US2376877A - Mass spectrometry - Google Patents

Description

May 29, 1945.

co//ec fed car/enf R. v. I ANGMUlR 2,376,877

MAss SPECTROMET'RY s Q N REGULATED POWER SUPPLY INVENTOR.r ROB5RT ll. LANGMU/R.

REGULATED F/LAME/vr SUPP/ r Filed May 25; 1942 s ,Sheets-sheet :v

INI/NToR'T ROBERT u LANGMu/R.

Patented ay 29, w45

asian atraen ss srac'raoaaarnr Robert V. Langmuir, Pasadena, Calif., assigner to Consolidated Engineering Corporation, Pasadena, Salif., a corporation oi' ilaiiiornia Application May 25, 194:2, Serial No. @4,4911

(Cl. 'i3-i8) 7 Claims.

This invention relates to mass spectrometry, and particularly to methods and apparatus for recording a mass spectrum.

In general mass spectrometry involves conversion of a material under analysis into ions, segregation or separation of these ions into ion beams according to their mass-to-charge ratios, and measurement of the intensities of the respective ion beams. 'Ihe rates of formation of the respective ions in said conversion process vary with a large number of factors, such as the nature and energy of the ionizing force as Well as the amount and composition'of the material being analyzed. In the analysis of a material existing in gaseous or vapor form for instance, the sample is generally ionized by bombardment with low velocity electrons at low pressure, the ions formed then being accelerated by an electric field and ions of different mass-to-charge ratios thenbeing segregated and successively directed by the combined action of magnetic and electric focussing forces through a collector slit onto a collector where the ions discharge, thereby producing a series of ion currents which vary according to the intensities of the corresponding beam, or according to the rates of formation of the respective ions. The ion currents are usually measured by means of an ion collector circuit and meter.

All mass spectrometers commonly in use at the present time are of xed apparatus geometry, that is, the dimensions and arrangement of the ionization chamber, electrodes, collector slit, etc., are constant during the measurement or recording of a given mass spectrum. These devices are sometimes operated in such a manner as to maintain the geometry of the magnetic and electric 'lelds also xed. While for simplicity I shall illustrate the application of my invention to mass spectrometers of fixed geometry, that is mass spectrometers having both apparatus and iield geometry xed, it is to be understood that my in-4 vention is also applicable to known form of mass spectrometers of changing geometry.

In all mass Spectrometers in whichions of different mass-to-charge ratio are successively detected, the mass-to-charge ratio may be ex- I pressed as some power function of an independwhere K is a constant depending on the construction and r is a linear dimension of the mass spec- In a mass spectrometer of fixed appa- In using trometer. ratus geometry, r is held lconstantl mass spectrometers of fixed geometry, beams of ions of diierent mass-tocharge ratios are moved past the collector slit in a direction transverse to the velocity of the ions at the collector slit so that ions of diierent mass-to-charge ratio are successively focussed at the collector .slit and detected at the collector. The movement of the beams past the collector slit may be accomplished by adjusting the electric and/or magnetic elds which cooperate in the separation and/or focussing of the ions. The velocity with which a beam moves past the collector slit position depends on the rate at which said field or fields are changed. This velocity will be referred to as the beam speed.

If the electric field is the control variable being changed, it can be shown from Equation l that the speed S, with which any given beam of ions moves past the collector slit is given by the equation:

' In 3 Swr *a (2) 1' dV S=r ai (3) -Similar equations apply if some other independent control variable is changed.

Due to numerous factors including var1ous physical limitations of the dimensions of the mass spectrometer, the beams have iinite widths at the collector slit. The intensity distribution of a beam as measured, for example, in terms of ionic charge density or electric current density in the beam varies :from point to point in the beam. In general, the intensity of any one beam is proportional to the surface integral of the current density across any cross-section thereof, and in a mass spectrometer of fixed geometry, is proportional to the maximum charge density or maximum current density in the beam cross-section at the collector slit, said beam cross-section preferably being taken perpendicular to the central path of the beam at the collectorslit.y For brevity, in :this specification, the percentage or relative distribution of ionic charge density or electric less of the mass-to-charge ratio of ions forming the beams being recorded.

A distinction is here made between the actual mass-to-charge ratio M of the ions which comprise a given ion beam, and the virtual masstio-.charge ratio m of ions which could if present, travel over a given path from the ionization lchamber to the collector. The virtual mass-tocharge ratio is given by Equation l, or some similar applicable equation. In this specification factual will be omitted when the actual mass-tocharg'e ratio is meant, but virtual will be used wherever the actual mass-to-charge ratio is not intended.

