate of the mass spectrometer tube at which the ion pulses arrive, correspondingto plate 7 of the preceding figures. P is connected through channel A to a coincidence device. The

pulses from the pulser, (which is not indicated in the diagram) travel through channel B and, after a properly chosen time delay, arrive at. the input of the coincidence circuit. The coincidence device is a circuit which gives an output only if two pulses, appear simultaneously in the individual channels A and B. The circuit can be constructed in such a way that the amplitude of the output is proportional to the product of the amplitudes of the coinciding pulses in the channels A and B. Since the pulses of channel B are always of the same amplitude, the output pulses of the coincidence device is proportional to the amplitude of A, i. e., the number of ions arriving at the plate in coincidence with the measuring delayed pulse in B. Channels A and B are constructed so as to have a very broad frequency response which can extend far beyond 1/1". A suitable method is to use concentric cables of very small attenuation up to frequencies 10 secf At the input of the coincidence device the cables are matched, to give no reflections. It is easy to design the inputs of the coincidence device-to have time constants of the order 10" sec. or less, so no appreciable memory of the pulses remains in the input after a time interval of 10- sec. The output of the coincidence device can easily be constructed to have a time constant of unlimited length, e. g., 10- sec., even up. to several seconds. In consequence of this, the output pulses can be amplified by means of audio amplifiers of very narrow band width, The measuring or indicating instrument can The use of the new method: results in a.

greases 9. be an oscilloscope or any current or voltage measuring device.

In order to obtain the entire mass spectrum of the mixture to be analyzed, the measurement is performed stepwise as previously described, changing either the accelerating voltage inside the mass spectrometer tube or changing the time delay of the measuring (gating) pulses in channel B and relating the mass of ions to the varied parameter.

The coincidence device in Fig. 12 can be any type of known circuits called in the literature, coincidence circuits or gating circuits. The special problem in connection with the present invention is to adapt such a circuit to indicate very small pulses, because great sensitivity is required of the mass spectrometer. A special circuit developed for that purpose is shown in Fig. 13.

In the circuit of Fig. 13, 41 and 42 are shielded cables leading to the two respective inputs of the coincidence circuit. Resistors 43, 44 and 45 are the so-called matching resistors chosen so as to avoid reflection. 46 and 47 are rectifier diodes (crystal diodes or tube diodes), the arrows indicating the direction of positive current through them. All the parts, 41-51 together, compose the fast part of the circuit capable of accepting and handling very short pulses if the elements are properly chosen. The magnitude of the time constant of this part of the circuit is the product of the matching resistance and the capacity of the diodes, e. g., if the matching resistance is in the order of 100 ohms and the capacitance of the diodes in the order of 1,u .L f., the time constant is sec. It should be noted that the condensers 4851, the capacity of which can be 10-100p41. f., do not contribute to the time constant because neither side of them is grounded.

The parts 52-60 on one side [and 61-69 on the otherside represent the slow parts of the circuit. Both sides are similar and either of them can be used separately or in combination. In the description of the operation of the device, only one of them, the parts of which are numbered 52-60, is discussed in connection with the fast part of the circuit.

The delayed gating pulses of the pulser appear in cable 41, the ion pulses of the plate 7 of the mass spectrometer appear in cable 42. To understand the operation of the device assume first that there are pulses in cable 41 but no pulses in cable 42. In this case point a is at ground potential, the same as point b. In this case, the branches consisting of 48, 46, 49, 44 on the one hand and 50, 47, 51, 45 on the other hand are identical and thus a short positive pulse in cable 41 causes the same amount of current through the diodes 46 and 47. These short current pulses through 46 and 47 charge the condensers 49 and 51 positively and the condensers 48 and negatively and if the diodes and the condensers are similar, all the charges will be of the same amount.

After the decay of the supposedly short gating pulse the condensers cannot discharge through the diodes because of the opposite polarity. Thus the charges flow over to the slow parts of the circuit. Consider now the lower part of the circuit diagram. The negative charge of. 48 flows through 52 to the RC circuit consisting of condenser 54 and resistor 56. Similarly, the positive charge of 51 flows through 53 to the RC circuit 55, 57. The time constants of these RC circuits are equal and they can be very long with respect to the pulse period of the gating pulses.