From Equation 1 it can be shown that the spacing between the centers of adjacent beams differing in mass-to-charge ratio by av unit amount and focussed at the exit slit decreases as the virtual mass-to-charge ratio increases.v Thus beams comprising ions of low mass-to-charge ratio are readily resolved. However, at large mass-tocharge ratios where the spacing between centers of adjacent beams approaches the width of the individual beams, adjacent beams and become difcult to resolve.

l In general two methods have been used heretofore for measuring the intensities of ion beams successively detected in a mass spectrometer.

l In the first, ion-beams comprising ions of predetermined m-ass-to-charge ratios were successively focussed at the collector slit by manual control of either the magnetic or electric segregating force or both. When using this method, the collected ion current is measured by means of a suitable circuit and meter connected to the ion current collector, after each ion beamis brought to rest at the collector slit. To study the shape of any beam with this manual control system, a

segregating force is set at 'a succession of values differing by small amounts and the corresponding collected ion currents measured. Such a method is slow, tedious, and cumbersome.

1A second method which has been used heretofore has involved automatically varying the acbegin to merge the accelerating voltage is so varied as a linear function of time. Asa result, in this system the recordingl speed is unnecessarily slow for ions of low mass-to-charge ratio, and the total time required for recording a mass spectrum is unnecessarily long.. As will become apparent hereinafter, with this system ion beams having the same shape are not recorded as peaks of the same shape on the mass spectrogram in this automatic recording system.

Accordingto my invention, ion beams of the same shape but comprising ions of differentmass-to-charge ratios may be recorded as peaks of the same shape by varying as an exponential function of time a control variable (suchA as H or V) in terms of which the mass-to-charge ratio is a power function, and recording the detected ion current on a recording'medium moving at constant speed past the recording point. If H or V is the variable, all beams are moved past the collector slit at the samespeed. In any event, according to my invention, the time in which a mass spectrogram may 'be recorded .is

decreased as compared with methods heretofore used, by changing the control variable at such a rate that the ratio of velocity with which different ion beams are moved past a collector slit to the mass-to-charge ratio of the ions comprising said beams is an inverse function of the recelerating voltage as a linear function of time in a `mass spectrometer of fixed geometry, and syn-E chronously recording the ion current produced as each ion beam is moved past the collector slit. In this method ofi Equation 3 is constant, and the respective ion beams move past the collector slit at a speed proportional to the mass-to-charge ratio of the ions comprising the respective beams. Due to the fact that time is required for the ion current collector circuit and meter to respond to the ion current,lthe rate at which the voltage can be changed is determined by the spacing between peaks corresponding to ions of the largest massto-charge ratio which are to be. recorded when April 6, 19.42, by

spective mass-to-charge ratios. For maximum recording speed consistent with high resolving power of beams comprising lons of large mass, I

move the respective detected beams past the collector rslit at speeds which are an inverse function of the mass-to-charge ratio of ions comprising the respective beams.

This application is a continuation-in-part of the patent application Serial No. 437,922, led

Edmund E. Hoskins, and Robert V. Langmuir.

The principal object of my invention is topro` vide an improved system for automatically re- Y cording a. mass spectrogram. Another object is to provide a rapid method of recording a mass spectrum throughout a large range of mass-tocharge ratios- Still another object is to record beams of a mass spectrum in such a manner that differences in recorded peak shapes represent differences in the intensity distribution in a crosssection of the individual ionbeams at the collector slit.

My invention possesses numerous other objects and features of advantage, some of which, to-

' gether with the foregoing,.will be set forth in the following description of specific apparatus embodying and utilizing my novel method. It is therefore to be understood that myrmethod is applicable to other apparatus, and that I do not limit myself, in any way, to the apparatus of the present application, as I may adopt various other apparatus embodiments, utilizing the method, within the scope of the appended claims. Referring to the drawings: f

Fig. 1 is a schematic diagram of a mass spectrometer incorporating a control circuit and a recorder operated in accordance with my invention. f

Figs. 2, 3, 4, 7, and 1-0 are drawings of peaks representing ion beams which may be measured under different conditions adopted in operating a mass spectrometer.

` Fig. 5 represents graphs useful in explaining my invention.