Assuming equal charges coming from 48 and 51 and equal time constants of the RC circuits, the voltages across the condensers 54 and 55 are equal, but of the opposite sign, during the entire discharging period of the circuit. This means that connecting the upper sides of 54 and 55 through the equal resistors 58 and 59, the common point of 58 and 59 remains'at zero poment of each spectral line.

tential during the entire discharging period. This point is considered as the output of the coincidence circuit and is connected by line 7% to the oscilloscope. It is possible that the characteristics of diodes 46 and 47 are not exactly equal in which case the said charges differ by a small amount. To balance the output 60, in the balance position potentiometers provided for 56 and 57 and 58 and 59 are connected to the moving taps achieving hereby the balancing of the output. Consequently, after balancing the circuit no output pulse appears, if there are pulses only in cable 41.

It is easy to see that there will be also no output pulses, if there are pulses only in cable 42. Assuming these pulses are positive, diode 46 will not conduct; therefore, no charges will be generated in the circuit.

It is also easy to see that in the case of coincidence there will be an output pulse because the voltage across diode 46 will differ from the voltage across diode 47; therefore, different charges will be generated in 48 and 51. The amplitude of the output pulses is, in this case, proportional to the pulse amplitude appearing in cable 42 and its period will be given by the time constant of the RC circuits 54, 56 and 55, 57. The resistors 52 and 53 are chosen to be one order of magnitude smaller than 56 and 57, and the resistors 58 and 59 one order of magnitude larger than 56 and 57.

The slow circuit 6169 in the upper part of Fig. 13 is built in a similar way to the lower part to assure symmetrical operation of the circuit. Its output, 69, gives pulses of the opposite polarity of the output 60.

Both of the outputs 60 and 69 can be used separately or they can be added after reversing the sign of one of them in the later amplifying process, as shown in Fig. 13.

It is clear that the amplifiers used to amplify the output pulses do not need to have a broader band width than the order of magnitude of l/RC seci. e., in the new method it is not necessary to use high speed amplifiers. It should be pointed out again that this is an essential feature of the invention and not only a matter of convenience. It means that small signals, i. e., a small number of ions, which are below the noise level of the high speed amplification process and which therefore are entirely undetectable with the old method, become detectable by the method and apparatus of the invention.

It should also be noted that there is no upper limit in the choice of the RC time constant and therefore no theoretical lower limit of the charge (number of ions) detectable with the new method.

A further advantage of the invention is the possibility of accumulation of signal charge when applied to the method of mass spectroscopy in which the ionizing pulses are repeated in time. It is clear from the above description of the operation of the coincidence circuit that the actions of all of the selected ion pulses which appear Within the time RC, will be added in the output. The time of accumulation is not even restricted to the RC time constant of the coincidence circuit proper. After amplification in the frequency range extending up to l/RC seerectification and prolongation of the pulse to a much greater time period can be effected and in this case the accumulation is effective within this much greater time interval. The time of this prolongation is theoretically limitless and it depends only on the time the experimenter wants to spend on the measure- It can be, e. g., 10 sec. in the case of scanning the mass spectrum on an oscilloscope, as described above; it can be extended to many seconds or minutes in the case of D. C. measurement of extremely weak components. It should be noted that this is not true of the methods of measurement shown in the prior art where the indicating device must come back to its zero position Within the time separation of the pulses in order to be able to indicate the next ion pulse.

In addition to the accuracy of the time measurement factor, the performance and realized accuracy of the complete spectrometer system depends also on correctly knowing the value of the ion accelerating potential and the precise construction of the spectrometer tube. Known pulse techniques together with the ion registering method and apparatus hereinbefore described are adapted to provide a time factor of satisfactory accuracy. Familiar voltage supply apparatus or power supplies for electronic apparatus are adequate to correctly provide an ion accelerating potential of any desired type. The direct precision required in the construction of the spectrometer tube does not exceed that employed in the manufacture of radio tubes; in fact, the spectrometer tube may be made in substantially the same way as a radio tube having either a glass or metal shell. But in a time of flight mass spectrometer tube, consideration must be given to the fact that ions are generated in a region rather than in a point generated line or surface. Therefore, the exact flight distance of ions cannot accurately be taken as the distance between two electrodes having a point generated line or surface configuration; In an ideal spectrometer tube of cylindrical construction, the ion flight distance would be the distance between two coaxial cylindrical electrodes of zero thickness.