Figs. 6 and 8 are schematlcdiagrams of control circuits.

Fig. 9 represents the response `characteristic of my ion current reproducing system.

In Fig. l I have illustrated the application ot my invention to a Dempster type mass spectrom eter, adapted to the analysis' of gaseous materials. A gas to be analyzed is flowed continuously from a sample region (not shown) through an inlet conduit l, into an ionization chamber 3, and

thence to a vacuum pump (not shown) through the discharge conduit 5. Electrons emitted from a heated lament l in the ionization chamber are directed in a beam along a line perpendicular to 'the face of magnetic pole 9 by the combined action of a magnetic field which is directed downward perpendicular to the plane of the drawing, and an electric field parallel to th'e magnetic field provided by suitable electrodes (not shown).

maenthese electrons encounter molecules of the gas, the molecules become ionized ina manner characteristic of the gas anddependent upon 'the conditions of ionization. The concentrations o? ions of different mass-'to-charge ratio thus' formed may be measured and utilized in studyina" the properties of the gas and in analyzing its composition. Positively charged ions are accelerated toward rst slit electrode Il by action of a small electrical potential which maintains said first vcollimating electrode negative with respect to a pusher electrode I3 on the opposite side of the electron beam. Some of the accelerated ions pass through a narrow slit l5 in said rst colliniating electrode il, and are thereupon accelerated by a 'large negative potential maintained between second collimating electrode lli and said rlrst collimating electrode. Some of the accelerated ions then pass through a second slit l@ in said second ccllimating electrode.

in an analysis the potential between electrodes il' and ll maybe varied from several hundred volts to several thousand volts. IIons which do not pass through the slits of the two electrodes il and il strike the walls of the ionization chamher and there give up their charge. The rate at which ions'of a. particular mass-to-charge ratio pase through the slits i5 and I9 is a .measure of the rate oi formation of such ions in the ionization chamber, and is also a measure of the concentration of such ions present there.

@Wing to the action of the magnetic field, positive ions passing through second slit i9, follow circular paths. To detect ions of predetermined mass-to-charge ratio a collector slit 22 is positioned on the circumference of a semi-circle 2t or radius r, and 180 from the midpoint between slits it and i9. The mass-to-charge ratio of any ions which are formed at a particular point in the ionization chamber with zero initial velocity, and which travel over the specific path 2Q, is given by the formula:

' the same path, nor over paths of the same radius.

Ions of a predetermined mass-to-charge ratio formed at different points in the ionization chamber or even at the same point with diierent initial velocities pass through slits I5, I1 at different angles iorming a divergent ion beam. Beams comprising ions o! different mass-tocharge ratios are focussed at diierent corresponding points 180 from the collimating slits I5 and IT. These beams may be successively focussed at the exit slit 22 by changing the accelerating voltage.

So'long as the relative concentrations with which the-ions are formed at different points within the ionization chamber, and the paths over which the ions travel in the analyzer tube 2| are the same, the percentage distribution of ionic charge density over a cross-section of the detected beam,for example at the position of the collector slit, is the same regardless of the massto-charge ratio of the ions. This will occur in a mass spectrometer of xed geometry when different ion beams are focussed at the collector slit, and some ions o the actual mass-to-charge ratio comprising each beam travel from a particular point in the ionization chamber over` a specic path, such as the centerpath represented by the semi-circle 2li in Fig. 1. (See Bleakneys theorem, American- Physics Teacher, t, 23, 1936.) The apparatus geometry may be maintained xed by keeping the dimensions of the apparatus constant; the magnetic field geometry may be maintained fixed by keeping the magnetic iield constant; and the electric field geometry Iii) ' dition may be readily made.

may be maintained fixed as beams of different mass-to-charge ratio are moved past the collector slit by proportionally varying the electric iields between the pair of electrodes Hand i3 and the utilized.

Ions which travel over path 26 or any adja.

cent paths and which pass through collector slit 22, fall on ion collector 23, there give up their chargmand produce an ion current which passes through shielded lead 25 into upper input terminal 2l of Dfi). amplifier 29, and thence to grounded input terminal ab. The D.C. current is applied to a galvanometer 41 after amplification by said D.C. amplifier. A

As the displacement of galvanometer mirror 50 l rom a neutral position is changed in response to changes in collected currents, the mirror displacementv is optically ampliiied by means of light ray 5l from light source tt. The ray is projected onto a moving medium in the form of a strip of photographic paper SS driven at constant speed by a motor 553. As the reflected ray'is displaced in a direction perpendicular to the direction of paper movement, a latent image is automatically formed on the paper representing a plot of the collected current as a function of time. If the speed of the recording medium past the recording point is not constant, correction for this con- Simultaneously with the recording of the galvanometer current,

light from source 59 is projected onto the photof graphic paper through slots 8l in mask 62. said slots lying in lines parallel to the direction of paper movement to provide indicia in the form i of trace displacement scale lines I4. A visible mass spectrogram is obtained when the paper isV developed.

In order time.

vpotentials to which the ions are subjected as an exponential function of time. Any deviation of the recorded peak shape from the normal will indicate an anomalous condition suchas may be associated for instance with peculiar .conditions sometimes arising in the ionization process where the ions `sometimes start with abnormally high' initial velocities. Since variations in the peak shape from the normal maybe utilized in a. study of the ionization process and in the identication of the gas producing the ions, it is 'important to reproduce the peak shapes in a uniform manner.

A control circuit utilized to achieve constant operating a mass spectrometer in accordance with my invention, the effect of variations in the speed with which any one ion beam is moved past vthe exit slit 22 in a radial direction indicated by arrowv 24, will now be explained.

In order that the recorded peaks .will not be sharp at the top and thus render the peak magnitude diilicult to measure, I prefer to make the exit slit at least as wide as the beam width or the space occupied at the exit slit by most of the ions of a beam. If the beam is moved slowly past the exit slit, the detected ion current will be represented somewhat according to a curve such as that shown in Fig. 2A. If the .beaml is moved past the exit slit faster, the current will be detected for a shorter time, and will be represented by a curve such as that shown in Fig. 2B. In both Figs. 2A and 2B, the amplifier and recorder v ,have been operated suiliciently slowly for each beam speed and hence uniform treatment of different beams may be in the form of a resistance R, condenser C. discharge circuit 3l. In this circuit resistance 32 provides Aa potentiometer having its positive end 33 connected to pusher electrode I3, its negative end 35 connected to second slit electrode I1 .through the metallic analyzer tube 2| and to ground and an intermediate point 3l near the positive end thereof connected to nrst collimator electrode Il. denser 31 is connected across said resistance. Battery 30 having its negative pole grounded may be connected to potentiometer end 33 through switch ll. After condenser 31 has been charged to a standard potential by closing `key- 4I,` thus Conestablishing predetermined positive potentials on l electrodes II and I3, said key may then be opened to discharge the condenser through poences between the electrodes Il, I3 and I'lv to decrease as an exponential function of time depending upon the time constant RC of the circuit tentiometer 32, thus causing the potential-differrather than' as the customary linear function of i Thus, at a time t after opening ke`y Il the accelerating voltage for ions formed at any particular point in the ionization chamberis given by the equation y I where Ic is the reciprocal of the time constants RC ratios are moved past the collector slit 22, variations -in the intensity of the collected current occur. In general, the magnitudes of these collected ion current variations correspond to the rates of formation of different corresponding ions. Expressed in other words, as beams of ions of diil'erent mass-to-charge ratios successivelyimpinge said ion collector, a series of ion currents is produced in which the magnitude of each ion current corresponds to the rate of formation of ions of the corresponding mass-to-charge ratios in the ionization chamber and the variation of the ion current with time depends on the beam shape. A mass spectrum of the sample is obtained by measuring heights of the recorded peaks, or the intensities of the ion currents.

In order to fully appreciatethe'advantages of variation in detected current to produce a corresponding full indication on the galvanometer, and both of the recorded peaks are flat topped, -thus simplifying accurate measurement of the ion current. However, if the beam is moved past the exit slit still faster, the amplifier and galvanometer may not react fast enough to produce a full indication of each variation of detected current and the recorded peak will be narrow and pointed somewhat as shown in Fig. 2C, even though the collector slit is wider than the beam. At a still greater speed, the maximum recorded displacement corresponding to the maximum value of the'ion current will be abnormally low, as indicated in Fig. 2D. The maximum attainable value may be approached to any .given degree of accuracy by moving the beam past the exit slit suiliciently slowly. To record at a speed corresponding to Fig. 2D would produce a recording which could not be easily used for accurate determination of beam intensities. at a speed corresponding to Fig. 2A would waste time. In practice a suitable intermediate speed is determined empirically by recordingat diilerent speeds and selecting a speed suitable for theVVV purpose at hand. I'he beam speed necessary to achieve any given degree of accuracy depends on' both the beam width and slit width. To measure ion currents to an accuracy of about 0.5 to 1%,