A time focusing idea of this invention is practically applied to make ion time of flight distances accurately uniform even though the ion originating points do not conform with ideal conditions. This ion focusing scheme may be explained with reference to Fig. 7 which is simplified to showing of electrodes 4, and 7 only of the spectrometer of Fig. l. The ideal ion flight distance would be the distance from grid 4 to plate 7; but, since ions are generated in the shaded region 20 of Fig. 7, ideal conditions cannot be realized. However, if ions lo, 11, and I: of the same mass but originating at different points in region 26 are considered, it will be clear that the-v can have the same flight times if a compensating difference in accelerating potential can be given to each ion. For example, if ion 10 starts from grid 4, ion 11 starts from the center of region 20, and 12 starts from' the outer edge of region 29, respective decreasing accelerating potentials V0, V1, and V3 can serve to equalize the ion flight times over their respective distances from points of origin to plate 7. All ions of equal mass originating anywhere in region then are time focused to have the same time of flight and distance relation of an ideal ion It) which travels the exact distance from grid 4 to plate 7.

Consider ions of the same mass M, which are accelerated in an electric field of the potential V(x). They start with zero velocity at different points of the coordinate x5 (x5 is variable) at the same time t=0 and arrive at the coordinate xii (x3, is the same for all ions). The time of fiight over the distance Jar-x8 is t(xs) =const. The mathematical solution of this problem is the parabolic potential function, i. e.

which is independent of 2:5.

To provide such an electric field a uniform space charge would be needed therefore it can not be realized in vacuum. The idea of this invention is that such an electric field can be approximated by the application of a few grids (instead of the space charges), or with only one grid between as and Xa and high accuracy can be maintained if (instead of the exact solution) we look for the solution of the following equation:

Having fulfilled this condition, the time of flight will be constant with a high degree of accuracy. Where the starting zone of ions is a relatively small part of the flight distance, the flight time will be substantially constant, or

t(xs) :const.

The simplest arrangement for the realization of time focusing is that represented in Fig. 7 where a homogeneous electric field is produced between grids 4 and 5 by an ion accelerating potential V0 and a field of free space exists between grid 5 and plate 7. Simple calculations show that the distance between 5 and 7 should be twice the distance between 4 and 5 to fulfill the condition It is true that grids 6 and 15 of the spectrometer of Fig. 1 are in this space between electrodes 5 and 7, and electrodes 6, 15 and 7 have small applied potentials relative to accelerating potential V0; but, these differences from the simplest arrangement have only a small order effect requiring no more than a minor change in electrode spacings. The major feature for the realization of satisfactory time focusing is the production of an electric field between grid 4 and plate 7 which closely approximates the effect of a field of parabolic potential distribution. The potential V at any point in a field of parabolic potential distribution between two electrodes increases proportionally with the square of the distance from the electrode which is at the lower potential.

In the case of an electrode arrangement having cylindrical symmetry, the field and potential distribution between time focusing elements of the spectrometer differ from the parallel plane condition of Fig. 7. Here the time focusing is already partly achieved by the fact that a hyperbolic field between cylindrical grids is itself an approximation of the parabolic field. Therefore, this condition dt dz,

Using also grid 5 (out of other reasons to obtain secondaries in a field free space as released by ion impact at 7) the distance between 5 and 7 should be less than twice of the distance between 4 and 5 which utilizes the focusing effect of the hyperbolic field. The relative radii of cylindrical elements 4 and 5, which may be selected as desired within an operative radii ratio range, determine the focus area and the radius of plate 7.

It is convenient herein to give tables of representative dimensions which have been derived for the time focusing elements of the spectrometer of this invention. If three concentric cylindrical elements 4, 5 and 7 are used for time focusing, it will be clear that ions originating in an area close to element 4 will be accelerated toward andthrough grid 5 and will arrive in time focus at plate 7. Ions are accelerated in the space between grids 4 and 5 but travel the remaining distance to plate 7 in free flight. For time focusing, then, the ion free flight distance 13 isproportioned with relation to the ratio of the radii selected for grids 4 and 5.

Where .7'4 is the radius of grid 4 of Figs. 1 and 2, m is the radius of grid 5, and r7 is the radius of plate 7, a first case, Table I, represents a design employing a fixed distance d(45) between grids 4 and 5, while a second case, Table II, represents a design wherein the ratio of radii ra/r-i' is a constant which is here made 3.

Table l Cm. 0111. CD1.

Table II n tin-) r5 11(5-7) T7 001-7 0m. CD1. CD1.