I prefer to utilize a beam speed which is justv If the exit slit isvery narrow compared to the I width that a beam has in the plane of the exit slit so that only a small cross-section of a beam is detected at any instant, and beams of different vmass-to-charge ratios move past said slit at a very slow speed, and the amounts of ions passing through unit area of this narrow slit per second are measured, then a graph such as that shown in Fig. 3 may be constructed representing a plot of the collected current as ordinate against the virtual mass-to-charge ratio m as abscissa. At each instant the collected current is equal to the current density in the beam integrated over the area of the collector slit. As a matter of fact all the ions producing any recorded peak P, are of the same actual mass-to-charge ratio, and the distance between centers of any two peaks on this graph is a multiple oi' one unit of mass-to-charge ratio. If the beams were sharply defined and of uniform intensity, the sides of the peaks are parallel to the current axis and the corners of the peaks at the top ,and bottom are right angles instead of being rounded as shown.

However, to record As shown hereinabove, if ions of different massto-charge ratios lcorresponding to the different peaks are formed in the ionization chamber in the same proportion at diiierent points thereof and passed through slits I and I9 of fixed dimen sions. then, while electric potentials acting on the beam are varied in constant proportions, beams of ions having initially zero or negligible velocity at the points of ionization (which is normally the case) will be of the same size and shape when focussed at the collector slit. The normal Width of such a beam at the exit slit is independent of the iield strengths and may be determined from the widths of slits l5 and I9, the distance between electrodes il and l1, and the field strengths.

Whereas the widths of the beams at the collector slit will normally be the same, the width of peaks in the graph of Fig. 3will be very nearly proportional to the mass-to-charge ratios of the ions forming the corresponding beam. For heavy ions, the plotted peaks begin to merge and for still heavier ions they actually overlap because the interval between centers of 'successive beams becomes less than the beam width.

In practice with a Bempster type mass spectrcmeter individual recorded peaks are slightly asymmetrical as shown in Fig. due to asymmetry of the ion beam at the exit slit.

If we introduce the Variable X, such that where t represents time, and Ic is a constant, then the height or any peak at any point thereof may. be represented as a function lDOI).

current which is detected by the ion collector, and the intensity of thedetected current varies as a function of time t, this current may be represented as the sum of sinusoidal current components of angular frequency c. If the origin of X is at the center of any one peak corresponding to the actual mass-to-charge ratio m, as indicated in Fig. e, the formula of that peak may be represented as:

where you) represents the Fourier transform of IMX), or the frequency spectrum of the ion current. This transform is denned by the equation:

(See Guillemin, Communication Networks, chapter XI, John Wiley, (1935).)

All normal peaks will yhave the same width and shape but will differ in height when plotted to the coordinates of Fig. 4. For beams of normal shape the function gw) will always be the same except for a multiplying factor proportional to the peak height. 'll'hc Fourier transform g(w) is characteristic oi any peak whether it be of normal or abnormal shape. Equation 'l will represent the actual shape of the recorded peak provided the apparatus responds uniformly to all frequency components w making any substantial contribution to the ion current. This will be true, for example, if the collector slit is very narrow compared to the beam width and the amplifier and galvanometer respond instantaneously to the detccted ion current. However,- the apparatus may not respond uniformly to all such frequency components, since the vcollector slit itself may be wide enough to modify the characteristic of the Since the peak displacement actually represents an electricdit ion current and the combination of amplifier 29 and galvanometer 41 is in fact a current transducer which has a finite time constant and may modify the proportions of the contributions of the different frequency components of any detected ion current to the actual recorded peak. If the width of the collector slit through which the beam passes is constant and large enough to modify the peak shape, the formula for the current to the ion collector may be represented as where, Bm) is a characteristic depending on the exit slit width and the speed of the beam past the exit slit. The product Bm) gw) is the Fourier transform of the ion current to the collector, or the rate of discharging the beam at the collector,

v and varies from one beam to another only if they are of different shapes. This current will be recorded as a peak represented by the formula where C( c) is a transducer characteristic depending on the electrical characteristics of the D.C. ampliiier and the electrical and mechanical characteristics of the galvanometer. Fig. 9 shows one such characteristic Cw) in which all frequency components l (fue of any ion current up to 4 cycles per sec. are reproduced substantially uniformly by the amplifier and galvanometer; above i cycles per sec. all components are attenuated in various amounts in.` creasing with frequency. At about 10 cycles per second Cw) is 0.70 oi' its maximum value. Expressed in other words, the time constant T of the transducer is 0.1 sec. It follows that if the virtual mass-t'o-charge ratio is varied in accordance with Equation 6, that; is according to the equation m==tceeM (ll) where m0 is the virtual mass-to-charge ratio of ions detectable at time t=0, all beams having the same shape will be recorded as peaks of the same shape regardless of the slit width, and the shapes of recorded peaks will differ only if the Fourier transform got) corresponding to said peaks differ. I have found that a mass spectrometer of xed geometry 'will uniformly reproduce beams of the same shape, if the accelerating voltage V is either increased or decreased as an exponential function of time. With the control circuit of Fig. l the voltage is decreased in accordance with the exponential function of time expressed by Equation 5. When the voltage varies exponentially, the instantaneous velocity with which ions vtraveling over the semi-circular path 20 move radially past the center of the collector slit is the same regardless of the mass-to-charge ratio of the ions and is given by the equation S=lcr (l2) In this case the operation may be described simply by saying that the speed of each beam is the same as it passes the collector slit. Recorded peaks bearing a uniform relation to the ion beam shapes will be produced as long as this condition is maintained during the recording of the respective peaks so that, as a matter of fact, the voltage may, it desired, be changed between peaks more exit slit. Curves b and b inA these figures repref sent cases in which the beam speed varies as an inverse function of the mass-to-charge ratio. In both cases represented in these gures the ratio. S/M of beam speed to mass-to-charge ratio is an inverse function of mass-to-charge ratio, that is, in the range of operation this ratio S/M decreases as the mass-to-charge ratio increases.

In a preferred form of control circuit which is shown in Fig. 6, the high voltage end of a regulated power supply is connected through two resistances '|3and 'l5 tothe accelerating voltage tual mass-to-charge ratio is an inverse function of th'e virtual mass-to-charge ratio. The beam speeds suitable for any purpose with mass spectrometers of different dimensions or with a mass sepectrometer of changing dimensions are proportional to a linear dimension of the mass spectrometer. i

Thus when uniform reproduction of beam shapes is not required, there is some advantage to be obtained by varying the beam speed as an inverse function of the mass-to-charge ratio of the ions. If the peaks of lower mass-to-charge ratio are being recorded atv a high speed satisfactorily consistent with accuracy, then at high masses where the peaks are crowded close together. that is, where the distance between sides Y of peaks is less than the peak widths, one peak may start to impinge the collector before the amplier and galvanometer have had suicient time to returnl to the neutral condition. This would produce a mass spectrum such as that shown in control potentiometer 32. Said potentiometer is connected to electrodes I3 and I1 in the same manner as potentiometer 33 shown in Fig. 1. In the present case a vacuum tube Tl is connected with its anode 19 and cathode 8| in shunting relation across one of said resistors and potentiometer 32. A control circuit in the form of a resistance condenser discharge circuit 3| is connected between grid 83 and filament type cathode 8| of tube Tl. A condenser 85 connected across potentiometer 32 cooperates with resistance 15 and said potentiometer to lter out any fluctuations in voltage that might appear across said potentiometer 32 due to the presence of any ripple in thelamentll current.. In this circuit the resistance vof the RC circuitis in a low voltage section of this apparatus and may therefore be several megohms, thus facilitating provision of a time constant of several minutes or more. With this circuit the beam speed may be varied if desired during the recording lby taking advantage of the non-linear characteristics of the vacuum'tube.` In Fig. 8 I have illustrated schematically a resistor condenser discharge control circuit 3|" by means of which the voltage is changed and the speed kr with which a beam is swept past the co1- lector slit maybe varied during the recording in any desired manner. controlled by varying onel of the elements of the resistance condenser discharge circuit 3|, such as the resistance 32 as a suitable predetermined function of time-or virtual mass-to-charge ratio by means of a suitably shaped motor driven cam 8|, thev operation of which is initiated in any convenient manner in timed relation to the moment that key 4| is opened to initiate the discharge of the condenser 31.