In Tables I and II, the radii of elements 4, 5 and 7 are given together with the distances d(4-5), d(5-7) and d (4-7) between respective elements. It will be observed that in Table I the ratio r4/r5 has three diflferent values and tables similar to Table II representing only a change in size is readily worked out for these differing ratio values. For three electrode focusing, the ratio r5/r4 may be selected in the value range between 1 and 3.55. Where the ratio is equal to 3.55, elements 5 and 7 coincide and time focusing occurs between only two cylinders.

There is, then, much engineering latitude allowable in the application of the time focusing principle of this invention. Well known in ion sources of small dimensions generate ions of too small a number for good results in a time of flight mass spectrometer. focusing, spectrometer accuracy is not limited to the use of a small ion generating source and the inner cylindrical element 4 of Figs; land 2 is readily made large enough to contain electronic means of sufficient power for the generation of ions in great number.

It will be understood that complications arise in the operation of a spectrometer device which set a natural limit to the accuracy of measurements. Inaccuracies arise from the thermal agitation and are the same for all ions independent of mass but will be under 1 percent as represented by: errors of 0.8%, 0.4% and 0.25% for, respectively, 100, 400 and 1000 volt values for the ion accelerating voltage V0. The spread of initial energy caused by ionizing electrons is negligibly small except for ions of very small mass where mass differences are relatively great and separation accuracies are unaffected. In view of these and other effects understood in the technique of mass spectrometry, some inaccuracies will exist but can be kept small or well within 1 percent. The design of the spectrometer of this invention permits inaccuracies to be reduced to a satisfactory low order with a minimum of difficulty.

The cylindrical symmetry employed facilitates the production of an intense ion beam afiording a high sensitivity in measurements. With this symmetry, it is comparatively easy to predict the potential distribution in the spectrometer tube and maintain uniform fields. Cylindrical design simplifies production problems in apparatus requiring precise construction. High grade insulation is readily provided to permit high accelerating voltages to be used. However, it will be apparent that the invention is not limited to a cylindrical construction as shown, but is equally applicable to a non-symmetrical construction as in other known types of mass spectrometers wherein With time the ion generator is at one end of the evacuated envelope This application. is a continuation-in-part of my application Serial No. 262,725, filed December 21, 1951.

I claim:

1. Apparatus for determining the atomic mass of a component of a mixture and the abundance of that component of the mixture comprising means for subjecting said mixture to a very brief ionizing impulse to generate a group of ions of said mixture, a plate electrode, means for establishing an electric field of predetermined distribution and magnitude between the area of generation of said ions and said plate electrode to impinge said ions upon said plate electrode at a fixed distance from thearea of generation, means for delaying a pulse derived from said ionizing impulse for a predetermined time interval, means for utilizing said delayed pulse to gate the detection of said ions by an ion detecting circuit, the delayed interval in relation to the field parameters being determinative of the atomic mass of the gated ions, means for measuring a function of the abundance of ions so gated, cyclical means whereby said steps are repeated rapidly to accumulate the gating effect, and additional means for measuring the effect of a large number of said repetitions.

2. The invention according to claim 1 and means whereby said field magnitude is periodically varied through a predetermined range so that the distribution of atomic masses may be determined.

3. The invention according to claim 1 and means whereby the delay interval is periodically varied through a predetermined range so that the distribution of atomic masses may be determined.

4. The invention according to claim 1, said gating means comprisingv a plate electrode circuit for transmitting signal pulses produced by impingement of ions upon said plate electrode, and a coincidence device having means for producing an output only upon coincidence of said signal pulses and said delayed pulses, said measuring means being connected to the output of said coincidence device.

5. The invention according to claim 4, said coincidence device having respective input circuits for said signal pulses and for said delayed pulses, said input circuits having a time constant not greater than the individual duration of said pulses.

6. The invention according to claim 5, said coincidence device having an output circuit with a much greater time constant than the individual duration of said pulses.

7. The invention according to claim 6, said coincidence. device comprising a balanced input circuit for one of said sets of pulses, said balanced circuit comprising two branches connected at a common point to a pulse circuit, each branch comprising two capacitors in series with a rectifier diode in series between them, said rectifier diodes being similarly oriented, an impedance between the common point of said two branches and ground, means for feeding the other of said sets of pulses into the opposite end of one of said branches, and impedance between said opposite end and ground, and detecting circuit means connected across said branches on respectively opposite sides of said diodes, said branch circuits constituting the input circuit and said detecting circuit constituting the output circuit of the coincidence device.