When peak shapes due to beams comprising ions of different mass-to-charge ratios are to be compared, the beam speed should be the same when recording the two corresponding peaks. However, if fidelity of reproduction of the respective beams is not required, but it is only desirable to measure the intensity of each of the beams, the beam speed can be made more rapid Where the spacing between peaks is less than or equal to the peak widths, and reduced for peaks which are closer together, responding to such speed changes are shown at b and b in Figs. 5a and 5b. For both curves the 3 .time rate of change of the logarithm of the virtua mass-to-charge ratio divided bythe vir- The beam speed may be Fig. 7, wherein the valleys v between' adjacent peaks p do not reach the zero displacement axis cd, even though the detected ion current may have reached a zero value momentarily. Accordingto my invention the galvanometer may be returned to its zero position atsuch high masses byl reducing the beam speed at the exit slit when detecting ions of hi-gh masses. When changing the ion beam speed in this manner, beams of the same shape are not recorded as peaks of the same shape. However, this does simplify the measurement of the maxiumum amount of ion current due to beams comprising ions of different high mass-to-charge ratios.

, If the beam is sharply defined and of uniform intensity distributionin a beam crosssection at the collector slit, it can be shownthat 'the fractional amount bywhich the recorded peak fails to achieve its maximum attainable value is given approximately by the following equation:

paratra) Where b a and the beam and collector slit widths are either a and b respectively or b and a respectively, as the case may be: From Equation 13 it is apparent that the beam speed S required to produce a given degree of accuracy as expressed in terms of A depends on both the beam width and the collector slit width, as well as the response time T of the ion current transducer. If a b, the velocity of the beam should be and the beam and collector slit widths are about the same,'a beam speedof about A10% of v the value computed from Formula 14 gives a high -Plots of curves cori degree of accuracy in analysis of mixtures.

In practice, for maximum recording speed each beam is `moved past the collector slit at a speed below that for which any harmful effects begin to appear due to' failure of the recorder either to produce maximum indication of the ion current or to accurately reproduce the beam shape according to the purpose athand. In all cases thel ratio of beam width to collector slit width should be maintained less than about unity to assurev normally achieving nat top peaks.

It is thusy seen that I have provided a ilexlble system in whichY the beam speed may be controlled in any suitable manner'such as may be required for example to achieve uniform reproduction of ion beams or maximum recording speed.

I' claim: 1. In a mass spectrometer of iixed geometry, an ion detector, a control device adapted to successively move beams comprising ions of a differentl mass-to-eharge ratio transversely to and in contact with said detectorvat speeds which vary as an inverse function of mass-to-charge ratio of the respective ions. and an ion current recorder connected to said detector. y 2. In mass spectrometry involving the production of a plurality of ion beams of substantially the same shape but different mass-to-charge ratios by propelling the ions with an accelerating voltage through a magnetic field and longitudinally along the beams successively onto an ion detector while shifting the beams with respect to the detector and transversely to the direction of ion travel, the improvement which comprises shifting the beams with respect to the detector" by varying as an exponential function of time a factor selected from the group consisting of the magnetic field vand the accelerating voltage. 3. Process according to claim 2 infwhich each ion beam is shifted transversely with respect to the detector at the same speed.

4. Process according. to claim 2 in whicnthe ing the respective beams.

factor is varied at a rate such that the ratio of velocity with which different ion beams are shifted transversely to the detector tothe mass-to-charse ratio of the ions comprising the respective beams is an inverse function of the respective mass-tocharge ratios. 5. Process according to claim 2 in Vwhich th respective ion beams are shifted transversely to the detector at speeds which are an inverse function of the mass-to-charge ratio of ions compris- 6. In a mass spectrometerl the combination which comprises means for producing a plurality of ion beams of different mass-to-charge ratios including means for producing a magnetic eld andelectrodes for impressing an accelerating voltage upon the ions to propel them in the field, an ion detector, means for bringing the beams successively into contact` with the detector by shifting the beams transversely to the direction of ion travel and means for controlling the rate of the shifting by varying. as an exponential function of time, -a factor selected from the group consisting of the accelerating voltage and the intensity of the magnetic ileld.

7. Apparatus according to claim 6 in which the means for controlling the rate of shift includes a control circuit with a condenser, a resistance, and means for discharging the condenser through the resistance. ROBERT V. LANGMUIR.

Images