8. The invention according to claim 7, said detecting circuit comprising a balanced circuit having two further branches of relatively long time constant for balancing 15 the outputs of said two first branches under conditions of non-coincidence of input pulses.

9. The invention according to claim 8 and means for adjusting said two further branches to obtain initial balance.

10. Apparatus for mass spectrometry comprising an evacuable gas impervious envelope for receiving a sample for analysis enclosing a plurality of spaced electrodes including a source generator of ionizing electrons having a grid anode and a plate electrode separated in space, means for pulsing the source generator of ionizing electrons, means for accelerating ions originating in the near region of the grid anode toward the plate electrode, means for collecting secondary electrons emitted from the plate electrode by the impact of ions, means for pulse controlling delayed in time the passage of secondary electrons from the plate electrode to the collecting electrode, and means for detecting the arrival of gated secondary electrons at the collecting electrode.

11. Apparatus for determining the mass of ions in a material in gaseous form, comprising means for subjecting said material to a very brief ionizing impulse to generate a group of ions of said material, means for producing an electric field of predetermined intensity between said generating means and a plate electrode located at a fixed distance from said generating means for accelerating said group of ions towards said plate to eject secondary electrons therefrom, a collecting grid for collecting ejected secondary electrons from said plate, detecting means associated with said collector grid for detecting secondary electrons received thereby, a gating grid between said plate and said collector grid normally biased to prevent passage of secondary electrons therethrough, and means for pulsing said gating grid with a biasing voltage of such value as to permit passage of secondary electrons therethrough to said collector grid, means for deriving biasing pulses from said ionizing impulses at a predetermined time delay with respect to said ionizing impulses.

12. The invention according to the preceding claim including means for varying cyclically the field intensity.

13. The invention according to claim 11 including additional grid means in said field, means for supplying electric potentials to said grid means to efiect the distribution of said field, the spacing and potential of said additional grid means being such that all ions of equal mass originating in the area of generation have the same time of flight between any point in said area and said plate electrode, the potential function defining the distribution of said field intensity being substantially parabolic.

14. Apparatus for mass spectrometry comprising an evacuable gas impervious envelope for receiving a sample for analysis enclosing a plurality of spaced electrodes including a source generator of pulsed ionizing electrons, a grid anode and a plate electrode for receiving ions spaced from said generator, means for producing an ion accelerating electric field between the grid anode and the plate electrode, the arrangement and geometry of the electrodes being such as to produce a time-focusing field distribution whereby ions of a given mass-charge at different locations in the vicinity of said generator are accelerated by said field toward said plate electrode so as to arrive at said plate electrode simultaneously, and means for detecting the arrival of ions at said plate electrode.

15. The invention according to claim 14, said detecting means comprising a collector electrode for receiving secondary electrons ejected from said plate electrode by the impact of ions, a gating grid between said plate and said collector electrode normally biased to prevent passage of secondary electrons, and means for pulsing said gating grid with a gating pulse of such magnitude as to permit passage of secondary electrons at a definite time interval subsequent to said first pulse.

16. Apparatus for mass spectrometry comprising an evacuable gas impervious envelope for receiving a sample for analysis enclosing a plurality of spaced electrodes including a source generator of pulsed ionizing electrons, a grid anode, a focusing grid electrode spaced a preselected distance from the grid anode, and a plate electrode for receiving ions spaced at a further distance from said grid anode, means for producing an ion accelerating electric field between the grid anode and the plate electrode, the arrangement and geometry of the electrode being such as to produce a time-focusing field distribution whereby ions of a given mass-charge at different locations in the vicinity of said generator are accelerated by said field toward said plate electrode so as to arrive at said plate electrode simultaneously, and means for detecting the arrival of ions at said plate electrode.

References Cited in the file of this patent UNITED STATES PATENTS 2,582,216 Koppius Jan. 15, 1952 2,642,535 Schroeder June 16, 1953 2,685,035 Wiley July 27, 1954 2,691,108 Berry Oct. 5, 1954 OTHER REFERENCES A Pulsed Mass Spectrometer with Time Dispersion, by Wolff and Stephens, published in The Review of Scientific Instruments, vol. 24, No. 8, August 1953, pages 616, 617.

Images